International Journal of Emerging Technologies in Computational and Applied Sciences issue 11 vol1 1

ISSN (ONLINE): 2279-0055 ISSN (PRINT): 2279-0047

Issue 11, Volume 1, 2 & 3 December-2014 to February-2015

International Journal of Emerging Technologies in Computational and Applied Sciences

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

STEM International Scientific Online Media and Publishing House Head Office: 148, Summit Drive, Byron, Georgia-31008, United States.

Offices Overseas: Germany, Australia, India, Netherlands, Canada.

Website: www.iasir.net, E-mail (s): iasir.journals@iasir.net, iasir.journals@gmail.com, ijetcas@gmail.com

PREFACE We are delighted to welcome you to the eleventh issue of the International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS). In recent years, advances in science, technology, engineering, and mathematics have radically expanded the data available to researchers and professionals in a wide variety of domains. This unique combination of theory with data has the potential to have broad impact on educational research and practice. IJETCAS is publishing high-quality, peer-reviewed papers covering topics such as computer science, artificial intelligence, pattern recognition, knowledge engineering, process control theory and applications, distributed systems, computer networks and software engineering, electrical engineering, electric machines modeling and design, control of electric drive systems, non-conventional energy conversion, sensors, electronics, communications, data transmission, energy converters, transducers modeling and design, electro-physics, nanotechnology, and quantum mechanics.

The editorial board of IJETCAS is composed of members of the Teachers & Researchers community who have expertise in a variety of disciplines, including computer science, cognitive science, learning sciences, artificial intelligence, electronics, soft computing, genetic

algorithms,

technology

management,

manufacturing

technology,

electrical

technology, applied mathematics, automatic control , nuclear engineering, computational physics, computational chemistry and other related disciplines of computational and applied sciences. In order to best serve our community, this Journal is available online as well as in hard-copy form. Because of the rapid advances in underlying technologies and the interdisciplinary nature of the field, we believe it is important to provide quality research articles promptly and to the widest possible audience.

We are happy that this Journal has continued to grow and develop. We have made every effort to evaluate and process submissions for reviews, and address queries from authors and the general public promptly. The Journal has strived to reflect the most recent and finest researchers in the field of emerging technologies especially related to computational and applied sciences. This Journal is completely refereed and indexed with major databases like: IndexCopernicus, Computer Science Directory, GetCITED, DOAJ, SSRN, TGDScholar, WorldWideScience, CiteSeerX, CRCnetBASE, Google Scholar, Microsoft Academic Search, INSPEC, ProQuest, ArnetMiner, Base, ChemXSeer, citebase, OpenJ-Gate, eLibrary, SafetyLit, SSRN, VADLO, OpenGrey, EBSCO, ProQuest, UlrichWeb, ISSUU, SPIE Digital Library,

arXiv,

ERIC,

EasyBib,

Infotopia,

WorldCat,

.docstoc

JURN,

Mendeley,

ResearchGate, cogprints, OCLC, iSEEK, Scribd, LOCKSS, CASSI, E-PrintNetwork, intute, and some other databases.

We are grateful to all of the individuals and agencies whose work and support made the Journal's success possible. We want to thank the executive board and core committee members of the IJETCAS for entrusting us with the important job. We are thankful to the members of the IJETCAS editorial board who have contributed energy and time to the Journal with their steadfast support, constructive advice, as well as reviews of submissions.

We are deeply indebted to the numerous anonymous reviewers who have contributed expertly evaluations of the submissions to help maintain the quality of the Journal. For this eleventh issue, we received 157 research papers and out of which only 48 research papers are published in three volumes as per the reviewers’ recommendations. We have highest respect to all the authors who have submitted articles to the Journal for their intellectual energy and creativity, and for their dedication to the field of computational and applied sciences.

This issue of the IJETCAS has attracted a large number of authors and researchers across worldwide and would provide an effective platform to all the intellectuals of different streams to put forth their suggestions and ideas which might prove beneficial for the accelerated pace of development of emerging technologies in computational and applied sciences and may open new area for research and development. We hope you will enjoy this eleventh issue of the International Journal of Emerging Technologies in Computational and Applied Sciences and are looking forward to hearing your feedback and receiving your contributions.

(Administrative Chief)

(Managing Director)

(Editorial Head)

--------------------------------------------------------------------------------------------------------------------------The International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS), ISSN(Online): 2279-0055, ISSN(Print): 2279-0047 (December-2014 to February-2015, Issue 11, Volume 1, 2 & 3).

---------------------------------------------------------------------------------------------------------------------------

BOARD MEMBERS

        

                 

EDITOR IN CHIEF Prof. (Dr.) Waressara Weerawat, Director of Logistics Innovation Center, Department of Industrial Engineering, Faculty of Engineering, Mahidol University, Thailand.

Prof. (Dr.) Yen-Chun Lin, Professor and Chair, Dept. of Computer Science and Information Engineering, Chang Jung Christian University, Kway Jen, Tainan, Taiwan.

Divya Sethi, GM Conferencing & VSAT Solutions, Enterprise Services, Bharti Airtel, Gurgaon, India.

CHIEF EDITOR (TECHNICAL) Prof. (Dr.) Atul K. Raturi, Head School of Engineering and Physics, Faculty of Science, Technology and Environment, The University of the South Pacific, Laucala campus, Suva, Fiji Islands.

Prof. (Dr.) Hadi Suwastio, College of Applied Science, Department of Information Technology, The Sultanate of Oman and Director of IETI-Research Institute-Bandung, Indonesia.

Dr. Nitin Jindal, Vice President, Max Coreth, North America Gas & Power Trading, New York, United States.

CHIEF EDITOR (GENERAL) Prof. (Dr.) Thanakorn Naenna, Department of Industrial Engineering, Faculty of Engineering, Mahidol University, Thailand.

Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Huiyun Liu, Department of Electronic & Electrical Engineering, University College London, Torrington Place, London.

ADVISORY BOARD Prof. (Dr.) Kimberly A. Freeman, Professor & Director of Undergraduate Programs, Stetson School of Business and Economics, Mercer University, Macon, Georgia, United States.

Prof. (Dr.) Klaus G. Troitzsch, Professor, Institute for IS Research, University of Koblenz-Landau, Germany.

Prof. (Dr.) T. Anthony Choi, Professor, Department of Electrical & Computer Engineering, Mercer University, Macon, Georgia, United States.

Prof. (Dr.) Fabrizio Gerli, Department of Management, Ca' Foscari University of Venice, Italy.

Prof. (Dr.) Jen-Wei Hsieh, Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taiwan.

Prof. (Dr.) Jose C. Martinez, Dept. Physical Chemistry, Faculty of Sciences, University of Granada, Spain.

Prof. (Dr.) Panayiotis Vafeas, Department of Engineering Sciences, University of Patras, Greece.

Prof. (Dr.) Soib Taib, School of Electrical & Electronics Engineering, University Science Malaysia, Malaysia.

Prof. (Dr.) Vit Vozenilek, Department of Geoinformatics, Palacky University, Olomouc, Czech Republic.

Prof. (Dr.) Sim Kwan Hua, School of Engineering, Computing and Science, Swinburne University of Technology, Sarawak, Malaysia.

Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Rafael Ignacio Alvarez Sanchez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Praneel Chand, Ph.D., M.IEEEC/O School of Engineering & Physics Faculty of Science & Technology The University of the South Pacific (USP) Laucala Campus, Private Mail Bag, Suva, Fiji.

Prof. (Dr.) Francisco Miguel Martinez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Antonio Zamora Gomez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Leandro Tortosa, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Samir Ananou, Department of Microbiology, Universidad de Granada, Granada, Spain.

Dr. Miguel Angel Bautista, Department de Matematica Aplicada y Analisis, Facultad de Matematicas, Universidad de Barcelona, Spain.

           

                  

Prof. (Dr.) Prof. Adam Baharum, School of Mathematical Sciences, University of Universiti Sains, Malaysia, Malaysia.

Dr. Cathryn J. Peoples, Faculty of Computing and Engineering, School of Computing and Information Engineering, University of Ulster, Coleraine, Northern Ireland, United Kingdom.

Prof. (Dr.) Pavel Lafata, Department of Telecommunication Engineering, Faculty of Electrical Engineering, Czech Technical University in Prague, Prague, 166 27, Czech Republic.

Prof. (Dr.) P. Bhanu Prasad, Vision Specialist, Matrix vision GmbH, Germany, Consultant, TIFACCORE for Machine Vision, Advisor, Kelenn Technology, France Advisor, Shubham Automation & Services, Ahmedabad, and Professor of C.S.E, Rajalakshmi Engineering College, India.

Prof. (Dr.) Anis Zarrad, Department of Computer Science and Information System, Prince Sultan University, Riyadh, Saudi Arabia.

Prof. (Dr.) Mohammed Ali Hussain, Professor, Dept. of Electronics and Computer Engineering, KL University, Green Fields, Vaddeswaram, Andhra Pradesh, India.

Dr. Cristiano De Magalhaes Barros, Governo do Estado de Minas Gerais, Brazil.

Prof. (Dr.) Md. Rizwan Beg, Professor & Head, Dean, Faculty of Computer Applications, Deptt. of Computer Sc. & Engg. & Information Technology, Integral University Kursi Road, Dasauli, Lucknow, India.

Prof. (Dr.) Vishnu Narayan Mishra, Assistant Professor of Mathematics, Sardar Vallabhbhai National Institute of Technology, Ichchhanath Mahadev Road, Surat, Surat-395007, Gujarat, India.

Dr. Jia Hu, Member Research Staff, Philips Research North America, New York Area, NY. Prof. Shashikant Shantilal Patil SVKM, MPSTME Shirpur Campus, NMIMS University Vile Parle Mumbai, India.

Prof. (Dr.) Bindhya Chal Yadav, Assistant Professor in Botany, Govt. Post Graduate College, Fatehabad, Agra, Uttar Pradesh, India.

REVIEW BOARD Prof. (Dr.) Kimberly A. Freeman, Professor & Director of Undergraduate Programs, Stetson School of Business and Economics, Mercer University, Macon, Georgia, United States.

Prof. (Dr.) Klaus G. Troitzsch, Professor, Institute for IS Research, University of Koblenz-Landau, Germany.

Prof. (Dr.) T. Anthony Choi, Professor, Department of Electrical & Computer Engineering, Mercer University, Macon, Georgia, United States.

Prof. (Dr.) Yen-Chun Lin, Professor and Chair, Dept. of Computer Science and Information Engineering, Chang Jung Christian University, Kway Jen, Tainan, Taiwan.

Prof. (Dr.) Jen-Wei Hsieh, Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taiwan.

Prof. (Dr.) Jose C. Martinez, Dept. Physical Chemistry, Faculty of Sciences, University of Granada, Spain.

Prof. (Dr.) Joel Saltz, Emory University, Atlanta, Georgia, United States.

Prof. (Dr.) Panayiotis Vafeas, Department of Engineering Sciences, University of Patras, Greece.

Prof. (Dr.) Soib Taib, School of Electrical & Electronics Engineering, University Science Malaysia, Malaysia.

Prof. (Dr.) Sim Kwan Hua, School of Engineering, Computing and Science, Swinburne University of Technology, Sarawak, Malaysia.

Prof. (Dr.) Jose Francisco Vicent Frances, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Rafael Ignacio Alvarez Sanchez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Francisco Miguel Martinez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Antonio Zamora Gomez, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Leandro Tortosa, Department of Science of the Computation and Artificial Intelligence, Universidad de Alicante, Alicante, Spain.

Prof. (Dr.) Samir Ananou, Department of Microbiology, Universidad de Granada, Granada, Spain.

Dr. Miguel Angel Bautista, Department de Matematica Aplicada y Analisis, Facultad de Matematicas, Universidad de Barcelona, Spain.

Prof. (Dr.) Prof. Adam Baharum, School of Mathematical Sciences, University of Universiti Sains, Malaysia, Malaysia.

Prof. (Dr.) Huiyun Liu, Department of Electronic & Electrical Engineering, University College London, Torrington Place, London.

                                

Dr. Cristiano De Magalhaes Barros, Governo do Estado de Minas Gerais, Brazil.

Prof. (Dr.) Pravin G. Ingole, Senior Researcher, Greenhouse Gas Research Center, Korea Institute of Energy Research (KIER), 152 Gajeong-ro, Yuseong-gu, Daejeon 305-343, KOREA Prof. (Dr.) Dilum Bandara, Dept. Computer Science & Engineering, University of Moratuwa, Sri Lanka.

Prof. (Dr.) Faudziah Ahmad, School of Computing, UUM College of Arts and Sciences, University Utara Malaysia, 06010 UUM Sintok, Kedah Darulaman Prof. (Dr.) G. Manoj Someswar, Principal, Dept. of CSE at Anwar-ul-uloom College of Engineering & Technology, Yennepally, Vikarabad, RR District., A.P., India.

Prof. (Dr.) Abdelghni Lakehal, Applied Mathematics, Rue 10 no 6 cite des fonctionnaires dokkarat 30010 Fes Marocco.

Dr. Kamal Kulshreshtha, Associate Professor & Head, Deptt. of Computer Sc. & Applications, Modi Institute of Management & Technology, Kota-324 009, Rajasthan, India.

Prof. (Dr.) Anukrati Sharma, Associate Professor, Faculty of Commerce and Management, University of Kota, Kota, Rajasthan, India.

Prof. (Dr.) S. Natarajan, Department of Electronics and Communication Engineering, SSM College of Engineering, NH 47, Salem Main Road, Komarapalayam, Namakkal District, Tamilnadu 638183, India.

Prof. (Dr.) J. Sadhik Basha, Department of Mechanical Engineering, King Khalid University, Abha, Kingdom of Saudi Arabia Prof. (Dr.) G. SAVITHRI, Department of Sericulture, S.P. Mahila Visvavidyalayam, Tirupati517502, Andhra Pradesh, India.

Prof. (Dr.) Shweta jain, Tolani College of Commerce, Andheri, Mumbai. 400001, India Prof. (Dr.) Abdullah M. Abdul-Jabbar, Department of Mathematics, College of Science, University of Salahaddin-Erbil, Kurdistan Region, Iraq.

Prof. (Dr.) P.Sujathamma, Department of Sericulture, S.P.Mahila Visvavidyalayam, Tirupati517502, India.

Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India.

Prof. (Dr.) Manjulatha, Dept of Biochemistry,School of Life Sciences,University of Hyderabad,Gachibowli, Hyderabad, India.

Prof. (Dr.) Upasani Dhananjay Eknath Advisor & Chief Coordinator, ALUMNI Association, Sinhgad Institute of Technology & Science, Narhe, Pune- 411 041, India.

Prof. (Dr.) Sudhindra Bhat, Professor & Finance Area Chair, School of Business, Alliance University Bangalore-562106.

Prof. Prasenjit Chatterjee , Dept. of Mechanical Engineering, MCKV Institute of Engineering West Bengal, India.

Prof. Rajesh Murukesan, Deptt. of Automobile Engineering, Rajalakshmi Engineering college, Chennai, India.

Prof. (Dr.) Parmil Kumar, Department of Statistics, University of Jammu, Jammu, India Prof. (Dr.) M.N. Shesha Prakash, Vice Principal, Professor & Head of Civil Engineering, Vidya Vikas Institute of Engineering and Technology, Alanahally, Mysore-570 028 Prof. (Dr.) Piyush Singhal, Mechanical Engineering Deptt., GLA University, India.

Prof. M. Mahbubur Rahman, School of Engineering & Information Technology, Murdoch University, Perth Western Australia 6150, Australia.

Prof. Nawaraj Chaulagain, Department of Religion, Illinois Wesleyan University, Bloomington, IL. Prof. Hassan Jafari, Faculty of Maritime Economics & Management, Khoramshahr University of Marine Science and Technology, khoramshahr, Khuzestan province, Iran Prof. (Dr.) Kantipudi MVV Prasad , Dept of EC, School of Engg, R.K.University,Kast urbhadham, Tramba, Rajkot-360020, India.

Prof. (Mrs.) P.Sujathamma, Department of Sericulture, S.P.Mahila Visvavidyalayam, ( Women's University), Tirupati-517502, India.

Prof. (Dr.) M A Rizvi, Dept. of Computer Engineering and Applications, National Institute of Technical Teachers' Training and Research, Bhopal M.P. India Prof. (Dr.) Mohsen Shafiei Nikabadi, Faculty of Economics and Management, Industrial Management Department, Semnan University, Semnan, Iran.

Prof. P.R.SivaSankar, Head, Dept. of Commerce, Vikrama Simhapuri University Post Graduate Centre, KAVALI - 524201, A.P. India.

Prof. (Dr.) Bhawna Dubey, Institute of Environmental Science( AIES), Amity University, Noida, India.

Prof. Manoj Chouhan, Deptt. of Information Technology, SVITS Indore, India.

                                

Prof. Yupal S Shukla, V M Patel College of Management Studies, Ganpat University, KhervaMehsana, India.

Prof. (Dr.) Amit Kohli, Head of the Department, Department of Mechanical Engineering, D.A.V.Institute of Engg. and Technology, Kabir Nagar, Jalandhar, Punjab(India) Prof. (Dr.) Kumar Irayya Maddani, and Head of the Department of Physics in SDM College of Engineering and Technology, Dhavalagiri, Dharwad, State: Karnataka (INDIA).

Prof. (Dr.) Shafi Phaniband, SDM College of Engineering and Technology, Dharwad, INDIA. Prof. M H Annaiah, Head, Department of Automobile Engineering, Acharya Institute of Technology, Soladevana Halli, Bangalore -560107, India.

Prof. (Dr.) Shriram K V, Faculty Computer Science and Engineering, Amrita Vishwa Vidhyapeetham University, Coimbatore, India.

Prof. (Dr.) Sohail Ayub, Department of Civil Engineering, Z.H College of Engineering & Technology, Aligarh Muslim University, Aligarh. 202002 UP-India Prof. (Dr.) Santosh Kumar Behera, Department of Education, Sidho-Kanho-Birsha University, Purulia, West Bengal, India.

Prof. (Dr.) Urmila Shrawankar, Department of Computer Science & Engineering, G H Raisoni College of Engineering, Nagpur (MS), India.

Prof. Anbu Kumar. S, Deptt. of Civil Engg., Delhi Technological University (Formerly Delhi College of Engineering) Delhi, India.

Prof. (Dr.) Meenakshi Sood, Vegetable Science, College of Horticulture, Mysore, University of Horticultural Sciences, Bagalkot, Karnataka (India) Prof. (Dr.) Prof. R. R. Patil, Director School Of Earth Science, Solapur University, Solapur, India.

Prof. (Dr.) Manoj Khandelwal, Dept. of Mining Engg, College of Technology & Engineering, Maharana Pratap University of Agriculture & Technology, Udaipur-313 001 (Rajasthan), India Prof. (Dr.) Kishor Chandra Satpathy, Librarian, National Institute of Technology, Silchar-788010, Assam, India.

Prof. (Dr.) Juhana Jaafar, Gas Engineering Department, Faculty of Petroleum and Renewable Energy Engineering (FPREE), Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor.

Prof. (Dr.) Rita Khare, Assistant Professor in chemistry, Govt. Women,s College, Gardanibagh, Patna, Bihar, India.

Prof. (Dr.) Raviraj Kusanur, Dept of Chemistry, R V College of Engineering, Bangalore-59, India.

Prof. (Dr.) Hameem Shanavas .I, M.V.J College of Engineering, Bangalore, India.

Prof. (Dr.) Sandhya Mehrotra, Department of Biological Sciences, Birla Institute of Technology and Sciences, Pilani, Rajasthan, India.

Prof. (Dr.) Dr. Ravindra Jilte, Head of the Department, Department of Mechanical Engineering,VCET, Thane-401202, India.

Prof. (Dr.) Sanjay Kumar, JKL University, Ajmer Road, Jaipur Prof. (Dr.) Pushp Lata Faculty of English and Communication, Department of Humanities and Languages, Nucleus Member, Publications and Media Relations Unit Editor, BITScan, BITS, PilaniIndia Prof. Arun Agarwal, Faculty of ECE Dept., ITER College, Siksha 'O' Anusandhan University Bhubaneswar, Odisha, India Prof. (Dr.) Pratima Tripathi, Department of Biosciences, SSSIHL, Anantapur Campus Anantapur515001 (A.P.) India.

Prof. (Dr.) Sudip Das, Department of Biotechnology, Haldia Institute of Technology, I.C.A.R.E. Complex, H.I.T. Campus, P.O. Hit, Haldia; Dist: Puba Medinipur, West Bengal, India.

Prof. (Dr.) ABHIJIT MITRA , Associate Professor and former Head, Department of Marine Science, University of Calcutta , India.

Prof. (Dr.) N.Ramu , Associate Professor , Department of Commerce, Annamalai University, AnnamalaiNadar-608 002, Chidambaram, Tamil Nadu , India.

Prof. (Dr.) Saber Mohamed Abd-Allah, Assistant Professor of Theriogenology , Faculty of Veterinary Medicine , Beni-Suef University , Egypt.

Prof. (Dr.) Ramel D. Tomaquin, Dean, College of Arts and Sciences Surigao Del Sur State University (SDSSU), Tandag City Surigao Del Sur, Philippines.

Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India.

Prof. (Dr.) R.K.Tiwari, Professor, S.O.S. in Physics, Jiwaji University, Gwalior, M.P.-474011, India.

Prof. (Dr.) Sandeep Gupta, Department of Computer Science & Engineering, Noida Institute of Engineering and Technology, Gr.Noida, India.

Prof. (Dr.) Mohammad Akram, Jazan University, Kingdom of Saudi Arabia.

                               

Prof. (Dr.) Sanjay Sharma, Dept. of Mathematics, BIT, Durg(C.G.), India.

Prof. (Dr.) Manas R. Panigrahi, Department of Physics, School of Applied Sciences, KIIT University, Bhubaneswar, India.

Prof. (Dr.) P.Kiran Sree, Dept of CSE, Jawaharlal Nehru Technological University, India Prof. (Dr.) Suvroma Gupta, Department of Biotechnology in Haldia Institute of Technology, Haldia, West Bengal, India.

Prof. (Dr.) SREEKANTH. K. J., Department of Mechanical Engineering at Mar Baselios College of Engineering & Technology, University of Kerala, Trivandrum, Kerala, India Prof. Bhubneshwar Sharma, Department of Electronics and Communication Engineering, Eternal University (H.P), India.

Prof. Love Kumar, Electronics and Communication Engineering, DAV Institute of Engineering and Technology, Jalandhar (Punjab), India.

Prof. S.KANNAN, Department of History, Annamalai University, Annamalainagar- 608002, Tamil Nadu, India.

Prof. (Dr.) Hasrinah Hasbullah, Faculty of Petroleum & Renewable Energy Engineering, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia.

Prof. Rajesh Duvvuru, Dept. of Computer Sc. & Engg., N.I.T. Jamshedpur, Jharkhand, India.

Prof. (Dr.) Bhargavi H. Goswami, Department of MCA, Sunshine Group of Institutes, Nr. Rangoli Park, Kalawad Road, Rajkot, Gujarat, India.

Prof. (Dr.) Essam H. Houssein, Computer Science Department, Faculty of Computers & Informatics, Benha University, Benha 13518, Qalyubia Governorate, Egypt.

Arash Shaghaghi, University College London, University of London, Great Britain.

Prof. Rajesh Duvvuru, Dept. of Computer Sc. & Engg., N.I.T. Jamshedpur, Jharkhand, India.

Prof. (Dr.) Anand Kumar, Head, Department of MCA, M.S. Engineering College, Navarathna Agrahara, Sadahalli Post, Bangalore, PIN 562110, Karnataka, INDIA. Prof. (Dr.) Venkata Raghavendra Miriampally, Electrical and Computer Engineering Dept, Adama Science & Technology University, Adama, Ethiopia.

Prof. (Dr.) Jatinderkumar R. Saini, Director (I.T.), GTU's Ankleshwar-Bharuch Innovation Sankul &Director I/C & Associate Professor, Narmada College of Computer Application, Zadeshwar, Bharuch, Gujarat, India.

Prof. Jaswinder Singh, Mechanical Engineering Department, University Institute Of Engineering & Technology, Panjab University SSG Regional Centre, Hoshiarpur, Punjab, India- 146001.

Prof. (Dr.) S.Kadhiravan, Head i/c, Department of Psychology, Periyar University, Salem- 636 011,Tamil Nadu, India.

Prof. (Dr.) Mohammad Israr, Principal, Balaji Engineering College,Junagadh, Gujarat-362014, India.

Prof. (Dr.) VENKATESWARLU B., Director of MCA in Sreenivasa Institute of Technology and Management Studies (SITAMS), Chittoor.

Prof. (Dr.) Deepak Paliwal, Faculty of Sociology, Uttarakhand Open University, Haldwani-Nainital Prof. (Dr.) Dr. Anil K Dwivedi, Faculty of Pollution & Environmental Assay Research Laboratory (PEARL), Department of Botany,DDU Gorakhpur University,Gorakhpur-273009, India.

Prof. R. Ravikumar, Department of Agricultural and Rural Management, TamilNadu Agricultural University, Coimbatore-641003,Tamil Nadu, India.

Prof. (Dr.) R.Raman, Professor of Agronomy, Faculty of Agriculture, Annamalai university, Annamalai Nagar 608 002Tamil Nadu, India.

Prof. (Dr.) Ahmed Khalafallah, Coordinator of the CM Degree Program, Department of Architectural and Manufacturing Sciences, Ogden College of Sciences and Engineering Western Kentucky University 1906 College Heights Blvd Bowling Green, KY 42103-1066 Prof. (Dr.) Asmita Das , Delhi Technological University (Formerly Delhi College of Engineering), Shahbad, Daulatpur, Delhi 110042, India.

Prof. (Dr.)Aniruddha Bhattacharjya, Assistant Professor (Senior Grade), CSE Department, Amrita School of Engineering , Amrita Vishwa VidyaPeetham (University), Kasavanahalli, Carmelaram P.O., Bangalore 560035, Karnataka, India Prof. (Dr.) S. Rama Krishna Pisipaty, Prof & Geoarchaeologist, Head of the Department of Sanskrit & Indian Culture, SCSVMV University, Enathur, Kanchipuram 631561, India Prof. (Dr.) Shubhasheesh Bhattacharya, Professor & HOD(HR), Symbiosis Institute of International Business (SIIB), Hinjewadi, Phase-I, Pune- 411 057 Prof. (Dr.) Vijay Kothari, Institute of Science, Nirma University, S-G Highway, Ahmedabad 382481, India.

Prof. (Dr.) Raja Sekhar Mamillapalli, Department of Civil Engineering at Sir Padampat Singhania University, Udaipur, India.

                             

Prof. (Dr.)B. M. Kunar, Department of Mining Engineering, Indian School of Mines, Dhanbad 826004, Jharkhand, India.

Prof. (Dr.) Prabir Sarkar, Assistant Professor, School of Mechanical, Materials and Energy Engineering, Room 307, Academic Block, Indian Institute of Technology, Ropar, Nangal Road, Rupnagar 140001, Punjab, India.

Prof. (Dr.) K.Srinivasmoorthy, Associate Professor, Department of Earth Sciences, School of Physical,Chemical and Applied Sciences, Pondicherry university, R.Venkataraman Nagar, Kalapet, Puducherry 605014, India.

Prof. (Dr.) Bhawna Dubey, Institute of Environmental Science (AIES), Amity University, Noida, India.

Prof. (Dr.) P. Bhanu Prasad, Vision Specialist, Matrix vision GmbH, Germany, Consultant, TIFACCORE for Machine Vision, Advisor, Kelenn Technology, France Advisor, Shubham Automation & Services, Ahmedabad, and Professor of C.S.E, Rajalakshmi Engineering College, India.

Prof. (Dr.)P.Raviraj, Professor & Head, Dept. of CSE, Kalaignar Karunanidhi, Institute of Technology, Coimbatore 641402,Tamilnadu,India.

Prof. (Dr.) Damodar Reddy Edla, Department of Computer Science & Engineering, Indian School of Mines, Dhanbad, Jharkhand 826004, India.

Prof. (Dr.) T.C. Manjunath, Principal in HKBK College of Engg., Bangalore, Karnataka, India.

Prof. (Dr.) Pankaj Bhambri, I.T. Deptt., Guru Nanak Dev Engineering College, Ludhiana 141006, Punjab, India .

Prof. Shashikant Shantilal Patil SVKM, MPSTME Shirpur Campus, NMIMS University Vile Parle Mumbai, India.

Prof. (Dr.) Shambhu Nath Choudhary, Department of Physics, T.M. Bhagalpur University, Bhagalpur 81200, Bihar, India.

Prof. (Dr.) Venkateshwarlu Sonnati, Professor & Head of EEED, Department of EEE, Sreenidhi Institute of Science & Technology, Ghatkesar, Hyderabad, Andhra Pradesh, India.

Prof. (Dr.) Saurabh Dalela, Department of Pure & Applied Physics, University of Kota, KOTA 324010, Rajasthan, India.

Prof. S. Arman Hashemi Monfared, Department of Civil Eng, University of Sistan & Baluchestan, Daneshgah St.,Zahedan, IRAN, P.C. 98155-987 Prof. (Dr.) R.S.Chanda, Dept. of Jute & Fibre Tech., University of Calcutta, Kolkata 700019, West Bengal, India.

Prof. V.S.VAKULA, Department of Electrical and Electronics Engineering, JNTUK, University College of Engg., Vizianagaram5 35003, Andhra Pradesh, India.

Prof. (Dr.) Nehal Gitesh Chitaliya, Sardar Vallabhbhai Patel Institute of Technology, Vasad 388 306, Gujarat, India.

Prof. (Dr.) D.R. Prajapati, Department of Mechanical Engineering, PEC University of Technology,Chandigarh 160012, India.

Dr. A. SENTHIL KUMAR, Postdoctoral Researcher, Centre for Energy and Electrical Power, Electrical Engineering Department, Faculty of Engineering and the Built Environment, Tshwane University of Technology, Pretoria 0001, South Africa.

Prof. (Dr.)Vijay Harishchandra Mankar, Department of Electronics & Telecommunication Engineering, Govt. Polytechnic, Mangalwari Bazar, Besa Road, Nagpur- 440027, India.

Prof. Varun.G.Menon, Department Of C.S.E, S.C.M.S School of Engineering, Karukutty, Ernakulam, Kerala 683544, India.

Prof. (Dr.) U C Srivastava, Department of Physics, Amity Institute of Applied Sciences, Amity University, Noida, U.P-203301.India.

Prof. (Dr.) Surendra Yadav, Professor and Head (Computer Science & Engineering Department), Maharashi Arvind College of Engineering and Research Centre (MACERC), Jaipur, Rajasthan, India.

Prof. (Dr.) Sunil Kumar, H.O.D. Applied Sciences & Humanities Dehradun Institute of Technology, (D.I.T. School of Engineering), 48 A K.P-3 Gr. Noida (U.P.) 201308 Prof. Naveen Jain, Dept. of Electrical Engineering, College of Technology and Engineering, Udaipur-313 001, India.

Prof. Veera Jyothi.B, CBIT ,Hyderabad, Andhra Pradesh, India.

Prof. Aritra Ghosh, Global Institute of Management and Technology, Krishnagar, Nadia, W.B. India Prof. Anuj K. Gupta, Head, Dept. of Computer Science & Engineering, RIMT Group of Institutions, Sirhind Mandi Gobindgarh, Punajb, India.

Prof. (Dr.) Varala Ravi, Head, Department of Chemistry, IIIT Basar Campus, Rajiv Gandhi University of Knowledge Technologies, Mudhole, Adilabad, Andhra Pradesh- 504 107, India Prof. (Dr.) Ravikumar C Baratakke, faculty of Biology,Govt. College, Saundatti - 591 126, India.

                              

Prof. (Dr.) NALIN BHARTI, School of Humanities and Social Science, Indian Institute of Technology Patna, India.

Prof. (Dr.) Shivanand S.Gornale, Head, Department of Studies in Computer Science, Government College (Autonomous), Mandya, Mandya-571 401-Karanataka Prof. (Dr.) Naveen.P.Badiger, Dept.Of Chemistry, S.D.M.College of Engg. & Technology, Dharwad-580002, Karnataka State, India.

Prof. (Dr.) Bimla Dhanda, Professor & Head, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India.

Prof. (Dr.) Tauqeer Ahmad Usmani, Faculty of IT, Salalah College of Technology, Salalah, Sultanate of Oman, Prof. (Dr.) Naresh Kr. Vats, Chairman, Department of Law, BGC Trust University Bangladesh Prof. (Dr.) Papita Das (Saha), Department of Environmental Science, University of Calcutta, Kolkata, India Prof. (Dr.) Rekha Govindan , Dept of Biotechnology, Aarupadai Veedu Institute of technology , Vinayaka Missions University , Paiyanoor , Kanchipuram Dt, Tamilnadu , India Prof. (Dr.) Lawrence Abraham Gojeh, Department of Information Science, Jimma University, P.o.Box 378, Jimma, Ethiopia Prof. (Dr.) M.N. Kalasad, Department of Physics, SDM College of Engineering & Technology, Dharwad, Karnataka, India Prof. Rab Nawaz Lodhi, Department of Management Sciences, COMSATS Institute of Information Technology Sahiwal Prof. (Dr.) Masoud Hajarian, Department of Mathematics, Faculty of Mathematical Sciences, Shahid Beheshti University, General Campus, Evin, Tehran 19839,Iran Prof. (Dr.) Chandra Kala Singh, Associate professor, Department of Human Development and Family Studies, College of Home Science, CCS, Haryana Agricultural University, Hisar- 125001 (Haryana) India Prof. (Dr.) J.Babu, Professor & Dean of research, St.Joseph's College of Engineering & Technology, Choondacherry, Palai,Kerala.

Prof. (Dr.) Pradip Kumar Roy, Department of Applied Mechanics, Birla Institute of Technology (BIT) Mesra, Ranchi-835215, Jharkhand, India.

Prof. (Dr.) P. Sanjeevi kumar, School of Electrical Engineering (SELECT), Vandalur Kelambakkam Road, VIT University, Chennai, India.

Prof. (Dr.) Debasis Patnaik, BITS-Pilani, Goa Campus, India.

Prof. (Dr.) SANDEEP BANSAL, Associate Professor, Department of Commerce, I.G.N. College, Haryana, India.

Dr. Radhakrishnan S V S, Department of Pharmacognosy, Faser Hall, The University of Mississippi Oxford, MS-38655, USA Prof. (Dr.) Megha Mittal, Faculty of Chemistry, Manav Rachna College of Engineering, Faridabad (HR), 121001, India.

Prof. (Dr.) Mihaela Simionescu (BRATU), BUCHAREST, District no. 6, Romania, member of the Romanian Society of Econometrics, Romanian Regional Science Association and General Association of Economists from Romania Prof. (Dr.) Atmani Hassan, Director Regional of Organization Entraide Nationale Prof. (Dr.) Deepshikha Gupta, Dept. of Chemistry, Amity Institute of Applied Sciences,Amity University, Sec.125, Noida, India Prof. (Dr.) Muhammad Kamruzzaman, Deaprtment of Infectious Diseases, The University of Sydney, Westmead Hospital, Westmead, NSW-2145.

Prof. (Dr.) Meghshyam K. Patil , Assistant Professor & Head, Department of Chemistry,Dr. Babasaheb Ambedkar Marathwada University,Sub-Campus, Osmanabad- 413 501, Maharashtra, INDIA Prof. (Dr.) Ashok Kr. Dargar, Department of Mechanical Engineering, School of Engineering, Sir Padampat Singhania University, Udaipur (Raj.) Prof. (Dr.) Sudarson Jena, Dept. of Information Technology, GITAM University, Hyderabad, India Prof. (Dr.) Jai Prakash Jaiswal, Department of Mathematics, Maulana Azad National Institute of Technology Bhopal-India Prof. (Dr.) S.Amutha, Dept. of Educational Technology, Bharathidasan University, Tiruchirappalli620 023, Tamil Nadu-India Prof. (Dr.) R. HEMA KRISHNA, Environmental chemistry, University of Toronto, Canada.

Prof. (Dr.) B.Swaminathan, Dept. of Agrl.Economics, Tamil Nadu Agricultural University, India.

                             

Prof. (Dr.) Meghshyam K. Patil, Assistant Professor & Head, Department of Chemistry, Dr. Babasaheb Ambedkar Marathwada University, Sub-Campus, Osmanabad- 413 501, Maharashtra, INDIA Prof. (Dr.) K. Ramesh, Department of Chemistry, C .B . I. T, Gandipet, Hyderabad-500075 Prof. (Dr.) Sunil Kumar, H.O.D. Applied Sciences &Humanities, JIMS Technical campus,(I.P. University,New Delhi), 48/4 ,K.P.-3,Gr.Noida (U.P.) Prof. (Dr.) G.V.S.R.Anjaneyulu, CHAIRMAN - P.G. BOS in Statistics & Deputy Coordinator UGC DRS-I Project, Executive Member ISPS-2013, Department of Statistics, Acharya Nagarjuna University, Nagarjuna Nagar-522510, Guntur, Andhra Pradesh, India Prof. (Dr.) Sribas Goswami, Department of Sociology, Serampore College, Serampore 712201, West Bengal, India.

Prof. (Dr.) Sunanda Sharma, Department of Veterinary Obstetrics Y Gynecology, College of Veterinary & Animal Science,Rajasthan University of Veterinary & Animal Sciences,Bikaner334001, India.

Prof. (Dr.) S.K. Tiwari, Department of Zoology, D.D.U. Gorakhpur University, Gorakhpur-273009 U.P., India.

Prof. (Dr.) Praveena Kuruva, Materials Research Centre, Indian Institute of Science, Bangalore560012, INDIA Prof. (Dr.) Rajesh Kumar, Department Of Applied Physics , Bhilai Institute Of Technology, Durg (C.G.) 491001 Prof. (Dr.) Y.P.Singh, (Director), Somany (PG) Institute of Technology and Management, Garhi Bolni Road, Delhi-Jaipur Highway No. 8, Beside 3 km from City Rewari, Rewari-123401, India.

Prof. (Dr.) MIR IQBAL FAHEEM, VICE PRINCIPAL &HEAD- Department of Civil Engineering & Professor of Civil Engineering, Deccan College of Engineering & Technology, Dar-us-Salam, Aghapura, Hyderabad (AP) 500 036.

Prof. (Dr.) Jitendra Gupta, Regional Head, Co-ordinator(U.P. State Representative)& Asstt. Prof., (Pharmaceutics), Institute of Pharmaceutical Research, GLA University, Mathura.

Prof. (Dr.) N. Sakthivel, Scientist - C,Research Extension Center,Central Silk Board, Government of India, Inam Karisal Kulam (Post), Srivilliputtur - 626 125,Tamil Nadu, India.

Prof. (Dr.) Omprakash Srivastav, Centre of Advanced Study, Department of History, Aligarh Muslim University, Aligarh-202 001, INDIA. Prof. (Dr.) K.V.L.N.Acharyulu, Associate Professor, Department of Mathematics, Bapatla Engineering college, Bapatla-522101, INDIA. Prof. (Dr.) Fateh Mebarek-Oudina, Assoc. Prof., Sciences Faculty,20 aout 1955-Skikda University, B.P 26 Route El-Hadaiek, 21000,Skikda, Algeria.

NagaLaxmi M. Raman, Project Support Officer, Amity International Centre for Postharvest, Technology & Cold Chain Management, Amity University Campus, Sector-125, Expressway, Noida Prof. (Dr.) V.SIVASANKAR, Associate Professor, Department Of Chemistry, Thiagarajar College Of Engineering (Autonomous), Madurai 625015, Tamil Nadu, India (Dr.) Ramkrishna Singh Solanki, School of Studies in Statistics, Vikram University, Ujjain, India Prof. (Dr.) M.A.Rabbani, Professor/Computer Applications, School of Computer, Information and Mathematical Sciences, B.S.Abdur Rahman University, Chennai, India Prof. (Dr.) P.P.Satya Paul Kumar, Associate Professor, Physical Education & Sports Sciences, University College of Physical Education & Sports, Sciences, Acharya Nagarjuna University, Guntur.

Prof. (Dr.) Fazal Shirazi, PostDoctoral Fellow, Infectious Disease, MD Anderson Cancer Center, Houston, Texas, USA Prof. (Dr.) Omprakash Srivastav, Department of Museology, Aligarh Muslim University, Aligarh202 001, INDIA. Prof. (Dr.) Mandeep Singh walia, A.P. E.C.E., Panjab University SSG Regional Centre Hoshiarpur, Una Road, V.P.O. Allahabad, Bajwara, Hoshiarpur Prof. (Dr.) Ho Soon Min, Senior Lecturer, Faculty of Applied Sciences, INTI International University, Persiaran Perdana BBN, Putra Nilai, 71800 Nilai, Negeri Sembilan, Malaysia Prof. (Dr.) L.Ganesamoorthy, Assistant Professor in Commerce, Annamalai University, Annamalai Nagar-608002, Chidambaram, Tamilnadu, India.

Prof. (Dr.) Vuda Sreenivasarao, Professor, School of Computing and Electrical Engineering, Bahir Dar University, Bahirdar,Ethiopia Prof. (Dr.) Umesh Sharma, Professor & HOD Applied Sciences & Humanities, Eshan college of Engineering, Mathura, India.

Prof. (Dr.) K. John Singh, School of Information Technology and Engineering, VIT University, Vellore, Tamil Nadu, India.

Prof. (Dr.) Sita Ram Pal (Asst.Prof.), Dept. of Special Education, Dr.BAOU, Ahmedabad, India.

                                 

Prof. Vishal S.Rana, H.O.D, Department of Business Administration, S.S.B.T'S College of Engineering & Technology, Bambhori,Jalgaon (M.S), India.

Prof. (Dr.) Chandrakant Badgaiyan, Department of Mechatronics and Engineering, Chhattisgarh.

Dr. (Mrs.) Shubhrata Gupta, Prof. (Electrical), NIT Raipur, India.

Prof. (Dr.) Usha Rani. Nelakuditi, Assoc. Prof., ECE Deptt., Vignan’s Engineering College, Vignan University, India.

Prof. (Dr.) S. Swathi, Asst. Professor, Department of Information Technology, Vardhaman college of Engineering(Autonomous) , Shamshabad, R.R District, India.

Prof. (Dr.) Raja Chakraverty, M Pharm (Pharmacology), BCPSR, Durgapur, West Bengal, India Prof. (Dr.) P. Sanjeevi Kumar, Electrical & Electronics Engineering, National Institute of Technology (NIT-Puducherry), An Institute of National Importance under MHRD (Govt. of India), Karaikal- 609 605, India.

Prof. (Dr.) Amitava Ghosh, Professor & Principal, Bengal College of Pharmaceutical Sciences and Research, B.R.B. Sarani, Bidhannagar, Durgapur, West Bengal- 713212.

Prof. (Dr.) Om Kumar Harsh, Group Director, Amritsar College of Engineering and Technology, Amritsar 143001 (Punjab), India.

Prof. (Dr.) Mansoor Maitah, Department of International Relations, Faculty of Economics and Management, Czech University of Life Sciences Prague, 165 21 Praha 6 Suchdol, Czech Republic.

Prof. (Dr.) Zahid Mahmood, Department of Management Sciences (Graduate Studies), Bahria University, Naval Complex, Sector, E-9, Islamabad, Pakistan.

Prof. (Dr.) N. Sandeep, Faculty Division of Fluid Dynamics, VIT University, Vellore-632 014.

Mr. Jiban Shrestha, Scientist (Plant Breeding and Genetics), Nepal Agricultural Research Council, National Maize Research Program, Rampur, Chitwan, Nepal.

Prof. (Dr.) Rakhi Garg, Banaras Hindu University, Varanasi, Uttar Pradesh, India.

Prof. (Dr.) Ramakant Pandey. Dept. of Biochemistry. Patna University Patna (Bihar)-India.

Prof. (Dr.) Nalah Augustine Bala, Behavioural Health Unit, Psychology Department, Nasarawa State University, Keffi, P.M.B. 1022 Keffi, Nasarawa State, Nigeria.

Prof. (Dr.) Mehdi Babaei, Department of Engineering, Faculty of Civil Engineering, University of Zanjan, Iran.

Prof. (Dr.) A. SENTHIL KUMAR., Professor/EEE, VELAMMAL ENGINEERING COLLEGE, CHENNAI Prof. (Dr.) Gudikandhula Narasimha Rao, Dept. of Computer Sc. & Engg., KKR & KSR Inst Of Tech & Sciences, Guntur, Andhra Pradesh, India.

Prof. (Dr.) Dhanesh singh, Department of Chemistry, K.G. Arts & Science College, Raigarh (C.G.) India.

Prof. (Dr.) Syed Umar , Dept. of Electronics and Computer Engineering, KL University, Guntur, A.P., India.

Prof. (Dr.) Rachna Goswami, Faculty in Bio-Science Department, IIIT Nuzvid (RGUKT), DistrictKrishna , Andhra Pradesh - 521201 Prof. (Dr.) Ahsas Goyal, FSRHCP, Founder & Vice president of Society of Researchers and Health Care Professionals Prof. (Dr.) Gagan Singh, School of Management Studies and Commerce, Department of Commerce, Uttarakhand Open University, Haldwani-Nainital, Uttarakhand (UK)-263139 (India) Prof. (Dr.) Solomon A. O. Iyekekpolor, Mathematics and Statistics, Federal University, WukariNigeria.

Prof. (Dr.) S. Saiganesh, Faculty of Marketing, Dayananda Sagar Business School, Bangalore, India.

Dr. K.C.Sivabalan, Field Enumerator and Data Analyst, Asian Vegetable Research Centre, The World Vegetable Centre, Taiwan Prof. (Dr.) Amit Kumar Mishra, Department of Environmntal Science and Energy Research, Weizmann Institute of Science, Rehovot, Israel Prof. (Dr.) Manisha N. Paliwal, Sinhgad Institute of Management, Vadgaon (Bk), Pune, India Prof. (Dr.) M. S. HIREMATH, Principal, K.L.ESOCIETY’S SCHOOL, ATHANI, India Prof. Manoj Dhawan, Department of Information Technology, Shri Vaishnav Institute of Technology & Science, Indore, (M. P.), India Prof. (Dr.) V.R.Naik, Professor & Head of Department, Mechancal Engineering , Textile & Engineering Institute, Ichalkaranji (Dist. Kolhapur), Maharashatra, India Prof. (Dr.) Jyotindra C. Prajapati,Head, Department of Mathematical Sciences, Faculty of Applied Sciences, Charotar University of Science and Technology, Changa Anand -388421, Gujarat, India Prof. (Dr.) Sarbjit Singh, Head, Department of Industrial & Production Engineering, Dr BR Ambedkar National Institute of Technology, Jalandhar, Punjab,India

                                

Prof. (Dr.) Professor Braja Gopal Bag, Department of Chemistry and Chemical Technology, Vidyasagar University, West Midnapore Prof. (Dr.) Ashok Kumar Chandra, Department of Management, Bhilai Institute of Technology, Bhilai House, Durg (C.G.) Prof. (Dr.) Amit Kumar, Assistant Professor, School of Chemistry, Shoolini University, Solan, Himachal Pradesh, India Prof. (Dr.) L. Suresh Kumar, Mechanical Department, Chaitanya Bharathi Institute of Technology, Hyderabad, India.

Scientist Sheeraz Saleem Bhat, Lac Production Division, Indian Institute of Natural Resins and Gums, Namkum, Ranchi, Jharkhand Prof. C.Divya , Centre for Information Technology and Engineering, Manonmaniam Sundaranar University, Tirunelveli - 627012, Tamilnadu , India Prof. T.D.Subash, Infant Jesus College Of Engineering and Technology, Thoothukudi Tamilnadu, India Prof. (Dr.) Vinay Nassa, Prof. E.C.E Deptt., Dronacharya.Engg. College, Gurgaon India.

Prof. Sunny Narayan, university of Roma Tre, Italy.

Prof. (Dr.) Sanjoy Deb, Dept. of ECE, BIT Sathy, Sathyamangalam, Tamilnadu-638401, India.

Prof. (Dr.) Reena Gupta, Institute of Pharmaceutical Research, GLA University, Mathura-India Prof. (Dr.) P.R.SivaSankar, Head Dept. of Commerce, Vikrama Simhapuri University Post Graduate Centre, KAVALI - 524201, A.P., India Prof. (Dr.) Mohsen Shafiei Nikabadi, Faculty of Economics and Management, Industrial Management Department, Semnan University, Semnan, Iran.

Prof. (Dr.) Praveen Kumar Rai, Department of Geography, Faculty of Science, Banaras Hindu University, Varanasi-221005, U.P. India Prof. (Dr.) Christine Jeyaseelan, Dept of Chemistry, Amity Institute of Applied Sciences, Amity University, Noida, India Prof. (Dr.) M A Rizvi, Dept. of Computer Engineering and Applications , National Institute of Technical Teachers' Training and Research, Bhopal M.P. India Prof. (Dr.) K.V.N.R.Sai Krishna, H O D in Computer Science, S.V.R.M.College,(Autonomous), Nagaram, Guntur(DT), Andhra Pradesh, India.

Prof. (Dr.) Ashok Kr. Dargar, Department of Mechanical Engineering, School of Engineering, Sir Padampat Singhania University, Udaipur (Raj.) Prof. (Dr.) Asim Kumar Sen, Principal , ST.Francis Institute of Technology (Engineering College) under University of Mumbai , MT. Poinsur, S.V.P Road, Borivali (W), Mumbai, 400103, India, Prof. (Dr.) Rahmathulla Noufal.E, Civil Engineering Department, Govt.Engg.College-Kozhikode Prof. (Dr.) N.Rajesh, Department of Agronomy, TamilNadu Agricultural University -Coimbatore, TamilNadu, India Prof. (Dr.) Har Mohan Rai, Professor, Electronics and Communication Engineering, N.I.T. Kurukshetra 136131,India Prof. (Dr.) Eng. Sutasn Thipprakmas from King Mongkut, University of Technology Thonburi, Thailand Prof. (Dr.) Kantipudi MVV Prasad, EC Department, RK University, Rajkot.

Prof. (Dr.) Jitendra Gupta,Faculty of Pharmaceutics, Institute of Pharmaceutical Research, GLA University, Mathura.

Prof. (Dr.) Swapnali Borah, HOD, Dept of Family Resource Management, College of Home Science, Central Agricultural University, Tura, Meghalaya, India Prof. (Dr.) N.Nazar Khan, Professor in Chemistry, BTK Institute of Technology, Dwarahat-263653 (Almora), Uttarakhand-India Prof. (Dr.) Rajiv Sharma, Department of Ocean Engineering, Indian Institute of Technology Madras, Chennai (TN) - 600 036, India.

Prof. (Dr.) Aparna Sarkar, PH.D. Physiology, AIPT, Amity University , F 1 Block, LGF, Sector125,Noida-201303, UP, India.

Prof. (Dr.) Manpreet Singh, Professor and Head, Department of Computer Engineering, Maharishi Markandeshwar University, Mullana, Haryana, India.

Prof. (Dr.) Sukumar Senthilkumar, Senior Researcher, Advanced Education Center of Jeonbuk for Electronics and Information Technology, Chon Buk National University, Chon Buk, 561-756, SOUTH KOREA. .

Prof. (Dr.) Hari Singh Dhillon, Assistant Professor, Department of Electronics and Communication Engineering, DAV Institute of Engineering and Technology, Jalandhar (Punjab), INDIA. .

Prof. (Dr.) Poonkuzhali, G., Department of Computer Science and Engineering, Rajalakshmi Engineering College, Chennai, INDIA. .

                                 

Prof. (Dr.) Bharath K N, Assistant Professor, Dept. of Mechanical Engineering, GM Institute of Technology, PB Road, Davangere 577006, Karnataka, India.

Prof. (Dr.) F.Alipanahi, Assistant Professor, Islamic Azad University, Zanjan Branch, Atemadeyeh, Moalem Street, Zanjan IRAN. Prof. Yogesh Rathore, Assistant Professor, Dept. of Computer Science & Engineering, RITEE, Raipur, India Prof. (Dr.) Ratneshwer, Department of Computer Science (MMV),Banaras Hindu University Varanasi-221005, India.

Prof. Pramod Kumar Pandey, Assistant Professor, Department Electronics & Instrumentation Engineering, ITM University, Gwalior, M.P., India.

Prof. (Dr.)Sudarson Jena, Associate Professor, Dept.of IT, GITAM University, Hyderabad, India Prof. (Dr.) Binod Kumar, PhD(CS), M.Phil(CS), MIEEE,MIAENG, Dean & Professor( MCA), Jayawant Technical Campus(JSPM's), Pune, India.

Prof. (Dr.) Mohan Singh Mehata, (JSPS fellow), Assistant Professor, Department of Applied Physics, Delhi Technological University, Delhi Prof. Ajay Kumar Agarwal, Asstt. Prof., Deptt. of Mech. Engg., Royal Institute of Management & Technology, Sonipat (Haryana), India.

Prof. (Dr.) Siddharth Sharma, University School of Management, Kurukshetra University, Kurukshetra, India.

Prof. (Dr.) Satish Chandra Dixit, Department of Chemistry, D.B.S.College, Govind Nagar,Kanpur208006, India.

Prof. (Dr.) Ajay Solkhe, Department of Management, Kurukshetra University, Kurukshetra, India.

Prof. (Dr.) Neeraj Sharma, Asst. Prof. Dept. of Chemistry, GLA University, Mathura, India.

Prof. (Dr.) Basant Lal, Department of Chemistry, G.L.A. University, Mathura, India.

Prof. (Dr.) T Venkat Narayana Rao, C.S.E, Guru Nanak Engineering College, Hyderabad, Andhra Pradesh, India.

Prof. (Dr.) Rajanarender Reddy Pingili, S.R. International Institute of Technology, Hyderabad, Andhra Pradesh, India.

Prof. (Dr.) V.S.Vairale, Department of Computer Engineering, All India Shri Shivaji Memorial Society College of Engineering, Kennedy Road, Pune-411 001, Maharashtra, India.

Prof. (Dr.) Vasavi Bande, Department of Computer Science & Engineering, Netaji Institute of Engineering and Technology, Hyderabad, Andhra Pradesh, India Prof. (Dr.) Hardeep Anand, Department of Chemistry, Kurukshetra University Kurukshetra, Haryana, India.

Prof. Aasheesh shukla, Asst Professor, Dept. of EC, GLA University, Mathura, India.

Prof. S.P.Anandaraj., CSE Dept, SREC, Warangal, India.

Prof. (Dr.) Chitranjan Agrawal, Department of Mechanical Engineering, College of Technology & Engineering, Maharana Pratap University of Agriculture & Technology, Udaipur- 313001, Rajasthan, India.

Prof. (Dr.) Rangnath Aher, Principal, New Arts, Commerce and Science College, Parner, DistAhmednagar, M.S. India.

Prof. (Dr.) Chandan Kumar Panda, Department of Agricultural Extension, College of Agriculture, Tripura, Lembucherra-799210 Prof. (Dr.) Latika Kharb, IP Faculty (MCA Deptt), Jagan Institute of Management Studies (JIMS), Sector-5, Rohini, Delhi, India.

Raj Mohan Raja Muthiah, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts.

Prof. (Dr.) Chhanda Chatterjee, Dept of Philosophy, Balurghat College, West Bengal, India.

Prof. (Dr.) Mihir Kumar Shome , H.O.D of Mathematics, Management and Humanities, National Institute of Technology, Arunachal Pradesh, India Prof. (Dr.) Muthukumar .Subramanyam, Registrar (I/C), Faculty, Computer Science and Engineering, National Institute of Technology, Puducherry, India.

Prof. (Dr.) Vinay Saxena, Department of Mathematics, Kisan Postgraduate College, Bahraich – 271801 UP, India.

Satya Rishi Takyar, Senior ISO Consultant, New Delhi, India.

Prof. Anuj K. Gupta, Head, Dept. of Computer Science & Engineering, RIMT Group of Institutions, Mandi Gobindgarh (PB) Prof. (Dr.) Harish Kumar, Department of Sports Science, Punjabi University, Patiala, Punjab, India.

Prof. (Dr.) Mohammed Ali Hussain, Professor, Dept. of Electronics and Computer Engineering, KL University, Green Fields, Vaddeswaram, Andhra Pradesh, India.

                                           

Prof. (Dr.) Manish Gupta, Department of Mechanical Engineering, GJU, Haryana, India.

Prof. Mridul Chawla, Department of Elect. and Comm. Engineering, Deenbandhu Chhotu Ram University of Science & Technology, Murthal, Haryana, India.

Prof. Seema Chawla, Department of Bio-medical Engineering, Deenbandhu Chhotu Ram University of Science & Technology, Murthal, Haryana, India.

Prof. (Dr.) Atul M. Gosai, Department of Computer Science, Saurashtra University, Rajkot, Gujarat, India.

Prof. (Dr.) Ajit Kr. Bansal, Department of Management, Shoolini University, H.P., India.

Prof. (Dr.) Sunil Vasistha, Mody Institute of Tecnology and Science, Sikar, Rajasthan, India.

Prof. Vivekta Singh, GNIT Girls Institute of Technology, Greater Noida, India.

Prof. Ajay Loura, Assistant Professor at Thapar University, Patiala, India.

Prof. Sushil Sharma, Department of Computer Science and Applications, Govt. P. G. College, Ambala Cantt., Haryana, India.

Prof. Sube Singh, Assistant Professor, Department of Computer Engineering, Govt. Polytechnic, Narnaul, Haryana, India.

Prof. Himanshu Arora, Delhi Institute of Technology and Management, New Delhi, India.

Dr. Sabina Amporful, Bibb Family Practice Association, Macon, Georgia, USA. Dr. Pawan K. Monga, Jindal Institute of Medical Sciences, Hisar, Haryana, India.

Dr. Sam Ampoful, Bibb Family Practice Association, Macon, Georgia, USA. Dr. Nagender Sangra, Director of Sangra Technologies, Chandigarh, India.

Vipin Gujral, CPA, New Jersey, USA. Sarfo Baffour, University of Ghana, Ghana.

Monique Vincon, Hype Softwaretechnik GmbH, Bonn, Germany.

Natasha Sigmund, Atlanta, USA. Marta Trochimowicz, Rhein-Zeitung, Koblenz, Germany.

Kamalesh Desai, Atlanta, USA. Vijay Attri, Software Developer Google, San Jose, California, USA. Neeraj Khillan, Wipro Technologies, Boston, USA. Ruchir Sachdeva, Software Engineer at Infosys, Pune, Maharashtra, India.

Anadi Charan, Senior Software Consultant at Capgemini, Mumbai, Maharashtra.

Pawan Monga, Senior Product Manager, LG Electronics India Pvt. Ltd., New Delhi, India.

Sunil Kumar, Senior Information Developer, Honeywell Technology Solutions, Inc., Bangalore, India.

Bharat Gambhir, Technical Architect, Tata Consultancy Services (TCS), Noida, India.

Vinay Chopra, Team Leader, Access Infotech Pvt Ltd. Chandigarh, India.

Sumit Sharma, Team Lead, American Express, New Delhi, India.

Vivek Gautam, Senior Software Engineer, Wipro, Noida, India.

Anirudh Trehan, Nagarro Software Gurgaon, Haryana, India.

Manjot Singh, Senior Software Engineer, HCL Technologies Delhi, India.

Rajat Adlakha, Senior Software Engineer, Tech Mahindra Ltd, Mumbai, Maharashtra, India.

Mohit Bhayana, Senior Software Engineer, Nagarro Software Pvt. Gurgaon, Haryana, India.

Dheeraj Sardana, Tech. Head, Nagarro Software, Gurgaon, Haryana, India.

Naresh Setia, Senior Software Engineer, Infogain, Noida, India.

Raj Agarwal Megh, Idhasoft Limited, Pune, Maharashtra, India.

Shrikant Bhardwaj, Senior Software Engineer, Mphasis an HP Company, Pune, Maharashtra, India.

Vikas Chawla, Technical Lead, Xavient Software Solutions, Noida, India.

Kapoor Singh, Sr. Executive at IBM, Gurgaon, Haryana, India.

Ashwani Rohilla, Senior SAP Consultant at TCS, Mumbai, India.

Anuj Chhabra, Sr. Software Engineer, McKinsey & Company, Faridabad, Haryana, India.

Jaspreet Singh, Business Analyst at HCL Technologies, Gurgaon, Haryana, India.

TOPICS OF INTEREST Topics of interest include, but are not limited to, the following:  Social networks and intelligence  Social science simulation  Information retrieval systems  Technology management  Digital libraries for e-learning  Web-based learning, wikis and blogs  Operational research  Ontologies and meta-data standards  Engineering problems and emerging application  Agent based modeling and systems  Ubiquitous computing  Wired and wireless data communication networks  Mobile Ad Hoc, sensor and mesh networks  Natural language processing and expert systems  Monte Carlo methods and applications  Fuzzy logic and soft computing  Data mining and warehousing  Software and web engineering  Distributed AI systems and architectures  Neural networks and applications  Search and meta-heuristics  Bioinformatics and scientific computing  Genetic network modeling and inference  Knowledge and information management techniques  Aspect-oriented programming  Formal and visual specification languages  Informatics and statistics research  Quantum computing  Automata and formal languages  Computer graphics and image processing  Web 3D and applications  Grid computing and cloud computing  Algorithms design  Genetic algorithms  Compilers and interpreters  Computer architecture & VLSI  Advanced database systems  Digital signal and image processing  Distributed and parallel processing  Information retrieval systems  Technology management  Automation and mobile robots  Manufacturing technology  Electrical technology  Applied mathematics  Automatic control  Nuclear engineering  Computational physics  Computational chemistry

TABLE OF CONTENTS International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) ISSN (Print): 2279-0047, ISSN (Online): 2279-0055 (December-2014 to February-2015, Issue 11, Volume 1, 2 & 3) Issue 11, Volume 1 Paper Code

Paper Title

Page No.

IJETCAS 15-109

Mathematical Methods for Separation of Overlapping Asymmetrical Peaks in Spectroscopy and Chromatography. Case study: One-Dimensional Signals J. Dubrovkin

01-08

IJETCAS 15-111

Content Based Image Retrieval System with Feature Extraction and Recently Retrieved Image Library Seema H. Jadhav, Dr. Sunita Singh, Dr. Hari Singh

09-16

IJETCAS 15-117

A Dual Security Approach for Medical Images using Encryption and Watermarking Optimized by Differential Evolution Algorithm Mr. CH.Venu Gopal Reddy, Dr. Siddaiah.P

17-29

IJETCAS 15-119

Energy Efficient Routing Protocol for MANET: A Survey GANESH GUPTA, MADNESH KUMAR GUPTA, ASHOK K. RAGHAV

30-37

IJETCAS 15-121

Solving Generalized Assignment Problem with Genetic Algorithm and Lower Bound Theory Mr. VikasThada, Mr. Utpal Shrivastava, Ms. Meenu Vijarania

38-42

IJETCAS 15-123

Priority Based Scheduling in a Federated Cloud Environment A. Stanislas, L. Arockiam

43-48

IJETCAS 15-125

Swarm Intelligence Techniques for Optimization in Data Clustering Ms. Dipali Kharche, Prof. A.D.Thakare

49-53

IJETCAS 15-129

Blind digital watermarking using AES technique for colour images Rahul Saxena, Nirupma Tiwari, Manoj Kumar Ramaiya

54-57

IJETCAS 15-130

Mapping and Partitioning of Task Graphs Using Kernighan-Lin/Fiduccia-Mattheyses Algorithm Ashish Mishra, Raja Jimit, Abhijit Rameshwar Asati, Kota Solomon Raju

58-61

IJETCAS 15-132

Ultrasonic and Pyroelectric Sensory Fusion System for Indoor Human/ Robot Localization and Monitoring Azhar K H

62-66

IJETCAS 15-134

Rough Set Techniques for Text Classification and Sentiment Analysis in Social Media G. K. Panda, Jayanta Mondal

67-74

IJETCAS 15-135

OFDM Channel Analysis between FFT and Wavelet Transform Techniques Quosay Jalil, S Nagakishore Bhavanam

75-79

IJETCAS 15-138

Bianchi Type IX Cosmological Model with Varying LambdaTerm R.K. Tiwari, D.K. Tiwari, C.Chauhan

80-83

IJETCAS 15-140

Thermo Physical Properties of Nano Ferro fluids L.S.V Prasad, Paul Douglas Sanasi, V.Srinivas

84-88

IJETCAS 15-143

A Novel Efficient and Accurate Detection Model to Detect Emerging Attacks in Network Supriya Gupta, Ankur Goyel

89-93

IJETCAS 15-149

Analysis of 5-Iodo-Uracil for its Infrared Spectra, Laser Raman Spectra and its Thermodynamic Functions Dr. Pradeep Kumar Sharma

94-101

Issue 11, Volume 2 Paper Code

Paper Title

Page No.

IJETCAS 15-152

Evaluation of the Peak Location Uncertainty in Spectra. Case Study: Exponentially Modified Asymmetrical Gaussian Doublets J. Dubrovkin

102-109

IJETCAS 15-155

LEBESGUE CONVERGENCE OF MODIFIED COMPLEX TRIGONOMETRIC SUMS JATINDERDEEP KAUR, SANDEEP KAUR AND S.S. BHATIA

110-114

IJETCAS 15-158

MULTIPURPOSE SMART CARD USING ADVANCED ENCRYPTION STANDARD ALGORITHM Nusrath A

115-118

IJETCAS 15-164

A Study on the Academic Performance of the Students by Applying Multiple Linear Regression Analysis using the method of Least Squares G. Narasinga Rao

119-121

IJETCAS 15-166

Analysis of Incompressible fluid flow over wedge with different angles Dr. Deepak Sharma, Mr. Swapnil Jain, Ankush Kumar

122-126

IJETCAS 15-174

Pairwise Key Distribution Mechanism for Heterogeneous Sensor Network Kanwalinderjit Gagneja

127-133

IJETCAS 15-175

Influence and Role of Technology on Stress: A Mathematical Analysis Dr. M.S.Saleem Basha, Dr. Esam Al Lawati, Mrs. Gargi Bhattacharya and Dr. Nasir Ahmed Khan

134-139

IJETCAS 15-176

Design of Unified Power Quality Conditioner (UPQC) To Improve the Power Quality Problems by Using Instantaneous Real & Reactive Power Theory S.Natarajan, Dr.M.AntoBennet, M. Manimaraboopathy,S.Sankararnarayan,N.Srinivasan

140-147

IJETCAS 15-177

Enhanced More Secure Trust Based Routing Scheme for DSR Protocol in MANETs Ankita Sahu, Vikas Sejwar

148-151

IJETCAS 15-178

Recognising Natural Emotions Using Artificial Neural Network In A Classroom Setting Rhea Mahajan, Remia Langer

152-156

IJETCAS 15-180

An Approach to Solve Single Machine Job Scheduling Problem using Heuristic Algorithm Satyasundara Mahapatra, Dr. Rati Ranjan Dash, Dr. Sateesh Kumar Pradhan

157-163

IJETCAS 15-181

Performance Comparison of Unicast and Multicast Routing Protocol in MANET using NS2 Mandeep Singh, Prakash Rao Ragiri

164-168

IJETCAS 15-183

A Study on Mining Approach under Cyber Crime Analysis Priyanka Maan, Meghna Sharma

169-173

IJETCAS 15-190

Design of Universal Gates Based on Reversible Logic Deepali Samtani, Naman Kumar Patel, Aditya Gupta, Shruti Jain

174-178

IJETCAS 15-198

Turbulent Heat Transfer and Pressure Drop in an Internally Ribbed Rectangular Duct Sohil Akahter, Ishwar Singh

179-182

IJETCAS 15-199

Brain Image Segmentation Using Chan-Vese algorithm using Active Contours and Level Set Functions Payal, Satinderjeet Singh

183-187

IJETCAS 15-203

Thermoluminescence of nanocrystalline Eu doped BAM Phosphor Vinit Kumar, M. K. Dhasmana, R.B.S. Rawat

188-189

Issue 11, Volume 3 Paper Code

Paper Title

Page No.

IJETCAS 15-204

Calculation of the Probability for CR-39 Alpha Particles Detectors Ali Farhan Nadir, Noori H.N. Al-Hashimi, Abdul Ridha.H.Subber

190-193

IJETCAS 15-208

Mathematical Analysis of Asymmetrical Spectral Lines.Case study: Exponentially Modified Functions J. Dubrovkin

194-201

IJETCAS 15-213

Bezier Surface Reconstruction using Artificial Neural Networks Kavita, Navin Rajpal

202-206

IJETCAS 15-219

A Dual Mode DNA Cryptography: Bio Inspired Approach for Information Security Fakhrayh Al Harrasi, Dr. M.S.Saleem Basha, Dr. A. Mohamed Abbas and Mohamed Jameel Hashmi

207-212

IJETCAS 15-223

Optimization of land and water resource in Nakatiya Minor Canal command area in Udham Singh Nagar, Uttarakhand, India

213-216

G.S.Yurembam and Vinod Kumar IJETCAS 15-229

Interactive Animations to Present Academic Subjects to Elementary School Children Cristian Javier Cauich Valle, Lizzie Narváez Díaz, Cinhtia M. González Segura

217-222

IJETCAS 15-230

Citizen Charter Validation A K Tripathy, M R Patra and S Pradhan

223-226

IJETCAS 15-231

Type 2 Diabetes, Therapeutic Targets and Potential Drugs: A Review Neha Verma and Usha Chouhan

227-233

IJETCAS 15-236

Database Management - Inline Queries vs. Stored Procedures–An Extended Analysis Er. Shobhit Gupta

234-238

IJETCAS 15-242

An Analytical Study of Weakly Nonlinear Convection in a Horizontal Mushy Layer Prof.Dr.P.K.Srimani, Mr. R.Parthasarathi

239-246

IJETCAS 15-243

Strategic Planning towards Effective Management of New Technology in Open Cast Mines Anand Pd. Sinha, Supriyo Roy

247-254

IJETCAS 15-244

ANALYSIS OF BANK’S UNSTRUCTURED DATA USING MAP REDUCE TECHNIQUE Jeny George, A. Karthika, Dr. Ulagamuthalvi V

255-257

IJETCAS 15-245

SOCIAL AWARENESS THROUGH NEW MEDIA Niket Mehta, Dr. Suparna Dutta, Dr. Asit Bandyopadhyay

258-261

IJETCAS 15-246

IDENTIFYING GENETIC MUTATION RARE GENETIC DISORDER BY ANALYZING CHARACTERISTICS OF GENOTYPE-PHENOTYPE BY IMPLEMENTING APRIORI ALGORITHM Bipin nair B J ,Ratheesh A, Koushik K S

262-265

IJETCAS 15-247

Mining Frequent and Similar Patterns with Attribute Oriented Induction High Level Emerging Pattern (AOI-HEP) Data Mining Technique Spits Warnars

266-276

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Mathematical Methods for Separation of Overlapping Asymmetrical Peaks in Spectroscopy and Chromatography. Case study: One-Dimensional Signals J. Dubrovkin Multidisciplinary Department, Western Galilee College, 2421 Acre, Israel Abstract: Mathematical methods for separation of overlapping asymmetrical peaks in spectroscopy and chromatography were reviewed. Evaluation of initial peak parameters and curve fitting algorithms has been discussed.

Keywords: separation of overlapping analytical signals; asymmetrical peak models; curve fitting; spectroscopy; chromatography.

I. Introduction Instrumental methods of analysis are widely used to determine the physicochemical properties of natural and synthetic materials. Data produced by analytical instruments is named as analytical signal [1] (AS). AS is the source of physicochemical information about the sample under study. AS of a multicomponent mixture is a superposition of the ASs of the mixture components. E. g., in spectroscopy [2] and chromatography [3] AS can be represented as a sum of elementary peaks. Peaks have symmetrical (Gaussian, Lorentzian and Voigt) and asymmetrical [4-6] bell-shaped shapes. Peaks are often strongly overlapped. The instrumental methods cannot completely resolve overlapping peaks; therefore, mathematical signal processing must be performed.

In order to avoid terminological confusion, a short description of some terms will follow.

Extraction of elementary peaks from badly resolved AS means complete separation of peaks. The term "separation" seems to us the most relevant to chromatography where complete separation of peaks indicates a good separation of pure components of a mixture. Resolution enhancement, often referred as “deconvolution”, causes partial separation due to the narrowing of the peaks. These terms have specific meaning. Resolution is one of the main concepts in spectroscopy [2] which describes the ability of a spectral instrument to resolve overlapping peaks. Deconvolution is the inversion of a convolution of "true" (undisturbed) AS with instrumental (kernel) function of analytical tool [7]. We considered resolution enhancement as pseudo deconvolution [6].

The separation problem should be considered, according to how much a priori information on the AS is available. If all elementary components of AS are known (the white box model), then the separation of overlapping components is not complicated. In contrary, blind source separation, without knowing the AS components (the black box model) and some constrains, in principle impossible [8]. Fortunately, in spectroscopy and chromatography the gray box model is valid [9]. This model is based on a priori information about peak shape and peak parameters; approximated values of these parameters can be evaluated in the pre-processing step.

Mathematical methods of separation of overlapping symmetrical peaks have been discussed in numerous articles (see references in [1] and reviews [9-13]). A smaller number of papers has been devoted to modeling and separation of overlapping asymmetrical peaks in spectroscopy, chromatography, polarography and related techniques [4-6, 9-49], since this problem is more complicated and less common than that of the processing of symmetrical peaks. Moreover, similar mathematical algorithms are often developed and used in different fields of instrumental analytical chemistry making it difficult to understand the separation problem as a whole.

Sometimes, chemist "cannot see forest for the trees" because of unnecessary technical details which should be included in the supplementary materials section of a manuscript.

In the present chemistry-oriented analytical review, the most common methods of separation of overlapping asymmetrical peaks in spectroscopy and chromatography are critically discussed. The advantages and disadvantages of these methods are demonstrated using sufficiently simple mathematical tools which should be clear to wide audience. Since the matrix approach to the separation problem was thoroughly reviewed in chemometrics literature, present study was mainly dedicated to the one-dimensional ASs. In what follows, for the sake of simplicity, term “peak” is used for short of phrases "spectral line and band" and "chromatographic peak".

Standard algebraic notations are used throughout the article. All calculations were performed and the plots were built using the MATLAB program.

IJETCAS 15-109; © 2015, IJETCAS All Rights Reserved

Page 1

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08

II. Theory a. Model In general, AS is modeled as a linear combination of peaks, a baseline (

) and a random noise ( ):

where is the maximal peak amplitude (peak height), is the shape of the th peak, is abscissa of the AS (e.g., time, wavelength), is the abscissa of the peak maximum, is the vector of the peak-shape parameters which define the full peak width at half-maximum (FWHM, and the peak asymmetry. It is common to assume that all peaks have the same shape and that has a normal distribution with zero mean.

Another noise models were considered in [4, 50, 51].

To simplify the analysis of asymmetrical peaks, has been transformed to the new dimensionless variable [6]: where is a peak shape parameter. Transformation (2) is applicable to all peak models that include dimensionless asymmetry coefficients. However, asymmetry coefficients which have dimension are also sometimes included for the sake of better fitting theoretical peak profile to experimental data. For example, the Parabolic-Variance Modified Gaussian model [33]: where contains coefficient , which unit of measure is .

The Exponential-Gaussian Hybrid (EGH) model [34] is a partial case of Eq. 3 when . The denominator of the fraction (Eq. 3) must have a positive value. Since the coefficient is physically unreasonable, the corresponding models were not considered in [6].

The bi-Gaussian peak model [35] can be transformed to the dimensionless variable (2):

where is the ratio of the FWHMs of the right and the left peak regions.

The Tail Modifier of symmetrical Gaussian, Lorentzian, and Voigt profiles [36] is also the asymmetrical multiplier used for improving AS fitting.

The generalized exponentially modified Gaussian function (GEMG) is a linear combination of two EMG peaks obtained by convolution of Gaussian peak with two different decay functions [19]. The essential drawback of GEMG is doubling of independent elementary components of AS. Some peak models have very limited application since they are focused only on a specific problem. For example, emission spectral lines of neutral atoms in plasmas were represented as convolution of Gaussian and complex integral asymmetrical function based on rigorous physical foundation [25].

Mathematical expressions of some non-integral peak models (for ) in the spatial and in the frequency domains are given in the Table.

Mathematical analysis of peak asymmetry was performed by decomposition of the theoretical peak shape ( ) to the sum and to the product of symmetrical and asymmetrical parts [6]. Asymmetry factor as a measure of peak tailing (AFPT) was evaluated using different empirical parameters [49]. The ratios of the extremal intensities of the first-order derivative and of the satellite intensities of the second-order derivative were taken as the asymmetry factor of the proximal and of the distal parts of the peak tails, respectively [6]. In another approach the measured chromatographic peak was approximated by the function [49]; the symmetrical peak ( ) with the same peak maximum position, height and width as was generated. Difference between and was used for calculating of the AFPT. However, since all asymmetrical factors are not related to the physical origin of peak asymmetry, they are only phenomenological characteristics of asymmetry.

Real-life model of AS includes systematic errors (unknown background or baseline) due to the minor unidentified components and slow instrumental drifts. Background uncertainties greatly complicate the curve fitting problem since additional background parameters must be included in the fitting algorithm.

In many cases baseline may be approximated by the low-order polynomial in the peak-free regions [48].

However, these regions are usually absent in chromatograms of complex mixtures due to the overlap between baseline and long tails of asymmetrical peaks [30]. Linear baseline can be suppressed by the second-order differentiation.

Since the tails of asymmetrical peaks go far from the peak center, establishing of integration limits of chromatographic peaks becomes serious problem in the presence of baseline. For this purpose the second-order derivative of chromatographic AS has been used [41].

Table. Peak shape models

IJETCAS 15-109; Š 2015, IJETCAS All Rights Reserved

Page 2

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08

Model

Mathematical expression

PMG [33] Stancik -G[42] EMG [4] Dobosh[43] Stancik -L [42] Log-normal[44]

Losev [46]

Exponentially Modified Losev BWF[47]

Pap[48] where L=

*

,

is the angular frequency and

Numerous background shapes were included in Casa Software program packet designed for processing X-ray photoelectric spectra [36]. Flexible non-linear background functions based on cubic spline polynomials were adjusted interactively using mouse control or were optimized in the fitting to the peak model. b. Curve fitting Extraction of pure components from overlapping peaks by fitting the model (1) to the experimental data is one of the main pre-processing tasks in instrumental analytical chemistry [1]. The reliability of the peak parameters obtained by curve fitting will affect the reliability of the results of the qualitative and quantitative analyses.

Simple approach to curve fitting: Non-linear Iterative Curve Fitting (NLICF) has been discussed by prof. Tom O'Haver in his practical guide to signal processing [52]. NLICF includes the following steps: 1. Determination of initial parameters of the model (1).

2. Curve fitting by minimizing the sum of the squared differences (the objective function) between the measured AC ( ) and fitted data ( ) with respect to the vector of unknown parameters of the peaks and baseline: 3. If the fitting error is greater than the specified error, then the model parameters, obtained in step 2 must be corrected, and a computer program loops back around to step 3.

Impact of statistical properties of the measured data and of the model parameters on minimization procedures has been thoroughly discussed by [53, 54].

The following object function takes into account a priory known values of the model parameters:

where is a weighting factor, is the initial value of parameter, is the standard deviation of the normally distributed which mean value is zero.

The main problem of the curve fitting is finding the global minimum of the objective function instead of the local minima or "plateaus" (when fitting parameters do not affect the objective function) [18]. Convergence of curve fitting algorithm in the neighborhood of the global solution point depends on the correctness of the model and on how the initial parameters are close to the true values [53]. Rate of convergence also depends on the numerical method of curve fitting [23]. It was pointed out that the probability of finding accurate values of peak

IJETCAS 15-109; Š 2015, IJETCAS All Rights Reserved

Page 3

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08

parameters (the global solution) increases with increasing resolution [23]. This conclusion can be explained using MATLAB function nlinfit as an example. It uses the Gauss-Newton algorithm with Levenberg-Marquardt (LM) modifications for global convergence which depends on the sensitivity of the objective function to small changes in the model parameter values. The sensitivity increases with increasing resolution. Since the resolution of overlapped ASs can be enhanced by differentiation, the convergence of curve fitting is improved using low-order derivatives of analytical signal [23]. However, the noise enhancement in derivatives is a serious disadvantage of this method.

The NLICF algorithms may be divided into two main groups: the classical optimization methods for small- and large residual problems [54] and the nature-inspired (genetic and immune) methods [18], e. g., a hybrid genetic algorithm (LOGA [20]). The LM and genetic algorithms were used for the modeling of HPLC retention surfaces [17, 18].

Brief description of genetic algorithm is given in the MATLAB software documentation and by [18]. This algorithm begins by creating a random set of initial parameters of AS. At the following steps the new "children" sets were produced from the "parent" sets using the ''best" (elite) estimated parameters and by modification (mutation) of other parameters. This process is the most critical to a rapid convergence of the algorithm to the global minimum of the objective function. Various stopping conditions of the genetic algorithm are listed in the documentation of the MATLAB function ga.m.

The first analytical application of the immune network method which models the defending process of an immune biological system against its invaders was given by [40]. Symmetrical Gaussians were used as models.

It was shown that the immune algorithm is more rapid than LM method.

Practical recommendations for choosing the quasi-optimal curve fitting algorithm have been discussed in details by [23]. However, in each particular case, the decision, what is the best algorithm for NLICF, should be made only by trial and error procedure.

Model correctness can be improved by using flexible models of asymmetrical peak incorporating more parameters [36]. Unfortunately, if a number of model parameters increases, then the fitting algorithm is complicated, and its rate of convergence significantly decreases.

Study [21] was dedicated to a priory prediction of the possibility of obtaining inaccurate parameters of deconvoluted AS. The prediction based on different resolution criteria was performed using retention data set of 16 aromatic compounds eluted in methanol-water and acetonitrile-water mixtures. It was found that the multivariate selectivity obtained from the net analyte signals of mixture components is the best failure prediction measure of deconvolution.

Peak fitting to the Gaussian shape was performed using adaptive digital filters in the spatial and frequency domains [55]. For baseline suppression and resolution enhancement the second-order derivatives were taken. In the case of weakly asymmetric peaks this method can be useful only for evaluation of initial peak parameters.

Peak fitting to the Log-Normal shape was performed using the pseudo-phase representation of parameters characterizing the shape of fluorescence spectra [44, 45]. This method based on the rigorous physical foundation is applicable only to fluorescence spectra. c. Determination of initial peak parameters 1. Peak detection Statistical methods for detecting peaks are based on the testing the null hypothesis ("some sample of AS belongs to the peak") against the alternative hypothesis ("this is a noise sample") [50]. The weighted sum of the baselinefree AS values is compared with a threshold. The drawback of this method is a broadening of AS. Another onepoint algorithm compares the first-order derivative of AS with some threshold. This algorithm does not require correction of the constant baseline. However, since smoothing of the noisy AS is needed the resolution of overlapping peaks decreases. Peak detection in the second-order derivatives has some advantages due to the removal of a linear baseline and to the resolution enhancement [50].

Peak detection algorithms in Liquid-Chromatography-Mass Spectrometry (LC/MC) for protein identification and quantification were reviewed by [56]. The LC elution profile of these objects demonstrates structured asymmetrical shape which has some local maxima. Software packages for LC/MS analysis perform peak recognition using polynomial approximation and wavelet transform. Peak shapes are Gaussian and EMG. 2. Fitting data to peak models Model-based estimation of approximate peak parameters is widespread in spectroscopy and chromatography; two examples are given in Appendix. In the Natural Logarithm Derivative Method (LNDM) the position of peak maxima was estimated using linearization of symmetrical Gaussian model [57]. In the case of overlapping nonequal-width peaks a non-linear system of equation must be solved. Therefore, LNDM has limited application.

3. Peak sharpening using self-deconvolution Narrowing (sharpening or deconvolution) of asymmetrical lines was firstly studied by [16] using EMG model of chromatographic peaks (Table, ). Fourier transform of the EMG peak ( ) was divided by the Fourier transform of the kernel function

IJETCAS 15-109; Š 2015, IJETCAS All Rights Reserved

Page 4

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08

Since the peak and the kernel have identical mathematical expression, this method is named Fourier selfdeconvolution [26]. Deconvoluted peak in the spatial domain obtained by the inverse Fourier transform contains Gauss error function (erf) which cannot be expressed using elementary functions.

Resolution enhancement and asymmetry of deconvoluted peaks depend on the kernel parameters and .

Change of asymmetry may cause apparent shifts of the maximum peak position from the accurate value.

However, these shifts are small compared to the shifts of the strongly overlapping peaks which are significantly eliminated by deconvolution.

It is well known that deconvolution of a noisy AS produces wrong side lobes (oscillations) which intensity increases with increasing resolution enhancement [7]. Therefore, deconvolution should be performed using special regularization procedures based on the compromise between resolution enhancement and acceptable level of side lobes [7].

4. Peak sharpening using derivatives of AS Mathematical differentiation of AS is widely used for qualitative and quantitative analysis in chemistry [58].

First- and second-order derivatives provide valuable information about initial parameters of overlapping peaks due to the resolution enhancement and reducing baseline signal. However, differentiation increases the noise and creates wrong structure (satellites). Also differentiation causes a redistribution of the peak intensities since the derivative peak intensity is inversely proportional to the peak width in the power of the order of the derivative.

Numerous algorithms for detection of symmetrical overlapped peaks and for approximate evaluation of their parameters have been developed in derivative spectroscopy [58, 59]. The most popular applications of the second-order derivative were dedicated to the evaluation of the maximum peak positions (e. g., as a first stage of the empirical and semi blind algorithm in chromatography [38], using modified Lorentz model [31]). In the study [22] chromatogram was separated into independent parts. The second-order derivative of each part was divided into zones located between zero points. The upper and the lower boundaries of peak parameters were estimated in these zones. Finally, the intensity of each individual peak was corrected by the iterative fitting the experimental chromatogram to the model. Wrong (negative) peaks were removed after each iteration. d. Matrix methods Chemometrics approach to the separation of the overlapping peaks based on the matrix methods which was thoroughly discussed in reviews [9, 12]. In this section we briefly note some characteristics of these methods used to extract asymmetrical peaks from composite ASs. 1. Dual detection Multidimensional data obtained by simultaneous dual detection of the UV-visible absorption and chiral dispersion spectra was used for the extraction of the pure-component chromatograms [27]. Decomposition was based on the ratio method which requires a priory information about chromatograms of the pure mixture components. This approach is similar to the Signal Ratio Resolution Method based on the single detection [28, 29]. As was pointed earlier [61], the ratio method has limited applications due to the strong impact of the noise and significant sensitivity to the selection of analytical points.

2. Multivariate curve resolution (MCR) The MCR is based on data matrix decomposition into the concentration matrix ( ) and the matrix of the purecomponent of AS ( ) [13]:

The convergence of the MCR algorithms to the exact solution strongly depends on the accurate estimation of the number of pure components in AS [24] and on a priory information about a set of possible solution (initial estimation of matrix ) and different constrains [13].

Due to the numerous limitation of MCR [13] each row of matrix is represented as the sum of the same number of asymmetrical peaks (3). The peak shapes are identical, but the maximum positions and intensities differ [20]: The advantages of the developed method were demonstrated by decomposition of binary chromatograms using LOGA method [20].

In conclusion it is important to underline that all decomposition algorithms were tested using relative small number of experimentally obtained or synthesized objects. Therefore, published results do not allow a priory error evaluation in determining parameters of elementary peaks using particular decomposition method. We intend to solve this problem in future studies.

References [1].

[2].

[3].

K. Danzer, Analytical chemistry. Theoretical and metrological fundamentals. Springer, 2005.

B. K. Sharma, Spectroscopy. 19th Ed. India, Meerut-Delhy: Goel Publishing House, 2007.

R. P. W. Scott, Principles and practice of chromatography, LIBRARYFORSCIENCE, LLC, 2003.

IJETCAS 15-109; Š 2015, IJETCAS All Rights Reserved

Page 5

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08 [4].

[5].

[6].

[7].

[8].

[9].

[10].

[11].

[12].

[13].

[14].

[15].

[16].

[17].

[18].

[19].

[20].

[21].

[22].

[23].

[24].

[25].

[26].

[27].

[28].

[29].

[30].

[31].

[32].

[33].

[34].

[35].

[36].

[37].

[38].

[39].

[40].

[41].

A. Felinger, Data analysis and signal processing in chromatography, Elsevier, 1998.

V. B. Di Marco and G. G. Bombi, “Mathematical functions for the representation of chromatographic peaks”, J. Chromatogr. A, vol. 931, 2001, pp. 1–30.

J. Dubrovkin, "Mathematical analysis of asymmetrical spectral lines", Journal of Emerging Technologies in Computational and Applied Sciences, vol. 1-8, 2014, pp. 27-36.

J. Dubrovkin, "Critical analysis of spectral deconvolution methods "Journal of Emerging Technologies in Computational and Applied Sciences, vol. 1-10, 2014, pp. 40-49 .

G. R. Naik and D. K. Kumar, "An overview of independent component analysis and its applications", Informatica, vol. 35, 2011, pp. 63–81.

J. A. Westerhuis, E. P. P. A. Derks, H. C. J. Hoefsloot and A. K. Smilde, "Grey component analysis", J. Chemometrics, vol. 21, 2007, pp.474–485.

A. C. Duarte, S. Capelo, "Application of chemometrics in separation science", Journal of Liquid Chromatography & Related Technologies, v. 29, 2006, pp.1143–1176.

V. A. Likić, " Extraction of pure components from overlapped signals in gas chromatography-mass spectrometry (GCMS)",BioData Mining, vol. 2:6, 2009.

Y. Ni, Y. Gu and S. Kokot, "Interpreting analytical chemistry data: recent advances in curve resolution with the aid of chemometrics", Anal. Lett, vol. 45, 2012, pp.933-948.

C. Ruckebusch and L. Blanchet, "Multivariate curve resolution: A review of advanced and tailored applications and challenges", Anal. Chim. Acta, vol. 765, 2013, pp. 28–36.

T. Bolanča, Š.Ukić, M. Novak, and M. Rogošić, "Computer assisted method development in liquid chromatography",Croat.

Chem. Acta, vol. 87, 2014, pp. 111–122. Š. Hatric and J. Hrouzek, " The use of general exponential function for the deconvolution of fused chromatographic peaks ", Chem. Papers, vol. 48,1994, pp. 376-380.

A. Felinger, "Deconvolution of overlapping skewed peaks", Anal. Chem., vol. 66, 1994, pp. 3066-3072.

P. Nikitas and A. Pappa-Louisi, "Non-linear least-squares fitting with Microsoft Excel Solver and related routines in HPLC modelling retention. 1. Considerations of the problems and the method", Chromatographia, vol. 52, 2000, pp.477-486.

P. Nikitas , A. Pappa-Louisi, A. Papageorgiou, and A. Zitrou, "On the use of genetic algorithms for response surface modeling in high-performance liquid chromatography and their combination with the Microsoft Solver", J.Chromatogr. A, vol.942, 2001, pp.

93–105.

P. Nikitas , A. Pappa-Louisi, and A. Papageorgiou, " On the equations describing chromatographic peaks and the problem of the deconvolution of overlapped peaks", J.Chromatogr. A, vol. 912, 2001, pp. 13–29.

G. Vivo-Truyols, J.R. Torres-Lapasio , R.D. Caballero, and M.C. Garcıa-Alvarez-Coque, " Peak deconvolution in onedimensional chromatography using a two-way data approach", J. Chromatogr. A, vol. 958, 2002, pp. 35–49.

G. Vivó-Truyols, J.R. Torres-Lapasió and M. C. García-Alvarez-Coque," Net analyte signal as a deconvolution-oriented resolution criterion in the optimisation of chromatographic techniques", J. Chromatogr. A, vol. 991, 2003, pp. 47-59.

G. Vivó-Truyols , J. R. Torres-Lapasió ,A. M. van Nederkassel , Y. Vander Heyden, and D. L., Massart , "Automatic program for peak detection and deconvolution of multi-overlapped chromatographic signals. Part I: Peak detection", J .Chromatogr. A., vol.1096, 2005, pp.133-45.

G. Vivó-Truyols , J. R. Torres-Lapasió ,A. M. van Nederkassel , Y. Vander Heyden, and D. L., Massart , "Automatic program for peak detection and deconvolution of multi-overlapped chromatographic signals. Part II: Peak model and deconvolution algorithms", J. Chromatogr. A., vol.1096, 2005, pp.146-155.

S. Peters, H.-G. Janssena, and G.Vivó-Truyols, " A new method for the automated selection of the number ofcomponents for deconvolving overlapping chromatographic peaks", Anal. Chim. Acta, vol. 799, 2013, p. 29–35.

D. Nikolic, Z. Mijatovic, S. Djurovic, R. Kobilarov, and N. Konjevic,"Deconvolution of plasma broadened non-hydrogenic neutral atom lines", J. Quantitative Spectroscopy & Radiative Transfer vol. 70, 2001, pp.67-74.

J. Kauppinen and J. Partanen, Fourier Transforms in Spectroscopy. Germany: WILEY-VCH Verlag, 2001.

A. Gergely, P. Horváth, and B. Noszál, "Deconvolution of composite chromatographic peaks by simultaneous dual detections", J. Chromatogr. Science, vol. 38, 2000, pp.425-429.

Z. Grabaric, B. S. Grabaric, M. Esteban and E. Casassas," Signals ratio method for resolution enhancement in differential pulse polarography and related techniques", Anal. Chim. Acta, vol. 312, 1995, pp. 27-34.

S. V. Romanenko, N. S. Shehovtzev and D.M. Karachakov, "Enhancement signal ratio resolution method (SRRM) for decomposition of inverse volt-amperometric peaks", Bulletin of the Tomsk Polytechnic University, vol. 312, 2008, pp. 48-53 [Russian].

J. H. Christensen, J. Mortensena, A. B. Hansenb and O. Andersen, "Chromatographic preprocessing of GC–MS data for analysis of complex chemical mixtures", J. Chromatogr. A, vol. 1062, 2005, pp. 113–123.

J. R.Torres-Lapasio, J. J. Baeza-Baeza, M. C. Garcia-Alvarez and M. C. Coque, "A model for the description simulation and deconvolution of skewed chromatographic peaks", Anal Chem., vol. 68, 1997, pp. 3822-3831.

R.D. Caballero, M.C. Garcıa-Alvarez-Coque, and J.J. Baeza-Baeza, "Parabolic-Lorentzian modified Gaussian model for describing and deconvolving chromatographic peaks", J.Chromatogr. A, vol. 954, 2002, pp. 59–76.

J. J. Baeza-Baeza, C. Ortiz-Bolsico, M. C. García-Álvarez-Coque, “New approaches based on modified Gaussian modelsfor the prediction of chromatographic peaks” , Anal. Chim. Acta, vol. 758, 2013, pp. 36–44.

K. Lan and J.W. Jorgenson, "A hybrid of exponential and Gaussian functions as a simple model of asymmetric chromatographic peaks", J. Chromatogr. A, vol. 915, 2001, pp. 1-13.

T. Yu and H. Peng," Quantification and deconvolution of asymmetric LC-MS peaks using the bi-Gaussian mixture model and statistical model selection", BMS Bioinformatics, vol. 11:559, 2010.

"Peak Fitting in XPS", http://www.casaxps.com/help_manual/manual_updates/peak_fitting_in_xps.pdf, 2006.

S. S. Fong, P. Rearden, C. Kanchagar, C. Sassetti, J.Trevejo, and R. G. Brereton, "Automated peak detection and matching algorithm for gas chromatography–differential mobility spectrometry", Anal. Chem.,vol. 83, 2011, pp.1537–1546.

M. Mecozzi and M. Scarpiniti,"An empirical and semi blind algorithm for resolving overlapped peaks in chromatography: application to the analysis of environmental samples", APCBEE Procedia, vol. 5, 2013, pp. 145 – 151.

Nature-inspired Methods in Chemometrics: Genetic Algorithms and Artificial Neural Networks, R. Leardi (Ed.), Elsiever,2003.

X. Shao, Z. Chen and X. Lin, "An immune algorithm for resolution of multicomponent overlapping chromatograms",Fresenius J. Anal. Chem., vol.366, 2000, pp.10–16.

E. Grushka and I. Atamna, "The use of derivatives for establishing integration limits of chromatographic peaks", Chromatographia, vol 24, 1987, pp. 226-232.

IJETCAS 15-109; © 2015, IJETCAS All Rights Reserved

Page 6

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08 [42].

[43].

[44].

[45].

[46].

[47].

[48].

[49].

[50].

[51].

[52].

[53].

[54].

[55].

[56].

[57].

[58].

[59].

[60].

[61].

A. L. Stancik and E. B. Brauns, “A simple asymmetric lineshape for fitting infrared absorption spectra”, Vibrational Spectrosc., vol. 47, 2008, pp. 66–69.

S. Dobosh et al, “Detection of ions with the energy larger than 100 keV, which are produced due to the interaction of the 60fslaser pulse with clusters”, J. Experim. Theor. Phys., vol. 115, 1999, pp. 2051-2066.

E. A. Burstein and V. I. Emelyanenko, “Log-normal description of fluorescence spectra of organic fluorophores”, Photochem.Photobiol., vol. 64,1996, pp. 316–320.

E. A. Burstein, S. M. Abornev and Y. K. Reshetnyak, ʺDecomposition of protein tryptophan florescence spectra into log-normal components. I. Decomposition algorithmsʺ, Biophys J, vol. 81, 2001, pp. 1699-709.

A. Losev, “On a model line shape for asymmetric spectral peaks”, Appl. Spectrosc., vol. 48, 1994, pp. 1289-129 M. V. Klein, in Light Scattering in Solids I. Berlin, Germany: Springer-Verlag, 1983, pp. 169–172.

T.L. Pap and Zs. Papai, "Application of a new mathematical function for describing chromatographic peaks", J. Chromatography A, vol. 953, 2002, pp. 31–38. , Zs. Papai and T.L. Pap, "Analysis of peak asymmetry in chromatography ", J. Chromatography A, vol. 953, 2002, pp. 31–38.

L. Gurevich, L. A. Rusinov, and N. A. Syagaev, Automatic Chromatographic Analysis. Russia: Leningrad, Khimiya, 1980 [Russian].

J. M. Dubrovkin and V. G. Belikov, Derivative Spectroscopy. Theory, Technics, Application. Russia: Rostov University, 1988 [Russian].

T. O'Haver, "An introduction to signal processing with applications in chemical analysis", http://terpconnect.umd.edu/~toh/spectrum/IntroToSignalProcessing.pdf.

D. G. Luenberger and Y. Ye, Linear and Nonlinear Programming, 3rd Ed., Springer, 2008.

G. A. F. Seber and C. J. Wild, Nonlinear Regression, Hoboken, NJ: Wiley-Interscience, 2003.

R. Danielsson, D. Bylund and K. E. Markides, “Matched filtering with background suppression for improved quality of base peak chromatograms and mass spectra in liquid chromatography-mass spectrometry”, Anal. Chim. Acta, vol. 454, 2002, pp. 167– 184.

J. Zhang, E. Gonzalez, T. Hestilow, W. Haskins and Y. Huang, "Review of peak detection algorithms in liquid-chromatographymass spectrometry", Current Genomics, vol. 10, 2009, pp. 388-401.

Fernández-González and JM. Montejo-Bernardo, "Natural Logarithm Derivative Method: a novel and easy methodology for finding maximums in overlapping experimental peaks", Spectrochim. Acta Part A: Mol. Biomol. Spectrosc., vol. 74, 2009, pp.

714-718.

J. Dubrovkin,"Chemical analysis with the use of the mathematical differentiation", Zhurnal .Analiticheskoi. Khimii, vol. 43, 1988, pp. 965-979 [Russian].

G. Talsky. Derivative Spectrophotometry. Low and Higher Order.Germany, Weinheim: VCH Verlagsgesellschaft, 1994.

M. J. Fredriksson, P. Petersson, B.-O. Axelsson and D. Bylund, "An automatic peak finding method for LS-MS data using Gaussian second derivative filtering", J. Sep. Sci, vol. 32, 2009, pp. 3906-3918.

J. Dubrovkin, "Error propagation of quantitative analysis based on ratio spectra", International Journal of Emerging Technologies in Computational and Applied Sciences, vol. 1-9, pp. 2014, 11-20.

APPENDIX A. Bi-Gaussian fitting of asymmetrical peak [11] Estimation of the peak parameters from the noisy data is performed using the log-ratios of the truncated zeroand second order moments of the left and the right peak sides:

where The maximum peak position (

The log-scale is used for partial linearization of the estimation procedure. ) is the solution of the equation

The widths: (A4) (A5) Peak intensity ( ) is calculated by a weighting average in the log-scale:

where

and

are evaluated (Eq. 3) and observed values at discrete point

, respectively.

B. Fitting of PVMG peak [33] Logarithm of Eq. 3 is equal to

Substituting the left

and the right

half width into Eq. B1, we have

IJETCAS 15-109; © 2015, IJETCAS All Rights Reserved

Page 7

J. Dubrovkin, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 01-08

Since the denominator of Eq. B1 must be positive, then

and

Empirical solution of Eq. B2 with constrains (Eq. B3) is obtained using the half widths at 10% peak height.

For EGN model ( ) [34]:

where is the peak intensity relative to the maximum intensity, at % peak height.

IJETCAS 15-109; Š 2015, IJETCAS All Rights Reserved

and

are the left and the right half widths

Page 8

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Content Based Image Retrieval System with Feature Extraction and Recently Retrieved Image Library Seema H. Jadhav1, Dr.Sunita Singh2, Dr.Hari Singh3 Associate Professor, Department of Instru & control, N.D.M.V.P.S’s K.B. Thakare College of Engineering, Gangapur Road, Nashik, Maharashtra, India 2,3 Professor, Bundelkhand University, University Campus, Jhansi, Kanpur Road, Uttar Pradesh, India

1

Abstract: One of the most important processes in image processing is image retrieval. The main process of image retrieval is retrieval of image from the database based on the query. Basically, retrieval is done based on low level features such as shape, texture and colour. If a shape based query image is given, all images related to the given query image are retrieved.

Content based image retrieval system is a fast growing research area, where the visual content of a query image is used to search images from large scale image databases. In this proposed an effective system, both the semantically and visually relevant features are used to retrieve the related images. The challenge for the CBIR system is how to efficiently capture the features of the query image for retrieval. In traditional content based retrieval system, the visual content features of the whole query image are used for the retrieval purpose. But in the proposed system, the object wise features of query image are utilized for the effective retrieval. Moreover, an active Recently Retrieved Image Library (RRI Library) is used, which increases the accuracy in each retrieval. An RRI library uses an index system, which maintains the recently retrieved images, and during the retrieval process, the proposed system searches the pertinent images from both the database as well as the RRI library and hence the retrieval precision is gradually increased in each retrieval. The proposed CBIR method is evaluated by querying diverse images and the retrieval efficacy is analyzed by calculating the precision-recall values for the retrieval results.

Keywords: Content Based Image Retrieval, Low level features, K-mean algorithm, High Level Features, Image Segmentation, Recently Image Library, Precision-recall values

I. INTRODUCTION Image retrieval systems browse, search and retrieve images from a huge database of digital images [7]. Pictorial queries based retrieval of image data is emerging as an interesting and challenging problem with the advancement of the multimedia network technology and the growth of image data. A method used for retrieving similar images from an image database, called Content Based Image Retrieval (CBIR) [3] [4]. has emerged as a hot topic in technical research [1]. CBIR has diverse applications in internet, multimedia, medical image archives, crime prevention, entertainment, and digital libraries [13] and it is an important field in image processing [2].

Visual contents, commonly called as features are used by CBIR to search images from large scale image databases according to the requests of the user which is provided in the form of a query image [12]. It is essential for features of an image to have a sound relationship with the semantic meaning of the image. By comparing the features of the query image with the features of the images present in the database the CBIR system retrieves relevant images from the image database for s given query image [14] [5]. Based on the low level or high level features used for retrieval, the CBIR systems can be classified into different types [11].

CBIR systems that use low-level features for retrieval identifies the data base images that have visual similarity with the query image by comparing the low-level image features [8] like color, texture, shape and structure that are extracted from the images [9] [10]. The high-level description is an attributed graph attained by the structural representation of the image [16]. Compared to low level features, extraction of high level features is more difficult, even though they are more preferable for retrieval of images, particularly where human perception is more important [6]. Bridging the gap between low-level feature layout and high-level semantic concepts is the most challenging aspect of CBIR [15].

CBIR system extracts the images that are appropriate to the specified query image based on the image content.

The feature sets that limit the retrieval efficiency is been extracted by a majority of the CBIR systems existing in the literature. The proposed CBIR system using homogeneity feature extraction effectively retrieves the images relevant to the query image from the data base as well as from the RRI library when compared to the other CBIR system.

II.RELATED WORK Chia -Hung Wei et al. [25] derived a novel content-based trademark retrieval system with a feasible set of feature descriptors, capable of depicting global shapes and interior /local features of the trademarks. They have also proposed an effective two-component rotation, translation, scaling and stretching. For image retrieval stage, a two-component matching strategy was used in feature matching. With those strategies, the images were

IJETCAS 15-111; Š 2015, IJETCAS All Rights Reserved

Page 9

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16

compared with the query image with their local and global features taken into account separately, and therefore enabling the system to be insensitive to noise or small regional changes.

Sankara Rao et al. [27] have proposed a CBIR method which is based on automatically-derived image features, for the retrieval semantically-relevant images from an image database. Hierarchical and multilayer network which incorporates RBFN has been the unique feature utilized in their system. Their proposed procedure consists of two stages. They have obtained better favored image results by using hierarchical clustering at the initial stage to filter most of the images before employing the clustered images to the RBFN network.

Ying Liu et al. [22] provided a comprehensive survey of the recent technical achievements in high-level semantic-based image retrieval. Their research had covered different aspects, including low-level image feature extraction, similarity measurement, and deriving high-level semantic features. Some major categories of the state-of-the-art techniques in narrowing down the semantic gaps were object ontology to define high-level concepts, machine learning methods to associate low-level features with query concepts, relevance feedback to learn user’s intention generating semantic template to support high-level image retrieval, fusing the evidences from HTML text and the visual content of images for WWW image retrieval. In addition, such as image test bed and retrieval performance evaluation were also discussed. Finally, based on existing technology and the demand from real-world applications, a few promising future research directions were implemented.

Amit Jain et al. [23] proposed an algorithm for retrieving images with respect to a database consisting of engineering/computer-aided design (CAD) models. A linear approximation procedure captured the depth information using the idea of shape from shading has been used. Retrieval of objects has been done using a similarity measure that combines shape and the depth information. The main contribution of this paper is the idea of combining shape (contour) obtained from the contour tracing along with the 3D embedding, the depth information at each point on the contour.

Ricardo da S. Torresa et al. [24] proposed a Genetic Programming framework to the design of combined similarity functions. Those methods allow nonlinear combination of image similarities and were validated through several experiments, where the retrieved images were based on the shape of their objects. The proposed framework was validated for shape-based image retrieval, through several experiments involving two image databases, and many simple descriptors and fitness functions.

Reddy et al. [26] have proposed a method called semantic indexing using high level features. The capability of their approach in retrieving images of diverse shape, colors and size has been evident from their results.

Secondly images possessing largest area have been retrieved correctly. Their system also has certain drawbacks.

Images having small segment size are not satisfactorily processed by their approach.

Jalil Abbas et al. [28] have proposed content based image retrieval and mainly concentrated on Text Based image retrieval (TBIR). Comparison of their results have shown that content based has been visual whereas text based has been semantic. Compared to Content Based image retrieval (CBIR), the text based image retrieval has been faster.

III.CBIR SYSTEM WITH RECENCY-BASED RETRIEVED IMAGE LIBRARY An explosive growth of digital images has increased the need for an efficient content-based image retrieval system. The performance of the CBIR system usually depends upon the features adopted to represent the images in the database. In order to retrieve image based on the given query image, either low level or high level features of the database may be used. In order to extract and retrieve the image an effective mechanism is needed. Here, we use a recently retrieved image library for the retrieval of the system which increases the precision of the retrieval. The feature extraction and retrieval phase are the two steps we considered in our proposed work. This system also uses a recently retrieved image library for the retrieval of the system which increases the precision of the retrieval in the proposed system. The Image retrieval system consists of two steps namely feature extraction and retrieval phase.

A. Low Level Feature Extraction: We focus on considering four popular features, namely, shape, texture, color and homogeneity features. The shape, color, texture and homogeneity features are the low level features used in this CBIR for retrieval. In the feature extraction process, the low level features shape, color, the texture and homogeneity features are extracted for the query image from the database images and also from recently retrieved image library. The image features are either extracted from the whole image or from the regions. After the completion of the feature extraction, the query image features compared with the database image features and recently retrieved image library. Hence the images which have similar low level features are retrieved.

A.1 Shape feature extraction: Shape is an important visual feature and also it is one of the primitive features for image content explanation.

Shape feature grabs a position to convey a complex set of ideas in a quick way, especially the content of images [1] [2]. It is a concept of wide understanding however difficult for formal definition [3]. Despite shape is a high level feature in the perspective of humans, it is a description with low level attributes as per the mathematical definitions. In the feature extraction process, the low level features shape, color and the texture features are extracted from the query image and also from the database images. Hence the images which have similar low

IJETCAS 15-111; Š 2015, IJETCAS All Rights Reserved

Page 10

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16

level features are retrieved. The use of low-level visual features to retrieve relevant information from image and video databases has drawn much research attention in recent years. Color is perhaps the most dominant and distinguishing visual feature. Color histogram is the most widely used color descriptor in content based retrieval research. The techniques of shape description can be categorized in various ways. Usually they are classified as boundary-or region-based methods, scalar or space domain methods, and information preserving or nonpreserving techniques [7] [8]. Let X be the image database which contains images of dimension P  Q . For the filterization process, the image X is converted to gray scale X g from RGB color space. Let, X R

XG, XB

be the R, G, B weights of the image X then,

X g  0.2989  X R  0.5870  X G  0.1140  X B

(1)

The Craig’s formula has been applied for the conversion of RGB to gray scale image and then the mean filter is applied on the converted gray scale image to remove the noises. The mean filter smoothens the image data, thus the noise has been eliminated. Using the grey level values, this filter performs spatial filtering on each individual pixel in an image in a square or rectangular window surrounding each pixel. And then the filtered image has been clustered to identify the various regions in the image and this can be discovered by identifying groups of pixels that have similar gray levels, colors or local textures utilizing clustering in the image analysis. the k-means method has been shown to be effective in producing good clustering results. The k-means method aims to minimize the sum of squared distances between all points and the cluster centre. It is obvious in this description that the final clustering will depend on the initial cluster centres chosen and on the value of K. In this work, we use k means clustering to identify the various regions in the image. The denoised image is then clustered by means of k means clustering. The steps for the k means clustering process is given below 1. The input is the denoised image I g with m  n pixels and K  number of clusters 2. Set of K-clusters Here, the input is the denoised image and k number of clusters and the output includes set of k clusters. The steps included in k means clustering are as follows Step 1: Arbitrarily select K data items from the input as initial centroid.

Step 2: Repeat and Assign the remaining data items apart from the selected initial centroid to the cluster K , which has the closest centroid..Calculate the new centroid for each cluster until convergence occurs.After applying the k means clustering algorithm, the clustered regions of the denoised image I g has been identified.

A.2 Color based Region Segmentation: In the proposed technique the different objects in the image are segmented on the basis of the colors. In our proposed segmentation technique the number of different colors present in the image is reduced to 128. The LGB vector quantization algorithm is used in proposed technique to obtain the set of different colors which will represent image colors in lab space (with respect to mean square error).

For each pixel a x , y , we can calculate the local color contrast.  x y as follows.

 x y  Where

 xy the

 xy   xy

(2)

 xy

average of a color in the small neighborhood around

 xy and .

represents the norm of the

vector. The pixel a x , y is considered as an edge if its contrast exceeds a predefined value threshold  . In the next step, to distinguish between the different uniform regions, texture areas, and contour points, we use a sliding window to estimate the different characteristics of image such as mean  , and variance  , of edge density for each pixel.

Finally, the amount of averaging is performed in the textured areas and is chosen based on the edge density, so that amount of averaging is higher for fine textures and lower for coarse textures. Thus the perceived color at location (x, y),  ( x, y ) is as follows,

 ( x, y)  (  * g xy ) ( x, y)

(3)

Where * is the convolution operator and g xy is the Gaussian kernel which is defined as

g xa ya  k exp ( x 2  y 2 /  2 ); g xa ya 1

(4)

B. High level feature extraction: This CBIR is based on the query keyword which is a high level feature in human understandable form. Relevant images are retrieved by this CBIR utilizing the keyword retrieved. Each image is indexed with its semantic meaning so that they could be identified by the CBIR.

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 11

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16

C. Recency-based retrieved image library: The image library consists of images used for further purposes in CBIR. Here in recency-based retrieved image library (RRI) contains retrieved images. After, each retrieval the retrieved image is saved in the RRI library for further process. Hence this process works as like during retrieval for the second query image it search first in the recency based retrieved image library and so the process get faster and retrieval timing decreases and performance of the whole system increases after completion of each and every retrieval.

D. Retrieval Phase: The features are extracted for all the images in the database and stored in feature database. The query image is match up to the images in the database for image retrieval. After the extraction of images using shape, color, texture and homogeneity feature extraction, the extracted images are stored in the feature database. A semantic name is given to all the images, which is stored in the database, are also one of the high-level features of the image. After extraction process the feature set from the database images, it is necessary to compare all of these feature sets with the given query image’s feature set. The relevant images which will satisfy the low level feature of the query images is retrieved firstly and stored in low feature image library and then images which satisfy the high level feature are extracted and stored in high feature image library. The system recovers the relevant images which exists in both the low level library, high level library and also in Recency-based Retrieved Image Library (RRI Library) are retrieved. The feature set of the recovered relevant images are stored with syntactic name index in RRI library for future reference.

IV. EXPERIMENTAL RESULTS OF CBIR SYSTEM WITH RECENCY-BASED RETRIEVED IMAGE LIBRARY In our research the content based image retrieval technique is used for retrieving the images from the database for the query images. Here the large database consists of various types of images and the image retrieval process is carried out by using our proposed method for the query of rose and elephant images. The below experimental results shows the retrieval process for the query of rose and elephant images. In our presented method we uses shape feature extraction, color, texture, color intensity and homogeneity feature extraction for low level feature extraction. we can note the retrieved images for the query of input image and our presented techniques shows effective approach towards image retrieval system. The figure 1 shows the retrieval of query image ‘rose’ and also the figure 2, 3 shows the retrieval of query image ‘elephant’. This paper discuss about low level feature extraction, high level feature extraction and common features. Therefore the precision, recall and the comparison of precision-recall with the retrieved images gives the effective accuracy of our presented technique while comparing with the existing method. Therefore below sections briefly explains the CBIR with the results obtained at every process. By using homogeneity feature extraction process the accuracy of the precision and recall increases with the total number of image retrieved.

A. Experimental results - retrieval using Common Feature: The retrieval using common features includes the features of low level and high level features. This common feature of image retrieval gives the common features included in the low level features as well as the high level features. Thus the study area of experimental results using common features is as follows:

Figure1.Retrieved result for the input query image ‘rose’

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 12

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16

The retrieved result for the input query image ‘rose’ is shown in figure 1. Here we can see the rose image is loaded and processed. After the process gets completed low level feature based output and high level feature based output with common image retrieved process is shown above The retrieved result for the input query image ‘elephant’ is shown in figure 2. Here we can see the elephant image is loaded and processed. After the process gets completed low level feature based output and high level feature based output with common image retrieved as 20, 37 and 55 images The retrieved result for the input query image ‘elephant’ is shown in figure 3. Here we can see the elephant image is loaded and processed. After the process gets completed low level feature based output and high level feature based output and common image retrieved as 20, 37 and 50 images V. OBSERVATION AND DISCUSSION The performance of the CBIR system usually depends upon the features adopted to represent the images in the database. The presented CBIR technique will use both the human perception as well as machine level perception. Proposed system also uses a recently retrieved image library for the retrieval of the system which increases the precision of the retrieval in the proposed system. The proposed Image retrieval system consists of two steps namely feature extraction and retrieval phase.

A. Experimental Discussion over CBIR system with RRI Library: This result with observation and discussion proved to be effective that the system shows its best retrieval process for retrieving images all other techniques. Thus our thesis provides a brief discussion under the topic content based image retrieval process. Hence our discussion proves the content based image retrieval is the best and effective techniques towards image retrieval process.

Table 1: Shows the relevant retrieved images from the total retrieved images for the query input images rose, elephant

Input query image Rose Elephant Elephant

Total number of images retrieved 50 50 55

Relevant images retrieved 35 35 37

B. Performance Evaluation of Low Level and High Level Feature Extraction: In the feature extraction process, the low level features shape, color and the texture features are extracted from the query image and also from the database images. Hence the images which have similar low level features are retrieved. The use of low-level visual features to retrieve relevant information from image and video databases has drawn much research attention in recent years. CBIR is based on the query keyword which is a high level feature in human understandable form. The Relevant images are retrieved by this CBIR utilizing the keyword retrieved. Each image is indexed with its semantic meaning so that they could be identified by the CBIR.

Figure 2 Retrieved result for the input query image ‘elephant’

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 13

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16 Table 2: Retrieved images using low level and high level feature extraction

Input query image Rose Elephant Elephant

Low level feature extraction 6 2 6

High level feature extracted 37 37 37

Figure 3 Retrieved result for the input query image ‘elephant’ 60 50

Retrieved images

40 30

total images retrieved relevant retrieved images

20 10 0 rose

elephant

elephant

Query images Figure 4 Relevant retrieved images from the total retrieved images for the query input images rose, elephant

5.3Performance Analysis: The performance analysis section briefly discuss about the CBIR system with RRI library and without RRI library. Hence our system with RRI library proved to be best of image retrieval system.

Table 3: Precision Recall values (presented work) Number of retrieval First Second Third

Precision 0.3333 0.5945 0.6667

Recall 0.32 0.88 0.88

The above figure shows the precision recall values for the number of retrieval in our presented work. Thus it shows three iterations with the precision and recall. After completion of iterations the precision and recall values increases and therefore it proves to be best. Also the presented method is effective and it shows best retrieval for the precision and recall values. Hence the precision and recall values shown in our presented method is best

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 14

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16

while comparing with the existing method. Here the precision values are 0.0416, 0.5135, and 0.6061 for the presented system.

PERFORMANCE

1

Evaluation matrix

0.9 0.8 0.7 0.6 0.5

Precision

0.4

Recall

0.3 0.2 0.1 0 1

2

3

Number of retrieval Figure 5: Precision Recall graph (presented method)

VI. Conclusion In this paper a content based image retrieval system was proposed for effective retrieval of the relevant images from the image database. The system is intended to use both the high and low level feature of the images for retrieval purpose which decrease the semantic gap between low level and high level features. The system was implemented and experimented with varying query images. The analytical results confirmed that the proposed technique showed better performance than the classical CBIR system. It also proved that the performance of the proposed system with RRI library was improving at remarkable rate in each successive retrieval. From all the afore described analytical results, it can be assertively concluded that the proposed system shows good performance than the conventional hierarchical system.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

Jianhua Wu, Zhaorong Wei and Youli Chang, "Color and Texture Feature for Content Based Image Retrieval", International Journal of Digital Content Technology and its Applications, Vol.4, No.3, pp: 43-49, June 2010.

Ch. Srinivasa Rao , S. Srinivas Kumar and B. Chandra Mohan," Content Based Image Retrieval using Exact Legendre Moments and Support Vector Machine", The International Journal of Multimedia and its Applications (IJMA), Vol.2, No.2,pp: 69-79, May 2010.

Hiremath and Jagadeesh Pujari, "Content Based Image Retrieval based on Color, Texture and Shape features using Image and its complement", International Journal of Computer Science and Security, Vol. 1, Issue. 4, pp: 25-35, 2007.

Suresh Pabboju and A. Venu Gopal Reddy, "A Novel Approach for Content-Based Image Indexing and Retrieval System using Global and Region Features", International Journal of Computer Science and Network Security, Vol.9 No.2, pp: 119-130, Feb 2009.

Ch. Srinivasa rao , S. Srinivas kumar and B.N. Chatterji, "Content Based Image Retrieval using Contourlet Transform", ICGST-GVIP Journal, Volume 7, Issue 3, pp:9-15, November 2007.

Awais Adnan, Muhammad Nawaz, Sajid Anwar, Tamleek Ali and Muhammad Ali, "Object Identification with Color, Texture, and Object-Correlation in CBIR System", World Academy of Science, Engineering and Technology, Vol.64, pp.117-122, 2010.

Murthy, Vamsidhar, Swarup Kumar and Sankara Rao, "Content Based Image Retrieval using Hierarchical and K-Means Clustering Techniques", International Journal of Engineering Science and Technology, Vol.2, No.3, pp.209-212, 2010.

G. Sasikala, R. Kowsalya, M. Punithavalli, “A Comparative Study of Dimension Reduction Techniques for Content-Based Image Retrieval”, The International journal of Multimedia & Its Applications (IJMA) Vol.2, No.3, pp: 40-47, August 2010.

Rajshree S. Dubey, Rajnish Choubey and Joy Bhattacharjee, "Multi Feature Content Based Image Retrieval", International Journal on Computer Science and Engineering, Vol. 2, No. 6, pp: 2145-2149, 2010.

Christoper C. Yang, "Content Based Image Retrieval: a Comparison between Query by Example and Image Brousing Map Approaches", Journal of Information Science, Vol.30, No.3, pp: 254-267, 2004.

Hui Hui Wang, Dzulkifli Mohamad, N.A Ismail, "Image Retrieval: Techniques, Challenge, and Trend", World Academy of Science, Engineering and Technology Vol.60, pp: 716-718, 2009.

Nandagopalan, Adiga and Deepak, "A Universal Model for Content-Based Image Retrieval", World Academy of Science, Engineering and Technology, Vol.46, 2008.

Srinivasa Rao and Srinivas Kumar, “Content Based Image Retrieval using Contourlet Sub band Decomposition,” In Proceedings of IEEE International Conference, SPIT Colloquium, Mumbai, pp.140-145, February 2008.

Thomas M. Deserno, Sameer Antani and Rodney Long, "Ontology of Gaps in Content-Based Image Retrieval", Journal of Digital Imaging, Vol. 22, No. 2, 2009 Hiremath and Jagadeesh Pujari, "Content Based Image Retrieval using Color Boosted Salient Points and Shape features of an Image", International Journal of Image Processing, Vol.2, No.1, pp. 10-17, 2008.

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 15

Seema H. Jadhav et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 09-16 [16] Mark Ewald, "Content-Based Image Indexing and Retrieval in an Image Database for Technical Domains", Transactions on Machine Learning and Data Mining, Vol. 2, No 1, pp: 3-22, 2009.

[17] Ganapathi Reddy, Babu and Somasekhar, "Image Retrieval by Semantic Indexing", Journal of Theoretical and Applied Information Technology, Vol.5, No.6, pp.745-750, 2005.

[18] Arun Kulkarni, Harikrisha Gunturu and Srikanth Datla, "Association-Based Image Retrieval", WSEAS Transactions on Signal Processing, Vol. 4, No.4, pp.183-189, 2008.

[19] Preeti Aggarwal, Sardana and Gagandeep Jindal, "Content Based Medical Image Retrieval: Theory, Gaps and Future Directions", ICGST-GVIP Journal, Vol.9, No.2, pp. 27-37, 2009.

[20] Wichian Premchaiswadi and Anucha Tungkatsathan, "On-line Content-Based Image Retrieval System using Joint Querying and Relevance Feedback Scheme", WSEAS Transactions on Computers, Vol.9, No.5, pp. 465-474, 2010.

[21] Hui Hui Wang, Dzulkifli Mohamad and Ismail, "Semantic Gap in CBIR: Automatic Objects Spatial Relationships Semantic Extraction and Representation", International Journal of Image Processing (IJIP), Vol.4, No.3, 192-204, 2010.

[22] Ying Liu, Dengsheng Zhang, Guojun Lu and Wei-Ying Ma, “A Survey of Content-Based Image Retrieval with High-level Semantics,” Journal of Pattern Recognition, Vol. 40, No. 1, pp. 262-282, Jan 2007.

[23] Amit Jain, Ramanathan Muthuganapathy and Karthik Ramani, “Content-Based Image Retrieval Using Shape and Depth from an Engineering Database,” In Proc. of the Third International Conference on Advances in Visual Computing, Vol.2, pp. 255-264, 2007.

[24] Ricardo da S. Torresa, Alexander X. Falcaoa, Marcos A. Gonçalvesb, Joao P. Papaa, Baoping Zhang, Weiguo Fanc and Edward A. Foxc, “A Genetic programming framework for content-based Image retrieval,” Journal of Pattern Recognition, Vol.42, No.2, pp. 283292, Feb. 2009.

[25] Chia-Hung Wei, Yue Li, Wing Yin Chau and Chang-Tsun Li, “Trademark Image Retrieval Using Synthetic Features for Describing Global Shape and Interior Structure,” Journal of Pattern Recognition, Vol. 42, No.3, pp. 386-394, Mar 2009.

[26] Ganapathi Reddy, Babu and Somasekhar, "Image Retrieval by Semantic Indexing", Journal of Theoretical and Applied Information Technology, Vol.5, No.6, pp.745-750, 2005.

[27] Sankara Rao, Vamsidhar, Samuel Vara Prasad Raju, Ravikanth SAtapati and Varma,"An Approach For Cbir System Through Multi Layer Neural Network", International Journal of Engineering Science and Technology, Vol.2, No.4, pp.559-563, 2010.

[28] Hui Hui Wang, Dzulkifli Mohamad and Ismail, "Semantic Gap in CBIR: Automatic Objects Spatial Relationships Semantic Extraction and Representation", International Journal Of Image Processing, Vol.4, No.3, pp.192-204, 2010.

IJETCAS 15-111; © 2015, IJETCAS All Rights Reserved

Page 16

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A Dual Security Approach for Medical Images using Encryption and Watermarking Optimized by Differential Evolution Algorithm Mr. CH.Venu Gopal Reddy1, Dr. Siddaiah.P2 Professor, Department of ECE, Priyadarshini College of Engineering and Technology, Nellore, A.P, INDIA1 Professor and Dean, Department of ECE, Principal, University College of Engineering and Technology, Acharya Nagarjuna University, Guntur, A.P, INDIA2 Abstract: Cryptography and watermarking combination appears as a good promising tool in regard to medical image security and management. In this paper a dual security approach of watermarking and encryption is proposed along with a Multi-objective function for medical image watermarking to ensure that the watermark maintains its structural integrity along with robustness and imperceptibility. Differential Evolution (DE) optimization is employed to optimize the objective function to choose a correct type of wavelet and scaling factor. The water marking is proposed to be implemented using both Discrete Wavelet Transforms (DWT) and Singular Value Decomposition (SVD) technique. The encryption is done using RSA and AES encryption algorithms. A Graphical User Interface (GUI) which enables the user to have ease of operation in loading the image, watermark it, encrypt it and also retrieve the original image whenever necessary is also designed and presented in this paper. The robustness and the integrity of the watermark is tested by measuring different performance parameters and subjecting it to various attacks. The performance of the optimization is compared with the optimsation results of Genetic Algorithm (GA).

Keywords: Medical Image, DWT, SVD, DE, GUI, RSA, AES.

I. Introduction Advent of multimedia combined with information and communication technology boost the potential of medical information handling and sharing with applications ranging from telediagnosis to telesurgery and cooperative and working session. Medical information protection derives from strict ethics and legislatives rules.

Regulations like USA's HIPAA and Europe's EC 95/46 Directive are expressions of such a constraint. Focusing on medical information records, which for a patient are a complex set of clinical examinations, diagnosis annotations and other findings and images centered in its EPR, we recall the three mandatory security characteristics; confidentiality, availability and reliability. In Medical Information Systems (MIS), these characteristics are maintained through five security services [1]: integrity, availability, authentication, confidentiality, and non-repudiation. If availability, integrity and confidentiality services have similar definition in respect with the corresponding security component, the authentication service is “designed to establish the validity of a transmission, message, or originator, or a means of verifying an individual's authorizations to receive specific categories of information” [1]. Non-repudiation service manages proofs of delivery and of the message sender's identity. At the interface between the information and the MIS security services, watermarking can improve information protection from the information side.

A watermarking method is usually designed depending on an application framework striking a compromise between different requirements: capacity (amount of information that can be embedded), robustness (a fragile watermark will not survive any image processing), privacy (secret knowledge for watermark content access usually a secret key) and imperceptibility. We can say that the higher the strength of the watermark signal, the more it is robust and/or of higher capacity albeit perceptibility is compromised. Consequently, if it is envisioned to process the image with an information loss, a robust watermark is desirable to authenticate the image origins, while at the same time the watermark should not interfere with the image content interpretation. However each property has its own limitation and conflict with each other. It will be a challenging task to design a good algorithm by coupling both the concept of reversibility and robustness with proper optimization. It is well known that the integrity and confidentiality of medical folders are a critical issue for ethical as well for legal reasons. Classical encryption technology is an important tool that can be used to protect data transmitted over computer networks but it does not solve all digital data protection problems. At the receiver’s side, decrypted content may be subjected to unauthorized use or manipulation.

In transform domain watermarking can be performed using DCT (Discrete Cosine Transform) [2] or IWT (Integer Wavelet Transform) [3].Different approaches have been proposed in order to improve the security of medical image transmission using watermarking, which gives one level security. A Tamper Assessment Factor (TAF) of the watermarked image with the physician’s signature and patient diagnosis information embedded into wavelet transform coefficients of the medical images is proposed in [4]. Similarly a novel blind

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 17

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

watermarking method with secret key is proposed by embedding Electrocardiograph (ECG) signals in medical images combined with the EZW-based wavelet coder (5). A blind watermarking scheme using the non-tensor product wavelet filter banks are used for copyright protection is presented in [6]. An efficient watermarking method based on the significant difference of wavelet coefficient quantization is proposed in [7]. A multiple, fragile image authentication scheme is proposed for DICOM images using discrete wavelet transform in [8] in this work multiple watermarks are embedded into wavelet domains, the multiple watermarks serve as reference watermarks. A novel watermarking algorithm based on singular value decomposition (SVD) is proposed in [9].

Both of the D and U components of SVD are explored for embedding the watermark in [10].

To enforce integrity and authenticity several works have been implemented that provides two level security for transmission of medical images. In joint encryption/watermarking [11] method, watermarking and encryption step processes are merged. Joint watermarking/encryption system is slower than simply encrypting the image but it provides reliability control functionalities. Watermarking is done by Quantization Index modulation.

(QIM) method and AES (Advanced Encryption standard) and RC4 (Rivest cipher 4) algorithms do encryption.

A Digital envelope (DE) method to assure data integrity and security that outlines the systematic evaluation, development, and deployment of the DE method for PACS environment is proposed in [12]. A new cryptographic means to improve the trustworthiness of medical images is implemented [13]. A comparative study of AES and RC4 algorithm is done in [14] in the case of AES algorithm, as the key size increases the encryption and decryption time increases, whereas for RC4 it remains almost constant and it is less than AES. Similarly, several methodologies have been proposed for medical image security [15].These methods can detect, whether the medical images are tampered or modified but cannot protect the images from tampering. In this work, the digital watermarking is done by using special symmetric matrices to construct the new nontensor product wavelet filter banks [16] which can capture singularities in all directions. Here, natural image is considered as original image and medical image is taken as watermark to avoid the attacker’s attention toward the medical information.

The optimization of watermark through evolutionary approaches has also been researched extensively. A new method for adaptive watermark strength optimization in Discrete Cosine Transform (DCT) domain in which watermark strength is intelligently selected through Genetic Algorithm (GA) is proposed in [17].A novel hybrid PSO, namely (HPSO) to improve the performance of fragile watermarking based DCT which results in enhancing both the quality of the watermarked image and the extracted watermark is implemented in [18]. A novel optimal watermarking scheme based on singular value decomposition (SVD) using differential evolution algorithm (DE) is explained in [19]. Differential evolution (DE) algorithm to balance the tradeoff between robustness and imperceptibility by exploring multiple scaling factors in image watermarking is proposed in [20].

In this work we have implemented a dual security approach for maintaining the data integrity of the medical images. Watermarking and encryption of watermarked image is proposed. In order to preempt any attack from attacker the medical image is considered as water mark and is embedded in to a natural image. A multiobjective optimization approach is proposed to maintain the fidelity of the watermark (medical image) as it contains valuable diagnostic information. This multiobjective approach ensures that there is an optimum tradeoff between robustness, imperceptibility and structural integrity of the watermark. Maintaining the structural integrity of the watermark is necessitated by the fact that most of the diagnostic approaches in medical image consider the morphological factors of the image to divulge precious information about the prognosis of a particular clinical condition. Different performance parameters like Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Normalized Coefficient (NC) and Structural Similarity Index (SSIM) is used to frame an objective function. This objective function is optimized using Differential Evolution (DE) to choose a particular wavelet in selected wavelet family and scaling factor of the Singular Value Decomposition (SVD). The medical image security is further enhanced by encrypting the watermarked image using Ron Rivest, Adi Shamir, and Leonard Adleman (RSA) Algorithm and Advanced Encryption Standard (AES) algorithms. Correlation Value (CV) between the watermarked image and the encrypted image is used to measure the efficacy of watermark.

The watermarked image is tested for different types of attacks like sharpening, smoothening, rotation , cropping and different types of noises which include speckle noise, salt and pepper noise, Gaussian noise and Poisson noise. To enable ease of use and seamless integration of the user a Graphical User Interface (GUI) is designed to automate the process of embedding, encrypting, decrypting and extracting. The tool helps user in analyzing and testing different scenarios and choose the best possible one for a watermarking a given medical image. The performance measures are compared and contrasted with that of the performance measures as achieved by Genetic Algorithm (GA).

II. Methodologies This work aims at exploiting the features of Discrete Wavelet Transforms (DWT) and Singular Value Decomposition (SVD) to provide a robust and imperceptible watermark.Similary RSA and AES algorithms are used for encrypting the watermarked images to provide an extra layer of security. This section dwells on these concepts and methods used in this research work.

IJETCAS 15-117; Š 2015, IJETCAS All Rights Reserved

Page 18

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

A. Discrete Wavelet Transforms The first recorded mention of what we now call a “wavelet” seems to be in 1909, in a thesis by Alfred Haar. The concept of wavelets in its present theoretical form was first proposed by Jean Morlet and the team at the Marseille Theoretical Physics Center working under Alex Grossmann in France. The methods of wavelet analysis have been developed mainly by Y. Meyer and his colleagues, who have ensured the methods’ dissemination. The main algorithm dates back to the work of Stephane Mallat in 1988 [21].In numerical analysis and functional analysis, a discrete wavelet transform (DWT) is any wavelet transform for which the wavelets are discretely sampled. As with other wavelet transforms, a key advantage it has over Fourier transforms is temporal resolution: it captures both frequency and location information (location in time).Thus discrete wavelet transform (DWT) is a linear transformation that operates on a data vector whose length is an integer power of two, transforming it into a numerically different vector of the same length. It is a tool that separates data into different frequency components, and then studies each component with resolution matched to its scale. DWT is computed with a cascade of filters followed by a factor 2 subsampling.

Figure 1: Discrete Wavelet Transform tree for two-dimensional image.

H and L denote high and low-pass filters respectively followed by subsampling. Outputs of these filters are given by equations (1) and (2)

a j 1  p  



 l n  2 p a n

n 

d j 1  p  

(1)

j



 h n  2 p a n

n 

(2)

j

Elements aj are used for next step (scale) of the transform and elements dj, called wavelet coefficients, determine output of the transform. l[n] and h[n] are coefficients of low and high-pas filters respectively One can assume that on scale j+1 there is only half from number of a and d elements on scale j. This causes that DWT can be done until only two aj elements remain in the analyzed signal. These elements are called scaling function coefficients.The types of wavelets used in this work are described here. Haar wavelet is discontinuous, and resembles a step function. It represents the same wavelet as Daubechies ‘db1’ Ingrid Daubechies, invented what are called compactly supported orthonormal wavelets —The names of the Daubechies family wavelets are written dbN, where N is the order, and db the “surname” of the wavelet. The db1 wavelet, as mentioned above, is the same as Haar wavelet. Biorthogonal family of wavelets exhibits the property of linear phase, which is needed for signal and image reconstruction. By using two wavelets, one for decomposition (on the left side) and the other for reconstruction (on the right side) instead of the same single one, interesting properties are derived.

The Symlets are nearly symmetrical wavelets proposed by Daubechies as modifications to the db family. The properties of the two wavelet families are similar. The Wavelets function psi of different wavelet families used in this work are represented in the below Figure (2).

‘Haar’

Daubechies (‘db5’)

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Symlets (sym3)

Page 19

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

Biorthogonal (‘bior3.7’) Figure 2: psi of different wavelet families used in this research work The main feature of DWT that makes it attractive for image processing applications is multiscale representation of function. By using the wavelets, given function can be analyzed at various levels of resolution. The DWT is also invertible and can be orthogonal.DWT involves decomposition of image into frequency channel of constant bandwidth. This causes the similarity of available decomposition at every level. DWT is implemented as multistage transformation. Level wise decomposition is done in multistage transformation. At level 1: Image is decomposed into four sub bands: LL, LH, HL, and HH where LL denotes the coarse level coefficient which is the low frequency part of the image. LH, HL, and HH denote the finest scale wavelet coefficient. The LL sub band can be decomposed further to obtain higher level of decomposition. This decomposition can continues until the desired level of decomposition is achieved for the application. The watermark can also be embedded in the remaining three sub bands to maintain the quality of image as the LL sub band is more sensitive to human eye.

B. Singular Value Decomposition (SVD) Among the methods to write a matrix as a product of matrices, Singular Value Decomposition (SVD) is a very useful method. Singular Value Decomposition (SVD) is said to be a significant topic in linear algebra by many renowned mathematicians. SVD has many practical and theoretical values; Special feature of SVD is that it can be performed on any real (m, n) matrix. Let’s say we have a matrix A with m rows and n columns, with rank r and r ≤ n ≤ m. Then the A can be factorized into three matrices: Since an image can be represented as a matrix of positive scalar values SVD for any image say A of size × is a factorization of the form given by.

U  u1 , u2

V  v1 , v2

A  USV T , . . . ur , ur 1 , . . . , um 

, . . . vr

(3) (4)

, vr 1 , . . . , vn 

(5)

Where U and V are orthogonal matrices in which columns of U are left singular vectors and columns of V are right singular vectors of image A. S is a diagonal matrix of singular values in decreasing order.

0 . . . 0  1 0 . . . 0 0  . . . 0 0 . . . 0  2   . . . . . . .    . . . . . .   .  . . . . . . .  (6)   0 0 . . . r 0 . . . 0  S 0 0 . . . 0  r 1 . . . 0    . . . . . .   .  . . . . . . .     . . . . . . .    0 . . . 0 0 . . . n  0  0 0 . . . 0 0 . . . 0  The basic idea behind SVD technique of watermarking is to find SVD of image and the altering the singular value to embed the watermark. In Digital watermarking schemes, SVD is used due to its main properties namely a) A small agitation added in the image, does not cause large variation in its singular values.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 20

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

b) The singular value represents intrinsic algebraic image properties.

C. Encryption Algorithms Ron Rivest, Adi Shamir, and Leonard Adleman (RSA) Algorithm. RSA is an asymmetric key encryption algorithm [22]. Over 1000 bits long numbers are used. Therefore, it can avoid attacks like brute force, man-inmiddle, and so on. RSA algorithm (Zhou et al., 2011) involves the following steps (a) Key (private, public) generation. (b) Encryption is performed using receiver’s public key c) At the receiver’s side decryption is performed using the receiver’s private key [22].Advanced Encryption Standard (AES) was published by NIST (National Institute of Standards and Technology) in 2001[23]. It has 128,192, or 256 bits variable key length.

AES encryption is fast and flexible in block ciphers and can be implemented on various platforms. AES (specifies a cryptographic algorithm that can be used to protect electronic data. AES algorithm is a symmetric block cipher, which can encrypt and decrypt the information. In this work 8 rounds and 256 bit key lengths are used. AES Encryption includes the following steps.1. Key Expansion, 2. Initial Round, 3. Nine Rounds, 4. Final Round. Initial round has only added round key operation. Each round has the following steps, a. Substitute Bytes, b. Shift Rows. Mix columns. Add Round Key. In the final round steps a, b, and d are performed, excluding step: c. AES Decryption part a 10 set of reverse rounds are performed to transform encrypted image into the watermarked images using the same encryption key [23].

III. Optimization using Differential Evolution Evolutionary methods for solving optimization problems have become a very popular research topic in recent years. There are three main processes in all evolutionary algorithms. The first process is the initialization process where the initial population of individuals is randomly generated according to some solution representation. Each individual represents a solution, directly or indirectly. If an indirect representation is used, each individual must first be decoded into a solution. Each solution in the population is then evaluated for fitness value in the second process. The fitness values can be used to calculate the average population fitness or to rank the individual solution within the population for the purpose of selection. The third process is the generation of a new population by perturbation of solutions in the existing population.

DE was proposed by Storn and Price (1995) for global optimization over continuous search space. Its theoretical framework is simple and requires a relatively few control variables but performs well in convergence. In DE algorithm, a solution is represented by a D-dimensional vector [24]. DE starts with a randomly generated initial population of size N of D-dimensional vectors. In DE, the values in the D-dimensional space are commonly represented as real numbers. Again, the concept of solution representation is applied in DE in the same way as it is applied in GA The key difference of DE from GA is in a new mechanism for generating new solutions. DE generates a new solution by combining several solutions with the candidate solution. The population of solutions in DE evolves through repeated cycles of three main DE operators: mutation, crossover, and selection.

However, the operators are not all exactly the same as those with the same names in GA. The key process in DE is the generation of trial vector. Consider a candidate or target vector in a population of size N of D-dimensional vectors. The generation of a trial vector is accomplished by the mutation and crossover operations and can be summarized as follows. 1) Create a mutant vector by mutation of three randomly selected vectors. 2) Create trial vector by the crossover of mutant vector and target vector.

Figure 3: The flow chart of Differential Evolution

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 21

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

As shown in Figure (3), the first process is the generation of a population of new solutions called vectors. Each vector in the population is evaluated for fitness value. Each vector takes turns as a candidate or target vector, and for each target vector, a trial vector is formed. The selection process simply chooses between the target vector and trial vector, i.e., the winning vector between the trial vector and the target vector survives into the next round while the losing vector is discarded. Several observations are made here. First, since a new solution would be selected only if it has better fitness, the average fitness of the population would be equal or better from iteration to iteration. Any improvement in the solution is immediately available to be randomly selected to form a mutant vector for the next target vector. This is different from GA and PSO where an improvement would take effect only after all the solutions has completed the iteration.

In contrast with GA where parent solutions are selected based on fitness, every solution in DE takes turns to be a target vector (one of the parents), and thus all vectors play a role as one of the parents with certainty. The second parent is the mutant vector which is formed from at least three different vectors. In other words, the trial vector is formed from at least four different vectors and would replace the target vector only if this new vector is better than the target vector; otherwise, it would be abandoned. This replacement takes place immediately without having to wait for the whole population to complete the iteration. This improved vector would then immediately be available for random selection of vectors to form the next mutant vector. In this work Differential Evolution is coded using Matlab. The Parameters used and the settings are mentioned as mentioned below: DEParamsDefault.CR = 0.7; DEParamsDefault.F = 0.8; DEParamsDefault.NP = 30; DEParamsDefault.strategy = 1; DEParamsDefault.minvalstddev = -1; DEParamsDefault.minparamstddev = -1; DEParamsDefault.nofevaliter = 10; DEParamsDefault.nochangeiter = 10; DEParamsDefault.maxiter = inf; DEParamsDefault.maxtime = inf; DEParamsDefault.refreshiter = 10; DEParamsDefault.refreshtime = 60; seconds IV. Problem Formulation for Multi -Objective Optimization Multi-objective optimization is an appropriate tool for handling different incommensurable objectives with conflicting/ supporting relations or not having any mathematical relation with each other. In this work the multiobjective optimization problem is transformed into a scalar optimization problem with different performance measures represented in it. This kind of scenario is typical of medical images in which it is of foremost importance maintain and preserve the diagnostic information in the medial image. Unlike regular watermarking scheme where in the original image is of importance to the user, in this proposed scheme the watermark (medical image) is of much value to the user. Different performance parameters like Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Normalized Coefficient (NC) and Structural Similarity Index (SSIM) is used to frame this multi- objective function Any watermarking scheme should provide robustness, imperceptibility and also should be capable of maintaining the structural integrity of the watermark (medical image). The watermark embedding parameters plays a very crucial role in defining these parameters. In this work the type of wavelet in a particular wavelet family of Discrete Wavelet Transform (DWT) and the scaling factor used in Singular Value Decomposition (SVD) are using the multi-objective optimization function. The fitness function used for this multi-objective optimization is Min {f = (100-PSNR) + (1-NC) + (1-SSIM) + MSE} (7) The Peak Signal to Noise Ratio (PSNR) is used to find the deviation of watermarked and attacked image from the original image. Equation (8) represents the PSNR. In this equation mean squared error (MSE) for two M * N monochrome images f and z and it is given by Equation (9). MaxBits gives the maximum possible pixel value (255) of the image.

PSNR  10 X log10

MSE 

1 MxN

MaxBits 2 MSE

(8)

M 1 N 1

 (( f ( x, y)  z( x, y))

2

(9)

x 0 y 0

Normalized Coefficient (NC) gives a measure of the robustness of watermarking. After extracting the watermark, the normalized correlation coefficient (NC) is computed between the original watermark and the extracted watermark using Equation (10). This is used to judge the existence of the watermark and to measure the correctness of the extracted watermark.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 22

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

 w(i,j)w (i,j) j

NC=

,

(10)

i

j

 w(i,j)  w (i,j) j

2

,

2

i

i

Where, w and w’ represent the original and extracted watermark, respectively.

Structural Similarity Index (SSIM) index is a method for measuring the similarity between embedded and extracted watermark images. The SSIM is measured between two windows X and Y of common size N*N on image using Equation (11).

SSIM ( x, y ) 

(2 x  y  c1 )(2 xy  c2 )

(11)

(  x2   y2  c1 )( x2   y2  c2 )

V. Proposed Algorithm The water marking is proposed to be implemented using a hybrid approach which encompasses Discrete Wavelet Transforms (DWT) and Singular Value Decomposition (SVD) techniques.

The resultant of multi-objective optimization in form type of wavelet in a particular wavelet family of Discrete Wavelet Transform (DWT) and the scaling factor used in Singular Value Decomposition (SVD) is used in the process of embedding and extracting the watermark. In this algorithm, Medical image is taken as the watermark and it is embedded in each block of the Natural image (cover image) by altering the wavelet coefficients of selected DWT sub bands. The steps involved in this process are described below. a) Watermark Embedding and Encryption.

Step 1: Obtain the medical image to be embedded and the input natural Step2: Perform DWT by using the optimized selection of wavelet obtained through optimization approach on the natural image to decompose it into four non-overlapping subbands: LL, HL, LH, and HH. Step 3: Apply SVD to sub band i.e., = where = Step 4: Apply SVD to the watermark i.e., = where = Step 5: Modify the singular value of by embedding singular value of such that , = + × ,Where is modified singular matrix of and denotes the scaling factor, is used to control the strength of watermark signal the value of which is optimized through Differential Evolution (DE) using the multi objective function.

Step 6: Then apply SVD to this modified singular matrix i.e., = _ _ _ and obtain the modified DWT coefficients, i.e., = × _ × Step 7: Obtain the watermarked image by applying inverse DWT using one modified and other non modified DWT coefficients.

Step 8: Then encrypt the watermarked image with RSA or AES algorithms in the time domain. b) Decryption and Watermark Extraction Step 1: Decrypt the encrypted image to obtain the watermarked image.

Step 2: Apply the chosen DWT to decompose the watermarked image in to four sub bands (i.e., , , , ).

Step 3: Apply SVD to sub band i.e., = , Where = Compute ∗= ( − )/ , where ∗ singular matrix of extracted watermark Step 4: Apply SVD to ∗ i.e., ∗= _ ∗ _ ∗ _ ∗ Step 5: Now Compute extracted watermark ∗ i.e., ∗= × _ ∗× VI.

The Graphical User Interface (GUI)

The functional icons present in the GUI can be described as below in reference to the Figure (4).

1) Functional icon used to load the natural image and the medical image to be watermarked and encrypted.

2) This functional icon is used to choose different wavelet techniques and method for the implementation of watermarking.

3) This functional icon enables the user to test the watermark images against a set of standard attacks.

4) Functional icon used to implement the encryption of the image.

5) Functions used to decrypt the image and retrieve the watermark which in this case is the medical image.

6) The resultant images of the process are displayed here.

7) The values of the validation parameters are displayed here.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 23

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

Figure 4: Screen Shot of the GUI. VII. Results and Discussion To validate the proposed approach, a Brain MRI image (MI1), a Knee MRI image (MI2), a Lung CT image (MI3) and an Ultrasound image (MI4) of fetus are considered is considered as the medical image that has to be used as the watermark image. The medical images are resized to have a size of 512* 512 to enable ease of computation and comparison of test results. The medical images used are indicatively represented in the below Figure (5). The results of Differential Evolution (DE) optimization is compared with that of Genetic Algorithm (GA).

Figure 5: Different Medical Images used in this work.

Three standard test images are used as natural images for embedding the watermark. The details of the images are enlisted in the Table I below.

IJETCAS 15-117; Š 2015, IJETCAS All Rights Reserved

Page 24

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

Table I Natural Images used for embedding the watermark Image Name Size ( Pixels) Image 1 Lena 512*512 Image 2 Fruits 512*512 Image 3 Pepper 512*512

S.No. 1 2 3

Memory (Kilo Bytes) 443 169 31.2

Four different Discrete Wavelet families namely Haar, Daubechies, Symlets and Bior splines are used in this work. RSA and AES encryption algorithms are used for encrypting the watermarked images. The Optimization algorithm Differential Evolution is used for optimization, the process can be initiated through the GUI. The below Figure (6) illustrates the steps involved in operation of the method and the tool designed.

The encryption algorithm is evaluated on the basis of correlation values. The correlation between two images refers to similarity in them. The correlation value is computed using E ( xy)  E ( x) E ( y )

CV 

(12)

E ( x )  ( E ( x)) E ( y )  ( E ( y )) Where x and y represents the input and encrypted image The Natural image, Image 2 is taken as a representative image for analysis, and the CT Lung image is considered to be the watermark. The watermark embedding process is optimized using Differential Evolution (DE) and the results presented below are the best of the ten trial runs. The results are also compared against the results of Genetic Algorithm based optimization. The below tabular column specifies performance of different types of wavelets families and the scalar function as optimized by the proposed approach.

2

2

2

2

Figure 6: From Top Left: Natural Inage, Watermark, Watermarked image, Cropped Watermarked image, Encrypted image, Extracted Watermark Table II Performance Measures PSNR, NC, MSE of Different Images using different DWT approaches Optimized using DE and GA DWT

Haar Daubechies

Type

Scalar Value

PSNR

DE Haar

GA Haar

DE 0.102353

GA

db10

db10

0.101333

0.108571 0.102857

MSE

NC

DE 59.5259

GA 58.357

DE 0.00259911

GA 0.00262427

DE 1

GA 1

60.2331

59.9389

0.00242163

0.00249393

1

1

Symlets

sym3

sym3

0.104706

0.102857

59.1847

59.5369

0.00268931

0.00259625

1

1

Bior Splines

bior2.4

bior3.9

0.102353

0.105714

59.7208

59.3307

0.00254893

0.00265034

1

1

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 25

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

Table III Performance Measures SSIM and CV of Different Images using different DWT approaches Optimized using DE and GA DWT

Haar

SSIM (RSA) DE GA 0.998978 0.998873

SSIM ( AES) DE 0.999155

CV (RSA) GA

CV (AES)

DE 0.00237855

GA 0.00271957

DE 0.00271957

GA 0.00258213

0.00270058

0.00269091

0.00275223

0.00274255

Daubechies

0.998848

0.998815

0.999024

0.998100 0.998991

Symlets

0.998896

0.998942

0.999072

0.999118

0.00284515

0.00285001

0.00289679

0.00290166

Bior Splines

0.999039

0.999051

0.999216

0.999227

0.00253048

0.00285998

0.00258213

0.00280833

From table (3) it can be observed that mutlti-objective oiptimsation has resulted in very high PSNR values and NC values. Both Genetic Algorithm (GA ) and Differential Evolution optimization performs well for this optimization problem, and the fact despite having such a high SNR, the NC value and the SSIM is very close to 1. This shows that, this kind of optimization approach is highly suitable in medical watermarking. The proposed algorithm is tested against different types of attacks namely, rotation, cropping , motion blur, sharpening and different types of noise attacks like, salt and pepper, Gaussian, speckle and poisson. The rotation operation performs a geometric transform which maps the position (x1, y1) of a picture element in an input image onto a position (x2, y2) in an output image by rotating it through user-specified angle about an origin O. The Figure (7) illustrates the different attacks on the watermarked image.

Figure 7: Watermarked image attacked with different types of attacks.

Table IV Performance of watermark under different attacks

IJETCAS 15-117; Š 2015, IJETCAS All Rights Reserved

Page 26

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29 DWT

Attack

DWT

Type Sharpening Smoothening

DE PSNR (db) 46.8779 59.9865

GA NC

NC

1

PSNR (db) 46.7961

1

59.702

1

0.983088 1 1 0.995355 1

52.4407 40.2511 44.1905 45.6927 46.7057

0.984671 1 1 0.996652 1

1 1 1 1 1

59.3022 52.2461 40.158 44.1144 45.5673

1 1 1 1 1

1

‘db 10’

Symlets

MotionBlur Salt & Pepper Noise Gaussian Noise Speckle Noise Sharpening

52.5703 40.2183 44.265 45.7445 46.6024

Smoothening Motion Blur Salt & Pepper Noise Gaussian Noise Speckle Noise

58.9608 52.0849 40.1723 44.0532 45.4582

From above Table IV it can be clearly observed that the watermarked preserves its integrity amidst different types of attacks. The PSNR value and the NC value both continue to be on the higher side, implying the fact that the watermarked image is both imperceptible as well as robust. It can also be observed that the ‘DE gives slightly better performance when compared to GA based optimization. Gaussian Noise is the most common form of noise encountered in most of the communication channels. The Gaussian noise with variance of 0.01, 0.05, 0.1, and 1 were added to the image for testing. Salt and pepper noise is also known as impulse noise. Salt and pepper noise with noise density of 0.001, 0.005, 0.01, 0.02, and 0.05 were added to the image for testing as shown in Figure (8) and( 9).

Figure 8: Reduction in PSNR of the watermarked image with increase in noise density of the salt and pepper noise.

Figure 9: Reduction in PSNR of the watermarked image with increase in variance of the Gaussian Noise.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 27

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29

To emphasize the fact that the choice of natural image also plays a role maintaining the integrity and the performance of the watermark. A single natural image, Image 3 is taken and all the 4 medical images are used as watermarks alternately to understand how it performs for different optimization,. For illustrative convenience ‘Haar’ wavelet is chosen for embedding the watermark.

Table V Performance Measures of a single chosen natural image when embedded with different medical images. – GA based optimization Image

DWT Type

Scalar Value

Natural Image : Image 3- Peppers MI1 Haar 0.105714 MI2 Haar 0.108571 MI3 Haar 0.108571 MI4 Haar 0.102857

PSNR (db)

MSE

NC

69.6959 63.402 60.1999 76.0814

0.000940038 0.00176394 0.00242969 0.000496393

0.949029 0.967536 0.97891 0.933832

SSIM RSA 0.89196 0.931073 0.947487 0.83076

AES 0.89213 0.931249 0.947663 0.830936

CV RSA 0.00270327 0.00185431 0.00268328 0.00213863

AES 0.00275492 0.00190595 0.00273492 0.00219028

Table VI Performance Measures of a single chosen natural image when embedded with different medical images. – DE based optimization.

Image

DWT Type

Scalar Value

Natural Image : Image 3- Peppers MI1 Haar 0.104706 MI2 Haar 0.104706 MI3 Haar 0.102353 MI4 Haar 0.102353

PSNR (db)

MSE

NC

69.8833 64.101 61.3406 76.1754

0.000922588 0.00164486 0.00216777 0.000491749

0.948566 0.965235 0.974583 0.933666

SSIM RSA 0.892037 0.931205 0.948741 0.830781

AES 0.892213 0.930531 0.948917 0.830957

CV RSA 0.00258404 0.0019132 0.00315385 0.00216586

AES 0.0026356 0.00182499 0.00320549 0.0022175

From the tables mentioned above it can be observed that both the methods of optimization provide results which converge and are in close agreement with each other. Although Differential Evolution (DE) based optimization yields results which slightly in better than the results obtained through Genetic Algorithm ( GA) From the above tables it can also be concluded that Medical Image 3 ( MI 3) is best suited for embedding in this natural image and the image MI4 results in very low SSIM if embedded in this natural image.

VIII. Conclusion A dual security approach using watermarking and encryption is proposed and implemented. The watermark embedding is optimized using a multi objective optimization function. The optimization is carried out using evolutionary approaches like Genetic Algorithm (GA) and Differential Evolution (DE). The results proves that this optimization procedure is capable providing high robustness and imperceptibility while maintaining the structural integrity of the medical images.

References [1] [2]

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

C.D. Schou, J. Frost, W.V. Maconachy, "Information Assurance in Biomedical Informatics Systems," IEEE Engineering in Medicine and Biology Magazine, vol. 23, no1, pp. 110–118, 2004.

M. Jiansheng, L. Sukang, and Tan Xiaomei, “A digital watermarking algorithm based on DCT and DWT”, Proceedings of the 2009 International Symposium on Web Information Systems and Applications (WISA ’09) Nanchang, PR China, May 22–24, (2009), pp. 104–107.

C. Piao, D. Woo, D. Park, and S. Han, “Medical image authentication using hash function and integer Wavelet transform”, Congress on Image and Signal Processing, (2008).

C.S. Woo, J. Du, and B. Pham, “Multiple watermark method for privacy control and tamper detection in medical images”, WDIC2005 pages, Australia, February, (2005), pp. 59–64.

M. Nambakhsh, A. Ahmadian, and H. Zaidi, “A contextual based double watermarking of PET images by Patient ID and ECG Signal”, Comput Meth Prog, 104 (2001), 341–353.

X. You, L. Du, Y. Cheung, and Q. Chen, “A blind watermarking scheme using new nontensor product wavelet filter bank”, IEEE Trans On Image Processing, 19 (2010).

W. Lin, S. Horng, T. Kao, P. Fan, C. Lee, and Y. Pan, “An efficient watermarking method based on significant difference of wavelet coefficient quantization”, IEEE Trans On Multimedia, 10 (2008), 746–757.

A. Kannammal and S. Subha Rani, “Double watermarking of DICOM medical images using wavelet decomposition technique”, Eur J Sci Res (1) (2012), 46–55 R.Z. Liu and T.N. Tan “An SVD-Based Watermarking Scheme for Protecting Rightful Ownership”, IEEE Trans. On Multimedia, Vol. 4, No. 1, pp. 121–128. ,2002.

Chin-Chen Chang, Piyu Tsai, Chia-Chen Lin, “SVD-based digital image watermarking scheme”, Pattern Recognition Letters, Volume 26, Issue 10, 15 July 2005, Pages 1577-1586 D. Bouslimi, G. Coatrieux, M. Cozic, “A joint encryption/watermarking systems for verifying the reliability of medical images”, IEEE Trans Information Technol Biomed, 16 (2012).

F. Cao, H.K. Huang, X.Q. Zhou, “Medical image security in HIPAA mandated PACS environment”, Comput Med Imaging Graphics 27 (2003), 185–196.

L.O.M. Kobayashi, S.S. Furuie, and P.S.L.M. Barreto, “Providing Integrity and Authenticity in DICOM Images: A Novel Approach”, IEEE TransInform Technol Biomed, 13 (2009).

E. Thambiraja, G. Ramesh, and R. Umarani, “A survey on various most common encryption techniques”, Int J Adv Res Comput Sci Software Eng, 2 (2012), 226–233.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 28

CH.Venu Gopal Reddy et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 17-29 [15] [16] [17]

[18] [19] [20] [21] [22] [23] [24]

W. Puech, J.M. Rodrigues, “A new crypto watermarking method for medical images safe transfer”, Proc of the 12th European signal processing conference,Vienna, Austria, 2004, 1481–1484.

X. You, L. Du, Y. Cheung, and Q. Chen, “A blind watermarking scheme using new nontensor product wavelet filter bank”, IEEE Trans On Image Processing, 19 (2010).

Sikander, B.; Ishtiaq, M.; Jaffar, M.A.; Tariq, M.; Mirza, A.M., "Adaptive Digital Watermarking of Images Using Genetic Algorithm," Information Science and Applications (ICISA), 2010 International Conference on , vol., no., pp.1,8, 21-23 April 2010.

Sawsan morkos gharghory “Hybrid Of Particle Swarm Optimization With Evolutionary Operators To Fragile Image Watermarking Based Dct” international journal of computer science & information technology (ijcsit), vol 3, no 3, june 2011.

V. Aslantas , “Optimal SVD based Robust Watermarking using Differential Evolution Algorithm” Proceedings of the World Congress on Engineering 2008 Vol I WCE 2008, July 2 - 4, 2008, London, U.K. Musrrat Ali, Chang Wook Ahn, Millie Pant, “A robust image watermarking technique using SVD and differential evolution in DCT domain”, Optik - International Journal for Light and Electron Optics, Volume 125, Issue 1, January 2014.

Matlab R 2012 a Wavelet Tool Box Reference Manual.

PKCS #1 v2.2: RSA Cryptography Standard RSA Laboratories October 27,, 2012.

Federal Information Processing Standards Publication 197, Advanced Encryption Standard (AES), November 26, 2001.

Differential Evolution, “A Practical Approach to Global Optimization”, Series: Natural Computing Series Price, Kenneth, Storn, Rainer M., Lampinen, Jouni A.2005, XX, 538 p.

Acknowledgements The author’s would like to acknowledge the Priyadarshini College of Engineering & Technology, Nellore and University College of Engineering and Technology, Acharya Nagarjuna University, Guntur for providing a platform for doing this research work.

Authors Biography CH. Venugopal Reddy is working as Professor in the department of ECE, Priyadarshini College of Engineering & Technology, Nellore, affiliated to JNTU Anantapur, A.P. Has more than 13 years of teaching experience. He got his B.Tech(ECE) from K.S.R.M. Engg College, Kadapa, A.P, affiliated to S.V University, Tirupathi, A.P. M.E (Commn.Sys) from Dr. M.G.R Engg College, Chennai, T.N affiliated to Anna University, Chennai. He is presently pursuing Ph.D in the area of Digital Image Water Marking, Acharya Nagarjuna University, Guntur. He has a good No. of research publications in his credit. His research interests are in the areas of Signal Processing, Image Processing and Communications.

Dr.P.Siddaiah obtained B.Tech degree in Electronics and Communication Engineering from JNTUA college of Engineering in 1988. He received his M.Tech degree from SV University, Tirupathi. He did his Ph.D program in JNTU, Hyderabad. He is the chief Investigator for several outstanding projects sponsored by Defense organizations and AICTE, UGC & ISRO. He is currently working as Principal, University College of Engineering and Technology, Acharya Nagarjuna University, Guntur, A.P,India. He has taught a wide variety of courses for UG & PG students and guided several projects. Several members successfully completed their PhD under his guidance. He has published several papers in National and International Journals and Conferences. He is a life member of FIETE, IE, and MISTE.

IJETCAS 15-117; © 2015, IJETCAS All Rights Reserved

Page 29

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Energy Efficient Routing Protocol for MANET: A Survey 1

GANESH GUPTA Asst. Professor, Amity School of Engg. & Technology Amity University, Gurgaon, Haryana, INDIA 2 MADNESH KUMAR GUPTA Asst. Professor, University College Kurukshetra University, Kurukshetra, Haryana, INDIA 3 ASHOK K. RAGHAV Director & Professor, Industrial Research and Development, Amity University Haryana, INDIA Abstract: This paper presents review on energy efficient routing protocol for MANET (Mobile Ad -Hoc Network). In MANET nodes are mobile so routing and power management is critical issue. Power consumption can occur due to receiving the data, transmitting the data traffic, mobility etc. Power failure of mobile node not only affects the node itself but also its ability to forward packets on behalf of others and hence overall network lifetime. It might not be possible to replace/recharge a mobil e node that is powered by batteries. To take full advantage of life time of nodes, traffic should be routed in a way that power consumption is minimized. Power Aware Routing is a consideration in a way that it minimizes the energy consumption while routing the traffic, aims at minimizing the total power consumption of all the nodes in the network, minimizing the overhead etc and thus, at maximizing the lifespan of the network using some Power Aware Routing Protocols. Although establishing correct and effici ent routes is an important design issue in mobile ad hoc networks (MANETs), a more challenging goal is to provide power efficient routes because mobile nodes operation time is the most critical limiting factor. Each protocol has definite advantages/disadvantages and is well suited for certain situations. The purpose of this paper is to facilitate the research efforts in combining the existing solutions to offer a more power efficient routing mechanism.

Keywords: Mobile ad hoc network, routing protocols, energy efficient routing, transmission power control, load distribution, sleep mode operation and global positioning system.

I. Introduction The computer network can be classified into two types: Wired or wireless. In wired network, data travels as electrical signals through wires, but in wireless medium, no wires are used and signals travel as electromagnetic waves through the air. With a wirelessly connected device, anyone can move around and still stay connected providing the person to be in range. This is great for office users or even students. In wired network, speed is fixed, where one will get the speed wire is capable of but in wireless network, speed is fluctuating and depends on the distance between nearest access point and whatever happens to be in between wireless device and access point itself. Further, the wireless network can be classified into two types: 1. Infrastructure based, 2.

Infrastructure less . In Infrastructure based wireless networks the mobile nodes can move while communicating with the base stations being fixed and as the node goes out of the range of a base station, it gets into the range of another base station. In Infrastructure less or Ad Hoc wireless networks the existing wireless infrastructure is expensive and inconvenient to use. An ad hoc network consists of a collection of autonomous mobile nodes formed by means of multi-hop wireless communication without using any pre-existing fixed network infrastructure. Ad-hoc networks can be classified into three categories based on applications; Mobile Ad-hoc Networks (MANETs), Wireless Mesh Networks (WMNs) and Wireless Sensor Networks (WSN) [3]. Mobile Ad-hoc Networks (MANETs) are becoming more popular these days in a wide spectrum of applications.

Wireless mobile ad-hoc networks are useful in many areas which are in TABLE I. The energy efficient routing may be the most important design criteria for MANETs, since mobile nodes will be powered by batteries with limited capacity [11]. Power failure of a mobile node not only affects the node itself but also its ability to forward packets on behalf of others and thus the overall network lifetime. This paper surveys and classifies numerous energy-efficient routing mechanisms proposed for MANETs. A mobile node consumes its battery energy not only when it actively sends or receives packets, but also when it stays idle listening to the wireless medium for any possible communication requests from other nodes. Thus, energy-efficient routing protocols minimize either the active communication energy required to transmit and receive data packets or the energy

IJETCAS 15-119; Š 2015, IJETCAS All Rights Reserved

Page 30

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

during inactive periods. The transmission power control approach can be extended to determine the optimal routing path that minimizes the total transmission energy required to deliver data packets to the destination [12].

For protocols that belong to the latter category, each node can save the inactivity energy by switching its mode of operation into sleep/power-down mode or simply turns it off when there is no data to transmit or receive. This leads to considerable energy savings, especially when the network environment is characterized with low duty cycle of communication activities. However, it requires a well-designed routing protocol to guarantee data delivery even if most of the nodes sleep and do not forward packets for other nodes. Another important approach to optimizing active communication energy is load distribution approach [5].

Table II Shows Useful Area in Ad-hoc Network 1.Military Environments • Special operations • Homeland defence • Soldiers, tanks, plants • Automated battlefield

2.Civilian environments • Disaster Recovery (flood, fire, earthquakes etc) • Law enforcement (crowd control) • Search and rescue in remote areas • Environment monitoring (sensors) • Space/planet exploration • Boats, small aircraft • Sports stadiums • Taxi cab network

3.Commercial • Sport events, festivals, conventions • Patient monitoring • Ad hoc collaborative computing (Bluetooth) • Sensors on cars (car navigation safety) • Vehicle to vehicle communications

This paper surveys and classifies numerous energy-efficient routing mechanisms for MANETs. The main focus on motivation, research challenges, recent development and modifications in this widely used field. And also, see how conventional routing protocols are modified to make them as energy efficient. We believe the survey can be a great source of information for researchers in ad-hoc networks. Finally, discuss the latest development, industry effort and the future direction for further research is identified problems. While it is not clear whether any particular algorithm or a class of algorithms is the best for all scenarios, each protocol has definite advantages/disadvantages and is well-suited for certain situations. However, it is possible to combine and integrate the existing solutions to offer a more energy-efficient routing mechanism. Since energy efficiency is also a critical issue in other network layers, considerable efforts have been devoted to developing energy-aware MAC and transport protocols [7]. Each layer is supposed to operate in isolation in layered network architecture but, as some recent studies suggested, the cross-layer design is essential to maximize the energy performance .

The paper is organized as follows; Introduction is presented in Section I. Section II presents classification of routing protocols then general discussions on energy efficiency which includes two sub sections, first sub section discusses the general definition of energy efficiency and the second sub section includes a discussion on routing algorithms and energy efficiency. Section III, survey on energy efficient routing protocol. Section IV presents result and comparison and finally conclusion is presented in Section V. II. Overview of Routing Protocols It is the process of establishing path and forwarding packets from source node to destination node. It consists of two steps, route selection for various source-sink pairs and delivery of data packets to the correct destination.

Various protocols and data structures (routing tables) are used to meet these two steps. This survey paper is focussed on finding and selecting energy efficient routes. We are going to discuss the main approaches in the routing are proactive, reactive, and hybrid then the general definition of energy efficiency as well as energy efficient routing algorithms.

A. Classification of Routing Protocols for MANET 1) Proactive (Table-Driven) routing protocols: A Proactive protocols continuously maintain fresh list of destinations in the network by exchanging topological information among the network nodes. Thus, when there is a need for a route to a destination, such route information is available immediately. Examples of the proactive protocols are - DSDV [2] (Destination-Sequenced Distance-Vector), Wireless Routing Protocol (WRP)[7], Hierarchical State Routing protocol, Optimized Link State Routing,(OLSR)[4] and Topology Dissemination based on Reverse-Path Forwarding routing protocol (TBRPF)[36]. The main disadvantages of such algorithms are high latency time in route finding and excessive flooding can lead to network clogging.

2) Reactive (On demand) routing protocols: The reactive routing protocols are based on some sort of queryreply dialog. It is also called on demand routing. It is more efficient than proactive routing and most of the current work and modifications have been done in this type of routing for making it more and more better. The main idea behind this type of routing is to find a route on demand between a source and destination whenever that route is needed. Discovering the route on demand avoids the cost of maintaining routes that are not being used and also controls the traffic of the network because it doesn’t send excessive control messages which significantly create a large difference between proactive and reactive protocols. Time delay in reactive protocols is greater comparative to proactive types since routes are calculated when it is required. e.g. Ad-hoc On Demand Distance Vector (AODV)[18], and Dynamic Source Routing (DSR)[7][9] etc.

3) Hybrid routing protocols: Both of the proactive and reactive routing methods have some advantages and shortcomings. In hybrid routing a well combination of proactive and reactive routing methods are used which

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 31

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

are better than the both used in isolation. It includes the advantages of both protocols. Examples of hybrid protocols are Zone Routing Protocol (ZRP)[38], Hazy Sighted Link State(HSLS)[40] etc. Some location based routing protocols are LAR [7][20], LEER [17], LEARN [23], DREAM [26] etc. Some others approaches are Flow oriented routing, Adaptive routing, Hierarchical routing etc. Some examples of flow oriented routing protocols are Link life Based Routing Protocol (LBR)[7], Lightweight Mobile Routing Protocol (LMR) [28], Lightweight Underlay Network Ad-hoc Routing Protocol (LUNER)[29], Link Quality Source Routing etc.

Some examples of Adaptive routing protocols are TORA [34] etc. Some examples of Hierarchical routing protocols are Cluster Based Routing Protocol (CBR)[32], Core Extraction Distributed Ad-hoc Routing (CEDAR)[30][33], Dynamic Address Routing (DART), Fisheye State Routing Protocol (FSR)[31], Global State Routing Protocol (GSR)[7], Hybrid Ad-hoc Routing Protocol (HARP)[7]etc.

B. Definition of Energy efficiency For a wireless networks, the devices operating on battery try to pursue the energy efficiency heuristically by reducing the energy they consumed, while maintaining acceptable performance of certain tasks. Using the power consumption is not only a single criterion for deciding energy efficiency. Actually, energy efficiency can be measured by the duration of the time over which the network can maintain a certain performance level, which is usually called as the network lifetime. Hence routing to maximize the lifetime of the network is different from minimum energy routing.

Minimum energy routes[1][6] sometimes attract more flows, and the nodes in these routes exhaust their energy very soon; hence the whole network cannot perform any task due to the failure on these nodes. In other words, the energy consumed is balanced consumed among nodes in the networks. Routing with maximum lifetime balances all the routes and nodes globally so that the network maintains certain performance level for a longer time. Hence, energy efficiency is not only measured by the power consumption but in more general it can be measured by the duration of time over which the network can maintain a certain performance level.

There are lots of ways to categorize routing algorithms One is flooding and broadcast routing, which is often necessary during the operation of the wireless network, such as to discover node failure and broadcast some information. The second kind is multicast routing, which is very common in wireless networks, to communicate in a one-to-group fashion. The last is unicast, which is always in an end-to-end fashion and the most common kind of routing in networks.

It goes without saying that node failure is very possible in the wireless network. Hence, saving energy when broadcasting in order to recover from the node failure or to re-routing around the failed nodes is essential. By the same token, multicast has the same challenge to achieve the energy efficiency [11][19]. For unicast, it is highly related to the node and link status, which require a wise way to do routing as well. Sometimes, shortest path routing is possibly not the best choice from the energy efficiency point of view. Energy is a limiting factor in case of Ad-hoc networks[12][13][24]. Routing in ad-hoc networks has some unique characteristics. Firstenergy of nodes is crucial and depends upon battery which has limited power supply. Second-nodes can move in an uncontrolled manner so frequent route failures are possible. Third-wireless channels have lower and more variable bandwidth compare to wired network. Energy efficient routing protocols are the only solution to above situation. Most of the work of making protocols energy efficient has been done on “on demand routing protocols” because these protocols are more energy efficient rather than proactive protocols but still these have some drawbacks which have been discussed in the next section. Energy efficiency can also be achieved by sensible flooding at the route discovery process in reactive protocols. And energy efficiency can also be achieved by using efficient metric for route selection such as cost function, node energy, battery level etc. Here energy efficiency doesn’t mean only the less power consumption here energy efficiency means increasing the time duration in which any network maintains certain performance level. We can achieve the state of energy efficient routing by increasing the network lifetime and performance and all the protocols discussed in this paper are based on this concept. Intensive research has been done in energy efficient communication to achieve power efficient, multi-hop communications in ad-hoc networks. Generally they can be classified into four types, which are described in detail in the following. First let us look at an simple example. Considering the multi-hop communication channels from A to B, and the intermediate node C between them, see Fig. 1.

Figure 1 A Simple Scenario for Energy Consumption in Multi-hop Networks

The possible working states for the wireless modules (including transmitter or receiver) could be transmitting, receiving, idle, sleeping. The corresponding power consumed in these states can be represented by Ptx(SNR(d)), Prx, Pid, and Ps, where SNR(d) is the signal to noise ratio for certain reliable transmission over some communication range. Therefore Ptx is the function of d, which is the transmission range such as the distance

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 32

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

between A and B, or C. Also, the energy consumed is the function of the data rates over the channel. Therefore, there are two possible ways to communicate between A and B. One is to directly transmit date from A to B; or we can relay with Node B. However, these two different methods lead to two different level of energy consumption, in which one must be better. Assume that the power consumed is linear with the data rate sent over the links. Then the two different methods have two levels of power consumption. Assume the link bandwidth is W. The first transmits data directly from A to B, with based on this observation, following ways are discussed to deal with this issue.

P 1 = R*(P tx (│AB│))/W + R*P rx /W + 2(1-R/W)*P id +P s If sending data through C, then P 2 = R*(P tx (│AC│))/W + R*(P tx (│CB│))/W + 2R*P rx /W +(3-4R/W)*P id.

First we show some protocols that only consider the control of transmission range. Then the sleep scheduling methods are described. Finally, a unified approach is proposed.

III. Survey on Energy Efficient Routing Algorithms and Protocols Just as its name implies, power aware routing [44][21] is to choose appropriate transmission range and routes to save energy for multihop packet delivery [6]. Let first discuss the five metrics for power aware routing:  Minimize Energy consumed per packet: the most intuitive metric, however not optimal for maximum lifetime  Maximize Time to Network Partition: important for mission critical applications, hard to maintain low delay and high throughput simultaneously  Minimize Variance in node power levels: balance the power consumption for all the node in the network, i.e., all nodes in the network have the same importance  Minimize Cost per packets: try to maximize the life of all the nodes  Minimize Maximum Node Cost: try to delay the nodes failures.

Then used them as the new power aware metric for determining the routes, which shows that per packet cost is reduced by 40-70% and mean time node failure increase significantly. Assuming the lifetime for each node is inversely proportional to the information going through that node, the authors in [8][9][10] use the optimal lifetime as the key metric, trying to maximize the minimum lifetime for individual nodes under the constraints of information flows at each nodes. In order to solve this problem, the authors proposed distributed algorithms using bisection search. One is the heuristic flow redirection algorithm [21], whose basic idea is to start from a feasible routing strategy, then redirect the flows from nodes with low lifetimes to nodes with higher lifetimes.

Another algorithm in [9] uses bisection search for successive feasible routing strategies.

A. Proactive energy aware routing protocols and algorithms First of all we will discuss about proactive routing protocols which are categorized further in a following way on the basis of the algorithms used.

A.1 Destination-sequenced distance vector (DSDV) DSDV [1] (destination sequenced distance vector) is the most obvious proactive protocol which was given by Perkins and bhagvat. It is based on bellman ford algorithm; it removed the shortcomings (loops, count to infinity problem) of contemporary distance vector protocol which was not suited for ad-hoc networks. It is a destination based distance vector routing protocol in which every node maintains a routing table. This routing table contains all available destinations, the next node to reach to destination, and the no of hops between it. Whenever any node changes its position it broadcast the routing updates to the other nodes. Sequence number is used to avoid loop freeness. Keeping the simplicity of distance vector protocol it guarantees loop freeness it reacts immediately on topology changes. Since the route for destination is always available at the routing table of each node so there is no latency caused by route discovery. But broadcasting of routing updates may cause high traffic load between the nodes if the density of the nodes are high. So this protocol is best suited if the density of the ad-hoc network is low. However if the mobility of the node is too high broadcasting updates may cause time delay.

A.2 Optimized Link state routing (OLSR) OLSR [8] is another link state proactive protocol which routes to all reachable nodes in the network with minimal delay. It was developed by IETF (Internet Engineering Task Force) which is an open standard organization. In this very protocol we use the concept of selective flooding which reduces the network traffic and power consumption for highly dense network since it allows only to the set of nodes (MPR’s) to broadcast the control messages whenever the topology changes. it removed the problem of unnecessary duplication of control messages. The main advantage of OLSR protocol was that it was good for dense network which was not supported by AODV protocol. In OLSR each node periodically broadcast hello messages to learn topology up to 2 hops. Based on this hello messages each node select its set of MPR’s. The problem in this type of protocol is to select a minimal set of MPR each time the topology changes which is a NP hard problem. However in this paper we are concerning on the energy efficient protocols the traditional OLSR protocol was not suitable for the viewpoint of energy efficiency which is a critical issue in case of mobile ad-hoc network. Several enhancements

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 33

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

have been done by the professionals for making it energy efficient which is as follows OLSR protocol does not take energy saving techniques into account proposed a new energy efficient unicast routing protocol EOLSR [3] which made it energy efficient. EOLSR increases the network lifetime by selecting the path having minimum cost where the cost is calculated on the basis of residual energy of each traversed node and the energy conserved on this path.

A.3 Energy-efficient broadcast OLSR A new protocol EBOLSR [8] is proposed in 2008 which adapts the OLSR protocol in order to maximize the network lifetime for broadcast communications. In EBOLSR energy efficient MPR [8][19] selection is done by the residual energy of nodes in this protocol we considers the weighted residual energy of energy efficient MPR candidate and its 1 hop neighbors. The basic phenomenon about this EBOLSR protocol was to select the energy efficient multipoint relays[MPR].

A.4 Energy-Efficient OLSR EEOLSR [7][8] is another enhancement of OLSR [4] for increasing the network lifetime without loss of performance. Two mechanisms are used in this protocol.  EA-Willingness Setting mechanism  Overhearing Exclusion In EA-Willingness setting mechanism we consider the energy state of the node in MPR selection. Every node shows the willingness for being an MPR heuristic value of the node (default, high, low) is used to determine which node can work as an MPR. The heuristic value is calculated with the help of battery capacity and predicted lifetime of a node. If the battery charge is low that node will have LOW heuristic value whereas if the battery is highly charged and there exist a low traffic in that node then the node will have HIGH heuristic value.

In the overhearing Exclusion device is turned off when neighborhood nodes exchanges message with each other.

This method saves significant amount of energy. In this way EEOLSR solved the problem of energy efficiency of conventional OLSR protocol.

B. Reactive energy aware routing protocols and algorithms In reactive routing, the routes are discovered only when need of that route arises. There are two types of reactive routing. Source Routing: In source routing, data packets carry the complete addresses from source to destination and no routing table in intermediate nodes. Some source routing protocols are: Dynamic Source Routing, Associatively Based Routing [16], and Signal Stability-based Adaptive Routing [17] B.1 Dynamic Source Routing DSR [9] is a source routing protocol it means the sender node knows the complete route to the destination.

These routes are stored in the route cache. If a node has data to send and no route is present then route discovery process will go on. Route discovery is basically based on flooding mechanism in which route request (RREQ) packets is sent to all its neighbours. Each intermediate node rebroadcasts it unless it is the destination or it has a route to the destination. This type of node replies to the request with a route reply packet that is routed back to the source node. If the node has already treated this route request it rejects the new received request. Route maintenance will go on if a link of route is broken then it deletes each route having this link from its cache, then it generates a route error packet to inform the source node and all intermediate nodes about this link failure until this route error packet reaches to the destination. After that a new route request launched by source to find a new route or check in its route cache. Due to caching DSR is more effective at low mobility and at low loads. But, it has many limitations such as it doesn’t take into consideration the capacity of each node as power computing and no security mechanism is defined for DSR. B.2 Weight based DSR (WBDSR) WBDSR [10] Weight Based DSR is an improvement of conventional DSR. In this protocol, the weight of each route is considered as metric for route selection. Weight of each route can be calculated as:  Compute the node weight of each node weight i= battery level of this node + Stability of this node  Compute the route-weight as the minimum of all node weights included in this route.  To select the main route the one having the maximum route-weight.  If two or more routes have the same route-weights then choose the route which has minimum hops.

Thus WBDSR gives always the longest network life time in both high mobile networks and static networks because it timely change the used route with another one which maintains the use of the nodes which enhances the network life time.

B.3 Energy Dependent DSR (EDDSR) DDSR [14] is energy dependent DSR algorithm which helps node from sharp and sudden drop of battery power.

EDDSR provides better power utilization compare to least energy aware routing (LEAR)[23] and minimum drain rate(MDR). EDDSR avoids use of node with less power supply and residual energy information of node is useful in discovery of route. Residual battery power of each node is computed by itself and if it is above the specific threshold value then node can participate in routing activities otherwise node delays the rebroadcasting of route request message by a time period which is inversely proportional to its predicted lifetime. With help of

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 34

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

ns-2 simulator author performed simulation which shows MDR and EDDSR is better than DSR in terms of node lifetime. EDDSR has further advantage over MDR because it can use route cache used by DSR. B.4 Dynamic Source Routing-Cache (DSR-C) DSR-Cache(DSR-C)[15] is the variant of existing DSR protocol which is based on energy efficient modification to this protocol .DSR works in two phase-first is route discovery and second is route maintenance. It is essential to reduce the cost of route discovery therefore each node maintains cache of source routes it has obtained through route discovery.

B.5 Associatively-Based Routing (ABR) ABR [16] Associatively-Based Routing Protocol is another protocol in which selecting the longer lived route is the main concern because it will help in reducing cost of reconstructing routes. The metric used instead of the shortest hop count is the Location Stability or the associatively between nodes. Moving nodes tend to break the associatively with their neighbours and hence they are not good candidates to carry routes. Nodes periodically broadcast beacons to signify their existence with their neighbours; Location Stability is determined by counting the periodic beacons that a node receives from its neighbours. Links between nodes are classified into Stable and Unstable links based on the count of beacons. When source node broadcasts route request packets, each neighbour will check if it received this request before or if its ID is in the list. If yes it will drop the packet. If not it will append its ID and the status of the link weather it is stable or not to the packet and rebroadcast the packet again. Destination node will select the route one with less unstable links while it is not shortest one.

B.6 Signal Stability-based adaptive routing (SSA) SSA [17] Signal Stability-based adaptive routing protocol is a derivative of ABR. In this protocol signal strength works as a prime metric for route selection. When source node broadcasts route request, route requests are forwarded through stable links only. Route requests received through unstable links are dropped.

Destination node once get the first route request over a stable links it will send a route reply to the source.

IV. Result and Comparison The comparison of routing protocols based upon the above discussion is as follows: DSR and LAR protocol has multiple routes available in its route table while ABR, SSA, AODV has single route.

Figure 2 Comparison of protocols based on several criteria

LAR, AODV and DSR uses shortest path as metric for route selection, ABR uses Link Stability as metric for route selection and SSA uses Signal Strength for route selection. ABR, SSA and AODV uses beacons for

IJETCAS 15-119; Š 2015, IJETCAS All Rights Reserved

Page 35

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37

monitoring the routes but DSR and LAR does not use beacons. The route discovery process in DSR, ABR, SSA and AODV is global but in LAR it is localized. So this protocol helps in prolonging the network life time. The main goal is to extend the average lifetime for each node while the balancing the total energy consumption among all nodes in network. As per discussion held in this survey, results slight improvement in all energy efficient protocols over conventional routing protocol in term of energy.

V. Conclusion This paper concludes that there is not a single protocol which can give the best performance in ad-hoc network.

Performance of the protocol varies according to the variation in the network parameters, as we know that in adhoc network properties continuously vary. Sometimes the mobility of the node of the network is high while sometimes energy of the node is our prime concern. So, we will choose the protocol in such a way that which perform best for that particular type of network. That’s why we have surveyed many types of conventional protocols and their modification which includes energy efficiency. Energy efficiency is one of the main problems in a MANET, especially in designing a routing protocol. Given paper surveyed and classified various conventional and energy efficient routing protocols. In many cases, it is complicated to compare them directly since each technique has a different objective with different assumptions and employs different means to achieve the objective. Our prime concern is energy efficiency and we have tried to discuss almost all possible approaches of energy efficient protocols.

References [1]

[2]

[3]

[4]

[5] [6]

[7] [8] [9] [10]

[11] [12] [13] [14] [15] [16] [17] [18] [19]

[20] [21] [22]

Jangsu Lee, Seunghwan Yoo, and Sungchun Kim, “Energy aware Routing in Location based Ad-hoc Networks,” Proceedings of the 4th International Symposium on Communications, Control and Signal Processing, ISCCSP 2010, Limassol, Cyprus, 3-5 March 2010.

Fotis Diamantopoulos and Anastasios A. Economides, "A Performance study of DSDV-based CLUSTERPOW and DSDV routing algorithms for sensor network applications," In proceeding of IEEE International Symposium on Wireless Pervasive Computing 2006 , pp. 1-6, January H. Zhang, A. Arora, Y. Choi and M. G. Gouda. 2005.Reliable Bursty Converge cast in Wireless Sensor Networks. Proceedings of the 6th ACM International Symposium on Mobile Ad Hoc Networking and Computin g (MobiHoc’05), Urbana-Champaign, IL, May P. Jacquet, P. Muhlethaler, T. Clausen, A. Laouiti, A. Qayyum and L. Viennot, “Optimized Link State Routing Protocol for Adhoc Networks,” Multi Topic Conference, 2001. IEEE INMIC 2001, Technology for the 21st Century. Proceedings. IEEE International. Issue Date: 2001.

Radhika D. Joshi and Priti P. rege, “Distributed Energy Efficient Routing in Ad-hoc Networks,” in 978-1-4244-3328-5/08 in IEEE 2008.

Xiaoying Zhang, Thomas Kunz, Li Li and Oliver Yang, “An Energy-efficient Broadcast Protocol in MANETs,” Communications Networks and Services Research Conference, Proceedings of the 2010 8th Annual Communication Networks and Services Research Conference, Pages: 199-206 SBN: 978-0-7695-4041-2.

Jones CE, Sivalingam KM, Agrawal P, and Chen JC. A survey of energy efficient network protocols for wireless networks.

Wireless Networks 2001; 7(4): 343–358.

Floriano, De Rango, Marco Fotino and Salvatore Marano, “EE-OLSR: Energy Efficient OLSR Routing Protocol For Mobile Adhoc Networks,” in Proceedings of Military Communications (MILCOM'08), San Diego, CA, USA, November 17-19, 2008.

David B. Johnson and David A. Maltz, ”Dynamic Source Routing in Ad-hoc Wireless Network,” The Kluwer International Series in Engineering and Computer Science, 1996, Volume 353, 153-181, DOI: 10.1007/978-0-585-29603-6_5 .

Benamar KADRI, Mohammed FEHAM and Abdallah M’HAMED, “Weight based DSR for Mobile Ad Hoc Networks,” in 3rd International Conference on Information and Communication Technologies: From Theory to Applications, 2008. ICTTA 2008. pp. 1- 6, 7-11 April 2008.

Goldsmith AJ, and Wicker SB. “Design challenges for energy-constrained ad hoc wireless networks.” IEEE Wireless Communications 2002; 9(4): 8–27.

V. Kauadia and P.R. Kumar, "Power Control and clustering in ad hoc networks," IEEE INFOCOM 2003.

Juan A. Sanchez and Pedro M. Ruiz, “ LEMA: Localized Energy-Efficient Multicast Algorithm based on Geographic Routing,” in Proceedings. Of 2006 31st IEEE Conference on Local Computer Networks in 2006.

J.-E. Garcia, A. Kallel, K. Kyamakya, K. Jobmann, J.-C. Cano and P. Manzoni, “A Novel DSR-based Energy-efficient Routing Algorithm for Mobile Ad-hoc Networks,” in vehicular technology conference 2003 IEEE. Hai-Keong Toh, “Associativity-Based Routing for Ad-Hoc Mobile Networks in Wireless Personal Communications,” An International Journal Volume 4 , Issue 2 (March 1997) Pages: 103 – 139.

Hai-Keong Toh, “Associativity-Based Routing for Ad-Hoc Mobile Networks in Wireless Personal Communications,” An International Journal Volume 4 , Issue 2 (March 1997) Pages: 103 – 139.

Rohit Dube, Cynthia D. Rais, Kuang Yeh Wang and Satish K. Tripathi, “Signal Stability Based Adaptive Routing for Ad-hoc mobile Networks,” in Science Technical Report Series; Vol. CS-TR-3646 in 1996.

Charles E. Perkins and Elizabeth M. Royer, “ Ad-hoc on demand distance vector Routing ,” in Proceedings of the Second IEEE Workshop on Mobile Computer Systems and Applications in 1999.

Tai Hieng Tie, Chong Eng Tan and Sei Ping Lau, “Alternate Link Maximum Energy Level Ad-hoc Distance Vector Scheme for Energy Efficient Ad-hoc Networks Routing,” in International Conference on Computer and Communication Engineering (ICCCE 2010), 11-13 may 2010,Kuala Lumpur, Malaysia.

Young-Bae Ko and Nitin H. Vaidya, “Location-Aided Routing (LAR) in mobile ad-hoc networks in Wireless Networks,” Volume 6 , Pages: 307 - 321 Issue 4 (July 2000).

Z. Liu and A. Sankar, "Maximum lifetime routing in wireless ad-hoc networks," IEEE INFOCOM 2004.

J. Choi, Y. Ko, and J. Kim, “Utilizing directionality information for power-efficient routing in ad hoc networks,” Proc. IEEE/IEE International Conference on Networking (ICN’04), Feb 2004.

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 36

Ganesh Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 30-37 [23]

[24] [25]

[26]

[27] [28] [29] [30] [31] [32] [33]

[34] [35] [36] [37]

[38] [39] [40] [41] [42] [43] [44]

Yu Wang, Wen-Zhan Song, Weizhao Wang, Xiang-Yang Li and Teresa A. Dahlberg, “ LEARN: Localized Energy Aware Restricted Neighborhood Routing for Ad-hoc Networks,” in Third Annual IEEE Communications Society Conference on Sensor, Mesh and Ad-hoc Communications (IEEE SECON 2006).

Dahai Du and Huagang Xiong, “A Location aided Energy-Efficient Routing Protocol for Ad-hoc Networks,” in wireless and optical communications conference (WOCC), 2010 19th annual.

S. Murthy, and J.J. GarciaLuna-Aveces, “An Efficient Routing Protocol for Wireless Networks,” AACM/Baltzer Journal on Mobile Networks and Applications, Special Issue on Routing in Mobile Communication Networks, Vol. 1, No. 2, pp 183-197, ACM, October 1996.

Stefano Basagni, Irnrich Chlamtac, Violet R. Syrotiuk and Barry A. Woodward, “ A Distance Routing Effect Algorithm for Mobility (DREAM),” in Proceedings of the 4th annual ACM/IEEE international conference on Mobile computing and networking texas, United States, pages :76-84 in 1998.

Yonghui chen, chunfeng zhang and zhiqin liu, “Energy Efficient Routing Protocol Based on Energy of node and Stability of Topology,” in 3rd International conference on Information and computing 2010.

M.S. CORSON AND A. EPHREMIDES, “Lightweight Mobile Routing protocol (LMR), A distributed routing algorithm for mobile wireless networks, Wireless Networks 1 (1995).

C.Tschudin and R. Gold, ”Lightweight Underlay Network Ad hoc Routing (LUNAR),” http://cn.cs.unibas.ch/projects/lunar/ 34 R. Sivakumar, P. Sinha, V. Bharghavan, “Core Extraction Distributed Ad hoc Routing (CEDAR) Specification,” Internet Draft, http://tools.ietf.org/html/draft-ietf-manet-cedar-spec.

M. Geria, Guangyu, X. Hong, and T. Chen, “Fisheye State Routing Protocol (FSR) for Ad Hoc Networks Internet Draft,” http://tools.ietf.org/html/draft-ietf-manet-fsr, work in progress, June 2001.

M. Jiang, J. Li, Y. C. Tay, “Cluster Based Routing Protocol (CBRP) Functional Specification Internet Draft,” http://tools.ietf.org/html/draft-ietf-manet-cbrp-spec, work in progress, June 1999.

P. Sinha, R. Sivakumar, V. Bharghavan, “CEDAR: A Core-Extraction Distributed Ad Hoc Routing Algorithm,” The 18th Annual Joint Conference of the IEEE Computer and Communications Societies, INFOCOM '99 New York, NY, USA, pp. 202-209 IEEE, March 19 .

A. Iwata, C. C. Chiang, G. Pei, M. Gerla, and T.W. Chen, "Scalable Routing Strategies for Ad-Hoc Wireless Networks" IEEE Journal on Selected Areas in Communications, Special Issue on Ad-Hoc Networks, Aug. 1999, pp.1369-79.

N. Nikanein, C. Bonnet, and N. Nikaein, “Hybrid Ad Hoc Routing Protocol - HARP,” in proceeding of IST 2001: International Symposium on Telecommunications http://www.eurecom.fr/~nikaeinn/harp.ps B. Bellur, R. G. Ogier, and F. L. Temlin,” Topology Dissemination Based on Reverse-Path Forwarding (TBRPF)”, RFC 3684, February 2004.

Saleh Ali K. Al-Omari and Putra Sumari, “ An over view of Mobile Ad Hoc Networks for Existing Protocols and Applications,” International Journal on applications of graph theory in wireless ad hoc networks and sensor networks (Graph-Hoc), Vol.2, March 2010.

Haas ZJ, Pearlman MR, Samar P (2002) The Zone Routing Protocol (ZRP) for Ad Hoc Networks. IETF draft, July 2002, available at http://tools.ietf.org/id/draft-ietf-manetzone-zrp-04.txt. Accessed 21 February 2008 94 A.-S.K. Pathan and C.S. Hong.

V. D. Park and S. Corson, “Temporarily-ordered routing algorithm (TORA) version 1 functional specification,”corsondraft-ietfanet.tora-spec-00.txt, IETF, Internet draft, 1997.

V. Rodoplu and T. H. Meng, "Minimum Energy Mobile Wireless Networks," IEEE J. Selected Areas in Communications 17(8), 1999.

V. Kauadia and P.R. Kumar, "Power Control and clustering in ad hoc networks," IEEE INFOCOM 2003.

P .Santi, "Topology Control in Wireless Ad Hoc and Sensor Networks," ACM Comp. Surveys, Vol. 37, n. 2, p. 164-194, June 2005.

B. Krishnamachari, "Networking Wireless Sensors," Cambridge University Press, December 2005.

S. Singh, M. Woo, and C.S. Raghavendra, "Power-Aware Routing in Mobile Ad Hoc Networks," ACM/IEEE MOBICOM 1998. .

IJETCAS 15-119; © 2015, IJETCAS All Rights Reserved

Page 37

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Solving Generalized Assignment Problem with Genetic Algorithm and Lower Bound Theory 1

Mr. VikasThada, 2Mr. Utpal Shrivastava, 3Ms. Meenu Vijarania 1, 2,3 Assistant Professor (CSE), Amity University, Gurgaon, Haryana, INDIA.

Abstract: In this paper an attempt has been made to solve the “Assignment problem” through genetic algorithm.

In the assignment problem given N men and N jobs the task is to minimize overall cost of assignment considering the fact that a single job can be assigned to only one person. The problem is solved using genetic algorithm using an encoding scheme along with Partially Matched Crossover (PMX) function. The experimental setup has been carried out using values of N from 5 to 100. Though population size and generations are kept fixed but same can be varied in each iteration. The source codes for the above have been developed in matlab.

For comparison results from lower bound theory and result from this experiment have analyzed. A comparison table for different values of N with results is presented that also contains deviation from results obtained through lower bound theory.

Keywords: Assignment, problem, genetic, algorithm, jobs I. Introduction The assignment problem popularly known also as generalized assignment problem (GAP) is a famous NPcomplete combinatorial optimization problem [1]. In the GAP the goal is to assign a set of tasks to a set of agents with minimum cost. It is assumed that each agent has a limited amount of a single resource and every task must be assigned to one and only one agent. The problem can also be considered N men N machine problem where we have to assign each machine to one and only one man requiring that total cost of assigning man to machine is minimized [2]. The GAP is best described using knapsack problems [3]. Given n items and m knapsacks, with P[i][j] as the cost associated with assigning item j to knapsack i, W[i][j] as the weight of assigning item j to knapsack i, and C[i] the capacity of knapsack i, assign each item j to exactly one knapsack i, not exceeding knapsack capacities. Then the GAP can be formulated as[3]: Min P[i][j] * X[i][j] Subject to

X[i][j] € {0,1}, i € M, j € N. We start with overview of Genetic Algorithm, formulation of GAP using GA and then present experimental results.

II. Overview of GA Genetic Algorithms [4] are based on the principle of heredity and evolution which claims “in each generation the stronger individual survives and the weaker dies”. Therefore, each new generation would contain stronger (fitter) individuals in contrast to its ancestors.

The process of Genetic Algorithm is as follows: a. Initialize Population b. Loop i.

Evaluation ii.

Selection iii.

Reproduction iv.

Croosover v.

Mutation c.

Convergence The initial population is usually represented as a number of individuals called chromosomes. The goal is to obtain a set of qualified chromosomes after some generations. The quality of a chromosome is measured by a fitness function. Each generation produces new children by applying genetic crossover and mutation operators. Usually, the process ends while two consecutive generations do not produce a significant fitness improvement or terminates after producing a certain number of new generations.

IJETCAS 15-121; © 2015, IJETCAS All Rights Reserved

Page 38

VikasThada et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 38-42

A. Selection Once the fitness evaluation process is done next step is to perform selection operation. Process of selection operation is based on the principle of ‘‘survival of the fittest’. Higher fitness valued chromosomes goes through reproduction. Lower fitness valued chromosomes are discarded. There are number of ways to implement this operator, but all relies on the concept that candidates with good fitness values are to be preferred over poor fitness values. The idea is to give preference to better individuals. This selection operation does the replication of candidate chromosomes with good fitness values and eliminating those with poor fitness values [5].

Figure 1: Selection Operator on a Population of 4 Individuals[9]

The research work uses roulette wheel selection method as selection operator [6]. It is also known as fitness proportionate selection method.

B. Crossover[7,8] In the crossover operation mating of two chromosomes is performed that gives birth to two new offspring. This operation of crossover always happens with one parameter that is known as probability of crossover (ProC). When ProC is say 0.8 it means only 80% of the total population goes for crossover operation. Rests 20% chromosomes remain abstain from this operation and has no effect of crossover. Motive behind performing crossover operation is to explore new solutions and exploit use of old solutions. GA forms an optimum solution by mating two fit chromosomes together. Chromosomes with higher fitness will always have good selection probability then others with lower fitness values, thus a good solution moves from one generation to next generation Figure 2: Single Point Crossover Explained

C. Mutation [7, 8] Mutation involves changing one bit of a chromosome from 0 to 1 or viceversa. This is performed under the constraint parameter called probability of mutation (ProM). For example if ProM is 0.10 then 10% genes of total chromosomes will go for mutation. The concept of mutation is based on this natural theory that varying breeds are possible only by varying gene values. After this operation fitness quality of new chromosomes may be high or low then old ones. In case new chromosomes are poor then old ones they are removed during selection process. The motive behind mutation is regaining the lost and discovering varying breeds. For example: randomly mutate chromosome at position 5.

Figure 3: Mutation Operation Explained

IJETCAS 15-121; © 2015, IJETCAS All Rights Reserved

Page 39

VikasThada et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 38-42

Figure 4: Basic Operation of Genetic Algorithm [9] III. Experimentation In this section work carried out in implementing the solution of assignment problem using GA is discussed. Its quite easy to find the lower bounds and upper bounds given cost matrix. Lower bound is sum obtained by selecting minimum value from each row and upper bound by selecting the diagonal elements of the cost matrix. We emphasis on implementation using GA mainly on crossover and fitness function.

A. Chromosome Representation The representation of the chromosome will be a tuple from 1 to N considering square cost matrix and in the population every chromosome will be a permutation from 1 to N. If the ith entry in the tuple is j then job number j will be assigned to ith person. Thus indexes of tuple element is row number and their value is corresponding column. For example consider the following matrix:

18 22 33 37 19 39 13 16 17 21 21 16 13 13 21 26 15 31 34 31 27 24 14 22 29 If solution happens to be tuple [ 1 5 4 2 3] that means job 1 with cost 18 will be assigned to first person, job 5 with cost 21 will be assigned to second person and so on.

B. Crossover Function[10,11] As elements in each tuple have to be distinct, it is very much possible that after croosoverduplicacy of elements may be there in new offsprings. That’s why general crossover functions like 1-point, 2-point or random point crossover methods does not work . Consider two tuples: P1=(8 3 6 7 5 1 2 4) P2=( 6 1 5 3 7 2 8 4) Assuming crossover point chosen is 3, new offsprings will be: C1=(8 3 6 3 7 2 8 4) C2=(6 1 5 7 5 1 2 4) Clearly the new chromosomes are not valid.

Thus we have to rely on new crossover techniques where chromosomes are some permutation of some numbers between 1 and N. Table given below shows some of the permutation respected crossover techniques.

IJETCAS 15-121; Š 2015, IJETCAS All Rights Reserved

Page 40

VikasThada et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 38-42

Table 1: Few Permutation Respected Crossover Techniques Partially Mapped Crossover (PMX) Goldberg and Lingle (1985) [15] Order - Crossover (OX1)

Davis (1985) [13]

Order Based Crossover (OX2)

Syswerda (1991) [19]

Position Based Crossover (POS)

Syswerda (1991) [19]

Heuristic Crossover (HX)

Grefenstette (1987) [17]

Edge Recombination Crossover (ER)

Whitley et al. (1989) [18]

Sorted Match Crossover (SMX)

Brady (1985) [12]

Maximal Preservative Crossover (MPX)

M¨uhlenbein et al. (1988) [17]

Voting Recombination Crossover (VR)

M¨uhlenbein (1989) [18]

Alternating - Position Crossover (AP)

Larranaga et al. (1996) [16]

In the research work PMX techniques have been used. In this technique after the crossover has been done all elements where there is a conflict are adjusted by some swapping operation [1]. Consider the two tuples 2 5 1 3 8 4 7 6 4

7

2

6

1

3

5

Using 2-point crossover new tuples are: 2 5

8

1

2

6

1

7

6

7

3

8

4

3

5

8

4

The conflicts are removed as: 1. 2 appear in first tuple at position 1 and 4, 3 at position 7 is swapped with 2 at position 1.

2. 1 at position 3 and at position 6 conflict. 1 at position 3 is swapped with 4 at position 2 in second tuple.

3. Similarly 6 in first tuple is swapped with 8 in second tuple.

The new tuples will be: 3 5 4 2 6 1 7 8 6

1

7

3

8

4

2

5

C. Mutation Function For mutation two random points have been generated between 1 and N and values at these positions have been swapped.

D. Fitness Function Tuples are generated randomly so fitness function is the sum of generated tuple multiplied by their corresponding. This has to be minimized under the constraint that tuple should not have duplicated values which can be easily detected as discussed in the crossover function. Best fitness function will be one which returns result close to lower bound.

E. Selection Function Populations were duplicated by 50% of top fitness chromosomes.

F. Results For a population of size=1000 ,generations=100 and taking values of N=5,10,15,20,25,30,35,40,45,50,60,70,80,90,100 random cost matrices were generated and saved in the file. The same matrices were then read using readmat function. For each of the value of N random 100 tuples were generated. The tuple size were depended on N. Using the code written in matlab for selection, crossover, mutation and fitness result were calculated which are shown in the table .

Matrix Size 5 10 15 20 25 30 35 40 45 50 60 70 80 90 100

Table 2: Simulation Results using Genetic Algorithm for GAP Lower Bound Upper Bound Simulation Result Deviation from Lower Bound 76 107 81 5 125 270 141 16 196 432 244 48 234 489 292 58 266 606 357 91 328 672 441 113 374 891 519 145 433 1055 615 182 491 1126 731 240 527 1296 846 319 626 1455 1016 390 727 1808 1191 464 819 1910 1382 563 919 2203 1603 684 1016 2590 1784 768

IJETCAS 15-121; © 2015, IJETCAS All Rights Reserved

Page 41

VikasThada et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 38-42

As can be seen from the table that as matrix size increases the deviation from actual lower bound increases but this deviation can be tolerated as simulation results are computed after 10 different runs of algorithm and can be considered as true. Further these results are highly deviated from upper bound.

IV.

Conclusion & Future Work

In this paper we have solved the GAP using genetic algorithm using bounds from lower bound theory. We have shown that results obtained using genetic algorithm is very satisfactory as compared to results obtained using lower bound theory. We have not used the effect of probability of mutation (PM) and probability of crossover (PC) in this experimental work.

Varying the PC and PM the results can be varied. We leave it as future work. Comparison of branch and bound technique, Hungarian technique and this present work can be compared for matrices of different size and results can be analyzed.

VI. [1] [2]

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

[16] [17] [18]

[19] [20]

References

A.Sahu, R.Tapadar, “Solving the Assignment problem using Genetic Algorithm and Simulated Annealing”, International Journal of Applied Mathematics, Vol 36 issue 1 P.Juell, A.S.Perera,K.E.Nygard,” Application of a Genetic Algorithm to improve an existing solution for the General Assignment Problem”, Proceedings of the 16th International Conference on Computer Applications in Industry and Engineering, Las Vegas,2003 L.A.N.Lorena, M.G.Narciso,J.E.Beasely, “A Constructive Genetic Algorithm For The Generalized Assignment Problem”, Proceedings of the 16th International Conference on Computer Applications in Industry and Engineering, Las Vegas,2005 M. Shokouhi, P.Chubak, , Z. Raeesy “ Enhancing focused crawling with genetic algorithms” Vol: 4-6, pp.503-508,2005 M.A.Kauser, M. Nasar, S.K.Singh, “A Detailed Study on Information Retrieval using Genetic Algorithm”, Journal of Industrial and Intelligent Information vol. 1, no. 3, pp.122-127 Sep 2013. http://en.wikipedia/wiki/Fitness_Proportionate_Selection B.Klabbankoh, O.Pinngern. “applied genetic algorithms in information retrieval” Proceeding of IEEE ,pp.702-711,Nov 2004 J.R. Koza, “ Survey Of Genetic Algorithms And Genetic Programming”, Proceedings of the Wescon, pp.589-595,1995 V.Thada, V.Jaglan, “Use of Genetic Algorithm in Web Information Retrieval”, International Journal of Emerging Technologies in Computational and Applied Sciences, vol.7,no.3,pp.278-281, Feb,2014 Kelly D. Crawford, “Solving n-Queen problem using genetic algorithms”SAC '92 Proceedings of the 1992 ACM/SIGAPP symposium on Applied computing,pp:1039-1047,1994 G. Üçoluk, “Genetic Algorithm Solution of the TSP Avoiding Special Crossover and Mutation”,Intelligent Automation & Soft Computing, vol. 8, no.3, 2002 Brady, R.M. “Optimization Strategies Gleaned from Biological Evolution.” Nature317, 1985, pp. 804.

Davis, L. “Applying Adaptive Algorithms to Epistatic Domains.” Proceedings of theInternational Joint Conference on Artificial Intelligence, 1985, pp. 162-164.

Grefenstette, J. “Incorporating Problem Specific Knowledge into Genetic Algorithms.”Genetic Algorithms and Simulated Annealing, edited by Davis L., Morgan Kaufmann,Los Altos, CA, pp. 42-60, 1987.

Goldberg, D.E., and R. Lingle. “Alleles, Loci, and the Traveling Salesman Problem.”Proceedings of the First International Conference on Genetic Algorithms and Their Application, edited by Grefenstette J., Lawrence Erlbaum Associates, Hillsdale, NJ,1985, pp. 154-159.

Larranaga, P., C.M.H. Kuijpers, M. Poza, and R.H. Murga. “Decomposing BayesianNetworks: Triangulation of the Moral Graph with Genetic Algorithms.” Statistics andComputing, 1996.

M¨uhlenbein, H., M. Gorges-Schleuter, and O. Kramer. “Evolution Algorithms in Combinatorial Optimization.” Parallel Computing, 7, 1988, pp. 65-85.

M¨uhlenbein, H. “Parallel Genetic Algorithms, Population Genetics and Combinatorial Optimization.” Proceedings on the Third International Conference on Genetic Algorithms, edited by Schaffer J., Morgan Kaufmann Publishers, Los Altos, CA, 1989,pp.

416-421.

Syswerda, G. “Schedule Optimization Using Genetic Algorithms.” Handbook of Genetic Algorithms, edited by Davis L., Van Nostrand Reinhold, New York, 1991,pp. 332-349.

Whitley, D., T. Starkweather, and D. Shaner. “The Traveling Salesman and Sequence Scheduling: Quality Solutions Using Genetic Edge Recombination.” Handbook of Genetic Algorithms, edited by Davis L., Van Nostrand Reinhold, New York, 1991,pp. 350-372.

Acknowledgments I thank Google for fast and efficient search and many other research papers that we’ve studied to help do this research work. Indirectly I thank to all my colleagues at my workplace, my supervisor and above all my family for constant support in all my endeavours.

IJETCAS 15-121; © 2015, IJETCAS All Rights Reserved

Page 42

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Priority Based Scheduling in a Federated Cloud Environment 1

A. Stanislas, 2L. Arockiam Ph. D. Scholar, Associate Professor, Department of Computer Science St. Joseph’s College (Autonomous), Tiruchirappalli, Tamil Nadu, INDIA __________________________________________________________________________________________ Abstract: Scheduling is a process of allocating available resources to the submitted requests. The main function of scheduling is to optimize the resource utilization and response time. Cloud is a new architecture, where a huge number of requests are submitted by various customers. Some of these requests need multi-tiers in federated cloud environment to perform effective execution. Since there are a huge number of requests, at times, the requests with the small scale will have to wait for a long time. In order to overcome such situation, a new scheduling algorithm is proposed based on the priorities to schedule the requests in a vertically scaled virtualized machine. This algorithm works based on the properties of requests such as depth of tier and memory size. The experimental results show that the performance of the proposed algorithm is better than the existing (FCFS) algorithm.

1

2

Keywords: Scheduling, Priority, Federated Cloud, Virtualization, Multi-Tier _________________________________________________________________________________________ I. Introduction Cloud computing is an attempt to realize the vision of utility computing through the provisioning of virtualized hardware, software platforms and software applications as the services over the Internet. It has mainly three different service models known as Software as a Service (SaaS), Platform as a Service (PaaS), and Infrastructure as a Service (IaaS). The focus is on IaaS, which intends to provide and to manage the system in terms of processor (CPU), memory (RAM), storage (HDD), and network connectivity (N/W)[1]. Scaling the resources up and down is the special feature in cloud computing. Virtualization techniques [2] are used for scaling resources and web applications over the internet. They are used on a single system while executing the processes. The main contribution of the virtualization technique is virtual machines. An effective resource management [3] is an important factor for high scalability to increase the system utilization and to reduce the response time on customer requirement in average time limit. So, scheduling the requests and the resources is the challenge in Infrastructure as a Service (IaaS).

Amazon’s Elastic Compute Cloud (EC2) is one of the most popular data center in cloud environment [4].

Amazon EC2 provides its customers to install virtual machines (VMs) on demand on Amazon’s infrastructure.

They deploy services on pay for use policy like, computing, storage, and network resources. Though a good number of papers explain, the virtualization of web applications is in the multi-tier architecture level. Moreover, virtualization applications share the request in limited time period. Lots of applications are used in virtual systems like resource virtualization and resource provisioning management based on virtual machines [5], [6].

These works help to utilize and to improve the performance of virtual resources.

A lot of scheduling models are used in a multi-tier application environment. The traditional Queuing model (FCFS) is one of the important models for virtualized multi-tier applications. The main aim of queuing model is to maximize the resource utilization with the minimum response time of the customer requirement. In any queuing model, scheduler is the main part of the queuing model [7]. The goal is to maximize the use of resources under a workload which fulfils different customer for the limits of average response time.

II. Motivation Customer requests for Infrastructure as a Service are received by the infrastructure provider. To increase the ability of the service on demand of the customer, it has to increase the ability of the hardware to run applications. The application of virtual technique deals with the VM resource requests [8], and also it deals with the application for managing computing resource demands of the infrastructure customers. Allocation of resources should be lesser resource cost and must save resources by distributing workload to virtualized multitier applications. Some dynamic provisioning models [9] are used in distributed workload requests to virtualized multi-tier applications in cloud data centers.

Dynamic resource provisioning is based on hybrid model and VM based cloud data center. The customer requests are received in queuing method. So, all the workload requests stored in queue will be raised in execution period. Dynamic provisioning applications are scheduler and dispatcher. Jobs are allocated with the help of scheduler. But, in this queuing model, each task is executed one by one on the First-Come First-Serve (FCFS) basis. So, a large number of tasks will be waiting for a long period of time. This hampers the effective scheduling of the requests and allocating of the resources and that is the challenge today. Therefore, a new

IJETCAS 15-123; © 2015, IJETCAS All Rights Reserved

Page 43

A. Stanislas et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 43-48

improved scheduling algorithm is developed based on the priorities in queuing model for virtualized multi-tier application in cloud data center. The main aim of the proposed approach is to minimize the response time depending on the execution tier. To achieve this goal, sorting methodologies will be used as parameter for sorting the requests on priority base in the queue.

III. Related Works Rodrigo N. Calheiros et al. [6] proposed a model based on the behavior and performance of applications and cloud based IT resources to adaptively serve end-user requests. To improve the efficiency of the system, they used analytical performance (queuing network system model) and workload information to supply intelligent input about system requirements to an application provider with limited information about the physical infrastructure. Christian Tilgner et al. [10] presented a declarative rule based scheduler component. They proposed a scheduler component that can handle more than one request at a time and also it can handle data base query processing techniques that give high quality of scheduling. Arshdeep Bahga et al. [11] presented a design and implementation of a synthetic workload generator that accepts both the benchmark and workload model specifications. They proposed the Georgia Tech Cloud Workload Specification Language (GT-CWSL) that provides a structured way for specification of application workloads.

Waheed Iqbal et al. [12] proposed a methodology for automatic detection and resolution of bottlenecks in a multi-tier web application. They also proposed a method for identifying and retracting over-provisioned resources in multi-tier cloud. Jing Bi et al. [7] presented a novel dynamic provisioning technique for a cluster based virtualization multi-tier applications. It used a hybrid queuing model to determine the number of virtual machines at each tier. They presented a cloud data center based on virtual machine to optimize the resource provisioning. Qi Zhang et al. [13] proposed a regression-based approximation model for a dynamic provisioning. They used the approximation data in an analytical model for network of queue. Bhuvan Urgaonkar et al. [14] proposed a novel dynamic provisioning technique for multi-tier Internet. They used queuing model to develop their technique. It determined the resource allocation to each tier of application and predictive and reactive methods to provision resources. Bhuvan Urgaonkar et al. [15] presented a model based on a network of queues which were represented in different tier applications. This model was amply to the behavior of tiers with significantly different performance characteristics and applications such as session-based workloads, tier replication, load imbalances across replicas, and catching at intermediate tiers.

IV. Priority Based Scheduling Model In the existing First Come First Served (FCFS) queuing model, the requests are submitted in the Request Queue (RQ).A virtualized multi-tier application in cloud computing environment is deployed on multiple virtual machines (VMs), and each tier provides certain functionality to its preceding tier. Here, an online e-commerce application that consists of n number of tiers is considered and it is denoted by T 1, T2, T3...Tn. It is assumed that there are parallel identical VMs in each tier of virtualized Multi-tier Application. A simple architecture of scheduler for multi-tier cloud is presented in the following Figure.

Figure 1: Simple Architecture of a Multi-Tier Cloud

RQ

MSDT

Scheduler Customers

VM 1

VM 2

RQ

MSDT

VM 1

VM 2

Scheduler

RQ

MSDT

VM 1

VM 2

Scheduler

VM 3

VM 3

VM 3

VM n

VM n

VM n

Tier-1

Tier-2

Tier-n

Figure 1 shows a new architecture with the addition of MSDT for scheduler in a virtualized multi-tier application. As shown in the figure with the addition of Memory Size and Depth of Tier based (MSDT) scheduler, a novel architecture for queuing multi-tier cloud is developed. The Requests are listed in Request Queue (RQ), and it provides the request to MSDT. The functionality of the MSDT is that each request has got

IJETCAS 15-123; Š 2015, IJETCAS All Rights Reserved

Page 44

A. Stanislas et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 43-48

some of its own properties in itself. That property is called a measurement which is used to identify the request among the number of requests. This property is needed to execute in some other data centers or web applications. The VMs for corresponding requests are searched in all the tiers starting from T1 to Tn. So, a small scale service request will have to wait for a long time if the VM is in the further tier. Therefore, the addition of MSDT simplifies the process of searching the corresponding VM based on the measures namely, Memory Size and Depth of Tier.

Hence, each request has got some common measurement attributes like request Memory Size (MS), Depth of the Tier (DT), Session, etc. So, this proposed model focuses on maximizing the resource provisioning for different customers by reducing the execution time, storage time, IO, power in virtualized multi-tier. In existing queue model, the requests are collected in the queue for scheduling. The requests in the queue then will be scheduled in the First-Come First-Served (FCFS) queuing method [16]. While using this methodology, the request with minimal resource requirement will have to wait for a long time which increases the response time.

For example, let us suppose that R0, R1, R2 requests are waiting in this QM. Each request has some properties like Depth-of Tiers, Memory Size and Request Session failure time rate. Let us consider that R0 is in 3rd tier with 200 MB and its response time is 20 sec, R1 is in 2nd tier with 500 MB and its response time is 15 sec, and R2 is in 1st tier with 200 MB and its response time is 5 sec. Suppose, if these requests are executed in QM, R0 request will be passed to 1st tier, and 1st tier will check the request parameters. If this request is to be executed in 3rd tier, it has to pass through 2nd tier to 3rd tier. If this intermediate time or request time supposes 10 sec, and the request is communicated via 3rd and 2nd tier and finally reached 1st tier, then, the total communication time of R0 is 20 sec. R1 request is also processed likewise which means the total communication time would be 10 sec.

However, R3 request session failure time rate is 20 sec. So, R3 may be failed in this QM. In a federated cloud environment, the system is designed for fast execution of service requests in a limited period of time because, the service will be delivered based on pay per use and QoS is also an important factor for cloud systems. So, the MSDT model is designed for this QM, particularly an algorithm for a scheduler. This algorithm, will work based on requests’ basic measurements like Memory Size, Depth of Tier. This is a metric based algorithm to schedule the requests and to allocate the virtual machines effectively in cloud data centers.

A. MSDT Framework Design The Memory Size and Depth of Tier (MSDT) design is added to the existing Queuing Model (QM) to enhance the scheduling effectively in a federated cloud environment. This model has various components like, MSDT Log Centre (MLC), Memory Size Log (MSL), Depth of Tier Log (DTL), sorted requests based on Depth of Tier (DOT SRT), sorted requests based on Memory Size (MS SRT). The submitted requests from the customers are processed in the proposed MSDT model and are scheduled through the existing FCFS queuing model. The overall basic architecture of the MSDT scheduling model is given below in Figure 2.

Figure 2: Architectural Diagram of MSDT MSDT FSRQ

DTL

Sche duler DT SRT MLC RQ

SRQ

Customers MSL

MS SRT

APL Pr-1

APL Pr-2

APL Pr-3

OS

OS

OS

Virtual Machine Monitor (VMM) Hardware

Figure 2 shows the overall architecture of MSDT and follows the processes in different phases that are in MSDT. In the first phase, the requests are collected from different customers from different data centers. These requests are gathered in the Request Queue (RQ) list. In the second phase, the requests are sent to MSDT Log

IJETCAS 15-123; Š 2015, IJETCAS All Rights Reserved

Page 45

A. Stanislas et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 43-48

Centre (MLC). Here, the requests are segregated based on the basic metric values, such as Depth-of Tier and Memory Size. In the third phase, MSDT Log center divides the requests and stores them in a separated Log Centre like, DTL and MSL. In the fourth phase, in Depth-of-Tier Log (DTL), each request is checked for the DT center log based on a minimum tier requirement for the faster execution in a multi-tier environment. So, DT Log requests are assigned and sorted out in ascending order and kept in DT sort. In the fifth phase, the requests that are stored in Memory Size (MS) Log are collected and sent as the unsorted requests to MS Sort. In the sixth phase, the requests of both MS log which is Memory Size based and DT Log which is Depth-of-Tier based are sorted in ascending order and are checked for the maximum memory requirement in DT Log request form and again, the requests are sorted based on the Memory Size and sent to the Final Sorted Request Queue (FSRQ) for scheduling the requests. Once the entire process of MSDT is complete, the FSRQ carries out the scheduling in the multitier cloud environment. This process of segregating and sorting out the requests according to the priorities based on the measures takes place in all the tiers in order to have the efficient way of scheduling.

B. Scheduling Algorithm The steps involved in MSDT Algorithm are listed in algorithm. To prove the performance of proposed algorithm, some sample values have been taken and evaluation has been done. The results obtained are discussed in the next section. In the following algorithm let us assume that, n is the number of requests, and R represents ith request. MCDOT (Ri) denotes the Metric Collector Depth-of-Tier. DOTLOG is denoted by the Depth of the tier Log. The requests are assigned in Step1. Step2 describes the requests Depth-of-Tier segregated from metric collector and stored in DOTLOG with i = 0, 1, 2… Step3 describes the request based on memory size which is stored in MSLOG with i = 0,1,2…n. SORTDOT (Ri, DOTLOG) function is used to sort the requests in the Depth-of-Tier list and that output will be sorted in SRQ (Sorted Request Queue). SORTMS (SRQ, MSLOG) is a function used to sort the memory size based on SRQ. The pseudo code for the algorithm is given below.

----------------------------------------------------------------------------MSDT Scheduling Algorithm ----------------------------------------------------------------------------//Parameters Declaration:MCDOT – Metric Collector of Request Depth-of-Tier MCMS – Metric Collector of Request Memory Size DOTLOG – Depth-of-Tier Log Centre MSLOG – Memory Size of Log Centre SORTDOT – Sort of Request’ Depth-of-Tier SORTMS – Sort of Request’s Memory Size Ri – ith Request FSRQ – Final Sort Request Queue 1.

Let n requests 2.

MCDOT (Ri) => DOTLOG µi = 0 to n 3.

MCMS (Ri) => MSLOG µi = 0 to n 4.

SORTDOT (Ri, DOTLOG) => SRQ µi = 0 to n 5.

SORTMS (SRQ, MSLOG) => FSRQ µi = 0 to n 6.

DISPATCH (FSRQ) -------------------------------------------------------------------------------------V. Evaluation and Results The evaluation and experimental results of the efficiency of the automatic service request scheduling technique for optimizing the number of VMs in a multi-tier cloud environment is presented. The results show that under the fine-grained resource provisioning, the providers achieved revenues which can be maximized while the customers’ operational cost is reduced as much as possible. The following experiments show the validation of the model.

A java tool has been developed to check the performance of the algorithm. Sample data randomly generated by the tool are taken as inputs in a data set. Let us assume that the number of resources (virtual machines) generated is 10 and the number of tiers generated is 5.The data set is compared to ten requests and five tiers in virtualized cloud environment. Each tier level has a number of virtual machines in a single tier system and each virtual machine has a specific shared memory from the physical machine. This experiment is compared with MSDT and traditional queuing model which works based on First-Come-First-Serve (FCFS). Table 1 describes the queuing model which is executed in multi-tier environment. This table represents the request that follows the allotment path and corresponding response time and Table 2 represents the MSDT based allocation of requests’ response time and following of the request path.

IJETCAS 15-123; © 2015, IJETCAS All Rights Reserved

Page 46

A. Stanislas et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 43-48

Table 1 Request Allotment of FCFS Queue Model Req. No. 0 1 2 3 4 5 6 7 8 9

Queuing Model Allotment Allotment Path Req.Time (sec) [0-0, 1-0, 2-0, 3-0, 4-0] 0 [0-1, 1-0, 2-1, 3-0, 4-1] 2.876 [0-2, 1-0, 2-0, 3-0] 3.617 [0-3, 1-0] 3.792 [0-0, 1-0, 2-1, 3-0] 3-855 [0-1, 1-0, 2-0, 3-0] 4.473 [0-2] 3.792 [0-3, 1-0, 2-1] 5.098 [0-0, 1-0, 2-0, 3-0] 6.918 [0-1, 1-0, 2-1, 3-0] 7.359

Resp.

Time (sec) 2.876 3.617 3.792 3.855 4.473 5.098 5.925 6.918 7.359 8.016

Table 2 Request Allotment of MSDT Req. No. 6 3 7 2 8 4 5 9 1 0

MSDT Model Allotment Allotment Path Req. Time (sec) [0-0] 0 [0-1, 1-0] 0 [0-3, 1-0, 2-0] 0.063 [0-2, 1-0, 2-1, 3-0] 1.26 [0-0, 1-0, 2-0, 3-0] 1.435 [0-1, 1-0, 2-1, 3-0] 1.875 [0-3, 1-0, 2-0, 3-0] 2.493 [0-2, 1-0, 2-1, 3-0] 3.118 [0-0, 1-0, 2-0, 3-0, 4-1] 3.775 [0-1, 1-0, 2-1, 3-0, 4-2] 4.516

Resp. Time (sec) 0.586 0.063 1.26 1.435 1.875 2.493 2.493 3.775 4.516 5.925

After tracing through the path of each request for its VMs, they are sorted according to the measuring parameters. This sorted order describes the sorted tier number and sorted memory size based on each request metric properties. In MSDT, the requests are sorted among their metric properties such as Depth of Tier and Memory Size. Finalized requests are stored in FSRQ. The requests from the FSRQ are assigned to the scheduler.

The scheduler watches the high priorities memory size in virtual resources that is sent to virtual machine based on high memory size. In this data set, 5 tier clouds have been taken.

Table 3 gives the comparison of response time between MSDT and the existing Queuing Model (FCFS). As it is observed, the queuing model has 8.016 sec. maximum response times to travel in 5 tier clouds while the MSDT takes only 5.925 sec. as the maximum response time. So, this model gives the better performance in comparison with the existing model.

Table 3 Comparison of Response Time of FCFS and MSDT for 10 Requests Req. No 0 1 2 3 4 5 6 7 8 9

Existing Queuing Model FCFS 2.876 3.617 3.792 3.855 4.473 5.098 5.925 6.918 7.359 8.016

Figure 3 Comparisons of Two Models

MSDT 0.586 0.063 1.260 1.435 1.875 2.493 2.493 3.775 4.516 5.925

Figure 3 represents the comparison of two models namely; existing queuing model (FCFS) and MSDT based queuing model. The comparison of response time of QM and MSDT for 10 requests in a 5 tier cloud environment which is in the table 3 is presented in the graph. This graph describes that the MSDT based queuing model has the minimum response time in comparison with the FCFS QM. From the observation of the Table 3 and its corresponding graph, the highest response time of FCFS queuing model is 8.016, whereas, MSDT queuing model ‘s highest response time is 5.925. So, it is evident that MSDT has a better performance.

VI. Conclusion From the study of literature, it is understood that resource provisioning of virtualized multi-tier applications raises new challenges which were not addressed by the previous work for scheduler technique in cloud environment. So, scheduler architecture for virtual machine in cloud data center is presented. A novel priority based scheduling model with an algorithm is proposed which is used to prioritize the requests based on the metrics namely, memory size and depth of tier. This queuing model enhances the scheduler to locate the resources or virtual machines exactly in the specified tier. This process of MSDT reduces the waiting time and response time of the request for a virtualized multi-tier application in the cloud data center. A comparison between the existing model (FCFS) and MSDT is made and the verified results show that the proposed queuing model works with minimum response time and waiting time in comparison with the existing model. Hence, the efficiency and flexibility for resource provisioning were improved in cloud environment. We compared and

IJETCAS 15-123; Š 2015, IJETCAS All Rights Reserved

Page 47

A. Stanislas et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 43-48

showed the distinction of the performance of multi-tier virtualized applications through experiments. Results have shown that under fine-grained resources provisioning, computing resource utilization is optimized.

VII. References [1].

[2].

[3].

[4].

[5].

[6].

[7].

[8].

[9].

[10].

[11].

[12].

[13].

[14].

[15].

[16].

Jim Machi, “Introduction to Cloud Computing”, Dialogic Corporation in White paper, 2010.

Ajay Gulati, Ganesha Shanmuganathan, Anne Holler, “Cloud-Scale Resource Management: Challenges and Techniques”, VMware Technical Journal, USEfix Conference, HotCloud, June 2011.

Jonathan Kupferman, Jeff Silverman, Patricio Jara, Jeff Browne, “Scaling into the Cloud”, CS270 – Advance Operation Systems, University of California, Santa Barbara, 2009.

M. Hasan Jamal, Abdul Qadeer, Waqar Mohmood, Abdul Waheed, Jianxun Jason Ding, “Virtual Machine Scalability on MultiCore Processors based Servers for Cloud Computing Workloads”, IEEE 9th International Conference on Networking, Architecture, and Storage, 2009.

Ming Mao, Jie Li, Marty Humphrey, “Cloud Auto-Scaling with Deadline and Budget Constraints”, Grid Computing (GRID), 11th IEEE/ACM International Conference, 2010.

Rodrigo N. Calheiros, Rajiv Ranjan, and Rajkumar Buyya, “Virtual Machine Provisioning Based on Analytical Performance and QoS in Cloud Computing Environments’, International Conference Parallel Processing (ICPP), 2011.

Jing bi, Zhiliang Zhu, Ruixiong Tian, Oringbo Wang, “Dynamic Provisioning Modelling for Virtualized Multi-Tier Applications in Cloud Data Center”, IEEE 3rd International Conference of Cloud Computing, 2010.

P. Bahram, B. Dragovic, K. Fraser, “Xen and the art of Virtualization”, Proceedings of the 19th ACM Symposium on Cluster Computing and the Grid, pp. 148-155, 2009.

X. Y. Wang, Z. H. DU, Y. N. Chen, “Virtualization based Autonomic Resource Management for Multi-Tier Web Application in Shared Data Center”, The Journal of System and Software, 2008, 81(9), pp. 1591-1608.

Christian Tilgner, “Declarative Scheduling in Highly Scalable Systems”, Extending Database Technology/International Conference on Database Theory (EDBT/ICDT), Lausanne, Switcherland, 2010.

Arshdeep Bahga, Vijay Krishna Madisetti, “Synthetic Workload Generation for Cloud Computing Applications”, Journal of Software Engineering and Applications, 2011.

Waheed Iqbal, Mathew N. Dailey, David Carrera, Paul Janecek, “Adaptive Resource Provisioning for Read Intensive Multi-tier Applications in the Cloud”, Future Generation Computer Systems, Volume 21, Issue 6, pp. 871-879, June 2011.

Qi Zhang, Ludmila Cherkasova, Evgenia Smirini, “A Regression-Based Analytic Model for Dynamic Resource Provisioning of Multi-Tier Applications”, Proceedings of the 4th International Conference on Autonomic Computing, IEEE Computing Society, Washington, USA, 2007 Bhuvan Urgaonkar, Giovanni Pacificy, Prashant Shenoy, Mike Spreitzery, and Asser Tantawi, “An Analytic Model for MultiTier Internet Services and its Applications”, ACM SIGMETRICS, International Conference on Measurement and Modeling of Computer Systems, Volume 33, Issue 1, pp. 291-302, 2005.

B. Urgaonkar, P. Shenoy and A. Chandra, “Agile Dynamic Provisioning of Multi-Tier Internet Applications”, ACM Transaction on autonomous and adaptive System, 3(1) pp. 1-39, 2008.

Hao-Pen CHEN, Shao-Chong LI, “A Queuing Based Model for Performance Management on Cloud”, Advanced Information Management and Service (IMS) 6th International Conference, pp.83-88, 2010.

IJETCAS 15-123; © 2015, IJETCAS All Rights Reserved

Page 48

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Swarm Intelligence Techniques for Optimization in Data Clustering Dipali Kharche1, A.D. Thakare2, C.A. Dhote3 Department of Computer Engineering, Pimpri Chinchwad College of Engineering, Nigdi, Pune, Maharashtra, INDIA

Abstract: Clustering represents the large datasets by a structured well defined number of clusters or prototypes.

K-Means is a useful technique to data clustering which partitions the data into K-Clusters. However, the results of k-means algorithm are based on the selection of initial seeds and converge to local optimum solution. The Swarm Intelligence (SI) is an algorithm to apply many simple agents behaviour which inn terns lead to an emergent global behaviour solution. Data mining tasks require fast and accurate partitioning of large datasets, which may come with a number of attributes or features. The Swarm Intelligence Optimization Techniques like ACO and PSO has successfully been applied to a number of real world clustering problems in order to meet the clustering requirement. This paper surveys the research work in the area of swarm intelligence for solving clustering problems. The new clustering model is proposed using the merits of ACO and PSO to overcome the drawback of k-means clustering algorithm.

Keywords: Swarm Intelligence (SI); Clustering; K-Means; Ant Colony Optimization (ACO); particle Swarm Optimization (PSO); Cluster Analysis

I. Introduction A. Data Clustering Technique Clustering is an important unsupervised learning method that deals with finding a structure in the collection of unlabeled data. A clustering is the process of grouping a set of data objects into multiple groups so that objects within the cluster that has some kind of similarity, but different objects in other clusters [9]. The k-means algorithm highly depends on the initial state and converges to a local optimum solution. Data mining tasks require accurate and fast partitioning of large datasets, which comes with a number of attributes or features. This gives the several computational requirements on the relevant clustering techniques. Clustering can be defined as the division of a dataset into groups of similar kind of objects. Clustering technique is widely used in the application of spatial data processing, satellite photo analysis, financial data classification, and medical figure auto-detection, etc.

Issues in Clustering: 1. Recent clustering techniques do not address all requirements concurrently.

2. Time Complexity is a major issue while dealing with a huge number of dimensions and data items.

3. Effectiveness of clustering techniques depends on the “distance”.

4. Traditional Clustering algorithms have some limitations - initial partition selection and local optima convergence.

B. Swarm Intelligence Swarm intelligence is a collective behavior of self-organized systems, decentralized, artificial or natural. The concept is based on work of genetic algorithms and artificial intelligence. Swarm Intelligence is an artificial Intelligence (AI) technique that gives the study of the collective behavior of a decentralized system made up by a population of the agents interacting locally with each other and with the environment.

A swarm is a distinct set of agents which are able to communicate directly or indirectly with each other, and together approves solving a problem of distributed nature. The swarm body is built on ten thousand nearly identical units such as a bee society. Swarm demonstrates the design of very efficient optimization and clustering algorithms. The skill of the clustering algorithms is based on Swarm Intelligence tools; Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO).

C. Ant Colony Optimization Ant Colony Optimization algorithm [7] is population-based meta-heuristic method that can be used to find approximate solutions to difficult optimization problems introduced by Dorigo is a probabilistic technique and class of optimization algorithm that are useful in problems that generally deals with discovering better paths through graph. The main objective of ACO is to search for a minimum cost path in the entire graph. Artificial

IJETCAS 15-125; © 2015, IJETCAS All Rights Reserved

Page 49

Dipali Kharche et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 49-53

ants are looking for the shortest paths while walking on the graph. Each ant has a simple behavior capable of finding relatively longest paths. Shortest paths are found as a result of global cooperation among ants in the colony.

An ant is a computational simple agent in the ant colony optimization. In each iteration of the algorithm, the ant moves from a state X to state Y, which belongs complete intermediate solution. For ant K, the probability of moving from state X to state Y depends on the combination of two values, viz., the attractiveness of the move, as computed by the priori desirability of that move and the trail level of the move that shows how efficient it has been in the past to make that particular move.

In general, the ant moves from state to state with probability (1) Where is the amount of pheromone deposited for transition from state X to Y, 0 ≤ α is a parameter to control the influence of , is the desirability of state transition XY and β ≥ 1 is a parameter to control the influence of .

And represent the attractiveness and trail level for the other possible state transitions.

If all ants have built their tours, then pheromone level is updated on all the ages as, (2) From Eqn (2), we know that the pheromone is updating attempts to accumulate a greater amount of pheromone to shorter tours. Where the amount of pheromone deposited for a state transition is XY, is the pheromone evaporation coefficient and is the amount of pheromone deposited by ant, (3) Where

is the cost of the

ant's tour (typically length) and

is a constant.

D. Particle Swarm Optimization The Particle Swarms initially introduced for simulating the human social behaviors; it become very popular for efficient search and optimization technique [8]. The Particle Swarm Optimization as it is called now optimized uses mathematical operators and is eventually very simple.

PSO is initialized with a group of random particles (solutions) and then searches for optima by updating generations. In all iteration, each particle is updated by following two “best” values. The first one is the best solution (fitness) it has achieved so far. (The fitness value is also sorted.) This value is called pbest. Another “best” value that is tracked by the particle swarm optimizer is the best value, obtained by some particle in the population. The global best value called gbest, when a particle becomes a part of the population as its neighbors; the local best value is called lbest. After finding the two best values (gbest and lbest), the particle updates velocity and positions with following equations, – (4) (5) Where is the particle velocity, is the current particle (solution). and are defined as stated before. is a random number between (0,1). C1, C2 are constants. Usually c1=c2=2.

II. Related Work The ACO differs from the classical ant system in the sense that the pheromone trails are updated in two ways. i.e. when an ant creates a path they locally change the amount of pheromone on the visited edges by updating amount of pheromone. Another way is after all the ants have built their individual paths, a global updating the pheromone level on the edges that belong to the best ant paths found so far. PSO is also tied to Evolutionary Computation, namely to Genetic Algorithms (GA) and to Evolutionary Programming [1].

A continuous Ant Colony Optimization named ACOR used for solving the continuous optimization problems.

The ACOR with PSO used to improve the searching ability, and gives four kinds of hybridization approach as follows: (a) sequence approach, (b) parallel approach, (c) sequence approach with an enlarged pheromoneparticle table, and (d) global best exchange. These all hybrid systems were applied to data clustering. From all these approach, the sequence approach with the enlarged pheromone table is better approach than other approaches because the enlarged pheromone table maintains the diversity of new solutions of ACOR and PSO, which prevent to traps into the local optimum [2].

Ant Colony Optimization (ACO) algorithm used to improve the k-means. Initially to avoid local minima, a simple and efficient method used to select initial centroids based on mode values of the data vector and k-means

IJETCAS 15-125; © 2015, IJETCAS All Rights Reserved

Page 50

Dipali Kharche et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 49-53

algorithm is applied to cluster the data vectors. An ant colony optimization algorithm is applied to refine the cluster to improve the quality. The basic reason of refinement is a cluster obtained by clustering algorithm never gives 100% quality. It gives a data item can be wrongly clustered. These all kinds of errors can be removed by using refinement algorithm. The quality of the clusters can be analyzed using two measures called entropy and F-measure [3].

A hybrid PSO+K-means document clustering algorithm used to performs fast document clustering and can avoid being trapped in a local optimal solution. The results indicate that the PSO+K-means algorithm can generate the best results in 50 iterations in comparison with the K-means algorithm and the PSO algorithm. In the PSO+K-means algorithm, the globalized searching ability of PSO, and the fast convergence ability of the Kmeans are combined. The PSO algorithm is used at the initial stage to help global search for discovering the optimal solution. The result of PSO is used as the initial center of the K-means algorithm, which is applied for generating and refining the final result. The evaluation of the cluster quality is evaluated by fitness function [4].

To overcome the limitations of k-means, the hybrid evolutionary algorithm is used for solving the clustering problem. The output of the hybrid algorithm is considered as the initial input of k-means. The hybrid evolutionary algorithm is the combination of FAPSO (fuzzy adaptive particle swarm optimization), ACO (ant colony optimization) and k-means algorithms, called FAPSO-ACO–K, which can used to find better cluster partition. The hybrid algorithm for solving the clustering problem which is based on the combination of PSO, ACO and k-means algorithms, The FAPSO–ACO–K algorithm used for converging to the optimal solution in almost runs. The FAPSO–ACO–K clustering algorithm applied when the number of clusters is known a prior.

The quality of cluster is measured by the two criteria: Total mean-square quantization error of a data point to all the centers, as the smaller the sum is, the higher the quality of clustering and F-Measure uses the ideas of precision and recall from information retrieval [5].

Hybrid Particle Swarm Optimization and Subtractive- (PSO) clustering algorithm that performs fast clustering.

The subtractive clustering helps the PSO to start with good initial cluster centroid to converge faster with small fitness function which means a compact result. The Subtractive module runs at the primarily to find initial cluster centroids. The result of the Subtractive module is used as the initial seed of the PSO to get the optimal solution globally. At the same time avoids consuming high computation. The clustering quality is evaluated by fitness function [6].

Table I shows the comparative analysis of different approaches of ACO, PSO and K-means for finding the optimal solution for clustering.

Table I: Comparative Analysis Different approaches of ACO, PSO and K-Means based on research papers

Publication

Elsevier 2013 Cheng-Lung Huang etal [2] JATIT C.Immaculate Mary etal [3] Research Gate Xiaohui Cui etal [4]

Paper Title Hybridization strategies for continuous ant colony optimization and particle swarm optimization applied to data clustering Refinement of Clusters From K-Means with Ant Colony Optimization Document Clustering Analysis Based on Hybrid PSO+K-means Algorithm

Elsevier 2010 Taher Niknam etal [5]

An efficient hybrid approach based on PSO, ACO and k-means for cluster analysis

IJORCS (2013) Mariam ElTarabily etal [6]

A PSO-Based Subtractive data clustering algorithm

PROS

CONS

The hybrid models can preserve diversity When generating new solutions.

Applicable only when the prior number of clusters are known

Clusters provide better results than conventional algorithm

Simple in execution but, the time complexity is more.

Ability of globalized searching and fast convergence.

Number of iterations and computational time requirement is more.

Ability for solving the clustering problem which is based on the combination of PSO, ACO, and k-means algorithms.

Convergence speed is high and fitness value minimum.

Several drawbacks due to its choice of initializations of clusters

The clusters are overlapped.

III. Evaluation of related work The purpose of swarm intelligence (SI) is the plan of smart multi-agent systems by taking motivation from the collective performance of social insects such as bees, ants, termites, particles, birds, and other societies such as fish schools or flocks of birds. SI algorithms are mainly stochastic optimization and search techniques, followed by the principles of self-organization and collective behavior of swarms. They are strong, well-organized and adaptive search methods producing near best possible solutions and have a large amount of indirect parallelism.

IJETCAS 15-125; © 2015, IJETCAS All Rights Reserved

Page 51

Dipali Kharche et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 49-53

On the other hand, data clustering is formulated as a difficult global optimization problem; thereby making the application of SI tools more obvious and appropriate.

Ant colony optimization (ACO) is a recent technique for optimization. The stimulating source of ACO is real ant colonies. ACO algorithm is based on the ants foraging behavior; This behavior is the communication among the ants in terms of pheromone trails which searches short paths between nest and food sources. Ant-based clustering is a distributed process that employs positive response. Ants are designed by simple agents that randomly move in their environment.

Particle Swarm Optimization (PSO) is a population-based search procedure referred to as particles that are grouped into a swarm. Each particle in the flock represents a candidate solution to the optimization problem.

The main strength of PSO is its convergence, which compares with many global optimization algorithms, for applying PSO effectively; main key issues are finding how to plan the problem solution into the PSO particle, which straight affects its performance and feasibility. An important benefit of the PSO is its aptitude to manage with local optima by recombining, comparing and maintaining several solutions. These methods are mainly suitable to perform an investigative analysis. Several investigations are performed in this field – the research nowadays concentrates on improving the performance. The other key features like stability, robustness, speed, convergence would allow us to apply these methods in real applications. Table II gives the general analysis of the optimization techniques which gives the common strengths and weakness of these techniques.

Table II: Analysis of Optimization Techniques.

Techniques

Strengths 1. Gives Better Performance

Ant Colony Optimization (ACO)

Particle Swarm Optimization (PSO)

K-Means

2. Memory of the entire colony is stored.

Weakness 1. Computationally very difficult for a large number of processing nodes.

2. Convergence is guaranteed, but time to convergence is uncertain

3. Poor initial Solution May affect 4. Easy to accommodate to another algorithm

3. Tradeoffs in evaluating Convergence

1. Simple, easy and derivative free algorithm

1. For large search space premature convergence to local optima

2. Few Parameters need to adjust and efficient global search ability 3.They have internal memory and each particle represent individual solution 1. Computationally Efficient 2. Easy to use

2. Weak local search 1. Not Suitable for all data types 2. Outliers and Noise are problem 3. Choice of seeds (Initial Centroids)

IV. Proposed Model for Clustering The Proposed Work integrates an Ant Colony Optimization and Particle Swarm Optimization techniques for data clustering. In order to overcome the drawbacks of existing clustering algorithms, Ant Colony Optimization (ACO) will be used. Ant Colony Optimization (ACO) has the capability to select the initial centroids for clustering in a more effective way.

Figure.1: Proposed System

IJETCAS 15-125; Š 2015, IJETCAS All Rights Reserved

Page 52

Dipali Kharche et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 49-53

On these initial centroids, Particle Swarm Optimization (PSO) will be used with multiple objective functions.

Since Particle Swarm Optimization (PSO) is a global optimization technique, the optimized results are expected by using three functions with Particle Swarm Optimization (PSO).

Figure1 represents the proposed model for Ant Colony Multi Objective Particle Swarm Optimization (ACMOPSO). It represents that; the proposed system will provide more effective results.

V.

Conclusion

Data clustering is well formulated as a difficult global optimization problem. SI algorithms are efficient, adaptive and robust search methods producing near optimal solutions and have a large amount of implicit parallelism which helps in achieving the global optimization. The problems in clustering are very well addressed by various SI algorithms to achieve the optimal clusters. As per comparative analysis, Ant Colony Optimization (ACO) Algorithm and Particle Swarm Optimization (PSO) Algorithm are used for finding the optimal solutions for the clustering problem and can be applied both when the number of clusters is known as well as when this number is unknown.

VI. References [1]

[2]

[3] [4] [5]

[6] [7] [8] [9] [10]

Mr. Pankaj K. Bharne, Mr. V. S. Gulhane, Miss. Shweta K. Yewale, ―Data Clustering Algorithms Based On Swarm Intelligence Electronics Computer Technology (ICECT), 2011 3rd International Conference on (Volume4) E-ISBN: 978-1-42448679-3.

Cheng-Lung Huang, Wen-Chen Huang, Hung-Yi Chang, Yi-Chun Yeh, Cheng-Yi Tsai Hybridization strategies for continuous ant colony optimization and particle swarm Optimization applied to data clustering Applied Soft Computing 13 (2013) 3864– 3872 Taher Niknam, Babak Amiri, ―An efficient hybrid approach based on PSO, ACO and kmeans for cluster analysis Applied Soft Computing 10 (2010) 183–197 C.Immaculate Mary, Dr. S.V.Kasmir Raja, ―Refinement of Clusters from K-Means with Ant Colony Optimization Journal of Theoretical and Applied Information Technology Xiaohui Cui, Thomas E. Potok ―Document Clustering Analysis Based on Hybrid PSO+Kmeans Algorithm Applied Software Engineering Research Group, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 378316085, USA Mariam El-Tarabily, Rehab Abdel-Kader, Mahmoud Marie, Gamal Abdel-Azeem, ―A PSO based subtractive data clustering algorithm International Journal of Research in Computer Science, eISSN 2249-8265 Volume 3 Issue 2 (2013) pp. 1-9 Christian Blum (2005a) Ant colony optimization: Introduction and recent trends, Physics of Life Reviews, 2(4), pp. 353—373, 2005 Christian Blum (2005b) Beam-ACO---hybridizing ant colony optimization with beam search: an application to open shop scheduling, Computers & Operations Research, pages 1565—1591, 2005.

Ajith Abraham, Swagatam Das, and Sandip Roy ―Swarm Intelligence Algorithms for Data Clustering A. Abraham, C. Grosan and V. Ramos (2006) (Eds.), Swarm Intelligence and Data Mining, Studies in Computational Intelligence, Springer Verlag, Germany, pages 270, ISBN: 3-540-34955-3.

IJETCAS 15-125; © 2015, IJETCAS All Rights Reserved

Page 53

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Blind digital watermarking using AES technique for colour images Rahul Saxena1, Nirupma Tiwari 2, Manoj Kumar Ramaiya3 Suresh Gyanvihar University, Jaipur, India 2, 3 Shriram College of Engineering & Management, Gwalior, India1 Abstract: Digital Multimedia contents like text, image, audio and video are widely used and are very easy to transmit over the Internet, and hence the copyright protection has been receiving and increasing attention these days. Other than copyright protection Digital Watermarking is now used for Digital Marketing and promotional Services. A Static Promotional media is now can be made highly productive Dynamic Promotional medium by using Digital watermarking. So for making it more secure, authentic and copyright a Blind digital watermarking using AES technique for colour images is proposed in this paper. This proposed technique enhances the security of transmitted secret information and also minimizes the distortion of the image. Simulations were done in MATLAB. The total PSNR is increased by after applying Advanced Encryption Standard (AES) technique and Noise filters.

Keywords: Digital Watermarking, Blind detection, two-level DCT, AES I.

Introduction

In 1993, Tirkel introduced the term “Digital Watermarking” and presented two watermarking techniques to hide the watermark data in the images [1]. Digital watermarking is a technique which allows attaining the objective of protecting the intellectual property rights by adding patent notices or other verification messages to digital media.

One of the applications of digital watermarking is image endorsement, which is used for authenticating the digital images. Its main objective is to provide a method to validate the image and assure the integrity of the image.

Watermarking is used for Proof of rights/copyrights protection, Data Hiding, Copying Prevention, Broadcast Monitoring etc. Watermark embedding, watermark detection and extraction are basic modules of Watermarking.

Digital watermarking technology has many applications in anti-counterfeit of the digital media, label of the user information, protection, certification, distribution etc. Information hiding is its important study area. Digital watermarks can be classified into two different groups: blind detection [2] (original data is not required) and nonblind detection (original data required) when extracting watermark.

As the Digital media is widely used in the current scenario for transmission over the internet the space required to store the image is a major factor, so here the concept of Image compression comes which means to reduce the amount of data required to represent a digital image. Popularly used techniques for image compression are DCT [3] and DWT [4] which are frequency based techniques and have its’ own advantages and disadvantage. As DWT gives better compression ratio without losing more information of image but it need more processing power.

While in DCT need low processing power but it has blocks artefacts means loss of some information but has most important feature of Energy Concentration.

Motivated by the above discussion, the two-level DCT is introduced to further concentrate the energy and a novel blind watermarking method based on two-level DCT for dual colour image was proposed by QingtangSua, YugangNiub, XianxiLiuc and Tao Yaoa [5]. This paper analyses the key technologies of digital watermarking and explores the application in the digital image copyright protection. Other than copyright protection Digital Watermarking is now used for Digital Marketing and promotional Services. A Static Promotional media is now can be made highly productive Dynamic Promotional medium by using Digital watermarking. So for making it more secure [6], authentic and copyright a blind digital watermarking using AES technique [7] and Noise filters for colour images is proposed in this paper.

The organization of the paper is as follows: Section 2 describes AES technique for the encryption and details of noise filters used. The Proposed Algorithm is shown in Section 3. In Section 4 simulation results are shown and Section 5 concludes the paper.

II.

Proposed Encryption (AES-128)

The proposed work presents an exclusive technique for Blind digital watermarking using AES technique for embedding colour image and text into colour image. AES is based on a design principle known as a substitutionpermutation network, combination of both substitution and permutation. AES has a key size of 128, 192, or 256 bits and fixed size block of 128 bits. The key size used for an AES cipher specifies the number of repetitions of transformation rounds that change the input called the plaintext, into the final output which is said to be the cipher text.

IJETCAS 15-129; © 2015, IJETCAS All Rights Reserved

Page 54

Rahul Saxena et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December-February, 2015, pp. 54-57

This proposed technique enhances the security of transmitted secret information and also minimizes the distortion of the image by using Advance Encryption Standard (AES) where we have used 128 bit block size of plaintext & 128 bits of Secrete key. The pre-processing provide high level of security as extraction of image is not possible without the knowledge of mapping rules of AES and secrete key .

Several different methods are used to eliminate speckle noise, based upon different mathematical models of the phenomenon. One method, for example, employs multiple-look processing (a.k.a. multi-look processing), averaging out the speckle noise by taking several "looks" at a target in a single radar sweep. The average is the incoherent average of the looks. Salt-and-pepper noise is a form of noise presents itself as sparsely occurring white and black pixels on images.

Median filter or a morphological filter is an effective noise reduction method for this type of noise which is a nonlinear digital filtering technique, very widely used in digital image processing because, under certain conditions, it preserves edges while removing noise. Such noise reduction is a typical pre-processing step to improve the results of later processing (for example, edge detection on an image).

III. 1.

2.

3.

4.

Proposed Work

All Watermark preprocessing: - compression & conversion. Encrypt the watermark using AES to make it more secure.

Embedding Watermark in image using two-level DCT. Watermark Extraction: Remove noise form watermarked image then extract watermark to get more accurate results.

Decrypt the watermark using AES to get it back and check for its validity.

Proposed Algorithm A. Watermark preprocessing Step 1: Image decomposition. The original watermark image is alienated into three components red (R), green (G), blue (B) by dimension-reduction treatment.

Step 2: DCT transform. Three component watermarks are divided into 8×8 non-overlapping blocks and each bock is transformed by one-level DCT, respectively.

Step 3: Compress coding in the Zig–Zag order, the sixteen coefficients (the DC coefficient and the first fifteen AC ones) are selected and other coefficients are discarded. Encoding sixteen consecutive pixel values from watermark image one block of 128 Bits is formed. Then, we obtain an encoded 1D binary sequence Let V. Base Step 4: Random permutation to further remove the space correlations among the component watermarks the binary sequence V should be disarranged before embedding. This is done with left or right shifting the binary bits.

Proposed Step 4: Encrypt the watermark using AES to make it more secure. The algorithm is supporting any combination of data and key size of 128/192/256 bits. The proposed method used its 128 bits version allow 128 bits block size and 128 bits key size.

B. Embedding the Watermark Step 1: Transform the image into YIQ colour model.

Step 2: Obtain the two level DCT transformed 4X4 blocks from the Y factor of the images.

Step 3: Select the Standard embedding positions for the watermark and manipulate the intensities accordingly…

C* = where Ak = q×k + T/4, Bk = q×k + 3×T/4, q=C/T

C1

C2

C3

C4

C5 C9

C6 C10

C7 C11

C8 C12

C13

C14

C15

C16

Fig.1: Positions selected for embedding watermark information.

Step 4: Inverse two-level DCT transform.

Step 5: Inverse YIQ colour transform.

C. Extraction of Watermark Base Step 1: Transform the image into YIQ colour model.

Proposed Step 1: Apply Filtration and then Base Step 1.

IJETCAS 15-129; © 2015, IJETCAS All Rights Reserved

Page 55

Rahul Saxena et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December-February, 2015, pp. 54-57

Table I: Filters used for common image processing attack Laplace Resize

Used Speckle Noise remover for noisy image.

Used Salt and pepper Noise remover for resized image.

Used Salt and pepper Noise remover for rotated image.

Used Motion Noise remover for Sharpened image.

Rotate Sharpen

Step 2: Obtain the two level DCT transformed 4X4 blocks from the Y factor of the images Blocks from the Y factor of the images.

Step 3: Apply following rules for Extracting watermark information from the DCT Coefficients.

X* = C** is the DCT coefficient at the low-frequency position of watermarked watermark and x* is the extracted watermark bit. Finally, three segments watermarks x(i)(k) with the same length of original watermark are extracted, where i = {1,2,3}, 1 ≤ k≤ Nw, and x(i)(k) presents the kth binary information of the ith segment.

Obtain the watermark information. According to the majority principle

X’ (k) = Step 5: Decrypt the watermark using AES to get it back.

Step 6: Use the inverse coding and the decompression technology that corresponding to the embedding algorithm to get the watermarks followed by R, G, B components, then reconstruct them to get the final extracted color watermark W’ in RGB color space. Check for the validity of the watermark.

In the proposed work, AES technique is applied on the Novel Blind Digital Watermarking Technique to further enhance the security of the transmitted secret information and also to minimize the distortion of image by using noise filters.

IV.

Simulation and Results

To evaluate the performance of the proposed watermarking algorithm, four standard 24-bit colour images with size of 512 × 512 are selected as original host images in Fig. 2(a) and one 24-bit Colour image with size of 64 × 64 is used as original watermark in Fig. 2(b). Fig.2(c) shows the AES watermarked image. An important thing is to verify the invisibility and robustness of the proposed watermarking scheme.

The peak signal-to-noise ratio (PSNR) is defined in Eq. (1), which may be used to evaluate perceptual distortion of the proposed scheme.

(1) where i = {1, 2, 3}, respectively, denotes the R, G, B components, and PSNRi presents the PSNR value of i component.

where H(x, y, i) and H’(x, y, i) are the pixel values location at (x, y) of i component of the original host image and the watermarked image. Generally, the larger the PSNR value is, the more invisible the watermark is.

(a) Cover Image

(b) Color WM

(c) AES Watermarked Image

Fig. 2: Cover, Color Watermark and AES Watermarked Images

IJETCAS 15-129; © 2015, IJETCAS All Rights Reserved

Page 56

Rahul Saxena et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December-February, 2015, pp. 54-57

Table IIIII: Experimental results of common image processing attack.

Attack

Attacked image

Extracted WM

PSNR (dB)

Blurred

78.166

Sharpened

24.424

Resized

27.541

Rotated

39.975

V.

Conclusions

A Blind digital watermarking using AES technique for colour images has been proposed in this work. The colour watermark image is embedded into a colour host image by modifying the AC coefficients of two-level DCT. In the proposed scheme, the watermark embedding after encryption makes it more secure, extracting processes are very simple and proposed Noise filters Remove noise form watermarked image to get more accurate results.

Therefore, we can conclude that the new proposed technique is more suitable for using colour image information to protect the copyright, best way to detect channel intrusion, Originality of data, Author Genuineness and more Copyright Protected. With the introduction of Digital Marketing and Promotional Services using Digital watermarking the Future Scope for this scheme will be for providing more secure algorithm and approach so as to protect from the Hackers and market competitors.

VI.

References

[1]

R.G. Schyndel, A. Tirkel and C. F. Osborne, A Digital Watermark, Proceedings of IEEE International conference on Image Processing, pp. 86-90, (1994).S. M. Metev and V. P. Veiko, Laser Assisted Microtechnology, 2nd ed., R. M. Osgood, Jr., Ed. Berlin, Germany: Springer-Verlag, 1998.

[2]

X.Y. Luo, D.S. Wang, P. Wang, F.L. Liu, A review on blind detection for image steganography, Signal Process. 88 (2008) 2138–2157.

[3]

Z. Yong, L.L. Cai, L.Q. Shen, J.Z. Tao, A blind watermarking algorithm based on block DCT for dual colour images, in: 2009 Second International Symposium on Electronic Commerce and Security, 2009, pp. 213–217.

[4]

Nirupma Tiwari, Manoj Kumar Ramaiya, Monika Sharma “Digital Watermarking using DWT and DES”2013 3rd IEEE International Advance Computing Conference (IACC).

[5]

QingtangSua, YugangNiub, XianxiLiuc and Tao Yaoa, A novel blind digital watermarking algorithm for embedding colour image into colour image, Optik Elsevier B., Vol. 124, pp.3254– 3259, (2012).

[6]

Metkar, S.P. ; Lichade, M.V., Digital image security improvement by integrating watermarking and encryption technique, Signal Processing, Computing and Control (ISPCC), 2013 IEEE International Conference, pp. 1-6.

[7]

Manoj Ramaiya, Naveen Hemrajani, Anil kishor Saxena “Secured Steganography approach by using AES” International Journal of Computer Science Engineering and Information Technology Research (IJCSEITR) ISSN 2249-6831Vol. 3, Issue 3, Aug 2013, 185192.

VI.

Acknowledgments

First of all, I would like to thank Almighty God for His blessings which helped me in completing my thesis work. . I would like to thank a number of people for their assistance and support during this thesis, especially Shri R. S. Sharma, Honorable Chairman, S.R.G.O.C., Banmore for his love and blessings. I am deeply indebted to gs tomar, for giving me the opportunity to carry out my thesis.

I am fortunate enough that I have got an opportunity to work under the guidance of Prof. Nirupma Tiwari, (Asst. Prof CSE Department), S.R.C.E.M, Banmore. I wish to express my sincere gratitude to her. Her painstaking guidance, inspiring supervision and keen interest, invaluable and tireless devotion, scientific approach and brilliant technological acumen have been a source of tremendous help throughout my research period. I would also like to thank Prof Manoj Ramaiya , HOD-CS, Department, S.R.C.E.M, for his valuable support.

I would like to thank my parents for their love and support. They always believed in me for whatever decision I have made in my life. I would also like to thank my friends for their love and encouragement over the years. Last but not the least, I am thankful to all the members of the CS Department, S.R.C.E.M, for their help and valuable support.

IJETCAS 15-129; © 2015, IJETCAS All Rights Reserved

Page 57

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Mapping and Partitioning of Task Graphs Using Kernighan-Lin/FiducciaMattheyses Algorithm 1, 2, 3

Ashish Mishra1, Raja Jimit2, Abhijit Rameshwar Asati3, Kota Solomon Raju4 Department of Electrical and Electronics Engineering, BITS-Pilani, Pilani, Rajasthan, India 4 Principal Scientist & Project Leader, Reconfigurable Computing Systems & Wireless Sensor Network Systems Lab, Digital System Group, CSIR-CEERI, Pilani, Rajasthan, India

Abstract: Hardware Software partitioning of a task graph refers to the mapping of task nodes to physical components such as processors, Application Specific Integrated Circuits, memory with an optimization of parameters involving execution time, physical area, cost of memory and other factors. This works evaluates the performance of iterative min-cut heuristic Fiduccia-Mattheyses for HW and SW partitioning. The results shows the algorithm can be very effective in giving the bi-partite partition.

Keywords: Fiduccia-Mattheyses(FM), Kernighan-Lin ,task graph, cell and move, Matlab.

I. Introduction The system-level partitioning problem refers to the assignment of operations or task nodes of a task graph to hardware or software components. Overall system performance is determined by the effect of hardware-software partition on the utilization of the processor, number of hardware components, bandwidth of the bus between the processor and application-specific hardware, amongst other related factors. Thus a partitioning scheme must attempt to capture and make use of its effect on system performance in making trade-offs between hardware and software implementations of an operation. Typically, the time constraints of the system won’t be met if it is all implemented as software on a generic processor, but it would be too expensive to design an application-specific hardware chip for the whole functionality. While pure hardware synthesis tools like Hardware Description Languages allow the implementation of systems on a chip from high-level specification, the resulting cost of implementation is a considerable drawback. On the other hand, off-the-shelf generic processors, and implementation in software, are much cheaper, but seldom suit the hard time constraints imposed on embedded systems. Thus, cost-effective designs should use a mixture of hardware and software to accomplish their goals.

The primary issue with a design-oriented approach, where allocation of tasks to hardware and software precedes the synthesis, is that it is not known whether the final system will meet its requirements. This suggests a synthesis-oriented approach, where constraints on performance and cost of the systems are specified, and a systematic constraint-driven exploration of the design space is done. A partitioning algorithm matches every functional object with a system component object and looks for partitions with constraint optimization. These constraints may be in the form of execution time deadlines, area bounds and number of components to use. There are many approaches to solving the partitioning problem like iterative algorithms, dynamic algorithms, heuristic algorithms and exact algorithms. Partitioning is an NP hard problem, and therefore exact solutions tend to be quite slow for bigger dimensions of the problem. Among the well-known heuristic algorithms, FiducciaMattheyses, Genetic Algorithm (GA), Simulated Annealing (SA), Tabu Search Algorithm, Multiway Greedy Algorithms and Ant Colony Optimization (ACO) are most common . Besides the heuristic algorithms referred to above, sometimes family of heuristics such as hierarchical clustering. Kernighan-Lin heuristics are equally useful for application in the partitioning problem. The Integer Linear Programming (ILP) is an exact algorithm subject to the provision that all system constraints can be expressed as a set of linear inequalities containing independent variables [1][2].

II. Fidducia-Mattheyses Algorithm Given a circuit with n elements, we wish to generate a balanced two-way partition of the circuit into sub-circuits of hardware and software. The number of elements in each partition is to be determined dynamically so as to minimize the area i.e., the cost function and a constraint of time are applied so as to ensure good performance.

The definitions used in algorithm are Cut state of a net: A net is said to be cut if it has cells in both blocks, and is uncut otherwise.

Gain of cell: The gain g(i) of a cell ‘i’ is the number of nets by which the cut set would decrease if cell ‘i’ were to be moved.

Cut set of partition: The cut set of a partition is the number of elements of the set of all nets with cut state equal to cut.

IJETCAS 15-130; © 2015, IJETCAS All Rights Reserved

Page 58

Ashish Mishra et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 58-61

Base cell: The cell selected for movement from one block to another is called base cell. It is the cell with maximum gain and the one whose movement will not violate the balance criterion.

Given a partition (A, B) of the cells, the main idea of the algorithm is to move a cell at a time from one block of the partition to the other in an attempt to minimize the cut-set of the final partition. The cell to be moved is called the base cell, is chosen both on the basis of the balance criterion and its effect on the size of the current cut set.

Define the gain g(i) of cell(i) as the number of nets by which the cut set would decrease were cell(i) to be moved from its current block to its complimentary block. Note that a cell's gain may be negative. During each move we must keep in mind the balance criterion to prevent all cells from migrating to one block of the partition.. Thus the balance criterion is used to select the block from which a cell of highest gain is to be moved. It will often be the case that this cell has a non-positive gain. In that case, we still move the cell with the expectation that the move will allow the algorithm to climb out of local minima. After all moves have been made, the best partition encountered during the pass is taken as the output of the pass. To prevent the cell-moving process going into an infinite loop, each base cell is immediately locked in its new block for the remainder of the pass. Thus only free cells are actually allowed to make one move during a pass, until either all cells become locked or the balancing criterion prevents further moves. The best partition encountered during the pass is then returned. Additional passes may then be performed until no further improvements are obtained. In practice this typically occurs quickly, in several passes.

Various steps in the algorithm are: 1. Compute the balance criterion r ∙ area (V) – areamax(V) <= area(A) <= r ∙ area(V) + areamax(V) 2. Compute gains of all cells.

The gain g(i) resulting from the movement of cell i from block A to block B is: g (i) = FS(i) - TE(i).

FS(i) = the number of nets connected to cell i and not connected to any other cell in the block of cell i and TE(i) = the number of nets that are connected to cell i and not crossing the cut.

3. Select base cell that has a maximum gain which satisfies balance criterion. If tie then use Size balance criterion and hierarchy.

4. Lock cell: Fix the base cell Update all cell gains that are connected to critical nets via the base cell.

4. If all cells are fixed, go to Step 5. Otherwise choose next base cell with maximum gain, and move this cell.

Iterate over, going to Step 3 5:.Determine the best move sequence c1, c2, c3... (1<m<i), so that total gain Gm is maximized.

If Gm> 0, go to Step 6, Otherwise exit.

6. Make all i moves permanent. Execute m moves, reset all fixed nodes. Start with a new pass, go to Step 1.

III. Problem Definition The task graph shown in the Fig. 1 is to be partitioned into a hardware software mapping subject with a sequential time constraint of 275 units. The objective of the partition is to yield a minimum area mapping. All the tasks have to be run sequentially on the time line. The Area and Time metrics of the different system components are shown in Table I. A maximum of two components are to be used initially with at least one hardware component and at least one software component Two graphs have been partitioned in this paper and the corresponding results are shown.

Fig. 1 shows the dummy given graph and fig. 2 shows the given architecture for which the mapping has to be carried out. Table 1 shows the parameters given for each node on four given components in the architecture.

Fig.1 Task graph 1

The parameters for a sample input task graph 1 is presented in the table below and the task graph is fig2.

Table1: sample input data Task Time Area A B C D

CPU1

CPU2

ASIC 1

ASIC 2

CPU1

CPU2

ASIC 1

ASIC 2

60 90 81 60

30 50 54 40

20 30 27 20

10 15 15 10

40 40 40 40

60 60 60 60

25 30 15 10

45 35 20 15

IJETCAS 15-130; © 2015, IJETCAS All Rights Reserved

Page 59

Ashish Mishra et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 58-61

E F G H

90 87 90 99

44 30 50 56

30 27 40 33

15 20 15 20

40 40 40 40

60 60 60 60

10 10 15 15

10 25 35 15

Fig. 2 Testing Architecture

IV. Results The results obtained prove that using FM algorithm the resultant output is very much dependent on the initial partition. We have implemented with a variety of initial partitions and some sample results are shown in the tables below: For task graph 1, Initial Partition1 = {ABCD}, {EFGH} Table 2 : Partitions with area and time taken A

B

C

D

E

F

G

H

Area

Time

3

1

1

3

3

3

3

3

125

264

3

3

3

2

3

3

3

3

180

207

4

4

2

4

4

4

2

2

190

185

3

3

4

3

3

3

4

4

155

127

Initial Partition2 = {ABCDEFGH}, {} Table 3 : Partitions with area and time taken A

B

C

D

E

F

G

H

AREA

TIME

3

3

3

3

3

3

2

2

160

186

3

3

3

3

3

3

3

2

175

250

4

4

2

2

4

2

2

2

150

255

4

4

4

2

4

2

2

2

170

201

4

4

4

4

4

2

2

2

185

201

3

3

4

3

3

3

4

4

155

147

Fig. 3 Scheduling of Graph on ASIC-1 and Scheduling of Graph on CPU-1 Since the results varied according to initial cut_set. It was better to give better cut_set with the help of Kernighan-Lin algorithm. Kernighanâ&#x20AC;&#x201C;Lin is a O(n3 ) heuristic algorithm[2] for solving the graph partitioning problem. The algorithm has important applications in the layout of digital circuits and components in VLSI. Graph partitioning problems arise in a variety of circumstances, particularly in computer science, but also in pure and applied mathematics, physics, and of course in the study of networks themselves. KL algorithm can be found in any CAD tools for VLSI design text book .

Fig. 4 Input graph and file format

IJETCAS 15-130; Š 2015, IJETCAS All Rights Reserved

Page 60

Ashish Mishra et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 58-61

Fig.5 Final output form KL algorithm

Fig. 6 Output for the given graph V. Conclusion FM Algorithm was implemented with specific initial cutsets and ratio factors on small task graphs, which give only locally optimum values depending on the provided initial cutsets. The minimum total area value for deadline satisfying partitions found using two initial cutsets was found to be 125 for the partition {BC}, {ADEFGH} using CPU1 and ASIC1. Different initial cutsets may be tried to give a more optimum value of area. Global Heuristic Algorithms have a much better performance than FM and can optimize more than one heuristic, but FM has the added advantage of being simple and less time consuming to get an approximate minimum cost. FM was implemented on directed Task Graphs which have at most one connection with each other. The future work may incorporate hypergraphs too. Also attempts can be made in our work to minimize the time complexity in the implementation of the algorithm.

References [1].

[2].

C.M.Fiduccia,R.M.Mattheyses, A linear time heuristic for network partition,19th Design Automation Conference.

Kernighan, B. W.; Lin, Shen (1970). "An efficient heuristic procedure for partitioning graphs". Bell Systems Technical Journal 49: 291â&#x20AC;&#x201C;307.

IJETCAS 15-130; Š 2015, IJETCAS All Rights Reserved

Page 61

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Ultrasonic and Pyroelectric Sensory Fusion System for Indoor Human/ Robot Localization and Monitoring Azhar K H Embedded System Technologies, Department of Electrical and Electronics Engineering, Nehru Institute of Engineering and Technology, Anna University, Coimbatore - 641105, Tamil Nadu, INDIA Abstract: Indoor localization and monitoring system is a key factor in various emerging fields. Localization aware computing has greater advantages in defence sectors. The technology is used all over in places where security problems are a major issue. Localization and monitoring in a real time remote environment is a strenuous process. Although there are several monitoring systems and approaches, issues and problem regarding their shortcomings remains consistent. The most widely used system includes the use of a pyroelectric sensor. The pyroelectric sensory system alone provides less accurate information of human localization and is restricted when there are multiple targets. This system again posses limited accuracy. In this paper a fusion system module of both ultrasonic and pyroelectric sensors are used. The system is experimentally demonstrated, this methods successfully tracked multiple targets in a real-time environment. This fusion method possesses good accuracy and efficiency. With this accurate localization mechanism for indoor environment, the provision of appropriate services for people can be realized, including medical needs.

Keywords: Ultrasonic sensor, Pyroelectric sensor, Real time system, Multiple targets.

I.

Introduction

Localization is a technique of determining the location of an object, a device or a human being. There are various localization techniques and types of localizations. In this study we are considering robot/human positioning or mapping. Many methods and technologies are used for these purposes. Identifying the position of an object or a person is an inevitable part in various security fields. For the outdoor purposes we have got many methods and technologies. RADAR, Tsunami buoys are most prominent among those categories. But these methods are not suitable in an indoor environment. And the technologies used for location tracking in an indoor environment are not efficient. The main problem faced in this section is about multiple detection and real-time tracking. In this paper the problems regarding these issues are solved.

Figure 1: A real-time tracking demo.

II.

Literature Review

Navigation and position detection works have been carried out by various criteria. Harter et al. [1] described a sensor driven platform that locates the mobile users, as they move around a building. Here the user will have to carry a tag, and the platform builds a dynamic model of environment. With the help of that tag the location can be plotted using software accurately. Rabb et al. [3] explained that, 3-axis generation and sensing of quasi static magnetic dipole provides information about relative positions based on the sensors. Kim et al. [5] developed an intelligent home service system consisting of a home network, sensors and a computing system that collect residential environmental information. The combined network used a Bayesian classifier based algorithm.

Koyuncu et al. [6] discussed many methods, they used RFID based on RSSI technologies. The problems they faced were depended on the irregular radiation patters due to RSSI. Mehrjerdi et al. [7] proposed a path following algorithm for multiple robots based on fuzzy logic. They employed a path follow and group

IJETCAS 15-132; Š 2015, IJETCAS All Rights Reserved

Page 62

Azhar K H, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 62-66

cooperation behavior. The working of the entire system was limited to kinematics. Ahn et al. [8] developed an algorithm for pedestrian localization using distributed wireless sensor networks. Authors used this method because it was difficult to put on hardware modules over human body and moreover wireless sensor networks suffered noises and signal disturbances. Djugash et al. [9] with the approach of extended kalman filter they proposed a decentralized algorithm for mapping the robots in sensor networks. In this method each node will give the information about its adjacent neighbor nodes. It is simply a message passing algorithm and it has its disadvantages. Krumm et al. [10] developed a personal tracking system with the use of two set of color stereo cameras. The stereo images are used for locating people, and the color images are used for maintaining their identities. This system is efficient but it cannot be used in places where cameras are not allowed because of security issues. Shrivastava et al. [11] with the help of spatial resolution techniques they studied about tracking an object in a 2D space and they researched about its limitations. They used a geometric and probabilistic analysis, but for non-ideal sensors exhibiting sensing errors, the geometric algorithm can yield poor performance. Bahl et al. [12] developed a radio frequency based system for indoor positioning; it employed multiple base stations for recording and processing of signals. It combined signal propagation and empirical measurements thus enabling for location aware services. Bingbing et al. [13] proposed a locating algorithm based on RFID technology This method is based on the characteristics of the break-through points, virtual nodes regional divisions and their judgment strategy.

Table 1: Comparison of indoor positioning system methods.

Technology

Sensor Used

RADAR Active badge Active bat Sonitor Motion star Optotrak

RFID IR Ultrasonic Sound Magnetic Light

Performance and Accuracy Moderate Good Moderate Moderate Good Good

III.

Definite Security

Cost

Moderate Limited Limited No Moderate No

Low Low High Low High High

Proposed System

In this project a fusion system is created using PIR and ultrasonic sensor. A number of such fusions can be created, here we are considering only 4 fusion systems. And they are properly arranged and installed in an area.

Ultrasonic sensors work on a principle similar to radar or sonar, which evaluate target by interpreting the echoes from sound waves. Active ultrasonic sensors generate high frequency sound waves and evaluate the sound which is received back by the sensor, by measuring the time interval between sending and receiving the signal it is possible to determine the distance of an object. Passive ultrasonic sensors are normal microphones that detect ultrasonic noise that is present under certain conditions.

A passive infrared sensor (PIR sensor) uses pyroelectricity. All objects with a temperature above absolute zero emit heat energy in the form of radiation. Usually this radiation is invisible to the human eye because it radiates at infrared wavelengths, but it can be detected by electronic devices. The term passive refers to the fact that PIR devices do not generate or radiate any energy for detection purposes. PIR sensors don't detect or measure heat, instead they detect the infrared radiation emitted or reflected from an object. A PIR-based motion detector is used to sense movement of people, animals, or other objects. .

Figure 2: Block diagram.

From the block diagram we can see that four sections are created and these sections can be placed in various ends.

This fusion of PIR and ultrasonic sensor is created and they are made as a module and the sections can be named as PSF sections and it can be fixed in the workspace. These sections will be connected to a microcontroller ATMEGA328, and it will be wired properly with other circuit components. Now this region will be studied, we

IJETCAS 15-132; Š 2015, IJETCAS All Rights Reserved

Page 63

Azhar K H, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 62-66

know that our workspace is equipped with these sensors and when any robot or device enters in that area it will be identified by the section which is in closer proximity with the robot. The source code for the system is written with the help of arduino software in embedded c language. It has a built in terminal window to show the running result. According to the architecture build in the workspace, naming will be given for all the devices.

RF Transceiver serves the purpose of transmitting and receiving the signals in the system. It is concentrated on the hardware side.

The hex file will be generated. As we know we are using microcontroller ATMEGA328, this hex file will be added to the microcontroller in the proteus workspace.

Combined modules of ultrasonic and PIR sensor will be placed on the left, bottom, top and right positions of the room or the indoor environment. Each ultrasonic and PIR on each section run a combined algorithm to plot the positions of the moving object. So the system will be programmed in such a way that any detection will be plotted by the corresponding fusion section. If the robot is in the vicinity of the left and top section, those sections will locate its position. When it moves to the vicinity of other section say bottom, then the bottom section will plot the location. Thus by considering the values obtained from each section we can exactly identify the position if the robot. In default all PIR will be set to one. When detection occurs it will be changed.

PIRL = 1 PIRB = 1 PIRT = 1 PIRR = 1 PIR show only binary values, but Ultrasonic sensors provide distance in units.

This indicates that there is no detection in any of the PIR. And if there is any detection it will be toggled. That is determined while writing the program. In the software environment we can see the change of values in real-time.

PIRL = 0 PIRB = 1 PIRT = 0 PIRR = 1 This indicates that PIR on the left and top have detected a robot or object. It is an ON condition and indicates that there is detection in the system. After it has been done, the ultrasonic will be sensing and plotting the distance of the corresponding robot. As the robot/human moves along the path we can see the real-time change of values in the terminal. In such a way we can find out the exact position of the device. The terminal will be running realtime and it will be showing the corresponding change in the distance values.

Basic code void setup() { Serial.begin(9600); // Set Baud rate for communication pinMode(pirL,INPUT); pinMode(pirB,INPUT); pinMode(pirR,INPUT); pinMode(pirT,INPUT); } void ultrasonicreadfn()//Function TO read Ultrasonic module { // read the analog in value: ultrasensevalueL = analogRead(ultrasonicL); ultrasensevalueB = analogRead(ultrasonicB); ultrasensevalueR = analogRead(ultrasonicR); ultrasensevalueT = analogRead(ultrasonicT); } void ultrasonicmappfn() // Function to map cp ordinates { ultramapL = map(ultrasensevalueL,0,1023,100,0); ultramapB = map(ultrasensevalueB,0,1023,100,0); ultramapR = map(ultrasensevalueR,0,1023,100,0); ultramapT = map(ultrasensevalueT,0,1023,100,0); } void pirreadfn() // Function to read PIR status { pirstatL = digitalRead(pirL); // read the digital pin status pirstatB = digitalRead(pirB); pirstatR = digitalRead(pirR); pirstatT = digitalRead(pirT); } IV.

Simulation Analysis

Ultrasonic and pyroelectric sensory fusion system for indoor human robot localization and monitoring was implemented in software level using proteus and arduino software. The coding for the implementation is done in embedded c language. The simulation in proteus is done by various tools. Literally a robot entering into a room

IJETCAS 15-132; Š 2015, IJETCAS All Rights Reserved

Page 64

Azhar K H, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 62-66

is considered as a page, and this page is divided into four quadrants, and each quadrant is separated. If any robot or devices enter in any region relating to any particular quadrant, the written algorithm will plot the position and distance of that robot in reference to that particular quadrant. Similarly if multiple robots enter into a room it will be identified and plotted in the same way. It is possible for a robot to be in the location of two sections, it can be implemented in hardware level. The proximity or the vicinity of these multiple robots can be easily identified and plotted. The real time movement of such a robot is implemented and its co-ordinates are calculated in an efficient way.

Figure 3: Detection on top left coordinate.

Here we can see that on the terminal it is written movement on top left coordinate.

U – Ultrasonic sensor, PIR – Pyroelectric sensor, L – Left, R – Right, T – Top, B – Bottom.

UL = 27 denotes that the distance is 27 units from the left ultrasonic sensor.

PIRL = 0 denotes that the pyroelectric infrared sensor is ON, it indicates there is a detection.

PIRT = 1 denotes that the pyroelectric infrared sensor is OFF, it indicates there is no detection.

Figure 4: Real-time detection of movement.

In figure UL = 27 and UR = 37.

This implies that there is detection in the upper left quadrant at a distance of 27 units and there is second detection in upper right quadrant at a distance of 37 units. So we have now solved the problem of multiple detection in the system. Now we can see as time moves on first robot is idle and the second robot is moving.

This is happening in the real time environment.

IJETCAS 15-132; © 2015, IJETCAS All Rights Reserved

Page 65

Azhar K H, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 62-66

Now UL = 27 and UR = 41.

In the next moment it again changes and distance is again plotted.

Now UR = 43.

Now we can see that the complete problems of the robotic position detections are solved. This implemented multiple detection and its real-time tracking.

V.

Conclusion

In this paper we found out the solution for the problems faced by many localization techniques. The confusions regarding real time and multiple tracking of humans and robot in an indoor environment are efficiently solved.

The shortcomings and the limited possibilities are identified. This paper introduced a suitable fusion system using ultrasonic and pyroelectric sensor modules. This technique provide high accuracy than normal PIR system or other localization technique, moreover the error bound is also reduced. The importance of this method is that it consumes only less power and moreover it has got high consistency. Since this method is cost efficient it can be implemented in any fields. Applications of this method are in security, defense areas, medical fields, navigation purposes etc.

VI.

References

[1]

A. Harter, A. Hopper, P. Steggles, A.Ward, and P.Webster, “The anatomy of a context-aware application,” in Proc. 5th Annu.

ACM/IEEE Int. Conf. Mobile Comput. Netw., 1999, pp. 59–68.

[2]

A. J. Coulson, A. G. Williamson, and R. G. Vaughan, “A statistical basis for lognormal shadowing effects in multipath fading channels,” IEEE Trans. Vehicle Tech., vol. 46, no. 4, pp. 494–502, Apr. 1998.

[3]

F. Raab, E. Blood, O. Steiner, and H. Jones, “Magnetic position and orientation tracking system,” IEEE Trans. Aerosp. Electron.

Syst, vol. 15, no. 5, pp. 709–717, Sep. 1979.

[4]

G. Durgin, T. S. Rappaport, and H. Xu, “Measurements and models for radio path loss and penetration loss in and around homes and trees at 5.85 GHz,” IEEE J. Sel. Areas Commun., vol. 46, no. 11, pp. 1484–1496, Nov. 1998.

[5]

H. H. Kim, K. N. Ha, S. Lee, and K. C. Lee, “Resident location recognition algorithm using a Bayesian classifier in the PIR sensor-based indoor location-aware system,” IEEE Trans. Syst., Man, Cybern. C, Appl. Rev. Archive, vol. 39, no. 2, pp. 240– 245, Mar. 2009.

[6]

H. Koyuncu and S. H. Yan, “A survey of indoor positioning and object locating systems,” IJCSNS Int. J. Comput. Sci. Netw.

Security, vol. 10,no. 5, pp. 121–128, May 2010.

[7]

H.Mehrjerdi,M. Saad, and J. Ghommam, “Hierarchical fuzzy cooperative control and path following for a team of mobile robots,” IEEE/ASME Trans. Mechatronics, vol. 16, no. 5, pp. 907–917, Oct. 2011.

[8]

H. S. Ahn and K. H. Ko, “Simple pedestrian localization algorithms based on distributed wireless sensor networks,” IEEE Trans.

Ind. Electron., vol. 56, no. 10, pp. 4296–4302, Oct. 2009.

[9]

J.Djugash, Sanjiv Singh and Benjamin Grocholsky, “Decentralized mapping of robot-Aided Sensor Networks” In National Science Foundation under Grant No.IIS-0426945.

[10]

J. Krumm, S. Harris, B. Myers, B. Brummit, M. Hale, and S. Shafer, “Multi-camera multi-person tracking for easy living,” presented at the 3rd IEEE Workshop on Visual Surveillance, Dublin, Ireland, 2000.

[11]

N. Shrivastava, R. Mudumbai, U. Madhow, and S. Suri, “Target tracking with binary proximity sensors,” ACM Trans. Sensor Netw., vol. 5, no. 4,art. no. 30, pp. 1–33, 2009.

[12]

P. Bahl and V. Padmanabhan, “RADAR: An in-building RF-based userlocation and tracking system,” in Proc. IEEE INFOCOM, 2000, pp. 775–784.

[13]

Yan Bingbing, Ren Wenbo, Yin Boli and Li Yang, “An indoor positioning algorithm and its experiment research based on RFID” in International journal on smart sensing and intelligent stsrems vol.7, NO.2.

IJETCAS 15-132; © 2015, IJETCAS All Rights Reserved

Page 66

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Rough Set Techniques for Text Classification and Sentiment Analysis in Social Media G. K. Panda1, Jayanta Mondal2 Department of Computer Science and Engineering MITS Engineering College Sriram Vihar, Bhujabala, Kolonara, Rayagada, Odisha-765017 INDIA Abstract: Sentiment Analysis (SA) is an ongoing research in the field of text mining and classification. SA finds a computational domain from opinions and subjectivity of text data in online social media. Sentiments are inherited in the form of simple lexicons with symbols and texts having noise of irregular texts in complex forms.

It is also seen that the high dimensional growth of lexical blends used by online users while expressing or responding their responses. These blends differ according to demographics and on the context of topics. The simplest approach to get rid of the noise data, adapted by number of studies is by simply removing the irregular lexicons, stopwords, emoticons and lexical blends. This paper investigates such effectiveness in sentiment classification. We assess the impact of rough set approach in classification of universe and apply to the raw datasets. Our earlier study on covering based approximation of classifications outperforms the general classification of universe. We apply roughest based classification process using MATLAB functions to the raw dataset before data pre-processing. Our results show that precompiled roughest classification has better accuracy and outperforms than some of earlier studies.

Keywords: Sentiment Analysis, Text Classification, Rough Sets, Approximations of Classification, Covering based Approximation of Classifications I. Introduction Social media have become effective and popular platforms for expressing oneâ&#x20AC;&#x2122;s online identity. Social networking sites serve as the de-facto internet portal for billions of users [15]. The popularity of smart phones and quantitative shift of internet users from desktop to mobile devices inferred a wonderful platform to the social media to grow in hours. Social networking sites have become a hub for research community to make a deeper analysis on social engineering. The derived sentiments of users from such social media on different context are becoming useful in recommendation systems including advertising, broadcasting, finance, trade, politics, security and many more. In finance data analytics, the interests and preferences of users are extrapolated with the analysis of the attributes of individuals with their preferences through unambiguous data.

Collaborative filtering extends the analysis in the context of related attribute selections on a domain of preferences of users. However, due to the difficulty of acquiring preference data they are considered especially valuable by the retailers and advertisers that hold them. For instance, companies such as Amazon and Netflix maintain large stores of information on usersâ&#x20AC;&#x2122; product and media preference [10]. These data are derived through user interactions on their respective websites or through customized Apps such as buying a book, watching a movie etc. They can then be used to anticipate the interests of other users whose interests may not be known to create personalized user experiences or to make recommendations.

As text messages express the state of minds of individual; mining such messages form a large population in different context and discovering user preferences through social media is a challenging task. Opinion mining, SA and subjectivity analysis are related fields sharing common goals of developing and applying computational techniques to process collections of opinionated texts or reviews. Other research goals are to generate heuristics or tools that can be used to classify, rank or summarize sentiments toward certain objects, events or topics. Most of the studies in this area are based on binary task of classifying sentiment into positive and negative classes discarding the neutral and other classes. For example, tools used to determine a thumbs up or thumbs down vote for specific movies from their reviews or to predict in-favor or in-worse of certain products or events.

In this paper, we look specifically at Twitter data, called tweets, to perform classification of texts and discover sentiment patterns embedded in the form natural languages. Our approach is to use all types of word set with intact of emoticons and not to discard the other types of word sets (discussed in section III) which are generally discarded because of ambiguity. We used Rough Set based classification techniques in MATLAB to label the dataset and improve the polarity of sentiments.

The rest of the paper is organized as follows. In section two, we discuss some of the related works in the field of text classification and sentiment analysis. Section three presents the methodology that we adapted. In section four, we present classification performances. We conclude and give future directions of research in section five.

IJETCAS 15-134; Š 2015, IJETCAS All Rights Reserved

Page 67

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

II. Related Work A. Text Classification Text classification (text categorization) is considered as the task of assigning predefined classes or categories to free-text documents. It provides conceptual view of document collections and has been remaining in front fort in number applications in the real world. The main task of text classification is how to label texts with a predefined set of categories. It has been applied in areas such as document indexing, document filtering, word sense disambiguation, etc. Early SA can be viewed as an application of text classification [11]. One of the central issues in text classification is how to represent the content of a text in order to facilitate an effective classification. From researches in information retrieval systems, one of the most popular and successful method is to represent a text by the collection of terms appear in it. The problem of text classification finds applications in a wide variety of domains in text mining with a high demand in News filtering and Opinion Mining ([3],[4],[9],[16]).

B. Sentiment Analysis Social media is finding the global popularity through micro blogging services like Twitter, Facebook, YouTube, MySpace, Tumblr etc. These social networking websites have evolved to become a source of varied kind of information. This is due to nature of microblogs on which people post real time messages about their opinions on a variety of topics, discuss current issues, complain and express sentiments for certain products, services, brands and business they come across in daily life. In fact, manufacturing and service companies have started to poll these microblogs to get a sense of general sentiment for their products and study user reactions.

SA has been handled as a natural language processing (NLP) at many levels of granularity. It has been migrating in the form of document level ([13], [23]), sentence level ([6], [8]) and phrase level ([1], [24]) classifications. It is seen that, SA is affected by the noisy nature of tweet data which includes abbreviations, irregular forms of text, stop words and lexical blends. Stopwords generally identified in the forms of prepositions and pronouns (such as been, have, is, being and so forth). Lexical blends are words formed in different forms such as, by combining a prefix of one source word with a suffix of another, or a mixture/absence of source words. Blends can also be formed from proper nouns. Blends are cute, creative formations and vary according to demographic.

Blends are also seen as a common type of new word. Some examples are: complisult (mean to compliment + insult), globesity (mean to global + Obesity), turducken (mean to turkey + dcuk + chicken) and so forth.

A popular procedure to reduce the noise of textual data is to remove stopwords by using pre-compiled stopword lists or more sophisticated methods for dynamic stopword identification. While some works support their removal [2]. Others claim that stopwords indeed carry sentiment information and removing them harms the performance of Twitter sentiment classifiers [7]. In this paper, we look the popular microblog Twitter data sets and build models for classifying tweets by discovering sentiments embedded in it. As discussed in the previous section, one of the key challenges that Twitter SA methods have to confront is the noisy nature of generated data from tweets and to find and summarize overall sentiment.

C. Rough Set and Classificaiton The rough set theory (RST) was introduced by Z. Pawalak in 1982 [14]. RST has been applied in many interesting area of research for solving problems including machine learning, intelligent systems, knowledge discovery, decision analysis, expert systems etc. Recently RST and its extensions have also been used for anonymisation techniques in social networking ([21], [22]). RST is an applied theory for reasoning on data represented as information systems or data sets. The information system used to represent a finite collection of data characterized by a finite set of attributes or features. RST does not need any implicit assumptions.

Classifications of universes play a central role in the study of basic rough set theory ([19], [20]).

A rough set is defined through two crisp sets, namely the lower and upper approximation of a set as, where an arbitrary set XU, it may not be possible to describe ‘X’ precisely in the approximation space aprR =(U, R).

Assume that x, y U and R is the equivalent relationship defined on the universe U. The lower-approximation set of the set X on R is R(X) (e.g., see Eq. 1) where RX is the minimum set of entities that for sure belong to the set X. RX is also referred to as positive region. The upper-approximation set of the X on R, RX (e.g., see Eq. 2) where Ø is an empty set. RX is the minimum set of entities which possibly belong to set X. The boundary set can be defined as BND(X) = RX -RX. If BND(X) is non-empty X is a rough set on R, otherwise X is a crisp (ordinary) set.

RX= Y  U R : Y  X  (1) RX=

Y  U

R:Y

X  

(2)

In recent works, we observed that, it is easier to get information in any quantity but a huge amount of unprocessed information leads to “data disasters”. RST facilitates the feature of attribute reduction, finding the

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 68

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

minimum set of conditional attributes without affecting the decisional attributes. In RST, the set of attributes which is common to all the reductions is referred to as the core set of R, denoted as Core(R).

Let B be a subset of A. The core of B is the set of all indispensable attributes of B. The following (see Eq. 3) is an important property, connecting the core and the reduct.

(3) CORE( B)  RED( B) where RED(B) is the set off all reducts of B, the core is the intersection of all reducts and will include in every reduct. Therefore, the core is an important subset of attributes.

Further, the approximations of sets have been extended to a family of sets. Tripathy and Panda ([19], [20]) introduced the notion of covering based approximations of classifications (CAC) and studied their properties.

CAC extends the notions of basic approximations of classifications introduced by Busse. We outline some of the theorems and corollaries in this context.

X i )*j  U , j = 1, 2, 3, 4.

X j )   then ( * iI Also, the result is true when we consider CB–rough approximation of T–type of F with respect to C. * Corollary1 Let i  N n . Then we have, if ( X i ) j  U then ( X k )*   for j = 1, 2, 3, 4. k i Theorem 2 Let I  N n . Then . * X )   if and only if ( X i )2  U . (i) ( c iI i * iI Theorem1 For any I  Nn , if (

(ii) (iii) (iv)

iI c iI c

* ( X i )3  U if (

* ( X i )1  U if (

iI

iI

X i )*   .

X i )*  



iI c

kI C

( X i )  U if (

iI

and condition(5.2) holds.

X i )  .

* X i ) j   then ( X i )*  U for each i  I c and j = 1, 2, 3, 4. iI  X i )   then ( X i )   U for each i  I c .

Also, if ( iI * (X )  U. Corollary3 For each i  N n , ( X i )*   if and only if k i k j Corollary2 For I  N n , if (

Also, i  N n , ( X i )    if and only if



(Xk )  U.

k i * Corollary4 For all i,1  i  n, ( X i ) j  U if and only if ( X )  . j i j * 

Also, for all i, 1  i  n, ( X i )  U if and only if ( X )  . j i j *

Corollary5 If there exists i  N n such that ( X i )*   then for each k ( i )  N n , ( X k )  U . 

Also, if there exists i  N n such that ( X i )    then for each k ( i )  N n , ( X k )  U . * Corollary6 If for all i  N n , ( X i )*   holds then ( X i ) j  U for all i  N n and j=1, 2, 3. 

Also, if for all i  N n , ( X i )    holds then ( X i )  U for all i  N n .

Microblog data of social media are represented as information system and has the characteristics of incomplete and imprecise. Microblog data sets contain a wide range of explicit and implicit topics or subjects. Thus application of RST (as an extension, CAC) for the sphere of Microblog analysis and in particular, sentimental analysis is promising; as we discussed in the beginning of this section.

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 69

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

D. Sentiment Classificaiton Techniques Most of the recent studies on sentiment classification techniques are based on binary sentiments, that is, positive or negative with the approaches of Machine Learning, Lexicon based or Hybrid. The machine learning approach uses machine learning algorithms (supervised and unsupervised learning) and linguistic features. The supervised methods make use of a large number of labeled training documents. The unsupervised methods are used when it is difficult to find the labeled training documents. The lexicon-based approach relies on a sentiment lexicon.

Sentiment lexicons are a collection of known and precompiled sentiment terms. This approach is further categorized into dictionary-based approach and corpus-based approach which use statistical or semantic methods to discover sentiment polarity. The dictionary-based approach which depends on finding opinion seed words, and then searches the dictionary of their synonyms and antonyms. The corpus based approach begins with a seed list of opinion words and then searches other opinion words in a large corpus to help in obtaining opinion words with context specific orientations. The hybrid approach combines both approaches and is very common with sentiment lexicons playing a key role in the majority of methods. The approaches and the most popular algorithms of Sentiment Classification techniques are outlined in Table I. Table I: Approaches for Sentiment Classification

Machine learning classifiers

Micro-blogging features

Supervised Learning  Decision Tree Classifiers  Linear Classifiers Support Vector Machine Neural Network  Rule-based Classifiers  Probabilistic Classifiers Naïve Bayes Maximum Entropy Unsupervised Learning Hashtags, emoticons, repeated letters

Lexicon-based Semantics

Dictionary-based Corpus-based  Statistical approach(LSA,PMI,S-HAL)  Semantic approach Lexicon-based and NLP

Feature sets

Unigrams, bi-grams, n-grams, part-of-speech tags

Sentiments

Positive, Negative, Neutral

III. Methodology As discussed in previous sections, the abstract outline of our approach is to detect linguistic expressions of user preference from social media, classify the dataset using rough set techniques for improved classification, discover the sentiments embedded in the expressions; predict the polarity of new expressions over the specified subject of interest with machine intelligence and finally compare the accuracy, F-positive, F-negative factors with improved accuracy, improved F-positive and improved F-negative factors.

A. Data Source: Twitter Twitter is a social networking based micro-blogging service that allows users to post quick and short messages (tweets) up to maximum length of 140 characters. Each user in Twitter has a set of other users (followers) who receive their messages; those who (user) follows are called friends. The follow relationship is directed and requires authorization from the followee only when the folowee has elected to make their account protected. The default setting is to allow all tweets to be publicly visible.

Messages posted by tweets looks to be an informal register. Twitter is therefore an excellent source of data to search for classification and analysis because it is expected to contain many expressions that would be unlikely to occur in more formal registers. Furthermore, a very large amount of data is available with Twitter [28] (over 37 billion tweets spanned in 7 years; reporting that as of Decemeber 2014 five hundred million tweets were being sent each day). Figure I shows the median tweet per user per month of last six years.

B. Sentiment Factors and Tweet We find the following factors which reflect on identification of sentiments in the sets of tweets. We try to highlight some basic terms concisely for the researchers in understanding the Twitter platform and interpreting results.

Twitters’ Terms of Service requires that any data shared about tweets is shared only in the form of a tweet_id, preserving privacy of User. In Twitter we can observe multiple screen names (like @Bachchan to @AmitabhBachchan for same user (user-id).

Tweets may replicate according to two observed tweeting behaviors. Firstly, a user re-shares one of her friends’ tweets with her own followers, termed as Retweets. In other words, semantics of “RT @username” and tweet also indicate a retweet. Secondly, a user replies to a tweet authored by one of her friends, termed as Replies. As a result, when calculating retweets, we need to look for both native retweets (Retweets) as well as manual retweets (RTs) and assign labels on the context, which is ignored in most of the classification techniques in literatures [18]. Twitter’s native content (entities) can be categorized into three types. A user shares a Link, termed as “Tweets with URLs”. A user states the topic of the tweet; termed as “hashtags” (such as #fail). A user

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 70

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

mentioning another user in the tweet (such as @Bachchan) termed as “tweets with mentions”. The syntax for specifying both hashtags and mentions was created by the Twitter users themselves.

Users use acronyms, make spelling mistakes, use emoticons and other characters that express special meanings and affects in SA like JK for just kidding, LOL for Laugh Out Loud YW for You’r welcome, Mocial for Mobile and social, Twiple for Twitter and People, Cooooool for Cool etc.

Emoticons are pictorially represented facial expressions with punctuation and letters; they express the user’s mood. Target users of Twitter use the symbol “@” to refer to other users on the microblogs. Users use hashtags to mark topics to increase the visibility of their tweets.

C. Data, Text Features and Classification We used publicly available dataset for our sample space. The Twitter Streaming API [26] allows to easily obtain tweets containing specific keywords as they are tweeted. The API provides with tweets containing all of the keywords in a query, but does not guarantee the order of the keywords in those tweets. We use simple MATLAB program to process the tweets to keep all similar lexicon that matches to the original regular expression depending on the need of the context. We used six data sets for classifications. The OMD represents to the Obama-McCain Debate DataSet, HCR represents the Health care reform DataSet, WAB is the Dialogue Earth Project DataSet and GASP is the Dialogue Earth Project DataSet. Figure II shows the number of tweets and vocabulary size of each dataset.

The first step in our approach is to mine explicit natural language expressions of preference from social media for the purposes of prediction and recommendation. These explicit linguistic expressions of preference are focused to mine from natural language data using regular expression patterns. We used lexicon-based methods to annotate the datasets on identifying tweets containing words or phrases that indicate whether a word is same or similar to the requirements. We then identify candidate words amongst those tweets. We use Stemming to reduce an irregular or misspelled word to its root form. Some of the common exemplary string matching regular expressions are shown in Table II. The second step is concerned on rearrange the data which is also termed as Data Preprocessing or Data Cleaning in some literatures. The common observed procedure in this step is to reduce the noise of textual data on removing stopwords by using pre-compiled stopword lists or more sophisticated methods for dynamic stopword identification.

Figure I: Median Tweets

Figure II: Datasets

Saif et al., [17] observed that, removing pre-compiled (classic) stopwords from the dataset negatively impacts the performance of Twitter sentiment classification. In their study, the Mutual Information Method (MI) stopword generation method obtains the best classification performance in the environment of human data annotation and threshold setup. It has a low impact on both the size of the feature space and the dataset sparsity degree. We adopt MI stopword generation method rather than simply removing singleton words.

The mutual information method [5] is a supervised method that works by computing the mutual information between a given term and a document class (e.g., positive, negative) providing an indication of how much information the term can tell about a given class. Low mutual information suggests that the term has low discrimination power and hence it should be easily removed. According to Xu et al., [25] the mutual information between two random variables representing a term t and a class c is calculated as:

I (T , C )   tT

 p(t, c) log ( p(t ). p(c) ) p(t , c)

cC

Where I(T,C) =The mutual information between T and C; T ={0, 1} and C = {0, 1}. In a document, if a term t occurs then T=1; otherwise T=0. If a document belongs to class c then C=1; otherwise C=0

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 71

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

Table II: Regular Expressions for Text Features Regular expression Coined the (term|word) Jargon for Known in \w+ (terms|speak) Known to \w+as New word Slang(expressio n|phrase) for

Table III: Set of emoticons

Examples of matching strings Coined the term, coined the word Jargon for Known in technical terms, known in computer speak Known to scientists as, known to geeks as New word Slang expression for, slang phrase for

Positive Sentiment

Negative Sentiment

Neutral Sentiment

:) :-) :] :D :P ;) ;-) ;] ;D ;P =) =-) =D amazing, awesome, birthday, bom, congrats, cute, enjoy, excited, glad, gracias, hello, loving, moon, smile, sweet, thx

:( :-( ;( ;-( =( =-( brittany, crying, died, headache, horrible huhu, hurts, murphy, nickjonas, poor, rip, sad, sakit, snow, stomach, sucks, throat, triste, ugh, upset

(_/) bunny ears 0w0 a cat’s face a face with wide eyes

The tweets of datasets are labeled with one of the three classes (positive, negative, other). In Table III, we highlight such labels. Each entity of the tweets is annotated with the same three sentiment classes and a sentiment weight factor (1-3) on the positivity of the class. The annotated tweets are further processed to classify using rough set based classification in MATLAB. This improves the quantification of binary polarity compared to non-classified datasets. The performance of improvements is presented in the next section. Some of the basic functions of RST-classification are listed below: Sl=rslower(A,a,S) % find the lower-approximation set Sl Su=rsupper(A,a,S) % find the upper-approximation set Su Sd=setdiff(Su,Sl) % find the boundary set Sd Function for Lower Approximation

Function for Upper Approximation

function sl=rslower(y,a,X) r_ind=ind(a,X); sl=[]; [p,q]=size(r_ind); for i=1:p sd=setdiff(r_ind(i,:),0); if ismember(sd,y) sl=cat(2,sl,sd); end end

function su=rsupper(y,a,X) r_ind=ind(a,X); su=[]; [p,q]=size(r_ind); for i=1:p sd=setdiff(r_ind(i,:),0); ict=intersect(sd,y); [p_ict,q_ict]=size(ict); if q_ict~=0 su=cat(2,su,sd); end end

Function for Indiscernibility relationship function aa=ind(a,x) [p,q]=size(x); [ap,aq]=size(a); z=1:q; tt=setdiff(z,a); x(:,tt(size(tt,2):-1:1))=-1; for r=q:-1:1 if x(1,r)==-1 x(:,r)=[]; end end for i=1:p v(i)=x(i,:)*10.^(aq-[1:aq]'); end

y=v'; [yy,I]=sort(y); y=[yy I]; [b k l]=unique(yy); y=[l I]; m=max(l); aa=zeros(m,p); for ii=1:m for j=1:p if l(j)==ii aa(ii,j)=I(j) end end end

IV.

Classification Performance

We used Weka 3.6.11 [27] and performed a binary sentiment classification on all the datasets with Naïve Bayes classification. The RST based classified datasets are processed for this classifications. Figure III represents the improved accuracy of classification. In WAB dataset 35% of the total data were not categorized to binary polarity and occupied in other and irrelevant classes which were further labeled and padded to binary polarity using RST classification and hence WAB outperformed. Figure IV shows the improvements of F-measure in positive and negative sentiment polarity compared to non RST classifications. It is observed that, the improvements are marginal. Figure V represents the classification performance of all the test datasets.

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 72

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74

Figure III: Improved Accuracy of Classification

Figure IV: Improved F-Measure of Classification

Figure V: Classification Improvements

V. Conclusion and Future Work In this paper, we studied relevant research applications of text classification and sentiment analysis. Discovery of sentiments from microblog data and classifying them with utmost accuracy is found to be challenging. The Rough Set theory established as a classical tool in solving real life problems pertaining to information systems or data sets. The RST based classifications introduced by Tripathy and Panda has its competency to tackle characteristics like implicit, explicit, incomplete, impreciseness of information system or data sets. As the microblog data sets contain a wide range of explicit and implicit topics or subjects; Rough Set techniques are used in MATLAB to classify six datasets and found the improvement of binary polarity. It is also observed that, presence of stopwords, irregular lexicons, emoticons, lexical blends in test datasets lead to higher accuracy level. Our results show that the improvements of accuracy in HCR and WAB datasets were impressive with marginal improvements in binary polarity. Sentiment analysis in foreign languages (other than English) is growing as there is still lack of resources and researches concerning on foreign languages. Applications of Deep Architecture may be used to infer how fast the tweets can pass in the context (ego-centric) to utmost users irrespective to demographics.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

Agarwal, A., Biadsy, F., and Mckeown K: “Contextual phrase-level polarity analysis using lexical affect scoring and syntactic ngrams”. Proceedings of the 12th Conference of the European Chapter of the ACL, 2009, pg. 24–32.

Asiaee, T. A., Tepper, M., Banerjee, A., and Sapiro, G.: “If you are happy and you know it… tweet. In proceedings of the 21st ACM international conference on Information and knowledge management”, 2012, ACM, pg. 1602-1606.

Carvalho, V. R. de, and Cohen, W.: “On the collective classification of email speech acts”, ACM SIGIR Conference, 2005.

Cohen, W.: “Learning rules that classify e-mail”. AAAI Conference, 1996.

Cover, T.M., and Thomas, J. A.:” Elements of information theory”. John Wiley & Sons, 2012.

Hu, M., and Liu, B.: “Mining and summarizing customer reviews”. KDD, 2004.

Hu, X., Tang, J., Gao, H., and Liu, H.: “Unsupervised sentiment analysis with emotional signals”. In proceedings of the 22nd Intl. Conf. on World Wide Web, 2013, pg.607-618.

Kim, S. M., and Hovy, E.: “Determining the sentiment of opinions”, Coling, 2004.

Lewis, D., and Knowles, K.: “Threading electronic mail: A preliminary study”. Information Processing and Management, 33(2), 1997, pg. 209–217.

Linden, G., Smith, B., and York, J.: “Amazon.com recommendations item-to-item collaborative filtering”, Internet computing, IEEE, 2003.

Maron, M.: Automatic indexing: an experimental inquiry. Journal of the Association for Computing Machinery 8, 3, 1961, pg.

404–417.

Pak, A. and Paroubek, P.: “Twitter as a corpus for sentiment analysis and opinion mining”. In Proceedings of LREC, 2010, Malta.

Pang, B., and Lee, L.: “A sentimental education: Sentiment analysis using subjectivity analysis using subjectivity summarization based on minimum cuts,” 2004, ACL. Pawlak Z.: Rough Sets, Intl. Journal of Information and Computer Science,II. Post, H.: Twitter Statistics. http://bit.ly/18KIwd2, 2014.

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 73

G. K. Panda et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 67-74 [16] [17] [18] [19]

[20]

[21] [22] [23]

[24] [25] [26] [27] [28]

Sahami, M., Dumais, S., Heckerman, D., and Horvitz, E.: “A Bayesian approach to filtering junk e-mail. AAAI Workshop on Learning for Text Categorization”. Tech. Rep. WS-98-05, AAAI Press. http://robotics.stanford. edu/users/ sahami/papers.html Saif, H., He, Y., and Alani, H.: “Semantic sentiment analysis of twitter”. In proceedings of the 11th international conference on The Semantic Web, 2012, MA. Speriosu, M., Sudan, N., Upadhyay, S., and Baldrige, J.: Twitter Polarity Classification with Label Propagation over Lexical Links and the Follower Graph, Tripathy, B. K., and Panda, G. K.: “On Covering Based Approximations of Classifications of Sets”, Proc.22nd Int. conf. of Industrial, Engineering and other applications of Applied Intelligent Systems, IEA/AIE 2009, Tainan, Taiwan, LNAI Series, Springer, 2009, pg.777-786 Tripathy, B.K., and Panda, G. K.: “On Some Properties of Covering based Approximations of Classifications of Sets”, Proceedings of the ACM Intl. Conf. on Interaction Sciences: Information Technology, Culture and Human,ICIS, Korea, Nov (2009), ACM (III), pg. 1227-1232.

Tripathy, B. K., and Panda, G. K.: “A New Approach to Manage Security Against Neighbourhood Attacks in Social Networks”, The Intl. Conf. on ASONAM 2010, University of Southern Denmark, Denmark, 2010, IEEE Computer Society, pg. 264-269.

Tripathy, B. K., Panda, G. K. and Kumaran, K.: “A Fast l-diversity Anonymisation Algorithm”, Proc. of the third Intl. Conf. on Computer Modeling and Simulation-ICCMS 2011, Mumbai, Jan, 2011,(2) pg.648-652.

Turney, P. D.: “Thumbs up or thumbs down? Semantic orientation applied to unsupervised classification of reviews”, In Proceedings of the 40th Annual Meetings of the Association for Computational Linguistics, Philadelphia, Pennsylvania, 2002, pg.417-424.

Wilson, T., Wiebe, J., and Hoffman, P.,: “Recognizing contextual polarity in phrase level sentiment analysis”. 2005, ACL. Xu, Y., Jones, G. J., Li, J., Wang, B., and Sun, C.: “A study on mutual information-based feature selection for text categorization.

Journal of Computational Information Systems”, 2007, 3(3): pg. 1007–1012. http://dev.twitter.com http://sourceforge.net/projects/weka/ http://www.internetlivestats.com/twitter-statistics/

IJETCAS 15-134; © 2015, IJETCAS All Rights Reserved

Page 74

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net OFDM Channel Analysis between FFT and Wavelet Transform Techniques Quosay Jalil1, S Nagakishore Bhavanam2 1 Student, M.Tech., 2Assistant Professor 1,2 University College of Engineering & Technology, Acharya Nagarjuna University, Andhra Pradesh, INDIA. Abstract: Nowadays In 4G (MIMO-OFDM) Communication is Role on World from this communicate channel we are going face problem on Interference between Transmitter Users for this In our project We are implement through Transformation method for OFDM systems of transmits and receive antennas with cyclic prefix (CP) .In OFDM multiple carriers are used and it provides higher level of spectral efficiency as compared to Frequency Division Multiplexing (FDM). In OFDM because of loss of orthogonality between the subcarriers there is inter carrier interference (ICI) and inter symbol interference (ISI) and to overcome this problem use of cyclic prefixing (CP) is required, which uses 20% of available bandwidth. Comparison between the conventional FFT based OFDM systems with DWT based OFDM system have been made according to some conventional and non-conventional modulation methods over AWGN. The wavelet families have been used and compared with FFT based OFDM system and found that DWT based OFDM system is better than FFT based OFDM system with regards to the bit error rate (BER) performance .

Keywords: MIMO-IOFDM, CP, FFT, BER I. Introduction Orthogonal frequency-division multiplexing (OFDM), is a frequency-division multiplexing (FDM) scheme used as a digital multi-carrier modulation method. OFDM meets the LTE requirement for spectrum flexibility and enables cost-efficient solutions for very wide carriers with high peak rates. Each user is allocated a number of socalled resource blocks in the time/frequency grid. The more resource blocks a user gets, and the higher the modulation used in the resource elements, the higher the bit-rate. Which resource blocks and how many the user gets at a given point in time depend on advanced scheduling mechanisms in the frequency and time dimensions.

The interfaces of both OFDM and OFDMA work by separating a single signal into subcarriers, or, in other words, by dividing one extremely fast signal into numerous slow signals that optimize mobile Access, as the sub channels can then transmit data without being subject to the same intensity of multipath distortion faced by single carrier transmission. The numerous subcarriers are then collected at the receiver and recombined to form one high speed transmission.

Fig. 1: The OFDM system block

IJETCAS 15-135; Š 2015, IJETCAS All Rights Reserved

Page 75

Quosay Jalil et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 75-79

The difference between OFDM and OFDMA is that OFDMA has the ability to dynamically assign a subset of those subcarriers to individual users, making this the multi-user version of OFDM, using either Time Division Multiple Access (TDMA) (separate time frames) or Frequency Division Multiple Access (FDMA) (separate channels) for multiple users. OFDMA simultaneously supports multiple users by assigning them specific subchannels for intervals of time. Point-to-point systems are OFDM, and do not support OFDMA. Point-tomultipoint fixed and mobile systems use OFDMA.OFDM technologies typically occupy nomadic, fixed and oneway transmission standards, ranging from TV transmission to Wi-Fi as well as fixed Wi-MAX and newer multicast wireless systems like Qualcomm’s Forward Link Only (FLO). OFDMA, however, adds true mobility to the mix, forming the backbone of many of the emerging technologies including LTE and mobile WiMAX. .

II. FFT Based OFDM OFDM transmitter was described using sinusoidal components. Generally, an OFDM signal can be represented as S (t) = symbols mapped to chosen constellation (BPSK/QPSK/QAM etc..,) F n = orthogonal frequency This equation can be thought of as an IFFT process (Inverse Fast Fourier Transform). The Fourier transform breaks a signal into different frequency bins by multiplying the signal with a series of sinusoids. This essentially translates the signal from time domain to frequency domain. But, we always view IFFT as a conversion process from frequency domain to time domain. FFT is represented by Whereas its dual, IFFT is given by The equation for FFT and IFFT differ by the co-efficient they take and the minus sign. Both equations do the same thing. They multiply the incoming signal with a series of sinusoids and separate them into bins. In fact, FFT and IFFT are dual and behave in a similar way. IFFT and FFT blocks are interchangeable. Since the OFDM signal (c (t) in the equation above) is in time domain, IFFT is the appropriate choice to use in the transmitter, which can be thought of as converting frequency domain samples to time domain samples. Well, you might ask: s (t) is not in frequency domain and they are already in time domain; so whats the need to convert it into time domain again? The answer is IFFT/FFT equation comes handy in implementing the conversion process and we can eliminate the individual sinusoidal multipliers required in the transmitter/receiver side. The following figure illustrates how the use of IFFT in the transmitter eliminates the need for separate sinusoidal converters. Always remember that IFFT and FFT blocks in the transmitter are interchangeable as long as their duals are used in receiver.

III. DWT Based OFDM System The wavelet transform is usually represented as MRA. The wavelet transform decomposes the signal using a set of basis function into different resolution subspaces …..V-2< V-1 < V0 < V1 <……The decomposition is done using a basis function and a wavelet function and there translation and dilation. The dilated and translated scaling function forms the basis of the various subspaces. i.e. {ø (t)} forms a basis for V0 . The wavelet functions forms a subspace orthogonal to the basis formed by the scaling function. The scaling and the wavelet function both satisfy some dilation equation. ø (t) = ∑ ø(2t-n)h(n) If ø (t) should be orthonormal to its translated then h[n] should satisfy the orthonormality condition ∑h[n]h[n-2m] = δ[m] and ∑(-1)n h[n] =0 Given a sequence we can find another sequence g[n] such that the function satisfying the Dilation equation Ψ (t) = ∑ Ψ (2t-n) g[n] This function is orthonormal to the scaling function is called the wavelet function. Using the wavelet and the scaling function we decompose the signal into two subspaces orthogonal to each other. Thus if the original signal is in space V0 then using the scaling and wavelet function we decompose it into subspaces V1 and W1 .

In the classical wavelet transform the subspace V1 is further decomposed into orthogonal subspaces V2 and W2.We see that ø (t) occupies only half the frequency space of ø (2t) and similarly for the wavelet function.

Thus this decomposition can be considered as decomposition into high and low frequency domain. In discrete wavelet transform we can represent the process of decomposition as low and high pass filtering and then down sampling by 2. The filter coefficients are given by g[-n] and h[-n]. The filter with coefficients h[-n] forms a low pass filter while the filter with g[-n] forms a high pass filter. Thus the wavelet transform can be constructed by using QMF filter banks. The low passed and the high passed signals are down sampled by 2. The low pass signal can again be decomposed into high and low pass signals.

IJETCAS 15-135; © 2015, IJETCAS All Rights Reserved

Page 76

Quosay Jalil et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 75-79

(A) Modulation In modulation, a message signal, which contains the information, is used to control the parameters of a carrier signal, so as to impress the information onto the carrier. analogue – denoted by m(t) digital – denoted by d(t) – i.e. sequences of 1's and 0's The message signal could also be a multilevel signal, rather than binary; this is not considered further at this stage.

Fig. 2: The tree structure of the Wavelet decomposition

Fig. 3: A Structure of wavelet decomposition

The synthesis side can be considered as up sampling by 2 and then filtering the low and high pass coefficient at the k-th level and then adding the two and this gives the low pass coefficient at the (k-1) level and similar structure at the subsequent levels give back the signals. The filter of the synthesis side can be determined from the analysis side filter by the perfect reconstruction condition. This gives a variety of filter’s and this leads to the various families of wavelets. A 1st level decomposition and reconstruction is shown below for QMF filters.

Fig. 4: The 2 channel QMF

In wavelet transform we do the decomposition of just the low pass coefficients. A generalization of this is the wavelet packet transform in which the decomposition is done along both the high and low pass coefficients.

Also we make use of the identities that when a signal is down sampled and then passed through a filter (H (z)) and it is equivalent to passing the signal through a filter (H(z2 )) and then down sampling. Thus the wavelet packet transform can be represented as below. The frequency time plot of Wavelet transform and wavelet packet transform is shown below. We see that at low frequencies the time-span is larger while at high frequencies the time-span is smaller. For the wavelet packet we can decide how to decompose the high and low frequencies parts as after each decomposition we can decide whether to decompose the signal in the low/high frequency domain or not.

Fig. 5: The transmitter and receiver filter DWT

The OFDM implemented by using IFFT’s and FFT’s have some problems. The OFDM suffers from ISI (inter symbol interference) –This is usually taken care of by using a adding a cyclic prefix greater than the channel length but this may not always be possible. This occurs due to loss of orthogonality due to channel effects. Time

IJETCAS 15-135; © 2015, IJETCAS All Rights Reserved

Page 77

Quosay Jalil et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 75-79

and Frequency Synchronization- the OFDM requires time and frequency synchronization to get a low bit error rate. Carrier Frequency Offset- The offset between the carrier frequency and the frequency of the local oscillator also causes a large bit error rate.

Due to these problems we need to look at other type of modulation to generate the carrier. One of these is the wavelet transform. The wavelet transform is proposed by many authors, it has a higher degree of side lobe suppression and the loss of orthogonality leads to lesser ISI and ICI. In Wavelet OFDM the FFT and IFFT is replace by DWT and IDWT respectively. For the Wavelet transform we see that from the time-frequency plot that the basic Wavelet transform offers lesser flexibility than the wavelet packet transform. For the wavelet packet transform we can construct an algorithm to do the decomposition such that the effect due to the noise (assuming that we know the frequency that is affected most by the noise and the time when it affected most).The transmitter and receiver are shown in the figure below. At the transmitter the data is first M-array modulated and then serial to parallel converted first then up sampled and then passed through an IDWPT filter bank. At the receiver the data is passed through an Analysis Filter bank and then parallel to serial converted and then M-array demodulated.

This is the case for wavelet packet transform; some researchers have also used wavelet transform. In such a case the IFFT and FFT’s in the OFDM are just replaced by the DWT/IDWT to give the DWT-OFDM The wavelet can be implemented using QMF filter banks’ but this cannot full fill linear phase filtering, therefore some authors have suggested the use of bi-orthogonal filter banks, bi-orthogonal filters provide linear phase filtering and the design is flexible. The design of the transmitter and receiver based on bi-orthogonal filter banks is similar to the above design with the filters being the bi-orthogonal filters.

Table 1: Modulation and symbol Rate relationship

IV. Conclusions We see that DWT-OFDM performs much better than the DFT-OFDM over AWGN and Rayleigh channel with low SNR. Also we find that use of Wavelet’s reduces the overhead thus giving a larger bandwidth. The wavelet packet is much better than the implementing just the wavelet transforms as it is more flexible. The bi-orthogonal wavelets though may provide some advantage and better flexibility doesn’t perform well for some wavelets considered to the theoretical case.

V. References [1] [2] [3] [4] [5] [6]

[7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

Communication Systems, 4th edition, Simon Haykin, John Wiley and Sons,Inc. C. V. Bouwel, et. al, Wavelet Packet Based Multicarrier Modulation, IEEECommunications and Vehicular Technology, SCVT 200, pp. 131-138, 2000.

B. G. Nagesh, H. Nikookar, Wavelet Based OFDM for Wireless Channels, IEEEVehicular Technology Conference, Vol. 1, pp.

688-691, 2001.

LI Wiehua, et.al, Bi-orthogonal Wavelet Packet based Multicarrier modulation.

Haixia Zhang, et. al, Research of DFT-OFDM and DWT-OFDM on DifferentTransmission Scenarios. Proceeding of the second international conference on Information Technology for Application (ICITA 2004) D.Karamehmedovic, M.K.Lakshmanan and H. Nikookar,”Performance of Wavelet packet modulation and OFDM in presence of carrier frequency and phase noise” Proceedings of the 1stEuropean Wireless Technology Conference, October 2008,Amsterdam, The Netherlands.

R.S. Bodhe, S.S. Narkhede, Shirish Joshi, “Design and Implementation of Baseband Processing for Wavelet OFDM”, National Conference e-PGCON2012, Pune.

Andre Ken Lee Ooi, MichealDrieberg and VarunJeoti, “DWT based FFT in practical OFDM systems” IEEE Region Ten Conference, TENCON2006, 14-17Nov2006, Hongkong.

S. Adhikari, S. L. Jansen, M. Kuschnerov, B. Inan, and W. Rosenkranz, “Analysis of spectrally shaped DFT-OFDM for fiber nonlinearity mitigation,” Opt. Express, to be published.

S. Nakajima, “Effects of spectral shaping on OFDM transmission performance in nonlinear channels,” in Proc. 16th ISTMWC 2007, Budapest, Hungary, Jul., pp. 1–5.

O. Gaete, L. Coelho, B. Spinnler, and N. Hanik, “Pulse shaping using the discrete Fourier transform for direct detection optical systems,” in Proc. ICTON, Stockholm, Sweden, Jun. 2011, pp. 1–4, paper We.A1.2.

C. Xia and D. van den Borne, “Impact of the channel count on the nonlinear tolerance in coherently-detected POLMUX-QPSK modulation,” in Proc. OFC 2011, Los Angeles, CA, Mar. 2011, pp. 1–3, paper OWO1.1.

Theodores Rappaport, Wireless communication principles and practice, Pearson education, 2nd edition.

Dr. KamiloFeher,Wireless digital communications modulation and spread spectrum applications, Pretice-Hall, 2ndedition.

DharmaprakashAgrawal and Qing Anzeng, Introduction to wireless and mobile system, Vikas publishing house.

JayRanade, series advisor, wireless and networked communication, Mcgraw hill Inc Mitra, Sanjit K. Digital Signal Processing: A Computer-Based Approach. NewYork: McGraw-Hill, 2001.

IJETCAS 15-135; © 2015, IJETCAS All Rights Reserved

Page 78

Quosay Jalil et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 75-79 [18] [19] [20] [21] [22] [23] [24] [25] [26]

M. Guillaud and D. Slock, “Full-rate full-diversity space-frequency coding for MIMO OFDM systems,” in Proc. IEEE Benelux Signal Processing Symp., 2002, pp. S02–14.

Y. (G.) Li, J. H. Winters, and N. R. Sollenberger, “MIMO-OFDM for wireless communications: Signal detection with enhanced channel estimation,” IEEE Trans. Commun., vol. 50, pp. 1471–1477, Sept. 2002.

S. Kaiser, “Spatial transmit diversity techniques for broadband OFDM systems,” in Proc. IEEE Globecom, 2000, pp. 1824–1828.

A. Dammann and S. Kaiser, “Standard conformable antenna diversity techniques for OFDM and its application to the DVB-T system,” in Proc. IEEE Globecom, 2001, pp. 3100–3105.

A. Dammann, P. Lusina, and M. Bossert, “On the equivalence of space–time block coding with multipath propagation and/or cyclic delay diversity in OFDM,” in Proc. Eur. Wireless 2002, pp. 847–851.

J. Tan and G. L. Stüber, “Multicarrier delay diversity modulation for MIMO systems,” IEEE Trans. Wireless Commun., to be published.

A. N. Mody and G. L. Stuber, “Efficient training and synchronization sequence structures for MIMO OFDM systems,” presented at the 6th OFDM Workshop 2001, Hamburg, Germany “Synchronization for MIMO OFDM systems,” in Proc. Globecom 2001, pp. 509–513.

V. Tarokh, H. Jafarkhani, and A. R. Calderbank, “Space–time block coding for wireless communications: Performance results,” IEEE J. Select. Areas Commun., vol. 17, pp. 451–460, Mar. 1999.

IJETCAS 15-135; © 2015, IJETCAS All Rights Reserved

Page 79

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Bianchi Type IX Cosmological Model with Varying Lambda Term R.K. Tiwari*, Devendra Kumar Tiwari *, C.Chauhan** *Dept. of Mathematics, Govt. Model Science College, Rewa-486001 (M.P.), India **Dept. of Mathematics, I.E.T. DAVV Indore (M.P.), India __________________________________________________________________________________________ Abstract: We have investigated Bianchi type IX space time in the presence of cosmic string in Einstein's general theory of relativity. To get a deterministic model, we have assumed the condition A = Bn (n is a constant),

  aR 2 (a is constant, R is scale factor) and     0. Physical and Kinematical behaviors are

discussed. It can be seen that the values of cosmological parameters resemble is in agreement with recent observations.

model and so the model

Keywords: Cosmic strings, Bianchi type IX Model, Variable Λ-term, Anisotropy. __________________________________________________________________________________________ I. INTRODUCTION The cosmological constant problem is a difficult and fascinating problem for cosmologists and quantum field theory researchers. Observations of the Cosmic microwave background radiations (CMBR) indicate that we live in a spatially flat universe with total energy density equal to the critical density. On the other hand, gravitational measurements of matter density in the galaxies lead to an average density of matter as the cosmological scale approximately equal to one third of the critical density. Observations suggest that the cosmological constant representing the density of vacuum is variable dynamic degree of freedom being initially very large came down to its small present value in an expanding universe. In this paper Bianchi type IX cosmological model allow not only expansion but also rotation and shear and in general are anisotropic. Many researchers have taken keen interest in studying Bianchi Type IX universes, because familiar solutions like Robertson Walker universe with positive curvature, the de-Sitter universe, the taub-Nat solutions etc. are particular case of Bianchi Type IX space time. In these models, neutrino viscosity does not guarantee isotropy at the present epoch.

Kibble [1] has investigated that strings are one of the sources of density perturbation, which are required for the formation of large scale structure of the universe. These strings possess stress energy and are coupled to the gravitational field. Letelier [2] and Satchel [3] have initiated the general relativistic treatment of the strings.

Consequently, many relativists have obtained exact solution which described homogeneous string cosmological models with different Bianchi symmetries Bali and Anjali [4] have investigated Bianchi type-I magnetized string cosmological model in general relativity. Rao et al. [5] studied exact Bianchi type II, VIII and IX string cosmological models in Saez-Ballester theory of gravitation. Singh [6] also studied string cosmology with electromagnetic fields in Bainchi type II, VIII and IX space times. Bali and Dave [7] have investigated Bianchi type IX string cosmological models with bulk viscosity. The study of cosmic strings has received considerable attention in cosmology, since they play an important role in structure formation and evolution of the universe [8,9]. The gravitational effects of cosmic strings have been extensively discussed by Vilenkin [9] and Gotez [10] in general relativity. Bianchi type-IX model has been intensively studied over the years. This model was first investigated by Belinski, Khalalatnikov and Lifschitz [11,12] and Misner [13]. One of the most appealing properties is its chaotic motion dynamics. The emergence of chaos helped in the understanding of the singularities in general relativity [14-17].

For historical reasons, the approach to study Bianchi model was to rewrite Einstein equations in the Hamiltonian form. This was intensively pursued by Misner [13,14] and several authors for which nice Hamiltonian system was constructed, although this specific form of Hamiltonian was not used in the proof of the collapse conjecture

by Lin and Wald [18,19].

On the other hand, the collapse conjecture for a Mixmaster space-time has been proved by Lin and Wald [18,19]. The collapse conjecture states that an initially expanding Mix-master space-time connot expand for an infinite time. It should, in a finite time, reach a maximum of expansion and then begin to contract. The proof given in reference [18] that there do not exist solutions which expand forever is divided in to steps. In the first step, it is proved that for any solution in which the universe expands forever, the dynamical trajectory in the anisotropy plane must escape to infinity along one of the channels of the mix-master potential. The second step uses detailed properties of the equation of motion to show that such escape along a channel is impossible.

IJETCAS 15-138; © 2015, IJETCAS All Rights Reserved

Page 80

R.K. Tiwari et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 80-83

The purpose of the present work is to obtain Bianchi type-IX string cosmological model with the help of relation, A = Bn,

  aR 2

and

   0 .

II. METRIC AND FIELD EQUATIONS We consider the Bianchi Type-IX metric as

ds 2   dt 2  A2 dx 2  B 2 dy 2  ( B 2 sin 2 y  A2 cos 2 y) dz 2  2 A2 cos y dx dz

(1) where A and B are function of t alone. The energy momentum tensor for a cloud of strings with perfect fluid distribution is taken as

Ti j  (   p)i j  pg ij  xi x j Where

i

and

xi

(2)

satisfy the conditions

i i   1   x i x i

(3)

i x  0

(4)

i



here p is the pressure,

is the proper energy density for clouds of string with particles attached to them,



is

the string tensor density,  is the four velocity of the particles and x is a unit space like vector representing the direction of the string.

In a commoving coordinate system, we have i

i

1 A

 

 i  (0,0,0,1) ; xi   , 0,0,0  If the particle density of configuration is denoted by

(5)

p

then

 p  

(6)

Einstein's field equations are (In gravitational units,8

are

1 Ri j  Rg ij   Ti j  g ij 2

(7)

for the line element(1), equation (7) leads to

B  2 2A B 1 A2  2  2  2    AB B B 4B

(8)

 2B B 2 1 3 A2  2  2   p  B B B 4 B4

(9)

 B  A  A2 A B      p  A B AB 4B 4

(10)

where the overhead dots stand for derivatives with respect to t. The average, scale factor ( R ) of Bianchi type IX model is defined as R= The expansion scalar

AB  2

()

2   3 The

field

3

shear

  3H  2

1

(11)

()

defined as

  A 2B  A B

(12)

 B  2 A     A B

equation

III. SOLUTION OF FIELD EQUATIONS (8)-(10) are a system of three equations

(13)

with

A, B, , p,  and . Now to close the system completely we need three more constraints. expansion  in the model is proportional to shear (). which leads to A = Bn Where n is constant

IJETCAS 15-138; © 2015, IJETCAS All Rights Reserved

six

unknowns

We assume that (14)

Page 81

R.K. Tiwari et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 80-83

Secondary we assume

  aR 2

(15)

From (11) and (14), we get

RB

( n 2) 3

(16)

from (16) and (15), We get

  aB

2 ( n  2 ) 3

(17) In order to overcome the indeterminacy of six unknowns involved in three independent field equations, we consider the Reddy string condition   0 (18) i.e. sum of the rest energy density and tension density for a cloud of string varnishes. from equations (8), (9), (10), (14) and (15), we get 2 ( n 2 )  B B 2 2 5 2 n4 2 (n  1)  (n  2n  2) 2  2  B  aB 3 B B B 4

 2  n  2  2 2(n2 2n2)  dB   2 5 B 2n2 6aB 3 B 2    aB n 1  2 dt 4 ( n  1 ) ( 6 n  2 n  4 ) ( n  2 n  2 ) ( 2 n  4 n  1 )  

(19) 1

2

(20)

Using equation (20), after some suitable transform of coordinates, metric (1), reduces to 1  2  n  2  2  2( n2 2n2)  2 5 T 2n2 6aT 3 n 1 ds    2    aT  dT 2 2  (n  2n  2) 4 (2n  4n  1) (n  1) (6n  2n  4)  2

 (T 2n dX 2  T 2 dY 2  (T 2 sin 2 Y  T 2n cos 2 Y ) dZ 2  2T 2n cos Y dX dZ The pressure p, energy density

,

expansion

(21)

, particle density  p , shear  and cosmological constant 

for the model (21), we get

  3(4n  1) 2 2  4n  1  1  (n  1) (6n  2n  4)  18   1  n T (1  n)  n 2  2n  2      T 2n4  (4n  1)  2(4n  1)  2 ( 3n  2 )  T (n  4)  5  4(n  1)  1 n (2n 2  4n  1)  p

2a 2 n3 2  T 1 n



     T 2 1  

2(2n  1)  T 2 n4  (n 2  2n  2)  4

(22)

2 ( n 2 )  5(2n  1)  3 1   aT  (2n 2  4n  1)   

2 ( n2 n3 )   6(2n  1) n1  1   a ( 2 n  1 ) T   (n  1) (6n  2n  4)  18  

The expansion scalar θ and particle density

(23)

for the model are obtained as

   2 ( n 2  n 3)  2T  2 5 T 2n4 6aT 3 n 1    (n  2)  2    aT 2  (n  2n  2) 4 (2n  4n  1) (n  1) (6n  2n  4)  18    2(n2)

1

2

(24)

 p      2

 2(2n  1)  T 2 n4 2T 2 1  2  4  (n  2n  2)   4 (2n  1) T   aB

2

2( n2)  5(2n  1)  3 1   2 aT   2 ( 2 n  4 n  1 )  

( n 2 n3) n1

2 ( n  2 ) 3

IJETCAS 15-138; © 2015, IJETCAS All Rights Reserved

   (2n  1)  (n  1)(6n  2n  4)  18    (25) (26)

Page 82

R.K. Tiwari et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014-February 2015, pp. 80-83  2 ( n 2  n  3)   2T  2 5 T 2n4 6aT 3  2    2T n 1  2 4 ( n  1 ) ( 6 n  2 n  4 )  18 ( n  2 n  2 ) ( 2 n  4 n  1 )     2 ( n  2 )

2  2  (n  1) 2 3

(27)

From (24) and (27), we get

2 2 (n  1) 2  3 3(n  2) 2 2

(28)

For the model (21), we observe that when T→0, p→∞, σ→∞, and σ/θ =constant ,As T→∞, Λ=0, σ = 0.

ρ=-λ→∞, θ→∞, ρp→∞, Λ→∞,

IV. DISCUSSION We observe that the model starts with a big-bang at T=0 where n < 2, the expansion in the model decreases as time increases. The expansion in the model stops at T = ∞, when -2< n< 2. It can also be seen that at T = 0, pressure p, energy density ρ, expansion scalar θ, particle density , vacuum density Λ and shear σ are all infinite when -2<n< 2, which means that at the initial time the whole universe is concentrated to one point mass, which is a sign of big-bang at T=0. Also at T=0, the anisotropy (σ/θ) ≠ 0. Thus the model is not isotopic at initial times. Again, at T = ∞; pressure p, energy density ρ, expansion scalar θ and particle density , are all zero when -2 < n < 2. Also anisotropy (σ/θ) ≠ 0 at T = ∞. Implies that the model does not approach isotropy at late times when -2 < n <2. Furthermore ,for n = 1, the shear scalar σ is zero and as T ∞. The model becomes isotopic for n = 1. Thus the model becomes isotropic at late times when -2< n < 2. The cosmological term is astronomical observations.

decreasing functions of cosmic time

, which is in agreement with recent

V. CONCLUSION In this paper we have investigated Bianchi type-IX cosmological model in presence of decaying vacuum density in general relativity. To get the deterministic model we have assumed three conditions: A=Bn, Λ=aR-2 (where R is the scale factor and a is constant) and ρ+λ=0. It has been seen that the universe starts with a big-bang at initial times having anisotropic behavior and still remain anisotropic at late times. However, for n = 1, the model isotropizes showing No-Hair conjecture studied by Wald [19]. It can be seen that the values of cosmological parameters resemble

model and so the model is in agreement with recent observations.

REFERENCES

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

Kibble, T.W.B. Topology of cosmic domains and strings, J. Phys, A. Math Gen. 9, 1387-1398 (1976).

Letelier, P.S. String Cosmologies Phys. Rev. D28, 2414-2419 (1983).

Satchel, J.: Phys. Rev. D 21, 2171 (1980).

Bali, R., Anjali : Bianchi type I Magnetized string cosmological model with bulk viscous fluid in general relativity. Astrophy Space Sci. 288, 399-405 (2003).

Rao, V.U.M., Shanti, M.V., Vinutha : Exach Bianchi type II, VIII and IX String Cosmological models in Saez-ballester Theory of Gravitation Astrophys Space Sci 314, 73-77 (2008).

Singh, G.P.: String Cosmology with Electromagnetic field in Bianchi type II, VIII and IX Space times Nuovogiomento B 110, 1463-1471 (1995).

Bali, R., Dave, S. Bianchi type IX String Cosmological models with bulk viscous fluid in general relativity Astrophys Space Sci. 288, 399-405 (2003).

Kibble, T.W.B., Turok, N.: Phys. Lett. B 116, 141 (1982).

Vilenkin, A.: Phys Rev. D23, 852 (1981).

Goetz, G.: Maths Phys 31, 2683 (1990).

V.A. Belinsky, I.M. Khalatnikov and E.M. Lifschitz, Adv. Phys. 19, 525 (1970).

I.M. Khalatnikov and E.M. Lifschitz, Phys. Rev. Lett. 24, 76 (1970).

C.W. Misner, Phys. Rev. Lett. 22, 1071 (1969).

N.J. Cornish and and J.J. Levin, Phys. Rev. Lett. 78, 998 (1997).

A.E. Motter and P.S. Letelier, Phys. Lett. A 285, 127 (2001).

D. Hobill, A. Burd and A.A. Coley, Editors, NATO Science Series: B Phys. Vol. 332 (1994).

C.W. Misner, Phys. Rev. 186, 139 (1969).

X. Lin and K.M. Wald, Phys. Rev. D 40, 3280 (1989).

X. Lin and R.W. Wald, Phys. Rev. D 41, 2444 (1990).

IJETCAS 15-138; © 2015, IJETCAS All Rights Reserved

Page 83

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Thermo Physical Properties of Nano Ferro fluids L.S.V Prasad1, Paul Douglas Sanasi2, V.Srinivas3 Associate Professor, Mechanical Engineering, AU College of Engineering (A), Andhra University, Visakhapatnam-530003, Andhra Pradesh, INDIA. 2 Assistant Professor, Engineering Chemistry, AU College of Engineering (A), Andhra University, Visakhapatnam-530003, Andhra Pradesh, INDIA. 3 Professor and Head, Mechanical Engineering, GITAM University, Visakhapatnam, Andhra Pradesh, INDIA. 1

_______________________________________________________________________________________ Abstract: Nanofluids have novel properties that make them potentially useful in many heat transfer applications. Nanofluids have already got the attention of research due to their unique thermo physical properties. Experimental results of many researchers indicate that the thermo physical properties of nanofluids depend on parameters including the thermal conductivities of the base fluid, the nanoparticles and volume fraction. This article presents preparation and experimental determination of thermal conductivity and viscosity of nano ferro fluid and development of a suitable correlation for determination of thermo physical properties.

Keywords: Nanofluid, Thermo Physical Properties, Thermal conductivity, Viscosity ________________________________________________________________________________________ I. Introduction Nano fluids are dilute liquid suspensions of nanoparticles with at least one critical dimension smaller than ~100nm. Nanofluids are a new class of fluids engineered by dispersing nanometer-sized materials in base fluids. Common base fluids include water, ethylene glycol and oil. For the past decade pioneering scientists and engineers have made phenomenal discoveries that a very small amount of guest nanoparticles can provide dramatic improvements in the thermal properties of the host fluids, such as anomalously high thermal conductivity at low nano particle concentrations, strong temperature and size dependent. Nanofluids are of great scientific interest because their unprecedented thermal transport phenomenon which surpass the fundamental limits of conventional macroscopic theories of suspensions. Therefore, numerous mechanisms and models have been proposed to account for these unexpected, intriguing thermal properties of nanofluids. Sarma et al., [1] developed a theoretical model to estimate the thermal conductivity of water based nanofluids dispersed with metal and metal oxides nanoparticles. Abbaspoursani et al., [2] developed an empirical model for effective thermal conductivity of TiO2-water and Al2O3-water nanofluids based on dimensionless variables indicated an enhancement in thermal conductivity. Shams et al., [3] proposed that thermo-physical properties primarily depends on the properties of base fluids and nanoparticles. Tajik Jamal et al., [4] experimental results showed that thermal conductivity of nanofluids were higher than the base fluid. Investigation of Das et al., [5] used temperature oscillation technique is utilized for the measurement of thermal diffusivity and thermal conductivity is calculated from it. The results indicate an increase of enhancement characteristics with temperature. The objective of the present investigation is to study the influence of ferro nanoparticles on thermal conductivity and viscosity of nanofluids and formulate a empirical correlation as a function of temperature and volume fraction.

II. Preparation of Nanofluids There are two fundamental methods to obtain nanofluids (1) Single step direct evaporation method – the dispersion of nanoparticle is obtained by direct evaporation of nanoparticle metal and condensation of nanoparticles in the base fluid.

(2) Two step method – first nanoparticles are obtained by different methods and then dispersed into the base fluid.

The suspensions obtained by either case should be well mixed, uniformly dispersed and stable in time.

Nanoparticles can be produced from several processes such as gas condensation, mechanical attrition or chemical precipitation techniques. As condensation processing has an advantage over other techniques. This is because the particles can be produced under cleaner conditions and its surface can be devoid of undesirable coatings. However, the particles produced by this technique occur with some agglomeration, which can be broken up into smaller clusters by supplying a small amount of energy. The precipitation of nanofluids begins by direct mixing of the base fluids with nanoparticles. The delicate preparation of a nanofluid is important because nanofluids need special requirements such as an even suspension, durable suspension, stable suspension with low agglomeration of particles.

The formation of Ferro fluid involves various types of forces that hold the components together. Magnetite is held together by ionic interactions. Ionic attractions between hydroxide anions and tetra methyl ammonium cations allow colloidal suspension of the magnetite in the solution. Without tetramethyl ammonium hydroxide

IJETCAS 15-140; © 2015, IJETCAS All Rights Reserved

Page 84

L.S.V Prasad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 84-88

as a surfactant, the magnetite nanoparticles tend to cluster together. Therefore it is necessary to have the appropriate surfactant to stabilize an aqueous ferrofluid.

III. Synthesis of Magnetite Nano crystals The following were the steps involved in the preparation of Ferro nanofluid

Fe (III) coordinates to 6 water molecules and Fe (II) coordinates to 4 water molecules (not shown) until the solid forms. The water molecules on the periphery of the magnetite are ultimately replaced by tetramethyl ammonium hydroxide. The solutions of Ferro Nano Fluids 2M FeCl3 and 1M FeCl2 are combined before titration as shown in fig.1 and titrated with ammonium hydroxide subjected to ultrasonication as shown in the fig 2. The magnetic particles settle down at the bottom of the beaker as shown in fig. 3 and water from the beaker is decanted. The precipitate is later washed repeatedly with water to remove excess ammonia and the prepared precipitate is later is used in the preparation of ferronanofluid. To this 1-2 ml of 25% tetramethyl ammonium hydroxide is added and the water content is adjusted till the spoke like arrangement of ferrofluid is obtained when a magnet is placed below the container. To this ferrofluid a measured quantity of ethylene glycol is added according to volume fraction required for this study.

A precise determination of important physical quantity such as particle diameter is very essential in the preparation of nanofluid. The particle diameter is determined using a particle analyzer and mean particle diameter is determined as 10.9 nm as shown in fig. 4. Ethylene glycol is considered as the base fluid and three different volume concentrations of ferronanofluid are prepared using the two step procedure.

Fig. 1 Combination of FeCl2 and FeCl3 before titration

\ Fig. 3: Separation of Magnetic Ferrofluid at the bottom of the container with a strong magnet

Fig. 2: Titration of FeCl2 and FeCl3 using Ultra Sonication.

Fig. 4 Particle analyze test IV. Thermo physical properties of nanofluids Dating back to the classical model of Maxwell (1881) on the properties for particle fluid mixtures, numerous theoretical and experimental studies have been conducted. To predict the anomalous thermal conductivity of nanofluids, it is observed from literature that thermal conductivity of nanofluids depends on several parameters which include thermal conductivity of the base fluid, temperature and volume fraction. In the current investigation, three different volume concentrations of ferro nanofluids are prepared independently. The volume concentrations selected for this study are Ď&#x2022; = 0.57, 0.66 and 0.8. The effective thermal conductivity of nanoparticle suspensions are studied for variable concentration of nanoparticles in the base fluid. Thermal conductivity of prepared nanofluid is measured with KD2 Pro a fully portable field and lab thermal properties

IJETCAS 15-140; Š 2015, IJETCAS All Rights Reserved

Page 85

L.S.V Prasad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 84-88

analyser. It uses the transient line heat source method to measure thermal diffusivity, specific heat (heat capacity), thermal conductivity, and thermal resistivity. The test tube containing the sample initially ethylene glycol and later nanofluid is transferred into a facility where heating of the sample is carried out using water bath. Variation of thermal conductivity with temperature is measured and used for further analysis.

Einstein (1956) was the first to calculate the effective viscosity of a suspension of spherical solids using the phenomenological hydrodynamic equations. Experimental data for the effective viscosity of nanofluids are limited to certain nanofluids. The ranges of the parameters (the particle volume concentration, temperature, etc.) are also limited. Still, the experimental data show the trend that the effective viscosities of nanofluids are higher than the existing theoretical predictions. Viscosity measurement is done with the help of with Ostwald's viscometer. Ostwald viscometer is a commonly used viscometer, which consists of a U-shaped glass tube held vertically. For more accurate measurements it is held in a controlled temperature bath. It is also known as a glass capillary viscometer. The sample liquid is allowed to flow through its capillary tube between two etched marks and the time of flow of the liquid is measured using a stopwatch. The variation of viscosity with temperature is experimentally recorded and the data is further used to correlate a non-linear regression equations.

Thermal Condcutivity, w/m k

V Results The experimental data shows that thermal conductivity increases with temperature for this nanofluid within the temperature range (55°C) of the experiment. To illustrate the magnitude of conductivity enhancement by the nanoparticle over the base fluid, it is observed from Fig. 5 that thermal conductivity increases for volume concentration ranging from 0.57 to 0.8. Thermal diffusion due to Soret effect and micro-convection setup by Brownian motion of nanoparticles can enhance thermal conductivity of nanofluid and an average improvement of 33% has been observed. It is also found that temperature, volume fraction have significant effects over viscosity of nanofluids. Although researchers deduce that viscosity decreases with increase in temperature, results indicate that viscosity increases as particle volume concentration increases as shown in Fig. 6. Literature reveals that presence of nanoparticles can influence the Newtonian behaviour of fluid enhancing viscosity with increase in volume fraction. An average enhancement of 52% is observed with increase in volume fraction.

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Base Fluid 0.571 % 0.667 % 0.8 % 30

40

50

Temperature,

60

oC

Viscosity, cP

Fig. 5 Variation of thermal conductivity of ferro nanofluids with temperature 40 35 30 25 20 15 10 5 0

Base Fluid 0.571 % 0.667 % 0.8 % 35

40

45

50

55

Temperature, oC Fig. 6 Variation of viscosity of ferro nanofluid with temperature

IJETCAS 15-140; Š 2015, IJETCAS All Rights Reserved

Page 86

L.S.V Prasad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 84-88

A generalized regression equation is developed for estimation of thermal conductivity and viscosity of ferro nanofluid and is validated in Fig. 7 and Fig. 8 respectively. The equation is valid up to a volume concentration of = 0.8. The equation developed for thermal conductivity is with an average deviation of 3.6% and standard deviation of 4.2% given by

Experimental thermal condcutivity ratio

2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.40

1.60 1.80 2.00 Theoretical thermal condcutivity ratio

2.20

Fig. 7 Validation of formulated regression equation for thermal conductivity of nanofluids

Experimental Viscosity ratio

Similarly the experimental data of viscosity is subjected to regression equation valid upto = 0.8 volume concentration and the equation is developed with an average deviation of 6.9% and standard deviation of 9.6% given by

2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.40

0.90 1.40 1.90 Theoretical Viscosity ratio

2.40

Fig. 8 Validation of formulated regression equation for viscosity of nanofluids VI Conclusion The following conclusions are made from the experimental observations  The thermal conductivity of ferronanofluid improved by an average of 33% within the given temperature range  The viscosity of the ferronanofluid is also found to increase by an average of 52% within the given temperature range.

IJETCAS 15-140; © 2015, IJETCAS All Rights Reserved

Page 87

L.S.V Prasad et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 84-88



Thermal conductivity and viscosity of ferronanofluid can be estimated from Eqs (4) and (5) respectively.

Nomenclature Thermal conductivity, W/mK Temperature, ºC Greek symbols Volume concentration of nanoparticle, % Subscripts nanofluid base fluid References [1]

[2] [3]

[4] [5] [6] [7] [8] [9] [10] [11]

Sharma, K.V., Sarma, P.K., Azmi, W.H., RizalmanMamat., Kadirgrama, K., “Correlations to predict friction and forced convection heat transfer coefficients of water based nano fluids for turbulent flow in a tube”, International journal of Microscale and nanoscale thermal fluid transport phenomena. Vol. 3, No. 4, pp. 1 – 25 (2009.)

Abbaspoursani, K., Allahyari, M., Rahmani, M., “An improved model for prediction of the effective thermal conductivity of nanofluids”, World Academy of Science, Engineering and Technology, Vol. 58, pp. 234 – 237 (2011).

Shams, Z., Mansouri, S.H., Baghban, M., “A proposed model for calculating effective thermal conductivity of nanofluids, Effect of nanolayer and Non-uniform size of nanoparticles”, Journal of Basic and Applied Scientific Research, Vol. 2, pp. 9370 – 9377 (2012).

Tajik Jamal-Abadi, M., Zamzamian, A.H., “Thermal conductivity of Copper and Aluminum-water nanofluids”, International Journal of Engineering, Vol. 26, No. 8, pp. 821 – 828 (2013).

Das, S.K., Putra, N., Thiesen, P., Roetzel, W., “Temperature dependence of thermal conductivity enhancement for nanofluids”, Journal of Heat Transfer, Vol. 125, pp. 567 – 574 (2003).

Nicholas B. Adelman., Katie J. Beckman., Dean J. Campbell., Arthur B. Ellis., and George C. Lisensky., “Preparation and properties of an aqueous ferrofluid”, JChemEd.chem.wisc.edu, Vol. 76, No. 7, pp. 943-948 (1999).

Scherer, C., and Figueiredo Neto, A.M., “Ferrofluids: Properties and Applications”, Brazilian Journal of Physics, Vol. 35, no. 3A, pp. 718-727. (2005).

Putra, N., Roetzel, W., Das, S.K., “Natural convection of nano-fluids”, International Journal of Heat Mass Transfer, Vol. 39, No. 8, pp. 775–784 (2003).

Abu Nada, E., “Effects of variable viscosity and thermal conductivity of Al2O3–water nanofluid on heat transfer enhancement in natural convection”. International Journal of Heat and Fluid Flow, Vol. 30, No. 4, pp. 679 – 690 (2009).

Calvin, H., Li., Peterson, G.P., “Experimental investigation of temperature and volume fraction variations on the effective thermal conductivity of nanoparticle suspensions (nanofluids)”, Journal of Applied Physics, Vol. 99, No. 8, pp. 4314 – 4321 (2006).

Koo, J., Kleinstreuer, C., “Impact analysis of nanoparticle motion mechanisms on thermal conductivity of nanofluids”, International Communications in Heat and Mass Transfer, Vol. 32, pp. 1111 – 1118 (2005).

IJETCAS 15-140; © 2015, IJETCAS All Rights Reserved

Page 88

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net A Novel Efficient and Accurate Detection Model to Detect Emerging Attacks in Network Supriya Gupta1, Ankur Goyel2 M.Tech Scholar, Department of Computer Science & Engineering, Y.I.T., Jaipur, Rajasthan, India 2 Associate Professor, Department of Computer Science & Engineering, Y.I.T., Jaipur, Rajasthan, India 1

Abstract: Over past few decades, the rapid popularity of computer networks and their related applications have increases the information security issues in rocket speed which pose challenges to develop an efficient network intrusion detection system that can detect a high rate of attacks with the acceptable reliability. In current scenario no one detached security system is capable to detect all types of attack with accuracy, day by day naive attackers launch powerful attacks which can bring down an entire network. However, in the last few decades a lot of approaches have been intensively proposed to defend the system, but complete network security still is a critical issue due to the wideness of the network system, the field lacks an integrated approach with high detection rate for minority attacks namely R2L and U2R. In this paper we have proposed a new advanced network security approach which can efficiently detect a wide variety of attacks with higher accuracy.

Keywords: Feature Reduction, Network Security, Anomaly Detection, Misuse Detection I. Introduction In recent few decades, the use of open network system has gain high interest of users to communicate the message without any barrier of distance. The advancement and increasing interest of communication technology have also augmented the safety fears significantly. Day by day new attackers comes with novel way to exposed more vulnerabilities of network. There are a large variety of methods in which the security of a system can be compromised. Typically a security system monitors suspicious activity in network and alerts against them but from last few decades, attacks have become more complex and difficult to detect using traditional detection approaches, demanding more effective solutions [1]. Thus, in vision of current advancement the networks and applications are far away from a state where they can be considered as secure. However, several of approaches have been introduced by researchers for improving the quality of security [2-5] but they have own problem in their performance, which can be enhanced by increasing the detection rates and reducing the false positives [6].

The false positive is an error which comes when a normal type action detected as an attack types. In general false alarms are known as false positive. Broadly detection system can be categories in two categories: (i) Host-based (ii) Network-based detection system [7]. Host-based attacks target a machine and try to gain access to privileged services or resources on that machine. Host-based detection usually uses routines that obtain system call data from an audit process, which tracks all system calls made on behalf of each user. Network-based attacks make it difficult for legitimate users to access various network services by deliberately occupying or sabotaging network resources and services. This can be done by sending large amounts of network traffic, exploiting well-known faults in networking services, and overloading network hosts.

II. Attack Classes Attacks are broadly classified into four classes which are A. Denial of Service (Dos) Attacks: In a denial-of-service (DoS) attack, an attacker attempts to prevent legitimate users from accessing information or services. The most common DoS attacks will target the computer's network bandwidth or connectivity. Bandwidth attacks flood the network with such a high volume of traffic that all available network resources are consumed and legitimate user requests cannot get through. Connectivity attacks flood a computer with such a high volume of connection requests, that all available operating system resources are consumed and the computer can no longer process legitimate user requests [8].

B. Probe Attacks: Typically before the attack launching the attacker search victimâ&#x20AC;&#x2122;s network or host for open ports. The attacker use extensive process across the hosts on a different network and within a single host for services to probe the target network or single machine for open ports. This process is referred to as Probe Attacks.

C. Remote to Local (R2L) Attacks:-In this type of attacks the attackers does not have an account on the victim machine, hence tries to gain access from a remote machine and exploits access in order to send packets over the network. These attacks involve in both network and the host level feature.

IJETCAS 15-143; Š 2015, IJETCAS All Rights Reserved

Page 89

Supriya Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 89-93

D. User to Root (U2R) Attacks:- User to root or user to super user exploits are a class of attacks where an attacker starts out with access to a normal user account on the system and then exploits vulnerability to gain root access. These attacks involve the semantic details which are very difficult to capture at an early stage at the network level. The most common exploits in this class of attacks are regular buffer overflows, which are caused by regular programming mistakes and incorrect environment assumptions.

III.

Related Work

To improve the detection rate of attacks a novel attack detection model has been proposed [9]. The proposed model performs well for all the classes of attack. In this framework authors use four tiers architecture to enhance the adaptability of the cyber attack detection. The data collection and preprocessing of the proposed model is included in first tier of proposed model. The Second tier is meant for the feature extraction technique, third tier is dedicated to classification of cyber attacks and fourth tier is dedicated to user interface for reporting the events.

The Second tier is meant for the feature extraction technique, third tier is dedicated to classification of cyber attacks and fourth tier is dedicated to user interface for reporting the events. Several of researchers proposed layered approaches [1,10,11] for improving the quality of detection system. In [1, 10] authors addressed the dual problem of accuracy and efficiency for building robust and efficient detection systems. They used CRFs for improving the attack detection rate and decreasing the FAR. Having a low FAR is very important for any intrusion detection system. Further, feature selection and implement the layered approach significantly to reduce the time required to train and test the model. The proposed system achieved high gain in attack detection accuracy, particularly, for the U2R attacks (34.8 percent improvement) and the R2L attacks (34.5 percent improvement). A combinational model is designed in [11] for attack detection mechanism. The meta-modeling applied in this for gaining better classification performance than any individual classifier. To test the results used NSL-KDD datasets; and also applied PCA for feature reduction that results in a significant improvement on learning algorithms. A novel framework [12, 13] proposed to satisfy the core purpose of attack detection system, and allows detecting the attacks as quickly as possible with available data using mobile agents. This framework was mainly designed to provide security for the network using mobile agent mechanisms to add mobility features to monitor the user processes from different computational systems. The experimental results have shown that the system can detect anomalous user activity effectively.

Adaboost algorithm for network attack detection system with combination of multiple weak classifiers is proposed in [14]. The classifiers such as Bayes Net, Naive Bayes and Decision Tree have been used as weak classifiers. A benchmark dataset is used in these experiments to demonstrate that boosting algorithm can greatly improve the classification accuracy of weak classification algorithms. The approach achieves higher detection rate with low false alarm rates and is scalable for large datasets, resulting in an effective detection system. In same context a new detection approach, especially for network attack detection, based on improved genetic algorithm (GA) and multi-ANN classifiers has proposed in [15]. The improved GA used energy entropy to select individuals, optimize the training procedure of the BPNN, RBF, PNN and Fuzzy-NN. Then, the satisfactory ANN models with proper structure parameters were attained. In addition, to alleviate the complexity of the input vector, the principal component analysis (PCA) has been employed to eliminate redundant features of the original disturbance data. The efficiency of the proposed method was evaluated with the practical data, and the experiment results show that the proposed approach offers a good detection rate, and performs better than the standard GAANN method. Designs of multi stage filter propose in [16], which dealing with various categories of attacks in networks. The first stage of the filter is designed using Enhanced Adaboost with Decision tree algorithm to detect the frequent attacks occurs in the network and the second stage of the filter is designed using enhanced Adaboost with NaĂŻve Byes algorithm to detect the moderate attacks occurs in the network. The final stage of the filter is used to detect the infrequent attack which is designed using the enhanced Adaboost algorithm with NaĂŻve Bayes as a base learner. Performance of this design is tested with the KDDCup99 dataset and is shown to have high detection rate with low false alarm rates.

IV.

Proposed Approach

Typically all the accessing attack detection system depends on its dataset with its accuracy. If detection system trained with accurate and rich content then proposed approach will produce high performance in attack detection process. Hence collecting accurate and efficient data is an important problem for training and testing of a system.

In order to maintain the high accuracy and detection rate while at the same time to lower down the false alarm rate, we have apply feature reduction technique on the original NSL KDDCUP dataset with 1,25,973 data connections. Feature reduction [17] techniques are commonly used as preprocessing to machine learning and statistical tasks of prediction, including pattern recognition and regression. Although such problems have been tackled by researchers for many years, there has been recently a renewed interest in feature extraction [18]. The feature space having reduced features that truly contributes to classification that cuts pre-processing costs and

IJETCAS 15-143; Š 2015, IJETCAS All Rights Reserved

Page 90

Supriya Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 89-93

minimizes the effects of the ‘peaking phenomenon’ in classification [19]. Thereby improving the overall performance of classifier based cyber attack detection. Typically, some of the conventional methods widely used for feature reduction such as Correlation-based Feature Selection (CFS) using best first search method, Gain Ratio (GR) using ranker method, Information Gain (IG) using ranker method. These automatic feature selection methods are not effective in each and every case, especially in case to design a efficient attack detection system to detect the novel/unseen attacks. As stated earlier that manually effort has done to reduce the feature set for proposed system. Table 1 depicts the number of features selected by each feature reduction method. The reduced datasets are further classified by using common classification algorithm called Naïve Bayes using WEKA on discretized values to measure the classification performance.

Table 1: Number of Features Selected By Each Feature Reduction Method # Attribute Selected 10

Feature Selection Technique CFS+ BestFirst

Selected Attributes 3,4,5,6,12,26,29,30,37,38

GR + Ranker

15

1,8,9,11,13,14,17,18,23,27,28,32,34,36,41

InfoGain + Ranker

7

2,11,14,15,18,19,41

Selected Feature set for proposed work

14

1,3,5,6,10,13,16,17,27,30,32,33,36,37

Fig. 1: Adopted Methodology for Attack Detection V.

Experimental Evaluation

In order to evaluate the performance of proposed algorithm, we performed classification using KDD99 intrusion detection benchmark dataset with 1,25,973 data connections. This dataset overcome the problem of KDD99cup dataset and contains four types of attacks such as DoS(Denial of Service), R2L (Remote to Local), U2R (User to Root), Probe and normal data. The DoS and Probe attacks belong to majority class whereas U2R and R2L belongs to minority class also called as rare class of attacks. In first experiment NBC executed with reduced (selected) feature set as present in table 1 and compare it results with the outcomes of other well known automatic feature selection techniques. Fig. 2 shows the result of experiment.

Fig. 2: Accuracy of Selected Feature Set with NBC The fig. 3 presents the comparative results to shows the effectiveness of proposed approach over the other accessible approaches [20-22].

IJETCAS 15-143; © 2015, IJETCAS All Rights Reserved

Page 91

Supriya Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 89-93

Fig. 3: Comparative result of proposed approach on base of attack detection accuracy

Fig. 4: False Positive Rate The results as depicted in fig. 2 to fig. 4 indicates that proposed approach has produce low false alarm in comparison of other attack detection algorithms. The low false rate has presents the effectiveness and the efficiency of the proposed algorithm.

VI.

Conclusion & Future Work

The experimental results of proposed approach as depicted in previous section clearly suggest that proposed approach can be used to build effective and efficient network security system against various types of attacks. The proposed approach also improves the ratio of correctly classification with low false positive rates. However, proposed approach has produced better results in comparison of other accessible approaches but use domain knowledge, which require further investigation. Another possible direction for future research is method for feature selection. The selection process can be further investigated and deployed using some other techniques to speed up the proposed system, which can be customized by introducing another interesting research in region of network security.

References [1]

[2] [3] [4] [5]

[6]

Kapil Kumar Gupta, Baikunth Nath, Kotagiri Ramamohanarao, and Ashraf Kazi. Attacking Confidentiality: An Agent Based Approach. In Proceedings of IEEE International Conference on Intelligence and Security Informatics, pages 285–296. Lecture Notes in Computer Science, Springer Verlag, Vol (3975), 2006.

Prabhjeet Kaur, Amit Kumar Sharma, Sudesh Kumar Prajapat ― Madam ID for intrusion detection using data mining‖ IJRIM volume 2, issue 2, February 2012 Yogendra Kumar Jain* and Upendra ―An Efficient Intrusion Detection Based on Decision Tree Classifier Using Feature Reduction‖ International Journal of Scientific and Research Publications, Volume 2, Issue 1, January 2012.

G.V. Nadiammai, S.Krishnaveni, M. Hemalatha ― A Comprehensive Analysis and study in Intrusion Detection System using Data Mining Techniques‖ International Journal of Computer Applications (0975 – 8887) Volume 35– No.8, December 2011 R.Shanmugavadivu, Dr.N.Nagarajan ―Learning of Intrusion Detector in Conceptual Approach of Fuzzy Towards Intrusion Methodology‖ International Journal of Advanced Research in Computer Science and Software Engineering, Volume 2, Issue 5, May 2012 Iftikhar Ahmad ―Feature Subset Selection In Intrusion Detection Using Soft Computing Techniques‖ A Thesis, Department of Computer and Information Sciences Universiti Teknologi Petronas Bandar Seri Iskandar Perak February 2011

IJETCAS 15-143; © 2015, IJETCAS All Rights Reserved

Page 92

Supriya Gupta et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 89-93 [7] [8] [9] [10]

[11] [12] [13] [14]

[15] [16] [17] [18] [19] [20] [21] [22]

Asmaa Shaker Ashoor, Prof. Sharad Gore ―Importance of Intrusion Detection System (IDS)‖ International Journal of Scientific & Engineering Research, Volume 2, Issue 1, January-2011 Frank Kargl, Jörn Maier, Stefan Schlott, Michael Weber “Protecting Web Servers from Distributed Denial of Service Attacks” ACM 1-58113-348-0/01/0005. May 1-5, 2001.

Shailendra Singh, Sanjay Silakari “An Ensemble Approach for Cyber Attack Detection System: A Generic Framework” 14th ACIS, IEEE 2013. Pp 79-85.

Mr.C.Saravanan, Mr.M.V.Shivsankar, Prof.P.Tamije Selvy, Mr.S.Anto ―An Optimized Feature Selection for Intrusion Detection using Layered Conditional Random Fields with MAFS‖ International Journal of Mobile Network Communications & Telematics (IJMNCT) Vol.2, No.3, June 2012 Ankita Gaur, Vineet Richariya ―A Layered Approach for Intrusion Detection Using Meta-modeling with Classification Techniques‖ International Journal of Computer Technology and Electronics Engineering (IJCTEE) Volume 1 , Issue 2 N.Jaisankar1 and R.Saravanan2 K. Durai Swamy ―intelligent intrusion detection System framework using mobile Agents‖ International Journal of Network Security & Its Applications (IJNSA), Vol 1, No 2, July 2009 B.Bhanu Chander*, K. Radhika, D. Jamuna ― An Approach On Layered Framework For Intrusion Detection System‖ Asian Journal of Computer Science And Information Technology 2: 8 (2012) 230 – 233.

P. Natesan, P. Balasubramanie, G. Gowrison ―Improving Attack Detection Rate in Network Intrusion Detection Using Adaboost Algorithm with Multiple Weak Classifiers‖ Journal of Information & Computational Science 9: 8 (2012) 2239–2251 Available at http://www.joics.com Yuesheng Gu, Yongchang Shi, Jianping Wang ―Efficient Intrusion Detection Based on Multiple Neural Network Classifiers with Improved Genetic Algorithm‖ JOURNAL OF SOFTWARE, VOL. 7, NO. 7, JULY 2012 P.Natesan1, P.Balasubramanie ―Multi Stage Filter Using Enhanced Adaboost for Network Intrusion Detection‖ International Journal of Network Security & Its Applications (IJNSA), Vol.4, No.3, May 2012 Gopi K. Kuchimanchi,Vir V. Phoha, Kiran S.Balagani, Shekhar R. Gaddam, Dimension Reduction Using Feature Extraction Methods for Real-time Misuse Detection Systems, Proceedings of the IEEE on Information, 2004.

Venkatachalam V, Selvan S. An approach for reducing the computational complexity of LAMSTAR intrusion detection system using principal component analysis. International Journal of Computer Science, 2(1): 76-84, 2007 C.Burges. A Tutorial on Support Vector Machines for Pattern Recognition. Data Mining and Knowledge Discovery, 2(2): 121167, 1998.

Chandrashekhar Azad* and Vijay Kumar Jha “Data Mining based Hybrid Intrusion Detection System”, Indian Journal of Science and Technology, Vol 7(6), 781–789, June 2014 Karan Bajaj1, Amit Arora2 “Dimension Reduction in Intrusion Detection Features Using Discriminative Machine Learning Approach” IJCSI International Journal of Computer Science Issues, Vol. 10, Issue 4, No 1, July 2013 Rowayda A. Sadek1,2 , M. Sami Soliman3 and Hagar S. Elsayed “Effective Anomaly Intrusion Detection System based on Neural Network with Indicator Variable and Rough set Reduction” IJCSI International Journal of Computer Science Issues, Vol. 10, Issue 6, No 2, November 2013.

IJETCAS 15-143; © 2015, IJETCAS All Rights Reserved

Page 93

International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research)

ISSN (Print): 2279-0047 ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational and Applied Sciences (IJETCAS) www.iasir.net Analysis of 5-Iodo-Uracil for its Infrared Spectra, Laser Raman Spectra and its Thermodynamic Functions Dr. Pradeep Kumar Sharma Professor and Head, Department of Physics University of Engineering & Management, Jaipur, Rajasthan (303807), INDIA I. Introduction The spectral interpretation of N-Hetrocyclics like Pyrimidine, cytosine, thymine, Uracil and their derivatives is difficult because of their high complexity and low symmetry. Since these molecules are of considerable biological interest, their detailed spectral investigations are of great importance and many workers 1-10 have studied the vibrational and electronic spectra of such molecules. The vibrational spectra of these compounds may provide some evidence about outstanding structural problems. One of these concerns the tautomerism in some molecules which is possible with the involvement of functional groups like –OH, -NH and –SH and which may effect their function in the metabolism. A full understanding of the vibrational spectra of nucleic acid base is of great importance in biophysical research11-14 . The position of substituents their tautomeric behaviour and hydrogen bonding ability of C-O bond in substituted uracils will be considerable biological evidence. The spectral studies of nucleic acid bases and their derivatives which contain uracil moieties have been carried out both experimentally and theoretically. In view of the above the present analysis reports the Infrared, Laser Raman Spectra and Thermodynamic Functions of Solid 5-Iodo-Uracil.

II. Experimental Spec-pure grade of 5-Iodo-Uracil (hereafter referred as 5-IU) was obtained. Its purity was confirmed by elemental analysis and melting point determination. Infrared spectrum was recorded with Perkin Elmer M-683 Spectrophotometer in the region 600-4000 cm–1 in KBr pallets. The laser Raman spectrum in the region 60-4000 cm–1 has been recorded on a spex-Ramalab spectrophotometer using 52MG argon-Krypton laser of wavelength 488 nm. Vapour phase ultraviolet spectra were tried on medium Quartz spectrograph but under best experimental conditions no reportable bands were observed.

III. Results and Discussion The structural diagram of the compound is given in the fig. 1. The infrared spectra are shown in fig. 2 and 3 and Vibrational Assignments are shown in table 1. Normalized Laser Raman Spectra are shown in fig.4. Tautomeric forms of neutral, cationic and anionic forms of 5-IU are presented in fig. 5 & 6. The statistically computed thermodynamic function viz. enthalpy function, free energy function, entropy and heat capacity with absolute temperature are shown in table 2 for 5-IU. The variation of enthalpy function and heat capacity with absolute temperature is shown in the fig. 7, while the variation of free energy function and entropy with absolute temperature is shown in fig. 8. Also other values of thermal energy and potential barrier are given in the table 3 and their variation with absolute temperature is recorded in fig.9.

FIG. 1: 5 – IU

FIG. 2: INFRARED SPECTRUM OF 5-IU

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 94

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

FIG. 3: COMPARISON OF NORMALIZED IR SPECTRA, (A) EXPERIMENTAL AND SCALED SIMULATED SPECTRUM OBTAINED BY USING (B) DFT (SCALING FACTOR X 0.96) AND (C) HF (SCALING FACTOR X 0.89), HARMONIC CALCULATIONS FOR URACIL.

FIG. 4: COMPARISON OF NORMALIZED RAMAN SPECTRA (A) EXPERIMENTAL AND SCALED SIMULATED SPECTRUM OBTAINED BY USING (B) DFT (SCALING FACTOR X 0.96) AND (C) HF (SCALING FACTOR X 0.89), HARMONIC CALCULATIONS FOR URACIL. In uracil, two C-H stretching frequencies are present. The molecule under investigation is tri-substituted pyrimidine. The aromatic structure of the ring compound shows the presence of C-H stretching vibration in the region 3100-3000 cm–1 15. These vibrations are characteristic of the aromatic ring itself and afford a ready identification for the structure. The bands are not appreciably affected by the nature of substituent. The C-H stretching frequencies in benzene derivatives arises from the mode of a 18 (3073 cm–1), e29 (3096 cm–1), b1u (3060 cm–1), and e1u(3080 cm–1) of benzene. In case of 5-IU only one hydrogen is attached with carbon atom of the

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 95

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

ring, hence it will involve one C-H valence oscillation in 3000-3100 cm–1 has been assigned to C-H stretching vibration, which in accordance with literature values16-19. However, Raman band at 650 cm–1 and infrared active band at 640 cm–1 have been well assigned to C-H in-plane bending vibration.

Pyrimidine and the tautomers of uracil have the usual four bands corresponding to coupled C-C, C-N stretching vibrations in the region 1600-1400 cm–1. The pair of absorption band at 1564 and 1568 cm–1 in pyrimidine were assigned as the C-C and C-N coupled vibrations20 analogous to the pair of bands originating from e1u (1485 cm–1) and e2g(1595 cm–1) modes of benzene. Under the present investigation these bands have been assigned at 1460, 1510, 1530 and 1555 cm–1 in infrared spectrum while at 1430, 1515 and 1530 cm–1 in Raman spectrum of the said molecule.

The frequency of ring breathing mode which corresponds to a 1g (999 cm–1) mode of benzene, under reduced symmetry is lowered much because of the interaction of closely lying vibration b1u (1010 cm–1). Jakobsen and Brewer21 assigned the frequencies between 820-860 cm–1 to the a19 mode. Kartha22 reported this mode around at 820 cm–1 in methyl pyrimidines.

TABLE 1: VIBRATIONAL ASSIGNMENT OF 5-IODO-URACIL

I.R. Bands (cm–1)

Raman Bands (cm–1)

Assignments

210 m

-

Lattice vibration

305 s

-

 (C-I)

330 m

335 m

(C-OH)

400 m

390 vs

(O-H), (C=O)

420 m

-

 ring

455 m

-

 (C-I)

550 vs

550 m

 ring,  (C=O)

585 w

-

 ring,  (C-H)

640 vs

650 m

(C=O),  (C-H)

760 s

-

 ring, (C-I)

810 vs

-

(N-H)

990 s

-

Ring breathing

1055 vs

1060 vs

 ring

1180 s

1175 m

 (C-OH)

1245 s

1235 vs

(C-H), (O-H)

1315 w

-

(C-OH)

1340 w

-

 ring

1460 m

1430 w

 ring

1510 ms

1515 w

 ring

1530 ms

1530 m

 ring

1550 ms

-

 ring

1615 w

-

 (C=C)

1645 m

-

 (C=O4)

1690 m

1680 vs

 (C=O2)

3010 w

-

 (N-H)

3045 w

-

 (C-H)

3180 m

-

 (N-H)

3410 m

-

 (O-H)

3540 m

-

 (O-H)

w = weak, m=medium, s=strong, vs = very strong, ms = medium strong,  = stretching,  = in-plane bending,  = out-of plane-bending.

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 96

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

The N-heterocyclic ring breathing vibrations occurs at 791, 779, 790 and 780 cm –1 in cytosine, cytosine-d3, 2thiocytosine and 1-methyl-cytosine respectively18,23,24, In view of this the infrared band at 760 cm–1 in 5-IU has been assigned to ring breathing mode.

The components of vibration e2g (608 cm–1) and e2u (404 cm–1) which correspond to C-C in- plane and out- of plane deformation modes have been well identified in their respective regions in present study which find support from literature values15,25.

Susi and Ard26 assigned the N-H stretching mode at 3145 cm–1 in 1-methyl uracil and at 3146 cm–1 in 1- methyl thymine, Sanyal et al10 have assigned this mode at 3178 and 3098 cm–1 in 5-bromo-uracil, at 3188, 3128 cm–1 in 5-metyl uracil. In view of this, the bands at 3010 and 3180 cm–1 have been assigned to N-H stretching vibration in infrared spectrum of the present molecule.

The neutral form of uracil and its substituted derivatives have three double bands and their stretching vibrations are expected to given rise to three strong bands in region 1600-1700 cm–1. Susi and Ard27 have assigned the bands at 1695, 1662 and 1621 cm–1 to C=O2, C=O4 and C = C stretching vibrations respectively in 1-methyl uracil. Sanyal et al28 have assigned these vibrations at 1690, 1676 and 1614 cm–1 respectively in 5-metyl uracil.

In view of this the bands observed at 1690, 1645 and 1615 cm–1 in infrared spectrum of the present molecule have been assigned as C=O2, C=O4 and C-C respectively double bond stretching vibration.

Uracil due to the presence of two C=O at position 2 and 4 and two N-H group at 1 and 3 generally tautomerises to give a dihydroxy aromatic compound viz 2,4-dihydroxy-5-iodo-pyrimidine fig. 5.

It is well known that the substitution of –OH group in pyrimidine may change its aromatic structure to Ketonic one. Goel et al29 have observed strong bands at 3469 and 3569 cm–1 in case of 5,6-dimethyl uracil, to O-H stretching modes, which is well agreed with literature values30,31. In present molecule the infrared bands at 3540 and 3410 cm–1 may be assigned to O-H stretching modes. Kletz and Price32 have assigned C-OH stretching mode around 1300 cm–1 in substituted phenol while Sanyal et al35 assigned this mode at 1305 and 1332 cm–1 in dihydroxy methyl –pyrimidine. A weak C-OH stretching mode has been clearly identified at 1315 cm–1 in infrared spectrum of 5-IU, while the low frequency infrared at 1245 and 1180 cm–1 has been taken to represent the O-H in-plane bending mode, while only one Raman band at 1235 cm–1 has been assigned to this mode.

Money and other workers34-37 has assigned the C-I vibrations in the frequency interval the 1129-480 cm–1. The higher frequency bands have been assigned to C-I deformation mode. In few of this, the infrared band at 760 cm–1 has been assigned to C-I stretching vibration. While the infrared bands at 455 and 305 cm–1 has been assigned to in- plane bending and out- of plane bending vibrations respectively to C-I vibrations.

It is interested to note additional N-H stretching frequencies in the present studies. Sullivan and Sadler 38 have suggested the occurrence of bands between 3120-2900 cm–1 in indole-3- aldehyde corresponding to N–H stretching frequencies due to strong intra-molecular hydrogen bonding. According to Szczsniak39 band observed in infrared spectra of 1-methyl uracil in the region 3100-2800 cm–1 are due to hydrogen bonded N-H stretching frequencies., Rao25 has suggested that N-H stretching frequencies in the region 3300-3510 cm–1 are due to intermolecular association through NH-O hydrogen bonding.

FIG. 5: TAUTOMERISM IN 5-IODO URACIL The band observed at 3010 cm–1 in 5-IU has been assigned to N-H stretching frequency.

Jakobsen40 has assigned at weak band at 190 cm–1 to stretching mode OH-O in P-Cresol. According to Mallic and Banerjee41 in substituted phenol a low frequency band at 138 cm-1 is assigned to OH-O stretching mode due to intermolecular hydrogen bonding presence of C-OH, C=O, stretching and bending fundamentals indicate that the molecule under study is in a state of tautomeric equilibria.

As suggested by Becker et al42 hydrogen bonding takes place between non-bonding electrons of solvent and solute. The non-bonding electros are localized upon more electronegative atoms of polar group. The hydrogen bonding which stabilizes the ground state will be weakened by the shift of electron density away from the nonbonding center upon excitation, and a blue shift of the band will be observed relative to the band position in a non-hydrogen bonding solvent.

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 97

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

The unbonded or free hydroxyl group absorbs strongly in the region 3650-3580 cm–1. But intra-molecular hydrogen bonding or intermolecular hydrogen bonding usually, shifts the absorption to lower frequency, a broad extended band is characteristic of intra-molecular-hydrogen bonding while a splitted band suggests intermolecular hydrogen bonding. In the spectra of molecule under study we do not observe the same effect indicating the existence of hydrogen bonding in this molecule.

FIG. 6: TAUTOMERIC FORM OF 5-IODO-URACIL (2,4- DIHYDROXY-5- IODO-PYRIMIDINE) IV. Thermodynamic Functions Thermodynamic functions viz. enthalpy function (Ho-Eoo)/T, heat capacity C0p free energy function (Fo-E0o)/T and entropy So of 5-idouracil have been computed using the standard expressions43,44, by taking Y-axis perpendicular to the molecular plane and Z-axis to pass through the para position. For determining rotational contribution, the following structural parameters were used 45,46.

N1 - C2 = 1.31Å  N2C4O2 = 112° C2-N2 = 1.34 Å  C5C4O2 = 129° N2-C4 = 1.39 Å  N1C2O1 = 124° C4-C5 = 1.32 Å  N2C2O1 = 120° C5-C6 = 1.39 Å  C4C5I = 118° C6-N1 = 1.29 Å  C6C5I = 124° C2-O1 = 1.21 Å  C2N1C6 = 125° C4-O2 = 1.24 Å  C2N2C4 = 120° C5-I = 1.43 Å All other angles were taken as 120° in the ring.

The thermodynamic functions have been calculated at different temperatures between 200-1500 0K using fundamental frequencies and assuming rigid rotor harmonic oscillator approximation. The calculations were performed for 1 mole of an ideal gas at 1 atmospheric pressure. The symmetry number for overall rotation has been taken as 2 and internal rotation as 2. The principal moments of inertia were found to be 91.58, 116.63 and 25.05 gm × cm2 in this molecule while the reduced moment of inertia is 1.58 gm x cm2. The barrier height V is found to be 17.73 K cal/Mole.

The vibration of enthalpy function

H

o



 E00 / T

and heat capacity CoP with absolute temperature have been

F

o



 E0 / T

0 shown graphically in fig. 7. While those of free energy function and entropy So in fig. 8. Which 47,48 are in agreement with the trend reported in literature . The variation of thermal energy and potential barrier with absolute temperature has also been recorded in table 3.

TABLE 2: THERMODYNAMIC FUNCTIONS (IN CAL/DEG. MOLE) OF 5-IODO URACIL Temperature Kelvin 200

Deg.

Enthalpy

Free Energy (–)

Entropy

Heat Capacity

13.54

59.09

72.63

19.45

273

15.57

63.92

79.49

22.77

300

16.27

65.52

81.78

23.91

400

18.68

70.82

89.50

27.87

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 98

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

Temperature Kelvin 500

Deg.

Enthalpy

Free Energy (–)

Entropy

Heat Capacity

20.87

75.45

96.32

31.29

600

22.85

79.61

102.47

34.08

700

24.62

83.43

108.05

36.30

800

26.20

86.95

113.15

38.06

900

27.60

90.24

117.83

39.45

1000

28.84

93.31

122.15

40.59

1100

29.95

96.21

126.16

41.52

1200

30.95

98.95

129.90

42.28

1300

31.84

101.54

133.38

42.91

1400

32.66

104.00

136.38

43.45

1500

33.40

106.35

139.74

43.89

TABLE 3: POTENTIAL BARRIER AND THERMAL ENERGY OF 5-IODO URACIL Temperature Deg. Kalvin

POT BARRIER

THERMAL ENERGY

200

16.23

0.403

273

11.89

0.345

300

10.82

0.330

400

8.11

0.284

500

6.50

0.255

600

5.41

0.232

700

4.64

0.215

800

4.10

0.201

900

3.61

0.190

1000

3.25

0.180

1100

2.95

0.172

1200

2.70

0.165

1300

2.50

0.158

1400

2.32

0.153

1500

2.16

0.147

FIG. 7: ENTHALPY / HEAT CAPACITY

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 99

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

FIG. 8: ENTROPY / FREE ENERGY

FIG. 9: THERMAL ENERGY/ POTENTIAL BARRIER REFERENCES [1].

[2].

[3].

[4].

[5].

[6].

[7].

[8].

[9].

[10].

[11].

[12].

[13].

[14].

[15].

[16].

[17].

[18].

[19].

[20].

[21].

[22].

[23].

M. Itto, R. Shimada, T. Kurashi and Mizushima, J. Chem. Phys. (1956), 25, 597 R. C. Lord, A.J. Marshan and F.A.Miller, spectrochim, Acta (1957), 9, 113.

J.D. Simmons and K.K. Innes, J. Molec. Spectrosc. (1964), 13, 435.

G. Sbrana, G. Adenbri and S. Califano, Spectrochim. Acta (1966), 22, 1831.

Y.A. Sabrana, Spectrochi., Acta (1974), 30A, 1801.

F.M. Nejad and H.D. Stidham, Spectrochim. Acta (1975), 31A, 1433.

R.K. Goel, N.K. Sanyal and S.L. Srivastava, Indian J. Pure and Appl. Phys. (1976),14, 842.

E. Allenstein, P. Kimele, J. Weldin and W. Podszun Spectrochim Acta (1977), 33A,189.

V.A. Job and S.B. Kartha, Proc. Indian Acad. Sci(1977), 85A, 476.

R.K. Goel and S.N. Sharma, Indian J. Pure and Appl. Phys. (1980), 18, 195.

G.J. Thomas Jr. and Y. Kyogoku, “Infrared and Raman Spectroscopy” (edited by E.G. Brame, Jr. and J.G. Graselli) Dekker, New York (1977).

R.C. Lord and J. Thomas, Spectrochim Acta (GB) (1967) 23A. G.I. Angel, J. Chem. Soc. (GB), (1961), 504.

M. Tsuboi and Y. Kyobogu, Advances in chemical Phys. (1964), 7, 435.

L.J. Bellamy, “Infrared spectra of complex molecules” Chapman and Hall. London (1975).

Y. Nishamura, M. Tsuboi, S. Kato and K. Morokuma, J. Am. Chem. Soci(1981), 103, 1354.

A.J. Barnnes, M.A. Stuckey and L. Legall, spectrochim Acta (1984), 40A, 419 Seema, Ph.D. Thesis, “Electronic and vibrational spectral studies of some biological molecules”, (1991).

S.P. Gupta, S.K. Sharma and R.K. Goel, Acta Physica Polonica (1983), 64A, 129.

K.K. Innes. J.P. Byrne and I.G. Ross, J. Molec. Specrosc. (1967), 22, 125.

R.J. Jakobsen and E.J. Brewer, J. Appl. Spectrosc. (1962), 16, 32.

S.B. Kartha, Spectrochim Acta. (1974), 30A, 1943.

R.C. Lord and G.J. Thomas Jr. Spectrochim. Acta (1967), 23A, 2551.

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 100

Pradeep Kumar Sharma, International Journal of Emerging Technologies in Computational and Applied Sciences, 11(1), December 2014February 2015, pp. 94-101

[24].

[25].

[26].

[27].

[28].

[29].

[30].

[31].

[32].

[33].

[34].

[35].

[36].

[37].

[38].

[39].

[40].

[41].

[42].

[43].

[44].

[45].

[46].

[47].

[48].

M. Tsuboi, S. Takahashsi and I. Harada, “Physico Chemical Properties of Nucleic Acids” (Vol. 2, Chapter-II, Academic Press, New York, (1973).

C.N.R. Rao, “Chemical applications of infrared Spectroscopy” Academic Press, New York (1963).

H. Susi and J.S. Ard., Spectrochim, Acta, (1971), 27A, 1549.

H. Susi and J.S. Ard., Spectrochim. Acta, (1974), 30A, 1843.

N.K. Sanyal, R.K. Goel and S.L. Srivastava, Indian J. Pure and Appl. (1976), 18, 842.

R.K. Goel and S.K. Gupta, Indian J. Pure and Appl. Phys. (1981), 19, 501.

S.M. Pandey and S.J. Singh, Indian J. Phys. (1974), 48, 961.

J.C. Evans, Spectrochim, Acta (1960), 16, 1382.

T.A. Kletz and W.C. Price, J. Chem. Soci., (1947), 644.

N.K, Sanyal, S.L. Srivastava and D.N. Verma, Indian J. Phys. (1976), 50, 865.

E.E. Mooney, Spectrochim, Acta, (1968), 19, 877.

U.P. Tewari and A.K. Srivastava, Indian J. Chem., (1988), 27A, 657.

B.S. yadav and S.D. Sharma, J. Indi. Council of Chem. (1988), IV, 3.

Yogesh Gupta, Ph.D. Thesis “Electronic and Vibrational spectra studies of some substituted pyrimidine” Meerut University, (1992).

D.G.O. Sullivan and P.W. Sadler, J. Chem, Soc. (1959), 876.

M. Szczesniak, M. J. Nowak, K. Szczesniak and W.B. Person, Spectrochim, Acta, (1985), 41A, 237.

R.J. Jakobsen, Spectrochim Acta (1965), 21, 433.

P.K. Mallic and S.B. Banerjee, Indian J Pure and Appl. Phys. (1973), 11, 609.

R.S. Becker, A.B.F. Duncan, F.A. Matsen, D.R. Scott and W.West “Chemical Applications of Spectroscopy” John Wiley and Sons, New York (1968), 232.

G. Herzberg, “Molecular Spectra and Molecular Structure” Vol. II, Infrared and Raman Spectra, D. Van, Nostrand, Princeton N.J. (1966), 511.

K. S. Pitzer and W.D. Gwinn, J. Chem. Phys. (U.S.A.) (1942), 10, 428.

N.L. Owen and R.E. Hester, Spectrochim Acta (1969) 25A, 343.

C.L. Chatterjee, P.P. Garg and R.M.P. Jaiswal, Spectrochim. Acta (1978), 34A, 943 R.K. Goel and M.L. Agarwal, Spectrochim Acta (1982), 38A, 583.

R.K. Goel and M.L. Agarwal, J.De Chim. Phys., (1982), 79, 765.

IJETCAS 15-149; © 2015, IJETCAS All Rights Reserved

Page 101