Optimization through Bio Inspired Algorithms in Wireless Sensor Network: Survey and Future Directio

Page 1

Volume 2, Spl. Issue 2 (2015)

e-ISSN: 1694-2310 | p-ISSN: 1694-2426

Optimization through Bio Inspired Algorithms in Wireless Sensor Network: Survey and Future Directions Chandni1, Anjali Bharti2, Jyoti3 ECE Deptt., BUEST, Baddi, India 1

Chandni.smiley08@gmail.com, 2anjalibharti312@gmail.com, 3Jyoti.student@gmail.com

Abstract—Wireless sensor networks (WSNs) are the networks of tiny nodes used for analyzing the target area. Developers of Wireless Sensor Networks face a number of challenges that arise from communication link failures and memory constraints, and limited source of energy. Many issues in WSNs are formulated as multidimensional optimization problems, and approached through bio-inspired techniques. This paper presents some biologically inspired optimization algorithms like Particle swarm optimization, Bee Colony Optimization, Ant Colony Optimization. For the past few years, numerous research efforts have been made in this particular area. These bio inspired algorithms have been applied to address WSN issues such as optimal deployment of nodes, their localization, network clustering and dataaggregation. Keywords—Wireless Sensor Network, Optimization, Bio Inspired Algorithm, PSO, ACO, BCO

I. INTRODUCTION Wireless sensor network (WSN) is a large collection of sensor nodes that are spatially deployed all around the region to be analyzed and they manage themselves to form a network which is capable of performing collaborative work for performing the general functionality.

Industrial monitoring, Patient monitoring and Environmental monitoring [1]. WSNs have incredible research interest due to their recent applications in Civil and Military field. Gathering of sensor data, transmission and controlling depending on the applications is in-built in these systems. The main objective of this paper is to present an overview of WSN Optimization using a number of evolutionary models and to discuss the existing papers related to this field. A. Design challenges for WSN There are several features of WSNs that [distinguish them from traditional wireless ad hoc networks. First, WSNs have specific data flow patterns in the form of multicast (one-to-many) and converge-cast (many to-one) trees [2]. Second, WSNs are usually made up of a number of small nodes equipped with low end processors , limited non-rechargeable battery, small bandwidth links and small memory. As a result of which, WSN protocol designers face strict constraints on the use and the availability of sensor node resources [3]. Third, the majority of applications for Wireless Sensor Networks require the deployment of the nodes in huge amount, ranging from thousands to millions. Hence, the scalability of the used protocols is also a major issue [4]. Fourth, individual sensor nodes can themselves produce huge amounts of data. The data transmission from the sensor nodes to a common sink node would consume a huge amount of energy, bandwidth, and power. This matter can be resolved by detecting and filtering the redundant information to reduce the network traffic. There are some specific characteristics of WSNs that the routing protocols must have in order to allow their use in real-world applications. 1) Memory requirements and minimal computation. 2) Automaticity and self-organization. 3) Energy efficiency. 4) Scalability. 5) Support for in-network data aggregation.

WSN is built with few to thousands number of nodes and each node has a radio transceiver system with an antenna, a microcontroller, an electronic circuit used for interfacing with sensors and an energy source. WSN applications can be classified as Tracking and Monitoring. Tracking includes enemy tracking, Human or Traffic tracking and monitoring involves security detection in Military, animal monitoring, business monitoring,

B. Need for Optimization in WSN Optimization is the process, act or methodology of obtaining the best results under a given condition. Network optimization is a very critical task and a number of optimization techniques are used to achieve desired goals in networking. Application requirement, Energy efficiency and cost are the challenges that are to be considered while designing a Wireless Sensor Network. To design efficient WSN, the optimization of both hardware and software is

243

BUEST, Baddi

Fig.1. Architecture of a general wireless sensor network

RIEECE -2015


Volume 2, Spl. Issue 2 (2015)

e-ISSN: 1694-2310 | p-ISSN: 1694-2426

required. There are several optimization algorithms to sort out different problems. Choosing a proper algorithm according to the requirement is very important in any optimization technique. The deployment of the sensor nodes is done either inside the target area or very close to it [5]. It is not feasible to replace or recharge the batteries used in these nodes since the sensor nodes are deployed in hostile target areas. Thus, the battery life of the sensor nodes end which leads to the end of network itself. Efficient use of battery energy is required to enhance the lifetime of the network. The sensor network protocols have to emphasize more primarily on power conservation issues. Other important requirements for a WSN are, achieving high quality QoS, low bandwidth, limited processing and storage in sensor node [6]. These are the different issues of WSN which are directly related to the problem of optimization. Sensor networks as well as biological systems need to adapt themselves according to the varying environmental conditions including the ability to selforganize, scalability and to provide robust operation for the long life of the sensor network. [7].

e.g., to perform olfactory sensing [10] or tasting [11] has attracted the attention of a number of scientists. A number of survey papers based on bio inspired networking and communication protocols and algorithms have been published recently [12]. These survey papers provide a brief review on the current bio inspired thinking in networking. Swarm intelligence is the most important area of bio inspired network engineering [13]. Two important approaches which are based on ant colony, called ant colony optimization (ACO), described by Dorigo et al. [8] and based on bird flocking, called particle swarm ACO: Ant Colony Optimization PSO: Particle Swarm Optimization BCO: Bee Colony Optimization BO: Bee Optimization BSO: Bee Swarm Optimization ABC: Artificial Bee Colony HBCO: Hybrid Bee Colony Optimization

Fig.3. Hierarchical representation of Evolutionary Algorithms

optimization (PSO) introduced by Kennedy and Eberhart [9] have been proposed. Both approaches have been studied by many researchers and scientists and their updated research work have been described and applied to solve real-world problems in different areas.

Fig.2. A simple optimization process

II. APPROACHES TO BIO INSPIRED NETWORKING Sensor networks as well as biological systems need to adapt themselves according to the varying environmental conditions including the ability to self-organize, scalability and to provide robust operation for the long life of the sensor network. [8]. By looking carefully into nature, the observation which came into picture is that the dynamics of many biological systems and the laws governing them are based on a surprisingly few number of simple rules. These rules results in collaborative mechanisms for management of resource, task allocation, synchronization without the requirement of any central controlling element. Bio inspired (or natural) computing [9] represents a class of algorithms focused on the efficient computing in applications like optimization processes and pattern recognition. The research and development of bio inspired sensor systems with the goals of mimicking animal senses, BUEST, Baddi

RIEECE-2015

A. Particle Swarm Optimization Particle swarm optimization (PSO) is a populationbased optimization technique proposed by Kennedy and Eberhart [9]. PSO technique is basically inspired by the social behavior of bird flock searching for the food. PSO has been extensively applied to a number of engineering fields for optimization due to its unique searching mechanism, computational efficiency and easy implementation. In PSO, the term particle refers to population of members which are mass-less and volumeless (or with an arbitrarily small mass or volume) [14]. Each particle in the flock represents a solution in a highdimensional space with four vectors, its velocity, the best position found so far, its current position, the best position found by its neighborhood particle and adjusts its position in the search space based on the best position reached by itself (pbest) and on the best position reached by its neighborhood particle (gbest) during the search process. In 244


Volume 2, Spl. Issue 2 (2015)

each iteration step, each particle updates its position and velocity according to these equations: xik+1= xik+ vik+1 vik+1= vik +c1r1(pik - xik)+ c2r2(pgk- xik) xik : indicates Particle position vik : specifies Particle velocity pik : specifies Best "remembered" position c1 c2 : indicates cognitive and social parameters, r1 r2 : are random numbers between 0 and 1 Steps of PSO algorithm : 1) Initialize the swarm by assigning a random position in the target space to each particle. 2) Evaluate the fitness function for each particle in the flock. 3) Compare the particle’s fitness value with its pbest for each individual particle. If the current value is better than the pbest value, then set this value as the pbest and the current particle’s position, xi, as pi. 4) Identify the particle in the flock that has the best fitness value. The value of its fitness function is identified as guest and its position as pg. 5) Update the velocities and positions of all the particles using step (1) and (2). 6) Repeat steps 2–5 until a stopping criterion is met (e.g., maximum number of iterations or a sufficiently good fitness value). B. Ant Colony Optimization Ant Colony Optimization ACO is among the most successful swarm based algorithms proposed by Dorigo & Di Car [15]. It is a meta heuristic inspired by the foraging behavior of ants, and, the process known as stigmergy, the term introduced by Grasse in 1959. It refers to the indirect communication amongst a self-organizing emergent system via individuals. The most important aspect of the collaborative behavior of several ant species is their ability to calculate the shortest paths between the nest and the food sources by analyzing the pheromone trails [16]. Then, ants choose a path to follow by a probabilistic decision determined by the amount of pheromone: the stronger is the pheromone trail, the higher is the desirability of a particular path. Because ants in turn laid the pheromone trail on the path they are following, this behavior leads to a self-reinforcing process leading to the formation of routes marked by high concentration of pheromone. By modeling and simulating the nest building, ant foraging behavior and selfassembling etc. algorithms can be designed that could be further used for complex, combinatorial optimization problems. The first ant algorithm, named Ant System (AS), was developed by Dorigo et al. [17] and implemented successfully. The ACO Meta heuristic was developed by Dorigo & Di Caro to generalize the overall procedure for solving complex problems by approximate solutions based on the generic behavior of ants. ACO is designed into three main functions as follows: 1. AntSolutionsConstruct 2. Pheromone Update 3. DeamonActions

245

e-ISSN: 1694-2310 | p-ISSN: 1694-2426

An alternative approach, called the ant colony system (ACS) has been introduced by Dorigo and Gambardella to further improve the performance of ant system. It is based on four modifications of ant system: a different transition rule, a different pheromone trail update rule, the use of local updates of pheromone trail to favor exploration, and the use of candidate list to restrict the choice [18].

Fig.4. Representation of ACO Algorithm With Respect to Ants Foraging Behavior

C. Bee Colony Optimization It is a part of Swarm Intelligence (SI) where the workers/members/agents to the group are honey bees. They communicate with each other by a mechanism called “Waggle Dance”. They exchange important information regarding the location of rich food source. BCO is the Swarm Intelligence system where the low level agent of the system is the honeybee [18]. This name (BCO) is given to the honey Bees because of their collective food foraging behavior. The bee system is a standard example of organized team work, labor division, simultaneous task performance, well coordinated interaction well coordinated interaction and specialized individuals. In a honey bee colony there are different types of honey bees. There is a queen bee, many male drone bees and many thousands of worker bees. Types of bees: The Queen’s responsibility is to lay eggs so that new colonies could be formed. The Drones are the male of the hive and are responsible to mate with the Queen Bee. They are discarded from the colony during their down fall. The worker bees are the females of the hive [19]. They are the important building blocks of the hive. They build the honey bee comb, clean it, maintain it, guard it, and feed the queen and drones. Apart from these duties, the main job of the worker bee is to search and collect the food. There are two types of worker bees namely scout bees and forager bees. Both of them are responsible for the collection of food but they have different roles. Scout Bee’s Role: The Scout bees fly randomly all around and search for food. They come back to their hive after their exhaustion of energy and distance limits. Upon arrival to their hive the

BUEST, Baddi

RIEECE -2015


Volume 2, Spl. Issue 2 (2015)

e-ISSN: 1694-2310 | p-ISSN: 1694-2426

scout bee shares their experience and a lot of important information with the forager bees. III. CONCLUSION AND FUTURE WORK Development of effective optimization algorithm is the key to improve the utilization of the limited resources of WSNs like energy, bandwidth, computational power etc. A number of bio-inspired algorithms have addressed a large number of issues such as design, deployment, localization, security, energy efficient routing and clustering, scheduling, data aggregation, and fusion etc. This paper provides an overview of a range of Bio inspired algorithms drawn from an evolutionary metaphor or natural phenomena like PSO, ACO, BCO etc. Although the presented algorithms address many issues associated with optimization in WSNs, some research questions remain relatively unexplored, such as QoS, energy efficiency and security. There is significant scope for future work in these areas. Realizing the importance of these issues in WSNs, the future work is focused on developing a framework which integrates QoS-awareness, energy efficiency and security for Wireless Sensor Networks. REFERENCES [1] Jennifer Yick, Biswanth Mukherjee, Dipak Ghosal, “Wireless Sensor network survey”, International Journal of Computer networks, Vol. 52, pp. 2292- 2330, 2008. [2] C. Raghavendra, K. Krishna, T. Znati, “Wireless Sensor Networks”, Springer-Verlag, 2004. [3] Kemal Akkaya, Mohamed Younis “A survey on routing protocols for wireless sensor networks”, IEEE Communication Magazine on Ad Hoc Networks, pp.325–349, 2005. [4] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, “A survey on sensor networks”, IEEE Communications Magazine, pp.102– 114, 2002.

BUEST, Baddi

RIEECE-2015

[5] Chandramouli H., Dr. Somashekar C Desai, K S Jagadeesh, Kashyap D Dhruve, “Elephant swarm Optimization for Wireless Sensor Networks – A cross Layer Mechanism”, IJCET, Vol 4 No.2, pp. 4560, 2013. [6] Md. Akhtaruzzaman Adnan, Mohammd Abdur Razzaque, Ishtiaque Ahmed, Ismail Fauzi Is nin, “Bio-mimic Optimization Strategies in Wireless Sensor Networks: A Survey”, Sensors, Vol.14, pp. 299235, 2014. [7] Kennedy, J.; Eberhart, R.C.; Shi, Y, “Swarm Intelligence”, Morgan Kaufmann Publishers,San Francisco, 2001. [8] Dorigo M, Colorni A, V Maniezzo, “Positive feedback as a search strategy”, Technical Report, pp. 91-016, 1991. [9] Kennedy J, Eberhart R,“Particle swarm optimization”, IEEE international conference on neural networks, pp 1942–1948, 1995. [10] Dressler, F., Akan, O.B, “Bio-inspired networking: from theory to practice”, IEEE Communication Mag., pp. 176–183, 2010. [11] Ehsani Zonouz,Xing,Vokkarane, V.M, Sun, “Application communication reliability of wireless sensor networks”, Wireless Sensor Systems, Vol. 5 No. 2, pp.58–67, 2015. [12] Dressler, F, Akan, O.B, “A survey on bio-inspired networking”, Computer Network, Vol. 54, pp.881–900, 2010. [13] Clerc, M. and Kennedy, J. “The particle swarm-explosion, stability, and convergence in a multidimensional complex space”, IEEE Transactions on Evolutionary Computation, Vol. 6, No. 1, pp. 5873. 2002. [14] Hongliang Ren, Max Q H Meng, “Bio Inspired approches for Wireless Sensor Networks”, IEEE Conference on Mechatronics and Automation A Survey”, pp. 762-768, 2014. [15] Bonabeau, E., Dorigo, M. and Theraulaz, “Swarm intelligence”, Oxford University Press, 1999. [16] R. Storn, K. Price, “Differential evolution – a simple and efficient heuristic for global optimization over continuous spaces”, Journal of Global Optimization, Vol. 11, pp. 341–359, 1997. [17] Dorigo, M., Maniezzo, & Colorni, “Ant System: Optimization by a colony of cooperating agents”, IEEE Transactions on Systems, Vol. 26, pp. 29–41, 1996. [18] A. Kaur and S. Goyal, “A Bee Colony Optimization Algorithm for Fault Coverage Based Regression Test Suite Prioritization”, International Journal of Advanced Science and Technology, Vol. 29, pp. 17-30, 2011. [19] P.Navrat, T. Jelinek, and L. Jastrzembska, “Bee hive at work: A problem solving, optimizing mechanism,” World Congress on Nature & Biologically Inspired Computing, pp. 122 - 127 , 2009.

246


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.