4 hdd2

Page 1

ISBN 974-229-997-2

ราคา 220 บาท

Strengthening The Hard Disk Industry in Thailand National Electornics and Computer Technology Center National Science and Technology Development Agency Ministry of Science and Technology 112 Thailand Science Park, Phahon Yothin Road, Klong Luang, Pathumthani 12120, THAILAND. Tel. +66(0)2-564-6900 Fax. +66(0)2-564-6901..2

การประมวลผลสัญญาณสำหรับการจัดเก็บข้อมูลดิจทิ ลั เล่ม 2 : การออกแบบวงจรภาครับ

โครงการเสริมสร้างความแข็งแกร่งให้กบั อุตสาหกรรมฮาร์ดดิสก์ไดรฟ์ในประเทศไทย ศูนย์เทคโนโลยีอเิ ล็กทรอนิกส์และคอมพิวเตอร์แห่งชาติ สำนักงานพัฒนาวิทยาศาสตร์และเทคโนโลยีแห่งชาติ กระทรวงวิทยาศาสตร์และเทคโนโลยี 112 อุทยานวิทยาศาสตร์ประเทศไทย ถนนพหลโยธิน ตำบลคลองหนึง่ อำเภอคลองหลวง จังหวัดปทุมธานี 12120 โทรศัพท์ 02-564-6900 โทรสาร 02-564-6901..2

ดร. ปิยะ โควินท์ทวีวฒ ั น์


ÊÓËÃѺ¤Ãͺ¤ÃÑÇ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì



¤Ó¹ÔÂÁ

¹Ñº ¨Ò¡» ¾.È. 2548 ÈÙ¹Âì à·¤â¹âÅÂÕ ÍÔàÅç¡·Ã͹ԡÊì áÅФÍÁ¾ÔÇàµÍÃì áËè§ªÒµÔ (๤෤) Êӹѡ§Ò¹ ¾Ñ²¹ÒÇÔ·ÂÒÈÒʵÃì áÅÐà·¤â¹âÅÂÕ áËè§ªÒµÔ (ÊÇ·ª.)

ä´é ÃѺ ÁͺËÁÒÂãËé à» ¹ à¨éÒÀÒ¾´Óà¹Ô¹§Ò¹ªØ´

â¤Ã§¡ÒÃàÊÃÔÁÊÃéÒ§¤ÇÒÁá¢ç§á¡Ãè§ãËéá¡èÍصÊÒË¡ÃÃÁÎÒÃì´´ÔÊ¡ìä´Ã¿ìã¹»ÃÐà·Èä·Ââ´ÂÁÕÇѵ¶Ø»ÃÐʧ¤ì ËÅÑ¡ ¤×Í ¡ÒáÃдѺ ¤ÇÒÁÊÒÁÒö·Ò§¡ÒÃá¢è§¢Ñ¹ ÃÐËÇèÒ§»ÃÐà·È áÅÐÊè§àÊÃÔÁ ãËé »ÃÐà·Èä·Â¤§ ¤ÇÒÁà» ¹¼Ùé¼ÅÔµÎÒÃì´´ÔÊ¡ìä´Ã¿ìáÅÐÊèǹ»ÃСͺ Íѹ´Ñº 1 ¢Í§âÅ¡µÅÍ´·ÈÇÃÃÉ˹éÒ à¾× ÍãËéºÃÃÅØ Çѵ¶Ø »ÃÐʧ¤ì ´Ñ§¡ÅèÒÇ à¹¤ à· ¤ ä´é ´Óà¹Ô¹§Ò¹ ¼èÒ¹ ¡Ô¨¡ÃÃÁ ÃٻẺ µèÒ§æ àªè¹ ¡Òà ʹѺʹع ·Ø¹ÇԨѠà¾× Í ¾Ñ²¹Òà·¤â¹âÅÂÕ ¡ÒüÅÔµ ã¹ÍصÊÒË¡ÃÃÁÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¡ÒÃʹѺʹع ·Ø¹ à¾× Í ¾Ñ²¹ÒºØ¤ÅÒ¡Ã ·Ñ §¡è͹ ËÅѧà¢éÒ·Ó§Ò¹ã¹ÍصÊÒË¡ÃÃÁÎÒÃì´´ÔÊ¡ìä´Ã¿ì ¡ÒþѲ¹ÒËèǧâ«èÍØ»·Ò¹ ÃÇÁ¶Ö§¡ÒþѲ¹Ò ¹âºÒÂ

ÁҵáÒÃ

áÅÐâ¤Ã§ÊÃéÒ§¾× ¹°Ò¹·Õ à¡Õ ÂÇ¢éͧ¡ÑºÍصÊÒË¡ÃÃÁÎÒÃì´´ÔÊ¡ìä´Ã¿ì

ÍÂèÒ§äáçµÒÁ

๤෤µÃÐ˹ѡ ´Õ ÇèÒ ¹Í¡à˹×Í ¨Ò¡¡Ô¨¡ÃÃÁµèÒ§æ àËÅèÒ¹Õ áÅéÇ ¡Ô¨¡ÃÃÁ·Õ ÊӤѭ ÍÕ¡ ÃÙ»áººË¹Ö § ¤×Í ¡ÒöèÒ·ʹͧ¤ì ¤ÇÒÁÃÙé áÅÐà·¤â¹âÅÂÕ µèÒ§æ ·Õ à¡Õ ÂÇ¢éͧ¡Ñº ÍصÊÒË¡ÃÃÁÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì à¾ÃÒСÒà ´Óà¹Ô¹§Ò¹ã´æ ÂèÍÁ¡èÍ ãËé à¡Ô´ ͧ¤ì ¤ÇÒÁÃÙé «Ö § ͧ¤ì ¤ÇÒÁÃÙé µèÒ§æ àËÅèÒ¹Õ ËÒ¡äÁè ä´é ¶Ù¡ ÃǺÃÇÁËÃ×Í ¨Ñ´ à¡çºÍÂèҧ໠¹Ãкº áÅÐäÁèä´éà¼Âá¾ÃèÍÍ¡ä» ÂèÍÁà·èҡѺÇèÒ ¡ÒôÓà¹Ô¹§Ò¹·Õ ¼èÒ¹ÁÒ¨Ðà¡Ô´»ÃÐâª¹ì µèÍ ¤¹ ·Ó§Ò¹ à¾Õ§ ¡ÅØèÁ Ë¹Ö § à·èÒ¹Ñ ¹ áµè ËÒ¡ ͧ¤ì ¤ÇÒÁ ÃÙé áÅÐ »ÃÐʺ¡Òóì àËÅèÒ¹Ñ ¹ ¶Ù¡ ÃǺÃÇÁ áÅÐ à¼Âá¾Ãè ÊÙè Êѧ¤Áã¹Ç§¡ÇéÒ§ â´Â੾ÒÐÍÂèÒ§ÂÔ § ¡ÒèѴ·Óà» ¹ ˹ѧÊ×Í/µÓÃÒ «Ö § ·Ø¡¤¹ÊÒÁÒöÈÖ¡ÉÒ àÃÕ¹ÃÙé áÅФ鹤ÇéÒàÃ× Í§ÃÒǵèÒ§æ à¾Ô ÁàµÔÁ ÂèÍÁà·èҡѺÇèÒͧ¤ì¤ÇÒÁÃÙé´Ñ§¡ÅèÒǨСèÍãËéà¡Ô´»ÃÐ⪹ìáÅÐ ¼Å¡ÃзºµèÍÊѧ¤Áâ´ÂÃÇÁµèÍä» ÍÂèÒ§äáçµÒÁ¡ÒÃà¢Õ¹˹ѧÊ×Í/µÓÃÒäÁèãªèàÃ× Í§§èÒ à¾ÃÒмÙéà¢Õ¹µéͧ·ØèÁà··Ñ §àÇÅÒáÅФÇÒÁÁҹР¾ÂÒÂÒÁÍÂèÒ§Áҡ㹡ÒÃÃǺÃÇÁ¤ÇÒÁ¤Ô´áÅТéÍÁÙŵèÒ§æ ¨¹ÊÒÁÒöà¢Õ¹໠¹Ë¹Ñ§Ê×Í/µÓÃÒä´éÊÑ¡àÅèÁ


Ë¹Ö § ¼Á¨Ö§¢Íª× ÁªÁ㹤ÇÒÁÁØè§ÁÑ ¹¢Í§¼Ùéà¢Õ¹ ¤×Í ´Ã.» ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì «Ö §à» ¹¼Ùé·Õ ÁÕ¤ÇÒÁÃÙé ¤ÇÒÁ ÊÒÁÒö áÅФÇÒÁªÓ¹Ò­·Ò§´éÒ¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ìà» ¹ÍÂèÒ§´Õ â´Â ´Ã.» ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì ä´é·ØèÁà·àÇÅÒáÅФÇÒÁÊÒÁÒö¨¹à¢Õ¹˹ѧÊ×ÍàÃ× Í§ ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ ÊÓËÃѺ¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ àÅèÁ 1 áÅÐàÅèÁ 2 ¹Õ ä´éÊÓàÃ稾ÃéÍÁ¡Ñ¹·Õà´ÕÂÇ·Ñ § 2 àÅèÁ ¹Í¡¨Ò¡¹Õ ¼Á Âѧ¢Íª× Áª¹ã¹à¨µ¹ÒÃÁ³ì¢Í§ ´Ã.» ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì «Ö §¹Í¡¨Ò¡µéͧ¡ÒöèÒ·ʹͧ¤ì¤ÇÒÁÃÙéáÅÐ »ÃÐʺ¡ÒÃ³ì ¡ÒÃÃèÇÁ§Ò¹¡Ñº ÍصÊÒË¡ÃÃÁÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ã¹µèÒ§»ÃÐà·ÈáÅéÇ ¡ÒÃà¢Õ¹˹ѧÊ×Í ·Ñ § 2 àÅèÁ¹Õ ´Ã.» ÂÐ ÂѧÁØè§à¹é¹¡ÒöèÒ·ʹ´éÇÂÀÒÉÒ áÅÐÇÔ¸Õ¡ÒÃ·Õ à¢éÒ㨧èÒ à¾× ÍãËé¼ÙéÍèÒ¹·Ø¡·èÒ¹ÊÒÁÒö ·Ó¤ÇÒÁà¢éÒã¨à¹× ÍËҢͧ˹ѧÊ×Í ·Ñ § 2 àÅèÁ ä´é ÍÂèÒ§¤Ãº¶éǹáÅÐÊÁºÙóì áÅÐàª× Í ä´é ÇèÒ Í§¤ì ¤ÇÒÁÃÙé ¨Ò¡Ë¹Ñ§Ê×Í·Ñ § 2 àÅèÁ¹Õ ¨ÐªèÇ¡ÃеØé¹áÅÐÊè§àÊÃÔÁãËéÁÕ¡ÒÃÈÖ¡ÉÒ ÇԨѠáÅФ鹤ÇéÒã¹ÈÒʵÃì´éÒ¹¡Òà »ÃÐÁÇżÅÊÑ­­Ò³ãËéá¾ÃèËÅÒÂÂÔ §¢Ö ¹µèÍä»

´Ã.¾Ñ¹¸ìÈÑ¡´Ô ÈÔÃÔÃѪµ¾§Éì ¼ÙéÍӹǡÒà ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìáÅФÍÁ¾ÔÇàµÍÃìáËè§ªÒµÔ 3 Á¡ÃÒ¤Á 2550


¤Ó¹Ó

ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì (hard disk drive) à» ¹ ¼ÅÔµÀѳ±ì ·Ò§ÍÔàÅç¡·Ã͹ԡÊì ·Õ ÁÕ ¤ÇÒÁÊӤѭ µèÍ ÀÒ¤¡ÒÃÊè§ ÍÍ¡¢Í§»ÃÐà·Èä·Â ¨Ò¡Ê¶ÔµÔ¢Í§¤³Ð¡ÃÃÁ¡ÒÃÊè§àÊÃÔÁ¡ÒÃŧ·Ø¹ (BOI) ã¹» 2547 ¾ºÇèÒ »ÃÐà·È ä·ÂÊÒÁÒöÊè§ÍÍ¡¢Í§ÍØ»¡Ã³ìÎÒÃì´´ÔÊ¡ìä´Ã¿ìáÅÐÊèǹ»ÃСͺä´éÊÙ§¶Ö§ 483,000 ÅéÒ¹ºÒ· áÅСèÍ ãËéà¡Ô´¡ÒèéÒ§§Ò¹ÁÒ¡¡ÇèÒ 100,000 ÍѵÃÒ ´Ñ§¹Ñ ¹ à¾× Íà» ¹¡ÒÃÊÃéÒ§¤ÇÒÁà¢éÁá¢ç§ãËé¡ÑºÍصÊÒË¡ÃÃÁ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ã¹»ÃÐà·È «Ö § »ÃСͺ仴éǺÃÔÉÑ· ¼Ùé¼ÅÔµ ªÑ ¹¹Ó¢Í§âÅ¡ 4 ÃÒ ¤×Í ºÃÔÉÑ· «Õà¡· à·¤â¹âÅÂÕ (»ÃÐà·Èä·Â) ¨Ó¡Ñ´, ºÃÔÉÑ· àÇÊà·ÔÃì¹ ´Ô¨Ô·ÑÅ (»ÃÐà·Èä·Â) ¨Ó¡Ñ´, ºÃÔÉÑ· ÎÔµÒªÔ â¡ÅàºÔŠʵÍàŨ à·¤â¹âÅÂÕ ¨Ó¡Ñ´, áÅкÃÔÉÑ· ¿Ù ¨Ôµ «Ö ¡ÃзÃǧÇÔ·ÂÒÈÒʵÃì áÅÐà·¤â¹âÅÂÕ â´ÂÊӹѡ§Ò¹ ¾Ñ²¹ÒÇÔ·ÂÒÈÒʵÃìáÅÐà·¤â¹âÅÂÕáËè§ªÒµÔ (ÊÇ·ª.) ÃèÇÁ¡Ñº¼Ùéà¡Õ ÂÇ¢éͧ·Ñ §¨Ò¡ÀÒ¤àÍ¡ª¹ áÅÐʶҺѹ ¡ÒÃÈÖ¡ÉÒä´é ´Óà¹Ô¹¡ÒèѴµÑ § ¤³Ð¡ÃÃÁ¡ÒþѲ¹ÒÇÔÊÒË¡Ô¨ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¢Ö ¹ ÀÒÂãµé ¡ÒôÓà¹Ô¹¡Òà ¢Í§ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìáÅФÍÁ¾ÔÇàµÍÃìáËè§ªÒµÔ (๤෤) â´ÂÁÕÇѵ¶Ø»ÃÐʧ¤ìà¾× ͪèÇÂã¹ ¡ÒÃʹѺʹع áÅÐÊè§àÊÃÔÁ ¡ÒþѲ¹ÒÍصÊÒË¡ÃÃÁ´Ñ§¡ÅèÒÇã¹»ÃÐà·Èä·ÂãËé ÁÕ ¤ÇÒÁà¢éÁá¢ç§ â´Â·Ò§ ๤෤¨Ö§ ä´é ¨Ñ´àµÃÕÂÁá¼¹¡ÒÃÊè§àÊÃÔÁ ´éÒ¹µèÒ§æ äÇé ËÅÒÂâ¤Ã§¡Òà ä´éá¡è â¤Ã§¡ÒþѲ¹ÒËÅÑ¡Êٵà áÅкؤÅҡ÷ҧ´éҹ෤â¹âÅÂÕ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì, â¤Ã§¡ÒÃÃèÇÁŧ·Ø¹ 㹡ÒÃÇԨѠáÅоѲ¹Òà·¤â¹âÅÂÕ ÎÒÃì´´ÔÊ¡ìä´Ã¿ì, áÅÐâ¤Ã§¡ÒèѴµÑ §Ê¶ÒºÑ¹ÇÔ¨ÑÂáÅоѲ¹Ò·Ò§´éÒ¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì à» ¹µé¹ ˹ѧÊ×Í ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÊÓËÃѺ¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ àÅèÁ 2:

¡ÒÃÍ͡Ẻǧ¨ÃÀÒ¤

ÃѺ ä´é ¶Ù¡ à¢Õ¹¢Ö ¹ ÁÒ â´ÂÁÕ ¨Ø´ÁØè§ËÁÒÂà¾× Í ãªé à» ¹ ͧ¤ì ¤ÇÒÁÃÙé ãËÁè 㹡ÒÃʹѺʹع â¤Ã§¡ÒþѲ¹Ò ´éÒ¹µèÒ§æ à¾× ÍÊè§àÊÃÔÁÍصÊÒË¡ÃÃÁÎÒÃì´´ÔÊ¡ìä´Ã¿ì¢Í§»ÃÐà·Èä·Â ˹ѧÊ×ÍàÅèÁ¹Õ àËÁÒÐÊÓËÃѺ¼Ùé·Õ ÁÕ ¾× ¹°Ò¹·Ò§´éÒ¹¡ÒûÃÐÁÇżÅÊÑ­­Ò³´Ô¨Ô·ÑÅ (digital signal processing) áÅÐÃкºÊ× ÍÊÒôԨԷÑÅ (digital communication) ÃÇÁ·Ñ §¼Ùé·Õ ʹ㨷ҧ´éÒ¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì


à¹× ͧ¨Ò¡ à¹× ÍËÒÀÒÂã¹Ë¹Ñ§Ê×ÍàÅèÁ¹Õ ¨ÐµèÍà¹× ͧÁҨҡ˹ѧÊ×Í ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÊÓËÃѺ¡Òà ¨Ñ´ à¡çº ¢éÍÁÙÅ ´Ô¨Ô·ÑÅ àÅèÁ 1:

¾× ¹°Ò¹ ªèͧÊÑ­­Ò³ ÍèÒ¹ à¢Õ¹ à¾ÃÒÐ©Ð¹Ñ ¹ ¼Ùéà¢Õ¹ ¢Í á¹Ð¹Ó

ãËé ¼ÙéÍèÒ¹ÈÖ¡ÉÒ˹ѧÊ×Í àÅèÁ ·Õ 1 ãËé à¢éÒ㨡è͹·Õ ¨ÐÈÖ¡ÉÒ˹ѧÊ×Í àÅèÁ ¹Õ à¾× Í·Õ ¨Ðä´é à¢éÒã¨ã¹º·àÃÕ¹ ÁÒ¡ÂÔ § ¢Ö ¹ â´Âã¹Ë¹Ñ§Ê×Í àÅèÁ¹Õ ¨Ð͸ԺÒ¶֧ ËÅÑ¡¡ÒÃÍ͡Ẻǧ¨ÃÀÒ¤ÃѺ ¢Í§Ãкº¡ÒûÃÐÁÇżŠÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ã¹ÃٻẺ·Õ §èÒµèÍ ¡ÒÃà¢éÒã¨áÅÐÈÖ¡ÉÒ´éǵ¹àͧ ¹Í¡¨Ò¡¹Õ ¼Ùéà¢Õ¹ Âѧ ä´é ãªé ˹ѧÊ×Í àÅèÁ ¹Õ »ÃСͺ¡ÒÃÊ͹ÇÔªÒ ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÊÓËÃѺ ¡ÒèѴ à¡çº ¢éÍÁÙÅ ´Ô¨Ô·ÑÅ (signal processing for digital data storage) «Ö §ã¹¢³Ð¹Õ ä´éà» ´Ê͹ã¹ËÅÑ¡ÊÙµÃÇÔÈÇ¡ÃÃÁÈÒʵà ºÑ³±Ôµ ªÑ ¹» ·Õ 4 ¢Í§ÁËÒÇÔ·ÂÒÅÑ ÃÒªÀѯ ¹¤Ã»°Á áÅÐʶҺѹ à·¤â¹âÅÂÕ ¾ÃШÍÁà¡ÅéÒ ¾Ãй¤Ã à˹×Í

¢éÒ¾à¨éÒä´éà¢Õ¹˹ѧÊ×ÍàÅèÁ¹Õ ¢Ö ¹ÁÒâ´ÂÍÒÈÑ»ÃÐʺ¡Òóì·Ñ §ËÁ´·Õ ä´éÃѺµÑ §áµèàÃÔ Á¡ÒÃÈÖ¡ÉÒÃдѺ ªÑ ¹»ÃÔ­­ÒàÍ¡·Õ Georgia Institute of Technology »ÃÐà·ÈÊËÃÑ°ÍàÁÃÔ¡Ò ¹Í¡¨Ò¡¹Õ ¢éÒ¾à¨éÒÂѧà¤Â ·Ó§Ò¹ÇԨѠ·Ò§´éÒ¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ³ ÈÙ¹ÂìÇԨѠ¢Í§«Õà¡· (Sea gate research center) ·Õ àÁ×ͧ Pittsburgh »ÃÐà·ÈÊËÃÑ°ÍàÁÃÔ¡Ò à» ¹ ÃÐÂÐàÇÅÒ 1 » ·ÓãËé ¢éÒ¾à¨éÒ ÁÕ ¤ÇÒÁÃÙé áÅФÇÒÁªÓ¹Ò­·Ò§´éÒ¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì â´Â㹺··Õ 1 ¨Ð¡ÅèÒǶ֧ ¾× ¹°Ò¹·Ñ Ç仢ͧ¡Òúѹ·Ö¡ ÃкºààÁè àËÅç¡ à຺´Ô¨Ô·ÑÅ, ¡Ãкǹ¡ÒÃà¢Õ¹áÅÐÍèÒ¹¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì, áÅÐà຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ ÃкºààÁè àËÅç¡ «Ö § ¨Ð¶Ù¡ ¹Óä»ãªé 㹡Òà Í͡Ẻǧ¨ÃÀÒ¤ÃѺ ¢Í§Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì º··Õ 2 ¨Ð͸ԺÒ¶֧ ¤ÇÒÁÊӤѭ, ËÅÑ¡¡Ò÷ӧҹ, áÅСÒÃÍ͡Ẻ ¢Í§Ãкºä·ÁÁÔ § ÃԤѿàÇÍÃÕ (timing recovery) º· ·Õ 3 ¨Ð¡ÅèÒǶ֧ ¢Ñ ¹µÍ¹¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ (target) ààÅÐÍÕ¤ÇÍäÅà«ÍÃì (equalizer) ·Õ ãªé 㹪Ի ªèͧÊÑ­­Ò³ÍèÒ¹ (read channel chip) º··Õ 4 ¨Ð͸ԺÒ¶֧ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ PRML (partial response maximum likelihood) º··Õ 5 ¨Ð¡ÅèÒǶ֧¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´ (er ror event) à¾× Í·Õ ¨Ð¹Ó¼ÅÅѾ¸ì ·Õ ä´é ÁÒãªé »ÃѺ»Ãا »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ ÎÒÃì´´ÔÊ¡ì ä´Ã¿ì º··Õ 6 áÅк··Õ 7 ¨Ð͸ԺÒ¶֧ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML (noise predictive maximum likelihood) áÅÐǧ¨ÃµÃǨËÒ PDNP (pattern dependent noise predictive) µÒÁÅӴѺ «Ö §à» ¹Ç§¨ÃµÃǨËÒ·Ñ §ÊͧẺ¹Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒǧ¨ÃµÃǨËÒ PRML â´Â੾ÒÐ ÍÂèÒ§ÂÔ § àÁ× Í ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÁÕ ¤ÇÒÁ¨Ø ÁÒ¡ áÅк··Õ 8 «Ö § à» ¹ º·ÊØ´·éÒ¢ͧ˹ѧÊ×Í àÅèÁ ¹Õ ¨ÐáÊ´§


¢Ñ ¹µÍ¹¡ÒÃÍÍ¡à຺ÃËÑÊ RLL (run length limited) «Ö § à» ¹ ÃËÑÊ ÁÍ´Ù àŪѹ ·Õ ¹ÔÂÁãªé ã¹ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ˹ѧÊ×ÍàÅèÁ¹Õ ¨ÐäÁèÊÒÁÒö·ÓãËéÊÓàÃç¨¢Ö ¹ÁÒä´éàŶéÒËÒ¡¢Ò´ºØ¤¤ÅµèÒ§æ ·Õ ¤ÍÂãËé¤ÇÒÁªèÇÂàËÅ×Í áÅÐà» ¹¡ÓÅѧã¨ãËé¢éÒ¾à¨éÒµÅÍ´ÁÒ

¢éÒ¾à¨éҢ͡ÃÒº¢Íº¾ÃФسÍÒ¨ÒÃÂì·Ø¡·èÒ¹·Õ ãËé¤ÇÒÁÃÙéáÅФÓ

»ÃÖ¡ÉÒµÅÍ´ÃÐÂÐàÇÅÒ¡ÒÃÈÖ¡ÉÒ â´Â੾ÒÐÍÂèÒ§ÂÔ § Prof. John R. Barry áÅÐ Prof. Steve W. McLaughlin ÃÇÁ·Ñ §¹Ñ¡ÇԨѨҡÈÙ¹ÂìÇԨѢͧ«Õà¡· àªè¹ Dr. Erozan M. Kurtas, Dr. M. Fatih Er den, áÅÐ Dr. Inci Ozgunes ·Õ ãËéâÍ¡ÒÊ¢éÒ¾à¨éÒä´é·Ó§Ò¹ÇԨѷҧ´éÒ¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì áÅÐ·Õ äÁè ÊÒÁÒö¨ÐÅ×Á ä´é ¡ç ¤×Í ·Ø¡æ ¤¹ã¹¤Ãͺ¤ÃÑÇ ¢Í§¢éÒ¾à¨éÒ ä´éá¡è ¤Ø³ à¡ÕÂÃµÔ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, ¤Ø³¾ÃÃ³Õ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, ¤Ø³ÃѪ¹ÔÈ â蹡Ԩ, ¤Ø³Í¹Ø·Ñȹì â蹡Ԩ, ¤Ø³ ©ÑµÃªÑ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, ¤Ø³¡ÔµµÔÈÑ¡´Ô â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, áÅФسÈÔÃÊØ´Ò âÊÁ¹ÑÊ ¹Í¡¨Ò¡¹Õ ¢éÒ¾à¨éÒ¢Í ¢Íº¤Ø³ ÁËÒÇÔ·ÂÒÅÑ ÃÒªÀѯ ¹¤Ã»°Á·Õ ãËé¡ÒÃʹѺʹع áÅÐãËé ¤ÇÒÁÊдǡ¢éÒ¾à¨éÒ µÅÍ´ÃÐÂÐàÇÅÒ ã¹¡ÒÃà¢Õ¹˹ѧÊ×ÍàÅèÁ¹Õ ·éÒÂÊØ´¹Õ ¢éÒ¾à¨éÒä´é¾ÂÒÂÒÁÍÂèÒ§ÂÔ §ã¹¡ÒÃ·Õ ¨Ð·ÓãËé˹ѧÊ×ÍàÅèÁ¹Õ §èÒµèÍ¡ÒÃàÃÕ¹ÃÙé à¾× ÍãËé¼ÙéÍèÒ¹ ÊÒÁÒö·Ó¤ÇÒÁà¢éÒã¨ä´é ´éǵ¹àͧÍÂèÒ§ÃÇ´àÃçÇ áÅÐÁÕ »ÃÐÊÔ·¸Ô¼Å Ëҡ˹ѧÊ×Í àÅèÁ ¹Õ ÁÕ ¢éͺ¡¾Ãèͧ »ÃСÒÃã´ ¢éÒ¾à¨éÒ ÁÕ ¤ÇÒÁÂÔ¹´Õ áÅШѡ ¢Íº¾ÃФس ÂÔ § ËÒ¡·èÒ¹¼Ùéãªé ˹ѧÊ×Í àÅèÁ ¹Õ ¨ÐÊè§ ¢éͤԴàËç¹ áÅÐ ¤Óá¹Ð¹Ó·Õ à» ¹»ÃÐ⪹ìÊÓËÃѺ¡ÒûÃѺ»Ãا˹ѧÊ×ÍàÅèÁ¹Õ ÁÒ·Õ ÍÕàÁÅì piya@npru.ac.th à¾× Í·Õ ¢éÒ¾à¨éÒ ¨Ðä´é ´Óà¹Ô¹¡ÒûÃѺ»Ãا áÅÐá¡éä¢ã¹¡ÒþÔÁ¾ì ¤ÃÑ § µèÍä» ÊÓËÃѺ ¢éÍÁÙÅ ¢èÒÇÊÒõèÒ§æ à¡Õ ÂǡѺ ˹ѧÊ×Í àÅèÁ¹Õ áÅÐÍ× ¹æ ÊÒÁÒöµÔ´µÒÁä´é·Õ àÇçºä«µì http://home.npru.ac.th/∼t3058

´Ã.» ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì â»Ãá¡ÃÁÇÔÈÇ¡ÃÃÁâ·Ã¤Á¹Ò¤Á ÁËÒÇÔ·ÂÒÅÑÂÃÒªÀѯ¹¤Ã»°Á Á¡ÃÒ¤Á 2550



ÊÒúѭ

1

2

º·¹Ó

1

1.1

¾× ¹°Ò¹¢Í§¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡à຺´Ô¨Ô·ÑÅ . . . . . . . . . . . . . . . . . . .

1

1.2

à຺¨ÓÅͧ¢Í§Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì . . . . . . . . . . . .

3

1.3

¡Ãкǹ¡ÒÃà¢Õ¹

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

1.4

¡Ãкǹ¡ÒÃÍèÒ¹ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

1.5

à຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡ . . . . . . . . . . . . . . . . . .

9

1.5.1

à຺¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§

. . . . . . . . . . . . . . . . . . . .

11

1.5.2

à຺¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

. . . . . . . . . . . . . . . . . . . . . .

12

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15

1.6

ÊÃØ»·éÒº·

1.7

à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15

ä·ÁÁÔ §ÃԤѿàÇÍÃÕ

17

2.1

º·¹Ó

17

2.2

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

2.3

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

2.4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19

. . . . . . . . . . . . . . . . . . . . .

23

2.3.1

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊ鹢ͧǧ¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö § . . . . . . . . . . . . . .

23

2.3.2

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊ鹢ͧǧ¨Ã PLL Íѹ´Ñº·Õ Êͧ

. . . . . . . . . . . . . .

28

2.3.3

¡ÒÃËÒàÊé¹â¤é§ÃÙ»µÑÇàÍÊ

. . . . . . . . . . . . . . . . . . . . . . . . . .

30

»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä» . . . . . . . . . . . . . . . . .

34

ix


3

4

2.5

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺´Ô¨Ô·ÑÅ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37

2.6

àà¹Çâ¹éÁ¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕã¹Í¹Ò¤µ . . . . . . . . . . . . . . . . . . . .

38

2.7

ÊÃØ»·éÒº·

41

2.8

à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµààÅÐÍÕ¤ÇÍäÅà«ÍÃì

43

3.1

º·¹Ó

43

3.2

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.2.1

à§× ͹䢺ѧ¤Ñºà຺âÁ¹Ô¡ (h0

3.2.2

à§× ͹䢺ѧ¤Ñºà຺

3.2.3

à§× ͹䢺ѧ¤Ñºà຺¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ (H

. . . . . . . . . . . .

53

3.2.4

à§× ͹䢺ѧ¤Ñºà຺·ÒÃìà¡çµà©¾ÒÐ . . . . . . . . . . . . . . . . . . . . . .

53

3.3

¼Å¡Ò÷´Åͧ

3.4

ÊÃØ»·éÒº·

3.5

à຺½ ¡ËÑ´·éÒº·

h1 = 1

= 1)

47

. . . . . . . . . . . . . . . . . . . . .

49

. . . . . . . . . . . . . . . . . . . . . . . . .

52

TH

= 1)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

62

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

64

ǧ¨ÃµÃǨËÒ PRML

65

4.1

º·¹Ó

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

65

4.2

ÍÕ¤ÇÍäÅà«ÍÃì . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66

4.2.1

ÍÕ¤ÇÍäÅà«ÍÃìà຺¼ÅµÍºÊ¹Í§àµçÁ

. . . . . . . . . . . . . . . . . . . .

67

4.2.2

ÍÕ¤ÇÍäÅà«ÍÃìà຺¼ÅµÍºÊ¹Í§ºÒ§Êèǹ . . . . . . . . . . . . . . . . . .

68

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69

4.3.1

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

71

4.3.2

à༹ÀÒ¾à·ÃÅÅÔÊ

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

72

4.3.3

ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75

4.3.4

¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ . . . . . . . . . . . . . . . . . . .

80

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ . . . . . . . . . . . . . . . . . . . . . . .

81

4.3

4.4


4.4.1

5

6

7

ÊÃػǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

. . . . . . . . . . . . . . . . . . . . . . . . .

87

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

88

4.5

ÊÃØ»·éÒº·

4.6

à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

89

¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´

91

5.1

º·¹Ó

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

91

5.2

¤ÇÒÁËÁÒ¢ͧà˵ءÒóì¢éͼԴ¾ÅÒ´ . . . . . . . . . . . . . . . . . . . . . . . .

93

5.3

ÃÐÂзҧÂؤÅÔ´ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97

5.4

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

99

5.5

¼Å¡Ò÷´Åͧ

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

5.5.1

¡ÒÃÇÔà¤ÃÒÐËìÃÐÂзҧ·Õ ¹éÍÂÊØ´ . . . . . . . . . . . . . . . . . . . . . . . 104

5.5.2

¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§

5.6

ÊÃØ»·éÒº·

5.7

à຺½ ¡ËÑ´·éÒº·

SNReff

ààÅÐ BER

. . . . . . . . . . . . . . . . . 107

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

ǧ¨ÃµÃǨËÒ NPML

115

6.1

º·¹Ó

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

6.2

¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹ . . . . . . . . . . . . . . . . . . . . . 117

6.3

¡ÒÃËÒ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ

6.4

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML . . . . . . . . . . . . . . . . . . . . . . 120

6.5

¼Å¡Ò÷´Åͧ

6.6

ÊÃØ»·éÒº·

6.7

à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . 118

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

ǧ¨ÃµÃǨËÒ PDNP

137

7.1

º·¹Ó

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

7.2

¡ÒÃ¢Ö ¹ÍÂÙè¡Ñºà຺¢éÍÁÙŢͧÊÑ­­Ò³Ãº¡Ç¹

7.3

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP

. . . . . . . . . . . . . . . . . . . . 138

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140


8

7.4

ÍÑÅ¡ÍÃÔ·ÖÁ PS PDNP

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144

7.5

¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ PDNP

7.6

¼Å¡Ò÷´Åͧ

7.7

ÊÃØ»·éÒº·

7.8

à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

¡ÒÃÍÍ¡à຺ÃËÑÊ RLL

151

. . . . . . . . . . . . . . . . . . . . . . . 146

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

8.1

º·¹Ó

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

8.2

¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

8.3

¤ÇÒÁ¨Ø¢Í§ÃËÑÊ RLL à຺

(d, k)

(d, k)

. . . . . . . . . . . 153

. . . . . . . . . . . . . . . . . . . . . . . . . 154

(d, k)

8.3.1

ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ¢Í§ÃËÑÊ RLL à຺

. . . . . . . . . . . 155

8.3.2

ÍѵÃÒ¤ÇÒÁ˹Òàà¹è¹ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

8.4

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL . . . . . . . . . . . . . . . . . . . . . . . . . . 156

8.5

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

8.5.1

¡ÒÃËÒÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ . . . . . . . . . . . . . . . . . . . . . . 160

8.5.2

ÅӴѺ¢éÍÁÙÅ·Õ ÊÍ´¤Åéͧ¡Ñºà¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL à຺

(d, k)

. . 160

8.6

¢Ñ ¹µÍ¹¡ÒÃÍÍ¡à຺ÃËÑÊ RLL . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

8.7

µÑÇÍÂèÒ§ÃËÑÊ RLL à຺µèÒ§æ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

8.8

ÃËÑÊ

8.9

ÊÃØ»·éÒº·

(0, G/I)

ÊÓËÃѺªèͧÊÑ­­Ò³ PRML

. . . . . . . . . . . . . . . . . . . . 166

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

8.10 à຺½ ¡ËÑ´·éÒº·

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

Q

¡

µÒÃÒ§¿ §¡ìªÑ¹

¢

Êٵä³ÔµÈÒʵÃì·Õ ÊӤѭ

171

175

¢.1

µÃÕ⡳ÁÔµÔ (Trigonometric)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

¢.2

»ÃԾѹ¸ìäÁè¨Ó¡Ñ´à¢µ (Inde nite Integral) . . . . . . . . . . . . . . . . . . . . . . 176


¤

¡ÒÃËÒ͹ؾѹ¸ì¢Í§àÇ¡àµÍÃìààÅÐàÁ·ÃÔ¡«ì

177

§

¤ÓÈѾ·ìà·¤¹Ô¤

179

ºÃóҹءÃÁ

191

´Ãê¹Õ

199



ÊÒúѭÃÙ»

1.1

ËÅÑ¡¡ÒÃ¾× ¹°Ò¹¢Í§¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡ . . . . . . . . . . . . . . . . . . . .

2

1.2

Ẻ¨ÓÅͧ·Ñ Ç仢ͧÃкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì [6] . . . . . . .

4

1.3

ÊÑ­­Ò³¾ÑÅÊì à»ÅÕ Â¹Ê¶Ò¹Ð ÊÓËÃѺ ¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻ á¹ÇµÑ § . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

1.4

¼ÅµÍºÊ¹Í§ä´ºÔµ ÊÓËÃѺ¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ § . . .

9

1.5

¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ ÊÓËÃѺ ¡Òúѹ·Ö¡ (a) Ẻá¹Ç ¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ § . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10

1.6

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§ . . . . . . . . . . . . . . . . . . . . . . . . .

11

1.7

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

13

1.8

¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ ẺµèÒ§æ ÊÓËÃѺ ªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ (a)

. . . . . . . . . . . . . . . . . . . . . . . . . .

Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §

. . . . . . . . . . . . . . . . . . . . . . . .

14

2.1

ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ¹ÔùÑ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19

2.2

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ ¾ÃéÍÁ¡Ñºä·ÁÁÔ §ÃԤѿàÇÍÃÕẺÍØ»¹Ñ . . . . . . . .

19

2.3

¤èÒÁÒ¡ÊØ´¢Í§

2.4

(a) ¤èÒ

αC

áµèÅÐ

d,

2.5

α

·Õ Âѧ¤§·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾ ÊÓËÃѺáµèÅФèÒ

·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾áÅÐÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒÀÒÂã¹

(a) ¼ÅµÍºÊ¹Í§¢Ñ ¹ºÑ¹ä´¢Í§Ãкº áÅÐ (b) ¼ÅµÍºÊ¹Í§¢éͼԴ¾ÅÒ´ÊÓËÃѺ

d=

14T

áÅФèÒ

α

µèÒ§æ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

25

ºÔµ ÊÓËÃѺ

30T

ÊÓËÃѺ¤èÒ

d

C

. . . . . .

¨Ò¡ 0 ¶Ö§

áÅÐ (b) ¼ÅµÍºÊ¹Í§¢Í§ÃкºàÁ× Íãªé

α100

d

26

27


E(z)

ºÔµ àÁ× Íãªé

d = 14

áÅÐ

αC

¢¹Ò´ÁÒ¡ÊØ´¢Í§

2.7

Ẻ¨ÓÅͧÊÓËÃѺ¡ÒÃËÒàÊé¹â¤é§ÃÙ»µÑÇàÍÊ

2.8

àÊé¹â¤é§ ÃÙ» µÑÇ àÍʢͧǧ¨Ã M&M TED ÊÓËÃѺ ªèͧÊÑ­­Ò³ PR4 ·Õ ãªé ä·ÁÁÔ §

2.9

ËÅѧ¨Ò¡·Õ ¢éÍÁÙżèÒ¹ä»

C

2.6

. .

29

. . . . . . . . . . . . . . . . . . . . .

31

ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

32

µÑÇÍÂèÒ§ÅѡɳТͧä«à¤ÔÅÊÅÔ» . . . . . . . . . . . . . . . . . . . . . . . . . . .

33

2.10 (a) ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒẺ RMS

σ² /T

ªèͧÊÑ­­Ò³ÍØ´Á¤µÔẺ PR4 ·Õ ÁÕ¤èÒ ¤ÇÒÁ¶Õ )

áÅÐ (b) »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ BER ÊÓËÃѺ

σw /T

µèÒ§æ ¡Ñ¹ (àÁ× ÍÃкºäÁèÁÕÍͿ૵·Ò§

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35

2.11 »ÃÐÊÔ·¸ÔÀÒ¾ BER ¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Ç仢ͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ Ẻ PR4 ÊÓËÃѺ

σw /T = 0.5%

áÅÐÍͿ૵·Ò§¤ÇÒÁ¶Õ 0.2%

. . . . . . . . .

36

2.12 â¤Ã§ÊÃéÒ§¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ»ÃÐÁÒ³¤èÒ㹪èǧ . . . . . . . . . . . . . . . .

38

2.13 »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺµèÒ§æ ¢Í§ªèͧÊÑ­­Ò³ÍØ´Á¤µÔẺ PR4 . . .

39

2.14 ÍѵÃÒ¡ÒÃÅÙèà¢éҢͧä·ÁÁÔ §ÃԤѿàÇÍÃÕẺµèÒ§æ àÁ× Íãªé

3.1

α50

·Õ

Eb /N0 = 10

dB

. . .

40

¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ ẺµèÒ§æ ÊÓËÃѺ ÃкººÑ¹·Ö¡ (a) Ẻá¹Ç ¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ § . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

46

3.2

Ẻ¨ÓÅͧÊÓËÃѺ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE . . . . . . . . . . . . .

48

3.3

»ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» BER ¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµµèÒ§æ ÊÓËÃѺ ND = 2

. . . . . . .

59

3.4

»ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» BER ¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµµèÒ§æ ·Õ ND = 2.5 . . . . . . . . .

60

3.5

¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµáººµèÒ§æ à·Õº¡ÑºªèͧÊÑ­­Ò³·Õ ND = 2.5

61

3.6

ÍѵÊËÊÑÁ¾Ñ¹¸ì¢Í§ÅӴѺ¢éÍÁÙÅ

3.7

(a) ¡ÃÒ¿ÃÐËÇèÒ§ SNR ·Õ µéͧ¡Òà áÅÐ ND àÁ× Í SNR ·Õ µéͧ¡Òà áÅÐ

σj

{wk }

.

ÊÓËÃѺÃкº·Õ ·ÒÃìà¡çµáººµèÒ§æ ·Õ ND = 2.5

σj = 0%

62

áÅÐ (b) ¡ÃÒ¿ÃÐËÇèÒ§

·Õ ND = 2.5 . . . . . . . . . . . . . . . . . . . . . . .

4.1

ËÅÑ¡¡ÒÃ¾× ¹°Ò¹¢Í§à·¤¹Ô¤ PRML

. . . . . . . . . . . . . . . . . . . . . . . .

4.2

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ

4.3

µÑÇÍÂèҧẺ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅÅѡɳеèÒ§æ

. . . . . . . . . . . . . . . . .

63

66 66 70


H(D) = 1 + D

4.4

á¼¹ÀÒ¾ªèͧÊÑ­­Ò³áºº PR1,

4.5

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ªèͧÊÑ­­Ò³ PR1,

4.6

á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ PR1,

4.7

á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ EPR4,

4.8

(a) á¼¹ÀÒ¾ºÅçÍ¡, (b) à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, áÅÐ (c) á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§ªèͧ ÊÑ­­Ò³

4.9

H(D) = 1 − D2

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³

. . . . . . . . . . . . . . . . .

H(D) = 1 + D

H(D) = 1 + D

. . . . . . . . . . . .

72

. . . . . . . . . . . . .

73

H(D) = 1 + D − D2 − D3

. . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

H(D)

¾ÃéÍÁǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

71

74

75

. . . . . . . . . . . .

76

4.10 ¤Ó͸ԺÒÂá¼¹ÀÒ¾à·ÃÅÅÔÊ . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

76

4.11 á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ PR4,

H(D) = 1 − D2

. . . . . . . . . . . .

77

4.12 ¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ . . . . . . . . . . . . . . . . . . . . . .

79

4.13 á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³

H(D) = 1 + 0.5D

. . . . . . . . . . . . . .

81

4.14 á¼¹ÀҾ͸ԺÒ¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔã¹áµèÅЪèǧàÇÅÒ . . . . . . . .

82

4.15 (a) á¼¹ÀÒ¾ºÅçÍ¡, (b) à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, áÅÐ (c) á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§ªèͧ ÊÑ­­Ò³

H(D) = 1 − D

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

86

4.16 á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ . . . . . . . . . . . . . .

87

5.1

µÑÇÍÂèÒ§¡Ò÷ӧҹÀÒÂã¹à·ÃÅÅÔʢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ . . . . . . . . . . . . .

92

5.2

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ GPR ẺÊÁÁÙÅ

. . . . . . . . . . . . . . . . . . . . .

93

5.3

µÑÇÍÂèÒ§¡ÒäӹdzËÒÅӴѺ¢éͼԴ¾ÅÒ´

. . . . . . . . . . . . . . . . . . . . . .

95

5.4

ÀÒ¾ÊͧÁÔµÔáÊ´§¡Ãкǹ¡ÒöʹÃËÑÊ¢éÍÁÙÅâ´Âãªéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

5.5

»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºã¹ÃÙ» (a) BER áÅÐ (b)

5.6

(a) ¡ÃÒ¿ BER áÅÐ

SNReff ,

SNReff

. . . . . .

98

¢Í§·ÒÃìà¡çµáºº GPR5 . . 109

(b) ¡ÃÒ¿ BER áÅÐ SNR ¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ

ẺµèÒ§æ ³ ÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡·Õ à¢Õ¹ÇèÒ jitter µèÒ§æ ·Õ ND = 2.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

6.1

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ

6.2

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³¾ÃéÍÁǧ¨ÃµÃǨËÒ NPML

. . . . . . . . . . . . . . 116

. . . . . . . . . . . . . . . . 117


6.3

Ẻ¨ÓÅͧ¡Ãкǹ¡ÒÃ㹡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ

6.4

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ PR4

6.5

â¤Ã§ÊÃéÒ§¢Í§Ç§¨ÃµÃǨËÒ NPML ·Õ ãªé ¡Ñº §Ò¹»ÃÐÂØ¡µì ·Õ µéͧ¡ÒäÇÒÁàÃçÇ ã¹¡Òà »ÃÐÁÇżÅÊÙ§

. . . . . . . . . . . 118

. . . . . . . . . . . . . . . . . . . . . . . . . . . 121

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

6.6

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³áººÊÁÁÙÅ ¾ÃéÍÁ·Ñ §Ç§¨ÃµÃǨËÒ NPML áÅÐ PRML

6.7

á¼¹ÀÒ¾à·ÃÅÅÔʢͧ (a) ·ÒÃìà¡çµ

Heff (D)

=

H(D)

1 − 0.247D − 0.753D2

=

1−D

. . 126

áÅÐ (b) ·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

·Õ ãªé㹡ÒöʹÃËÑÊ¢éÍÁÙŢͧÃкº PRML

áÅÐ NPML µÒÁÅӴѺ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 6.8

á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÓËÃѺÃкº PRML

. . . 128

6.9

á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÓËÃѺÃкº NPML . . . 129

6.10 Ẻ¨ÓÅͧªèͧÊÑ­­Ò³¢Í§Ãкº¡Òúѹ·Ö¡áÁèàËÅç¡ . . . . . . . . . . . . . . . . 130 6.11 »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2

. . . . . . . . . . . . . . . . . . . . . . 131

6.12 »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ·Õ ãªé¨Ó¹Ç¹á·ç»µèÒ§¡Ñ¹ ·Õ SNR = 17 dB . . . . 132 6.13 »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2.5

. . . . . . . . . . . . . . . . . . . . . 133

6.14 ¼Å¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2.5 . . . . . . . . . . . . 134

7.1

Ẻ¨ÓÅͧªèͧÊÑ­­Ò³

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

7.2

¡ÓÅѧÊÑ­­Ò³Ãº¡Ç¹·Õ ¢Ö ¹¡Ñºáºº¢éÍÁÙÅ ³ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃì·Õ ¶Ù¡Í͡Ẻ ÊÓËÃѺ·ÒÃìà¡çµ EEPR2 [1 4 6 4 1] ¢Í§Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ·Õ ND = 2.5, SNR = 30 dB, áÅÐ

7.3

σj /T = 10%

. . . . . . . . . . . . . . . . . . . . . . . . 140

(a) á¼¹ÀÒ¾à·ÃÅÅÔÊÊÓËÃѺ·ÒÃìà¡çµáºº PR4 áÅÐ (b) ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ¢éÍÁÙÅ ´éÇÂá¼¹ÀÒ¾à·ÃÅÅÔÊ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

7.4

»ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» ¢Í§ BER ¢Í§Ç§¨ÃµÃǨËÒµèÒ§æ ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ § ·Õ ND = 2.5 áÅÐ

σj /T = 10%

. . . . . . . 149

8.1

Ẻ¨ÓÅͧ¡ÒÃà¢éÒÃËÑÊ RLL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

8.2

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(d, k)

. . . . . . . . . . . . . . . . . . . 157


8.3

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(1, 3)

. . . . . . . . . . . . . . . . . . . 158

8.4

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(0, 3)

. . . . . . . . . . . . . . . . . . . 160

8.5

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(0, 2)

. . . . . . . . . . . . . . . . . . . 161

8.6

µÑÇÍÂèÒ§ÃËÑÊ RLL ẺµèÒ§æ ·Õ ãªéã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì . . . . . . . . . . . . . . . . 165



ÊÒúѭµÒÃÒ§

3.1

µÑÇÍÂèÒ§·ÒÃìà¡çµáºº PR ·Õ à» ¹·Õ ÂÍÁÃѺ¡Ñ¹ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

4.1

¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§

4.2

µÑÇÍÂèÒ§áÊ´§¨Ó¹Ç¹Ê¶Ò¹Ð·Õ µéͧãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔʢͧ·ÒÃìà¡çµáººµèÒ§æ

5.1

ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ

ak

áÅÐ

5.2

. . .

71 88

εa (D) ¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáºº GPR3 ·Õ ND = 2.5

εa (D)

¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ Ẻ PR2 ·Õ ND=2.5

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ ÃкºÁÕ BER =

. . . . . . .

d2 {εa (D)}, ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ d2eff {εa (D)},

áÅÐÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

5.3

H(D) = 1 + D

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ

áÅÐ SNR=22 dB

¢Í§ªèͧÊÑ­­Ò³

45

d2 {εa (D)}, ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ d2eff {εa (D)},

áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ áÅÐ SNR = 22 dB

rk

. . . . . . . . . . .

10−4

εa (D)

¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáÅÐ

σj /T

ẺµèÒ§æ ³ ¨Ø´

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

7.1

¨Ó¹Ç¹¢Í§µÑÇ´Óà¹Ô¹¡ÒÃ·Õ ãªéµèÍ¢éÍÁÙÅ 1 ºÔµ ¢Í§Ç§¨ÃµÃǨËÒ PDNP áÅÐ PS PDNP147

8.1

µÑÇÍÂèÒ§¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´ ºÑ§¤Ñº

(d, ∞)

Nd (L)

·Õ ÁÕ¤ÇÒÁÂÒÇ

L

·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹ä¢

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

(d, k)

8.2

ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ¢Í§ÃËÑÊ RLL Ẻ

8.3

¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¤ÇÒÁ¨Ø

8.4

µÒÃÒ§¤é¹ËÒÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ RLL Ẻ

C(d, k)

µèÒ§æ

. . . . . . . . . . . . 156

áÅÐÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR . . . . . . . . . 157

xxi

(0, 2)

. . . . . . . . . . . . . 163



º··Õ 1

º·¹Ó

㹺·¹Õ ¨Ð͸ԺÒ¶֧ÀÒ¾ÃÇÁ¢Í§Ãкº¡Òúѹ·Ö¡¢éÍÁÙŢͧÍØ»¡Ã³ìÎÒÃì´´ÔÊ¡ìä´Ã¿ì (hard disk drive) à¾× Í à» ¹ ¡ÒÃàµÃÕÂÁ¤ÇÒÁ¾ÃéÍÁãËé ¼ÙéÍèÒ¹à¢éÒ㨶֧ ¾× ¹°Ò¹µèÒ§æ ·Õ à¡Õ ÂÇ¢éͧ¡Ñº Ãкº¡Òúѹ·Ö¡ ¢éÍÁÙŠẺ´Ô¨Ô·ÑÅ ÃÇÁ·Ñ § ËÅÑ¡¡Ò÷ӧҹ¢Í§¡Ãкǹ¡ÒÃà¢Õ¹áÅСÒÃÍèÒ¹¢éÍÁÙÅ ã¹ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¡è͹ ·Õ ͸ԺÒÂà¡Õ ÂǡѺ ¡ÒÃÇÔà¤ÃÒÐËì Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³áÅСÒÃÍ͡Ẻǧ¨ÃÀÒ¤ÃѺ ¢Í§ÎÒÃì´ ´ÔÊ¡ìä´Ã¿ì㹺·µèÍä»

1.1

¾× ¹°Ò¹¢Í§¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡à຺´Ô¨Ô·ÑÅ

¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡ (magnetic recording) ¤×Í ¡ÒèѴà¡çº¢éÍÁÙźԵãËéÍÂÙèã¹ÃÙ»¢Í§¡ÒÃà»ÅÕ Â¹ á»Å§ÃдѺÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ (magnetization) ã¹Ê× ÍºÑ¹·Ö¡ «Ö §ÊÒÁÒöáºè§ÍÍ¡à» ¹ 2 Ẻ [1] ¤×Í áººá͹ÐÅçÍ¡ (analog) áÅÐẺ´Ô¨Ô·ÑÅ (digital) ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒСÒà ºÑ¹·Ö¡ ÃкºáÁèàËÅç¡ áºº´Ô¨Ô·ÑÅ ·Õ ãªé ÊÓËÃѺ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì à·èÒ¹Ñ ¹ â´Â·Õ ¡Òúѹ·Ö¡ ÃкºáÁèàËÅç¡ áºº´Ô¨Ô·ÑŨÐãªé»ÃÐ⪹ì¨Ò¡ÊÁºÑµÔ¢Í§¤ÇÒÁà» ¹áÁèàËÅ硢ͧÇÑʴغҧª¹Ô´ ·Õ àÁ× ÍÍÂÙèã¹Ê¶Ò¹ÐÍÔ ÁµÑÇ (saturated) áÅéÇ ¨Ð·ÓãËéÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ã¹Ê× ÍºÑ¹·Ö¡ÁÕ·ÔÈ·Ò§ªÕ ä»ã¹·Ôȷҧ㴷ÔÈ·Ò§Ë¹Ö § ËÃ×Í ã¹·ÔÈ·Ò§µÃ§¡Ñ¹¢éÒÁ «Ö § ÅѡɳСÒúѹ·Ö¡ ¢éÍÁÙŠẺ¹Õ ¨ÐàËÁÒÐÊÓËÃѺ ¡ÒÃà¡çº ¢éÍÁÙÅ ´Ô¨Ô·ÑÅ ·Õ ÁÕ 2 ʶҹР¤×Í ºÔµ 1 áÅкԵ 0 ËÃ×Í·Õ àÃÕ¡¡Ñ¹ÇèÒ ¢éÍÁÙÅ亹ÒÃÕ (binary data) à¾ÃÒÐ©Ð¹Ñ ¹ 1


2

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

write signal

read-back signal write head

read head

medium disk motion

write current

read voltage

ÃÙ»·Õ 1.1: ËÅÑ¡¡ÒÃ¾× ¹°Ò¹¢Í§¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡

ÇÑÊ´Ø àËÅèÒ¹Õ ¨Ö§ ¶Ù¡ ¹ÓÁÒ·Óà» ¹ Ê× Í ºÑ¹·Ö¡ à¾× Í à¡çº ¢éÍÁÙŠ亹ÒÃÕ à¹× ͧ¨Ò¡¢éÍÁÙÅ ã¹» ¨¨ØºÑ¹ ÊèǹÁÒ¡¨Ð ÍÂÙè ã¹ÃÙ» ¢Í§¢éÍÁÙÅ ´Ô¨Ô·ÑÅ àªè¹ ¢éÍÁÙÅ ã¹à¤Ã× Í§¤ÍÁ¾ÔÇàµÍÃì áÅТéÍÁÙÅ ·Õ ÃѺÊè§ ¼èÒ¹à¤Ã×Í¢èÒÂÍÔ¹ à·ÍÃì à¹çµ à» ¹µé¹ ¹Í¡¨Ò¡¹Õ ¢éÍÁÙÅá͹ÐÅçÍ¡¡çÊÒÁÒö·Õ ¨Ð¶Ù¡á»Å§ãËéÍÂÙèã¹ÃÙ»¢Í§¢éÍÁÙÅ´Ô¨Ô·ÑÅä´éà¾× ÍãËé §èÒµèÍ ¡ÒèѴ à¡çº ¢éÍÁÙÅ â´Â¼èÒ¹¢Ñ ¹µÍ¹¡ÒÃ¡Å Ó ÃËÑÊ ¾ÑÅÊì (PCM: pulse code modulation) [2] à¾ÃÒÐ©Ð¹Ñ ¹ ¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡áºº´Ô¨Ô·ÑŨ֧àËÁÒÐÊÁ¡Ñº¡ÒÃà¡çº¢éÍÁÙÅã¹» ¨¨ØºÑ¹ ã¹» ¨¨ØºÑ¹¹Õ ¤ÇÒÁµéͧ¡ÒÃà¹× Í·Õ ã¹¡ÒèѴà¡çº¢éÍÁÙŢͧÍØ»¡Ã³ìÍÔàÅç¡·Ã͹ԡÊìµèÒ§æ ä´éá¡è ¤ÍÁ ¾ÔÇàµÍÃì, â·ÃÈѾ·ìà¤Å× Í¹·Õ , à¤Ã× Í§àÅè¹ à¾Å§áºº¾¡¾Ò, áÅСÅéͧ¶èÒÂÃÙ» ´Ô¨Ô·ÑÅ à» ¹µé¹ ÁÕ ÁÒ¡¢Ö ¹ àÃ× ÍÂæ à·¤â¹âÅÂÕ ¡Òúѹ·Ö¡ ÃкºáÁèàËÅç¡ áºº´Ô¨Ô·ÑÅ ¶×Í ä´é ÇèÒ à» ¹ ÇÔ¸Õ¡ÒÃËÅÑ¡ ·Õ ãªé 㹡ÒèѴ à¡çº ¢éÍÁÙŢͧ§Ò¹»ÃÐÂØ¡µì (application) µèÒ§æ ÃÇÁ件֧ á¼è¹ºÑ¹·Ö¡áÁèàËÅç¡ (magnetic oppy disk), ᶺºÑ¹·Ö¡ áÁèàËÅç¡ (magnetic tape), ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì, á¼è¹ «Õ ´Õ (CD: compact disc), áÅÐá¼è¹ ´ÕÇÕ´Õ (DVD: digital versatile disc) à» ¹µé¹ ÍÂèÒ§äáçµÒÁ ·Ø¡§Ò¹»ÃÐÂØ¡µì¨ÐµÑ §ÍÂÙ躹¾× ¹°Ò¹¢Í§ ËÅÑ¡¡Ò÷ӧҹà´ÕÂǡѹ«Ö §à¡Õ ÂÇ¢éͧ¡Ñº ËÑÇÍèÒ¹ (read head), ËÑÇà¢Õ¹ (write head), áÅÐÊ× ÍºÑ¹·Ö¡ áÁèàËÅç¡ (magnetic media) ´Ñ§áÊ´§ã¹ÃÙ»·Õ 1.1 àÁ× Í ËÑÇÍèÒ¹áÅÐËÑÇà¢Õ¹ẺÍÔ¹´Ñ¡·Õ¿ (inductive head) ¨Ð·ÓÁÒ¨Ò¡ÊÒÃáÁèàËÅç¡ÃÙ»à¡×Í¡ÁéÒ·Õ ÁÕ¤èÒÊÀҾźÅéÒ§áÁèàËÅç¡ (coercivity) µ Ó áÅФèÒÊÀÒ¾


1.2.

à຺¨ÓÅͧ¢Í§Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

3

ãËé«ÖÁ¼èÒ¹ä´é (permeability) ÊÙ§ [1, 3] â´Â¨ÐÁÕ¢´ÅÇ´¾Ñ¹ÍÂÙèÃͺæ áÅÐÊ× ÍºÑ¹·Ö¡¨Ð·ÓÁÒ¨Ò¡ÊÒà áÁèàËÅç¡·Õ ÁÕ¤èÒÊÀҾźÅéÒ§áÁèàËÅç¡ÊÙ§ ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ¨Ð¡ÅèÒǶ֧ ੾ÒÐà·¤â¹âÅÂÕ ¡Òúѹ·Ö¡ ¢éÍÁÙÅ 2 Ẻ ¤×Í ¡Òúѹ·Ö¡ Ẻá¹Ç ¹Í¹ (longitudinal recording) áÅСÒúѹ·Ö¡áººá¹ÇµÑ § (perpendicular recording) â´Â·Õ ෤⹠âÅÂÕ ¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹à» ¹ à·¤â¹âÅÂÕ ·Õ ãªé 㹡Òúѹ·Ö¡ ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì µÑ §áµè Í´Õµ ¨¹ ¶Ö§ » ¨¨ØºÑ¹ ¹Ñ ¹¤×Í ÊÀÒ¾¤ÇÒÁà» ¹ áÁèàËÅç¡ ¢Í§Ê× Í ºÑ¹·Ö¡ ¨Ð¢¹Ò¹¡Ñº ÃйҺ¢Í§¨Ò¹ºÑ¹·Ö¡ áÁèàËÅç¡ (magnetic disk) ´Ñ§·Õ áÊ´§ã¹ÃÙ»·Õ 1.1 ã¹¢³Ð·Õ à·¤â¹âÅÂÕ¡Òúѹ·Ö¡áººá¹ÇµÑ §ä´éàÃÔ Á·Õ ¨Ð¹ÓÁÒãªé ÊÓËÃѺ ¡Òúѹ·Ö¡ ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ã¹» ¨¨ØºÑ¹ â´ÂÊÀÒ¾¤ÇÒÁà» ¹ áÁèàËÅç¡ ¢Í§Ê× Í ºÑ¹·Ö¡ ¨Ð µÑ §©Ò¡¡ÑºÃйҺ¢Í§¨Ò¹ºÑ¹·Ö¡áÁèàËÅç¡ «Ö §ã¹» ¨¨ØºÑ¹¹Õ §Ò¹ÇԨѷҧ´éҹ෤â¹âÅÂÕ¡Òúѹ·Ö¡¢éÍÁÙŠẺá¹ÇµÑ §ä´é´Óà¹Ô¹ä»ÍÂèÒ§ÃÇ´àÃçÇ à¾ÃÒÐÇèÒ à·¤â¹âÅÂÕ¡Òúѹ·Ö¡áººá¹Ç¹Í¹à¢éÒã¡Åé ¢Õ´¨Ó¡Ñ´ «Ùà»ÍÃì¾ÒÃÒáÁ¡à¹µÔ¡ (superparamagnetic limit) [1, 3, 4, 10] ·ÓãËéäÁèÊÒÁÒöà¾Ô Á¤ÇÒÁ¨Ø¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ä´é ÁÒ¡¡ÇèÒ 1 à·ÃÐ亵ì (TB: terabyte) ¹Í¡¨Ò¡¹Õ à·¤â¹âÅÂÕ ¡Òúѹ·Ö¡ ¢éÍÁÙŠẺá¹ÇµÑ § ÊÒÁÒöªèÇÂà¾Ô Á ¢¹Ò´¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ä´é ËÅÒÂÊÔº à·èÒ àÁ× Í à·Õº¡Ñº ¡Òà ãªéà·¤â¹âÅÂÕ¡Òúѹ·Ö¡¢éÍÁÙÅẺá¹Ç¹Í¹ [3, 5]

1.2

à຺¨ÓÅͧ¢Í§Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ (digital data storage system) ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÒÁÒö·Õ ¨Ð¨ÓÅͧ໠¹ á¼¹ÀÒ¾·Ñ Çä»ä´é µÒÁÃÙ»·Õ 1.2 àÁ× Í ºÔµ¢èÒÇÊÒà (message bits) ¨Ð¶Ù¡·Ó¡ÒÃà¢éÒÃËÑÊâ´Â ǧ¨Ãà¢éÒ ÃËÑÊá¡é䢢éͼԴ¾ÅÒ´ (error correction code (ECC) encoder) â´Â·Õ ÃËÑÊ RS (Reed Solomon code) [7, 8] à» ¹ ÃËÑÊ ·Õ ¹ÔÂÁ¹ÓÁÒãªé 㹡ÒÃà¢éÒ ÃËÑÊ á¡é䢢éͼԴ¾ÅÒ´¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ã¹» ¨¨ØºÑ¹ ¨Ò¡¹Ñ ¹ ¢éÍÁÙÅ·Õ à¢éÒÃËÑÊáÅéÇ¡ç¨Ð¶Ù¡·Ó¡ÒÃà¢éÒÃËÑÊÍÕ¡¤ÃÑ §Ë¹Ö §´éÇ ǧ¨Ãà¢éÒÃËÑÊÁÍ´ÙàŪѹ (modu lation encoder) à¾× Í·Ó˹éÒ·Õ ã¹¡ÒûÃѺ¤Ø³ÊÁºÑµÔ¢Í§¢éÍÁÙÅãËéàËÁÒÐÊÁ¡ÑºªèͧÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ìä´Ã¿ì àªè¹ ·ÓãËéÅӴѺ¢éÍÁÙÅ (data sequence) ÁÕÃٻẺµÒÁ·Õ µéͧ¡Òà ËÃ×Í·ÓãËéÅӴѺ¢éÍÁÙÅäÁè ÁÕÊèǹ»ÃСͺ俿 Ò¡ÃÐáʵç (d.c. component) à» ¹µé¹ ÃËÑÊ·Õ ¹ÔÂÁãªéã¹Ç§¨Ãà¢éÒÃËÑÊÁÍ´ÙàŪѹ ¤×Í ÃËÑÊ RLL (run length limited code) [9] ¢éÍÁÙÅ àÍÒµì¾Øµ ·Õ ä´é ¨Ò¡Ç§¨Ãà¢éÒ ÃËÑÊ ÁÍ´ÙàŪѹ ¨Ð¶×Í ÇèÒ à» ¹ ¢éÍÁÙÅ ·Õ ¨Ð¶Ù¡ à¢Õ¹à¢éÒ ä»ã¹Ê× Í ºÑ¹·Ö¡ «Ö § ¨ÐàÃÕ¡¡Ñ¹ ÇèÒ ºÔµ ·Õ ¨Ð¶Ù¡ ºÑ¹·Ö¡ (recorded bit)


4

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

message bits

ECC encoder

modulation encoder

(e.g., RS codes)

(e.g., RLL codes)

write current waveform

recorded bits

modulator

write head/medium/read head assembly estimated message bits

ECC decoder

modulation decoder

reproduced bits

read channel read-back voltage waveform

A ÃÙ»·Õ 1.2: Ẻ¨ÓÅͧ·Ñ Ç仢ͧÃкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì [6]

ËÅѧ¨Ò¡¹Ñ ¹ ºÔµ ·Õ ¨Ð¶Ù¡ ºÑ¹·Ö¡ ¡ç ¨Ð¶Ù¡ Êè§ ä»Âѧ ǧ¨ÃÁÍ´ÙàÅàµÍÃì (modulator) à¾× Í á»Å§¢éÍÁÙÅ ºÔµ ãËé ÍÂÙè ã¹ÃÙ» ¤Å× ¹ ¡ÃÐáÊä¿¿ Ò à¢Õ¹ (write current waveform) ¨Ò¡¹Ñ ¹ ÃÙ» ¤Å× ¹ ¡ÃÐáÊä¿¿ Ò à¢Õ¹¡ç ¨Ð¶Ù¡» ͹ä»ÂѧËÑÇà¢Õ¹ à¾× Í·Ó¡ÒÃà¢Õ¹¢éÍÁÙÅŧä»ã¹Ê× ÍºÑ¹·Ö¡ ÊÓËÃѺ ¢Ñ ¹µÍ¹ã¹¡ÒÃÍèÒ¹¢éÍÁÙÅ ËÑÇÍèÒ¹¨Ð·Ó¡ÒÃÍèÒ¹¢éÍÁÙÅ ¨Ò¡Ê× Í ºÑ¹·Ö¡ àÁ× Í ËÑÇÍèÒ¹à¤Å× Í¹·Õ 1

ÁÒ¶Ö§ ºÃÔàdz·Õ ÁÕ ¡ÒÃà»ÅÕ Â¹á»Å§ÊÀÒ¾¤ÇÒÁà» ¹ áÁèàËÅç¡

(´Ù ÃÙ» ·Õ 1.1) ¨Ðä´é ¼ÅÅѾ¸ì ÍÍ¡ÁÒà» ¹

ÊÑ­­Ò³ÃÙ»¤Å× ¹áç´Ñ¹ä¿¿ Ò ·Õ àÃÕ¡¡Ñ¹ÇèÒ ÊÑ­­Ò³ read back ¨Ò¡¹Ñ ¹ ÊÑ­­Ò³ read back ¡ç¨Ð ¶Ù¡Êè§à¢éÒä»·Ó¡ÒûÃÐÁÇżÅ㹪èͧÊÑ­­Ò³ÍèÒ¹ (read channel) «Ö §»ÃСͺ仴éÇÂÊèǹ»ÃСͺ µèÒ§æ ä´éá¡è ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó (LPF: low pass lter), ǧ¨ÃªÑ¡ µÑÇÍÂèÒ§ (sampler ËÃ×Í analog to digital converter), ÍÕ¤ÇÍäÅà«ÍÃì (equalizer), áÅÐǧ¨ÃµÃǨËÒ (detector) à» ¹µé¹ â´Â¢éÍÁÙÅ àÍÒµì¾Øµ ·Õ ä´é ¡ç ¨Ð¶Ù¡ ·Ó¡ÒöʹÃËÑÊ ´éÇ ǧ¨Ã¶Í´ÃËÑÊ ÁÍ´ÙàŪѹ (modulation decoder) áÅÐǧ¨Ã ¶Í´ÃËÑÊá¡é䢢éͼԴ¾ÅÒ´ (ECC decoder) à¾× ÍËÒ¤èÒ»ÃÐÁÒ³¢Í§ºÔµ¢èÒÇÊÒÃ·Õ Êè§ÁÒ¨Ò¡µé¹·Ò§ 1

ã¹·Ò§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ºÃÔàdz·Õ ÁÕ¡ÒÃà»ÅÕ Â¹á»Å§ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ ¨Ð¶Ù¡á·¹´éÇ¢éÍÁÙźԵ 1 áÅкÃÔàdz

·Õ ÁÕäÁè¡ÒÃà»ÅÕ Â¹á»Å§ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ ¨Ð¶Ù¡á·¹´éÇ¢éÍÁÙźԵ 0 â´Â·Õ ÃٻẺ¢éÍÁÙÅÅÑ¡É³Ð¹Õ ¨ÐàÃÕ¡¡Ñ¹ÇèÒ ÃٻẺ NRZI (non return to zero interleaved) àÁ× Í ¢éÍÁÙÅ ºÔµ 1 ËÁÒ¶֧ ÁÕ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð (transition) áÅÐ ¢éÍÁÙźԵ 0 ËÁÒ¶֧ äÁèÁÕ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð


1.3.

¡Ãкǹ¡ÒÃà¢Õ¹

1.3

5

¡Ãкǹ¡ÒÃà¢Õ¹

ã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒÃà¢Õ¹¢éÍÁÙÅ [10] ¢éÍÁÙźԵ¨Ð¶Ù¡á»Å§ãËéÍÂÙèã¹ÃÙ»¤Å× ¹¡ÃÐáÊä¿¿ ÒÃÙ»ÊÕ àËÅÕ ÂÁ (rectangular current waveform) ·Õ àÃÕ¡¡Ñ¹ ÇèÒ ¡ÃÐáÊä¿¿ Ò à¢Õ¹ (write current) (´Ù ÃÙ» ·Õ 1.1) â´Âǧ¨ÃÁÍ´ÙàÅàµÍÃì (modulator) [1, 4] ¨Ò¡¹Ñ ¹ ¡ÃÐáÊä¿¿ Ò à¢Õ¹¨Ð¶Ù¡ » ͹ä»Âѧ ¢´ÅÇ´¢Í§ËÑÇ à¢Õ¹ (write head) ·ÓãËéà¡Ô´à» ¹Ê¹ÒÁà¢Õ¹áÁèàËÅç¡ (magnetic write eld) ºÃÔàdzªèͧÇèÒ§ (gap) ÃÐËÇèÒ§ Ê× Í ºÑ¹·Ö¡ ¡Ñº ËÑÇ à¢Õ¹ â´Â ·Ñ Çä» Ê¹ÒÁ à¢Õ¹ áÁèàËÅç¡ ¨Ð µéͧ ÁÕ ¢¹Ò´ ËÃ×Í ¤ÇÒÁ à¢éÁ ÁÒ¡ ¡ÇèÒ ÊÀҾźÅéÒ§áÁèàËÅ硢ͧÊ× Í ºÑ¹·Ö¡ à¾× Í·Õ ¨Ðä´é ÊÒÁÒö·ÓãËé Ê× Í ºÑ¹·Ö¡ ³ ºÃÔàdz¹Ñ ¹ ÁÕ ÊÀÒ¾¤ÇÒÁ à» è¹ áÁèàËÅç¡ µÒÁ·ÔÈ·Ò§¢Í§Ê¹ÒÁà¢Õ¹áÁèàËÅç¡ ·Õ » ͹à¢éÒ ä» ¹Í¡¨Ò¡¹Õ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐÊÀÒ¾ ¤ÇÒÁà» ¹ áÁèàËÅç¡ (magnetization transition) ¢Í§Ê× Í ºÑ¹·Ö¡ ÊÒÁÒö·Óä´é â´Â¡ÒÃà»ÅÕ Â¹á»Å§ ·ÔÈ·Ò§¢Í§Ê¹ÒÁáÁèàËÅç¡ÊÓËÃѺà¢Õ¹ (ËÃ×Í·ÔÈ·Ò§¢Í§¡ÃÐáÊä¿¿ Òà¢Õ¹) à¾× ÍãËéÊÍ´¤Åéͧ¡Ñº¡Òà à¢Õ¹¢éÍÁÙźԵ 0 áÅкԵ 1 ã¹·Ò§»¯ÔºÑµÔ Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì¨Ðãªé ¡Òúѹ·Ö¡áººäº¹ÒÃÕ (bina 2

ry recording) ¹Ñ ¹¤×Í ÊÀÒ¾¤ÇÒÁà» ¹ áÁèàËÅç¡ ·Õ ÍÂÙè ã¹Ê× Í ºÑ¹·Ö¡ ¨ÐÁÕ à¾Õ§ 2 ·ÔÈ·Ò§à·èÒ¹Ñ ¹

ËÃ×Í

¡ÅèÒÇÍÕ¡ ¹ÑÂ Ë¹Ö § ¤×Í ÃкºÊÒÁÒöºÑ¹·Ö¡ ¢éÍÁÙÅ ä´é à¾Õ§ 2 ÃдѺ (ËÃ×Í 2 ¤èÒ) à·èÒ¹Ñ ¹ «Ö § µèÒ§¨Ò¡ ¡Òúѹ·Ö¡¢éÍÁÙŢͧ´ÕÇÕ´Õ (DVD) ·Õ ÊÒÁÒöºÑ¹·Ö¡¢éÍÁÙÅä´éËÅÒÂæ ÃдѺ ·Ñ §¹Õ à» ¹à¾ÃÒÐÇèÒ â´Â»¡µÔ ¡Ãкǹ¡ÒÃà¢Õ¹¢éÍÁÙÅÁÕ¤ÇÒÁäÁèà» ¹àªÔ§àÊé¹ (nonlinearity) ÍÂÙè¾ÍÊÁ¤Çà ´Ñ§¹Ñ ¹ ¶éÒ·Ó¡Òúѹ·Ö¡ ¢éÍÁÙÅÁÒ¡¡ÇèÒ 2 ÃдѺŧä»ã¹Ê× ÍºÑ¹·Ö¡¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ¼Å¡Ãзº·Õ à¡Ô´¨Ò¡¤ÇÒÁäÁèà» ¹àªÔ§àÊé¹ ¡ç¨ÐÂÔ §ÁÕ¤ÇÒÁÃعáçÁÒ¡¢Ö ¹ «Ö §¨ÐÊ觼ŷÓãËé»ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§ÃкºáÂèŧÁÒ¡ [4]

1.4

¡Ãкǹ¡ÒÃÍèÒ¹

ã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒÃÍèÒ¹¢éÍÁÙÅ [10] ËÑÇÍèÒ¹¨Ð·Ó¡ÒõÃǨ¨Ñº¡ÒÃà»ÅÕ Â¹á»Å§¿ÅÑ¡«ìáÁèàËÅç¡ (mag netic ux) ³ µÓáË¹è§·Õ ÁÕ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ «Ö §à» ¹¼Å·ÓãËéà¡Ô´à» ¹ÊÑ­­Ò³ ¾ÑÅÊì áç´Ñ¹ä¿¿ Ò àË¹Õ ÂǹÓã¹¢´ÅÇ´ µÒÁ¡®¢Í§¿ÒÃÒà´Âì (Faraday s law) ÊÓËÃѺ ºÃÔàdz·Õ ÁÕ ¡Òà 2

¶éÒ ¾Ô¨ÒóҷÔÈ·Ò§¢Í§Ê¹ÒÁáÁèàËÅç¡ ¨Ò¡¢Ñ Ç à˹×Í ä»¢Ñ Ç ãµé (ËÃ×Í ¨Ò¡¢Ñ ÇºÇ¡ä»¢Ñ Çź) ¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹¨ÐÁÕ

ÅѡɳÐà» ¹áºº¢ÇÒ仫éÒ ËÃ×Í«éÒÂ仢ÇÒ ã¹¢³Ð·Õ ¡Òúѹ·Ö¡áººá¹ÇµÑ §¨Ðà» ¹áººº¹Å§ÅèÒ§ ËÃ×ÍÅèÒ§¢Ö ¹º¹ à·èÒ¹Ñ ¹


6

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

à»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È (isolated transition) ËÑÇÍèÒ¹¨ÐãËé ÊÑ­­Ò³¾ÑÅÊì áç´Ñ¹ä¿¿ Ò ·Õ àÃÕ¡¡Ñ¹ ÇèÒ ÊÑ­­Ò³¾ÑÅÊì à»ÅÕ Â¹Ê¶Ò¹Ð (transition pulse)

g(t)

ËÃ×Í

−g(t)

â´Â¨Ð¢Ö ¹ ÍÂÙè ¡Ñº ·ÔÈ·Ò§¢Í§

ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ã¹Ê× ÍºÑ¹·Ö¡ (´ÙÃÙ»·Õ 1.1) ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð (ËÃ×Í ÊÑ­­Ò³¾ÑÅÊì Lorent zian) ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìä´é ¤×Í [4]

g(t) =

³ 1+

PW50

àÁ× Í

¤×Í ¤ÇÒÁ¡ÇéÒ§¢Í§ÊÑ­­Ò³¾ÑÅÊì

g(t)

1 2t PW50

´2

(1.1)

ÇÑ´ ³ µÓáË¹è§ ·Õ ÊÑ­­Ò³¾ÑÅÊì ÁÕ ¤ÇÒÁÊÙ§ à» ¹

¤ÃÖ §Ë¹Ö § ¢Í§¤ÇÒÁÊÙ§ ÊÙ§ÊØ´ áÅÐÊÓËÃѺ ¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § ÊÑ­­Ò³¾ÑÅÊì à»ÅÕ Â¹Ê¶Ò¹Ð¨ÐÁÕ ÃÙ» ÊÁ¡Òà ¤×Í [11]

ln(·)

àÁ× Í

¤×Í ÅÍ¡ÒÃÔ·ÖÁ¸ÃÃÁªÒµÔ (natural logarithm),

tion) «Ö §¹ÔÂÒÁâ´Â ¢Í§

g(t)

à √ ! 2t ln 2 g(t) = erf PW50

erf(x) =

√2 π

Rx 0

e

−t2

erf(·)

dt, áÅÐ PW50

(1.2)

¤×Í ¿ §¡ìªÑ¹¢éͼԴ¾ÅÒ´ (error func

¤×Í ¤ÇÒÁ¡ÇéÒ§¢Í§¾ÑÅÊì

g 0 (t) ËÃ×Í Í¹Ø¾Ñ¹¸ì

ÇÑ´ ³ µÓáË¹è§·Õ ÊÑ­­Ò³¾ÑÅÊìÁÕ¤ÇÒÁÊ٧໠¹¤ÃÖ §Ë¹Ö §¢Í§¤ÇÒÁÊÙ§ÊÙ§ÊØ´

ã¹ Ãкº ¡Òà ºÑ¹·Ö¡ ¢éÍÁÙÅ ¢Í§ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¤ÇÒÁ ˹Òá¹è¹ ¢Í§ ¡Òà ºÑ¹·Ö¡ Ẻ ¹ÍÃì ÁÍ ÅäÅ«ì (ND: normalized recording density) ËÃ×Í ¤ÇÒÁ˹Òá¹è¹¢Í§¡Òúѹ·Ö¡¢éÍÁÙÅ [4] ¨Ð¹ÔÂÒÁâ´Â

ND = àÁ× Í

T

(1.3)

¤×Í ¤ÒºàÇÅҢͧ¢éÍÁÙÅË¹Ö §ºÔµ ËÃ×Í·Õ àÃÕ¡¡Ñ¹ÇèÒ ºÔµà«ÅÅì (bit cell) «Ö §¨Ðà» ¹µÑǺ觺͡

ÇèÒ ºÃÔàdz ¤èÒ

PW50 T

PW50

PW50

ÊÒÁÒö·Õ ¨Ð¨Ñ´à¡çº¢éÍÁÙÅä´é¨Ó¹Ç¹¡Õ ºÔµ ´Ñ§¹Ñ ¹ ¶éÒ¡Ó˹´ãËé

T

à» ¹¤èÒ¤§·Õ àÁ× Í

ËÃ×Í ND à¾Ô Á ¢Ö ¹ ¡ç ËÁÒ¤ÇÒÁÇèÒ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÊÒÁÒö¨Ø ¢éÍÁÙÅ ä´é ÁÒ¡¢Ö ¹ ÃÙ» ·Õ 1.3

áÊ´§¼ÅµÍºÊ¹Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐÊÓËÃѺ¡Òúѹ·Ö¡áººá¹Ç¹Í¹áÅÐẺá¹ÇµÑ § ³ ÃдѺ ND µèÒ§æ ¨ÐàËç¹ä´éÇèÒ ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð¢Í§·Ñ § 2 Ãкº¨Ð¤Ãͺ¤ÅØÁªèǧàÇÅÒËÅÒÂæ ºÔµà«ÅÅì â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í ND ÁÕ ¤èÒ à¾Ô Á ¢Ö ¹ ËÃ×Í ÍÒ¨¨Ð¡ÅèÒÇä´é ÇèÒ ¡ÒÃá·Ã¡ÊÍ´ÃÐËÇèÒ§ÊÑ­Åѡɳì (ISI: intersymbol interference) ã¹ÊÑ­­Ò³ read back ¨ÐÁÕ ¤ÇÒÁÃعáçÁÒ¡¢Ö ¹ àÁ× Í ND ÁÕ ¤èÒ


1.4.

¡Ãкǹ¡ÒÃÍèÒ¹

7

ND = 2

1

ND = 2.5 ND = 3

Amplitude

0.8

0.6

0.4

0.2

0 −5

0

5

(a) t/T 1 0.8

ND = 2 ND = 2.5 ND = 3

0.6

Amplitude

0.4 0.2 0 −0.2 −0.4 −0.6 −0.8 −1 −5

0

5

(b) t/T

ÃÙ»·Õ 1.3: ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð ÊÓËÃѺ¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §

à¾Ô Á¢Ö ¹ à¹× ͧ¨Ò¡ âÍ¡ÒÊ·Õ ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ÍÂÙèã¡Åé¡Ñ¹¨ÐÁÒ«é͹àËÅ× ÍÁ (overlap) ¡Ñ¹ÁÕ ¤ÇÒÁà» ¹ä»ä´éÊÙ§ ã¹¡Ã³Õ·Õ ËÑÇÍèÒ¹à¤Å× Í¹·Õ ÁÒ¶Ö§ºÃÔàdz·Õ ÁÕµÓá˹觡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐµÔ´¡Ñ¹ 2 ¤ÃÑ § ÊÑ­­Ò³¾ÑÅÊì


8

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊØ·¸Ô ·Õ ä´é ¨ÐàÃÕ¡¡Ñ¹ ÇèÒ ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ (dibit pulse) ËÃ×Í ¼ÅµÍºÊ¹Í§ä´ºÔµ (dibit re sponse) [4] «Ö §ÁÕ¤èÒà·èҡѺ

m(t) = g(t) − g(t − T )

(1.4)

´Ñ§áÊ´§ã¹ÃÙ»·Õ 1.4 ¶éÒãªé¡ÒÃá»Å§¿ÙàÃÕÂÃì·Õ µèÍà¹× ͧ·Ò§àÇÅÒ (continuous time Fourier transform) [12] ¡ÑºÊÑ­­Ò³

m(t)

m(t)

ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡

M (Ω) = exp{−π|Ω|ND} (1 − exp{−j2πΩ})

(1.5)

¨Ðä´é ÇèÒ ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ (frequency response) ¢Í§

Ẻá¹Ç¹Í¹ ¤×Í

àÁ× Í

exp{·}

m(t)

¤×Í ¿ §¡ìªÑ¹àÅ¢ªÕ ¡ÓÅѧ (exponential function) ã¹¢³Ð·Õ ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§

ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ¤×Í

½ 2 2 ¾ π Ω ND2 T exp − (1 − exp{−j2πΩ}) M (Ω) = jπΩ ln(16) àÁ× Í

Ω = fT

(1.6)

f ¤×Í ¤ÇÒÁ¶Õ ÁÕ˹èÇÂà» ¹ √ j = −1 ¤×Í Ë¹èǨԹµÀÒ¾

¤×Í ¤ÇÒÁ¶Õ Ẻ¹ÍÃìÁÍÅäÅ«ì (normalized frequency),

àÎÔõ«ì (Hertz),

|x|

¤×Í ¤èÒÊÑÁºÙóì (absolute value) ¢Í§

x,

áÅÐ

(imadinary unit) ÃÙ»·Õ 1.5 áÊ´§¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ ¨ÐàËç¹ä´éÇèÒ àÁ× Í ND à¾Ô Á ¢Ö ¹ ÃÙ»ÃèÒ§¢Í§¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ ·Ñ § 2 Ẻ¨Ð¶Ù¡ ºÕº ãËé ÁÒ ÍÂÙè ³ ºÃÔàdz¤ÇÒÁ¶Õ µ Ó ¹Í¡¨Ò¡¹Õ ªèͧÊÑ­­Ò³¢Í§¡Òúѹ·Ö¡áººá¹Ç¹Í¹¨ÐÁÕÊ໡µÃÑÁ¤èÒÈÙ¹Âì (spectral null) ³ µÓáË¹è§·Õ ¤ÇÒÁ¶Õ

f = 0 «Ö §ËÁÒ¶֧ äÁèÁÕÊèǹ»ÃСͺ俿 Ò¡ÃÐáʵç ã¹¢³Ð·Õ

ªèͧÊÑ­­Ò³¢Í§¡Òúѹ·Ö¡áººá¹ÇµÑ §¨ÐÁÕÊèǹ»ÃСͺ俿 Ò¡ÃÐáʵç ËÁÒÂà˵Ø

3

ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ¨Ðãªé â»Ãá¡ÃÁ SCILAB

[14] 㹡ÒÃÇÒ´ÃÙ» ¡ÃÒ¿¢Í§ÊÑ­­Ò³µèÒ§æ

ÃÇÁ·Ñ § ¼Å¡Ò÷´Åͧ·Õ ä´é ¨Ò¡¡Ò÷ӡÒèÓÅͧ (simulation) Ãкº ¼ÙéÍèÒ¹¤ÇÃ·Õ ¨Ð¾ÂÒÂÒÁ·´Åͧ ÇÒ´ÃÙ»¡ÃÒ¿µèÒ§æ ã¹Ë¹Ñ§Ê×ÍàÅèÁ¹Õ à¾× Í·Õ ¨Ðä´éªèÇ·ÓãËéà¢éÒã¨ã¹º·àÃÕ¹ÁÒ¡ÂÔ §¢Ö ¹ 3

â»Ãá¡ÃÁ SCILAB à» ¹â»Ãá¡ÃÁ·Õ ÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ã¡Åéà¤Õ§¡Ñºâ»Ãá¡ÃÁ MATLAB [13] «Ö §

¤èÒÅÔ¢ÊÔ·¸Ô «Í¿µìáÇÃì ¢Í§â»Ãá¡ÃÁ MATLAB ÁÕ ÃÒ¤ÒᾧÁÒ¡ áµè â»Ãá¡ÃÁ SCILAB à» ¹ â»Ãá¡ÃÁ·Õ ãËé¿ÃÕ (freeware) ¼ÙéÍèÒ¹ÊÒÁÒö´ÒǹìâËÅ´µÑÇâ»Ãá¡ÃÁä´é¨Ò¡ http//www.scilab.org ËÃ×Í http://home.npru.ac.th/∼t3058/Scilab.html


1.5.

à຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡

0.5

9

ND = 2 ND = 2.5

0.4

ND = 3 0.3

Amplitude

0.2 0.1 0 −0.1 −0.2 −0.3 −0.4 −0.5 −5

0

5

(a) t/T 1 ND = 2

0.9

ND = 2.5 0.8

ND = 3

Amplitude

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 −5

0

5

(b) t/T

ÃÙ»·Õ 1.4: ¼ÅµÍºÊ¹Í§ä´ºÔµ ÊÓËÃѺ¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §

1.5

à຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡

â´Â·Ñ Çä» ªèͧÊÑ­­Ò³¢Í§Ãкº¡Òúѹ·Ö¡ áÁèàËÅç¡ ÊÒÁÒö¨ÓÅͧä´é à» ¹ 2 Ẻ ¤×Í áºº¨ÓÅͧ ªèͧÊÑ­­Ò³ àÊÁ×͹ ¨ÃÔ§ (realistic channel model) áÅРẺ¨ÓÅͧ ªèͧÊÑ­­Ò³ ÍØ´Á¤µÔ (ideal


10

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ND = 2

1

Normalized magnitude

ND = 2.5 ND = 3 0.8

0.6

0.4

0.2

0

0

0.1

0.2

0.3

0.4

0.5

(a) Normalized frequency (fT)

ND = 2

1

ND = 2.5

Normalized magnitude

ND = 3 0.8

0.6

0.4

0.2

0

0

0.1

0.2

0.3

0.4

0.5

(b) Normalized frequency (fT)

ÃÙ»·Õ 1.5: ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ ÊÓËÃѺ¡Òúѹ·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §


1.5.

à຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡

11

n(t)

ak

1–D

bk

p(t)

g(t)

LPF

symbol detector

equalizer

â k

m(t) timing recovery target response H(D) ÃÙ»·Õ 1.6: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§

channel model) â´Â·Õ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§ [10] ¨ÐÁÕÅѡɳСÒ÷ӧҹã¡Åéà¤Õ§¡Ñº Ãкº¨ÃÔ§ à¹× ͧ¨Ò¡»ÃСͺ仴éÇ·ء Êèǹ»ÃСͺ·Õ ÊӤѭ ·Õ ÁÕ ÍÂÙè 㹠ʶһ µÂ¡ÃÃÁªèͧÊÑ­­Ò³ ÍèÒ¹ (read channel architecture) [8, 10] ¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ã¹¢³Ð·Õ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ ÍØ´Á¤µÔÁÑ¡¨Ð¹ÔÂÁãªé㹡ÒÃÈÖ¡ÉÒ áÅÐÇÔà¤ÃÒÐËì¾× ¹°Ò¹¡Ò÷ӧҹ¢Í§Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³ ¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì à¹× ͧ¨Ò¡ à» ¹áºº¨ÓÅͧ·Õ äÁè«Ñº«é͹áÅЧèÒµèÍ¡Ò÷ӤÇÒÁà¢éÒã¨

1.5.1

Êèǹ

A

à຺¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§

ã¹ÃÙ»·Õ 1.2 ÊÒÁÒö·Õ ¨ÐáÊ´§ãËéÍÂÙèã¹ÃÙ»¢Í§áºº¨ÓÅͧ·Ò§¤³ÔµÈÒʵÃìä´é µÒÁÃÙ»·Õ 1.6

¡ÅèÒǤ×Í ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ Ẻ亹ÒÃÕ

ak ∈ {0, 1}

·Õ ÁÕ ¤ÒºàÇÅҢͧºÔµ (bit period) à·èÒ ¡Ñº

¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧǧ¨ÃËÒ͹ؾѹ¸ìÍØ´Á¤µÔ (ideal di erentiator) ˹èǧàÇÅÒ

T

˹èÇ ·ÓãËéä´éà» ¹ÅӴѺ¢éÍÁÙÅà»ÅÕ Â¹Ê¶Ò¹Ð

1−D

àÁ× Í

D

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ

bk ∈ {−1, 0, 1} àÁ× Í bk = ±1 ËÁÒ¶֧

¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐẺºÇ¡ (positive transition) ËÃ×Í áººÅº (negative transition) áÅÐ ËÁÒ¶֧ äÁèÁÕ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ÅӴѺ¢éÍÁÙÅà»ÅÕ Â¹Ê¶Ò¹Ð á·¹´éǼŵͺʹͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð read back,

p(t),

g(t)

T

bk

bk = 0

¨Ð¶Ù¡Ê觼èÒ¹ä»ÂѧªèͧÊÑ­­Ò³·Õ ¶Ù¡

áÅж١ ú¡Ç¹´éÇÂÊÑ­­Ò³Ãº¡Ç¹

n(t)

ÊÑ­­Ò³

¨Ð¶Ù¡¡Ãͧ´éÇÂǧ¨Ã¡Ãͧ¼èÒ¹µ Ó (LPF) à¾× ͡ӨѴÊÑ­­Ò³Ãº¡Ç¹·Õ ÍÂÙè¹Í¡á¶º

¤ÇÒÁ¶Õ (out of band noise) ¨Ò¡¹Ñ ¹ ¡ç ¨Ð¶Ù¡ ·Ó¡Òêѡ µÑÇÍÂèÒ§ (sampling) ³ àÇÅÒ·Õ ¶Ù¡ ¤Çº¤ØÁ


12

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

â´ÂÃкºä·ÁÁÔ § ÃԤѿàÇÍÃÕ (timing recovery) ¢éÍÁÙÅ àÍÒµì¾Øµ ¢Í§Ç§¨ÃªÑ¡ µÑÇÍÂèÒ§¨Ð¶Ù¡ Êè§ ¼èÒ¹ä»Âѧ ÍÕ¤ÇÍäÅà«ÍÃì áÅÐǧ¨ÃµÃǨËÒÊÑ­Åѡɳì (symbol detector) à¾× ÍËÒÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ à» ¹ä»ä´é ÁÒ¡·Õ ÊØ´ (most likely input sequence) ¹Ñ ¹¤×Í ËÒ¤èÒ»ÃÐÁÒ³¢Í§

ak

ËÃ×Í

âk

ǧ¨ÃµÃǨËÒÊÑ­Åѡɳì·Õ ¹ÔÂÁãªéã¹Ãкº¡Òúѹ·Ö¡áÁèàËÅç¡ ¤×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ (Viter bi detector) [15] ÍÂèÒ§äáçµÒÁ à¹× ͧ¨Ò¡ ¤ÇÒÁ«Ñº«é͹ (complexity) ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ðà¾Ô Á ¢Ö ¹ ẺàÅ¢ªÕ ¡ÓÅѧ µÒÁ¨Ó¹Ç¹Ë¹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³ (channel memory) ´Ñ§¹Ñ ¹ ÍÕ¤ÇÍäÅà«ÍÃì¨Ö§ à» ¹ ÊÔ § ¨Óà» ¹ ·Õ ¨Ðµéͧ¶Ù¡ ¹ÓÁÒãªé §Ò¹ à¾× Í ·Ó˹éÒ·Õ ã¹¡ÒûÃѺ ÃÙ»ÃèÒ§¼ÅµÍºÊ¹Í§ 4

ÃÇÁ¢Í§·Ñ § ÃкºãËé à» ¹ ¼ÅµÍºÊ¹Í§·Õ µéͧ¡Òà ·Õ àÃÕ¡¡Ñ¹ ÇèÒ ¼ÅµÍºÊ¹Í§·ÒÃìà¡çµ sponse)

(target re

H(D) [4] áÅЪèÇ·ÓãËé¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔŴŧä´é (ÈÖ¡ÉÒÃÒÂÅÐàÍÕ´

à¾Ô ÁàµÔÁä´é㹺··Õ 3)

1.5.2

à຺¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

¶éÒÊÁÁصÔãËé ÃкºÁÕ¡Ãкǹ¡ÒÃÍÕ¤ÇÍäÅ૪ѹẺÊÁºÙóì (perfect equalization) Ẻ¨ÓÅͧã¹ÃÙ» ·Õ 1.6 ¨ÐÊÒÁÒöŴÃÙ» ä´é à» ¹ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ ´Ñ§ áÊ´§ã¹ÃÙ» ·Õ 1.7 â´Â·Õ ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ Ẻ亹ÒÃÕ

ak

·Õ ÁÕ ¤ÒºàÇÅҢͧºÔµ à·èÒ ¡Ñº

ÊÑ­­Ò³¾ÑÅÊì 乤ÇÔµÊì ÍØ´Á¤µÔ (ideal Nyquist pulse) ú¡Ç¹´éÇÂÊÑ­­Ò³Ãº¡Ç¹

n(t)

T

q(t)

¨Ð¶Ù¡ ¡Å Ó ÊÑ­­Ò³ (modulate) ¡Ñº =

sin(πt/T )/(πt/T )

ÊÑ­­Ò³·Õ ǧ¨ÃÀÒ¤ÃѺ ä´é ÃѺ

p(t)

[16] áÅж١

¨Ð¶Ù¡ Êè§ ¼èÒ¹ä»Âѧ ǧ¨Ã¡Ãͧ

¼èÒ¹µ Ó à¾× ͡ӨѴÊÑ­­Ò³Ãº¡Ç¹·Õ ÍÂÙè¹Í¡á¶º¤ÇÒÁ¶Õ ¨Ò¡¹Ñ ¹ ¡ç¨Ð¶Ù¡·Ó¡ÒêѡµÑÇÍÂèÒ§ ³ àÇÅÒ·Õ ¶Ù¡¤Çº¤ØÁâ´ÂÃкºä·ÁÁÔ §ÃԤѿàÇÍÃÕ ¨Ò¡¹Ñ ¹ ¢éÍÁÙÅàÍÒµì¾Øµ¢Í§Ç§¨ÃªÑ¡µÑÇÍÂèÒ§¡ç¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧ ǧ¨ÃµÃǨËÒÊÑ­Åѡɳì à¾× Í·Ó¡ÒÃËÒÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ à» ¹ä»ä´éÁÒ¡·Õ ÊØ´ ·ÒÃìà¡çµáºº¼ÅµÍºÊ¹Í§ºÒ§Êèǹ ËÃ×Í ·Õ àÃÕ¡ÇèÒ ·ÒÃìà¡çµáºº PR (partial response) [17] ·Õ à» ¹·Õ ÂÍÁÃѺ㹡Òúѹ·Ö¡áººá¹Ç¹Í¹ ¨ÐÁÕÃÙ»ÊÁ¡ÒÃà» ¹ [4]

H(D) = (1 − D)(1 + D)n 4

(1.7)

·ÒÃìà¡çµ (target) ã¹·Ò§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ËÁÒ¶֧ ǧ¨Ã¡ÃͧẺàªÔ§àÊé¹·Õ Áըӹǹá·ç» (tap) ¹éÍ áÅж١Í͡Ẻ

ãËéÁռŵͺʹͧàªÔ§¤ÇÒÁ¶Õ àËÁ×͹¡Ñº¼ÅµÍºÊ¹Í§¢Í§ªèͧÊÑ­­Ò³ãËéÁÒ¡·Õ ÊØ´ â´Â»ÃÒȨҡ¡ÒâÂÒÂÊÑ­­Ò³Ãº¡Ç¹


1.5.

à຺¨ÓÅͧªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ÃкºààÁèàËÅç¡

13

n(t)

ak

rk

H(D)

p(t)

q(t)

âk

symbol detector

LPF

timing recovery

ÃÙ»·Õ 1.7: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

ã¹¢³Ð·Õ ·ÒÃìà¡çµáºº PR ·Õ à» ¹·Õ ÂÍÁÃѺ㹡Òúѹ·Ö¡áººá¹ÇµÑ § ¨ÐÍÂÙèã¹ÃÙ»¢Í§ [18]

H(D) = (1 + D)n

àÁ× Í

n

(1.8)

¤×Í àÅ¢¨Ó¹Ç¹àµçÁ ºÇ¡ ¨Ò¡ÊÁ¡Òà (1.8) ¨Ð¾ºÇèÒ Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § ¨ÐäÁè ÁÕ ¾¨¹ì

(1 − D)

·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § ÁÕ Êèǹ»ÃСͺ俿 Ò¡ÃÐáʵç (´Ù

ÃÙ» ·Õ 1.5) ÃÙ» ·Õ 1.8 à»ÃÕºà·Õº¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ ẺµèÒ§æ â´Â·Õ µÑÇàÅ¢·Õ ÍÂÙè ã¹à¤Ã× Í§ËÁÒÂǧàÅçº ÊÕ àËÅÕ ÂÁ

[1 0 − 1]

D

áÊ´§¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô áµèÅÐá·ç» ¢Í§·ÒÃìà¡çµ µÑÇÍÂèÒ§àªè¹ PR4

ËÁÒ¶֧ ·ÒÃìà¡çµ Ẻ PR4 (PR class IV) ·Õ ÁÕ ¿ §¡ìªÑ¹ ¶èÒÂâ͹ã¹â´àÁ¹

H(D) = 1 − D2 ã¹â´àÁ¹

[. . .]

¤×Í

ËÃ×Í EEPR2

[1 4 6 4 1]

D

[10] ¤×Í

ËÁÒ¶֧ ·ÒÃìà¡çµáºº EEPR2 ·Õ ÁÕ¿ §¡ìªÑ¹¶èÒÂâ͹

H(D) = 1 + 4D + 6D2 + 4D3 + D4

à» ¹µé¹

¨Ò¡ÃÙ» ·Õ 1.8 ¨Ð¾ºÇèÒ àÁ× Í ªèͧÊÑ­­Ò³ÁÕ ¤èÒ ND à¾Ô Á ¢Ö ¹ ·ÒÃìà¡çµ·Õ ãªé ¡ç ¤ÇÃ·Õ ¨ÐÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡¢Ö ¹ (ÁÕ ¤èÒ

n

ÁÒ¡¢Ö ¹) à¾× Í·Õ ¨Ð·ÓãËé ¼ÅµÍºÊ¹Í§·ÒÃìà¡çµ ÁÕ ÅѡɳÐã¡Åéà¤Õ§¡Ñº ¼ÅµÍºÊ¹Í§

¢Í§ ªèͧÊÑ­­Ò³ ãËé ÁÒ¡ ·Õ ÊØ´ «Ö § ¨Ð Êè§ ¼Å ·Ó ãËé ǧ¨Ã µÃǨËÒ ÇÕà·ÍÃìºÔ ·Ó§Ò¹ ä´é ÍÂèÒ§ ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ ÁÒ¡¢Ö ¹ (ÈÖ¡ÉÒÃÒÂÅÐàÍÕ´à¾Ô ÁàµÔÁ ä´é 㹺··Õ 3) ¹Í¡¨Ò¡¹Õ ¨Ò¡ÊÁ¡Òà (1.7) áÅÐ (1.8) ¨Ð¾º ÇèÒ ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§·ÒÃìà¡çµ Ẻ PR ·Ø¡ µÑÇ ¨Ðà» ¹ àÅ¢¨Ó¹Ç¹àµçÁ ÍÂèÒ§äáçµÒÁ ¶éÒ ãªé ·ÒÃìà¡çµ ·Õ ÁÕ ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô à» ¹àÅ¢¨Ó¹Ç¹¨ÃÔ§ «Ö §¨ÐàÃÕ¡ÇèÒ ·ÒÃìà¡çµáºº GPR (generalized partial response target) »ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкº·Õ ä´é¨ÐÁÕÁÒ¡¡ÇèÒ¡ÒÃãªé·ÒÃìà¡çµáºº PR [18, 19]


14

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

Normalized magnitude

1

0.8

0.6

0.4

Channel response (ND = 2) Channel response (ND = 2.5) PR4 [1 0 −1] (n = 1)

0.2

EPR4 [1 1 −1 −1] (n = 2) EEPR4 [1 2 0 −2 1] (n = 3) 0

0

0.1

0.2

0.3

0.4

0.5

(a) Normalized frequency (fT)

Channel response (ND = 2)

1

Channel response (ND = 2.5)

Normalized magnitude

PR2 [1 2 1] (n = 2) EPR2 [1 3 3 1] (n = 3)

0.8

EEPR2 [1 4 6 4 1] (n = 4) 0.6

0.4

0.2

0

0

0.1

0.2

0.3

0.4

0.5

(b) Normalized frequency (fT)

ÃÙ»·Õ 1.8: ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ ẺµèÒ§æ ÊÓËÃѺ ªèͧÊÑ­­Ò³¡Òúѹ·Ö¡ (a) Ẻ á¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §


1.6.

ÊÃØ»·éÒº·

1.6

15

ÊÃØ»·éÒº·

㹺·¹Õ ä´é¡ÅèÒǶ֧¾× ¹°Ò¹¢Í§¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡áºº´Ô¨Ô·ÑÅ ÃÇÁ·Ñ §ËÅÑ¡¡Ò÷ӧҹ¢Í§¡Ãкǹ ¡ÒÃà¢Õ¹áÅСÒÃÍèÒ¹¢éÍÁÙÅã¹ÍØ»¡Ã³ìÎÒÃì´´ÔÊ¡ìä´Ã¿ì ¹Í¡¨Ò¡¹Õ Âѧä´é͸ԺÒ¶֧Ẻ¨ÓÅͧ¡Ò÷ӧҹ ¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ·Ñ §áºº¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§ áÅÐẺ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ â´Â·Õ Ẻ¨ÓÅͧ·Ñ §Êͧ¹Õ ¨Ð¶Ù¡¹ÓÁÒãªé㹡ÒèÓÅͧÃкº (system simulation) 㹺·µèÍæ ä» à¹× ͧ¨Ò¡ ¡ÒÃÇÔà¤ÃÒÐËìÃкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì¨ÐÍÒÈÑÂ¾× ¹°Ò¹·Ò§¤³Ôµ ÈÒʵÃì·Õ à¡Õ ÂÇ¢éͧ¡Ñº¡ÒûÃÐÁÇżÅÊÑ­­Ò³ áÅСÒÃÊ× ÍÊÒôԨԷÑŤè͹¢éÒ§ÁÒ¡ ´Ñ§¹Ñ ¹ ¼ÙéÍèÒ¹¤ÇÃ·Õ ¨Ð ·º·Ç¹¤ÇÒÁÃÙéàËÅèÒ¹Õ ãËéà¢éÒ㨡è͹·Õ ¨ÐÈÖ¡ÉÒ˹ѧÊ×ÍàÅèÁ¹Õ «Ö §¤ÇÒÁÃÙé¾× ¹°Ò¹·Ò§¤³ÔµÈÒʵÃì´éÒ¹µèÒ§æ ·Õ à¡Õ ÂÇ¢éͧ¡Ñº¡ÒÃÇÔà¤ÃÒÐËìÃкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ÊÒÁÒö·Õ ¨ÐÈÖ¡ÉÒä´é¨Ò¡ ˹ѧÊ×Í ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÊÓËÃѺ¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ àÅèÁ 1: ¾× ¹°Ò¹ªèͧÊÑ­­Ò³ÍèÒ¹ à¢Õ¹ [10]

1.7

à຺½ ¡ËÑ´·éÒº·

1. ¨§Í¸ÔºÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡

2. ¨§Í¸ÔºÒ¢éÍᵡµèÒ§ÃÐËÇèÒ§Ãкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹áÅÐẺá¹ÇµÑ §

3. ¨§Í¸ÔºÒÂËÅÑ¡¡ÒÃàº× ͧµé¹¢Í§¡Ãкǹ¡ÒÃà¢Õ¹áÅСÒÃÍèÒ¹¢éÍÁÙÅã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

4. ¨§Í¸ÔºÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§áºº¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§

5. ¨§Í¸ÔºÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§áºº¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

6. ¨§Â¡µÑÇÍÂèÒ§·ÒÃìà¡çµáºº PR ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ÁÒÍÂèÒ§¹éÍ 4 ·ÒÃìà¡çµ

7. ¨§Â¡µÑÇÍÂèÒ§·ÒÃìà¡çµáºº PR ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §ÁÒÍÂèÒ§¹éÍ 4 ·ÒÃìà¡çµ


16

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ


º··Õ 2

ä·ÁÁÔ §ÃԤѿàÇÍÃÕ

㹺·¹Õ ¨Ð͸ԺÒ¶֧ ¤ÇÒÁÊӤѭ ¢Í§ä·ÁÁÔ § ÃԤѿàÇÍÃÕ (timing recovery) ¾ÃéÍÁ·Ñ § ͸ԺÒÂËÅÑ¡¡Òà ·Ó§Ò¹¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» (conventional timing recovery) «Ö §ÍÂÙ躹¾× ¹°Ò¹¢Í§ ǧ¨Ãà¿ÊÅçÍ¡ÅÙ» (PLL: phased lock loop) ¹Í¡¨Ò¡¹Õ Âѧ͸ԺÒ¶֧ ÇÔ¸Õ¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅÙ» â´ÂãªéẺ¨ÓÅͧ¡Ò÷ӧҹ¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅٻẺàªÔ§àÊé¹ (linearized PLL model) ÃÇÁ·Ñ §áÊ´§¼Å¡Ò÷´Åͧà¾× ÍãËéàË繶֧¤ÇÒÁÊӤѭ¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕ

2.1

º·¹Ó

ã¹» ¨¨ØºÑ¹¹Õ ¤ÇÒÁà¨ÃÔ­¡éÒÇ˹éÒ ·Ò§´éҹ෤â¹âÅÂÕ ¡ÒÃÊ× ÍÊÒà (communication) ä´é ¾Ñ²¹Òä»ÍÂèÒ§ ÃÇ´àÃçÇ â´Â੾ÒÐÍÂèÒ§ÂÔ §¡ÒÃÊ× ÍÊÒôԨԷÑÅ·Õ ãªéã¹ÃкºµèÒ§æ àªè¹ Ãкºâ·ÃÈѾ·ìà¤Å× Í¹·Õ áÅÐÃкº ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÀÒÂã¹ÍØ»¡Ã³ìÎÒÃì´´ÔÊ¡ìä´Ã¿ì à» ¹µé¹ à¾ÃÒÐ©Ð¹Ñ ¹ ¤ÇÒÁµéͧ¡ÒÃ·Õ ¨Ðà¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾â´ÂÃÇÁ¢Í§Ãкº¨Ö§ à» ¹ ÊÔ § ·Õ ¨Óà» ¹ ÁÒ¡ à¾ÃÒÐÇèÒ ¨Ðä´é ÊÒÁÒöÃѺ áÅÐÊè§ ¢éÍÁÙÅ ä´é ÍÂèÒ§ ¹èÒàª× Ͷ×ÍÁÒ¡ÂÔ §¢Ö ¹ ã¹ÃкºÊ× ÍÊÒôԨԷÑÅ (digital communication system) ¡ÒÃÊè§ÊÑ­­Ò³¨Ò¡µé¹·Ò§ä»Âѧ»ÅÒ ·Ò§ ÍØ»¡Ã³ì µé¹·Ò§¨Ð·Ó˹éÒ·Õ à»ÅÕ Â¹ÊÑ­­Ò³´Ô¨Ô·ÑÅ (digital) ãËé à» ¹ ÊÑ­­Ò³á͹ÐÅçÍ¡ (ana log) ¡è͹·Õ ¨Ð¶Ù¡Êè§ÍÍ¡ä»Âѧ»ÅÒ·ҧ àÁ× ÍÊÑ­­Ò³á͹ÐÅçÍ¡ÁÒ¶Ö§·Õ ÍØ»¡Ã³ì»ÅÒ·ҧ ¡ç¨Ð¶Ù¡Êè§ 17


18

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ä»Âѧ ǧ¨ÃªÑ¡ µÑÇÍÂèÒ§ (sampler) à¾× Í ·Ó¡ÒÃá»Å§ÊÑ­­Ò³á͹ÐÅçÍ¡ãËé ¡ÅѺ ä»à» ¹ ÊÑ­­Ò³´Ô¨Ô·ÑÅ ã¹ÃÙ» ¢Í§ ¢éÍÁÙÅ ÇÔÂص (discrete data) ËÃ×Í ·Õ àÃÕ¡¡Ñ¹ ÇèÒ á«Áà» Å (sample) ¡Òêѡ µÑÇÍÂèÒ§ (sampling) ÊÑ­­Ò³á͹ÐÅçÍ¡·Õ ¼Ô´¨Ñ§ËÇШзÓãËéà¡Ô´¼ÅàÊÕÂËÒÂÍÂèÒ§ÁÒ¡¡Ñº»ÃÐÊÔ·¸ÔÀÒ¾â´ÂÃÇÁ ¢Í§Ãкº ä·ÁÁÔ §ÃԤѿàÇÍÃըзÓ˹éÒ·Õ ã¹¡ÒÃà¢éҨѧËÇÐ (synchronize) ǧ¨ÃªÑ¡µÑÇÍÂèÒ§¡ÑºÊÑ­­Ò³ á͹ÐÅçÍ¡·Õ ä´é ÃѺ à¾× ÍãËé ä´é ¢éÍÁÙÅ á«Áà» Å ·Õ ´Õ ·Õ ÊØ´ ÍÍ¡ÁÒ ¡è͹·Õ ¨Ð¹Ó¢éÍÁÙÅ á«Áà» Å àËÅèÒ¹Ñ ¹ ä»·Ó ¡ÒûÃÐÁÇÅ¼Å¢Ñ ¹ µèÍä» àªè¹ Êè§ µèÍä»Âѧ ÍÕ¤ÇÍäÅà«ÍÃì (equalizer) áÅÐǧ¨Ã¶Í´ÃËÑÊ (decoder) à» ¹µé¹ ´Ñ§¹Ñ ¹ä·ÁÁÔ §ÃԤѿàÇÍÃÕ¨Ö§¹Ñºä´éÇèÒà» ¹Í§¤ì»ÃСͺ·Õ ¤ÇÒÁÊӤѭÁÒ¡ÍÂèÒ§Ë¹Ö §ã¹ÃкºÊ× ÍÊÒà ´Ô¨Ô·ÑÅ à¾ÃÒÐÇèÒ ¶éÒ Ç§¨ÃªÑ¡ µÑÇÍÂèÒ§·Ó§Ò¹äÁè ´Õ ¡ç ¨ÐÊè§ ¼Å·ÓãËé ¢éÍÁÙÅ á«Áà» Å ·Õ ä´é äÁè ÁÕ ¤Ø³ÀÒ¾ËÃ×Í ÁÕ ¢éͼԴ¾ÅÒ´ÁÒ¡ áÅÐàÁ× Í Êè§ ¢éÍÁÙÅ á«Áà» Å àËÅèÒ¹Õ ä»Âѧ ÍÕ¤ÇÍäÅà«ÍÃì áÅÐǧ¨Ã¶Í´ÃËÑÊ ¡ç ¨Ð·ÓãËé ¼ÅÅѾ¸ì·Õ ä´éÁÕ¢éͼԴ¾ÅÒ´ÁÒ¡ ¹Ñ ¹¤×Í ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ (BER: bit error rate) ¢Í§Ãкº¨ÐÁÕ¤èÒ ÊÙ§ «Ö §à» ¹ÊÔ §·Õ ¤ÇèÐËÅÕ¡àÅÕ Â§ â´Â·Ñ Çä»ä·ÁÁÔ §ÃԤѿàÇÍÃըзӧҹÍÂÙ躹¾× ¹°Ò¹¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅÙ» (PLL) [4] «Ö §»ÃÐ¡Íºä» ´éÇ ǧ¨ÃµÃǨËÒ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ (TED: timing error detector), ǧ¨Ã¡ÃͧÅÙ» (loop lter), áÅÐǧ¨Ã VCO (voltage controlled oscillator) ã¹·Ò§»¯ÔºÑµÔ â¤Ã§ÊÃéÒ§¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕ¨ÐÁÕÍÂÙè 2 ÃٻẺ ¤×Í áºº¹ÔùÑ (deductive) áÅÐẺÍØ»¹Ñ (inductive) â´Â¨Ð¢Ö ¹ÍÂÙè¡ÑºÇèÒ ¢éÍÁÙÅ·Ò§àÇÅÒ ¶Ù¡ ´Ö§ ÍÍ¡ÁÒãªé ³ µÓáË¹è§ ¡è͹ËÃ×Í ËÅѧ ǧ¨ÃªÑ¡ µÑÇÍÂèÒ§ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº¹ÔùÑ (deductive timing recovery) ¨Ð´Ö§¢éÍÁÙÅ·Ò§àÇÅÒ ·Õ àÃÕ¡¡Ñ¹ÇèÒ timing tone [16] ¨Ò¡ÊÑ­­Ò³á͹ÐÅçÍ¡·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃªÑ¡µÑÇÍÂèÒ§ µÒÁÃÙ»·Õ 2.1 â´Â·Õ ǧ¨Ã PLL ¶Ù¡¹Óãªéà¾× ÍÅ´¼Å¡Ãзº¢Í§ä·ÁÁÔ § 1

¨ÔµàµÍÃì

(timing jitter) ·Õ ὧÍÂÙèã¹ÊÑ­­Ò³á͹ÐÅçÍ¡ ã¹¢³Ð·Õ ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺÍØ»¹Ñ (in

ductive timing recovery) ¨Ðãªéǧ¨Ã PLL Ẻ» ͹¡ÅѺ (feedback) à¾× Í·Ó˹éÒ·Õ ã¹¡Òô֧¢éÍÁÙÅ ·Ò§àÇÅÒ¨Ò¡ÊÑ­­Ò³á͹ÐÅçÍ¡·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨ÃªÑ¡ µÑÇÍÂèÒ§ ´Ñ§·Õ áÊ´§ã¹ÃÙ» ·Õ 2.2 ¢éÍ´Õ ¢Í§ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áººÍØ»¹Ñ ¤×Í ·Ø¡ Êèǹ»ÃСͺ¢Í§ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº¹Õ ÊÒÁÒö·Õ ¨Ð¶Ù¡ ÊÃéÒ§ãËé à» ¹áºº´Ô¨Ô·ÑÅä´é à¹× ͧ¨Ò¡ ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺÍØ»¹ÑÂà» ¹·Õ ¹ÔÂÁãªé§Ò¹¡Ñ¹ÁÒ¡ ´Ñ§¹Ñ ¹ ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ¨ÐàÃÕ¡ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº¹Õ ÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä» (conventional timing

1

ä·ÁÁÔ § ¨Ôµ àµÍÃì ¤×Í ÊÑ­­Ò³Ãº¡Ç¹ÀÒÂã¹Ç§¨ÃªÑ¡ µÑÇÍÂèÒ§

¤ÃÑ §à¡Ô´¤ÇÒÁ¤ÅÒ´à¤Å× Í¹ä»¨Ò¡µÓáË¹è§·Õ µéͧ¡ÒÃ

«Ö § ¨Ð·ÓãËé ¨Ñ§ËÇТͧàÇÅÒ·Õ ¨Ð·Ó¡Òêѡ µÑÇÍÂèÒ§áµèÅÐ


2.2.

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

19

yk

received signal

y(t)

to data detection

t k = kT + τˆk timing tone detector

loop filter

TED

VCO

ÃÙ»·Õ 2.1: ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ¹ÔùÑÂ

ak

H(D)

rk

q(t)

τk

n(t) p(t)

y(t) LPF

yk = y (kT + τˆk )

t k = kT + τˆk

Viterbi detector

âk

z − d symbol detector

VCO D 1− D

loop filter

α+

β εˆk 1− D

rˆk −d

yk − d TED

ÃÙ»·Õ 2.2: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ ¾ÃéÍÁ¡Ñºä·ÁÁÔ §ÃԤѿàÇÍÃÕẺÍØ»¹ÑÂ

recovery) ÊÓËÃѺÃÒÂÅÐàÍÕ´à¡Õ ÂǡѺä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ¹ÔùÑÂÊÒÁÒöÈÖ¡ÉÒà¾Ô ÁàµÔÁä´é¨Ò¡ [20]

2.2

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ (ideal channel model) ã¹ÃÙ» ·Õ 2.2 ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

ak ∈ {±1}

«Ö §ÁÕ¤ÒºàÇÅҢͧºÔµ

¤×Í ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô µÑÇ·Õ

k

T

¶Ù¡Ê觼èÒ¹ä»ÂѧªèͧÊÑ­­Ò³

¢Í§ªèͧÊÑ­­Ò³,

D

H(D) =

k=0 hk D

k â´Â·Õ

hk

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ˹èǧàÇÅÒ (delay operator), áÅÐ


20

ν

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¤×Í Ë¹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³ ´Ñ§¹Ñ ¹ ÊÑ­­Ò³ read back ÊÒÁÒöà¢Õ¹໠¹ÊÁ¡ÒÃä´é ¤×Í

p(t) =

X

rk q(t − kT − τk ) + n(t)

(2.1)

k â´Â·Õ

rk =

P

i ak−i hi ¤×Í ¢éÍÁÙÅ àÍÒµì¾Øµ ¢Í§ªèͧÊÑ­­Ò³·Õ »ÃÒȨҡÊÑ­­Ò³Ãº¡Ç¹,

sin(πt/T )/(πt/T )

q(t) =

¤×Í ¿ §¡ìªÑ¹«Ô§¡ì (sinc function) ËÃ×ÍÊÑ­­Ò³¾ÑÅÊì乤ÇÔµÊì (Nyquist pulse)

·Õ ÁÕ áº¹´ìÇÔ´·ì à¡Ô¹ à» ¹ ÈÙ¹Âì (zero excess bandwidth) [16], (unknown timing o set) µÑÇ ·Õ

k,

áÅÐ

n(t)

τk

¤×Í ÍͿ૵·Ò§àÇÅÒ·Õ äÁè ·ÃÒº¤èÒ

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN:

additive white Gaussian noise) ·Õ ÁÕ ¤ÇÒÁ˹Òá¹è¹ Ê໡µÃÑÁ ¡ÓÅѧ (power spectrum density) ẺÊͧ´éÒ¹à·èÒ ¡Ñº 2

à´Ô¹áººÊØèÁ

N0 /2

ã¹Ë¹Ñ§Ê×Í ¹Õ ÍͿ૵·Ò§àÇÅÒ

τk

¨Ð¶Ù¡ ¨ÓÅͧãËé ÁÕ ÅѡɳÐà» ¹ ¡ÒÃ

(random walk) «Ö §¹ÔÂÒÁâ´Â [21]

τk+1 = τk + wk àÁ× Í

wk

¤×Í µÑÇá»ÃÊØèÁà¡ÒÊìà«Õ¹Ẻ

i.i.d. (independent and identically distributed) ·Õ ÁÕ¤èÒà©ÅÕ Â

(mean) à·èҡѺ¤èÒÈÙ¹Âì áÅÐÁÕ¤èÒ¤ÇÒÁá»Ã»Ãǹ (variance) à·èҡѺ

2) N (0, σw

â´Â¤èÒ

σw

(2.2)

2 σw

ËÃ×Íà¢Õ¹᷹ä´é´éÇÂ

¨Ðà» ¹µÑÇ¡Ó˹´ÃдѺ¤ÇÒÁÃعáç¢Í§ä·ÁÁÔ §¨ÔµàµÍÃì 3

·Õ ǧ¨ÃÀÒ¤ÃѺ ÊÑ­­Ò³ read back ¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧǧ¨Ã¡Ãͧ¼èÒ¹µ Ó ÍÔÁ¾ÑÅÊìà·èҡѺ

q(t)/T

wk ∼

¹Ñ ¹¤×Í ÁÕ¤ÇÒÁ¶Õ µÑ´ (cut o frequency) à·èҡѺ

(LPF) ·Õ Áռŵͺʹͧ

1/(2T )

à¾× Í·Ó˹éÒ·Õ ¡Ó¨Ñ´

ÊÑ­­Ò³Ãº¡Ç¹·Õ ÍÂÙè¹Í¡á¶º¤ÇÒÁ¶Õ (out of band noise) ¨Ò¡¹Ñ ¹ ¡ç¨Ð·Ó¡ÒêѡµÑÇÍÂèÒ§ÊÑ­­Ò³ read back ·Õ àÇÅÒ

kT + τ̂k

à¾× Í·ÓãËéä´éà» ¹¢éÍÁÙÅá«Áà» Å

yk = y(kT + τ̂k ) =

X

ri q(kT + τ̂k − iT − τi ) + nk

(2.3)

i 2

Ẻ¨ÓÅͧ¡ÒÃà´Ô¹ ẺÊØèÁ ¹Õ ¶Ù¡ ¹Óãªé à¾ÃÒÐÇèÒ à» ¹ Ẻ¨ÓÅͧ·Õ §èÒ áÅÐÊÒÁÒöãªé á·¹ÅѡɳТͧªèͧÊÑ­­Ò³

µèÒ§æ ä´é§èÒ â´Â¡ÒÃà»ÅÕ Â¹á»Å§¤èÒ¾ÒÃÒÁÔàµÍÃì 3

2 σw

à¾Õ§µÑÇà´ÕÂÇà·èÒ¹Ñ ¹

ÊÓËÃѺ Ãкº·Õ ÁÕ á¶º¤ÇÒÁ¶Õ ¨Ó¡Ñ´ (band limited system) ¹Ñ ¹¤×Í ¾Åѧ§Ò¹¢Í§ÊÑ­­Ò³¨Ð¶Ù¡ ¨Ó¡Ñ´ ãËé ÍÂÙè 㹪èǧᶺ

¤ÇÒÁ¶Õ

|f | ≤ 1/(2T )

ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó ¨Ð·ÓãËé ¢éÍÁÙÅ àÍÒµì¾Øµ ·Õ ä´é ÁÕ ¤èÒ ·Ò§Ê¶ÔµÔ ·Õ ¾Íà¾Õ§ (su cient statistic) [23]

àËÁ×͹¡Ñº¡ÒÃãªéǧ¨Ã¡ÃͧàËÁÒÐÊØ´ (matched lter) [16]


2.2.

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

τ̂k

àÁ× Í

¤×Í ¤èÒ »ÃÐÁÒ³¢Í§

τk

ËÃ×Í ·Õ àÃÕ¡ÇèÒ ÍͿ૵·Ò§à¿Ê (phase o set) µÑÇ ·Õ

ªÑ¡ µÑÇÍÂèÒ§, áÅÐ

nk

N0 /(2T )

nk ∼ N (0, σn2 )

¹Ñ ¹¤×Í

21

k

¢Í§¡ÒÃ

¤×Í AWGN ·Õ ÁÕ ¤èÒà©ÅÕ Â à·èÒ ¡Ñº ¤èÒ ÈÙ¹Âì áÅФèÒ ¤ÇÒÁá»Ã»Ãǹà·èÒ ¡Ñº

σn2 =

ǧ¨Ã TED ¨Ð·Ó˹éÒ·Õ ã¹¡ÒäӹdzËÒ¤èÒ»ÃÐÁÒ³¢Í§¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ (timing error) =

τk − τ̂k

²k

«Ö §¡ç¤×Í ¤èÒ¤ÇÒÁäÁèµÃ§¡Ñ¹ÃÐËÇèÒ§à¿Ê¢Í§ÊÑ­­Ò³á͹ÐÅçÍ¡·Õ ä´éÃѺ¡Ñºà¿Ê¢Í§ÊÑ­­Ò³

¹ÒÌ ¡Ò¢Í§Ç§¨Ã PLL ã¹·Ò§»¯ÔºÑµÔ áÅéÇ Ç§¨Ã TED ÁÕ ËÅÒ»ÃÐàÀ· [4] â´Â¨Ð¢Ö ¹ ÍÂÙè ¡Ñº ÅѡɳР¡ÒùӢéÍÁÙÅ·Õ ´éÒ¹¢Òà¢éҢͧǧ¨Ã TED ÁÒãªé§Ò¹ «Ö §â´Â·Ñ Çä»áÅéÇ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕ ¨Ð¢Ö ¹ ÍÂÙè ¡Ñº ¤Ø³ÀÒ¾¢Í§Ç§¨Ã TED ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒÐǧ¨Ã TED ·Õ ¹ÔÂÁãªé §Ò¹¡Ñ¹ ¹Ñ ¹¤×Í Ç§¨Ã TED Ẻ Mueller and Müller ËÃ×ÍàÃÕ¡ÊÑ ¹æ ÇèÒ M&M TED [24] «Ö §¨Ð¤Ó¹Ç³ ËÒ¤èÒ»ÃÐÁÒ³¢Í§¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ ´Ñ§¹Õ

²̂k = KT {yk r̂k−1 − yk−1 r̂k } â´Â·Õ =

²

r̂k

¤×Í ¤èÒ»ÃÐÁÒ³¢Í§

rk , KT

(2.4)

¤×Í ¤èÒ¤§µÑÇ (constant) ·Õ ¶Ù¡ãªéà¾× Í·ÓãËéÁÑ ¹ã¨ä´éÇèÒ

E[²̂k |²]

àÁ× Í ÃдѺ ÍѵÃÒÊèǹ¤èÒ ¡ÓÅѧ à©ÅÕ Â ¢Í§ÊÑ­­Ò³·Õ µéͧ¡ÒõèÍ ¤èÒ ¡ÓÅѧ à©ÅÕ Â ¢Í§ÊÑ­­Ò³Ãº¡Ç¹

(SNR: signal to noise ratio) ÁÕ¤èÒÊÙ§ ËÃ×ÍÍÒ¨¨Ð¡ÅèÒÇä´éÇèÒ ¤èÒ

KT

¶Ù¡¹ÓÁÒãªéà¾× Í·ÓãËé¤ÇÒÁªÑ¹

¢Í§àÊé¹â¤é§ÃÙ»µÑÇàÍÊ (S curve) [4] ÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § ³ ¨Ø´¡Óà¹Ô´, áÅÐ

E[·]

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ

¤èÒ ¤Ò´ËÁÒ (expectation operator) [10, 25, 26] ¨Ò¡ÊÁ¡Òà (2.4) ¨ÐàËç¹ä´éÇèÒ »ÃÐÊÔ·¸ÔÀÒ¾ ¢Í§Ç§¨Ã TED ¨Ð¢Ö ¹ ÍÂÙè ¡Ñº ¤èÒ µÑ´ÊԹ㨠(decision)

{r̂k }

´Ñ§¹Ñ ¹ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ § ÃԤѿàÇÍÃÕ

¨Ðà» ¹¿ §¡ìªÑ¹¢Í§¤ÇÒÁ¹èÒàª× Ͷ×ͧ͢¤èҵѴÊÔ¹ã¨áÅÐ SNR ·Õ ãªé ¨Ö§à» ¹à˵ؼÅÇèÒ ·ÓäÁǧ¨ÃµÃǨËÒ 4

ÊÑ­Åѡɳì (symbol detector) ·Õ ãªéã¹ä·ÁÁÔ §ÃԤѿàÇÍÃÕ ¤×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ [15] ·Õ ÁÕ»ÃÔÁÒ³¡ÒÃ˹èǧàÇÅÒÊÓËÃѺ¡ÒõѴÊԹ㨠(decision delay) à·èҡѺ

dT

(Viterbi detector)

˹èÇ (àªè¹

d = 4)

á·¹¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹áººËÅÒÂÃдѺ (multi level threshold detector) [27] ËÅѧ¨Ò¡¹Ñ ¹ ¤èÒ»ÃÐÁÒ³¢Í§¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ

²̂k

¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧǧ¨Ã¡ÃͧÅÙ» à¾× ͡ӨѴ

ÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙè ã¹ÊÑ­­Ò³¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ áÅÐÍͿ૵·Ò§à¿Ê¢Í§¡Òêѡ µÑÇÍÂèÒ§ (sampling phase o set) µÑÇ ¶Ñ´ ä» 4

τ̂k+1

¡ç ¨Ð¶Ù¡ »ÃѺ ¤èÒ (update) â´Âǧ¨Ã PLL Íѹ´Ñº ·Õ Êͧ

ÈÖ¡ÉÒÃÒÂÅÐàÍÕ´ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ä´é㹺··Õ 4


22

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

(second order PLL) µÒÁ¤ÇÒÁÊÑÁ¾Ñ¹¸ì´Ñ§¹Õ [4]

àÁ× Í

θ̂k

θ̂k+1 = θ̂k + β²̂k ,

(2.5)

τ̂k+1 = τ̂k + α²̂k + θ̂k+1

(2.6)

¤×Í ¤èÒ »ÃÐÁÒ³¢Í§¢éͼԴ¾ÅÒ´·Ò§¤ÇÒÁ¶Õ (frequency error) [27], áÅÐ

α

áÅÐ

β

¤×Í

¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL [4] «Ö §¨Ðà» ¹µÑÇ¡Ó˹´ ẹ´ìÇÔ´·ì¢Í§ÅÙ» (loop bandwidth) áÅÐÍѵÃÒ ¡ÒÃÅÙèà¢éÒ (convergence rate) ¡ÅèÒǤ×Í ¶éÒ¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL ÂÔ §ÁÒ¡ ẹ´ìÇÔ´·ì¢Í§ÅÙ»¡ç ¨Ð¡ÇéÒ§ «Ö §¨Ð·ÓãËéÍѵÃÒ¡ÒÃÅÙèà¢éÒ¡ç¨ÐàÃçÇ áµèÊÑ­­Ò³Ãº¡Ç¹·Õ à¢éÒÁÒã¹Ç§¨Ã PLL ¡ç¨ÐÁÕÁÒ¡ ÊÓËÃѺ ã¹¡Ã³Õ·Õ ÃкºÁÕ੾ÒТéͼԴ¾ÅÒ´·Ò§à¿Ê (phase error) à·èÒ¹Ñ ¹ ǧ¨ÃÀÒ¤ÃѺÍÒ¨¨Ð¹Óǧ¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö § ( rst order PLL) ÁÒãªé§Ò¹á·¹Ç§¨Ã PLL Íѹ´Ñº·Õ Êͧ¡çä´é â´Â·Õ ÍͿ૵·Ò§à¿Ê ¢Í§¡ÒêѡµÑÇÍÂèÒ§µÑǶѴä»

τ̂k+1

¨Ð¶Ù¡»ÃѺ¤èÒµÒÁ¤ÇÒÁÊÑÁ¾Ñ¹¸ì´Ñ§¹Õ [4]

τ̂k+1 = τ̂k + α²̂k

(2.7)

â´Â·Ñ Çä» ä·ÁÁÔ §ÃԤѿàÇÍÃըзӧҹ໠¹ 2 ÀÒÇÐ (mode) ¤×Í

1) ÀÒÇСÒÃä´é ÁÒ (acquisition mode) ¨Ð·Ó§Ò¹ã¹µÍ¹àÃÔ Áµé¹ ¢Í§¡Ãкǹ¡ÒÃà¢éÒ ¨Ñ§ËÇдéÇ ¤ÇÒÁªèÇÂàËÅ×Í ¢Í§áºº¢éÍÁÙÅ (data pattern) ·Õ àÃÕ¡ÇèÒ preamble [27] à¹× ͧ¨Ò¡ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ ÃÙé á¹è¹Í¹ÇèÒ preamble ÁÕ ÅѡɳÐà» ¹ ÍÂèÒ§äà ¨Ö§ ·ÓãËé ÊÒÁÒö·ÃÒºä´é ÇèÒ ¤èÒ ¶Ù¡µéͧ¤×ͤèÒÍÐäà ´Ñ§¹Ñ ¹ã¹ªèǧÀÒÇСÒÃä´éÁÒ¹Õ Ç§¨Ã PLL ¨Ðãªé¤èÒ ËÒ¤èÒ

²̂k

r̂k = rk

r̂k

·Õ

㹡Òäӹdz

µÒÁÊÁ¡Òà (2.4) (ǧ¨ÃµÃǨËÒÊÑ­Åѡɳì·Õ ãªéã¹ä·ÁÁÔ §ÃԤѿàÇÍÃÕ¨ÐÂѧäÁè¶Ù¡ãªé§Ò¹

㹪èǧ¹Õ ) «Ö § ¨Ð·ÓãËé ä´é ¤èÒ ·Õ ¶Ù¡µéͧ à¾ÃÒÐ©Ð¹Ñ ¹ ¡Ãкǹ¡ÒÃä·ÁÁÔ § ÃԤѿàÇÍÃÕ ã¹ªèǧ¹Õ ¨Ö§ ÁÕ ¤ÇÒÁ¹èÒàª× Ͷ×ÍÁÒ¡ â´Â¨Ø´»ÃÐʧ¤ìËÅÑ¡¢Í§ÀÒÇСÒÃä´éÁÒ ¡ç¤×Í ¡ÒÃËÒ¤èÒ»ÃÐÁÒ³àÃÔ Áµé¹¢Í§ ÍͿ૵·Ò§à¿ÊáÅÐÍͿ૵·Ò§¤ÇÒÁ¶Õ (frequency o set) ·Õ ὧÍÂÙè ã¹ÊÑ­­Ò³á͹ÐÅçÍ¡ ·Õ ¨Ð·Ó¡ÒêѡµÑÇÍÂèÒ§

2) ÀÒÇСÒõԴµÒÁ (tracking mode) ¨Ð·Ó§Ò¹µèͨҡÀÒÇСÒÃä´éÁÒ â´Âã¹¢Ñ ¹µÍ¹¹Õ ¤èÒ ãªé㹡ÒäӹdzËÒ¤èÒ

²̂k

r̂k

·Õ

µÒÁÊÁ¡Òà (2.4) ¨Ðä´éÁҨҡǧ¨ÃµÃǨËÒÊÑ­Åѡɳì·Õ ãªéã¹ä·ÁÁÔ §

ÃԤѿàÇÍÃÕ («Ö §ÍÒ¨¨ÐÁդسÀÒ¾äÁè´Õ àÁ× Íà·Õº¡Ñº¡ÒÃãªé

rk

¨ÃÔ§æ) ´Ñ§¹Ñ ¹ ¨Ø´»ÃÐʧ¤ìËÅÑ¡¢Í§


2.3.

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

23

ÀÒÇСÒõԴµÒÁ ¡ç ¤×Í à» ¹ ¡ÒÃá¡éä¢áÅлÃѺ»Ãا ¤èÒ àÃÔ Áµé¹ ¢Í§ÍͿ૵·Ò§à¿ÊáÅÐÍͿ૵ ·Ò§¤ÇÒÁ¶Õ ·Õ ä´éÁÒ¨Ò¡ÀÒÇСÒÃä´éÁÒ

¨ÐàËç¹ä´éÇèÒ㹪èǧÀÒÇСÒÃä´éÁÒ Ç§¨Ã PLL ·ÃÒºá¹è¹Í¹ÇèÒ preamble ¤×ÍÍÐäà ´Ñ§¹Ñ ¹ ǧ¨Ã PLL ¨Ö§ÊÒÁÒö·Õ ¨Ðãªé¤èÒ

α

áÅÐ

β

α

áÅÐ

β

·Õ ÁÕ¤èÒÁÒ¡ä´é à¾× ͪèÇ·ÓãËéÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ ÍÂèÒ§äáçµÒÁ ¤èÒ

·Õ ãªé¤ÇÃ·Õ ¨ÐÁÕ¤èÒŴŧ àÁ× Íä·ÁÁÔ §ÃԤѿàÇÍÃÕà¢éÒÊÙèªèǧÀÒÇСÒõԴµÒÁ à¾× ÍÅ´¼Å¡Ãзº¢Í§

ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ðà¢éÒÁÒã¹Ç§¨Ã PLL [28] ´Ñ§¹Ñ ¹ ¹Ñ¡Í͡ẺÃкº¨Ðµéͧ»ÃйջÃйÍÁÃÐËÇèÒ§ ẹ´ìÇÔ´·ì¢Í§ÅÙ»áÅÐÍѵÃÒ¡ÒÃÅÙèà¢éÒ ã¹ÃÐËÇèÒ§·Õ ·Ó¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì

2.3

α

áÅÐβ

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

¨Ò¡¼Å¡Ò÷´Åͧ·Õ ä´é ÃѺ ¨Ò¡¡ÒèÓÅͧÃкº (system simulation) ¾ºÇèÒ ÇÔ¸Õ¡ÒÃ·Õ ´Õ ·Õ ÊØ´ 㹡Òà Í͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL (·Ñ §

α áÅÐ β ) ¤×Í ¡ÒÃàÅ×Í¡ãªé¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL ·Õ

·ÓãËéÃкºÁÕ BER ¹éÍÂÊØ´ ³ ·Õ ǧ¨ÃÀÒ¤ÃѺ ÍÂèÒ§äáçµÒÁ ÇÔ¸Õ¹Õ äÁèÊÒÁÒöãªé§Ò¹ä´é¨Ãԧ㹷ҧ»¯ÔºÑµÔ à¹× ͧ¨Ò¡ µéͧãªé ÃÐÂÐàÇÅҹҹ㹡ÒÃ·Õ ¨ÐËÒ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL ·Õ àËÁÒÐÊØ´ ´Ñ§¹Ñ ¹ â´Â ·Ñ Çä»áÅéÇ áºº¨ÓÅͧ¡Ò÷ӧҹ¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅٻẺàªÔ§àÊé¹ÁÑ¡¨Ð¶Ù¡¹ÓÁÒãªé㹡ÒÃÍ͡Ẻ¤èÒ ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL [4] ࡳ±ì (criterion) Ë¹Ö §·Õ à» ¹ä»ä´é ¡ç¤×Í ¡ÒÃàÅ×Í¡ãªé¤èÒ

α

áÅÐ

β

·Õ

·ÓãËé ¼ÅµÍºÊ¹Í§¢Í§Ãкº (system response) ÊÒÁÒö·Õ ¨ÐµÒÁ·Ñ¹ ¡ÒÃà»ÅÕ Â¹á»Å§¢Í§à¿ÊáÅÐ ¤ÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³ read back ÀÒÂ㹪èǧ C ºÔµ ËÃ×Í ¤ÒºàÇÅҢͧºÔµ (bit period) «Ö § ÍÒ¨¨Ð ¾Ô¨ÒóÒä´éÇèÒ à¡³±ì¹Õ ÊÍ´¤Åéͧ¡ÑºÍѵÃÒ¡ÒÃÅÙèà¢éÒ ¡ÅèÒǤ×Í ¶éÒ¤èÒ

C

ÂÔ §¹éÍ ¡çËÁÒ¤ÇÒÁÇèÒ ä·ÁÁÔ §

ÃԤѿàÇÍÃÕ¨ÐÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒ·Õ ÂÔ §àÃçÇ

2.3.1

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊ鹢ͧǧ¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö §

ǧ¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö § ( rst order PLL) ¨ÐÊÒÁÒö¨Ñ´¡ÒáѺ¢éͼԴ¾ÅÒ´·Ò§à¿Êä´éà¾Õ§ÍÂèÒ§à´ÕÂÇ (äÁè ÊÒÁÒö¨Ñ´¡ÒáѺ ¢éͼԴ¾ÅÒ´·Ò§¤ÇÒÁ¶Õ ä´é) ã¹Êèǹ¹Õ ¨Ð͸ԺÒ¡ÒÃÍ͡Ẻ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§ ǧ¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö § «Ö §¨Ðà» ¹»ÃÐ⪹ìµèÍ¡ÒÃàÃÕ¹ÃÙéǧ¨Ã PLL Íѹ´ÑºÊÙ§æ µèÍä» ÊÓËÃѺ㹡Òà ÇÔà¤ÃÒÐËì¹Õ ¢éͼԴ¾ÅÒ´·Ò§à¿Ê¨Ð¶Ù¡¨ÓÅͧãËéà» ¹¿ §¡ìªÑ¹¢Ñ ¹ºÑ¹ä´ (step function) ¹Ñ ¹¤×Í

τk = T


24

àÁ× Í

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

k≥0

áÅÐ

τk = 0

àÁ× Í

k<0

¾Ô¨ÒóÒÊÁ¡ÒûÃѺ¤èÒÍͿ૵·Ò§à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§µÑǶѴä»

τ̂k+1

¢Í§Ç§¨Ã PLL Íѹ´Ñº

·Õ Ë¹Ö § ´Ñ§µèÍ仹Õ

τ̂k+1 = τ̂k + α²̂k−d àÁ× Í

d

(2.8)

¤×Í »ÃÔÁҳ˹èǧàÇÅÒã¹ä·ÁÁÔ §ÅÙ» (timing loop) ÁÕ˹èÇÂà» ¹ºÔµà«ÅÅì

¤èÒ»ÃÐÁÒ³¢Í§¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ, timing error), áÅÐ

vk

²k

=

τk − τ̂k

T , ²̂k

=

²k + vk

¤×Í

¤×Í ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ·Õ ËŧàËÅ×ÍÍÂÙè (residual

¤×Í ÊÑ­­Ò³Ãº¡Ç¹ã¹Ç§¨Ã TED ¶éÒÊÁÁصÔãËé

vk

ÁÕ¤èÒ¹éÍÂÁÒ¡ (ÊÒÁÒö·Õ

¨Ðà¾Ô¡à©Âä´é) ´Ñ§¹Ñ ¹ ¿ §¡ìªÑ¹¶èÒÂâ͹ (transfer function) [12] ¢Í§ÃкºµÒÁÊÁ¡Òà (2.8) ÊÒÁÒö ·Õ ¨ÐËÒä´éâ´Â¡ÒÃãªé¡ÒÃá»Å§«Õ (Z transform) [12, 16] ¹Ñ ¹¤×Í

G(z) = àÁ× Í

Γ̂(z)

áÅÐ

¢éͼԴ¾ÅÒ´

²k

Γ(z) =

Γ̂(z) αz −(d+1) = Γ(z) 1 − z −1 + αz −(d+1)

¤×Í ¼Å¡ÒÃá»Å§«Õ¢Í§

τk − τ̂k

τ̂k

áÅÐ

E(z)

µÒÁÅӴѺ à¾ÃÒÐ©Ð¹Ñ ¹ ¿ §¡ìªÑ¹¶èÒÂâ͹¢Í§

ÊÒÁÒöà¢Õ¹ä´éà» ¹

E(z) = Γ(z) − Γ̂(z) = àÁ× Í

τk

(2.9)

¤×Í ¼Å¡ÒÃá»Å§«Õ¢Í§

1 − z −1 Γ(z) 1 − z −1 + αz −(d+1)

(2.10)

²k

ࡳ±ì ·Õ ÍÒ¨¨Ð¹ÓÁÒãªé 㹡ÒÃàÅ×Í¡¤èÒ

α

¤×Í ¡ÒÃàÅ×Í¡¤èÒ

α

·Õ ·ÓãËé àʶÕÂÃÀÒ¾ (stable) ¢Í§

ÃкºáÅÐÍѵÃÒ¡ÒÃÅÙèà¢éÒà» ¹·Õ ¹èÒ¾Í㨠ÊÓËÃѺáµèÅФèҢͧ»ÃÔÁҳ˹èǧàÇÅÒã¹ÅÙ»

d

(loop delay)

·Õ ¡Ó˹´ãËéÁÒ ÇÔ¸Õ¡ÒÃ¹Õ ÊÒÁÒö·Óä´éâ´Âãªé·Ñ §ÊÁ¡Òà (2.9) ËÃ×Í (2.10) ¡çä´é ¡ÅèÒǤ×Í ¢Ñ ¹µÍ¹áá ãËé ËÒ¤èÒ

α

·Ñ §ËÁ´·Õ ·ÓãËé ÃкºÁÕ ¤ÇÒÁàʶÕÂÃÀÒ¾ ¨Ò¡ÊÁ¡Òà (2.9) áÅÐ (2.10) Ãкº¨ÐÁÕ ¤ÇÒÁ

àʶÕÂÃÀÒ¾ ¡ç µèÍàÁ× Í ·Ø¡ â¾Å (all poles) ËÃ×Í ÃÒ¡¤ÓµÍº (root) ¢Í§µÑÇ Êèǹ ¢Í§ÊÁ¡Òà (2.9) ËÃ×Í (2.10) ÍÂÙèÀÒÂã¹Ç§¡ÅÁË¹Ö §Ë¹èÇ [16] ¨Ò¡¡ÒÃá¡éÊÁ¡ÒèÐä´éÇèÒ ¤èÒ

α

·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁ

àʶÕÂÃÀÒ¾ ÊÒÁÒöËÒä´é¨Ò¡ [4]

µ 0 < α < 2 sin ÃÙ» ·Õ 2.3 áÊ´§¤èÒ ÁÒ¡ÊØ´ ¢Í§

α

π 4d + 2

·Õ Âѧ¤§·ÓãËé ÃкºÁÕ ¤ÇÒÁàʶÕÂÃÀÒ¾ ÊÓËÃѺ áµèÅФèÒ

(2.11)

d

¨ÐàËç¹


2.3.

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

25

2

α (PLL gain parameter)

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4

α

max

0.2 0

0

10

20

30

40

50

Normalized loop delay, d

ÃÙ»·Õ 2.3: ¤èÒÁÒ¡ÊØ´¢Í§

α

ä´é ªÑ´à¨¹ÇèÒ ªèǧàʶÕÂÃÀÒ¾¢Í§¤èÒ ¤èÒ

α

·Õ Âѧ¤§·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾ ÊÓËÃѺáµèÅФèÒ

α

¨ÐŴŧÍÂèÒ§ÃÇ´àÃçÇ àÁ× Í

ËÅÒ¤èÒ ·Õ ·ÓãËé ÃкºÁÕ ¤ÇÒÁàʶÕÂÃÀÒ¾ áµè ¨ÐàÅ×Í¡ãªé

α

d

ÁÕ ¤èÒ à¾Ô Á ¢Ö ¹ áÅж֧áÁéÇèÒ ¨ÐÁÕ

à¾Õ§¤èÒ à´ÕÂÇ·Õ ·ÓãËé ¼ÅµÍºÊ¹Í§

¢Í§ÃкºÊÒÁÒö·Õ ¨ÐµÔ´µÒÁ¼ÅµÍºÊ¹Í§¢Ñ ¹ºÑ¹ä´ (step response) ÀÒÂã¹ ¤ÇÒÁ¤ÅÒ´à¤Å× Í¹ä´é

α

±5%

C

ºÔµ â´ÂÂÍÁãËé ÁÕ

â´Â·Õ µÑÇàÅ¢ 5% ¹Õ ¶Ù¡ ¹Óãªé à¾× Í à» ¹ ¡Òüè͹»Ã¹à¡³±ì ¡ÒÃÍ͡Ẻ

à¾× Í·Õ ¨Ðä´éÅ´¼Å¡Ãзº¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ðà¢éÒÁÒã¹Ç§¨Ã PLL ÃÙ»·Õ 2.4(a) áÊ´§¤èÒ

·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾áÅÐÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒÀÒÂã¹ (¹Ñ ¹¤×Í ¤èÒ

C

¾ºÇèÒÁÕà¾Õ§ ÊÓËÃѺ

d

α

ÁÕ¤èÒ¹éÍÂ) ¤èÒ

α

C

¡ç¨ÐÂÔ §ÁÕ¤èÒÁÒ¡ ¨Ò¡·Õ áÊ´§ã¹ÃÙ»·Õ 2.4(a) àÁ× Í¡Ó˹´

ºÒ§¤èÒ·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾ÊÓËÃѺ¤èÒ

µÑ §áµè ¤èÒ 0 ¶Ö§

30T

= 0 ¶Ö§

30T

à¹× ͧ¨Ò¡

·Õ

d

C

ÁÒãËé ¨Ð

¤èÒË¹Ö § µÑÇÍÂèÒ§àªè¹ ¨ÐÁÕ¤èÒ

α100

à·èÒ¹Ñ ¹ ·Õ ·ÓãËé ÃкºÁÕ ¤ÇÒÁàʶÕÂÃÀÒ¾áÅÐÁÕ ÍѵÃÒ¡ÒÃÅÙèà¢éÒ ÀÒÂã¹

«Ö §ÊÍ´¤Åéͧ¡Ñºà¡³±ì¡ÒÃÍ͡Ẻ

α100

αC

ºÔµ Êѧࡵ¨Ð¾ºÇèÒ ÍѵÃÒ¡ÒÃÅÙèà¢éÒÂÔ §àÃçÇ

100 ºÔµ ÃÙ»·Õ 2.4(b) áÊ´§¼ÅµÍºÊ¹Í§¢Ñ ¹ºÑ¹ä´¢Í§ÃкºµÒÁÊÁ¡Òà (2.9) â´Âãªé

d

d

α

α100

ÊÓËÃѺ

·Õ µÑ §äÇé

¶Ù¡Í͡ẺÁÒà¾× ÍãËéÃкºÊÒÁÒö·Õ ¨ÐµÔ´µÒÁ¼ÅµÍºÊ¹Í§¢Ñ ¹ºÑ¹ä´ÀÒÂã¹ 100

ºÔµ ´éǤèÒ¤ÇÒÁ¤ÅÒ´à¤Å× Í¹ÂÔ¹ÂÍÁ (tolerance)

±5% ´Ñ§¹Ñ ¹ ¤èÒÊÑÁºÙóì¢Í§¢¹Ò´¢Í§¼ÅµÍºÊ¹Í§


26

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0.05

α (PLL gain parameter)

0.045 0.04 0.035 0.03

α

0.025

64

0.02

0.01 0.005

α

α200

0.015

100

α300 0

10

20

30

40

50

(a) Normalized loop delay, d

d increases 1.05 0.95

Magnitude

± 5% tolerance

d increases

Convergence rate within 100 samples

0

100

200

300

400

500

(b) Time (in bit periods)

ÃÙ»·Õ 2.4: (a) ¤èÒ

d,

αC

·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾áÅÐÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒÀÒÂã¹

áÅÐ (b) ¼ÅµÍºÊ¹Í§¢Í§ÃкºàÁ× Íãªé

¢éͼԴ¾ÅÒ´ (error response)

E(z)

α100

ÊÓËÃѺ¤èÒ

d

¨Ò¡ 0 ¶Ö§

C

ºÔµ ÊÓËÃѺáµèÅÐ

30T

µÒÁÊÁ¡Òà (2.10) ¤ÇÃ·Õ ¨ÐÁÕ ¢¹Ò´¹éÍ¡ÇèÒ 0.05 ËÅѧ¨Ò¡·Õ

¢éÍÁÙżèÒ¹ä» 100 ºÔµ µÒÁ·Õ áÊ´§ã¹ÃÙ»·Õ 2.5 ÊÓËÃѺ

d = 14T

Êѧࡵ¨Ð¾ºÇèÒ ÁÕ¤èÒ

α100

¨Ó¹Ç¹


2.3.

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

1.4

27

α = 0.055 α = 0.035

Magnitude

1.05 1 0.95

α = 0.0218 α = 0.01 α = 0.005

0

0

14

50

100

150

200

250

(a) Time (in bit periods) 1

Magnitude

α = 0.005 α = 0.0218

α = 0.01

0.05 0 −0.05

α = 0.035 α = 0.055 −0.4

0

50

100

150

200

250

(b) Time (in bit periods)

ÃÙ»·Õ 2.5: (a) ¼ÅµÍºÊ¹Í§¢Ñ ¹ºÑ¹ä´¢Í§Ãкº áÅÐ (b) ¼ÅµÍºÊ¹Í§¢éͼԴ¾ÅÒ´ÊÓËÃѺ áÅФèÒ

α

2 ¤èÒ ¤×Í

d = 14T

µèÒ§æ

α = 0.0218

¤ÅÒ´à¤Å× Í¹ÂÔ¹ÂÍÁ

áÅÐ

±5%

α = 0.055

·Õ ·ÓãËéÃкºÁÕÍѵÃÒ¡ÒÃÅÙèà¢éÒÀÒÂã¹ 100 ºÔµ ´éǤèÒ¤ÇÒÁ

ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ ¤èÒ

α

·Õ ÁÕ¤èÒ¹éÍ¡ÇèÒ ¨Ð¶Ù¡àÅ×Í¡ÁÒãªé§Ò¹ã¹


28

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ǧ¨Ã PLL à¾× Í·Õ ¨Ð·ÓãËéẹ´ìÇÔ´·ì¢Í§ÅÙ»ÁÕ¤èÒ¹éÍ «Ö §¨ÐªèÇÂÅ´¼Å¡Ãзº¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ð à¢éèÒÁÒã¹Ç§¨Ã PLL ä´é

2.3.2

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊ鹢ͧǧ¨Ã PLL Íѹ´Ñº·Õ Êͧ

ã¹¡Ã³Õ ·Õ ÃкºÁÕ Í§¤ì»ÃСͺ¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ (frequency o set) ǧ¨Ã PLL Íѹ´Ñº ·Õ Êͧ áÅÐ

β

ÊÓËÃѺ¤èÒ

d

¨Ðµéͧ¶Ù¡¹ÓÁÒãªé§Ò¹á·¹Ç§¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö § «Ö §¾ÒÃÒÁÔàµÍÃì·Õ ¨Ðµéͧ¤Ó¹Ç³ËÒ ¤×Í ã¹¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL Íѹ´Ñº·Õ Êͧ¹Õ ¨ÐàÃÔ Áµé¹¨Ò¡¡ÒÃËÒ¤èÒ áÅÐ

C

α

α

·Õ ¡Ó˹´ÁÒãËé â´ÂÊÁÁصÔÇèÒ ã¹ÃкºÁÕà¾Õ§á¤è¢éͼԴ¾ÅÒ´·Ò§à¿Êà·èÒ¹Ñ ¹ (¹Ñ ¹¤×Í ãªéÇÔ¸Õ¡ÒÃ

Í͡Ẻ¤èÒ

α

µÒÁ·Õ ͸ԺÒÂã¹ËÑÇ¢éÍ ·Õ 2.3.1) ¨Ò¡¹Ñ ¹ àÁ× Í ä´é ¤èÒ

α

·Õ µéͧ¡ÒÃáÅéÇ ¡ç ¨ÐàÃÔ Á ËÒ¤èÒ

β

â´Âãªé¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊ鹢ͧǧ¨Ã PLL Íѹ´Ñº·Õ Êͧ ÊÓËÃѺ»ÃÔÁÒ³¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ ·Õ ãËéÁÒ

τk

=

¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ ÁÕ˹èÇÂà» ¹à»ÍÃìà«ç¹µì (percent) ¢Í§ºÔµà«ÅÅì

T

«Ö §ã¹¡ÒÃÇÔà¤ÃÒÐËì¹Õ ¢éͼԴ¾ÅÒ´·Ò§¤ÇÒÁ¶Õ ¨Ð¶Ù¡¨ÓÅͧãËéÁÕ¤èÒà» ¹

¾Ô¨ÒóÒÊÁ¡ÒûÃѺ¤èÒÍͿ૵·Ò§à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§µÑǶѴä»

kfd

τ̂k+1

àÁ× Í

fd

¤×Í »ÃÔÁÒ³

¢Í§Ç§¨Ã PLL Íѹ´Ñº

·Õ Êͧ ´Ñ§µèÍ仹Õ

θ̂k+1 = θ̂k + β²̂k−d

(2.12)

τ̂k+1 = τ̂k + α²̂k−d + θ̂k+1

(2.13)

àªè¹à´ÕÂǡѹ ¶éÒÊÁÁصÔÇèÒäÁèÁÕÊÑ­­Ò³Ãº¡Ç¹ã¹Ç§¨Ã TED ¹Ñ ¹¤×Í

²̂k

=

²k

=

τk − τ̂k ,

à¾ÃÒÐ©Ð¹Ñ ¹

¿ §¡ìªÑ¹ ¶èÒÂâ͹¢Í§ÃкºµÒÁÊÁ¡Òà (2.12) áÅÐ (2.13) ÊÒÁÒö·Õ ¨ÐËÒä´é â´Â¡ÒÃãªé ¡ÒÃá»Å§«Õ ´Ñ§¹Õ

G(z) =

Γ̂(z) (α + β)z −(d+1) − αz −(d+2) = Γ(z) 1 − 2z −1 + z −2 + (α + β)z −(d+1) − αz −(d+2)

(2.14)

áÅп §¡ìªÑ¹¶èÒÂâ͹¢Í§¢éͼԴ¾ÅÒ´ ¤×Í

E(z) = Γ(z) − Γ̂(z) =

1 − 2z −1 + z −2 Γ(z) 1 − 2z −1 + z −2 + (α + β)z −(d+1) − αz −(d+2)

㹷ӹͧà´ÕÂǡѹ ࡳ±ì·Õ ÍÒ¨¨Ð¹ÓÁÒãªé㹡ÒÃàÅ×Í¡¤èÒ

β

¤×Í ¡ÒÃàÅ×Í¡¤èÒ

β

(2.15)

·Õ ·ÓãËéàʶÕÂÃÀÒ¾

¢Í§ÃкºáÅÐÍѵÃÒ¡ÒÃÅÙèà¢éÒà» ¹·Õ ¹èÒ¾Í㨠ÊÓËÃѺ»ÃÔÁÒ³¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ ·Õ ¡Ó˹´ÁÒãËé¢Í§


¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

Maximum magnitude of E(z) after C samples

2.3.

0.2

C = 100

0.18

C = 50 0.16

0.5% frequency offset

0.14 0.12 0.1 0.08 0.06 0.04 0.02 0

0.2% frequency offset 0

0.5

ÃÙ»·Õ 2.6: ¢¹Ò´ÁÒ¡ÊØ´¢Í§

áµèÅÐ

d, C

áÅÐ

29

αC

E(z)

1

β (PLL gain parameter)

ËÅѧ¨Ò¡·Õ ¢éÍÁÙżèÒ¹ä»

C

β

2 −3

x 10

ºÔµ àÁ× Íãªé

«Ö § ÊÒÁÒö·Õ ¨Ð¤Ó¹Ç³ËÒä´é ´Ñ§¹Õ ÊÓËÃѺ ¤èÒ

¢Ñ ¹µÍ¹áá ¤×Í ¡ÒÃàÅ×Í¡¤èÒ

1.5

d, C

d = 14

áÅÐ

αC

αC

áÅÐ

·Õ ¡Ó˹´ÁÒãËé

·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾ ¨Ò¡ÊÁ¡Òà (2.14) áÅÐ (2.15) Ãкº

¨ÐÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾¡çµèÍàÁ× Í ·Ø¡â¾ÅËÃ×ÍÃÒ¡¤ÓµÍº¢Í§µÑÇÊèǹ¢Í§ÊÁ¡Òà (2.14) ËÃ×Í (2.15) ÍÂÙè ÀÒÂã¹Ç§¡ÅÁË¹Ö §Ë¹èÇ áÅж֧áÁéÇèÒ¨ÐÁÕ¤èÒ

β

à¾Õ§¤èÒà´ÕÂÇ·Õ ·ÓãËé

E(z)

β

ËÅÒ¤èÒ·Õ ·ÓãËéÃкºÁÕ¤ÇÒÁàʶÕÂÃÀÒ¾ áµè¨ÐàÅ×Í¡

ÁÕ¢¹Ò´¹éÍÂÊØ´ ËÅѧ¨Ò¡·Õ ¢éÍÁÙżèÒ¹ä»

C

ºÔµ à¾× Í·Õ ¨ÐÅ´¼Å¡Ãзº

¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ðà¢éèÒÁÒã¹Ç§¨Ã PLL ÃÙ»·Õ 2.6 áÊ´§¢¹Ò´ÁÒ¡ÊØ´ (maximum magnitude) ¢Í§

E(z)

µÒÁÊÁ¡Òà (2.15) ËÅѧ¨Ò¡·Õ ¢éÍÁÙżèÒ¹ä»

¾ºÇèÒ ÇÔ¸Õ¡ÒÃÇÔà¤ÃÒÐËì¹Õ ¨ÐãËéä´é¤èÒ à¾Õ§áµè¢¹Ò´¢Í§

ËÁÒÂà˵Ø

E(z)

β

C

ºÔµ ÊÓËÃѺ

d = 14T

áÅÐ

αC

Êѧࡵ¨Ð

à» ¹¤èÒà´ÕÂǡѹ â´ÂäÁè¤Ó¹Ö§¶Ö§»ÃÔÁÒ³¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ

¨ÐµèÒ§¡Ñ¹à·èÒ¹Ñ ¹ µÒÁ»ÃÔÁÒ³¢Í§ÍͿ૵·Ò§¤ÇÒÁ¶Õ

ÇÔ¸Õ¡ÒÃÍ͡Ẻ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL µÒÁ·Õ ¡ÅèÒÇÁÒ¢éÒ§µé¹ ¹Õ ¨ÐÍÂÙè º¹¾× ¹°Ò¹

¢Í§ÊÁÁص԰ҹ·Õ ÇèÒ ¤ÇÒÁªÑ¹ ¢Í§àÊé¹â¤é§ ÃÙ» µÑÇ àÍÊ (S curve) ¢Í§ÊÁ¡Òà (2.4) ÁÕ ¤èÒ à» ¹ ¤èÒ Ë¹Ö § ³ ¨Ø´¡Óà¹Ô´ áÅÐäÁè ÁÕ ÊÑ­­Ò³Ãº¡Ç¹ÀÒÂã¹Ç§¨Ã TED à¾ÃÒÐ©Ð¹Ñ ¹ ¡è͹·Õ ¨Ð¹Ó¤èÒ

α

áÅÐ

β

·Õ ä´é


30

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¨Ò¡ÇÔ¸Õ¡ÒÃÍ͡ẺµÒÁ·Õ ¡ÅèÒÇÁÒ¢éÒ§µé¹ ¹Õ ä»ãªé §Ò¹ ¼Ùéãªé ¨Ðµéͧ·ÓãËé ¤ÇÒÁªÑ¹ ¢Í§àÊé¹â¤é§ ÃÙ» µÑÇ àÍÊ ¢Í§Ç§¨Ã TED ÁÕ¤èÒà» ¹¤èÒË¹Ö § ³ ¨Ø´¡Óà¹Ô´¡è͹àÊÁÍ

2.3.3

¡ÒÃËÒàÊé¹â¤é§ÃÙ»µÑÇàÍÊ

ä·ÁÁÔ § ¿ §¡ìªÑ¹ (timing function) ËÃ×Í àÊé¹â¤é§ ÃÙ» µÑÇ àÍÊ (S curve) [22] ¨Ð¹ÔÂÒÁâ´Â ¤èÒà©ÅÕ Â (mean) ¢Í§ =

rk

{²̂k }â´ÂÊÁÁصÔÇèÒ

ÊÓËÃѺ·Ø¡¤èÒ

k,

¤èÒ

r̂k

·Õ ä´é¨Ò¡Ç§¨ÃµÃǨËÒÊÑ­ÅѡɳìÁÕ¤èÒ¶Ù¡µéͧ·Ñ §ËÁ´ ¹Ñ ¹¤×Í

r̂k

áÅТéÍÁÙÅÍÔ¹¾ØµáµèÅкԵäÁèÁÕÊËÊÑÁ¾Ñ¹¸ì¡Ñ¹ (uncorrelated) áÅÐÁÕ¾Åѧ§Ò¹

à·èҡѺ 1 ˹èÇ ´Ñ§¹Ñ ¹

STED (²) = E[²̂k | ², r̂k = rk for ∀k] àÁ× Í

² = τ − τ̂

(2.16)

¤×Í ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ à¹× ͧ¨Ò¡ ÃÙ» ¡ÃÒ¿¢Í§ä·ÁÁÔ § ¿ §¡ìªÑ¹ ÁÕ ÅѡɳФÅéÒÂ

µÑÇÍÑ¡Éà S (àÁ× ÍËÁعÃÙ»¡ÃÒ¿ 90 ͧÈÒ) ´Ñ§¹Ñ ¹ ¨Ö§àÃÕ¡¡Ñ¹ÇèÒ àÊé¹â¤é§ÃÙ»µÑÇàÍÊ (S curve) «Ö § ÊÒÁÒö·Õ ¨Ð¹ÓÁÒãªéÇÑ´»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨Ã TED ä´é ¹Í¡¨Ò¡¹Õ ã¹¡Ã³Õ ·Õ ·ÃÒºÇèÒ ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¢Í§ªèͧÊÑ­­Ò³¤×Í ÍÐäà àÊé¹â¤é§ ÃÙ» µÑÇ àÍÊ ÊÒÁÒö·Õ ¨Ð¤Ó¹Ç³ËÒä´é â´ÂµÃ§ [22] µÑÇÍÂèÒ§àªè¹ ãËé ¾Ô¨ÒóÒẺ¨ÓÅͧ¢Í§ªèͧÊÑ­­Ò³ PR4 (partial response class IV) ã¹ÃÙ»·Õ 2.2 àÁ× Í

H(D)

1 − D2

=

´Ñ§¹Ñ ¹ àÊé¹â¤é§ÃÙ»µÑÇàÍʢͧǧ¨Ã

M&M TED ÊÓËÃѺªèͧÊÑ­­Ò³¹Õ ËÒä´é¨Ò¡

STED (²) = E[²̂k | ², r̂k−1 = rk−1 , r̂k = rk ] = KT E[rk−1 yk − rk yk−1 ] X = KT E[(ak−1 − ak−3 ) ai h(kT − iT − ²) i

− (ak − ak−2 )

X

ai h(kT − T − iT − ²)]

i

= â´Â·Õ

P

rk = ak −ak−2

i ai h(kT

− iT − ²)

3T {−h(−T − ²) + 2h(T − ²) − h(3T − ²)} 16

(2.17)

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµµÑÇ·Õ

k

¢Í§ªèͧÊÑ­­Ò³·Õ »ÃÒȨҡÊÑ­­Ò³Ãº¡Ç¹,

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµµÑÇ·Õ

k

¢Í§Ç§¨ÃªÑ¡µÑÇÍÂèÒ§, áÅÐ

h(t)

=

yk

=

q(t) − q(t − 2T )


2.3.

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

31

n(t) PR-IV pulse

ak

h(t) = q(t) - q(t-2T)

τ

p(t)

LPF

y( t )

yk t k = kT

to data detection

z −d

symbol detector

yk −d

rˆk −d

εˆk

TED

ÃÙ»·Õ 2.7: Ẻ¨ÓÅͧÊÓËÃѺ¡ÒÃËÒàÊé¹â¤é§ÃÙ»µÑÇàÍÊ

¤×Í ÊÑ­­Ò³¾ÑÅÊìẺ PR4 ¤èÒ¤§µÑÇ

KT

=

3T /16

¶Ù¡ãªéà¾× ͪèÇ·ÓãËé¤ÇÒÁªÑ¹¢Í§àÊé¹â¤é§ÃÙ»µÑÇ

àÍÊã¹ÊÁ¡Òà (2.17) ÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § ³ ¨Ø´¡Óà¹Ô´

ÊÓËÃÑºã¹¡Ã³Õ·Õ äÁè·ÃÒº¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¢Í§ªèͧÊÑ­­Ò³ àÊé¹â¤é§ÃÙ»µÑÇàÍÊ¡çÂѧÊÒÁÒö·Õ ¨ÐËÒä´éâ´Â¡Ò÷ӡÒèÓÅͧÃкº «Ö §·Óä´éâ´Â¡ÒÃà» ´ä·ÁÁÔ §ÅÙ»¢Í§áºº¨ÓÅͧã¹ÃÙ»·Õ 2.2 (¡ÅèÒǤ×Í µÑ´Ç§¨Ã¡ÃͧÅÙ»áÅÐǧ¨Ã VCO ÍÍ¡¨Ò¡áºº¨ÓÅͧ) «Ö §¨Ð·ÓãËéä´éà» ¹áºº¨ÓÅͧãËÁèµÒÁÃÙ»·Õ 2.7 ¨Ò¡ÃÙ» ÊÑ­­Ò³

τ

´éÇÂ

²

¨Ð¶Ù¡ ·Ó¡Òêѡ µÑÇÍÂèÒ§·Õ àÇÅÒ

kT

(¹Ñ ¹¤×Í ¡Ó˹´ãËé

¨Ò¡¹Ñ ¹ ·Ó¡ÒäӹdzËÒ¤èÒà©ÅÕ Â ·Ò§àÇÅÒ (time average) ¢Í§

à¾× ÍãËéä´éà» ¹¤èÒ

−0.5T

y(t)

¶Ö§¤èÒ

STED (²)

0.5T

¤èÒà´ÕÂÇ ·ÓÅÑ¡É³Ð¹Õ ä»àÃ× ÍÂæ à¾× ÍËÒ¤èÒ

ÊØ´·éÒ¡ç·Ó¡ÒÃÇÒ´¡ÃÒ¿ÃÐËÇèÒ§

²/T

áÅÐ

τ̂ = 0)

{²̂k }

STED (²)

STED (²)/T

áÅÐá·¹¤èÒ

ÊÓËÃѺ áµèÅФèÒ

ÊÓËÃѺ

²

²

ÊÓËÃѺ¤èÒ

à¾× ÍãËéä´éà» ¹àÊé¹â¤é§

ÃÙ»µÑÇàÍÊ

ÃÙ» ·Õ 2.8 áÊ´§àÊé¹â¤é§ ÃÙ» µÑÇ àÍʢͧǧ¨Ã M&M TED ÊÓËÃѺ ªèͧÊÑ­­Ò³áºº PR4 ·Õ ãªé ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä» ³ ÃдѺ SNR µèÍ ºÔµ ËÃ×Í

Eb /N0

µèÒ§æ ÁÕ Ë¹èÇÂà» ¹ à´«ÔàºÅ

(dB) â´Âãªéǧ¨Ã PLL ¨Ðãªé¤èҵѴÊԹ㨢³ÐË¹Ö §áººá¢ç§ (instantaneous hard decision) ·Õ ä´éÃѺ ¨Ò¡Ç§¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹·Õ äÁèÁÕ˹èǤÇÒÁ¨Ó (memoryless threshold detector) Ẻ 3 ÃдѺ


32

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0.5 0.4 0.3

10 dB

S TED(ε)/T

0.2

20 dB 0.1

Mean 0

Eb/N0 = 5 dB −0.1 −0.2 −0.3 −0.4

Normalized timing funtion −0.5 −0.5

0

0.5

Normalized timing error (ε/T)

ÃÙ»·Õ 2.8: àÊé¹â¤é§ÃÙ»µÑÇàÍʢͧǧ¨Ã M&M TED ÊÓËÃѺªèͧÊÑ­­Ò³ PR4 ·Õ ãªéä·ÁÁÔ §ÃԤѿàÇÍÃÕ áºº·Õ ãªé¡Ñ¹·Ñ Çä»

â´ÂÁÕÃдѺ¢Õ´àÃÔ Áà»ÅÕ Â¹ (threshold level) ·Õ ¤èÒ

     r̂k =

   

±1

¹Ñ ¹¤×Í

2 if yk > 1 −2 if yk < −1

(2.18)

0 else

àÊ鹡ÃÒ¿¢Í§ ä·ÁÁÔ §¿ §¡ìªÑ¹áºº¹ÍÃìÁÍÅäÅ«ì (normalized timing function) ¨Ðä´éÁÒ¨Ò¡ÊÁ¡Òà (2.17) ¨Ò¡ÃÙ»·Õ 2.8 ¨Ð¾ºÇèÒ àÊé¹â¤é§ÃÙ»µÑÇàÍÊÁÕÅѡɳÐÊÁÁÒµÃẺ¤Õ (odd symmetric) àÁ× Íà·Õº ¡Ñº

²=0

«Ö §ËÁÒ¤ÇÒÁËÁÒÂÇèÒ ÍͿ૵·Ò§à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§·Õ ¶Ù¡»ÃѺ¤èÒ´éÇÂǧ¨Ã PLL ¨Ð

ÊÔ ¹ÊØ´ ³ ¨Ø´àʶÕÂÃÀÒ¾ (stable point) ·Õ

²=0

Êѧࡵ¨Ð¾ºÇèÒ àÊé¹â¤é§ÃÙ»µÑÇàÍÊ

ä´é¨Ò¡¡ÒèÓÅͧÃкº¨ÐÊÍ´¤Åéͧ¡Ñºä·ÁÁÔ §¿ §¡ìªÑ¹áºº¹ÍÃìÁÍÅäÅ«ìàÁ× Í à¾ÃÒÐÇèÒ ÊÁÁص԰ҹ·Õ ¡Ó˹´ãËé

r̂k

à» ¹ à˵ؼÅÇèÒ ·ÓäÁªèǧ¢Í§¡ÃÒ¿·Õ

=

rk

ÊÓËÃѺ·Ø¡¤èÒ

STED (²)/T

k

¨Ðãªéä´éäÁè´Õ àÁ× Í

²/T ²/T

STED (²)/T

·Õ

ÁÕ¤èÒ¹éÍ ·Ñ §¹Õ à» ¹ ÁÕ¤èÒÁÒ¡ ´Ñ§¹Ñ ¹¨Ö§

ÊÍ´¤Åéͧ¡Ñº ä·ÁÁÔ § ¿ §¡ìªÑ¹ Ẻ¹ÍÃìÁÍÅäÅ«ì àÁ× Í


2.3.

¡ÒÃÍÍ¡à຺¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL

33

T

Timing estimate

0.5T

Estimated τ 0

T

−0.5T

Actual τ −T

0

500

1000

1500

2000

2500

3000

3500

4000

Time (in bit periods)

ÃÙ»·Õ 2.9: µÑÇÍÂèÒ§ÅѡɳТͧä«à¤ÔÅÊÅÔ»

Eb /N0

ÁÕ¤èÒÊÙ§ ¨Ö§ÁÕ¤ÇÒÁ¡ÇéÒ§ÁÒ¡¡ÇèÒªèǧ¢Í§¡ÃÒ¿ àÁ× Í

¹Í¡¨Ò¡¹Õ ¨Ø´ ·Õ àÊé¹â¤é§ ÃÙ» µÑÇ àÍʵѴ ¡Ñº àÊé¹ á¡¹

x

Eb /N0

ÁÕ¤èÒ¹éÍÂ

¹Ñ ¹¤×Í àÁ× Í

STED (²)/T

0

=

¨ÐàÃÕ¡ÇèÒ

¨Ø´ÊÁ´ØÅ (equilibrium point) ¢Í§¡Ò÷ӧҹ «Ö §à» ¹¨Ø´·Õ ǧ¨Ã PLL ¨ÐÊÒÁÒöµÔ´µÒÁÍͿ૵ ·Ò§àÇÅÒä´é à» ¹ ÍÂèÒ§´Õ ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ ¨Ø´ ÊÁ´ØÅ ¨ÐÁÕ ËÅÒµÓáË¹è§ ¤×Í

±2T, . . . , ±nT

àÁ× Í

²

=

0, ±T ,

n ¤×Í àÅ¢¨Ó¹Ç¹àµçÁ à¹× ͧ¨Ò¡ ÊÑ­­Ò³Ãº¡Ç¹áÅСÒÃú¡Ç¹ (disturbance)

ã¹ÃкºÍÒ¨¨Ð·ÓãËé à¡Ô´ ¢éͼԴ¾ÅÒ´¢¹Ò´ãË­è ÃÐËÇèÒ§¡Ãкǹ¡ÒûÃѺ ¤èÒ ÍͿ૵·Ò§à¿Ê¢Í§¡Òà ªÑ¡ µÑÇÍÂèÒ§ «Ö § à» ¹ ¼Å·ÓãËé¡Ò÷ӧҹ¢Í§Ç§¨Ã PLL à¡Ô´ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡¨Ø´ ÊÁ´ØÅ ¨Ø´ Ë¹Ö § ä» Âѧ ¨Ø´ ÊÁ´ØÅ ÍÕ¡ ¨Ø´ Ë¹Ö § à˵ءÒóì àªè¹¹Õ ¨ÐàÃÕ¡¡Ñ¹ ÇèÒ ä«à¤ÔÅÊÅÔ» (cycle slip) «Ö § ¨Ð·ÓãËé à¡Ô´ ¢éͼԴ¾ÅÒ´¨Ó¹Ç¹ÁÒ¡·Õ ǧ¨ÃµÃǨËÒ (detector) ÃÙ» ·Õ 2.9 áÊ´§µÑÇÍÂèÒ§ÅѡɳТͧä«à¤ÔÅÊÅÔ» ¨Ò¡ÃÙ»¨Ð¾ºÇèÒ Ç§¨Ã PLL ÊÒÁÒöµÔ´µÒÁ¡ÒÃà»ÅÕ Â¹á»Å§¢Í§

τ

ä´é´Õ ³ ¨Ø´àÃÔ Áµé¹¢Í§¡ÅØèÁ¢éÍÁÙÅ

(data packet) áµè àÁ× Í ÁÕ ä«à¤ÔÅÊÅÔ» à¡Ô´ ¢Ö ¹ ǧ¨Ã PLL ¨Ð¤èÍÂæ ÊÙ­àÊÕ ¡ÒõԴµÒÁ¤èÒ

τ

¨¹¡ÃÐ·Ñ §

ǧ¨Ã PLL à¢éÒ ÊÙè ¨Ø´ ÊÁ´ØÅ ãËÁè ÍÕ¡ ¨Ø´ Ë¹Ö § ´Ñ§¹Ñ ¹ ÍÒ¨¨Ð¡ÅèÒÇä´é ÇèÒ ä«à¤ÔÅÊÅÔ» à» ¹ ÊÒà赯 ·ÓãËé ǧ¨Ã


34

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

PLL ·Ó§Ò¹·Õ ¨Ø´ÊÁ´ØÅÍÕ¡¨Ø´Ë¹Ö § «Ö §à» ¹à˵ؼÅÇèÒ·ÓäÁ

τ̂

¨Ö§ÁÕ¤èÒµèÒ§¨Ò¡

τ

»ÃÐÁÒ³

T

àÁ× ÍÊÔ ¹ÊØ´

¢Í§¡ÅØèÁ ¢éÍÁÙÅ ¨ÐàËç¹ä´éÇèÒ ä«à¤ÔÅÊÅԻ໠¹ ÊÔ § ·Õ ÍѹµÃÒÂÁÒ¡ÊÓËÃѺ Ãкºä·ÁÁÔ § ÃԤѿàÇÍÃÕ ´Ñ§¹Ñ ¹ ¹Ñ¡ÇԨѠ¨Ö§ ä´é àʹÍÇÔ¸Õ ¡ÒõèÒ§æ ·Õ ¨Ð¹Óãªé 㹡ÒèѴ¡ÒáѺ ä«à¤ÔÅÊÅÔ» ÊÓËÃѺ ¼Ùé ʹã¨ÊÒÁÒöÈÖ¡ÉÒ ÃÒÂÅÐàÍÕ´ä´éã¹ [29, 30, 31, 32, 33]

2.4

»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ PR4 ·Õ áÊ´§ã¹ÃÙ» ·Õ 2.2 ¨ÐàËç¹ä´éÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áÅÐǧ¨Ã µÃǨËÒÊÑ­ÅÑ¡É³ì ¨Ð·Ó§Ò¹á¡¨Ò¡¡Ñ¹ ´Ñ§¹Ñ ¹ »ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкº·Õ äÁè ä´é ¶Ù¡ à¢éÒ ÃËÑÊ (un coded system) ¨Ð¢Ö ¹ÍÂÙè¡Ñº¤Ø³ÀÒ¾¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕ ã¹Êèǹ¹Õ ¨ÐáÊ´§»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» àÁ× Í·Ó§Ò¹ã¹Ãкº·Õ ÁÕáÅÐ äÁè ÁÕ ÍͿ૵·Ò§¤ÇÒÁ¶Õ ¹Í¡¨Ò¡¹Õ ǧ¨ÃµÃǨËÒÊÑ­ÅÑ¡É³ì ·Õ ãªé ã¹Ç§¨Ã PLL ¤×Í Ç§¨ÃµÃǨËÒ ¢Õ´ àÃÔ Á à»ÅÕ Â¹·Õ äÁè ÁÕ Ë¹èǤÇÒÁ¨ÓẺËÅÒÂÃдѺ (multi level memoryless threshold detector) â´Â¢éÍÁÙÅ àÍÒµì¾Øµ ·Õ ä´é ¨Ðà» ¹ 仵ÒÁÊÁ¡Òà (2.18) ÊÓËÃѺ ¡Ãкǹ¡ÒõÃǨËÒ¢éÍÁÙÅ ÅӴѺ ¢éÍÁÙÅ

{yk }

¨Ð¶Ù¡Êè§ä»Âѧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ·Õ ÁÕ»ÃÔÁҳ˹èǧàÇÅÒÊÓËÃѺ¡ÒõѴÊÔ¹ã¨à·èҡѺ

60T

à¾× Í

ËÒÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ ·Õ à» ¹ ä»ä´é ÁÒ¡·Õ ÊØ´ ¹Í¡¨Ò¡¹Õ BER áµèÅФèÒ ·Õ ä´é ¨Ð¶Ù¡ ¤Ó¹Ç³â´Âãªé ¡ÅØèÁ ¢éÍÁÙŨӹǹÁÒ¡¨¹¡ÇèÒ¨Ðà¡Ô´¢éͼԴ¾ÅÒ´ÃÇÁ·Ñ §ËÁ´ 1000 ºÔµ ÊÓËÃѺ Ãкº·Õ äÁè ÁÕ ÍͿ૵·Ò§¤ÇÒÁ¶Õ ǧ¨ÃÀÒ¤ÃѺ ÊÒÁÒö·Õ ¨Ðãªé ǧ¨Ã PLL Íѹ´Ñº ·Õ Ë¹Ö § ä´é ã¹¡Ã³Õ¹Õ ¨ÐÊÁÁصÔãËé ÃкºÁÕ¡ÒÃà¢éҨѧËÇÐ㹪èǧÀÒÇСÒÃä´éÁÒẺÊÁºÙóì (perfect acquisition) «Ö § ·Óä´é â´Â¡ÒáÓ˹´ãËé

τ0 = 0

à¾× Í ·Õ ÇèÒ ¨Ðä´é äÁè µéͧãªé ¢éÍÁÙÅ preamble áÅСÓ˹´ãËé ¢éÍÁÙÅ

Ë¹Ö § ¡ÅØèÁ ÁÕ ¨Ó¹Ç¹ 4096 ºÔµ ÃÙ» ·Õ 2.10 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§¢éͼԴ¾ÅÒ´·Ò§àÇÅÒẺÃÒ¡ ¡ÓÅѧ Êͧà©ÅÕ Â (RMS: root mean square) ¹Ñ ¹¤×Í

σ²

p

=

E[(τk − τ̂k )2 ]

áÅÐ BER â´Â¤èÒ

¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL (α) ·Õ ãªé¨Ð¶Ù¡Í͡ẺÁÒà¾× ÍãËéÊÒÁÒöµÔ´µÒÁ¡ÒÃà»ÅÕ Â¹á»Å§·Ò§à¿Ê ä´é ÀÒÂã¹ 100 ºÔµ (¹Ñ ¹¤×Í

α100

= 0.0295) µÒÁ·Õ ä´é ͸ԺÒÂã¹ËÑÇ¢éÍ ·Õ 2.3.1 ¨ÐàËç¹ä´éÇèÒ àÁ× Í

ÃдѺ ¤ÇÒÁÃعáç¢Í§¨Ôµ àµÍÃì ·Ò§àÇÅÒ ¢Í§

σ² /T

σ² /T

σw /T

áÅÐ BER) ¹Í¡¨Ò¡¹Õ ¶éÒÃкºÁÕ¤èÒ

ÂÔ § ÁÒ¡ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº¡ç ¨ÐÂÔ § áÂè (·Ñ § ã¹ÃÙ»

σ² /T

¹éÍ Ãкº¡ç¨ÐÁÕ BER µ Ó ´Ñ§¹Ñ ¹ ¾ÒÃÒÁÔàµÍÃì

áÅÐ BER ¨Ö§ÊÒÁÒö·Õ ¨Ð¹ÓÁÒãªéà» ¹à¡³±ì㹡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºä·ÁÁÔ §


2.4.

»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺·Õ ãªé¡Ñ¹·Ñ Çä»

35

18

RMS timing error σε/T (%)

16 14 12 10

σw/T = 1%

8 6 4

σ /T = 0.1% w

2 0

4

5

6

7

8

9

10

9

10

(a) Eb/N0 (dB) −1

10

−2

BER

10

σw/T = 1% σw/T = 0.1%

−3

10

−4

10

−5

10

4

5

6

7

8

(b) Eb/N0 (dB)

ÃÙ»·Õ 2.10:

(a) ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒẺ RMS

ªèͧÊÑ­­Ò³ÍØ´Á¤µÔẺ PR4 ·Õ ÁÕ¤èÒ

σw /T

σ² /T

áÅÐ (b) »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ BER ÊÓËÃѺ

µèÒ§æ ¡Ñ¹ (àÁ× ÍÃкºäÁèÁÕÍͿ૵·Ò§¤ÇÒÁ¶Õ )

ÃԤѿàÇÍÃÕẺµèÒ§æ ä´é 㹷ӹͧà´ÕÂǡѹ ÊÓËÃѺÃкº·Õ ÁÕÍͿ૵·Ò§¤ÇÒÁ¶Õ ǧ¨ÃÀÒ¤ÃѺ¨Ðµéͧãªéǧ¨Ã PLL Íѹ´Ñº·Õ


36

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0

10

Conventional timing recovery with hard decision −1

10

C = 50 C = 100

−2

10

BER

C = 256 Perfect timing

−3

10

−4

10

−5

10

5

6

7

8

9

10

Eb/N0 (dB)

ÃÙ»·Õ 2.11: »ÃÐÊÔ·¸ÔÀÒ¾ BER ¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Ç仢ͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔẺ PR4 ÊÓËÃѺ

σw /T = 0.5%

áÅÐÍͿ૵·Ò§¤ÇÒÁ¶Õ 0.2%

Êͧ㹡ÒèѴ¡ÒáѺÍͿ૵·Ò§¤ÇÒÁ¶Õ ¹Õ â´Â¨Ð¾Ô¨ÒóÒÃкº·Õ ·Ó§Ò¹ã¹ÊÀÒÇлҹ¡ÅÒ§ (moder ate condition) ¹Ñ ¹¤×ÍÁÕ áÅÐ

C

β

σw /T = 0.5%

áÅÐÍͿ૵·Ò§¤ÇÒÁ¶Õ 0.2% àªè¹à´ÕÂǡѹ ¾ÒÃÒÁÔàµÍÃì

·Õ ãªé¨Ð¶Ù¡Í͡ẺÁÒà¾× ÍãËéÊÒÁÒöµÔ´µÒÁ¡ÒÃà»ÅÕ Â¹á»Å§·Ò§à¿ÊáÅзҧ¤ÇÒÁ¶Õ ä´éÀÒÂã¹

ºÔµ µÒÁ·Õ ä´é ͸ԺÒÂã¹ËÑÇ¢éÍ ·Õ 2.3.2 «Ö § ¨Ðä´é ÇèÒ ¤èÒ

α

·Õ ¶Ù¡ Í͡ẺÊÓËÃѺ

= 50, 100, áÅÐ 256 ¤×Í 0.012, 0.029, áÅÐ 0.058, µÒÁÅӴѺ ã¹¢³Ð·Õ ¤èÒ ÊÓËÃѺ

α

d=0

áÅÐ

C

d = 0 β

áÅÐ

C

·Õ ¶Ù¡ Í͡Ẻ

= 50, 100, áÅÐ 256 ¤×Í 0.00015, 0.000885, áÅÐ 0.00325, µÒÁÅӴѺ

¹Í¡¨Ò¡¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒÐ¡Ã³Õ·Õ Ç§¨Ã PLL ãªé¤èÒ

α

áÅФèÒ

β

à´ÕÂǡѹ ÊÓËÃѺ·Ñ §ªèǧÀÒÇСÒÃä´é

ÁÒáÅÐÀÒÇСÒõԴµÒÁ â´Â¢éÍÁÙÅË¹Ö §¡ÅØèÁ¨Ð»ÃСͺ仴éÇ preamble ¨Ó¹Ç¹

C

ºÔµ áÅÐÊèǹ·Õ

à» ¹ºÔµ¢èÒÇÊÒèӹǹ 4096 ºÔµ ÃÙ»·Õ 2.11 áÊ´§»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» àÁ× Íãªé¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL ·Õ Í͡ẺÊÓËÃѺáµèÅÐ ãªé á·¹ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä» ·Õ ãªé

τ̂k

=

τk

C

àÊ鹡ÃÒ¿·Õ à¢Õ¹ÇèÒ Perfect timing

ÊÓËÃѺ ¡Òêѡ µÑÇÍÂèÒ§ÊÑ­­Ò³

y(t)

¨Ð


2.5.

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺´Ô¨Ô·ÑÅ

37

àËç¹ä´éÇèÒ ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä»äÁèÊÒÁÒö·Ó§Ò¹ä´é´Õ àÁ× Í·Ó§Ò¹ã¹Ãкº·Õ µéͧ¡ÒÃÍѵÃÒ ¡ÒÃÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ ËÃ×Í ÍÕ¡ ¹ÑÂ Ë¹Ö § ¡ç ¤×Í àÁ× Í ãªé §Ò¹¡Ñº ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL ·Õ ¶Ù¡ Í͡Ẻ ÊÓËÃѺ

C

2.5

ä·ÁÁÔ §ÃԤѿàÇÍÃÕà຺´Ô¨Ô·ÑÅ

¹éÍÂæ

ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä»·Õ Í¸ÔºÒÂ㹺·¹Õ ¨ÐÁÕ ÅѡɳСÒ÷ӧҹ໠¹ Ẻ¼ÊÁ (hybrid) ¹Ñ ¹¤×Í ÁÕ ·Ñ § Êèǹ·Õ ·Ó§Ò¹¡Ñº ÊÑ­­Ò³á͹ÐÅçÍ¡ áÅÐÊèǹ·Õ ·Ó§Ò¹¡Ñº ÊÑ­­Ò³´Ô¨Ô·ÑÅ ¾Ô¨ÒóҨҡ ÃÙ» ·Õ 2.2 â´Â·Ñ Çä» Ç§¨Ã VCO ÁÑ¡¨Ðà» ¹ ǧ¨Ãá͹ÐÅçÍ¡«Ö § ÁÕ ÅѡɳСÒ÷ӧҹ·Õ «Ñº«é͹áÅÐÁÕ ÃÒ¤Òᾧ ã¹Êèǹ¹Õ ¨Ð¡ÅèÒǶ֧Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕ·Õ ÁÕÅѡɳÐà» ¹áºº´Ô¨Ô·ÑÅ·Ñ §ËÁ´ «Ö §ÁÕãªé㹪Ի ªèͧÊÑ­­Ò³ÍèÒ¹ (read channel chip) ºÒ§ÃØè¹ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº´Ô¨Ô·ÑÅ ¨Ðãªé ÍѵÃÒ¡Òêѡ µÑÇÍÂèÒ§ (sampling rate) ·Õ äÁè à¢éÒ ¨Ñ§ËÇÐ (asyn chronous) ¡ÑºÊÑ­­Ò³á͹ÐÅçÍ¡·Õ ä´éÃѺ à¾Õ§áµè¢ÍãËéÁÕ¤ÇÒÁ¶Õ ¡ÒêѡµÑÇÍÂèÒ§ (sampling frequen 5

cy) ÊÙ§¡ÇèÒ¤ÇÒÁ¶Õ 乤ÇÔµÊì

(Nyquist frequency) [2, 10] ¢Í§ÊÑ­­Ò³á͹ÐÅçÍ¡ ËÃ×ÍÍÒ¨¨Ð¡ÅèÒÇ 6

ä´éÇèÒ Ç§¨ÃªÑ¡µÑÇÍÂèÒ§ãªéÍѵÃÒ¡ÒêѡµÑÇÍÂèҧẺà¡Ô¹¨ÃÔ§

(oversampling rate) [34, 35] à¹× ͧ¨Ò¡

¢éÍÁÙÅá«Áà» Å·Õ ä´é¨Ò¡Ç§¨ÃªÑ¡µÑÇÍÂèÒ§¹Õ ¨ÐäÁèà¢éҨѧËÇСѺ¢éÍÁÙźԵ·Õ Êè§ÁÒ¨Ò¡µé¹·Ò§ ´Ñ§¹Ñ ¹ ¨Ö§µéͧ ÁÕ ¡ÒûÃѺ¤èÒ·Ò§àÇÅÒ (timing adjustment) ´éÇÂÇÔ¸Õ¡Ò÷ҧ´Ô¨Ô·ÑÅ·Õ àÃÕ¡ÇèÒ à·¤¹Ô¤¡ÒûÃÐÁÒ³ ¤èÒ ã¹ªèǧ (interpolation technique) à¾× ÍãËé ä´é ¢éÍÁÙÅ á«Áà» Å ·Õ ÊÍ´¤Åéͧ¡Ñº ¢éÍÁÙÅ ºÔµ ·Õ Êè§ ÁÒ¨Ò¡ µé¹·Ò§ Ãкºä·ÁÁÔ § ÃԤѿàÇÍÃÕ ·Õ ãªé à·¤¹Ô¤ ¹Õ ¨ÐàÃÕ¡¡Ñ¹ ·Ñ Çä»ÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº»ÃÐÁÒ³¤èÒ ã¹ªèǧ (interpolated timing recovery) «Ö § ÁÕ â¤Ã§ÊÃéÒ§µÒÁÃÙ» ·Õ 2.12 â´ÂÊèǹ»ÃСͺ·Ø¡ Êèǹ ã¹Ç§¨Ã PLL ¨ÐÁÕ ÅѡɳСÒ÷ӧҹ໠¹ Ẻ´Ô¨Ô·ÑÅ ·Ñ §ËÁ´ «Ö § ¨ÐªèÇ·ÓãËé ÊÒÁÒöŴ¤èÒãªé¨èÒÂã¹ ¡ÒÃÊÃéÒ§Ãкºä·ÁÁÔ § ÃԤѿàÇÍÃÕ ä´é ¨Ò¡¡Ò÷´Åͧ¾ºÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº»ÃÐÁÒ³¤èÒ ã¹ªèǧÁÕ »ÃÐÊÔ·¸ÔÀÒ¾à·Õºà·èÒ ¡Ñº ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä» ¶éÒ ÁÕ ¡ÒÃàÅ×Í¡ãªé ǧ¨Ã¡Ãͧ¡ÒûÃÐÁÒ³ ¤èÒ ã¹ ªèǧ (interpolation lter) ·Õ àËÁÒÐÊÁ ÊÓËÃѺ ¼Ùé ʹ㨠ÊÒÁÒö ÈÖ¡ÉÒ ÃÒÂÅÐàÍÕ´ ¢Í§ ä·ÁÁÔ § ÃԤѿàÇÍÃÕẺ»ÃÐÁÒ³¤èÒ㹪èǧà¾Ô ÁàµÔÁä´é¨Ò¡ [34, 35, 36, 37] 5

¤ÇÒÁ¶Õ 乤ÇÔµÊì¢Í§ÊÑ­­Ò³á͹ÐÅçÍ¡ ÁÕ¤èÒà·èҡѺÊͧà·èҢͧ¤ÇÒÁ¶Õ ÊÙ§ÊØ´¢Í§ÊÑ­­Ò³á͹ÐÅçÍ¡¹Ñ ¹

6

¨Ó¹Ç¹¢éÍÁÙÅá«Áà» Å·Õ ä´é¨Ò¡Ç§¨ÃªÑ¡µÑÇÍÂèÒ§ ¨ÐÁըӹǹÁÒ¡¡ÇèÒ¢éÍÁÙÅÍÔ¹¾ØµºÔµ·Õ Êè§ÁҨҡǧ¨ÃÀÒ¤Êè§


38

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

A/D converter

y (t )

interpolation filter

t i = iTs

yk

Viterbi detector

âk

t k = ( mk + µ k )Ts

fixed sampling frequency (1 / Ts )

τˆk digital accumulator

symbol detector

z −d

interpolator control unit

loop filter

εˆk

yk − d

rˆk −d

TED

ÃÙ»·Õ 2.12: â¤Ã§ÊÃéÒ§¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ»ÃÐÁÒ³¤èÒ㹪èǧ

2.6

àà¹Çâ¹éÁ¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕã¹Í¹Ò¤µ

¨Ò¡¼ÅÅѾ¸ì ·Õ áÊ´§ã¹ÃÙ» ·Õ 2.10 áÅÐ 2.11 ÊÃØ» ä´é ÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Ç仨зӧҹ ä´é äÁè ´Õ ¶éÒ ÃкºÁÕ ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒÁÒ¡ ËÃ×Í àÁ× Í ·Ó§Ò¹ã¹Ãкº·Õ µéͧ¡ÒÃÍѵÃÒ¡ÒÃÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ ÇÔ¸Õ¡ÒÃá¡é䢻 ­ËÒ·Õ §èÒÂ·Õ ÊØ´ 㹡ÒÃà¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ § ÃÔ ¤Ñ¿ àÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä» ¡ç ¤×Í ¡ÒÃà»ÅÕ Â¹Ç§¨ÃµÃǨËÒÊÑ­ÅÑ¡É³ì ·Õ ãªé ã¹ä·ÁÁÔ § ÅÙ» ¨Ò¡ ǧ¨ÃµÃǨËÒ¢Õ´ àÃÔ Á à»ÅÕ Â¹áººá¢ç§ (hard threshold detector) ä»à» ¹Ç§¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹áººÍè͹ (soft threshold detector) [33] ËÃ×Í ÍÒ¨¨Ðãªé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Õ ÁÕ »ÃÔÁҳ˹èǧàÇÅÒÊÓËÃѺ ¡ÒõѴÊÔ¹ã¨

dT

ÊÑ ¹ ¡çä´é [27] ÍÂèÒ§äÃ

¡çµÒÁ ÇÔ¸Õ¡ÒÃ·Õ ¡ÅèÒÇÁÒàËÅèÒ¹Õ ¨ÐªèÇÂà¾Ô Á»ÃÐÊÔ·¸ÔÀÒ¾·Õ ä´éà¾Õ§àÅ硹éÍÂà·èÒ¹Ñ ¹ ´Ñ§¹Ñ ¹ Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕẺãËÁè·Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» ¨Ö§à» ¹ÊÔ §·Õ µéͧ¡ÒÃÍÂèÒ§ÁÒ¡ ã¹ [30, 38] ä´é¹ÓàʹÍä·ÁÁÔ §ÃԤѿàÇÍÃÕÃٻẺãËÁè·Õ àÃÕ¡ÇèÒ à¾Íà«ÍÃì äÇàÇÍÃìä·ÁÁÔ §ÃԤѿàÇÍÃÕ (per survivor timing recovery) «Ö §ãªé§Ò¹¡ÑºÃкº·Õ äÁèä´é¶Ù¡à¢éÒÃËÑÊ «Ö §ÁÕ

y

¨ÐºÍ¡

10−4

ËÃ×ÍÍÕ¡

»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» ´Ñ§áÊ´§ã¹ÃÙ»·Õ 2.13 â´Â·Õ àÊé¹á¡¹ ¶Ö§»ÃÔÁÒ³

Eb /N0

(ÁÕ˹èÇÂà» ¹ dB) ·Õ Ãкºµéͧ¡Òà 㹡ÒÃ·Õ ¨Ð·ÓãËéÃкºÁÕ BER =

¹ÑÂË¹Ö §¡ç¤×Í àÊé¹á¡¹

y

¨ÐºÍ¡¶Ö§¡ÓÅѧ (power) ·Õ ǧ¨ÃÀÒ¤Ê觵éͧãªé㹡ÒÃÊ觢éÍÁÙÅ à¾× Í·Õ ¨Ð·ÓãËé


àà¹Çâ¹éÁ¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕã¹Í¹Ò¤µ

39

11 Conventional timing recovery with hard decision (d = 0) Conventional timing recovery with tentative decision (d = 4) Per−survivor timing recovery (d = 0) Genie−aided detector (d = 0)

10.8 10.6

Eb/N0 required to achieve BER = 10

−4

(in dB)

2.6.

10.4 10.2 10 9.8 9.6 9.4 9.2 9 0.1

0.2

0.3

0.4

0.5

0.6

σw/T (%)

0.7

0.8

0.9

1

ÃÙ»·Õ 2.13: »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺµèÒ§æ ¢Í§ªèͧÊÑ­­Ò³ÍØ´Á¤µÔẺ PR4

ǧ¨ÃÀÒ¤ÃѺÁÕ BER =

10−4

¨Ò¡ÃÙ»·Õ 2.13 àÊ鹡ÃÒ¿·Õ à¢Õ¹ÇèÒ Genie aided detector ËÁÒ¶֧

ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä»·Õ Ç§¨Ã PLL ãªé

r̂k

=

rk

(´ÙÃÙ»·Õ 2.2) 㹡ÒûÃѺ¤èÒÍͿ૵·Ò§

à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§µÑǶѴä», áÅÐ tentative decision (d = 4) ËÁÒ¶֧ ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ ·Õ ãªé ¡Ñ¹ ·Ñ Çä»·Õ Ç§¨ÃµÃǨËÒÊÑ­ÅÑ¡É³ì ·Õ ãªéè ã¹ä·ÁÁÔ § ÅÙ» ¤×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Õ ÁÕ »ÃÔÁҳ˹èǧ àÇÅÒ

d = 4T

¨Ð¼Å¡Ò÷´Åͧ¾ºÇèÒ à¾Íà«ÍÃì äÇàÇÍÃì ä·ÁÁÔ § ÃԤѿàÇÍÃÕ ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾´Õ ¡ÇèÒ ä·ÁÁÔ §

ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» (à¹× ͧ¨Ò¡ ãªé

Eb /N0

¹éÍ¡ÇèÒ ã¹¡ÒÃ·Õ ¨Ð·ÓãËéÃкºÁÕ BER =

à·èҡѹ) â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í·Ó§Ò¹·Õ ÃдѺ¤ÇÒÁÃعáç¢Í§ä·ÁÁÔ §¨ÔµàµÍÃì

σw /T

10−4

ÊÙ§

㹡ÒõÃǨÊͺÍѵÃÒ¡ÒÃÅÙèà¢éÒ ¢Í§ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áººµèÒ§æ ¨Ðãªé Ẻ¨ÓÅͧã¹ÃÙ» ·Õ 2.2 â´Â ¡Ó˹´ãËé

σw /T = 0%, τ̂0 = 0.5T ,

PLL ¨Ð¶Ù¡Í͡ẺÁÒÊÓËÃѺ

C

ÍͿ૵·Ò§¤ÇÒÁ¶Õ à·èҡѺ 0%, áÅФèÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã

= 50 ¹Ñ ¹¤×Í

α50

ÊÓËÃѺ

d

= 0 áÅÐ

4T

¤×Í 0.058 áÅÐ 0.049

µÒÁÅӴѺ ÃÙ»·Õ 2.14 à»ÃÕºà·ÕºÍѵÃÒ¡ÒÃÅÙèà¢éҢͧä·ÁÁÔ §ÃԤѿàÇÍÃÕẺµèÒ§æ â´Â¤Ô´à©ÅÕ Â¨Ò¡¡ÅØèÁ ¢éÍÁÙÅ 50000 ¡ÅØèÁ â´Â·Õ Ãкº¨Ð¶Ù¡¾Ô¨ÒóÒÇèÒ»ÃÐʺ¤ÇÒÁÊÓàÃç¨ã¹¡ÒÃÅÙèà¢éÒ ³ àÇÅÒ·Õ

k

¡çµèÍàÁ× Í


40

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

100

Genie−aided detector

Percentage of convergence

90 80 70

PSP−MM

60

Conventional timing recovery with hard decision (d = 0)

50 40 30

Conventional timing recovery with tentative decision (d = 4)

20 10 50

100

150

200

250

300

350

400

450

500

Time (in bit periods)

ÃÙ»·Õ 2.14: ÍѵÃÒ¡ÒÃÅÙèà¢éҢͧä·ÁÁÔ §ÃԤѿàÇÍÃÕẺµèÒ§æ àÁ× Íãªé

τ̂i

ÊÓËÃѺ

i≥k

ÁÕ ¤èÒ à·èÒ ¡Ñº ¤èÒ 0 ËÃ×Í

T

α50

·Õ

Eb /N0 = 10

´éǤèÒ ¤ÇÒÁ¤ÅÒ´à¤Å× Í¹ÂÔ¹ÂÍÁ

±10%

ÇèÒ Ãкº·Õ ãªé Genie aided detector ¨ÐÅÙèà¢éÒÀÒÂã¹ 50 ºÔµ «Ö §¨ÐÊÍ´¤Åéͧ¡Ñº ǧ¨Ã PLL ¢Í§ Genie aided detector ãªé¤èÒ·Õ ¶Ù¡µéͧ (¹Ñ ¹¤×Í

r̂k

=

rk )

α50

dB

¨Ò¡ÃÙ» ¨Ð¾º ·Õ ãªé à¾ÃÒÐÇèÒ

㹡ÒûÃѺ¤èÒÍͿ૵

·Ò§à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§µÑÇ¶Ñ´ä» ¹Í¡¨Ò¡¹Õ Âѧ¾ºÇèÒ à¾Íà«ÍÃìäÇàÇÍÃìä·ÁÁÔ §ÃԤѿàÇÍÃÕÁÕÍѵÃÒ¡Òà ÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ àÁ× Íà·Õº¡Ñºä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» »ÃÐÊÔ·¸ÔÀÒ¾¢Í§à¾Íà«ÍÃì äÇàÇÍÃì ä·ÁÁÔ § ÃԤѿàÇÍÃÕ ÊÒÁÒö·Õ ¨Ð·ÓãËé à¾Ô Á ¢Ö ¹ ä´é ÍÕ¡ â´Â¡ÒùÓä» ãªé §Ò¹ÃèÇÁ¡Ñ¹ ¡Ñº ÃËÑÊ á¡é䢢éͼԴ¾ÅÒ´ (ECC: error correction code) «Ö § ¼ÅÅѾ¸ì ·Õ ä´é ¨ÐàÃÕ¡ÇèÒ à¾Íà«ÍÃì äÇàÇÍÃì ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Ó§Ò¹« Ó (per survivor iterative timing recovery) [30, 39, 40] â´Â¨Ðãªé §Ò¹¡Ñº Ãкº·Õ ¶Ù¡ à¢éÒ ÃËÑÊ (coded system) «Ö § ¨Ò¡¼Å¡Ò÷´Åͧ¾ºÇèÒ à¾Íà«ÍÃì äÇàÇÍÃì ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Ó§Ò¹« Ó ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Çä»ÁÒ¡ â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í ÃкºÁÕ ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒÁÒ¡ ËÃ×Í àÁ× Í ·Ó§Ò¹ã¹Ãкº·Õ µéͧ¡ÒÃÍѵÃÒ¡Òà ÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ ÊÓËÃѺ¼Ùéʹã¨ÊÒÁÒöÈÖ¡ÉÒÃÒÂÅÐàÍÕ´à¾Ô ÁàµÔÁä´éã¹ [30, 39, 40] «Ö §ÊÒÁÒö´Òǹì âËÅ´àÍ¡ÊÒÃàËÅèÒ¹Õ ä´é·Õ http://home.npru.ac.th/∼t3058


2.7.

ÊÃØ»·éÒº·

2.7

41

ÊÃØ»·éÒº·

㹺·¹Õ ä´é͸ԺÒ¶֧ËÅÑ¡¡Ò÷ӧҹ¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» ÃÇÁ件֧ÇÔ¸Õ¡ÒÃÍ͡Ẻ ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL â´Âãªé Ẻ¨ÓÅͧǧ¨Ãà¿ÊÅçÍ¡ÅÙ» ẺàªÔ§ àÊé¹ ¨Ò¡¼Å¡Ò÷´Åͧ¾º ÇèÒ ¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL ·Õ ä´é ¨Ò¡¡ÒÃÍ͡ẺµÒÁࡳ±ì ·Õ ¡Ó˹´äÇé ÁÕ ¤èÒ à´ÕÂǡѹ äÁèÇèÒ ã¹ Ãкº¨ÐÁÕ »ÃÔÁÒ³ÍͿ૵·Ò§¤ÇÒÁ¶Õ à·èÒã´ áÅСè͹·Õ ¨Ð¹Ó¤èÒ ¾ÒÃÒÁÔàµÍÃì ¢Í§Ç§¨Ã PLL ·Õ ä´é ¨Ò¡ ÇÔ¸Õ¡ÒÃÍ͡ẺµÒÁ·Õ ͸ԺÒÂ㹺·¹Õ ä»ãªé §Ò¹ ¼Ùéãªé ¨Ðµéͧ·ÓãËé ¤ÇÒÁªÑ¹ ¢Í§àÊé¹â¤é§ ÃÙ» µÑÇ àÍʢͧ ǧ¨Ã TED ÁÕ¤èÒà» ¹¤èÒË¹Ö § ³ ¨Ø´¡Óà¹Ô´¡è͹àÊÁÍ ¹Í¡¨Ò¡¹Õ ¼Å¡Ò÷´ÅͧáÊ´§ãËéàËç¹ÇèÒ ä·ÁÁÔ § ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä»·Ó§Ò¹ä´éäÁè´Õ àÁ× ÍÃкºÁÕ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒÁÒ¡ ËÃ×ÍàÁ× Í·Ó§Ò¹ã¹Ãкº ·Õ µéͧ¡ÒÃÍѵÃÒ¡ÒÃÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ ä·ÁÁÔ § ÃԤѿàÇÍÃÕ áºº·Õ ãªé ¡Ñ¹ ·Ñ Ç仡ç Âѧ à» ¹ ·Õ ¹ÔÂÁãªé §Ò¹¡Ñ¹ ÁÒ¡ã¹ à¡×ͺ¨Ð·Ø¡§Ò¹»ÃÐÂØ¡µì à¹× ͧ¨Ò¡à» ¹Ç§¨Ã·Õ §èÒµèÍ¡ÒÃÊÃéÒ§áÅÐÊÒÁÒö·Ó§Ò¹ä´é´Õà¾Õ§¾Í ¶éÒàÅ×Í¡ ãªé¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL ·Õ àËÁÒÐÊÁ ÃÇÁ·Ñ §ÃкºÁÕ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒäÁèÁÒ¡¹Ñ¡ áÅÐÃкº äÁèÁÕ¤ÇÒÁµéͧ¡ÒÃÍѵÃÒ¡ÒÃÅÙèà¢éÒ·Õ ÃÇ´àÃçÇ

2.8

à຺½ ¡ËÑ´·éÒº·

1. ¨§Í¸ÔºÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅÙ» (PLL: phased lock loop) ÁÒ¾ÍÊѧࢻ

2. ¨§Í¸ÔºÒ墄 ¹µÍ¹¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL Íѹ´Ñº·Õ Ë¹Ö §

3. ¨§Í¸ÔºÒ墄 ¹µÍ¹¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì¢Í§Ç§¨Ã PLL Íѹ´Ñº·Õ Êͧ

4. ¨§Í¸ÔºÒ¤سÊÁºÑµÔáÅлÃÐ⪹ì¢Í§àÊé¹â¤é§ÃÙ»µÑÇàÍÊ

5. ¨§¤Ó¹Ç³ËÒÊÁ¡ÒÃàÊé¹â¤é§ ÃÙ» µÑÇ àÍʢͧǧ¨Ã M&M TED áÅÐÇÒ´ÃÙ» ä·ÁÁÔ § ¿ §¡ìªÑ¹ ¢Í§ ªèͧÊÑ­­Ò³

H(D)

µèÍ仹Õ

5.1)

H(D)

=

1−D

5.2)

H(D)

=

1+D


42

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

5.3)

H(D)

=

1 + 2D + D2

5.4)

H(D)

=

1 + D − D2 − D3

5.5)

H(D)

=

1 + 3D + 3D2 + D3

6. ¨§à»ÃÕºà·ÕºËÅÑ¡¡Ò÷ӧҹ¢Í§ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä»áÅÐä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ »ÃÐÁÒ³¤èÒ㹪èǧ ÁÒ¾ÍÊѧࢻ


º··Õ 3

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµààÅÐÍÕ¤ÇÍäÅà«ÍÃì

㹺·¹Õ ¨Ð͸ԺÒ¶֧ ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ (target) áÅÐÍÕ¤ÇÍäÅà«ÍÃì (equalizer) ãËé àËÁÒÐÊÁ ¡Ñº ªèͧÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì â´Â·Õ ·ÒÃìà¡çµ·Õ ´Õ ¨ÐµéͧÁÕ ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ·Õ ã¡Åéà¤Õ§ ¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³ãËé ÁÒ¡·Õ ÊØ´ ã¹·Ò§»¯ÔºÑµÔ ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ áÅÐ ÍÕ¤ÇÍäÅà«ÍÃì ÊÒÁÒö·Óä´é ËÅÒÂÇÔ¸Õ áµè 㹺·¹Õ ¨Ð¡ÅèÒǶ֧ ੾ÒÐÇÔ¸Õ¡Òà ¢éͼԴ¾ÅÒ´¡ÓÅѧ Êͧà©ÅÕ Â ·Õ ¹éÍÂÊØ´ (MMSE: minimum mean squared error) à·èÒ¹Ñ ¹ [19] à¹× ͧ¨Ò¡ à» ¹ÇÔ¸Õ·Õ §èÒµèÍ¡ÒÃ¹Ó ä»ãªé §Ò¹¨ÃÔ§ ¾ÃéÍÁ·Ñ § à»ÃÕºà·Õº¼Å¡Ò÷´Åͧ·Õ ä´é ¨Ò¡¡ÒÃãªé ·ÒÃìà¡çµ ẺµèÒ§æ 㹪èͧÊÑ­­Ò³ ¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

3.1

º·¹Ó

ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ (Viterbi detector) [15] ¨Ð¶Ù¡ ¹ÓÁÒãªé §Ò¹ÃèÇÁ¡Ñ¹ ¡Ñº ÍÕ¤ÇÍäÅà«ÍÃì Ẻ PR

(partial response equalizer) «Ö § ÍÕ¤ÇÍäÅà«ÍÃì

¹Õ ¡ç¤×Í Ç§¨Ã¡ÃͧẺàªÔ§àÊé¹ (linear lter) ·Õ ·Ó˹éÒ·Õ ã¹¡ÒûÃѺÃÙ»ÃèÒ§¢Í§¼ÅµÍºÊ¹Í§ÃÇÁ¢Í§ 1

ªèͧÊÑ­­Ò³ãËéÍÂÙèã¹ÃÙ»¢Í§¼ÅµÍºÊ¹Í§·Õ µéͧ¡Òà ËÃ×Í·Õ àÃÕ¡¡Ñ¹ÇèÒ ·ÒÃìà¡çµ 1

(target) ¨Ò¡¹Ñ ¹

·ÒÃìà¡çµ ¤×Í Ç§¨Ã¡ÃͧẺàªÔ§ àÊé¹ ·Õ ÁÕ ¨Ó¹Ç¹á·ç» (tap) ¹éÍ àÁ× Í à·Õº¡Ñº ¨Ó¹Ç¹á·ç» ¢Í§ªèͧÊÑ­­Ò³ áµè ÁÕ ¼Å

µÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ã¡Åéà¤Õ§¡Ñº¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³

43


44

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¡ç ¨Ð¹ÓàÍÒ¢éÍÁÙÅ àÍÒµì¾Øµ ¢Í§ÍÕ¤ÇÍäÅà«ÍÃì ÁÒ·Ó¡ÒõÃǨËÒÅӴѺ (sequence detection) Ẻ¤ÇèÐà» ¹ ÁÒ¡ÊØ´ (ML: maximum likelihood) à¾× Í ¤Ó¹Ç³ËÒÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ ·Õ Êè§ ÁҨҡǧ¨ÃÀÒ¤Êè§ ¢Ñ ¹µÍ¹·Ñ § 2 ¹Õ ÃÇÁàÃÕ¡¡Ñ¹ ÇèÒ à·¤¹Ô¤ ¼ÅµÍºÊ¹Í§ºÒ§Êèǹ¤ÇèÐà» ¹ ÁÒ¡ÊØ´ (PRML: partial response maximum likelihood) «Ö § ¶×Í ÇèÒ à» ¹ ËÑÇã¨ÊӤѭ ¢Í§Ãкº¡Òà »ÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ìã¹» ¨¨ØºÑ¹ ·ÒÃìà¡çµáºº PR ·Õ à» ¹·Õ ÂÍÁÃѺ¡Ñ¹ã¹Ãкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ (longitudinal recording) ¨ÐÍÂÙèã¹ÃÙ»¢Í§¾ËعÒÁ (polynomial)

H(D) = (1 − D)(1 + D)n â´Â·Õ

n

¤×Í àÅ¢¨Ó¹Ç¹àµçÁºÇ¡ áÅÐ

D

(3.1)

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ˹èǧàÇÅÒ (delay operator) ¨Ò¡ÊÁ¡ÒÃ

(3.1) ¨Ð¾ºÇèÒ ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ¹Õ ¨ÐÁÕÊ໡µÃÑÁ¤èÒÈÙ¹Âì (spectral null) ³ ¤ÇÒÁ¶Õ ¤èÒÈÙ¹ÂìáÅФÇÒÁ¶Õ 乤ÇÔµÊì (Nyquist frequency) à¹× ͧ¨Ò¡ ÁÕ¾¨¹ì·Õ à» ¹

(1−D)

áÅÐ

(1+D)

µÒÁ

ÅӴѺ ã¹¢³Ð·Õ ÊÑ­­Ò³ read back ¢Í§Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § (perpendicular recording) ¨ÐÁÕͧ¤ì»ÃСͺ¢Í§ä¿¿ Ò¡ÃÐáʵç (d.c. component) ´Ñ§¹Ñ ¹ ¾¨¹ì

(1 − D)

¨Ö§äÁè¨Óà» ¹ÊÓËÃѺ

Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § à¾ÃÒÐ©Ð¹Ñ ¹ ·ÒÃìà¡çµ Ẻ PR ·Õ à» ¹ ·Õ ÂÍÁÃѺ ã¹Ãкº¡Òúѹ·Ö¡ Ẻ á¹ÇµÑ § ¨ÐÍÂÙèã¹ÃÙ»¢Í§¾ËعÒÁ

H(D) = (1 + D)n

(3.2)

µÒÃÒ§·Õ 3.1 áÊ´§·ÒÃìà¡çµ Ẻ PR ·Õ à» ¹ ·Õ ÂÍÁÃѺ ¡Ñ¹ ·Ñ Çä» ã¹Ãкº¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ áÅÐẺá¹ÇµÑ § áÅÐÃÙ»·Õ 3.1 áÊ´§¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµáººµèÒ§æ àÁ× Íà»ÃÕºà·Õº ¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³·Õ ND = 2 áÅÐ 2.5 ¨Ò¡ÃÙ» ·Õ 3.1 ¨Ð¾ºÇèÒ àÁ× Í ¤èÒ ND ¢Í§ªèͧÊÑ­­Ò³ÁÕ ¤èÒ à¾Ô Á ¢Ö ¹ ·ÒÃìà¡çµ·Õ ãªé ¡ç ¤ÇÃ·Õ ¨ÐÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡¢Ö ¹ (¤èÒ

n

ÁÒ¡¢Ö ¹)

à¾× Í·Õ ¨Ð·ÓãËé ÁÕ ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ·Õ ÊÍ´¤Åéͧ¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³ãËé ÁÒ¡·Õ ÊØ´ ÍÂèÒ§äáçµÒÁ ·ÒÃìà¡çµ·Õ ãªé äÁè ¤ÇÃ·Õ ¨ÐÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡à¡Ô¹ ¤ÇÒÁ¨Óà» ¹ à¾ÃÒШÐÊè§ ¼Å·Ó ãËéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÁÕ¤ÇÒÁ«Ñº«é͹ (complexity) ÁÒ¡¢Ö ¹ «Ö §¨Ð͸ԺÒµèÍä»ã¹º··Õ 4 ¨Ò¡ÊÁ¡Òà (3.1) áÅÐ (3.2) Êѧࡵ¨Ð¾ºÇèÒ·ÒÃìà¡çµáºº PR ¨ÐÁÕ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§áµèÅÐá·ç» à» ¹àÅ¢¨Ó¹Ç¹àµçÁ áµè¶éÒÃкº PRML ÂÍÁãªé·ÒÃìà¡çµ·Õ ÁÕ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§áµèÅÐá·ç»à» ¹àÅ¢¨Ó¹Ç¹ ¨ÃÔ§¨Ð¾ºÇèÒ·ÒÃìà¡çµáºº¹Õ ¨ÐÊÒÁÒöªèÇÂà¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкºä´é ÁÒ¡ â´Â੾ÒÐÍÂèÒ§ÂÔ §


3.1.

º·¹Ó

45

µÒÃÒ§·Õ 3.1: µÑÇÍÂèÒ§·ÒÃìà¡çµáºº PR ·Õ à» ¹·Õ ÂÍÁÃѺ¡Ñ¹ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

n=1

·ÒÃìà¡çµáºº PR Ãкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹

PR4

n=2

[1 0 − 1]

1 − D2 Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ §

PR1

EPR4

n=3

[1 1 − 1 − 1]

1 + D − D2 − D3

[1 1]

PR2

[1 2 1]

[1 2 0 − 2 − 1]

1 + 2D − 2D3 − D4 EPR2

1 + 2D + D2

1+D

EEPR4

[1 3 3 1]

1 + 3D + 3D2 + D3

·Õ ND ÊÙ§æ â´Â·ÒÃìà¡çµ ÅÑ¡É³Ð¹Õ ¨ÐàÃÕ¡¡Ñ¹ ·Ñ Çä»ÇèÒ ·ÒÃìà¡çµáºº GPR (generalized partial response) «Ö §ÊÒÁÒö·Õ ¨ÐËÒä´é¨Ò¡ËÅÒÂÇÔ¸Õ¡Òà àªè¹ [41, 42, 43, 44, 45, 46]

1) ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ â´ÂãËé ¼ÅµÍºÊ¹Í§·ÒÃìà¡çµ (target response) ÁÕ ÃÙ»ÃèÒ§àËÁ×͹¡Ñº ¼Å µÍºÊ¹Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð (transition response) ËÃ×ͼŵͺʹͧ䴺Ե (dibit response) ¢Í§ªèͧÊÑ­­Ò³ ·Ñ §ã¹â´àÁ¹àÇÅÒ (time domain) áÅÐâ´àÁ¹¤ÇÒÁ¶Õ (frequency domain) ÇÔ¸Õ¡ÒÃ¹Õ ¨Ð·Ó¡ÒÃËÒ·ÒÃìà¡çµ·Õ ÁÕÃÙ»ÃèÒ§¢Í§¼ÅµÍºÊ¹Í§·ÒÃìà¡çµàËÁ×͹¡Ñº¼ÅµÍºÊ¹Í§ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐËÃ×Í ¼ÅµÍºÊ¹Í§ä´ºÔµ¢Í§ªèͧÊÑ­­Ò³áµèÅÐ ND ·Ñ § ã¹â´àÁ¹àÇÅÒáÅÐ â´àÁ¹¤ÇÒÁ¶Õ µÑÇÍÂèÒ§àªè¹ã¹Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ·ÒÃìà¡çµ·Õ ÊÍ´¤Åéͧ¡Ñº¼ÅµÍºÊ¹Í§ ä´ºÔµ [43] ã¹â´àÁ¹àÇÅÒ ä´éá¡è [1

−4

1], [1

−2 −2

1], áÅÐ [1 0

−8

0 1] à» ¹µé¹

¹Í¡¨Ò¡¹Õ ·ÒÃìà¡çµ ºÒ§áººÍÒ¨¨ÐäÁè ÊÍ´¤Åéͧ¡Ñº ¼ÅµÍºÊ¹Í§¢Í§ªèͧÊÑ­­Ò³ã¹â´àÁ¹ àÇÅÒ áµè¨ÐÊÍ´¤Åéͧ¡Ñº¼ÅµÍºÊ¹Í§¢Í§ªèͧÊÑ­­Ò³ã¹â´àÁ¹¤ÇÒÁ¶Õ ÁÒ¡¡çä´é ã¹·Ò§»¯ÔºÑµÔ ¤ÇÒÁÊÍ´¤Åéͧ¡Ñº ¼ÅµÍºÊ¹Í§¢Í§ªèͧÊÑ­­Ò³ã¹â´àÁ¹¤ÇÒÁ¶Õ à» ¹ ÊÔ § ·Õ µéͧ¡ÒÃÁÒ¡¡ÇèÒ ¤ÇÒÁÊÍ´¤Åéͧã¹â´àÁ¹àÇÅÒ à¹× ͧ¨Ò¡ ¨ÐªèǺ͡ãËé ·ÃÒº¶Ö§ ¤Ø³ÊÁºÑµÔ à¡Õ ÂǡѺ ÍѵÃÒ¡Òà ¢ÂÒÂÊÑ­­Ò³Ãº¡Ç¹ (noise enhancement) ä´é

2) ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ·Õ ·ÓãËé ¡ÓÅѧ ÃÇÁ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃì ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ ÇÔ¸Õ¡ÒÃ¹Õ ¨ÐÊÁÁØµÔ ÇèÒ ¹Ñ¡Í͡ẺÃкº·ÃÒºÇèÒ ªèͧÊÑ­­Ò³¤×Í ÍÐäà à¾× Í·Õ ¨Ðä´é ¹ÓàÍÒ


46

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

1.2

Channel (ND = 2) Channel (ND = 2.5) PR4 [1 0 −1] EPR4 [1 1 −1 −1] EEPR4 [1 2 0 −2 −1]

Normalized magnitude

1.0

0.8

0.6

0.4

0.2

0.0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

(a) Normalized frequency (fT) 1.0

Channel (ND = 2) Channel (ND = 2.5) PR2 [1 2 1] EPR2 [1 3 3 1] EEPR2 [1 4 6 4 1]

Normalized magnitude

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

(b) Normalized frequency (fT)

ÃÙ»·Õ 3.1: ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµáººµèÒ§æ ÊÓËÃѺÃкººÑ¹·Ö¡ (a) Ẻá¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ §

ªèͧÊÑ­­Ò³¹Ñ ¹ ÁÒãªé 㹡ÒäӹdzËÒ¿ §¡ìªÑ¹ ¶èÒÂâ͹¢Í§ÍÕ¤ÇÍäÅà«ÍÃì ¨Ò¡¹Ñ ¹ ¡ç ¨Ð·Ó¡Òà ¤Ó¹Ç³ËÒ¡ÓÅѧ ÃÇÁ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃì áÅÐàÁ× Í ä´é ¡ÓÅѧ ÃÇÁ


3.2.

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

47

¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ÍÂÙè ã¹ÃÙ» ¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃì áÅéÇ ¡ç ¨Ð·Ó¡ÒÃËÒ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§·ÒÃìà¡çµ ·Õ ·ÓãËé ¡ÓÅѧ ÃÇÁ¢Í§ÊÑ­­Ò³Ãº¡Ç¹ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ â´Âãªé à·¤¹Ô¤ ¡ÒÃËÒ͹ؾѹ¸ì (di erentiation) ÊÓËÃѺÃÒÂÅÐàÍÕ´¢Í§ÇÔ¸Õ¡ÒÃ¹Õ ÊÒÁÒöÈÖ¡ÉÒà¾Ô ÁàµÔÁä´é¨Ò¡ [41]

3) ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ·Õ ·ÓãËé ¤èÒ SNR »ÃÐÊÔ·¸Ô¼Å (e ective SNR) ·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨Ã µÃǨËÒÇÕà·ÍÃìºÔÁÕ¤èÒÁÒ¡·Õ ÊØ´ ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ẺµèÒ§æ µÒÁ·Õ ¡ÅèÒÇÁÒ¢éÒ§µé¹ ¹Õ äÁè ä´é ÃѺ»ÃСѹ ÇèÒ »ÃÐÊÔ·¸Ô ÀÒ¾ÃÇÁ¢Í§Ãкºã¹ÃÙ» ¢Í§ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ (BER: bit error rate) ·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨ÐÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ ã¹ [19, 42] ä´é àʹÍÇÔ¸Õ¡ÒÃËÒ ·ÒÃìà¡çµ·Õ àËÁÒÐ·Õ ÊØ´ (optimal target) ·Õ ¨Ð·ÓãËé ¤èÒ SNR »ÃÐÊÔ·¸Ô¼Å·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÕ ¤èÒ ÁÒ¡·Õ ÊØ´ ËÃ×ÍÍÕ¡¹ÑÂË¹Ö §¡ç¤×Í ·ÒÃìà¡çµ·Õ ·ÓãËé BER ¢Í§Ãкº àÁ× ÍÇÑ´·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨Ã µÃǨËÒÇÕà·ÍÃìºÔÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ ÊÓËÃѺ¼Ùéʹã¨ÊÒÁÒöÈÖ¡ÉÒÃÒÂÅÐàÍÕ´à¾Ô ÁàµÔÁä´éã¹ [19, 42]

4) ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ·Õ ·ÓãËé¢éͼԴ¾ÅÒ´¡ÓÅѧÊͧà©ÅÕ Â (MSE: mean squared error) ÃÐËÇèÒ§ ÊÑ­­Ò³·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃìáÅÐÊÑ­­Ò³·Õ µéͧ¡Òà (¹Ñ ¹¤×Í ÊÑ­­Ò³µÒÁ·ÒÃìà¡çµ ·Õ µéͧ¡ÒÃ) ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ ¨Ò¡¡ÒÃÈÖ¡ÉÒ¾ºÇèÒ ÇÔ¸Õ¡ÒÃ¹Õ à» ¹ ÇÔ¸Õ¡ÒÃ·Õ §èÒÂáÅÐàËÁÒÐÊÓËÃѺ ¡ÒùÓÁÒãªé §Ò¹¨ÃÔ§ ã¹ ·Ò§»¯ÔºÑµÔ [19] ¹Í¡¨Ò¡¹Õ ·ÒÃìà¡çµ·Õ ä´é ¨ÐÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ã¡Åéà¤Õ§¡Ñº ·ÒÃìà¡çµ ·Õ àËÁÒÐ·Õ ÊØ´ ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ Ẻ¹Õ ¨ÐÃÙé¨Ñ¡ ¡Ñ¹ ã¹ª× Í ÇèÒ ÇÔ¸Õ¡Òà ¢éͼԴ¾ÅÒ´¡ÓÅѧ Êͧà©ÅÕ Â ·Õ ¹éÍ ÊØ´ (MMSE: minimum mean squared error) «Ö §¨Ð͸ԺÒÂÃÒÂÅÐàÍÕ´ÇÔ¸Õ¡ÒÃ¹Õ ´Ñ§µèÍ仹Õ

3.2

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE [19] ¨Ð·ÓãËéäé´é·ÒÃìà¡çµËÅÒÂÃٻẺµÒÁà§× ͹䢺ѧ¤Ñº (con straint) ·Õ ¡Ó˹´Å§ä»ã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒÃÍ͡Ẻ ãËé¾Ô¨ÒóÒẺ¨ÓÅͧÃкºã¹ÃÙ»·Õ 3.2 â´Â ·Õ ÍÕ¤ÇÍäÅà«ÍÃì¨Ð¾ÂÒÂÒÁÊÃéÒ§¢éÍÁÙÅàÍÒµì¾Øµ

yk

ãËéÁÕ¤èÒã¡Åéà¤Õ§¡Ñº¢éÍÁÙÅ·Õ µéͧ¡ÒÃ

rk

â´Â»ÃÒȨҡ¡ÒâÂÒÂÊÑ­­Ò³Ãº¡Ç¹ ¶éÒ¡Ó˹´ãËéÍÕ¤ÇÍäÅà«ÍÃìÁըӹǹá·ç»à·èҡѺ

ãËéÁÒ¡·Õ ÊØ´

N = 2K +1


48

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

n(t) equalizer

ak

1 − D bk g(t) {±1} 2

p(t)

∆t k

sk

s(t)

LPF

F(D)

t k = kT

y k Viterbi âk detector

target

rk

H(D)

wk

ÃÙ»·Õ 3.2: Ẻ¨ÓÅͧÊÓËÃѺ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

á·ç» áÅÐÊÁÁØµÔ ãËé á·ç» ÈÙ¹Âì¡ÅÒ§ÍÂÙè ·Õ àÇÅÒ ÍÂÙèã¹ÃÙ»ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìã¹â´àÁ¹

D

k =0

à¾ÃÒÐ©Ð¹Ñ ¹ ÍÕ¤ÇÍäÅà«ÍÃì ÊÒÁÒö·Õ ¨Ðà¢Õ¹ãËé

ä´é ¤×Í

K X

F (D) =

fk Dk

(3.3)

k=−K àÁ× Í

D

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ˹èǧàÇÅÒ

T

˹èÇ 㹷ӹͧà´ÕÂǡѹ ·ÒÃìà¡çµ·Õ Áըӹǹá·ç»à·èҡѺ

á·ç» ¡ç¨ÐÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§¿ §¡ìªÑ¹ã¹â´àÁ¹

H(D) =

L−1 X

D

L

ä´é ¤×Í

hk Dk

(3.4)

k=0 â´Â·Õ

fk

áÅÐ

hk

à» ¹¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ·Õ à» ¹àÅ¢¨Ó¹Ç¹¨ÃÔ§ã¹áµèÅÐá·ç»¢Í§ÍÕ¤ÇÍäÅà«ÍÃìáÅзÒÃìà¡çµ

µÒÁÅӴѺ ¨Ø´»ÃÐʧ¤ì 㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ´éÇÂÇÔ¸Õ¡Òà MMSE ¤×Í ¨Ð·Ó¡ÒäӹdzËÒ¤èÒ ÊÑÁ»ÃÐ ÊÔ·¸Ô ¢Í§

F (D)

áÅÐ

H(D) yk

àÍÒµì¾Øµ ¢Í§ÍÕ¤ÇÍäÅà«ÍÃì ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô

fk

and

hk

ä» ¾ÃéÍÁ¡Ñ¹ ã¹ àÇÅÒ à´ÕÂǡѹ â´Â ¡Òà ·Ó ãËé ¤èÒ MSE ÃÐËÇèÒ§ ¢éÍÁÙÅ

áÅТéÍÁÙÅ àÍÒµì¾Øµ ¢Í§·ÒÃìà¡çµ

rk

ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ ËÃ×Í ÍÕ¡ ¹ÑÂ Ë¹Ö § ¤×Í

¨Ð¶Ù¡àÅ×Í¡ à¾× Í·Õ ·ÓãËé¤èÒ

£ ¤ £ ¤ E wk2 = E {(sk ∗ fk ) − (ak ∗ hk )}2 ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ àÁ× Í

wk

=

yk − rk

¤×Í ¢éͼԴ¾ÅÒ´·Õ ä´é¨Ò¡¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ,

¤Í¹âÇÅ٪ѹ (convolution operator), áÅÐ tor)

E[·]

(3.5)

¤×Í µÑÇ´Óà¹Ô¹¡ÒÃ

¤×Í µÑÇ´Óà¹Ô¹¡ÒäèÒ¤Ò´ËÁÒ (expectation opera


3.2.

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

¶éÒ¡Ó˹´ãËéàÇ¡àµÍÃìá¹ÇµÑ § â´Â·Õ

hk

and

fk

H

49

[h0 , h1 , · · · , hL−1 ]T

=

¤×Í ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§

H(D)

áÅÐ

F (D)

and

¢éÍÁÙÅ

sk , A

R

i

¤×Í àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì¢¹Ò´

áÅÐá¹ÇµÑ §·Õ

j)

K = 10

[·]T

¤×Í à¤Ã× Í§ËÁÒÂ

(ÍÕ¤ÇÍäÅà«ÍÃìÁÕ·Ñ §ËÁ´ 21

¤×Í àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì (auto correlation matrix) ¢¹Ò´

¢éÒÁ (cross correlation matrix) ¢¹Ò´ ·Õ

[f−K , · · · , f0 , · · · , fK ]T

=

µÒÁÅӴѺ, áÅÐ

àÁ·ÃÔ¡«ìÊÅѺà»ÅÕ Â¹ (transpose matrix) 㹺·¹Õ ¨Ð¡Ó˹´ãËé á·ç») ¶éÒ¡Ó˹´ãËé

F

L×L

N ×L

¢Í§¢éÍÁÙÅ

ÃÐËÇèÒ§¢éÍÁÙÅ

ak ,

sk

áÅÐ

áÅÐ

ak

P

N ×N

¢Í§

¤×Í àÁ·ÃÔ¡«ìÊËÊÑÁ¾Ñ¹¸ì

â´Â·Õ ÊÁÒªÔ¡

(i, j)

(á¶Ç

¢Í§àÁ·ÃÔ¡«ì·Ñ §ÊÒÁ¹Õ ¤×Í

R(i, j) = E

"S−1 X

# sk+K−i sk+K−j ,

k=0

A(i, j) = E P(i, j) = E

"S−1 X k=0 "S−1 X

−K ≤ i, j ≤ K

(3.6)

# ak−i ak−j ,

0 ≤ i, j ≤ L − 1

(3.7)

# sk+K−i ak−j ,

−K ≤ i ≤ K, 0 ≤ j ≤ L − 1

(3.8)

k=0

S

àÁ× Í

¤×Í ¤ÇÒÁÂÒÇ (ËÃ×ͨӹǹºÔµ) ¢Í§ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{ak }

¨Ò¡µÑÇá»Ã·Õ ¡Ó˹´ãËé¢éÒ§µé¹¹Õ ÊÁ¡Òà (3.5) ÊÒÁÒö·Õ ¨Ðà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§àÁ·ÃÔ¡«ìä´é ´Ñ§¹Õ

E[w2 ] = FT RF + HT AH − 2FT PH 㹡Ò÷ÓãËé¤èÒ

E[w2 ]

ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´â´Âà·Õº¡Ñº

F

áÅÐ

H

(3.9)

¨ÐµéͧÁÕ¡ÒáÓ˹´à§× ͹䢺ѧ¤Ñºà¢éÒä»

ã¹ÃÐËÇèÒ§¡Ãкǹ¡Ò÷ÓãËé ÁÕ ¤èÒ ¹éÍÂÊØ´ (minimization process) à¾× Í·Õ ¨ÐËÅÕ¡àÅÕ Â§¡ÒÃä´é ¼ÅÅѾ¸ì à» ¹

F=0

áÅÐ

G

3.2.1

áÅÐ

H=0

ã¹Êèǹ¹Õ ¨Ð͸ԺÒ¶֧à§× ͹䢺ѧ¤Ñº·Õ ¹èÒʹ㨠·Õ ãªé㹡ÒäӹdzËÒ¤èÒ

F

´Ñ§µèÍ仹Õ

à§× ͹䢺ѧ¤Ñºà຺âÁ¹Ô¡ (h0

= 1)

à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡ (monic constraint) ¨Ð¡Ó˹´ãËé¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§á·ç»µÑÇáá¢Í§·ÒÃìà¡çµ ÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § ¹Ñ ¹¤×Í

h0 = 1

[19] ¶éÒ¡Ó˹´ãËéàÇ¡àµÍÃìá¹ÇµÑ §

I

ááÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § ÊèǹÊÁÒªÔ¡µÑÇÍ× ¹æ ÁÕ¤èÒà·èҡѺ¤èÒÈÙ¹Âì ¡ÅèÒǤ×Í à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡¹Õ ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§àÁ·ÃÔ¡«ìä´é ¤×Í

¢¹Ò´

I

=

L×1

·Õ ÁÕÊÁÒªÔ¡µÑÇ

[1, 0, · · · , 0]T

IT H = 1

´Ñ§¹Ñ ¹


50

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡Ãкǹ¡ÒÃ㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ´éÇÂÇÔ¸Õ¡ÒÃ¹Õ ¤×Í ¡Ò÷ÓãËé ¤èÒ MSE ã¹ÊÁ¡Òà (3.9) ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ â´Â¾ÂÒÂÒÁÃÑ¡ÉÒãËé¤èÒ

IT H

=

1

ÍÂÙèµÅÍ´àÇÅÒ ¹Ñ ¹¤×Í ¡Ãкǹ¡ÒÃ¹Õ ¨Ð·ÓãËé¤èÒ

E[w2 ] = FT RF + HT AH − 2FT PH − 2λ(IT H − 1) ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ â´Â·Õ

λ

(3.10)

¤×Í µÑǤٳ ÅÒ¡ÃÒ¹¨ì (Lagrange multiplier) à» ¹¤èÒÊà¡ÅÒÃì (scalar) ¡ÒÃ

·ÓãËé ÊÁ¡Òà (3.10) ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ ÊÒÁÒö·Óä´é â´Â¡ÒÃËÒ͹ؾѹ¸ì (di erentiation) ¢Í§ÊÁ¡Òà (3.10) à·Õº¡Ñº F, H, áÅÐ

λ

µÒÁÅӴѺ «Ö §¨Ðä´é¼ÅÅѾ¸ì´Ñ§¹Õ (ÃÒÂÅÐàÍÕ´ã¹ÀÒ¤¼¹Ç¡ ¤)

¡ ¢ ∂ E[w2 ] ¡ ∂F ¢ ∂ E[w2 ] ¡ ∂H ¢ ∂ E[w2 ] ∂λ

= 2RF − 2PH

(3.11)

= 2AH − 2PT F − 2λI

(3.12)

= −2IT H + 2

(3.13)

¨Ò¡¹Ñ ¹¡ç¡Ó˹´ãËé ¼ÅÅѾ¸ì ¢Í§Í¹Ø¾Ñ¹¸ì ·Ñ §ËÁ´·Õ ä´é ¨Ò¡ÊÁ¡Òà (3.11) (3.13) ÁÕ ¤èÒ à» ¹ ¤èÒ ÈÙ¹Âì ¹Ñ ¹¤×Í ¡Ó˹´ãËéÊÁ¡Òà (3.11) ÁÕ¤èÒà·èҡѺ¤èÒÈÙ¹Âì ¨Ðä´éÇèÒ

2RF − 2PH = 0 RF = PH F = R−1 PH

(3.14)

áÅСÓ˹´ãËéÊÁ¡Òà (3.12) ÁÕ¤èÒà·èҡѺ¤èÒÈÙ¹Âì ¨Ðä´éÇèÒ

2AH − 2PT F − 2λI = 0 AH − PT F = λI á·¹¤èÒ

F

(3.15)

¨Ò¡ÊÁ¡Òà (3.14) ŧã¹ÊÁ¡Òà (3.15) ¨Ðä´é

¡ ¢ AH − PT R−1 PH = λI ¡ ¢ A − PT R−1 P H = λI ¡ ¢−1 I H = λ A − PT R−1 P

(3.16)


3.2.

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

51

㹷ӹͧà´ÕÂǡѹ ¡Ó˹´ãËéÊÁ¡Òà (3.13) ÁÕ¤èÒà·èҡѺ¤èÒÈÙ¹Âì ¨Ðä´éÇèÒ

−2IT H + 2 = 0 IT H = 1 á·¹¤èÒ

H

(3.17)

¨Ò¡ÊÁ¡Òà (3.16) ŧã¹ÊÁ¡Òà (3.17) ¨Ðä´é

¡ ¢−1 IT λ A − PT R−1 P I = 1 λ =

IT (A

1 − P R−1 P)−1 I T

(3.18)

à¾ÃÒÐ©Ð¹Ñ ¹â´ÂÊÃØ»áÅéÇ ¢Ñ ¹µÍ¹ã¹¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃìÊÓËÃѺà§× ͹䢺ѧ¤Ñºà຺ âÁ¹Ô¡ (h0

= 1)

¤×Í

1) ¡Ó˹´¨Ó¹Ç¹á·ç»¢Í§·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì (¤èÒ

L

áÅÐ

N)

¹Ñ ¹¤×ÍÊÃéèÒ§àÇ¡àµÍÃì F áÅÐ

H

2) ËÒ¤èÒàÁ·ÃÔ¡«ì R, A, áÅÐ P ¨Ò¡ÊÁ¡Òà (3.6) (3.8)

3) ÊÃéÒ§àÇ¡àµÍÃì I =

[1, 0, · · · , 0]T

4) ËÒ¤èÒµÑǤٳÅÒ¡ÃÒ¹¨ì

λ

¢¹Ò´

¨Ò¡ÊÁ¡Òà (3.18) ¹Ñ ¹¤×Í

1 − P R−1 P)−1 I

(3.19)

H = λ(A − PT R−1 P)−1 I

(3.20)

λ=

5) ËÒ¤èÒ·ÒÃìà¡çµ

H

L×1

IT (A

T

¨Ò¡ÊÁ¡Òà (3.16) ¹Ñ ¹¤×Í

6) ËÒ¤èÒÍÕ¤ÇÍäÅà«ÍÃì

F

¨Ò¡ÊÁ¡Òà (3.14) ¹Ñ ¹¤×Í

F = R−1 PH

(3.21)


52

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¤èÒ

λ

·Õ ä´é¨Ò¡ÊÁ¡Òà (3.19) à» ¹¤èÒ MMSE ·Õ ä´éÀÒÂãµéà§× ͹䢺ѧ¤Ñº¹Õ «Ö §ÊÒÁÒö¾ÔÊÙ¨¹ìä´éâ´Â

¡ÒÃá·¹¤èÒ F and H ¨Ò¡ÊÁ¡Òà (3.20) (3.21) ŧã¹ÊÁ¡Òà (3.9) ¹Ñ ¹¤×Í

E[w2 ] = (R−1 PH)T R(R−1 PH) + HT AH − 2(R−1 PH)T PH = HT PT (R−1 )T PH + HT AH − 2HT PT (R−1 )T PH = HT (A − PT R−1 P)H © ª = λ2 [IT (A − PT R−1 P)−1 ]T (A − PT R−1 P)(A − PT R−1 P)−1 I © ª = λ2 IT (A − PT R−1 P)−1 I = λ

(3.22)

ÊÁ¡Òà (3.22) ËÒä´éâ´ÂÍÒÈÑÂËÅÑ¡¤ÇÒÁ¨ÃÔ§·Õ ÇèÒ àÁ·ÃÔ¡«ì

R

áÅÐ

A

à» ¹àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ìẺ 2

ÊÁÁҵà ¹Í¡¨Ò¡¹Õ àÁ·ÃÔ¡«ì R, A áÅÐ P Âѧ໠¹àÁ·ÃÔ¡«ìẺ Toeplitz =

R−1

3.2.2

áÅÐ

£ ¤T (A − PT R−1 P)−1

à§× ͹䢺ѧ¤Ñºà຺

à§× ͹䢺ѧ¤Ñºáºº ¹Ñ ¹¤×Í

h1 = 1

=

«Ö §¨Ð·ÓãËéä´éÇèÒ

(R−1 )T

(A − PT R−1 P)−1

h1 = 1

h1 = 1 ¨Ð¡Ó˹´ãËé¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§á·ç»µÑÇÊͧ¢Í§·ÒÃìà¡çµÁÕ¤èÒà·èҡѺ¤èÒË¹Ö §

[19] Êèǹ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§á·ç» µÑÇ Í× ¹æ ¨Ðà» ¹ ¤èÒ ÍÐäáçä´é â´Âà§× ͹䢺ѧ¤Ñº ¹Õ ÁÕ

Çѵ¶Ø»ÃÐʧ¤ìà¾× Íãªé㹡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§·ÒÃìà¡çµáººµèÒ§æ ¶éÒ¡Ó˹´ãËéàÇ¡àµÍÃìá¹Ç

J

µÑ §

J

=

¢¹Ò´

L×1

·Õ ÊÁÒªÔ¡µÑÇ·Õ ÊͧÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § ÊèǹÊÁÒªÔ¡Í× ¹æ ÁÕ¤èÒà» ¹¤èÒÈÙ¹Âì ¡ÅèÒǤ×Í

[0, 1, 0, · · · , 0]T

´Ñ§¹Ñ ¹ à§× ͹䢺ѧ¤Ñº Ẻ

h1 = 1

ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè ã¹ÃÙ» ¢Í§àÁ·ÃÔ¡«ì ä´é

JT H = 1

¤×Í

¡Ãкǹ¡ÒÃ㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµâ´Âà§× ͹䢺ѧ¤Ñº¹Õ ¨ÐàËÁ×͹¡Ñº¡ÒÃÍ͡Ẻ·ÒÃìà¡çµâ´Âãªé à§× ͹䢺ѧ¤Ñº ẺâÁ¹Ô¡ à¾Õ§áµè à»ÅÕ Â¹¾¨¹ì ÊØ´·éÒÂã¹ÊÁ¡Òà (3.10) ¨Ò¡

2λ(JT H − 1) à» ¹àÇ¡àµÍÃì 2

2λ(IT H − 1)

ä»à» ¹

´Ñ§¹Ñ ¹ ¼ÅÅѾ¸ì·Õ ä´é¡ç¨ÐàËÁ×͹¡ÑºÊÁ¡Òà (3.19) (3.21) à¾Õ§áµèà»ÅÕ Â¹àÇ¡àµÍÃì

J

àÁ·ÃÔ¡«ìẺ Toeplitz ¤×Í àÁ·ÃÔ¡«ì·Õ ÊÁÒªÔ¡ã¹á¹ÇàÊé¹·á§ÁØÁà» ¹¤èÒ¤§·Õ ¤èÒà´ÕÂǡѹ

I


3.2.

¡ÒÃÍÍ¡à຺·ÒÃìà¡çµ´éÇÂÇÔ¸Õ¡Òà MMSE

3.2.3

53

à§× ͹䢺ѧ¤Ñºà຺¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ (H

T

H = 1)

à§× ͹䢺ѧ¤Ñº Ẻ¾Åѧ§Ò¹Ë¹Ö § ˹èÇ (unit energy) ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè ã¹ÃÙ» ¢Í§àÁ·ÃÔ¡«ì ä´é ¤×Í

HT H

=

1

«Ö §à» ¹¡ÒáÓ˹´ãËé¾Åѧ§Ò¹¢Í§·ÒÃìà¡çµÁÕ¤èÒà·èҡѺ¤èÒË¹Ö § [19, 46]

¡Ãкǹ¡ÒÃ㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ´éÇÂÇÔ¸Õ¡ÒÃ¹Õ ¤×Í ¡Ò÷ÓãËé ¤èÒ MSE ã¹ÊÁ¡Òà (3.9) ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ â´Â·Õ ¾ÂÒÂÒÁÃÑ¡ÉÒãËé¤èÒ

HT H

=

1

µÅÍ´àÇÅÒ ¡ÅèÒǤ×Í ¡Ãкǹ¡ÒÃ¹Õ ¨Ð·ÓãËé¤èÒ

E[w2 ] = FT RF + HT AH − 2FT PH − 2λ(HT H − 1) ÁÕ¤èÒµ ÓÊØ´ «Ö §ÊÒÁÒö·Óä´éâ´Â¡ÒÃËÒ͹ؾѹ¸ì¢Í§ÊÁ¡Òà (3.23) à·Õº¡Ñº F, H, áÅÐ

(3.23)

λ

µÒÁÅӴѺ

áÅéÇ ¡Ó˹´ãËé ¼ÅÅѾ¸ì ¢Í§Í¹Ø¾Ñ¹¸ì ·Ñ §ËÁ´·Õ ä´é ÁÕ ¤èÒ à» ¹ ¤èÒ ÈÙ¹Âì â´ÂàÁ× Í ·ÓµÒÁ¢Ñ ¹µÍ¹¹Õ áÅéÇ ¨Ð¾º ÇèÒ ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì ÊÒÁÒöËÒä´é¨Ò¡¡ÒÃá¡éÊÁ¡ÒÃ

(A − PT R−1 P)H = λH

(3.24)

«Ö §¡ÒÃá¡éÊÁ¡ÒÃ¹Õ ÁÕÅѡɳФÅéÒ¡Ѻ ¡ÒÃá¡é» ­ËÒ¤èÒÅѡɳÐ੾ÒÐ (eigenvalue problem) [12, 25] â´Â·Õ ¤èÒ (3.24) ¤èÒ

λ

λ

ã¹ÊÁ¡Òà (3.24) ÊÒÁÒö¾ÔÊÙ¨¹ìä´éÇèÒ ¤×Í ¤èÒ MMSE ¹Ñ ¹àͧ ´Ñ§¹Ñ ¹ ¨Ò¡ÊÁ¡ÒÃ

¨ÃÔ§æ áÅéÇ ¡ç ¤×Í ¤èÒ ÅѡɳÐ੾ÒÐ·Õ ¹éÍÂ·Õ ÊØ´ (minimum eigenvalue) ¢Í§àÁ·ÃÔ¡«ì

(A − PT R−1 P), H

¤×Í àÇ¡àµÍÃì ÅѡɳÐ੾ÒÐẺ¹ÍÃì ÁÍÅäÅ«ì(normalized eigenvector) ·Õ

ÊÍ´¤Åéͧ¡Ñº¤èÒÅѡɳÐ੾ÒÐ·Õ ¹éÍÂ·Õ ÊØ´, áÅÐ

3.2.4

F

¨ÐÁÕ¤èÒàËÁ×͹¡ÑºÊÁ¡Òà (3.21)

à§× ͹䢺ѧ¤Ñºà຺·ÒÃìà¡çµà©¾ÒÐ

ÊÓËÃѺ à§× ͹䢺ѧ¤Ñº Ẻ·ÒÃìà¡çµ ੾ÒÐ¹Õ ( xed target) ·ÒÃìà¡çµ¨Ð¶Ù¡ ¡Ó˹´ÁÒãËé µÑ §áµè àÃÔ Ááá ÊÔ § ·Õ Ãкºµéͧ¡Òà ¤×Í ÍÕ¤ÇÍäÅà«ÍÃì

F

·Õ àËÁÒÐÊÁ¡Ñº ·ÒÃìà¡çµ ·Õ ¡Ó˹´ÁÒãËé «Ö § ÊÒÁÒöËÒä´é ¨Ò¡

ÊÁ¡Òà (3.21) ¹Ñ ¹¤×Í

F = R−1 PG

(3.25)

â´Â¨ÐÃѺ»ÃСѹä´éÇèÒ¤èÒ MMSE ·Õ ä´é¨ÐÁÕ¤èÒà·èҡѺ¤èÒã¹ÊÁ¡Òà (3.9) ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáºº¹Õ ÁÕ»ÃÐ⪹ìÁÒ¡ à¹× ͧ¨Ò¡ã¹¡ÒÃãªé§Ò¹¨ÃÔ§ ªÔ»»ÃÐÁÇżÅÊÑ­­Ò³ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¨Ð¶Ù¡ Í͡ẺÁÒãËé ãªé §Ò¹¡Ñº ·ÒÃìà¡çµ Ẻã´áººË¹Ö § à·èÒ¹Ñ ¹ ´Ñ§¹Ñ ¹ ¼Ùéãªé §Ò¹


54

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¨Ðµéͧ¾ÂÒÂÒÁ»ÃѺ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô áµèÅÐá·ç»¢Í§ÍÕ¤ÇÍäÅà«ÍÃìãËéÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ·Õ ¡Ó˹´ÁÒãËé à¾× ÍãËé ä´é ÊÑ­­Ò³·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃìé ÁÕ ÃÙ»ÃèÒ§àËÁ×͹¡Ñº ÊÑ­­Ò³·Õ µéͧ¡Òà «Ö § ¨ÐªèÇÂ·Ó ãËé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Ó§Ò¹ä´é §èÒ墅 ¹ ã¹·Ò§µÃ§¡Ñ¹¢éÒÁ ¶éÒ ·Ó¡ÒûÃѺ ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô áµèÅÐá·ç» ¢Í§ÍÕ¤ÇÍäÅà«ÍÃìẺÅͧ¼Ô´Åͧ¶Ù¡ (trial and error) ¡ç¨Ð·ÓãËéàÊÕÂàÇÅÒÁÒ¡ áÅÐÍÕ¤ÇÍäÅà«ÍÃì·Õ ä´é ¡çÁÑ¡¨ÐäÁèÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ·Õ ¡Ó˹´ÁÒãËé

ËÁÒÂà˵Ø

㹡ÒÃËÒ·ÒÃìà¡çµ H ¨Ò¡ÊÁ¡Òà (3.20) áÅÐÍÕ¤ÇÍäÅà«ÍÃì

F

¨Ò¡ÊÁ¡Òà (3.21) ¢Ñ ¹µÍ¹

áá·Õ ¤ÇÃ·Ó ¤×Í ¡ÒÃËÒ¤èÒ àÁ·ÃÔ¡«ì R, A, áÅÐ P «Ö § àÁ·ÃÔ¡«ì ·Ñ § ÊÒÁ¹Õ ÊÒÁÒöËÒä´é §èÒÂã¹·Ò§ »¯ÔºÑµÔ ¡ÅèÒǤ×Í ¨Ò¡áºº¨ÓÅͧ·Õ ãªé 㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµ µÒÁÃÙ» ·Õ 3.2 ãËé ·Ó¡ÒÃÊè§ ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

{ak }

¨Ó¹Ç¹Ë¹Ö §à«¡àµÍÃì (sector) ËÃ×Í 4096 ºÔµ à¢éÒä»ã¹Ãкº à¾× Í·Ó¡ÒÃà¢Õ¹¢éÍÁÙÅŧ

ä»ã¹Ê× ÍºÑ¹·Ö¡ (¹Ñ ¹¤×Í ¼Ùéãªé·ÃÒºá¹è¹Í¹ÇèÒ

{ak }

¤×ÍÍÐäÃ) ¨Ò¡¹Ñ ¹ ¡çãËéËÑÇÍèÒ¹·Ó¡ÒÃÍèÒ¹¢éÍÁÙÅ

¨Ò¡Ê× Í ºÑ¹·Ö¡ áÅéÇ Êè§ ÊÑ­­Ò³ read back ·Õ ä´é ¼èÒ¹ä»Âѧ ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó áÅÐǧ¨ÃªÑ¡ µÑÇÍÂèÒ§ ·Ó ãËéä´é¼ÅÅѾ¸ìà» ¹ÅӴѺ¢éÍÁÙÅ à¾ÃÒÐ©Ð¹Ñ ¹ àÁ× Íä´é¢éÍÁÙÅ

{sk }

{ak }

«Ö §ã¹·Ò§»¯ÔºÑµÔ ¼ÙéãªéÊÒÁÒö·Õ ¨Ð·ÃÒºä´éÇèÒ¢éÍÁÙÅ

áÅÐ

{sk }

{sk }

¤×ÍÍÐäÃ

áÅéÇ ¡ç·Ó¡ÒäӹdzËÒ¤èÒàÁ·ÃÔ¡«ì R, A, áÅÐ P â´Â

ãªéâ»Ãá¡ÃÁ MATLAB [13] ËÃ×Í SCILAB [14] ¨ÐàËç¹ä´éÇèÒ ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáºº MMSE ÊÒÁÒö·Óä´é§èÒÂã¹·Ò§»¯ÔºÑµÔ àÁ× Íà·Õº¡ÑºÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáººÍ× ¹æ

3.3

¼Å¡Ò÷´Åͧ

ã¹Êèǹ¹Õ ¨Ð·Ó¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ ẺµèÒ§æ â´Âãªé Ẻ¨ÓÅͧã¹ÃÙ» 3

·Õ 3.2 ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ §

¨Ò¡ÃÙ» ·Õ 3.2 ÊÑ­­Ò³ read back ·Õ ä´é ¨Ò¡ËÑÇÍèÒ¹

ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Òä³ÔµÈÒʵÃìä´é ¤×Í [40]

p(t) =

S−1 X

bk g(t − kT + ∆tk ) + n(t)

(3.26)

k=0

3

¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ ẺµèÒ§æ ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ ÊÒÁÒöÈÖ¡ÉÒ

ÃÒÂÅÐàÍÕ´ä´éã¹ [19]


3.3.

¼Å¡Ò÷´Åͧ

55

â´Â·Õ

bk = (ak − ak−1 )/2

ʶҹкǡËÃ×Íź áÅÐ ºÔµµÑÇ·Õ

k

¤×Í ºÔµà»ÅÕ Â¹Ê¶Ò¹Ð (àÁ× Í

bk = 0

∆t

¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § µÒÁÊÁ¡ÒÃ·Õ (1.2),

n(t)

ÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹á»Å§

ËÁÒ¶֧ äÁèÁÕ¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹Ð),

·Õ Áըӹǹ·Ñ §ËÁ´ 4096 ºÔµ (1 à«¡àµÍÃì),

a jitter noise), áÅÐ

bk = ±1

ak ∈ ±1

¤×Í ¢éÍÁÙÅÍÔ¹¾Øµ

g(t) ¤×Í ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð¢Í§Ãкº

¤×Í ÊÑ­­Ò³Ãº¡Ç¹¨Ôµ àµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ (medi

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN) ·Õ ÁÕ ¤ÇÒÁ˹Òá¹è¹

Ê໡µÃÑÁ ¡ÓÅѧẺÊͧ´éÒ¹à·èÒ ¡Ñº

N0 /2

ã¹Ë¹Ñ§Ê×Í àÅèÁ ¹Õ ÊÑ­­Ò³Ãº¡Ç¹¨Ôµ àµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡

¨Ð¶Ù¡ ¨ÓÅͧãËé ÁÕ ÅѡɳÐà» ¹ ¡ÒÃàÅ× Í¹µÓáË¹è§ ¢Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐẺÊØèÁ (random transi tion shift) «Ö §ÁÕ¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹ (Gaussian probability den sity function) ·Õ ÁÕ ¤èÒà©ÅÕ Â à·èÒ ¡Ñº ¤èÒ ÈÙ¹Âì áÅФèÒ ¤ÇÒÁá»Ã»Ãǹà·èÒ ¡Ñº

N (0, |bk |σj2 ))

áÅж١¨Ó¡Ñ´ãËéÁÕ¤èÒäÁèà¡Ô¹

T

¤×Í ¤èÒÊÑÁºÙóì¢Í§

à«ÅÅì

áÅÐ

|bk |

ÊÑ­­Ò³ read back,

p(t),

T /2)

àÁ× Í

σj

|bk |σj2

(¹Ñ ¹¤×Í

∆tk ∼

¨Ð¶Ù¡¡Ó˹´à» ¹¨Ó¹Ç¹à»ÍÃìà«ç¹µì¢Í§ºÔµ

bk

¨Ð¶Ù¡ Êè§ ¼èÒ¹ä»Âѧ ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó ºÑµà·ÍÃìàÇÔÃìµ (Butterworth)

Íѹ´Ñº·Õ 7 áÅШзӡÒêѡµÑÇÍÂèÒ§ÊÑ­­Ò³

p(t)

´éÇÂÍѵÃÒ¤ÇÒÁ¶Õ à·èҡѺ

1/T

(¹Ñ ¹¤×Í ¢éÍÁÙÅá«Á

à» ÅáµèÅÐá«Áà» Å·Õ ä´é¨Ò¡¡ÒêѡµÑÇÍÂèÒ§¨ÐÍÂÙèËèÒ§¡Ñ¹ 1 ºÔµà«ÅÅì) â´Âã¹·Õ ¹Õ ¨ÐÊÁÁصÔÇèÒ ¡Ãкǹ¡Òà 4

à¢éÒ ¨Ñ§ËÇÐÃÐËÇèÒ§ÊÑ­­Ò³ read back áÅÐǧ¨ÃªÑ¡ µÑÇÍÂèÒ§ à» ¹ ẺÊÁºÙóì nization) ¨Ò¡¹Ñ ¹ ÅӴѺ¢éÍÁÙÅ

{sk }

(perfect synchro

¡ç¨Ð¶Ù¡Êè§ä»ÂѧÍÕ¤ÇÍäÅà«ÍÃì à¾× Í»ÃѺÃÙ»ÃèÒ§ÊÑ­­Ò³ãËéà» ¹ä»

µÒÁ·ÒÃìà¡çµ ·Õ Ãкºµéͧ¡Òà áÅÐǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¡ç ¨Ð·Ó¡ÒöʹÃËÑÊ ÅӴѺ ¢éÍÁÙÅ ÅӴѺ¢éÍÁÙÅ

{ak }

{yk }

à¾× Í ËÒ

·Õ à» ¹ä»ä´éÁÒ¡·Õ ÊØ´

㹡Ò÷´Åͧ ¨Ð¹ÔÂÒÁ Electronics SNR (ËÃ×ÍàÃÕ¡ÊÑ ¹æ ÇèÒ SNR) ãËéÁÕ¤èÒà·èҡѺ

µ SNR = 10 log10 â´Â·Õ

Vp = g(∞) = 1

pulse) ³ àÇÅÒ

t

=

Vp 2 σ2

¶ (dB)

¤×Í ¢¹Ò´¢Í§ÊÑ­­Ò³¾ÑÅÊì à»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È áÅÐ

σn2

=

N0 /(2T )

(3.27)

(isolated transition

¤×Í ¡ÓÅѧ ¢Í§ÊÑ­­Ò³Ãº¡Ç¹

n(t)

¹Í¡¨Ò¡¹Õ

áµèÅШش ¢Í§ BER ¶Ù¡ ¤Ó¹Ç³â´Âãªé ¢éÍÁÙÅ ËÅÒÂæ à«¡àµÍÃì ¨¹¡ÇèÒ ¨Ðä´é ¢éͼԴ¾ÅÒ´ÃÇÁäÁè ¹éÍ¡ÇèÒ 1000 ºÔµ áÅзÒÃìà¡çµ áÅÐÍÕ¤ÇÍäÅà«ÍÃì ·Õ ãªé ¨Ð¶Ù¡ Í͡ẺãËé àËÁÒÐÊÁ¡Ñº ¡Ò÷ӧҹã¹áµèÅÐ ND 4

¹Ñ ¹¤×Í Ç§¨ÃªÑ¡µÑÇÍÂèÒ§·ÃÒºÇèÒ ¨Ð·Ó¡ÒêѡµÑÇÍÂèÒ§ÊÑ­­Ò³ read back ·Õ µÓá˹è§ã´ à¾× ÍãËé¢éÍÁÙÅá«Áà» Å·Õ ä´éÍÍ¡

ÁÒÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´


56

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

³ ÃдѺ BER =

µÑÇÍÂèÒ§·Õ 3.1

10−5

µÑÇÍÂèÒ§¹Õ ¨ÐáÊ´§¢Ñ ¹µÍ¹¡ÒäӹdzËÒ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì µÒÁÊÁ¡Òà (3.19)

(3.21) ãËé¾Ô¨ÒóÒẺ¨ÓÅͧÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §ã¹ÃÙ»·Õ 3.2 ÊÓËÃѺ ND = 1.5 áÅÐ S NR = 20 dB ¶éÒ¡Ó˹´ãËéÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ ¢Òà¢éҢͧÍÕ¤ÇÍäÅà«ÍÃì ·Õ ÊÍ´¤Åéͧ¡Ñº

{ak }

{ak }

¤×Í

{1,

=

{sk }

=

1, 1,

−1,

1,

−1}

áÅÐÅӴѺ¢éÍÁÙÅ´éÒ¹

{−0.5337, −0.0662,

0.8821, 0.8122,

0.3219, 0.0260} ¨§¤Ó¹Ç³ËÒ·ÒÃìà¡çµáºº GPR ¢¹Ò´ 3 á·ç» áÅÐÍÕ¤ÇÍäÅà«ÍÃ좹Ҵ 5 á·ç» ·Õ ÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ µÒÁà§× ͹䢺ѧ¤Ñº ´Ñ§µèÍä»¹Õ ¡) à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡

¢) à§× ͹䢺ѧ¤Ñºáºº

h0 = 1

h1 = 1

¤) à§× ͹䢺ѧ¤Ñºáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇÂ

§) à§× ͹䢺ѧ¤Ñºáºº·ÒÃìà¡çµà©¾ÒÐ àÁ× Í¡Ó˹´ãËé ÇÔ¸Õ ·Ó

H(D) = 1 + 2D + D2

¢Ñ ¹µÍ¹áá㹡ÒÃËÒ·ÒÃìà¡çµ áÅÐÍÕ¤ÇÍäÅà«ÍÃì ¤×Í ¡ÒÃËÒ¤èÒ àÁ·ÃÔ¡«ì R, A, áÅÐ P µÒÁ

ÊÁ¡Òà (3.6) (3.8) àÁ× Í

S = 6, K = 2,

áÅÐ

L=3

«Ö § ¨Ò¡¡Òäӹdz¨Ðä´éÇèÒ àÁ·ÃÔ¡«ì

ÁÕ¤èÒà·èҡѺ

 0.3052

0.1926

−0.0549 −0.1440 −0.0868

1.0000

−0.2000

     0.1926 0.3052 0.1926 −0.0549 −0.1440      R =  −0.0549 0.1926 0.3052 0.1926 −0.0549       −0.1440 −0.0549 0.1926 0.3052 0.1926    −0.0868 −0.1440 −0.0549 0.1926 0.3052 àÁ·ÃÔ¡«ì

A

ÁÕ¤èÒà·èҡѺ

 0.5000

    A =  −0.2000 1.0000 −0.2000    0.5000 −0.2000 1.0000

R


3.3.

¼Å¡Ò÷´Åͧ

áÅÐàÁ·ÃÔ¡«ì

P

57

ÁÕ¤èÒà·èҡѺ

 0.4976

0.3867

0.1740

   0.2664 0.4976 0.3867   P =  −0.0390 0.2664 0.4976    −0.1983 −0.0390 0.2664  −0.0994 −0.1983 −0.0390 â´Âã¹·Õ ¹Õ ¨Ð¢ÍáÊ´§µÑÇÍÂèÒ§¡ÒäӹdzËÒÊÁÒªÔ¡ ËÒä´é¨Ò¡

A(i, j) = E

" 5 X

(i, j)

         

ºÒ§µÑÇ ¢Í§àÁ·ÃÔ¡«ì A ´Ñ§¹Õ àÁ·ÃÔ¡«ì A

# ak−i ak−j , 0 ≤ i, j ≤ 2

k=0

{ak } = {a0 , a1 , a2 , a3 , a4 , a5 } # " 5 X ak ak A(0, 0) = E

à¹× ͧ¨Ò¡ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

=

{1,

1, 1,

−1,

1,

−1}

´Ñ§¹Ñ ¹

k=0

= E [a0 a0 + a1 a1 + a2 a2 + a3 a3 + a4 a4 + a5 a5 ] 1 = [(1)(1) + (1)(1) + (1)(1) + (−1)(−1) + (1)(1) + (−1)(−1)] 6 = 1 A(0, 1) = E

" 5 X

# ak ak−1

k=0

= E [a1 a0 + a2 a1 + a3 a2 + a4 a3 + a5 a4 ] 1 = [(1)(1) + (1)(1) + (−1)(1) + (1)(−1) + (−1)(1)] 5 = −0.2 A(0, 2) = E

" 5 X

# ak ak−2

k=0

= E [a2 a0 + a3 a1 + a4 a2 + a5 a3 ] 1 = [(1)(1) + (−1)(1) + (1)(1) + (−1)(−1)] 4 = 0.5


58

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊÓËÃѺ ÊÁÒªÔ¡ ¢Í§àÁ·ÃÔ¡«ìÍ× ¹æ ¡ç ÊÒÁÒö¤Ó¹Ç³ËÒä´é ã¹ÅѡɳÐà´ÕÂǡѹ àÁ× Í ä´é àÁ·ÃÔ¡«ì R, A, áÅÐ P áÅéÇ áÅШҡ⨷Âì¨Ðä´éÇèÒ

H

=

[h0 , h1 , h2 ]T

áÅÐ

F

=

[f−2 , f−1 , f0 , f1 , f2 ]T

à¾ÃÒÐ©Ð¹Ñ ¹

·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃìÊÒÁÒöËÒ¤èÒä´éµÒÁÊÁ¡Òà (3.19) (3.21) µÒÁà§× ͹䢺ѧ¤ÑºµèÒ§æ ´Ñ§¹Õ

¡) à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡

−0.8587]T ,

áÅÐ

¢) à§× ͹䢺ѧ¤Ñº Ẻ

−1.6613]T ,

áÅÐ

F

=

h0 = 1

=

=

[1, 0, 0]T , λ = −0.0837, H

¨Ðä´é ÇèÒ

J

[0, 1, 0]T , λ = −2.2814, H

=

=

[4.7796, 1,

[7.0507, 1.8111, −4.6115, 3.0835, −1.9107]T

¤) à§× ͹䢺ѧ¤Ñºáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ ¨Ðä´éÇèÒ¤èÒÅѡɳÐ੾ÒТͧàÁ·ÃÔ¡«ì ¤×Í

[1, 0.1754,

=

[1.0936, −0.0722, −0.9953, −0.2400, −0.0805]T

h1 = 1 F

I

¨Ðä´éÇèÒ

{1.5182, −0.0476, −1.0211}, H

=

[0.3843, 0.7335, 0.5607]T ,

(A − PT R−1 P)

áÅÐ

F

=

[1.7138,

1.1288, 0.0256, 1.5285, −0.6140]T §) à§× ͹䢺ѧ¤Ñºáºº·ÒÃìà¡çµà©¾ÒÐ àÁ× Í

H

=

[1, 2, 1]T

¨Ðä´éÇèÒ

F

=

[4.2431, 2.6064, 0.1106,

3.1983, −1.3164]T

㹡Ò÷´Åͧ¹Õ ¨Ð·Ó¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ ẺµèÒ§æ ·Ñ § 5 Ẻ ¤×Í

1) ·ÒÃìà¡çµáºº PR1 ¹Ñ ¹¤×Í

2) ·ÒÃìà¡çµ PR2 ¹Ñ ¹¤×Í

H(D) = 1 + D

ËÃ×Íà¢Õ¹᷹´éÇÂ

H(D) = 1 + 2D + D2

3) ·ÒÃìà¡çµ·Õ Í͡ẺµÒÁà§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡

4) ·ÒÃìà¡çµ·Õ Í͡ẺµÒÁà§× ͹䢺ѧ¤Ñºáºº

[1, 1]

ËÃ×Íà¢Õ¹᷹´éÇÂ

[1, 2, 1]

h0 = 1

h1 = 1

5) ·ÒÃìà¡çµ·Õ Í͡ẺµÒÁà§× ͹䢺ѧ¤Ñºáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇÂ

ÃÙ» ·Õ 3.3 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµµèÒ§æ ÊÓËÃѺ ND = 2 â´Â·Õ Ãкº¨Ðãªé ·ÒÃìà¡çµáºº GPR ·Õ Áըӹǹ 3 á·ç» «Ö Ö§ã¹·Õ ¹Õ ¨Ðä´éÇèÒ ·ÒÃìà¡çµáººâÁ¹Ô¡

h0 = 1

¤×Í

1 + 1.15D


3.3.

¼Å¡Ò÷´Åͧ

59

−1

10

PR1 [1 1] PR2 [1 2 1] h0 = 1 h1 = 1 Unit energy

−2

BER

10

−3

10

−4

10

−5

10 14

15

16

17

18

19

20

SNR (dB)

ÃÙ»·Õ 3.3: »ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» BER ¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµµèÒ§æ ÊÓËÃѺ ND = 2

+

0.48D2 ,

·ÒÃìà¡çµáºº

0.45 + 0.77D

+

h1 = 1

0.45D2

¤×Í

0.55 + D

+

0.55D2 ,

áÅзÒÃìà¡çµáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ ¤×Í

¨Ò¡ÃÙ»·Õ 3.3 ¨ÐàËç¹ä´éÇèÒ â´ÂÀÒ¾ÃÇÁáÅéÇÃкº·Õ ãªé·ÒÃìà¡çµáºº GPR

¨ÐÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒÃкº·Õ ãªé·ÒÃìà¡çµáºº PR áÅÐÃкº·Õ ãªé·ÒÃìà¡çµáººâÁ¹Ô¡¨ÐãËé»ÃÐÊÔ·¸ÔÀÒ¾ ´Õ·Õ ÊØ´ â´Â੾ÒÐÍÂèÒ§ÂÔ §·Õ ND ÊÙ§æ à¾× Íà» ¹¡ÒÃÂ×¹ÂѹÇèÒ ·ÒÃìà¡çµáºº GPR ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒ·ÒÃìà¡çµáºº PR ¨Ð·Ó¡ÒÃà»ÃÕº à·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ND = 2.5 áµè ¤ÃÒÇ¹Õ ¨Ðãªé ·ÒÃìà¡çµ Ẻ GPR ·Õ ÁÕ ¨Ó¹Ç¹ 5 á·ç» (¨Ó¹Ç¹á·ç»ÁÒ¡¢Ö ¹ à¾× ÍÃͧÃѺ ND ·Õ à¾Ô Á¢Ö ¹) «Ö §ã¹¡Ã³Õ¹Õ ¨Ðä´éÇèÒ ·ÒÃìà¡çµáººâÁ¹Ô¡

1 + 1.42D + 1.06D2 + 0.43D3 + 0.08D4 , 0.53D3 + 0.14D4 , 0.2D4

·ÒÃìà¡çµáºº

áÅзÒÃìà¡çµáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ ¤×Í

h1 = 1

¤×Í

h0 = 1

¤×Í

0.47 + D + 0.96D2 +

0.20 + 0.5D + 0.66D2 + 0.5D3 +

ÃÙ» ·Õ 3.4 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµµèÒ§æ ·Õ ND = 2.5 ¨Ò¡¼ÅÅѾ¸ì ·Õ

ä´é¨Ð¾ºÇèÒ Ãкº·Õ ãªé·ÒÃìà¡çµáºº

h0 = 1

¨ÐãËé»ÃÐÊÔ·¸ÔÀÒ¾´Õ·Õ ÊØ´ ·Ñ §¹Õ à» ¹à¾ÃÒÐÇèÒ ·ÒÃìà¡çµáºº

GPR (â´Â੾ÒÐÍÂèÒ§ÂÔ § ·ÒÃìà¡çµáººâÁ¹Ô¡

h0 = 1)

¨ÐÁÕ ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ã¡Åéà¤Õ§¡Ñº ¼Å


60

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

−1

10

PR1 [1 1] PR2 [1 2 1] h0 = 1 h1 = 1 Unit energy

−2

BER

10

−3

10

−4

10

−5

10 19.0

19.5

20.0

20.5

21.0

21.5

22.0

SNR (dB)

ÃÙ»·Õ 3.4: »ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» BER ¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµµèÒ§æ ·Õ ND = 2.5

µÍºÊ¹Í§¢Í§ªèͧÊÑ­­Ò³ÁÒ¡¡ÇèÒ·ÒÃìà¡çµáºº PR ´Ñ§áÊ´§ã¹ÃÙ»·Õ 3.5 ¹Í¡¨Ò¡¹Õ ·ÒÃìà¡çµ Ẻ GPR ÁÕ á¹Çâ¹éÁ ·Õ ¨ÐªèÇ·ÓãËé ͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹ ¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇ (white noise) «Ö §à» ¹¢éÍ¡Ó˹´ ËÅÑ¡·Õ ¨ÐªèÇ·ÓãËéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´ [15] ¡ÒõÃǨ ÊͺÊÑ­­Ò³Ãº¡Ç¹

wk

ÇèÒÁդسÊÁºÑµÔà» ¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇËÃ×ÍäÁè ·Óä´éâ´Â¡ÒÃËÒ¤èÒÍѵÊË

ÊÑÁ¾Ñ¹¸ì (auto correlation) ¢Í§ÅӴѺ¢éÍÁÙÅ

{wk }

(´ÙÃÙ»·Õ 3.2) ¶éÒ¼ÅÅѾ¸ì·Õ ä´éÁÕ¤èÒ੾ÒеÓá˹è§

·Õ ¼ÅµèÒ§¢Í§àÇÅÒÁÕ¤èÒà·èҡѺ¤èÒÈÙ¹Âì áÅÐÁÕ¤èÒÈÙ¹Âì (ËÃ×ͤèÒã¡ÅéÈÙ¹Âì) àÁ× Í¼ÅµèÒ§¢Í§àÇÅÒÁÕ¤èÒà» ¹¤èÒ Í× ¹æ ¨ÐÊÃØ» ä´é ÇèÒ ÅӴѺ ¢éÍÁÙÅ ÊÑÁ¾Ñ¹¸ì ¢Í§ÅӴѺ ¢éÍÁÙÅ

{wk }

Ãкº·Õ ãªé·ÒÃìà¡çµáººâÁ¹Ô¡

wk

ÁÕ ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹ÊÕ ¢ÒÇ ÃÙ» ·Õ 3.6 áÊ´§¤èÒ ÍѵÊË

ÊÓËÃѺ Ãкº·Õ ãªé ·ÒÃìà¡çµ ẺµèÒ§æ ·Õ ND = 2.5 ¨Ò¡ÃÙ» ¨Ð¾ºÇèÒ

h0 = 1

¨ÐÁռŷÓãËéͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇÁÒ¡·Õ ÊØ´ àÁ× Íà·Õº¡Ñº·ÒÃìà¡çµáººÍ× ¹æ à¹× ͧ¨Ò¡ ·ÒÃìà¡çµáººâÁ¹Ô¡

h0 = 1

ãËé »ÃÐÊÔ·¸ÔÀÒ¾´Õ ·Õ ÊØ´ àÁ× Í à»ÃÕºà·Õº¡Ñº ·ÒÃìà¡çµ Ẻ


3.3.

¼Å¡Ò÷´Åͧ

61

1.0

Channel ND=2.5 PR1 [1 1] PR2 [1 2 1] h0 = 1 h1 = 1

0.9

Normalized magnitude

0.8 0.7

Unit energy 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

Normalized frequency (fT)

ÃÙ»·Õ 3.5: ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµáººµèÒ§æ à·Õº¡ÑºªèͧÊÑ­­Ò³·Õ ND = 2.5

GPR Í× ¹æ ã¹ÊèǹµèÍä»¹Õ ÊÑ­Åѡɳì GPRn ¨Ð¶Ù¡ãªéá·¹·ÒÃìà¡çµáººâÁ¹Ô¡

n

h0 = 1

·Õ Áըӹǹ

á·ç» ÃÙ» ·Õ 3.7(a) à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§·ÒÃìà¡çµ ẺµèÒ§æ ·Õ ND µèÒ§æ ÊÓËÃѺ ¡Ã³Õ ·Õ

äÁèÁÕÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡ (σj /T µéͧ¡ÒÃà¾× Í·Õ ¨Ð·ÓãËéä´é BER =

10−4

= 0%)

â´Â·Õ àÊé¹á¡¹

y

áÊ´§ SNR ·Õ Ãкº

´Ñ§¹Ñ ¹ Ãкº·Õ ãªé SNR ¹éÍ¡ÇèÒ¡çáÊ´§ÇèÒÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ

¡ÇèÒ ¨Ò¡ÃÙ»¨ÐàËç¹ä´éÇèÒ ·ÒÃìà¡çµáºº GPR ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒ·ÒÃìà¡çµáºº PR â´Â੾ÒÐÍÂèÒ§ÂÔ § ·Õ ND ÊÙ§æ ·Ñ §¹Õ à» ¹à¾ÃÒÐÇèÒ ·ÒÃìà¡çµáºº GPR ÁռŵͺʹͧàªÔ§¤ÇÒÁ¶Õ ã¡Åéà¤Õ§¡ÑºªèͧÊÑ­­Ò³ ÁÒ¡¡ÇèÒ·ÒÃìà¡çµáºº PR (´ÙÃÙ»·Õ 3.5) ÃÙ»·Õ 3.7(b) à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§·ÒÃìà¡çµáººµèÒ§æ ·Õ ÃдѺ¤ÇÒÁÃعáç¢Í§

σj /T

µèÒ§æ

ÊÓËÃѺ ND = 2.5 㹷ӹͧà´ÕÂǡѹ¨Ðä´éÇèÒ ·ÒÃìà¡çµáºº GPR ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒ (µéͧ¡Òà S NR ¹éÍ¡ÇèÒ à¾× ÍãËéä´é BER =

10−4

à·èҡѹ) ·ÒÃìà¡çµáºº PR ³ ·Ø¡ÃдѺ¤ÇÒÁÃعáç¢Í§

σj /T

Êѧࡵ¨Ð¾ºÇèÒ ·ÒÃìà¡çµáºº PR2 µéͧ¡Òà SNR ¹éÍ¡ÇèÒ (ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒ) ·ÒÃìà¡çµáºº EPR2 áÅÐ EEPR2 àÁ× Í ¤ÇÒÁÃعáç¢Í§

σj /T

ÁÕ ¤èÒ ÁÒ¡ ·Ñ §¹Õ ÍÒ¨¨Ðà» ¹ à¾ÃÒÐÇèÒ ·ÒÃìà¡çµ Ẻ PR ·Õ ÁÕ


62

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

1.2

PR1 [1 1] PR2 [1 2 1] Monic (h0 = 1) h1 = 1 Unit energy

1.0

Normalized amplitude

0.8

0.6

0.4

0.2

0.0

−0.2

−0.4

−0.6 0

1

2

3

4

5

6

7

8

9

Time difference (in bit period)

ÃÙ»·Õ 3.6: ÍѵÊËÊÑÁ¾Ñ¹¸ì¢Í§ÅӴѺ¢éÍÁÙÅ

{wk }

ÊÓËÃѺÃкº·Õ ·ÒÃìà¡çµáººµèÒ§æ ·Õ ND = 2.5

¨Ó¹Ç¹á·ç» ¹éͨÐÁÕ ¤ÇÒÁÍè͹äËÇ (sensitive) ¡Ñº ÊÑ­­Ò³Ãº¡Ç¹¨Ôµ àµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ ¹éÍ¡ÇèÒ ·ÒÃìà¡çµ Ẻ PR ·Õ ÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡ ã¹·Ò§µÃ§¡Ñ¹¢éÒÁ ·ÒÃìà¡çµáºº GPR ¨ÐãËé »ÃÐÊÔ·¸ÔÀÒ¾·Õ ´Õ ¢Ö ¹àÊÁÍ àÁ× Í ·ÒÃìà¡çµ·Õ ãªéÁըӹǹá·ç»ÁÒ¡¢Ö ¹ â´ÂäÁè¤Ó¹Ö§¶Ö§ÃдѺ¤ÇÒÁÃعáç¢Í§

3.4

σj /T

ÊÃØ»·éÒº·

·ÒÃìà¡çµáºº GPR ãËé »ÃÐÊÔ·¸ÔÀÒ¾´Õ ¡ÇèÒ ·ÒÃìà¡çµ Ẻ PR â´Â੾ÒÐÍÂèÒ§ÂÔ § ·Õ ND ÊÙ§æ ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ·ÒÃìà¡çµáºº GPR ÁÕ ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ã¡Åéà¤Õ§¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧ ÊÑ­­Ò³ÁÒ¡¡ÇèÒ ·ÒÃìà¡çµ Ẻ PR ¹Í¡¨Ò¡¹Õ ·ÒÃìà¡çµáºº GPR Âѧ ÁÕ á¹Çâ¹éÁ ·Õ ¨ÐªèÇ·ÓãËé ͧ¤ì »ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÕ ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹ÊÕ ¢ÒÇ «Ö §Ê觼ŷÓãËéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡ÂÔ §¢Ö ¹ ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ áÅÐÍÕ¤ÇÍäÅà«ÍÃì ÊÒÁÒö·Óä´é ËÅÒÂÇÔ¸Õ¡Òà áµè ã¹·Ò§»¯ÔºÑµÔ ¨Ð¾ºÇèÒ¡ÒÃ


3.4.

ÊÃØ»·éÒº·

63

ND 2 2.5 3

25

SNR required to achieve BER = 10

−4

in dB

26

5−tap GPR targets [1 1.14 0.58 0.16 0.03] [1 1.34 0.99 0.43 0.09] [1 1.44 1.31 0.74 0.22]

24

23

22

PR2 [1 2 1] EPR2 [1 3 3 1] EEPR2 [1 4 6 4 1] GPR5

21

20

2

2.1

2.2

2.3

(a)

2.4

2.5

2.6

2.7

2.8

2.9

3

Normalized density (ND)

SNR required to achieve BER = 10−4 in dB

30

Jitter (%) 0 3 6 9

29

28

[1 [1 [1 [1

5−tap GPR targets 1.34 0.99 0.43 0.09] 1.33 0.94 0.36 0.06] 1.27 0.72 0.13 −0.03] 1.02 0.15 −0.16 −0.01]

27

26

25

24

23

PR2 [1 2 1] EPR2 [1 3 3 1] EEPR2 [1 4 6 4 1] GPR5

22

21

20

0

1

2

3

4

(b)

ÃÙ»·Õ 3.7: (a) ¡ÃÒ¿ÃÐËÇèÒ§ SNR ·Õ µéͧ¡Òà áÅÐ ND àÁ× Í ·Õ µéͧ¡Òà áÅÐ

σj

·Õ ND = 2.5

5

6

7

8

Jitter (%)

σj = 0%

áÅÐ (b) ¡ÃÒ¿ÃÐËÇèÒ§ SNR


64

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

Í͡Ẻ·ÒÃìà¡çµ ´éÇÂÇÔ¸Õ¡Òà MMSE ÊÒÁÒö·Óä´é §èÒ¡ÇèÒ ÇÔ¸Õ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ẺÍ× ¹æ áÅÐ à¹× ͧ¨Ò¡¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ´éÇÂÇÔ¸Õ¡Òà MMSE ÁÕ à§× ͹䢺ѧ¤Ñº ËÅÒÂẺ áµè ¨Ò¡¡Ò÷´Åͧ¾º ÇèÒ ·ÒÃìà¡çµ·Õ Í͡Ẻâ´Âà§× ͹䢺ѧ¤Ñº ẺâÁ¹Ô¡ ¨ÐãËé »ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒ ·ÒÃìà¡çµ ·Õ Í͡Ẻâ´Â à§× ͹䢺ѧ¤ÑºÍ× ¹ ¹Í¡¨Ò¡¹Õ ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ·Õ ´Õ¡ç¤ÇÃ·Õ ¨ÐÍ͡Ẻ·ÒÃìà¡çµãËéàËÁÒÐÊÁ¡ÑºÊÀÒ¾ áÇ´ÅéÍÁ㹡Ò÷ӧҹ¢Í§µÑǪԻªèͧÊÑ­­Ò³ÍèÒ¹ (read channel chip) àªè¹ Í͡Ẻ·ÒÃìà¡çµÊÓËÃѺ áµèÅÐ ND, SNR, áÅÐ

σj /T

à» ¹µé¹ à¾× Í·Õ ¨Ð·ÓãËé ä´é ·ÒÃìà¡çµ ·Õ ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´ÀÒÂãµé

ÊÀÒ¾áÇ´ÅéÍÁ¡Ò÷ӧҹ·Õ ¡Ó˹´

3.5

à຺½ ¡ËÑ´·éÒº·

1. ¨§¾ÔÊÙ¨¹ìÊÁ¡Òà (3.9)

2. ¨§¾ÔÊÙ¨¹ìÊÁ¡Òà (3.24)

3. ¾Ô¨ÒóÒẺ¨ÓÅͧµÒÁÃÙ»·Õ 3.2 áµèà» ¹Ãкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ (longitudinal record ing) ·Õ ND = 2 áÅÐ SNR = 20 dB ¶éÒ¡Ó˹´ãËéÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ 1, 1} áÅÐÅӴѺ¢éÍÁÙÅ´éÒ¹¢Òà¢éҢͧÍÕ¤ÇÍäÅà«ÍÃì·Õ ÊÍ´¤Åéͧ¡Ñº 0.6184,

−0.1537,

0.0469,

−0.2469,

{ak }

{ak }

¤×Í

=

{1, −1,

{sk }

=

1,

−1,

{0.0767,

0.2096} ¨§¤Ó¹Ç³ËÒ·ÒÃìà¡çµ Ẻ GPR ¢¹Ò´ 3

á·ç» áÅÐÍÕ¤ÇÍäÅà«ÍÃ좹Ҵ 5 á·ç» ·Õ ÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ µÒÁà§× ͹䢺ѧ¤ÑºµèÍ仹Õ

¡) à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡ ¢) à§× ͹䢺ѧ¤Ñºáºº

h0 = 1

h1 = 1

¤) à§× ͹䢺ѧ¤Ñºáºº¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇ §) à§× ͹䢺ѧ¤Ñºáºº·ÒÃìà¡çµà©¾ÒÐ àÁ× Í¡Ó˹´ãËé

H(D) = 1 − D2

4. ¨§à¢Õ¹â»Ãá¡ÃÁ´éÇÂÀÒÉÒ SCILAB ËÃ×Í MATLAB à¾× ÍÇÒ´ÃÙ»ÊÑ­­Ò³¢Í§·ÒÃìà¡çµáºº [1

−4

1], [1

−2 −2

1], áÅÐ [1 0

−8

0 1] à¾× ;ÔÊÙ¨¹ìÇèÒ ·ÒÃìà¡çµàËÅèÒ¹Õ ÊÍ´¤Åéͧ¡Ñº¼Å

µÍºÊ¹Í§ä´ºÔµã¹â´àÁ¹àÇÅҢͧÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §


º··Õ 4

ǧ¨ÃµÃǨËÒ PRML

㹺·¹Õ ¨Ð͸ԺÒÂ¾× ¹°Ò¹¡Ò÷ӧҹ¢Í§à·¤¹Ô¤ ¼ÅµÍºÊ¹Í§ºÒ§Êèǹ¤ÇèÐà» ¹ ÁÒ¡ÊØ´ (PRML: partial response maximum likelihood) [27] «Ö §à» ¹à·¤¹Ô¤ËÅÑ¡·Õ ãªé㹡ÒõÃǨËÒ¢éÍÁÙŢͧÃкº ¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì â´Â¨Ðà¹é¹ ä»·Õ ËÅÑ¡¡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ (Viterbi algorithm) «Ö § ¶×Í ÇèÒ à» ¹ ÍÑÅ ¡ÍÃÔÖ ·ÖÁ ·Õ ãªé 㹡ÒöʹÃËÑÊ ¢éÍÁÙÅ ·Õ ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ãªé ã¹ ÎÒÃì´´ÔÊ¡ìä´Ã¿ì» ¨¨ØºÑ¹

4.1

º·¹Ó

à» ¹ ·Õ ·ÃÒº¡Ñ¹ ÇèÒ àÁ× Í ¤ÇÒÁ¨Ø ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì à¾Ô Á ¢Ö ¹ ¼Å¡Ãзº·Õ à¡Ô´ ¨Ò¡¡ÒÃá·Ã¡ÊÍ´ÃÐËÇèÒ§ ÊÑ­Åѡɳì (ISI) ¡ç¨ÐÂÔ §ÁÒ¡¢Ö ¹ ·ÓãËéäÁèÊÒÁÒöãªé§Ò¹Ç§¨ÃµÃǨËҨشÊÙ§ÊØ´ (peak detector) ä´éÍÕ¡ µèÍä» ´Ñ§¹Ñ ¹ à¾× ͨѴ¡ÒáѺ ISI ¨Ó¹Ç¹ÁÒ¡àËÅèÒ¹Õ Ç§¨ÃµÃǨËÒ PRML [27] ¨Ö§ä´é¶Ù¡¾Ñ²¹ÒáÅÐ ¹ÓÁÒãªé§Ò¹ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì¨¹¶Ö§·Ø¡Çѹ¹Õ â´Â·Ñ Çä» ¤ÓÇèÒ PRML ËÁÒ¶֧ à·¤¹Ô¤¡ÒÃãªé§Ò¹ÃèÇÁ ¡Ñ¹ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃìẺ PR áÅÐǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ µÒÁÃÙ»·Õ 4.1 «Ö §ÊÒÁÒöáºè§¢Ñ ¹µÍ¹¡Òà ·Ó§Ò¹ÍÍ¡à» ¹ 2 ¢Ñ ¹µÍ¹ ¤×Í

1) »ÃѺÃÙ»ÃèÒ§¢Í§ÊÑ­­Ò³ãËéà» ¹ä»µÒÁ·ÒÃìà¡çµ·Õ µéͧ¡ÒÃ

2) ¶Í´ÃËÑÊ¢éÍÁÙÅâ´Âãªéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ·Õ ÊÃéÒ§¨Ò¡·ÒÃìà¡çµ·Õ ¡Ó˹´äÇé 65


66

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

noise

ak

{± 1}

receiving filter

channel

detector

equalizer

âk

target response

ÃÙ»·Õ 4.1: ËÅÑ¡¡ÒÃ¾× ¹°Ò¹¢Í§à·¤¹Ô¤ PRML

A( D)

N (D)

C (D)

F ( D)

detector

 ( D )

ÃÙ»·Õ 4.2: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ

¢éʹբͧ¡ÒÃãªéà·¤¹Ô¤ PRML ¤×Í Ãкº¨Ð༪ԭ¡Ñº¡ÒâÂÒÂÊÑ­­Ò³Ãº¡Ç¹ (noise enhancement) ·Õ µ Ó áÅФÇÒÁ«Ñº«é͹¢Í§Ãкº¨Ð¹éÍÂŧ ã¹ÊèǹµèÍä»¹Õ ¨Ð͸ԺÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§ÍÕ¤ÇÍäÅà«ÍÃì Ẻ PR áÅÐÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔÍÂèÒ§ÅÐàÍÕ´

4.2

ÍÕ¤ÇÍäÅà«ÍÃì

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁè µèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ (equivalent discrete time chan nel model) µÒÁÃÙ» ·Õ 4.2 â´Â¢éÍÁÙÅ µèÒ§æ ¨ÐÍÂÙè ã¹â´àÁ¹

D

áÅÐÊÁÁØµÔ ãËé

N (D)

ú¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN) ¨Ò¡ÃÙ» ¨Ðä´é ÇèÒ ÊÑ­­Ò³·Õ ǧ¨ÃÀÒ¤ÃѺ ä´é ÃѺ

à» ¹ ÊÑ­­Ò³

P (D)

ÊÒÁÒö

à¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃì ¤×Í

P (D) = A(D)C(D) + N (D)

(4.1)


4.2.

ÍÕ¤ÇÍäÅà«ÍÃì

â´Â·Ñ Ç仪èͧÊÑ­­Ò³

67

C(D)

¨ÐÁÕ ÅѡɳÐà» ¹ ǧ¨Ã¡Ãͧ·Õ ÁÕ ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì ¨Ó¡Ñ´ (FIR:

nite impulse response) áÅÐÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡ (¡èÍ ãËé à¡Ô´ ISI ÁÒ¡) ¶éÒ äÁè ÁÕ ¡ÒÃãªé ÍÕ¤ÇÍäÅà«ÍÃì

F (D)

à¾× ÍÅ´¼Å¡Ãзº¢Í§ ISI ãËé¹éÍÂŧ ǧ¨ÃµÃǨËÒ (detector) ·Õ ãªé¨ÐµéͧÁÕ¤ÇÒÁ«Ñº«é͹ÁÒ¡

à¾× Í ¨Ñ´¡ÒáѺ ISI ¨Ó¹Ç¹ÁÒ¡ à¾ÃÒÐ©Ð¹Ñ ¹ à¾× Í Å´¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ ¨Ö§ ä´é ÁÕ ¡ÒÃ¹Ó ÍÕ¤ÇÍäÅà«ÍÃì ÁÒãªé §Ò¹ à¾× Í »ÃѺ ÃÙ»ÃèÒ§¢Í§ÊÑ­­Ò³ãËé à» ¹ 仵ÒÁ·ÒÃìà¡çµ ·Õ µéͧ¡Òà (à» ¹ ǧ¨Ã¡Ãͧ Ẻ FIR ·Õ Áըӹǹá·ç»¹éÍÂ) «Ö §¨ÐªèÇÂÅ´¼Å¡Ãзº¢Í§ ISI ãËé¹éÍÂŧä´é ÍÂèÒ§äáçµÒÁ ¡ÒÃ¹Ó ÍÕ¤ÇÍäÅà«ÍÃìÁÒãªé§Ò¹ÁÕ¢éÍàÊÕ ¤×Í (¶éÒÇÒ§ÍÕ¤ÇÍäÅà«ÍÃìäÇéËÅѧǧ¨ÃªÑ¡µÑÇÍÂèÒ§) ¨Ð·ÓãËéà¡Ô´»ÃÔÁÒ³ ˹èǧàÇÅÒ (delay) ¨Ó¹Ç¹ÁÒ¡ã¹ä·ÁÁÔ §ÅÙ» ¡ÅèÒǤ×Í ¨Ó¹Ç¹á·ç»¢Í§ÍÕ¤ÇÍäÅà«ÍÃìÂÔ §ÁÒ¡ »ÃÔÁÒ³ ˹èǧàÇÅÒ¡ç¨ÐÂÔ §ÁÒ¡ «Ö §¨ÐÊ觼ŷÓãËéÍѵÃÒ¡ÒÃÅÙèà¢éÒ (convergence rate) ¢Í§Ãкºä·ÁÁÔ §ÃԤѿàÇÍÃÕ ªéÒŧ ·ÓãËéǧ¨Ãà¿ÊÅçÍ¡ÅÙ» (PLL) äÁèÊÒÁÒöµÔ´µÒÁ¡ÒÃà»ÅÕ Â¹á»Å§à¿ÊáÅФÇÒÁ¶Õ ¢Í§ÊÑ­­Ò³ á͹ÐÅçÍ¡·Õ ¨Ð·Ó¡ÒêѡµÑÇÍÂèÒ§ä´é·Ñ¹ «Ö §ÍÒ¨¨ÐÊ觼ŷÓãËéÕà¡Ô´¡ÒÃÊÙ­àÊÕ¡Ãкǹ¡ÒÃà¢éҨѧËÇÐä´é

4.2.1

ÍÕ¤ÇÍäÅà«ÍÃìà຺¼ÅµÍºÊ¹Í§àµçÁ

ÍÕ¤ÇÍäÅà«ÍÃìẺ¼ÅµÍºÊ¹Í§àµçÁ (full response equalizer) ËÁÒ¶֧ÍÕ¤ÇÍäÅà«ÍÃì·Õ ¨Ð·ÓãËé¢éÍÁÙÅ àÍÒµì¾Øµ·Õ ä´éÁÕ¤èÒà·èҡѺ ¢éÍÁÙÅÍÔ¹¾Øµ

A(D)

ºÇ¡¡ÑºÊÑ­­Ò³Ãº¡Ç¹

W (D)

´Ñ§¹Ñ ¹ ¨Ò¡ÃÙ»·Õ 4.2 ¨Ð

ä´é ÇèÒ ÍÕ¤ÇÍäÅà«ÍÃì Ẻ¼ÅµÍºÊ¹Í§àµçÁ ¨ÐÁÕ ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì (impulse response) ã¹â´àÁ¹

D

¤×Í

F (D) = áÅТéÍÁÙÅàÍÒµì¾Øµ

Y (D)

1 C(D)

(4.2)

¢Í§ÍÕ¤ÇÍäÅà«ÍÃì¹Õ ¤×Í

Y (D) = P (D)F (D) á·¹¤èÒ

F (D)

(4.3)

¨Ò¡ÊÁ¡Òà (4.2) ŧã¹ÊÁ¡Òà (4.1) ¨Ðä´é

Y (D) = {A(D)C(D) + N (D)} = A(D) +

N (D) C(D) | {z } W (D)

1 C(D) (4.4)


68

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¹Ñ ¹¤×Í Í§¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ðà¢éÒ ä»ã¹Ç§¨ÃµÃǨËÒÊÑ­Åѡɳì (symbol detector) ¤×Í

W (D) = N (D)/C(D)

¶éÒ ÊÁÁØµÔ ÇèÒ

W (D)

ÁÕ ¤èÒ ¹éÍÂÁÒ¡ ǧ¨ÃµÃǨËÒÊÑ­ÅÑ¡É³ì ·Õ ãªé ¡ç

ÊÒÁÒö໠¹ Ẻ§èÒÂæ ä´é àªè¹ ǧ¨ÃµÃǨËÒ¢Õ´ àÊé¹ áºè§ ẺËÅÒÂÃдѺ (multi level threshold detector) à¾× Í ·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ

Y (D)

ÍÂèÒ§äáçµÒÁ ¢éÍàÊÕ ¢Í§¡ÒÃãªé ÍÕ¤ÇÍäÅà«ÍÃì Ẻ¼Å

µÍºÊ¹Í§àµçÁ ¡ç ¤×Í ÊÑ­­Ò³Ãº¡Ç¹

W (D)

·Õ ËŧàËÅ×Í ÍÂÙè ÍÒ¨¨Ð¡èÍ ãËé à¡Ô´ »ÃÒ¡®¡ÒÃ³ì ¡ÒâÂÒÂ

ÊÑ­­Ò³Ãº¡Ç¹ ¹Ñ ¹¤×Í

W (D)

ÁÕ ¤èÒ à» ¹ ¤èÒ Í¹Ñ¹µì ¶éÒ ªèͧÊÑ­­Ò³

C(D)

ÁÕ Ê໡µÃÑÁ ¤èÒ ÈÙ¹Âì

(spectral null) ·Õ ¤ÇÒÁ¶Õ ã´æ à¾ÃÒÐ©Ð¹Ñ ¹ ã¹·Ò§»¯ÔºÑµÔ ¨Ö§äÁè¹ÔÂÁ¹ÓÍÕ¤ÇÍäÅà«ÍÃìẺ¼ÅµÍºÊ¹Í§ àµçÁÁÒãªé§Ò¹ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

4.2.2

ÍÕ¤ÇÍäÅà«ÍÃìà຺¼ÅµÍºÊ¹Í§ºÒ§Êèǹ

ÍÕ¤ÇÍäÅà«ÍÃì Ẻ¼ÅµÍºÊ¹Í§ºÒ§Êèǹ (partial response equalizer) ¤×Í ÍÕ¤ÇÍäÅà«ÍÃì ·Õ ÊÒÁÒö à¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìä´é´Ñ§¹Õ

F (D) =

â´Â·Õ

H(D)

H(D) C(D)

(4.5)

¤×Í ¼ÅµÍºÊ¹Í§·ÒÃìà¡çµ (target response) ·Õ µéͧ¡Òà áÅÐàÁ× Íá·¹¤èÒ

F (D)

¹Õ ŧ

ã¹ÊÁ¡Òà (4.3) ¨Ðä´é

H(D) C(D) H(D) = A(D)H(D) + N (D) | {z } C(D) | {z } wanted signal

Y (D) = {A(D)C(D) + N (D)}

(4.6)

W (D)

¹Ñ ¹¤×Í ¢éÍÁÙÅàÍÒµì¾Øµ¢Í§ÍÕ¤ÇÍäÅà«ÍÃìẺ¼ÅµÍºÊ¹Í§ºÒ§Êèǹ¨Ð»ÃСͺ仴éÇ ¢éÍÁÙÅ·Õ µéͧ¡ÒÃ

A(D)H(D)

áÅÐÊÑ­­Ò³Ãº¡Ç¹

W (D) = N (D)H(D)/C(D)

¨Ò¡ÊÁ¡Òà (4.6) ¨ÐàËç¹ä´éÇèÒ

¢éÍÁÙÅ·Õ µéͧ¡ÒèÐÁÕ ISI ὧÍÂÙè áµèà¹× ͧ¨Ò¡ ǧ¨ÃÀÒ¤ÃѺ·ÃÒºÇèÒ ISI ¹Õ ¤×ÍÍÐäà (à¾ÃÒÐÇèÒà» ¹ ISI ·Õ à¡Ô´¨Ò¡·ÒÃìà¡çµ) ´Ñ§¹Ñ ¹ ISI ¹Õ ÊÒÁÒö·Õ ¨Ð¶Ù¡¨Ñ´¡ÒÃä´é´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ «Ö §¨Ð͸ԺÒµèÍä» ã¹ËÑÇ¢éÍ·Õ 4.3


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

69

¹Í¡¨Ò¡¹Õ àÁ× Í ¾Ô¨ÒóÒÊèǹ¢Í§ÊÑ­­Ò³Ãº¡Ç¹

W (D)

ã¹ÊÁ¡Òà (4.6) ¨Ð¾ºÇèÒ ÊÒà赯 ·Õ

Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì µéͧ¡ÒÃ·Õ ¨Ðä´é ·ÒÃìà¡çµ àªÔ§ ¤ÇÒÁ¶Õ àËÁ×͹¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³ ·ÓãËé

W (D)

ÁÕ ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇ

H(D) = C(D)

áÅéÇ ¨Ðä´é ÇèÒ

W (D) = N (D)

C(D)

N (D)

H(D)

·Õ ÁÕ ¼ÅµÍºÊ¹Í§

ãËé ÁÒ¡·Õ ÊØ´ ¡ç à¾× Í ·Õ ÇèÒ¨Ðä´é

ãËé ÁÒ¡·Õ ÊØ´ ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ¶éÒ

«Ö § ¶×Í ÇèÒ à» ¹ à§× ͹ä¢ËÅÑ¡ ·Õ ¨Ð·ÓãËé ǧ¨ÃµÃǨËÒ

ÇÕà·ÍÃìºÔÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´ ËÃ×Í¡ÅèÒÇÍÕ¡¹ÑÂË¹Ö §¤×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð¶Ù¡¾Ô¨ÒóÒÇèÒà» ¹ ǧ¨ÃµÃǨËÒ·Õ àËÁÒÐ·Õ ÊØ´ (opimal detector) ¶éÒͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³ ú¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔà» ¹ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇ [15] ´Ñ§¹Ñ ¹ ÍÒ¨¨ÐÊÃØ» ä´éÇèÒ ¶éҼŵͺʹͧàªÔ§¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµàËÁ×͹¡Ñº¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³ÁÒ¡ à·èÒã´ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºã¹ÃÙ»¢Í§ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ (BER) ÇÑ´·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨ÃµÃǨËÒ ÇÕà·ÍÃìºÔ¡ç¨Ð´ÕÁÒ¡¢Ö ¹à·èÒ¹Ñ ¹

4.3

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¤×Í Ç§¨ÃµÃǨËÒÅӴѺ (sequence detector) ·Õ ÊÃéÒ§â´Âãªé ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ (Viterbi algorithm) [15] à¾× Íãªé㹡ÒöʹÃËÑÊ¢éÍÁÙÅ·Õ ¶Ù¡à¢éÒÃËÑÊ´éÇ ÃËÑʤ͹âÇÅ٪ѹ (convo lutional code) [7] à·èÒ¹Ñ ¹ ã¹·Ò§»¯ÔºÑµÔáÅéÇ ªèͧÊÑ­­Ò³ÊÒÁÒö·Õ ¨Ð¶Ù¡¾Ô¨ÒóÒÇèÒà» ¹ÃËÑʤ͹ âÇÅ٪ѹ»ÃÐàÀ·Ë¹Ö §·Õ ÁÕÍѵÃÒÃËÑÊ (code rate) à·èҡѺ¤èÒË¹Ö § (¹Ñ ¹¤×Í ¢éÍÁÙÅÍÔ¹¾Øµ 1 ºÔµ àÁ× Íà¢éÒÃËÑÊ áÅéǨÐä´é¢éÍÁÙÅàÍÒµì¾ØµÍÍ¡ÁÒ 1 ºÔµàªè¹¡Ñ¹) ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÁÕ¤ÇÒÁÊÒÁÒö·Õ ¨Ñ´¡ÒáѺ ISI ·Õ ὧÍÂÙèã¹¢éÍÁÙÅ·Õ ¨Ð·Ó¡ÒöʹÃËÑÊä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ â´Â·Õ ¶éÒ ISI ÂÔ §ÁÒ¡ ¤ÇÒÁ«Ñº«é͹¢Í§ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¡ç¨ÐÂÔ §ÁÒ¡ áÅжéÒ ISI ¹éÍ ¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¡ç¨Ð¹éÍ à¹× ͧ¨Ò¡ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÕ ¤ÇÒÁ«Ñº«é͹ÁÒ¡¡ÇèÒ Ç§¨ÃµÃǨËÒẺ§èÒ (simple detector) àªè¹ ǧ¨ÃµÃǨËÒ¢Õ´ àÊé¹ áºè§ ẺËÅÒÂÃдѺ 㹡ÒÃ·Õ ¨ÐµÑ´ÊÔ¹ã¨ÇèÒ ¨Ð¹Óǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÒ ãªé §Ò¹ã¹ÃкºËÃ×Í äÁè ¹Ñ ¹ ãËé ¾Ô¨ÒóҨҡÃÙ» ·Õ 4.3 ´Ñ§µèÍä»¹Õ ¨Ò¡ÃÙ» ·Õ 4.3(a) ¶éÒ ªèͧÊÑ­­Ò³äÁè ÁÕ ISI ǧ¨ÃÀÒ¤ÃѺ ¡ç ÊÒÁÒö¹Óǧ¨ÃµÃǨËÒẺ§èÒÂÁÒãªé §Ò¹ä´é àÅ ¶éÒ ªèͧÊÑ­­Ò³ÁÕ ISI ¹éÍ µÒÁÃÙ» ·Õ 4.3(b) ǧ¨ÃÀÒ¤ÃѺ ¡ç ÊÒÁÒö¹Óǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÒãªé §Ò¹ä´é àÅ áµè ¶éÒ ªèͧÊÑ­­Ò³ ÁÕ ISI ¨Ó¹Ç¹ÁÒ¡µÒÁÃÙ» ·Õ 4.3(c) ǧ¨ÃÀÒ¤ÃѺ ¡ç ¤ÇÃ·Õ ¹ÓÍÕ¤ÇÍäÅà«ÍÃì ÁÒãªé §Ò¹ à¾× Í Å´¼Å¡Ãзº


70

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

AWGN

(a)

ak

simple detector

channel no ISI

âk

AWGN

(b)

ak

channel with low ISI C(D)

Viterbi detector C(D)

âk

AWGN

ak (c)

channel with severe ISI C(D)

equalizer

Viterbi detector H(D)

âk

target H(D) ÃÙ»·Õ 4.3: µÑÇÍÂèҧẺ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅÅѡɳеèÒ§æ

¢Í§ ISI ãËé ¹éÍÂŧ ¨Ò¡¹Ñ ¹ ¨Ö§ ¤èÍÂÊè§ ¢éÍÁÙÅ àÍÒµì¾Øµ ·Õ ä´é ¨Ò¡ÍÕ¤ÇÍäÅà«ÍÃì ä»·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¶×ÍÇèÒà» ¹Ç§¨ÃµÃǨËÒ¢éÍÁÙÅ·Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ áÅж١¹ÓÁÒãªé§Ò¹ã¹ËÅÒÂæ §Ò¹»ÃÐÂØ¡µì ÃÇÁ·Ñ §ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì â´Â·Õ ËÅÑ¡¡Ò÷ӧҹ¢Í§ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨ÐÍÂÙè º¹¾× ¹°Ò¹¢Í§ á¼¹ÀÒ¾à·ÃÅÅÔÊ à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´

(trellis diagram) «Ö § ÊÃéÒ§ÁÒ¨Ò¡

(FSM: nite state machine) ´Ñ§¹Ñ ¹ ¡è͹·Õ ¨Ð͸ԺÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§

ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ¼ÙéÍèÒ¹¤ÇèзӤÇÒÁà¢éÒã¨à¡Õ ÂǡѺ ÇÔ¸Õ¡ÒÃÊÃéÒ§à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ áÅÐá¼¹ÀÒ¾ à·ÃÅÅÔÊ¡è͹ «Ö §ÁÕÃÒÂÅÐàÍÕ´´Ñ§¹Õ


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

71

rk

H(D)

ak

D

{0,1} ÃÙ»·Õ 4.4: á¼¹ÀÒ¾ªèͧÊÑ­­Ò³áºº PR1,

µÒÃÒ§·Õ 4.1: ¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§

4.3.1

ak

áÅÐ

rk

H(D) = 1 + D

¢Í§ªèͧÊÑ­­Ò³

ak

ak−1

rk

0

0

0

0

1

1

1

0

1

1

1

2

H(D) = 1 + D

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´

¾Ô¨ÒóҪèͧÊÑ­­Ò³

H(D) = 1 + D

ã¹ÃÙ»·Õ 4.4 àÁ× Í ¢éÍÁÙźԵÍÔ¹¾Øµ

ak ∈ {0, 1}

à¢éÒä»ã¹ªèͧÊÑ­­Ò³·ÓãËéä´éà» ¹¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³ (channel output) à¢Õ¹µÒÃÒ§áÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾Øµ â´Â·Õ ¢éÍÁÙźԵ

ak−1

register) 㹺ÅçÍ¡

ak

áÅТéÍÁÙÅàÍÒµì¾Øµ

rk

¶Ù¡Ê觼èÒ¹

rk ∈ {0, 1, 2}

¶éÒ

¨Ðä´éµÒÁµÒÃÒ§·Õ 4.1

ÍÒ¨¨Ð¶Ù¡¾Ô¨ÒóÒÇèÒà» ¹¢éÍÁÙÅ·Õ ËŧàËÅ×ÍÍÂÙèã¹ àèÔÊàµÍÃìẺàÅ× Í¹ (shift

D

¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§

ak

áÅÐ

rk

ã¹µÒÃÒ§·Õ 4.1 ÊÒÁÒöáÊ´§ãËéÍÂÙèã¹ÃÙ»¢Í§ à¤Ã× Í§Ê¶Ò¹Ð

¨Ó¡Ñ´ (FSM: nite state machine) «Ö § ¡ç ¤×Í áºº¨ÓÅͧ¢Í§¡Òäӹdz·Õ áÊ´§ãËé àËç¹ ¶Ö§ ¡Òà à»ÅÕ Â¹á»Å§¢Í§ ¢éÍÁÙÅÍÔ¹¾Øµ, ʶҹÐàÃÔ Áµé¹ (start state), ʶҹеèÍä» (next state), áÅТéÍÁÙÅ àÍÒµì¾ØµªèͧÊÑ­­Ò³ ´Ñ§áÊ´§ã¹ÃÙ»·Õ 4.5 â´Â·Õ ¢éÍÁÙźԵ 0 áÅкԵ 1 ·Õ ÍÂÙèã¹Ç§¡ÅÁ ¤×Í ¤èÒ·Õ ÍÂÙè


72

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

1/1 X/Y is denoted as:

1

0

0/0

1/2

ak Y = the output bit rk X = the input bit

0/1 ÃÙ»·Õ 4.5: à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ªèͧÊÑ­­Ò³ PR1,

ã¹àèÔÊàµÍÃìẺàÅ× Í¹ ³ àÇÅÒ¢³Ð¹Ñ ¹, ªèͧÊÑ­­Ò³

rk

X

¤×Í ¤èÒ¢éÍÁÙÅÍÔ¹¾Øµ

H(D) = 1 + D

ak ,

áÅÐ

Y

¤×Í ¤èÒ¢éÍÁÙÅàÍÒµì¾Øµ

µÑÇÍÂèÒ§àªè¹ ÊÁÁصÔÇèÒ ³ µÍ¹¹Õ ¤èÒ·Õ ÍÂÙèã¹àèÔÊàµÍÃìẺàÅ× Í¹ ¤×Í ¤èÒ 1 ¶éÒ¢éÍÁÙÅ

ÍÔ¹¾Øµ ·Õ à¢éÒ ÁÒ㹪èͧÊÑ­­Ò³ ¤×Í

ak = 0

¨Ðä´é ÇèÒ ¤èÒ ·Õ ÍÂÙè ã¹àèÔÊàµÍÃì ẺàÅ× Í¹àÁ× Í àÇÅÒ¼èÒ¹ä»

Ë¹Ö §Ë¹èÇ (ËÃ×Í Ê¶Ò¹ÐµèÍä») ¤×Í ¤èÒ 0 â´Â¨Ðä´é¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

rk = 1

㹷ӹͧ

à´ÕÂǡѹ ¶éҵ͹¹Õ ¤èÒ·Õ ÍÂÙèã¹àèÔÊàµÍÃìẺàÅ× Í¹ ¤×Í ¤èÒ 0 áÅТéÍÁÙÅÍÔ¹¾Øµ·Õ à¢éÒÁÒ㹪èͧÊÑ­­Ò³ ¤×Í

ak = 0

¨Ðä´é ÇèÒ ¤èÒ ·Õ ÍÂÙè ã¹àèÔÊàµÍÃì ẺàÅ× Í¹àÁ× Í àÇÅÒ¼èÒ¹ä»Ë¹Ö § ˹èÇ ¤×Í ¤èÒ 0 â´Â¨Ðä´é

rk = 0

¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

ËÁÒÂà˵Ø

¶éÒ ¡Ó˹´ãËé

|A|

à» ¹µé¹

¤×Í ¨Ó¹Ç¹¤èÒ ·Õ à» ¹ ä»ä´é ·Ñ §ËÁ´¢Í§¢éÍÁÙÅ ÍÔ¹¾Øµ

ak ,

áÅÐ

ν

˹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³ ´Ñ§¹Ñ ¹ ¨Ó¹Ç¹Ê¶Ò¹Ð (state) ·Ñ §ËÁ´ã¹à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ ¤×Í àªè¹ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ»·Õ 4.4 ¨Ðä´éÇèÒ

|A| = 2

¨Ó¡Ñ´¢Í§ªèͧÊÑ­­Ò³ PR1 ¨ÐÁÕ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´

4.3.2

áÅÐ

21 = 2

ν=1

¤×Í

|A|ν

à¾ÃÒÐ©Ð¹Ñ ¹ à¤Ã× Í§Ê¶Ò¹Ð

ʶҹÐ

à༹ÀÒ¾à·ÃÅÅÔÊ

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´·Õ áÊ´§ã¹ÃÙ»·Õ 4.5 ÊÒÁÒöáÊ´§ãËéÍÂÙèã¹ÃÙ»¢Í§á¼¹ÀÒ¾à·ÃÅÅÔÊä´é ´Ñ§áÊ´§ã¹ ÃÙ»·Õ 4.6 â´Â·Õ ¨Ø´µèÍ (node) ·Ò§´éÒ¹«éÒÂÁ×Í ¤×Í Ê¶Ò¹ÐàÃÔ Áµé¹ («Ö §ÁÕ¤èÒà·èҡѺ ´éÒ¹¢ÇÒÁ×Í ¤×Í Ê¶Ò¹ÐµèÍä», ÅÙ¡ÈÃàÊé¹»Ðãªéá·¹¡ÒÃÊ觢éÍÁÙÅÍÔ¹¾Øµ ¡ÒÃÊ觢éÍÁÙÅÍÔ¹¾Øµ

ak = 1,

ak = 0,

ak−1 ),

¨Ø´µèÍ·Ò§

ÅÙ¡ÈÃàÊé¹·Öºãªéá·¹

áÅеÑÇàÅ¢·Õ áÊ´§ÍÂÙèã¹áµèÅÐàÊé¹ÊÒ¢Ò (branch) ¤×Í ¤èÒ¢éÍÁÙÅàÍÒµì¾Øµ


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

73

[ ak −1]

0

0

0

1

ak = 0

1

1

ak = 1

1

2

ÃÙ»·Õ 4.6: á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ PR1,

ªèͧÊÑ­­Ò³

rk

µÑÇÍÂèÒ§àªè¹ ¶éÒʶҹÐàÃÔ Áµé¹ ¤×ͤèÒ 0 àÁ× Í¢éÍÁÙÅÍÔ¹¾Øµ

·ÓãËéä´é¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

rk = 1

àÃÔ Áµé¹ ¤×Í ¤èÒ 1 àÁ× Í ¢éÍÁÙÅ ÍÔ¹¾Øµ

ak = 1

rk = 2

H(D) = 1 + D

ak = 1

à¢éÒÁÒã¹Ãкº ¨Ð

áÅÐʶҹеèÍä» ¤×ͤèÒ 1 㹷ӹͧà´ÕÂǡѹ ¶éÒʶҹРà¢éÒ ÁÒã¹Ãкº ¨Ð·ÓãËé ä´é ¢éÍÁÙÅ àÍÒµì¾Øµ ªèͧÊÑ­­Ò³

áÅÐʶҹеèÍä» ¤×ͤèÒ 1

㹷ӹͧà´ÕÂǡѹ ÃÙ»·Õ 4.7 áÊ´§µÑÇÍÂèҧἹÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ EPR4 (extended PR4),

H(D) = 1 + D − D2 − D3 ,

àÁ× Í ¢éÍÁÙÅÍÔ¹¾Øµ

¢Í§á¼¹ÀÒ¾à·ÃÅÅÔʨÐÁըӹǹà·èҡѺ

{ak } ∈ {0, 1}

|A|ν = 23 = 8

«Ö §ã¹¡Ã³Õ¹Õ ¨Ó¹Ç¹Ê¶Ò¹Ð

ʶҹРà¹× ͧ¨Ò¡ ªèͧÊÑ­­Ò³ EPR4 ÁÕ

¨Ó¹Ç¹Ë¹èǤÇÒÁ¨Óà·èҡѺ 3 ˹èÇÂ

µÑÇÍÂèÒ§·Õ 4.1

¡Ó˹´ãËé¢éÍÁÙÅÍÔ¹¾Øµ

H(D) = 1 − D2 ,

ak ∈ {−1, 1}

áÅéÇä´é¢éÍÁÙÅàÍÒµì¾Øµà» ¹

¶Ù¡Ê觼èÒ¹à¢éÒä»ÂѧªèͧÊÑ­­Ò³áºº PR4,

rk ∈ {−2, 0, 2}

¨§

¡) ÇÒ´á¼¹ÀÒ¾ºÅçÍ¡¢Í§ªèͧÊÑ­­Ò³áºº PR4 ¹Õ (¤ÅéÒ¡ѺÃÙ»·Õ 4.4)

¢) à¢Õ¹ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾ØµáÅТéÍÁÙÅàÍÒµì¾Øµ¢Í§ªèͧÊÑ­­Ò³¹Õ

¤) ÊÃéÒ§à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´

§) ÊÃéҧἹÀÒ¾à·ÃÅÅÔÊ


74

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

[ak −1 ak −2 ak −3] (0)

0

000

1

(1)

100

(2)

010

(3)

110

(4)

001

(5)

101

1 2 -1 0 1 -1

0 0 0 1 -2

(6)

011

(7)

111

-1 -1

ak = 1 ak = 0

0

yk ÃÙ»·Õ 4.7: á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ EPR4,

H(D) = 1 + D − D2 − D3

ÇÔ¸Õ·Ó ¡)

¨Ò¡ªèͧÊÑ­­Ò³·Õ ¡Ó˹´ãËé ÊÒÁÒöáÊ´§à» ¹á¼¹ÀÒ¾ºÅçÍ¡¢Í§ªèͧÊÑ­­Ò³áºº PR4 ä´é

´Ñ§ÃÙ»·Õ 4.8(a) ¢)

à¹× ͧ¨Ò¡

H(D) =

P k

hk Dk = 1 − D2

´Ñ§¹Ñ ¹ ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ì ÃÐËÇèÒ§¢éÍÁÙÅ

ÍÔ¹¾ØµáÅТéÍÁÙÅàÍÒµì¾Øµ¢Í§ªèͧÊÑ­­Ò³¹Õ ¤×Í

rk = ak ∗ hk = ak − ak−2 ¤)

¨Ò¡ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾ØµáÅТéÍÁÙÅàÍÒµì¾Øµ·Õ ä´éã¹¢éÍ ¢) à¤Ã× Í§Ê¶Ò¹Ð

¨Ó¡Ñ´ÊÒÁÒöáÊ´§ä´éµÒÁÃÙ»·Õ 4.8(b)


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

75

[ ak −1ak −2 ]

rk ak

D

{±1} (a) Block diagram: PR4

1 -11

1/2 -1/0

-1 - 1-1 -

1/0

-1/0

0

-1 -1

D

2

1 -1

1/2

-1 1 11

0 -2 0

1/0

-1/-2

-1/-2

0 2 -2

1 1

ak = -1 ak = 1

-1111

(b) FSM

(c) Trellis diagram

ÃÙ»·Õ 4.8: (a) á¼¹ÀÒ¾ºÅçÍ¡, (b) à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, áÅÐ (c) á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§ªèͧÊÑ­­Ò³

H(D) = 1 − D2

§)

á¼¹ÀÒ¾à·ÃÅÅÔÊÊÒÁÒöÊÃéÒ§ä´é¨Ò¡à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ ´Ñ§áÊ´§ã¹ÃÙ»·Õ 4.8(c)

4.3.3

ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ

¨Ò¡à«µ¢Í§ªØ´µÑÇÍÑ¡ÉÃ

H(D) µÒÁÃÙ»·Õ 4.9 â´Â·Õ ak ¤×Í ¢éÍÁÙźԵÍÔ¹¾Øµ·Õ ¶Ù¡àÅ×Í¡ÁÒ P A, H(D) = νk=0 hk Dk ¤×Í ªèͧÊÑ­­Ò³ÇÔÂص (discrete channel), hk

¤×Í ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô µÑÇ ·Õ

k, ν

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³

¤×Í Ë¹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³,

nk

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ

¢ÒÇẺºÇ¡ (AWGN) ·Õ ÁÕ¤èÒà©ÅÕ Âà·èҡѺ¤èÒÈÙ¹Âì áÅФÇÒÁá»Ã»Ãǹà·èҡѺ¤èÒ ä´é´éÇÂ

nk ∼ N (0, σ 2 ), rk

ÍÔ¹¾Øµ

{ak }

(â´Â·Ñ Çä»

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³, áÅÐ

L = 4096

L

σ2

ËÃ×Íà¢Õ¹᷹

¤×Í ¤ÇÒÁÂÒǢͧÅӴѺ¢éÍÁÙÅ

ºÔµ ÊÓËÃѺ ¢éÍÁÙÅ 1 à«¡àµÍÃì), áÅÐ

âk

¤×Í ¤èÒ »ÃÐÁÒ³¢Í§


76

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

nk ak

rk

H(D)

yk

ÃÙ»·Õ 4.9: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³

Viterbi detector

H(D)

time k

âk

¾ÃéÍÁǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

k+1

u

(u, q) q k-th stage

ÃÙ»·Õ 4.10: ¤Ó͸ԺÒÂá¼¹ÀÒ¾à·ÃÅÅÔÊ

¢éÍÁÙźԵÍÔ¹¾Øµ

ak

·Õ ä´é¨Ò¡¡ÒöʹÃËÑÊ¢éÍÁÙÅâ´Âǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¡è͹·Õ ¨Ð͸ԺÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ¨Ð¹ÔÂÒÁÊÑ­ÅÑ¡É³ì µÒÁÃÙ» ·Õ 4.10 ´Ñ§¹Õ ãËé

Ψk = [ak ak−1 . . . ak−ν+1 ]

³ àÇÅÒ

k ), Q = |A|ν

¢Í§¢éÍÁÙÅ ÍÔ¹¾Øµ,

ν

¤×Í Ê¶Ò¹Ð ³ àÇÅÒ

u

(ËÃ×ͤèÒ·Õ ÍÂÙèã¹àèÔÊàµÍÃìẺàÅ× Í¹·Ñ §ËÁ´

¤×Í ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´·Õ à» ¹ä»ä´é,

|A|

¤×Í Ë¹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³, ÃÐÂзÕ

ÊÒ¢Ò·Õ à» ¹ ä»ä´é ·Ñ §ËÁ´ ³ àÇÅÒ·Õ Ê¶Ò¹Ð

k

ä»ÂѧʶҹÐ

k,

áÅÐ

(u, q)

k (k th

stage) ¤×Í ¡ÅØèÁ ¢Í§àÊé¹

¤×Í ÊÑ­ÅÑ¡É³ì ·Õ ãªé á·¹¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡

q

¾Ô¨ÒóÒá¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ PR4 ¹Ñ ¹¤×Í ÁÕ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´

¤×Í ¨Ó¹Ç¹¤èÒ·Õ à» ¹ä»ä´é·Ñ §ËÁ´

Q = 22 = 4

H(D) = 1 − D2

µÒÁÃÙ»·Õ 4.11 «Ö §

ʶҹР·Õ áÊ´§´éÇÂÊÑ­Åѡɳì (0), (1), (2), áÅÐ (3) àÁ× Í


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

77

time k (0) -1 -1

k+1 0 2

(1) 1 -1

0 2 -2

(2) -1 1

λk (1, 2)

Φ + (2)

0

k 1

π k +1 (2)

-2

(3) 1 1

0

ak = 1 ak = -1

yk ÃÙ»·Õ 4.11: á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³ PR4,

¢éÍÁÙÅ ÍÔ¹¾Øµ

ak ∈ {−1, 1}

㹡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ÊÔ § ·Õ µéͧ¤Ó¹Ç³·Ø¡ ªèǧàÇÅÒ ¤×Í

¤èÒ àÁµÃÔ¡ ÊÒ¢Ò (branch metric) ³ àÇÅÒ

λk (u, q),

k

¢Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡Ê¶Ò¹Ð

k+1

¤èÒ àÁµÃÔ¡ àÊé¹·Ò§ (path metric) ³ àÇÅÒ

(predecessor) ÊÓËÃѺʶҹÐ

H(D) = 1 − D2

q

³ àÇÅÒ

k + 1, πk+1 (q),

·Õ ʶҹÐ

u

ä»Âѧ ʶҹÐ

q , Φk+1 (q),

áÅеÑǹÓ˹éÒ

«Ö §¨Ðà¡çº¤èÒʶҹÐàÃÔ Áµé¹·Õ à» ¹¼Å·ÓãËé

à¡Ô´àÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ (best transition) àªè¹ ¾Ô¨ÒÃ³Ò·Õ Ê¶Ò¹Ð (2) ³ àÇÅÒ ¨ÐÁÕàÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð 2 àÊé¹·Ò§ ¤×Í

(1, 2)

áÅÐ

(3, 2)

¡ÒÃàÅ×Í¡àÊé¹·Ò§à¾Õ§àÊé¹·Ò§à´ÕÂÇ·Õ ÁҶ֧ʶҹР(2) ³ àÇÅÒ àÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ ¨Ðä´éÇèÒ

πk+1 (2)

q,

k+1

ÊÁÁصÔÇèÒ ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ¨Ð·Ó

k+1

ÊÁÁصÔÇèÒ àÊé¹·Ò§

(1, 2)

¤×Í

= 1 ¹Ñ ¹àͧ

à» ¹ ·Õ ·ÃÒº¡Ñ¹ ÇèÒ Ç§¨ÃµÃǨËÒ·Õ ·ÓãËé ¤ÇÒÁ¹èÒ¨Ðà» ¹ ¢Í§¢éͼԴ¾ÅÒ´¢Í§ÅӴѺ ¢éÍÁÙÅ ·Ñ § ÅӴѺ ÁÕ ¤èÒ¹éÍÂ·Õ ÊØ´ ¤×Í Ç§¨ÃµÃǨËÒÅӴѺ·Õ ¤ÇèÐà» ¹ÁÒ¡ÊØ´ (MLSD: maximum likelihood sequence detector) «Ö §ÊÒÁÒöÊÃéÒ§ä´éâ´ÂãªéÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ ¨Ò¡áºº¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ»·Õ 4.9 ǧ¨Ã µÃǨËÒÇÕà·ÍÃìºÔ ¨ÐàÅ×Í¡ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

{ak }

·Õ ·ÓãËé ¤ÇÒÁ¹èÒ¨Ðà» ¹ ¢Í§ÅӴѺ ¢éÍÁÙÅ

{yk }

àÁ× Í


78

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡Ó˹´ÅӴѺ¢éÍÁÙÅ

{ak }

ÁÒãËé ¹Ñ ¹¤×Í

(

L+ν−1 1 X exp − 2 |yk − rk |2 2σ

1

p(y|a) = ³√ ´L+ν 2πσ 2

a

(4.7)

k=0

ÁÕ¤èÒÁÒ¡·Õ ÊØ´ ÊÁ¡Òà (4.7) ä´éÁÒ¨Ò¡¤ÇÒÁ¨ÃÔ§·Õ ÇèÒ ÊÑ­­Ò³Ãº¡Ç¹ ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

)

nk ∼ N (0, σ 2 ) áÅÐàÁ× Í¡Ó˹´

ÁÒãËé áÊ´§ÇèÒ Ãкº·ÃÒºÇèÒ ÅӴѺ ¢éÍÁÙÅ àÍÒµì¾Øµ

r

¤×Í ÍÐäà ´Ñ§¹Ñ ¹ ¢éÍÁÙÅ

yk

¨ÐÁÕ¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹ (Gaussian probability density function) àËÁ×͹¡Ñº

nk

·Õ ÁÕ¤èÒà©ÅÕ Âà·èҡѺ¤èÒ

rk

áÅФèÒ¤ÇÒÁá»Ã»Ãǹà·èҡѺ

σ2

àÁ× ÍãÊèÅÍ¡ÒÃÔ·ÖÁ¸ÃÃÁªÒµÔ (natural logarithm) ·Ñ §Êͧ¢éÒ§¢Í§ÊÁ¡Òà (4.7) ¨Ðä´éà» ¹

   ln {p(y|a)} = ln

  

1

³√ ´L+ν     2πσ 2

L+ν−1 1 X |yk − rk |2 2σ 2

(4.8)

k=0

Êѧࡵ¨Ð¾ºÇèÒ ¡Ò÷ÓãËé ÊÁ¡Òà (4.8) ÁÕ ¤èÒ ÁÒ¡·Õ ÊØ´ ÁÕ ¼Åà·Õºà·èÒ ¡Ñº ¡Ò÷ÓãËé ¾¨¹ì ·Õ Êͧ·Ò§´éÒ¹ ¢ÇÒÁ×Í ¢Í§ÊÁ¡Òà (4.8) ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ à¹× ͧ¨Ò¡ ¾¨¹ì ·Õ Ë¹Ö § à»ÃÕºàÊÁ×͹¡Ñº ¤èÒ¤§·Õ à¾ÃÒÐ©Ð¹Ñ ¹ ¡Ò÷ÓãËéÊÁ¡Òà (4.8) ÁÕ¤èÒÁÒ¡·Õ ÊØ´¨ÐÁÕ¤èÒà·èҡѺ¡Ò÷ÓãËéàÁµÃÔ¡ (metric)

L+ν−1 X

|yk − rk |2

(4.9)

k=0

ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ ´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨ÐàÅ×Í¡ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ ·ÓãËéÊÁ¡Òà (4.9) ÁÕ¤èÒ¹éÍ ·Õ ÊØ´ Êѧࡵ¨Ò¡ÊÁ¡Òà (4.9) ¨Ð¾ºÇèÒ ¾¨¹ì

|yk − rk |2

1

¡ç ¤×Í ¤èÒ ÃÐÂзҧ¡ÓÅѧ Êͧà©ÅÕ Â

(MSD:

mean squared distance) [10] àÁµÃÔ¡ã¹ÊÁ¡Òà (4.9) ÊÒÁÒö·ÓãËéÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ä´é â´Â¡Òäé¹ËÒ àÊé¹·Ò§ (path) ·Õ ÁÕ ¤èÒ àÁµÃÔ¡ ¹éÍÂ·Õ ÊØ´ µÒÁá¼¹ÀÒ¾à·ÃÅÅÔÊ àÁ× Í àÁµÃÔ¡ àÊé¹·Ò§ÁÕ ¤èÒ à·èÒ ¡Ñº ¼ÅÃÇÁ ¢Í§àÁµÃÔ¡ÊÒ¢Ò â´Â·Õ àÁµÃÔ¡ÊҢҢͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡Ê¶Ò¹Ð

λk (u, q) = |yk − r̂k (u, q)|2 1

ÈÖ¡ÉÒÃÒÂÅÐàÍÕ´à¾Ô ÁàµÔÁä´éã¹ËÑÇ¢éÍ·Õ 6.6.11 ã¹ [10]

u

ä»ÂѧʶҹÐ

q

¨Ð¹ÔÂÒÁâ´Â

(4.10)


4.3.

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

79

(A 1) ¡Ó˹´¤èÒàÃÔ Áµé¹¢Í§àÁµÃÔ¡àÊé¹·Ò§ (A 2) For

k = 0, 1, . . . , L + ν − 1

(A 3)

For

Φ0 (p) = 0

ÊÓËÃѺ·Ø¡¤èÒ

q = 0, 1, . . . , Q − 1

(A 4)

λk (p, q) = |yk − r̂(p, q)|2

(A 5)

πk+1 (q) = arg minp {Φk (p) + λk (p, q)}

(A 6)

Φk+1 (q) = Φk (πk+1 (q)) + λk (πk+1 (q), q)

(A 7)

Sk+1 (q) = [Sk (πk+1 (q)) | πk+1 (q)]

(A 8)

p

for

∀p

End

(A 9) End (A 10) ¶Í´ÃËÑÊ¢éÍÁÙÅÍÔ¹¾Øµ

¨Ò¡àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè·Õ ÁÕ¤èÒ

ΦL+ν

¹éÍÂ·Õ ÊØ´

ÃÙ»·Õ 4.12: ¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ

àÁ× Í

r̂k (u, q)

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³·Õ ÊÍ´¤Åéͧ¡Ñº

(u, q)

áÅÐàÁµÃÔ¡àÊé¹·Ò§ÊÒÁÒöËÒä´é

¨Ò¡

Φk+1 (q) =

k X

λi

(4.11)

i=0 ÃÙ» ·Õ 4.12 áÊ´§¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ¢Í§ªèͧÊÑ­­Ò³

H(D) = 1 − D2

µÑÇÍÂèÒ§àªè¹ ¨Ò¡á¼¹ÀÒ¾à·ÃÅÅÔÊ

ã¹ÃÙ»·Õ 4.11 ãËé¾Ô¨ÒóÒÃÐÂзÕ

k (k th

à·ÃÅÅÔÊ ¨Ð¾ºÇèÒ ÁÕ ¡ÒÃàÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð 2 àÊé¹·Ò§·Õ ÁÒ¶Ö§ ʶҹР¹Ñ ¹¤×Í

(1, 2)

λk (3, 2)

áÅÐ

(3, 2)

stage) ¢Í§á¼¹ÀÒ¾

(2)

³ àÇÅÒ

ãËé ·Ó¡Òäӹdz¤èÒ àÁµÃÔ¡ ÊÒ¢Ò·Ñ § 2 àÊé¹·Ò§ ¹Ñ ¹¤×Í

k+1

λk (1, 2)

áÅÐ

µÒÁ¢Ñ ¹µÍ¹·Õ (A 4) ¨Ò¡¹Ñ ¹ ʶҹÐàÃÔ Áµé¹·Õ ÊÍ´¤¤Åéͧ¡ÑºàÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ

·Õ ØÊØ´·Õ ÁҶ֧ʶҹР(2) ³ àÇÅÒ

k+1

¨Ð¶Ù¡àÅ×Í¡µÒÁ¢Ñ ¹µÍ¹·Õ (A 5) ÊÁÁصÔÇèÒ

(1, 2)

¤×Í àÊé¹·Ò§


80

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ØÊØ´·Õ ÁҶ֧ʶҹР(2) ³ àÇÅÒ ¹Ñ ¹ àÁµÃÔ¡àÊé¹·Ò§·Õ ÁҶ֧ʶҹР(2) ³ àÇÅÒ áÅÐàÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè

k+1

´Ñ§¹Ñ ¹¨Ðä´éÇèÒ

k + 1, Φk+1 (2),

πk+1 (2) = 1

ËÅѧ¨Ò¡

¨Ð¶Ù¡»ÃѺ¤èÒµÒÁ¢Ñ ¹µÍ¹·Õ (A 6)

(survivor path) ·Õ ÁÒ¶Ö§ ʶҹР(2) ³ àÇÅÒ

k + 1, Sk+1 (2),

¨Ð¶Ù¡

»ÃѺ¤èÒµÒÁ¢Ñ ¹µÍ¹·Õ (A 7) ãËé·ÓµÒÁ¢Ñ ¹µÍ¹µèÒ§æ àËÅèÒ¹Õ µÒÁÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ仨¹ÊÔ ¹ÊØ´ÅӴѺ ¢éÍÁÙÅ

{yk }

·Õ ä´éÃѺÁÒ áÅÐ¢Ñ ¹µÍ¹ÊØ´·éÒ¡ç¤×Í ¡ÒõѴÊԹ㨨ж١¡ÃзÓâ´Â¡ÒÃàÅ×Í¡àÊé¹·Ò§·Õ ÂѧÁÕ

ªÕÇÔµÍÂÙè·Õ ÁÕ¤èÒàÁµÃÔ¡àÊé¹·Ò§ ³ àÇÅÒ

4.3.4

L + ν , ΦL+ν ,

¹éÍÂ·Õ ÊØ´

¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

à¹× ͧ¨Ò¡ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·Ó¡ÒûÃÐÁÇżÅÅӴѺ ¢éÍÁÙÅ ·Ñ §ËÁ´¡è͹·Õ ¨ÐµÑ´ÊÔ¹ã¨ÇèÒ ÅӴѺ ¢éÍÁÙÅ·Õ ä´éÃѺ¤ÇèÐà» ¹ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµã´ÁÒ¡·Õ ÊØ´ ´Ñ§¹Ñ ¹ ¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ö§¢Ö ¹ÍÂÙè¡ÑºËÅÒ» ¨¨Ñ ´Ñ§¹Õ

1) ¨Ó¹Ç¹¤èÒ·Õ à» ¹ä»ä´é·Ñ §ËÁ´¢Í§¢éÍÁÙÅÍÔ¹¾Øµ

2) ¤ÇÒÁÂÒǢͧÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

3) ˹èǤÇÒÁ¨Ó¢Í§ªèͧ·ÒÃìà¡çµ

|A|

L

ν

ËÃ×ÍÍÒ¨¨ÐÊÃØ»ä´éÇèÒǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔẺ·Õ ãªé§Ò¹¡Ñ¹·Ñ Ç仨ÐÁÕ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´ã¹á¼¹ÀÒ¾ à·ÃÅÅÔÊà·èҡѺ ¹éÍÂà·èҡѺ

|A|ν

áÅеéͧ¡ÒÃãªé¨Ó¹Ç¹Ë¹èǤÇÒÁ¨Ó㹡ÒÃà¡çº¢éÍÁÙŵèÒ§æ àªè¹ ¤èÒ

(L + 1)|A|ν

{πk }

ÍÂèÒ§

˹èÇÂ

¨ÐàËç¹ä´éÇèÒ Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔẺ·Õ ãªé§Ò¹¡Ñ¹·Ñ Çä»äÁèÊÒÁÒö¹ÓÁÒãªé§Ò¹¨ÃÔ§ä´éã¹·Ò§»¯ÔºÑµÔ à¹× ͧ¨Ò¡µéͧ¡ÒÃ˹èǤÇÒÁ¨Óà» ¹¨Ó¹Ç¹ÁÒ¡ à¾ÃÒÐ©Ð¹Ñ ¹ÇÔ¸Õ¡ÒÃÅ´¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ ÇÕà·ÍÃìºÔ ¤×Í¡ÒÃãªé¾ÒÃÒÁÔàµÍÃì·Õ àÃÕ¡ÇèÒ ¤ÇÒÁÅÖ¡¡ÒöʹÃËÑÊ ÇÕà·ÍÃìºÔ àÁ× Í

T

dT

(decoding depth) ã¹ÍÑÅ¡ÍÃÔ·ÖÁ

¤×Í ¤ÒºàÇÅҢͧºÔµ (bit period) ¡ÅèÒǤ×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð·Ó¡ÒõѴÊÔ¹ã¨

¢éÍÁÙÅ ·ÕÅкԵ ËÅѧ¨Ò¡·Õ àÇÅÒ¼èÒ¹ä»

dT

˹èÇ ´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Õ ÁÕ ¡ÒÃãªé ¤ÇÒÁÅÖ¡ ¡ÒÃ

¶Í´ÃËÑʨеéͧ¡Òèӹǹ˹èǤÇÒÁ¨Ó㹡ÒÃà¡çº¢éÍÁÙŵèÒ§æ ÍÂèÒ§¹éÍÂà·èҡѺ ·Ñ Çä»áÅéÇÁÑ¡¨Ðãªé

d ≥ 5(ν + 1)

[27]

(d+1)|A|ν

«Ö §â´Â


4.4.

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

81

0

(a) 0

0

1

ak = 1

0.5

(b) 1

ak = 0

1

1.5

ÃÙ»·Õ 4.13: á¼¹ÀÒ¾à·ÃÅÅÔʢͧªèͧÊÑ­­Ò³

4.4

H(D) = 1 + 0.5D

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ã¹Êèǹ¹Õ ¨ÐáÊ´§µÑÇÍÂèÒ§¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÍÂèÒ§ÅÐàÍÕ´ ´Ñ§µèÍ仹Õ

µÑÇÍÂèÒ§·Õ 4.2

{0,

¨Ò¡áºº¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ» ·Õ 4.9 ¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

0, 1}, ªèͧÊÑ­­Ò³

hk = δk + 0.5δk−1

delta function), ÊÑ­­Ò³Ãº¡Ç¹ ¢éÍÁÙÅ

ÇÔ¸Õ·Ó

{yk }

{nk }

=

àÁ× Í

{0.2,

δk

{ak }

=

¤×Í ¿ §¡ìªÑ¹â¤Ã๤à¡ÍÃìà´ÅµÒ (Kronecker

0.5, 0,

−0.35}

¨§áÊ´§¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ

´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¨Ò¡·Õ ⨷Âì¡Ó˹´ ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

rk

ËÒä´é¨Ò¡

rk = ak ∗ hk = {r0 , r1 , r2 , r3 } = {0, 0, 1, 0.5} àÁ× Í

¤×Í µÑÇ´Óà¹Ô¹¡Òä͹âÇÅ٪ѹ (convolution operator) áÅÐ

yk = rk + nk = {0.2, 0.5, 1, 0.15} ¨Ò¡¹Ñ ¹ ãËéÊÃéҧἹÀÒ¾à·ÃÅÅÔʨҡªèͧÊÑ­­Ò³

hk = δk + 0.5δk−1

¹Ñ ¹¤×Í

H(D) = 1 + 0.5D

«Ö § ¨Ð ä´é µÒÁ ÃÙ» ·Õ 4.13 â´Â ã¹ ·Õ ¹Õ á¼¹ÀÒ¾ à· ÃÅÅÔ ÊÁÕ ·Ñ §ËÁ´ 2 ʶҹР¤×Í Ê¶Ò¹Ð (a) áÅРʶҹР(b) ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ¢éÍÁÙÅÊÒÁÒöáºè§ÍÍ¡à» ¹ªèǧàÇÅÒµèÒ§æ ´Ñ§áÊ´§ã¹ÃÙ»·Õ 4.14 «Ö § ÁÕÃÒÂÅÐàÍÕ´´Ñ§¹Õ


82

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0

0.04

Φ1 ( a ) = 0.04

0.64 0.09 0

1.69

Φ1 ( b ) = 0.64 0.04

0.25

Φ ( a ) = 0.29 2

0.25 0 0.64

1

Φ ( b ) = 0.29 2

0.29

1

0 0.25 0.29

0.25

Φ3 ( a ) = 0.54

Φ3 ( b ) = 0.29 0.54

0.02

Φ 4 ( a ) = 0.41

0.72 0.12 0.29

y 0 = 0.2

y1 = 0.5

y2 = 1

1.82

Φ 4 ( b ) = 1.26

y3 = 0.15

ÃÙ»·Õ 4.14: á¼¹ÀҾ͸ԺÒ¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔã¹áµèÅЪèǧàÇÅÒ


4.4.

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ªèǧàÇÅÒ·Õ 0

àÁ× Í àÃÔ Áµé¹ ÃѺ ¢éÍÁÙÅ

y0 = 0.2

83

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·Ó¡ÒáÓ˹´¤èÒ àÃÔ Áµé¹ ¢Í§

àÁµÃÔ¡àÊé¹·Ò§ãËéà·èҡѺ¤èÒÈÙ¹Âì µÒÁ¢Ñ ¹µÍ¹·Õ (A 1) ã¹ÃÙ»·Õ 4.12 ¹Ñ ¹¤×Í

Φ0 (a) = 0

áÅÐ

Φ0 (b) = 0

¨Ò¡¹Ñ ¹ ¡ç¨Ð·Ó¡ÒäӹdzàÁµÃÔ¡ÊÒ¢Ò·Ø¡àÊé¹ÊÒ¢ÒµÒÁ¢Ñ ¹µÍ¹·Õ (A 4) ã¹ÃÙ»·Õ 4.12 ´Ñ§¹Õ

λ0 (a, a) = |0.2 − 0|2 = 0.04 λ0 (a, b) = |0.2 − 1|2 = 0.64 λ0 (b, a) = |0.2 − 0.5|2 = 0.09 λ0 (b, b) = |0.2 − 1.5|2 = 1.69 µÒÁ·Õ à¢Õ¹áÊ´§äÇéã¹áµèÅÐàÊé¹ÊÒ¢Òã¹ÃÙ»·Õ 4.14 ¨Ò¡¹Ñ ¹ ·Õ áµèÅШشµèÍ¡ç¨Ð·Ó¡ÒÃàÅ×Í¡àÊé¹·Ò§¡Òà à»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ áÅéÇ¡ç·Ó¡ÒûÃѺ¤èÒàÁµÃÔ¡àÊé¹·Ò§ µÒÁ¢Ñ ¹µÍ¹·Õ (A 5) áÅÐ (A 6) ã¹ÃÙ»·Õ 4.12 ¹Ñ ¹¤×Í

Φ1 (a) = min{0 + 0.04, 0 + 0.09} = 0.04 Φ1 (b) = min{0 + 0.64, 0 + 1.69} = 0.64 µÒÁ·Õ à¢Õ¹áÊ´§äÇé㹨شµèÍã¹ÃÙ»·Õ 4.14 â´Â·Õ àÊé¹ÅÙ¡ÈÃÊÕ´Ó ¤×Í àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè (survivor path) ÊèǹàÊé¹ÅÙ¡ÈÃÊÕà·Ò ¤×Í àÊé¹·Ò§·Õ ¶Ù¡µÑ´·Ô §

ªèǧàÇÅÒ·Õ 1

àÁ× Í ÃѺ ¢éÍÁÙÅ

y1 = 0.5

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·Ó¡ÒäӹdzàÁµÃÔ¡ ÊÒ¢Ò·Ñ §ËÁ´

´Ñ§¹Õ

λ1 (a, a) = |0.5 − 0|2 = 0.25 λ1 (a, b) = |0.5 − 1|2 = 0.25 λ1 (b, a) = |0.5 − 0.5|2 = 0 λ1 (b, b) = |0.5 − 1.5|2 = 1


84

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

µÒÁ·Õ à¢Õ¹áÊ´§äÇéã¹áµèÅÐàÊé¹ÊÒ¢Òã¹ÃÙ»·Õ 4.14 ¨Ò¡¹Ñ ¹ ·Õ áµèÅШشµèÍ¡ç¨Ð·Ó¡ÒÃàÅ×Í¡àÊé¹·Ò§¡Òà à»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ áÅéÇ¡ç·Ó¡ÒûÃѺ¤èÒàÁµÃÔ¡àÊé¹·Ò§ ¹Ñ ¹¤×Í

Φ2 (a) = min{0.04 + 0.25, 0.64 + 0} = 0.29 Φ2 (b) = min{0.04 + 0.25, 0.64 + 1} = 0.29 µÒÁ·Õ à¢Õ¹áÊ´§äÇé㹨شµèÍã¹ÃÙ»·Õ 4.14

ªèǧàÇÅÒ·Õ 2

àÁ× ÍÃѺ¢éÍÁÙÅ

y2 = 1

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð·Ó¡ÒäӹdzàÁµÃÔ¡ÊÒ¢Ò·Ñ §ËÁ´ ´Ñ§¹Õ

λ2 (a, a) = |1 − 0|2 = 1 λ2 (a, b) = |1 − 1|2 = 0 λ2 (b, a) = |1 − 0.5|2 = 0.25 λ2 (b, b) = |1 − 1.5|2 = 0.25 µÒÁ·Õ à¢Õ¹áÊ´§äÇéã¹áµèÅÐàÊé¹ÊÒ¢Òã¹ÃÙ»·Õ 4.14 ¨Ò¡¹Ñ ¹ ·Õ áµèÅШشµèÍ¡ç¨Ð·Ó¡ÒÃàÅ×Í¡àÊé¹·Ò§¡Òà à»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ áÅéÇ¡ç·Ó¡ÒûÃѺ¤èÒàÁµÃÔ¡àÊé¹·Ò§ ¹Ñ ¹¤×Í

Φ3 (a) = min{0.29 + 1, 0.29 + 0.25} = 0.54 Φ3 (b) = min{0.29 + 0, 0.29 + 0.25} = 0.29 µÒÁ·Õ à¢Õ¹áÊ´§äÇé㹨شµèÍã¹ÃÙ»·Õ 4.14

ªèǧàÇÅÒ·Õ 3

àÁ× Í ÃѺ ¢éÍÁÙÅ

y3 = 0.15

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·Ó¡ÒäӹdzàÁµÃÔ¡ ÊÒ¢Ò·Ñ §ËÁ´

´Ñ§¹Õ

λ3 (a, a) = |0.15 − 0|2 = 0.02 λ3 (a, b) = |0.15 − 1|2 = 0.72 λ3 (b, a) = |0.15 − 0.5|2 = 0.12 λ3 (b, b) = |0.15 − 1.5|2 = 1.82


4.4.

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

85

µÒÁ·Õ à¢Õ¹áÊ´§äÇéã¹áµèÅÐàÊé¹ÊÒ¢Òã¹ÃÙ»·Õ 4.14 ¨Ò¡¹Ñ ¹ ·Õ áµèÅШشµèÍ¡ç¨Ð·Ó¡ÒÃàÅ×Í¡àÊé¹·Ò§¡Òà à»ÅÕ Â¹Ê¶Ò¹Ð·Õ ´Õ·Õ ÊØ´ áÅéÇ¡ç·Ó¡ÒûÃѺ¤èÒàÁµÃÔ¡àÊé¹·Ò§ ¹Ñ ¹¤×Í

Φ4 (a) = min{0.54 + 0.02, 0.29 + 0.12} = 0.41 Φ4 (b) = min{0.54 + 0.72, 0.29 + 1.82} = 1.26 µÒÁ·Õ à¢Õ¹áÊ´§äÇé㹨شµèÍã¹ÃÙ»·Õ 4.14

ËÅѧ¨Ò¡·ÓµÒÁ¢Ñ ¹µÍ¹¢Í§ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ¨¹ÊÔ ¹ÊØ´ÅӴѺ¢éÍÁÙÅ·Õ ä´éÃѺ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¡ç ¨Ð·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ÍÔ¹¾Øµ ¨Ðä´éÇèÒ

Φ4 (a) = 0.41

¨Ò¡àÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè ·Õ ÁÕ ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§¹éÍÂ·Õ ÊØ´ ã¹·Õ ¹Õ

ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ à¾ÃÒÐ©Ð¹Ñ ¹ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨ÐÁͧÂé͹¡ÅÑºä» (trace

back) µÒÁàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè·Õ ÁÒ¶Ö§¨Ø´µèÍ

Φ4 (a)

¡ç¨Ð¾ºÇèÒ ¢éÍÁÙÅÍÔ¹¾Øµ

{âk }

·Õ ÊÍ´¤Åéͧ¡Ñº

àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè¹Õ ¤×Í

{âk } = {â0 , â1 , â2 } = {0, 0, 1} «Ö §µÃ§¡Ñº¢éÍÁÙÅÍÔ¹¾Øµ

{ak }

·Õ Êè§ÁÒ¨Ò¡µé¹·Ò§¨ÃÔ§ áÊ´§ÇèÒ¡ÒöʹÃËÑÊ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¹Õ

äÁèÁÕ¢éͼԴ¾ÅÒ´à¡Ô´¢Ö ¹ã¹Ãкº ËÁÒÂà˵Ø

−1)

¢éÍÁÙÅ ÍÔ¹¾Øµ µÑÇ ÊØ´·éÒÂ·Õ ÊÒÁÒö¶Í´ÃËÑÊ ä´é ¨Ò¡Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ (ã¹·Õ ¹Õ ¤×Í

a3 =

äÁè¨Óà» ¹µéͧ¶Í´ÃËÑÊ à¾ÃÒж×ÍÇèÒ à» ¹¢éÍÁÙźԵÊèǹà¡Ô¹·Õ ä´é¨Ò¡¡Ò÷Ӥ͹âÇÅ٪ѹ¢Í§¢éÍÁÙÅ

ÍÔ¹¾Øµ ¡Ñº ªèͧÊÑ­­Ò³ áµè ¢éÍÁÙÅ

y3

à» ¹ ¢éÍÁÙÅ ·Õ ¨Óà» ¹ ·Õ ¨Ðµéͧ¹ÓÁÒãªé 㹡ÒöʹÃËÑÊ ¢éÍÁÙÅ ¢Í§

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

µÑÇÍÂèÒ§·Õ 4.3

{−1, −1,

1,

¨Ò¡áºº¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ» ·Õ 4.9 ¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

−1},

ÊÑ­­Ò³Ãº¡Ç¹

{nk }

ªèͧÊÑ­­Ò³ =

hk = δk − δk−1

{0.5, −0.4,

0.1, 0.7,

¡) ÇÒ´á¼¹ÀÒ¾ºÅçÍ¡¢Í§ªèͧÊÑ­­Ò³

−0.3}

àÁ× Í

δk

{ak }

=

¤×Í ¿ §¡ìªÑ¹ â¤Ã๤à¡ÍÃìà´ÅµÒ,

¨§

hk

¢) à¢Õ¹ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ì ÃÐËÇèÒ§¢éÍÁÙÅ ÍÔ¹¾Øµ áÅТéÍÁÙÅ àÍÒµì¾Øµ ¢Í§ªèͧÊÑ­­Ò³ áÅÐ ¤Ó¹Ç³ËÒ¤èÒ

{yk }


86

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

rk ak

0 2

-1

D

(a) Block diagram

-2

1

1/2

--1 1-

-1/0

-1

111

1

0

ak = 1 ak = -1

1/0

(c) Trellis diagram -1/-2

(b) FSM ÃÙ»·Õ 4.15: (a) á¼¹ÀÒ¾ºÅçÍ¡, (b) à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, áÅÐ (c) á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§ªèͧÊÑ­­Ò³

H(D) = 1 − D

¤) ÊÃéÒ§à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ áÅÐÇÒ´á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§ªèͧÊÑ­­Ò³

§) ¶Í´ÃËÑÊ¢éÍÁÙÅ

{yk }

hk

´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ÇÔ¸Õ·Ó ¡)

¨Ò¡ªèͧÊÑ­­Ò³·Õ ¡Ó˹´ÁÒãËé

hk = δk − δk−1

¹Ñ ¹¤×Í

H(D) = 1 − D

ÊÒÁÒöáÊ´§à» ¹

á¼¹ÀÒ¾ºÅçÍ¡¢Í§ªèͧÊÑ­­Ò³ä´éµÒÁÃÙ»·Õ 4.15(a) ¢)

à¹× ͧ¨Ò¡

àÍÒµì¾Øµ

rk

H(D) = 1 − D

´Ñ§¹Ñ ¹ ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾Øµ

ak

áÅТéÍÁÙÅ

¢Í§ªèͧÊÑ­­Ò³¹Õ ¤×Í

rk = ak − ak−1 «Ö §¨Ðä´éÇèÒ

rk

=

{−1,

0, 2,

−2, 1}

´Ñ§¹Ñ ¹

yk = rk + nk

=

{−0.5, −0.4,

2.1,

−1.3,

0.7}


4.4.

µÑÇÍÂèÒ§¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

0

-1

0.25

0.25

0.41

0.16

87

4.41

0.91

4.82 0.49

1

0.25

0.25

0.16

0.41

0.42

1.69

1.40

1.69

0.01

0

0.49

2.11

0.49

2.60

ÃÙ»·Õ 4.16: á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¤)

¨Ò¡ÊÁ¡ÒÃáÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾ØµáÅТéÍÁÙÅàÍÒµì¾Øµ·Õ ä´éã¹¢éÍ ¢) à¤Ã× Í§Ê¶Ò¹Ð

¨Ó¡Ñ´áÅÐá¼¹ÀÒ¾à·ÃÅÅÔÊ ÊÒÁÒöáÊ´§ä´éµÒÁÃÙ»·Õ 4.15(b) áÅÐ 4.15(c) µÒÁÅӴѺ §)

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔãªéá¼¹ÀÒ¾à·ÃÅÅÔÊ㹡ÒöʹÃËÑÊ¢éÍÁÙÅ

{yk } â´Â·Õ ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ

¢éÍÁÙÅ ÊÒÁÒöÊÃØ» ä´é µÒÁÃÙ» ·Õ 4.16 àÁ× Í µÑÇàÅ¢·Õ áÊ´§ÍÂÙè º¹¨Ø´µèÍ áµèÅШش ¤×Í ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§ ·Õ ÁÒ¶Ö§ ³ ¨Ø´µèÍ ¹Ñ ¹ áÅеÑÇàÅ¢·Õ áÊ´§ÍÂÙè º¹àÊé¹ ÊÒ¢ÒáµèÅÐàÊé¹ ¤×Í ¤èÒ àÁµÃÔ¡ ÊҢҢͧáµèÅÐàÊé¹ ÊÒ¢Ò·Õ ´Õ ·Õ ÊØ´ ·Õ ÁÒ¶Ö§ ·Õ ¨Ø´µèÍ ¹Ñ ¹æ ¨Ò¡ÃÙ» ·Õ 4.16 ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§·Õ ¹éÍÂ·Õ ÊØ´ ¤×Í ¤èÒ 1.40 ´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð¶Í´ÃËÑÊ¢éÍÁÙÅâ´Â¡ÒÃÁͧÂé͹¡ÅѺ仵ÒÁàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè·Õ ÁÒ¶Ö§¨Ø´µèÍ ·Õ ÁÕ¤èÒàÁµÃÔ¡àÊé¹·Ò§à·èҡѺ 1.40 «Ö §¨Ð¾ºÇèÒ ¤èÒ»ÃÐÁÒ³¢Í§ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{âk }

·Õ ÊÍ´¤Åéͧ

¡ÑºàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè¹Õ ¤×Í

{âk } = {â0 , â1 , â2 , â3 } = {−1, −1, 1, −1} «Ö §ÁÕ¤èҵç¡ÑºÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{ak }

·Õ Êè§ÁҨҡǧ¨ÃÀÒ¤Êè§ à¾ÃÒÐ©Ð¹Ñ ¹ ¡ÒöʹÃËÑÊ´éÇÂǧ¨Ã

µÃǨËÒÇÕà·ÍÃìºÔã¹µÑÇÍÂèÒ§¢éÍ¹Õ ¨Ö§äÁèÁÕ¢éͼԴ¾ÅÒ´à¡Ô´¢Ö ¹

4.4.1

ÊÃػǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·Ó§Ò¹ä´é ÍÂèÒ§ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´ ¡ç µèÍàÁ× Í Í§¤ì»ÃСͺ¢Í§ÊÑ­­Ò³ ú¡Ç¹·Õ ὧÍÂÙèã¹¢éÍÁÙÅ·Õ ¨Ð·Ó¡ÒöʹÃËÑÊ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹ à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN) [15] ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ä´é ÁÒ¨Ò¡ÊÁÁص԰ҹ·Õ ÇèÒ


88

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

µÒÃÒ§·Õ 4.2: µÑÇÍÂèÒ§áÊ´§¨Ó¹Ç¹Ê¶Ò¹Ð·Õ µéͧãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔʢͧ·ÒÃìà¡çµáººµèÒ§æ

H(D)

·ÒÃìà¡çµáºº PR PR4

[1 0

EPR4 EEPR4

−1]

[1 1

−1 −1]

[1 2 0

−2 −1]

˹èǤÇÒÁ¨Ó

ν

¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´

(1 − D)(1 + D)

2

22 = 4

(1 − D)(1 + D)2

3

23 = 8

(1 − D)(1 + D)3

4

24 = 16

ͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔà» ¹ÊÑ­­Ò³Ãº¡Ç¹áººà¡ÒÊì ÊÕ¢ÒÇẺºÇ¡ ˹éÒ·Õ ËÅÑ¡ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¤×Í ¨Ð·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ â´ÂãËé ÁÕ ¤èÒ¤ÇÒÁ¹èÒ¨Ðà» ¹ ¢Í§ ¢éÍ ¼Ô´¾ÅÒ´ÅӴѺ (probability of sequence error) ¹éÍÂ·Õ ÊØ´ áÅÐâ´Â·Ñ Ç令ÇÒÁ«Ñº«é͹¢Í§Ç§¨Ã µÃǨËÒÇÕà·ÍÃìºÔ¨Ð¢Ö ¹ÍÂÙè¡Ñº¨Ó¹Ç¹Ê¶Ò¹Ð·Õ ãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ ´Ñ§áÊ´§ã¹µÒÃÒ§·Õ 4.2 ¨Ð¾ºÇèÒ àÁ× Í ·ÒÃìà¡çµ ·Õ ãªé ÁÕ ¨Ó¹Ç¹á·ç» (ËÃ×Í Ë¹èǤÇÒÁ¨Ó) ÁÒ¡¢Ö ¹ ¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¡ç ¨ÐÁÒ¡¢Ö ¹ à¹× ͧ¨Ò¡ ¨Ó¹Ç¹Ê¶Ò¹Ð·Õ ãªé ã¹á¼¹ÀÒ¾à·ÃÅÅÔ ÊÁÕ ÁÒ¡¢Ö ¹ ÍÂèÒ§äáçµÒÁ àÁ× Í ¤ÇÒÁ¨Ø ¢éÍÁÙŢͧÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÙ§¢Ö ¹ ·ÒÃìà¡çµ·Õ ¨Ð¹ÓÁÒãªé¡ç¤ÇÃ·Õ ¨ÐµéͧÁըӹǹá·ç»ÁÒ¡¢Ö ¹ à¾× Í·ÓãËé¼Å µÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§·ÒÃìà¡çµ ÊÍ´¤Åéͧ¡Ñº ¼ÅµÍºÊ¹Í§àªÔ§ ¤ÇÒÁ¶Õ ¢Í§ªèͧÊÑ­­Ò³ ´Ñ§¹Ñ ¹ã¹ ¡ÒþԨÒóÒÇèÒ ¨Ð¹Ó·ÒÃìà¡çµ ·Õ ÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡ÁÒãªé §Ò¹ËÃ×Í äÁè ¹Ñ ¹ ¨Ðµéͧ»ÃйջÃйÍÁÃÐËÇèÒ§ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ¨Ðä´éÃѺ áÅФÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

4.5

ÊÃØ»·éÒº·

à·¤¹Ô¤ PRML à» ¹ ¡ÒÃãªé §Ò¹ÃèÇÁ¡Ñ¹ ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃì Ẻ PR áÅÐǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ «Ö § à» ¹·Õ ¹ÔÂÁãªé§Ò¹¡Ñ¹ÁÒ¡ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ÊÒà˵ØË¹Ö §ÍÒ¨¨Ðà» ¹ à¾ÃÒÐÇèÒǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¶×Íä´éÇèÒà» ¹Ç§¨ÃµÃǨËÒ¢éÍÁÙÅ·Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡ â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í Í§¤ì »ÃСͺ¢Í§ÊÑ­ ­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÁÕ ÅѡɳÐà» ¹ ÊÑ­­Ò³ ú¡Ç¹áººà¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ ËÅÑ¡ ¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨ÐÍÂÙè º¹¾× ¹ °Ò¹¢Í§á¼¹ ÀÒ¾à·ÃÅÅÔÊ«Ö §ÍÒ¨¨ÐÊÃéÒ§ä´é¨Ò¡à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ áÅÐâ´Â·Ñ Çä»áÅéǤÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ


4.6.

à຺½ ¡ËÑ´·éÒº·

89

ÇÕà·ÍÃìºÔ¨Ð¢Ö ¹ÍÂÙè¡Ñº¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´·Õ µéͧãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ «Ö §¨Ó¹Ç¹Ê¶Ò¹Ð·Õ ãªéã¹á¼¹ ÀÒ¾à·ÃÅÅÔÊ¨Ð¢Ö ¹ÍÂÙè¡Ñº ¨Ó¹Ç¹¤èÒ·Õ à» ¹ä»ä´é·Ñ §ËÁ´¢Í§¢éÍÁÙÅÍÔ¹¾Øµ áÅÐ˹èǤÇÒÁ¨Ó¢Í§·ÒÃìà¡çµ ´Ñ§¹Ñ ¹ ¡ÒèйӷÒÃìà¡çµ ã´ÁÒãªé §Ò¹ã¹ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ¨Ðµéͧ»ÃйջÃйÍÁÃÐËÇèÒ§»ÃÐÊÔ·¸ÔÀÒ¾ ¢Í§Ãкº·Õ ¨Ðä´éÃѺ áÅФÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

4.6

à຺½ ¡ËÑ´·éÒº·

1. ¨§Í¸ÔºÒ¢éÍᵡµèÒ§ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃìẺ¼ÅµÍºÊ¹Í§àµçÁ (full response) áÅÐẺ¼Å µÍºÊ¹Í§ºÒ§Êèǹ (partial response) 2. ¡Ó˹´ãËé¢éÍÁÙÅÍÔ¹¾Øµ

ak ∈ {−1, 1}

¨§ÇÒ´á¼¹ÀÒ¾ºÅçÍ¡ (block diagram), à¤Ã× Í§Ê¶Ò¹Ð

¨Ó¡Ñ´ (FSM), áÅÐá¼¹ÀÒ¾à·ÃÅÅÔÊ (Trellis diagram) ¢Í§ªèͧÊÑ­­Ò³ 2.1)

H(D) = 1 − 0.5D

2.2)

H(D) = 1 + 2D + D2

2.3)

H(D) = 1 − D3

2.4)

H(D) = 1 + 3D + 3D2 + D3

3. ¨Ò¡áºº¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ» ·Õ 4.9 ¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

−1,

1} ¶Ù¡ Êè§ ¼èÒ¹ªèͧÊÑ­­Ò³

−0.4,

0.1, 0.7,

−0.3,

H(D)

H(D)

{ak }

«Ö § ¶Ù¡ ú¡Ç¹´éÇÂÊÑ­­Ò³Ãº¡Ç¹

0.4} ¨§¶Í´ÃËÑÊ¢éÍÁÙÅ

{yk }

µèÍ仹Õ

{1, −1,

=

{nk }

=

{0.5,

â´Âãªéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÓËÃѺªèͧ

ÊÑ­­Ò³µèÍä»¹Õ 3.1)

H(D) = 1 + 2D + D2

3.2)

H(D) = 1 − D2

4. ¨Ò¡áºº¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ» ·Õ 4.9 ¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ 0} ¶Ù¡Ê觼èÒ¹ªèͧÊÑ­­Ò³ 0.2, 0.6,

−0.4, −0.3,

ÊÑ­­Ò³µèÍ仹Õ

H(D)

«Ö §¶Ù¡Ãº¡Ç¹´éÇÂÊÑ­­Ò³Ãº¡Ç¹

0.5} ¨§¶Í´ÃËÑÊ¢éÍÁÙÅ

{yk }

{ak }

{nk }

=

=

{1,

0, 1,

{0.3, −0.5,

â´Âãªéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÓËÃѺªèͧ


90

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

4.1)

H(D) = 1 + 3D + 3D2 + D3

4.2)

H(D) = 1 + D − D2 − D3


º··Õ 5

¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´

㹺·¹Õ ¨Ð͸ԺÒ¶֧ ËÅÑ¡¡ÒÃÇÔà¤ÃÒÐËì à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´ (error event) «Ö § à» ¹ ¡Òäé¹ËÒÃٻẺ ¢Í§¢éͼԴ¾ÅÒ´·Õ à¡Ô´ ¢Ö ¹ ºèÍ ã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ áÅÐ àÁ× Í ÊÒÁÒöÇèÒ ¢éͼԴ¾ÅÒ´ÃÙ»áººã´·Õ à¡Ô´ ¢Ö ¹ ºèÍ ¹Ñ¡Í͡ẺÃкº¡ç ÊÒÁÒö·Õ ¨Ð·Ó¡ÒÃÍ͡Ẻ ÃËÑÊ RLL (run length limited)

[9] ËÃ×Í Ç§¨Ãà¢éÒÃËÑÊ¡è͹

(precoder) [47] à¾× Íãªéà¢éÒÃËÑÊ¢éÍÁÙÅ

¢èÒÇÊÒáè͹·Õ ¨Ð·Ó¡ÒÃà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡ à¾× ÍËÅÕ¡àÅÕ Â§¡ÒÃà¡Ô´¢éͼԴ¾ÅÒ´àËÅèÒ¹Ñ ¹ÃÐËÇèÒ§¡Òà ¶Í´ÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ «Ö § ¨ÐÊè§ ¼Å·ÓãËé »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºã¹ÃÙ» ¢Í§ÍѵÃÒ¢éÍ ¼Ô´¾ÅÒ´ºÔµ (BER) ¹éÍÂŧÁÒ¡ ¹Í¡¨Ò¡¹Õ ¶éÒ·ÃÒºÇèÒ¢éͼԴ¾ÅÒ´ÃÙ»áººã´·Õ à¡Ô´¢Ö ¹ºèÍÂã¹Ãкº ¡çÊÒÁÒö·Õ ¨Ð¹Ó¢éÍÁÙÅ¹Õ ÁÒãªé㹡ÒÃÇÔà¤ÃÒÐËìáÅÐà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§·ÒÃìà¡çµáººµèÒ§æ ä´é â´Âãªé ¾ÒÃÒÁÔàµÍÃì ·Õ àÃÕ¡ÇèÒ SNR »ÃÐÊÔ·¸Ô¼Å (e ective SNR) [48] «Ö § ÁÕ ¼ÅÅѾ¸ì à·Õºà·èÒ ¡Ñº ¡Òà ãªé¾ÒÃÒÁÔàµÍÃì BER 㹡ÒÃà»ÃÕºà·Õº áµèãªéàÇÅÒ㹡ÒäӹdzËÒ¤èÒ SNR »ÃÐÊÔ·¸Ô¼Å¹Õ ¹éÍ¡ÇèÒ ÁÒ¡ ´Ñ§ÃÒÂÅÐàÍÕ´·Õ ¨Ð͸ԺÒµèÍä»ã¹º·¹Õ

5.1

º·¹Ó

µÒÁ·Õ ͸ԺÒÂã¹ËÑÇ¢éÍ·Õ 4.3 ¾ºÇèÒǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð·Ó§Ò¹ÍÂÙ躹¾× ¹°Ò¹¢Í§á¼¹ÀÒ¾à·ÃÅÅÔÊ (trellis diagram) «Ö §ã¹¡ÒöʹÃËÑÊ¢éÍÁÙźҧ¤ÃÑ § ÁÕ¤ÇÒÁà» ¹ä»ä´é·Õ ¨Ðà¡Ô´à˵ءÒóìµÒÁÃÙ»·Õ 5.1 91


92

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

B

A

C

E

D F #1

-1

1

-1

1

-1

-1

1

1

#2

-1

1

1

-1

1

-1

1

1

Error 0 sequence

0

-2

2

-2

0

0

0

ÃÙ»·Õ 5.1: µÑÇÍÂèÒ§¡Ò÷ӧҹÀÒÂã¹à·ÃÅÅÔʢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¹Ñ ¹¤×Í ã¹¢³Ð·Õ ·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ¼èҹἹÀÒ¾à·ÃÅÅÔÊ ¨¹ÊÔ ¹ÊØ´ ÅӴѺ ¢éÍÁÙÅ ·Õ ä´é ÃѺ »ÃÒ¡®ÇèÒ àÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè (survivor path) ·Õ ´Õ ·Õ ÊØ´ à» ¹ 仵ÒÁÃÙ» ·Õ 5.1 ¡ÅèÒǤ×Í àÁ× Í ¤Ó¹Ç³ËÒ¤èÒ àÁµÃÔ¡àÊé¹·Ò§ (path metric) ·Õ ¨Ø´ E à¾× Í·Ó¡ÒÃàÅ×Í¡àÊé¹·Ò§·Õ ´Õ·Õ ÊØ´·Õ ÁÒ¶Ö§¨Ø´ E ¾ºÇèÒÁÕàÊé¹·Ò§ 2 àÊé¹·Ò§ ¤×Í àÊé¹·Ò§ #1 (àÊé¹·Ò§ ABCEF) áÅÐàÊé¹·Ò§ #2 (àÊé¹·Ò§ ABDEF) ·Õ ÁÕ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§ ³ ¨Ø´ E à·èҡѹ (àÊé¹·Ò§·Ñ §ÊͧÁÕ¤ÇÒÁà» ¹ä»ä´éà·èҡѹ·Õ ¨Ð¶Ù¡àÅ×Í¡â´ÂÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ) à¾ÃÒÐ©Ð¹Ñ ¹ ã¹¡Ã³Õ ¹Õ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÒÁÒö·Õ ¨ÐµÑ´ÊÔ¹ã¨àÅ×Í¡àÊ鹷ҧ㴡çä´é à¾× Í ãªé à» ¹ ÊèÇ¹Ë¹Ö § ¢Í§àÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè ·Õ ´Õ ·Õ ÊØ´ ÊÓËÃѺ ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ µÑÇÍÂèÒ§àªè¹ ¶éÒ Ç§¨ÃµÃǨËÒ ÇÕà·ÍÃìºÔàÅ×Í¡àÊé¹·Ò§ #1 ¡ç¨Ðä´é¶Í´ÃËÑÊ¢éÍÁÙÅä´éà» ¹ àÊé¹·Ò§ #2 ¡ç¨Ðä´é¼ÅÅѾ¸ìà» ¹ ¨ÃÔ§ ¤×Í

{−1,

¢éÍÁÙÅà» ¹

1,

{−1,

−1, 1, 1,

1,

{−1,

−1, −1,

−1,

1,

−1,

1, 1,

−1,

1,

{−1,

−1,

1,

−1,

1,

−1, −1,

1, 1} áµè¶éÒàÅ×Í¡

1, 1} ´Ñ§¹Ñ ¹¶éÒÊÁÁصÔÇèÒ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

1, 1} ¡çËÁÒ¤ÇÒÁÇèÒ âÍ¡ÒÊ·Õ Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð¶Í´ÃËÑÊ 1, 1} ¡çÁÕ¤ÇÒÁà» ¹ä»ä´éÊÙ§ (ËÃ×Íã¹·Ò§¡ÅѺ¡Ñ¹) à¾ÃÒÐ©Ð¹Ñ ¹

¶éÒ¹ÓàÍÒÅӴѺ¢éÍÁÙÅ·Ñ §Êͧ (ÅӴѺ·Õ #1 áÅÐ #2) ÁÒź¡Ñ¹ àªè¹ ÅӴѺ¢éÍÁÙÅ #1 ź´éÇÂÅӴѺ¢éÍÁÙÅ #2 ¡ç ¨Ðä´é ¼ÅÅѾ¸ì à» ¹

{0,

0,

−2,

2,

−2,

0, 0, 0} ËÃ×Í ¶éÒ ¹ÓàÍÒÅӴѺ ¢éÍÁÙÅ #2 ź´éÇÂÅӴѺ


5.2.

¤ÇÒÁËÁÒ¢ͧà˵ءÒóì¢éͼԴ¾ÅÒ´

93

W(D) A(D)

H(D)

X(D)

Y(D)

Viterbi detector

Aˆ ( D )

ÃÙ»·Õ 5.2: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ GPR ẺÊÁÁÙÅ

¢éÍÁÙÅ #1 ¡ç¨Ðä´é¼ÅÅѾ¸ìà» ¹

{2, −2, 2}

{0,

0, 2,

−2,

2, 0, 0, 0} ´Ñ§¹Ñ ¹ ¼ÅÅѾ¸ì·Õ ä´é

{−2,

2,

−2}

ËÃ×Í

¨ÐàÃÕ¡¡Ñ¹ ÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´ (error sequence) [15, 49, 50] «Ö § àÍÒäÇé ãªé 㹡ÒÃ

¤Ó¹Ç³ËÒà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº ¡ÒÃÇÔà¤ÃÒÐËì à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´¢Í§Ãкº¨ÐªèÇ·ÓãËé ·ÃÒºÇèÒ ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ ÃÙ»áººã´·Õ ¨ÐÊè§ ¼Å¡Ãзº·ÓãËé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ à¡Ô´ ¤ÇÒÁÊѺʹã¹ÃÐËÇèÒ§·Õ ·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ (à¡Ô´ »ÃÒ¡®¡ÒóìµÒÁÃÙ»·Õ 5.1) ´Ñ§¹Ñ ¹ ¶éÒÊÒÁÒö·Õ ¨ÐËÅÕ¡àÅÕ Â§¡ÒÃÊè§ÅӴѺ¢éÍÁÙÅÍÔ¹¾ØµÃٻẺàËÅèÒ¹Ñ ¹ à¢éÒä»ã¹Ãкºä´é ¡ç¨ÐªèÇ·ÓãËéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔà¡Ô´¢éͼԴ¾ÅҴ㹡ÒöʹÃËÑÊ¢éÍÁÙŹéÍÂŧ «Ö § ¨ÐÊ觼ŷÓãËé»ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкº´ÕÂÔ §¢Ö ¹ [49, 50]

5.2

¤ÇÒÁËÁÒ¢ͧà˵ءÒóì¢éͼԴ¾ÅÒ´

Ẻ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµã¹ÃÙ»·Õ 3.2 ÊÒÁÒö·Õ ¨Ð¨Ñ´ãËéÍÂÙèã¹ÃÙ»¢Í§áºº¨ÓÅͧªèͧÊÑ­­Ò³ GPR (generalized partial response) ẺÊÁÁÙÅ ´Ñ§áÊ´§ã¹ÃÙ»·Õ 5.2 àÁ× Í ÅӴѺ¢éͼԴ¾ÅÒ´ ã¹ÃÙ» ·Õ 3.2 ÊÒÁÒö·Õ ¨Ð¶Ù¡ ¾Ô¨ÒóÒÇèÒ à» ¹ ÊÑ­­Ò³Ãº¡Ç¹

W (D)

wk

·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒ

ÇÕà·ÍÃìºÔ µÒÁÃÙ»·Õ 5.2 ã¹·Ò§»¯ÔºÑµÔ (â´Â੾ÒÐÍÂèÒ§ÂÔ § ·Õ ¤ÇÒÁ¨Ø¢éÍÁÙŢͧÎÒÃì´´ÔÊ¡ìä´Ã¿ì ND ÊÙ§) ÊÑ­­Ò³Ãº¡Ç¹

wk

¹Õ ÁÑ¡¨ÐÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ (colored noise) ·Õ ÁÕ¤èÒà©ÅÕ Âà·èÒ

¡Ñº¤èÒÈÙ¹ÂìáÅФèÒ¤ÇÒÁá»Ã»Ãǹà·èҡѺ

2 σw

¨Ò¡áºº¨ÓÅͧã¹ÃÙ» ·Õ 5.2 ÊÑ­­Ò³ÍÔ¹¾Øµ ·Õ ´éÒ¹¢Òà¢éÒ Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè


94

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ã¹ÃÙ»ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìã¹â´àÁ¹

D

ä´é´Ñ§¹Õ

Y (D) = A(D)H(D) + W (D) = X(D) + W (D) â´Â·Õ

A(D) ¤×Í ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ, H(D) =

¢Í§·ÒÃìà¡çµ,

X(D)

=

A(D)H(D)

PL−1

hk Dk

k=0

(5.1)

¤×Í ·ÒÃìà¡çµ,

L ¤×Í ¨Ó¹Ç¹á·ç»·Ñ §ËÁ´

¤×Í ¢éÍÁÙÅ àÍÒµì¾Øµ ªèͧÊÑ­­Ò³, áÅÐ

ú¡Ç¹áººÊÕ·Õ ÁÕ¤èÒà©ÅÕ Âà·èҡѺ¤èÒÈÙ¹Âì áÅФèÒ¤ÇÒÁá»Ã»Ãǹà·èҡѺ ãËé ¾Ô¨ÒóÒÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ 2 Ẻ ¤×Í

A1 (D)

áÅÐ

W (D)

¤×Í ÊÑ­­Ò³

2 σw

A2 (D)

«Ö § à» ¹ ÅӴѺ ¢éÍÁÙÅ ·Õ ·ÓãËé

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ à¡Ô´ ¤ÇÒÁÊѺʹã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ ÁÒ¡·Õ ÊØ´ â´Â·Õ ÊÁÒªÔ¡ áµèÅÐ µÑÇã¹ÅӴѺ¢éÍÁÙÅÍÔ¹¾ØµÁÕ¤èÒ·Õ à» ¹ä»ä´é ¤×Í

{−1, 1}

´Ñ§¹Ñ ¹ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÊÍ´¤Åéͧ¡Ñº

à˵ءÒóì¢éͼԴ¾ÅÒ´ ¨Ð¹ÔÂÒÁâ´Â

εa (D) = A1 (D) − A2 (D) àÁ× Í ÊÁÒªÔ¡áµèÅеÑÇã¹

εa (D)

à» ¹ÊÁÒªÔ¡¢Í§

{−2,

(5.2)

0, 2} ¶éÒ¡Ó˹´ãËéÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾ØµÁÕ

ÃÙ»¢Í§¾ËعÒÁ (polynomial) ¤×Í

εa (D) =

p−1 X

εa,k Dk

k=0

= εa,0 + εa,1 D + · · · + εa,p−1 Dp−1

(5.3)

ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ (valid input error sequence) ¨ÐµéͧÊÍ´¤Åéͧ¡Ñº ¤Ø³ÊÁºÑµÔ 2 ¢éÍ ´Ñ§µèÍ仹Õ

1) ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¾¨¹ìáááÅо¨¹ìÊØ´·éÒ¢ͧ

0

áÅÐ

εa (D)

µéͧÁÕ¤èÒäÁèà·èҡѺ¤èÒÈÙ¹Âì ¹Ñ ¹¤×Í

εa,0 6=

εa,p−1 6= 0

2) ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

εa (D)

·Õ ¶Ù¡µéͧ¨ÐµéͧäÁè Á¤ Õ èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ·Õ ÁÕ ¤èÒ à·èÒ ¡Ñº ¤èÒ ÈÙ¹Âì ã¹

ÅӴѺ¢éͼԴ¾ÅÒ´àÃÕ§µÔ´¡Ñ¹à» ¹¨Ó¹Ç¹ÁÒ¡¡ÇèÒËÃ×Íà·èҡѺ ä´é¨ÐÊ觼ŷÓãËéà¡Ô´à˵ءÒóì¢éͼԴ¾ÅÒ´·Õ àËÁ×͹¡Ñ¹ [49]

L−1

µÑÇ ÁÔ©Ð¹Ñ ¹áÅéÇ ¼ÅÅѾ¸ì·Õ


5.2.

¤ÇÒÁËÁÒ¢ͧà˵ءÒóì¢éͼԴ¾ÅÒ´

95

{ 1 1 -1 1 -1 1 1 -1 -1 1 1 -1 1 -1 1 1 -1 1 -1 1 } A2 ( D ) = { 1 -1 1 -1 -1 1 -1 1 -1 -1 1 -1 -1 1 1 1 1 -1 1 1 } A1 ( D ) =

ε a ( D) =

{ 0 2 -2 2 0 0 2 -2 0 2 0 0 2 -2 0 0 -2 2 -2 0 } ÃÙ»·Õ 5.3: µÑÇÍÂèÒ§¡ÒäӹdzËÒÅӴѺ¢éͼԴ¾ÅÒ´

àÁ× Í ä´é ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ ¶Ù¡µéͧáÅéÇ à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´ (error event) ¢Í§Ãкº¨Ð ¹ÔÂÒÁâ´Â

εx (D) = εa (D)H(D)

µÑÇÍÂèÒ§·Õ 5.1

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ PR4,

H(D) = 1 − D2 ,

ÍÔ¹¾ØµÁÕ¤ÇÒÁÂÒÇà·èҡѺ 20 ºÔµ â´Â·Õ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ Ë¹Ö §

−1, −1, −1,

1,

1, 1,

−1,

−1, −1,

1,

1,

−1,

−1,

1,

1, 1,

−1,

−1, −1,

−1, 1}

1,

1,

−1, −1,

(5.4)

A1 (D)

¶éÒ¡Ó˹´ãËéÅӴѺ¢éÍÁÙÅ

¤×Í

{1,

1, 1, 1,

áÅÐÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ Êͧ 1, 1, 1, 1,

−1,

−1,

A2 (D)

1, 1,

¤×Í

{1,

1, 1} ¨§ËÒÅӴѺ¢éͼԴ¾ÅÒ´

ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ·Ñ §ËÁ´·Õ ¾ºã¹Ãкº¹Õ

ÇÔ¸Õ ·Ó

àÃÔ Áµé¹ ãËé ¤Ó¹Ç³ËÒÅӴѺ ¢éͼԴ¾ÅÒ´

εa (D)

¨Ò¡ÊÁ¡Òà (5.2) «Ö § ¨Ðä´é ¼ÅÅѾ¸ì ´Ñ§ áÊ´§

ã¹ÃÙ» ·Õ 5.3 ¨Ò¡¹Ñ ¹ ãËé ·Ó¡ÒÃËÒÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ ¶Ù¡µéͧ «Ö § ¨Ò¡ÃÙ» ·Õ 5.3 ¨Ðä´é ÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧÁÕ·Ñ §ËÁ´ 4 µÑÇ ¤×Í

{2, −2, 2},

{2, −2},

{2, −2, 2}

áÅÐ

{−2,

A1 (D) − A2 (D)

ËÃ×Í

A2 (D) − A1 (D)

áµè ÅӴѺ ¢éͼԴ¾ÅÒ´ ¨Ð¾Ô¨ÒóҨҡ

{2, −2, 0, 2}, 2,

−2}

{−2, 2, −2}

¶×Í ÇèÒ à» ¹ µÑÇ à´ÕÂǡѹ à¹× ͧ¨Ò¡¢Ö ¹ ÍÂÙè ¡Ñº ÇèÒ à¾ÃÒÐ©Ð¹Ñ ¹ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ

¶Ù¡µéͧã¹Ãкº¨Ð¶×ÍÇèÒÁÕ·Ñ §ËÁ´ 3 Ẻ ¤×Í

{2, −2, 2},

{2, −2, 0, 2},

{2, −2}


96

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ËÃ×ÍÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Òä³ÔµÈÒʵÃìã¹â´àÁ¹

2D3 ,

áÅÐ

ËÁÒÂà˵Ø

2 − 2D

ä´é ¤×Í

2 − 2D + 2D2 , 2 − 2D +

µÒÁÅӴѺ

¨ÐàËç¹ä´éÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ

¢Ö ¹ÍÂÙè¡ÑºÇèҨоԨÒóҨҡ

µÑÇÍÂèÒ§·Õ 5.2

D

A1 (D) − A2 (D)

ËÃ×Í

εa (D)

¨ÐÁÕ¼ÅÅѾ¸ìà·èҡѺ

−εa (D)

·Ñ §¹Õ

{−1, 1}

à» ¹

A2 (D) − A1 (D)

¡Ó˹´ãËéÊÁÒªÔ¡áµèÅеÑǢͧÅӴѺ¢éÍÁÙÅÍÔ¹¾ØµÁÕ¤èÒ·Õ à» ¹ä»ä´é ¤×Í

·Õ ·ÃÒº¡Ñ¹ÇèÒ ã¹Ãкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ (longitudinal recording) ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍÂã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¤×Í ¤Ó¹Ç³ËÒà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµáºº PR4,

ÇÔ¸Õ·Ó

¨Ò¡·Õ ⨷Âì¡Ó˹´ãËé ¨Ðä´éÇèÒ

εa (D) = 2 − 2D + 2D2

{2, −2, 2}

[19] ¨§

H(D) = 1 − D2

´Ñ§¹Ñ ¹ à˵ءÒóì¢éͼԴ¾ÅÒ´

εx (D)

ÊÒÁÒöËÒä´é¨Ò¡ÊÁ¡Òà (5.4) ¹Ñ ¹¤×Í

εx (D) = εa (D)H(D) = (2 − 2D + 2D2 )(1 − D2 ) = 2 − 2D + 2D3 − 2D4 à¾ÃÒÐ©Ð¹Ñ ¹ à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´¢Í§Ãкº¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ ·Õ ãªé ·ÒÃìà¡çµ Ẻ PR4 ¤×Í

εx (D) = 2 − 2D + 2D3 − 2D4

µÑÇÍÂèÒ§·Õ 5.3

¡Ó˹´ãËéÊÁÒªÔ¡áµèÅеÑǢͧÅӴѺ¢éÍÁÙÅÍÔ¹¾ØµÁÕ¤èÒ·Õ à» ¹ä»ä´é ¤×Í

{−1, 1}

à» ¹

·Õ ·ÃÒº¡Ñ¹ ÇèÒ ã¹Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § (perpendicular recording) ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ à¡Ô´¢Ö ¹ºèÍÂã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¤×Í ¤Ó¹Ç³ËÒà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµáºº PR2,

ÇÔ¸Õ ·Ó

¨Ò¡·Õ ⨷Âì ¡Ó˹´ãËé ¨Ðä´é ÇèÒ

εa (D) = 2 − 2D

{2, −2}

[18] ¨§

H(D) = 1 + 2D + D2

´Ñ§¹Ñ ¹ à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´

εx (D)


5.3.

ÃÐÂзҧÂؤÅÔ´

97

ÊÒÁÒöËÒä´é¨Ò¡ÊÁ¡Òà (5.4) ¹Ñ ¹¤×Í

εx (D) = εa (D)H(D) = (2 − 2D)(1 + 2D + D2 ) = 2 + 2D − 2D2 − 2D3 à¾ÃÒÐ©Ð¹Ñ ¹ à˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ·Õ ãªé·ÒÃìà¡çµáºº PR2 ¤×Í =

εx (D)

2 − 2D + 2D3 − 2D4

5.3

ÃÐÂзҧÂؤÅÔ´

¨Ò¡ÃÙ» ·Õ 5.2 ÅӴѺ ¢éͼԴ¾ÅÒ´¢Í§ÊÑ­­Ò³·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ËÃ×Í ·Õ àÃÕ¡¡Ñ¹ ÇèÒ à˵ءÒóì¢éͼԴ¾ÅÒ´ (error event) ¢Í§Ãкº ¤×Í

εx (D) = [A1 (D) − A2 (D)] H(D) = X1 (D) − X2 (D)

(5.5)

¡Ãкǹ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ â´Âãªé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÒÁÒö·Õ ¨Ð¶Ù¡ ͸ԺÒÂâ´Âãªé ÃÙ»ÀÒ¾ÊͧÁÔµÔ µÒÁÃÙ»·Õ 5.4 àÁ× Í

X1 (D)

¤×Í ÊÑ­­Ò³·Õ ¶Ù¡µééͧ áÅÐ

X2 (D)

ÇÕà·ÍÃìºÔ¨ÐµÑ´ÊԹ㨼Դ¾ÅÒ´ ¶éÒ¢¹Ò´¢Í§ÊÑ­­Ò³Ãº¡Ç¹ ä»Âѧ

X2 (D)

ÃÐËÇèÒ§

ÁÕ¤èÒÁÒ¡¡ÇèÒ

X1 (D)

áÅÐ

d/2

â´Â·Õ

d

¤×Í ÊÑ­­Ò³·Õ ¼Ô´¾ÅÒ´ ǧ¨ÃµÃǨËÒ

W (D) (¹Ñ ¹¤×Í ϕ) ã¹·ÔÈ·Ò§¨Ò¡ X1 (D)

¤×Í ÃÐÂзҧÂؤÅÔ´ (Euclidean distance)

εx (D)

·Õ ÇÑ´

X2 (D)

¶éÒ¡Ó˹´ãËéÅӴѺà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§ÃкºÁÕÃÙ»¢Í§¾ËعÒÁ ¤×Í

εx (D) =

n X

εx,k Dk

k=0

= εx,0 + εx,1 D + · · · + εx,n Dn ´Ñ§¹Ñ ¹ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ (squared Euclidean distance) ¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´,

(5.6)

d2 {εa (D)},


98

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

Y ( D) W ( D)

ϕ

X 1 ( D)

X 2 ( D)

d Decision boundary ÃÙ»·Õ 5.4: ÀÒ¾ÊͧÁÔµÔáÊ´§¡Ãкǹ¡ÒöʹÃËÑÊ¢éÍÁÙÅâ´Âãªéǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¨Ð¶Ù¡¹ÔÂÒÁãËéÁÕ¤èÒà·èҡѺ ¤èÒ¾Åѧ§Ò¹¢Í§

εx (D)

[15] ¹Ñ ¹¤×Í

d2 {εa (D)} = kεx (D)k2 n X ε2x,k = k=0

= ε Tε â´Â·Õ

k·k

·Õ ÁÕÊÁÒªÔ¡

¤×Í ¡ÒÃËÒ¤èÒ¹ÍÃìÁ (norm) áÅÐ

ε

¤×Í àÇ¡àµÍÃìá¹ÇµÑ §¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´

n+1 µÑÇ µÑÇÍÂèÒ§àªè¹ ¶éÒ εx (D) = 2−3D+4D3 −5D4

µÑÇÍÂèÒ§·Õ 5.4

(5.7)

¨Ðä´éÇèÒ

εx (D)

ε = [2, −3, 0, 4, −5]T

¨Ò¡¢éÍÁÙÅ ã¹µÑÇÍÂèÒ§·Õ 5.3 ¨§¤Ó¹Ç³ËÒÃÐÂзҧÂؤÅÔ´ ¡ÓÅѧ Êͧ¢Í§à˵ءÒóì

¢éͼԴ¾ÅÒ´¢Í§Ãкº

ÇÔ¸Õ·Ó

¨Ò¡µÑÇÍÂèÒ§·Õ 5.3 ¨Ðä´éÇèÒ

2D + 2D3 − 2D4

εa (D) = 2 − 2D, H(D) = 1 + 2D + D2 ,

´Ñ§¹Ñ ¹ ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´

áÅÐ

εx (D) = 2 −

d2 {εa (D)}

ËÒä´é¨Ò¡ÊÁ¡Òà (5.7) ´Ñ§µèÍ仹Õ

2

d {εa (D)} =

n X i=0

ε2x,i = (2)2 + (−2)2 + (0)2 + (2)2 + (−2)2 = 16

ÊÒÁÒö


5.4.

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å

99

ËÃ×ÍÍÒ¨¨ÐËÒä´é¨Ò¡

d2 {εa (D)} = ε Tε = [2, −2, 0, 2, −2] · [2, −2, 0, 2, −2]T = 16 à¾ÃÒÐ©Ð¹Ñ ¹ ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´

εx (D) = 2 − 2D + 2D3 − 2D4

ÁÕ¤èÒà·èҡѺ 16

¶éÒÊÁÁصÔÇèÒ ÊÑ­­Ò³Ãº¡Ç¹ ¤èÒ¤ÇÒÁá»Ã»Ãǹ¢Í§

W (D)

W (D)

à» ¹ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺºÇ¡ (AWGN) ¨Ðä´éÇèÒ

¨ÐÁÕ¢¹Ò´à·èҡѹ㹷ء·ÔÈ·Ò§ ´Ñ§¹Ñ ¹ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨ÃµÃǨËÒ

ÇÕà·ÍÃìºÔã¹ÃÙ»¢Í§ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ (BER) ¨Ð¢Ö ¹ÍÂÙè¡ÑºÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·ÓãËéà¡Ô´ÅӴѺ¢éͼԴ¾ÅÒ´

εx (D)

εa (D)

·Õ à» ¹¼Å

·Õ ÁÕ¤èÒÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ·Õ ¹éÍÂÊØ´ [15] «Ö §¹ÔÂÒÁâ´Â

d2min =

min

valid εa (D)

£ 2 ¤ d {εa (D)}

(5.8)

à¾ÃÒÐ©Ð¹Ñ ¹ ¤ÇÒÁ¹èÒ¨Ðà» ¹ ¢Í§¢éͼԴ¾ÅÒ´ (probability of error) ËÃ×Í BER ·Õ ´éÒ¹¢ÒÍÍ¡¢Í§ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÒÁÒö»ÃÐÁÒ³¤èÒä´é´Ñ§¹Õ [15]

µ Pe ≈ K1 Q àÁ× Í

K1

¤×Í ¤èÒ¤§µÑÇ·Õ äÁè¢Ö ¹¡Ñº¤èÒ

σw

áÅÐ

Q(x) =

dmin 2σw

√1 2π

R∞ x

(5.9)

e−

u2 2

du ´Ñ§¹Ñ ¹ ã¹¡Ã³Õ·Õ Í§¤ì»ÃСͺ

¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔà» ¹ÊÑ­­Ò³Ãº¡Ç¹ AWGN ¤èÒ ÃÐÂзҧ ÂؤÅÔ´·Õ ¹éÍÂÊØ´

dmin

ÊÒÁÒö·Õ ¨Ð¹ÓÁÒãªé㹡ÒûÃÐÁÒ³¤èÒ¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´¢Í§Ãкº

ä´é â´Â·Õ ¼ÅÅѾ¸ì ·Õ ä´é ¨ÐÁÕ ¤ÇÒÁ¹èÒ àª× Ͷ×Í ÁÒ¡ÂÔ § ¢Ö ¹ ¶éÒ ÃкºÁÕ à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´·Õ â´´à´è¹ (dominant error event) à¾Õ§µÑÇà´ÕÂÇ ¹Ñ ¹¤×ÍÁÕ¤èÒ

d2 {εa (D)} ¹éÍÂÁÒ¡ àÁ× Íà·Õº¡Ñº¤èÒ d2 {εa (D)}

¢Í§ÅӴѺ¢éͼԴ¾ÅÒ´Í× ¹æ ·Õ ¾ºã¹Ãкº

5.4

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å

¶éÒÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

W (D)

à» ¹ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ («Ö §ã¹

·Ò§»¯ÔºÑµÔáÅéÇÁÑ¡¨Ðà» ¹àªè¹¹Õ â´Â੾ÒÐÍÂèÒ§ÂÔ §·Õ ND ÊÙ§) ¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´¨Ð¢Ö ¹ÍÂÙè


100

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡Ñº ·Ñ § ÃÐÂзҧÂؤÅÔ´ áÅФèÒ ¤ÇÒÁá»Ã»Ãǹ¢Í§ÊÑ­­Ò³Ãº¡Ç¹ã¹·ÔÈ·Ò§¢Í§ ã¹¡Ã³Õ¹Õ ¤èÒ

dmin

dmin

X1 (D)

ä»Âѧ

X2 (D)

ε

W (D)

àÇ¡àµÍÃì á¹ÇµÑ § ¢Í§ÊÑ­­Ò³Ãº¡Ç¹ ¨Ðä´éÇèÒ

W (D)

(5.10)

εx (D)

·Õ ÁÕ ÊÁÒªÔ¡

·Õ ÁÕ ÊÁÒªÔ¡

n+1

n+1

µÑÇ àªè¹ ¶éÒ

µÑÇ, áÅÐ

w

W (D) = 0.82 −

ϕ

ÊÒÁÒöËÒä´é¨Ò¡ [48]

σϕ2 = E[ϕ2 ] · T T T ¸ (w ε ) (w ε ) = E kεεk2 £ ¤ ε T E wwT ε = ε Tε T ε Rwwε = ε Tε E[·]

¤×Í

w = [0.82, −1.3, 0, 0.2]T

´Ñ§¹Ñ ¹ ¤èÒ¤ÇÒÁá»Ã»Ãǹ¢Í§

àÁ× Í

ϕ

ŧº¹àÇ¡àµÍÃìË¹Ö §Ë¹èÇ (unit vector)

w Tε kεεk

¤×Í àÇ¡àµÍÃì á¹ÇµÑ § ¢Í§à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´

1.3D + 0.2D3

[48, 49]

¹Ñ ¹¤×Í

ϕ= àÁ× Í

deff {εa (D)}

ÊÓËÃѺ¡ÒûÃÐÁÒ³¤èÒ¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´¢Í§Ãкº ¨Ò¡ÃÙ»·Õ 5.4 ¤èÒ

ÁÕ¤èÒà·èҡѺ¡ÒéÒ (projection) ÊÑ­­Ò³Ãº¡Ç¹ ·Õ ÁÕ·ÔÈ·Ò§¨Ò¡

´Ñ§¹Ñ ¹

¨Ö§äÁèÊÒÁÒö¹ÓÁÒãªé㹡ÒûÃÐÁÒ³¤èÒ¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´·Õ à¡Ô´¢Ö ¹ã¹

Ãкºä´é ´Ñ§¹Ñ ¹ ¨Ö§ä´éÁÕ¡ÒÃ¹Ó ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å (e ective distance) ÁÒãªéá·¹

εx (D), ϕ,

¤×Í µÑÇ´Óà¹Ô¹¡ÒäèÒ¤Ò´ËÁÒ (expectation operator),

(auto correlation matrix) ¢Í§

W (D)

ËÒä´é¨Ò¡

Rww (i, j) = E

â´Â·Õ ÊÁÒªÔ¡á¶Ç·Õ

"S−1 X

i

(5.11)

Rww

¤×Í àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì

áÅÐá¹ÇµÑ §·Õ

j

¢Í§àÁ·ÃÔ¡«ì

Rww

# wk−i wk−j ,

0 ≤ i, j ≤ n + 1

(5.12)

k=0 àÁ× Í

S

¤×Í ¤ÇÒÁÂÒǢͧÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

SNRevent {εa (D)},

{ak }

ã¹ [49] ¤èÒ SNR ¢Í§à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´,

¨Ð¹ÔÂÒÁâ´Â

SNRevent {εa (D)} =

d2 {εa (D)} 2 σw

(5.13)


5.4.

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å

áÅÐ SNR »ÃÐÊÔ·¸Ô¼Å,

101

SNReff ,

¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´ ¤×Í

d2 {εa (D)} σϕ2

SNReff {εa (D)} =

(5.14)

à¾× ÍãËéÊÒÁÒö»ÃÐÁÒ³¤èÒ¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´·Õ à¡Ô´¢Ö ¹ã¹Ãкºä´éÍÂèÒ§¹èÒàª× Ͷ×ÍÁÒ¡ ¡ÇèÒ ¡ÒÃãªé ¾ÒÃÒÁÔàµÍÃì

dmin

µÒÁÊÁ¡Òà (5.9) ¨Ö§ ä´é ÁÕ ¡ÒùÔÂÒÁ¤èÒ ÃÐÂзҧẺãËÁè ¢Ö ¹ ÁÒ·Õ àÃÕ¡

ÇèÒ ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å (e ective distance), á»Ã»Ãǹ

σϕ2

à¢éÒä»´éÇÂáÅéÇ ´Ñ§¹Ñ ¹ àÁ× Íá·¹¤èÒ

áÅéÇ ¨Ðà» ¹¼Å·ÓãËé

SNRevent {εa (D)}

deff {εa (D)}, d2 {εa (D)}

ÁÕ¤èÒà·èҡѺ

SNReff {εa (D)} =

«Ö § ¨ÐÃÇÁ¼Å¡Ãзº¢Í§¤èÒ ¤ÇÒÁ

d2eff {εa (D)}

´éÇÂ

SNReff {εa (D)}

ã¹ÊÁ¡Òà (5.13)

¹Ñ ¹¤×Í

d2eff {εa (D)} d2 {εa (D)} = 2 σw σϕ2

(5.15)

¨Ò¡ÊÁ¡Òà (5.15) ¨Ðä´éÇèÒ ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ (squared e ective distance) ÁÕ¤èÒà·èҡѺ

2 d2eff {εa (D)} = σw 2 = σw

d2 {εa (D)} σϕ2 (εεTε )2 ε T Rwwε

(5.16)

àªè¹à´ÕÂǡѹ ¤èÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ·Õ ¹éÍÂÊØ´ ¨Ð¹ÔÂÒÁâ´Â

d2effmin =

£

min

valid εa (D)

¤ d2eff {εa (D)}

(5.17)

áÅФÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´ã¹ (5.9) ¨Ðà»ÅÕ Â¹ä»à» ¹ [15]

µ Pe ≈ K2 Q àÁ× Í

K2

¤×Í ¤èÒ¤§µÑÇ·Õ äÁè¢Ö ¹¡Ñº¤èÒ

σw

deffmin 2σw

¨Ò¡¡Ò÷´Åͧ¾ºÇèÒ

(5.18)

deffmin

ÊÒÁÒö¹Óãªé㹡ÒûÃÐÁÒ³

¤èÒ ¤ÇÒÁ¹èÒ¨Ðà» ¹ ¢Í§¢éͼԴ¾ÅÒ´¢Í§Ãкºä´é ÍÂèÒ§¹èÒ àª× Ͷ×Í ÁÒ¡¡ÇèÒ ¡ÒÃãªé

deffmin

dmin

·Ñ §¹Õ à¹× ͧÁÒ¨Ò¡

ä´éÃÇÁ¼Å¡Ãзº·Õ à¡Ô´¨Ò¡ÊÑ­­Ò³Ãº¡Ç¹áººÊÕäÇéáÅéÇ

µÑÇÍÂèÒ§·Õ 5.5

¾Ô¨ÒóÒẺ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ã¹ÃÙ» ·Õ 3.2 ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ Ẻ

á¹ÇµÑ § ·Õ ND = 2 áÅÐ SNR = 22 dB â´Â·Õ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

ak ∈ {−1, 1}

áÅлÃÒ¡®ÇèÒ·Ó


102

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ãËéä´é ÅӴѺ¢éͼԴ¾ÅÒ´

−12.72}

{wk }

=

{−3.64, −4.34,

¶éÒÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

Ẻ PR2,

H(D) = 1 + 2D + D2

εa (D)

2.96,

−1.56, −3.70,

·Õ à¡Ô´¢Ö ¹ºèÍÂÃкº ¤×Í

0.80, 8.52,

{2, −2}

−3.76,

6.10,

áÅÐÃкºãªé·ÒÃìà¡çµ

¨§¤Ó¹Ç³ËÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ

d2eff {εa (D)}

¢Í§

ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

ÇÔ¸Õ·Ó

¨Ò¡·Õ ⨷Âì¡Ó˹´ à˵ءÒóì¢éͼԴ¾ÅÒ´

εx (D)

¢Í§Ãкº¹Õ ËÒä´é¨Ò¡

εx (D) = εa (D)H(D) = (2 − 2D)(1 + 2D + D2 ) = 2 + 2D − 2D2 − 2D3 ¹Ñ ¹¤×Í

ε = [2, 2, −2, −2]T

¨Ò¡ÅӴѺ ¢éͼԴ¾ÅÒ´

{wk }

·Õ ¡Ó˹´ãËé ¨Ðä´é ÇèÒ àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì

Rww

ÊÒÁÒöËÒä´é

¨Ò¡ÊÁ¡Òà (5.12) «Ö §ÁÕ¤èÒà·èҡѺ

 34.33 −13.84

Rww

áÅÐ

6.13 −11.25

     −13.84 34.33 −13.84 6.13    =   6.13 −13.84 34.33 −13.84    −11.25 6.13 −13.84 34.33

2 = R σw ww (0, 0) = 34.33

´Ñ§¹Ñ ¹ ¤èÒ ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧ Êͧ

d2eff {εa (D)}

ËÒä´é ¨Ò¡

ÊÁ¡Òà (5.16) ¹Ñ ¹¤×Í

(εεTε )2 ε T Rwwε µ ¶ 256 = 34.33 430.48 = 20.41

2 d2eff {εa (D)} = σw

à¾ÃÒÐ©Ð¹Ñ ¹ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

(5.19)

εa (D) = {2, −2} ã¹Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ §·Õ ãªé·ÒÃìà¡çµ

Ẻ PR2 ¨ÐÁÕ¤èÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧà·èҡѺ 20.41 ˹èÇÂ


5.5.

¼Å¡Ò÷´Åͧ

103

ËÅѧ¨Ò¡¤Ó¹Ç³ËÒ¤èÒ

deffmin

deffmin

ä´é áÅéÇ ´Ñ§¹Ñ ¹ ¤èÒ SNR »ÃÐÊÔ·¸Ô¼Å

¨Ð¹ÔÂÒÁâ´Â

SNReff =

SNReff

d2effmin (εεTε )2 = 2 σw ε T Rwwε

¨Ò¡¡Ò÷´Åͧ·Õ ¨ÐáÊ´§ã¹ËÑÇ¢éÍ·Õ 5.5 ¨Ð¾ºÇèÒ

SNReff

·Õ ÊÍ´¤Åéͧ¡Ñº

(5.20)

ÊÒÁÒö¹ÓÁÒãªé㹡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸Ô

ÀÒ¾¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ ẺµèÒ§æ ä´é ÍÂèÒ§¹èÒ àª× Ͷ×Í àªè¹à´ÕÂǡѹ ¡Ñº ¡ÒÃãªé ¾ÒÃÒÁÔàµÍÃì BER ã¹ ¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº â´Â¼ÅÅѾ¸ì ·Õ ä´é ¨ÐÁÕ ¤ÇÒÁ¹èÒ àª× Ͷ×Í ÁÒ¡ÂÔ § ¢Ö ¹ ¶éÒ ã¹ÃкºÁÕ à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´·Õ â´´à´è¹ à¾Õ§µÑÇ à´ÕÂÇà·èÒ¹Ñ ¹ ¹Ñ ¹¤×Í àÁ× Í Ãкº·Ó§Ò¹ ³ ÃдѺ SNR ·Õ ÊÙ§ à¾Õ§¾Í àªè¹ àÁ× ÍÃкºÁÕ BER

5.5

< 10−4

¼Å¡Ò÷´Åͧ

㹡Ò÷´Åͧà¾× Í ·Ó¡ÒÃÇÔà¤ÃÒÐËì à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´ ¨Ð¾Ô¨ÒóÒ੾ÒÐÃкº¡Òúѹ·Ö¡ Ẻá¹Ç µÑ § (perepdicular recording) à·èÒ¹Ñ ¹ ÊÓËÃѺ¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ (longitudinal recording) ÊÒÁÒöÈÖ¡ÉÒä´é¨Ò¡ [49] ãËé¾Ô¨ÒóÒẺ¨ÓÅͧÃкºã¹ ÃÙ»·Õ 3.2 â´Â·Õ ÊÑ­­Ò³ read back ¨Ðà» ¹ä»µÒÁÊÁ¡Òà (3.26) ¹Ñ ¹¤×Í

p(t) =

S−1 X

bk g(t − kT + ∆tk ) + n(t)

(5.21)

k=0 àÁ× Í

bk = (ak − ak−1 )/2

ʶҹкǡËÃ×Íź áÅÐ ·Õ

k

·Õ Áըӹǹ·Ñ §ËÁ´

¤×Í ºÔµ à»ÅÕ Â¹Ê¶Ò¹Ð (àÁ× Í

bk = 0

bk = ±1

¤×ÍäÁèÁÕ¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹Ð),

S = 4096

ºÔµ (1 à«¡àµÍÃì),

¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § µÒÁÊÁ¡Òà (1.2),

n(t)

g(t)

ÊÍ´¤Åéͧ¡Ñº ¡ÒÃà»ÅÕ Â¹á»Å§

ak ∈ ±1

¤×Í ¢éÍÁÙÅÍÔ¹¾ØµºÔµµÑÇ

¤×Í ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð¢Í§Ãкº

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN)

·Õ ÁÕ¤ÇÒÁ˹Òá¹è¹Ê໡µÃÑÁ¡ÓÅѧẺÊͧ´éÒ¹à·èҡѺ

N0 /2,

áÅÐ

∆t

¤×Í ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì

¢Í§Ê× ÍºÑ¹·Ö¡ (media jitter noise) ·Õ ¶Ù¡¨ÓÅͧãËéÁÕÅѡɳÐà» ¹ ¡ÒÃàÅ× Í¹µÓá˹觢ͧ¡ÒÃà»ÅÕ Â¹ ʶҹÐẺÊØèÁ «Ö §ÁÕ¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹·Õ ÁÕ¤èÒà©ÅÕ Âà·èҡѺ¤èÒÈÙ¹Âì áÅФèÒ ¤ÇÒÁá»Ã»Ãǹà·èÒ ¡Ñº

|bk |σj2

áÅж١ ¨Ó¡Ñ´ ãËé ÁÕ ¤èÒ äÁè à¡Ô¹

¨Ó¹Ç¹à»ÍÃìà«ç¹µì¢Í§ºÔµà«ÅÅì

T

|bk |

áÅÐ

¤×Í ¤èÒÊÑÁºÙóì¢Í§

bk

T /2

àÁ× Í

σj

¨Ð¶Ù¡ ¡Ó˹´à» ¹


104

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊÑ­­Ò³ read back ¨Ð¶Ù¡ Êè§ ¼èÒ¹ä»Âѧ ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó ºÑµà·ÍÃìàÇÔÃìµ Íѹ´Ñº ·Õ 7 áÅж١ ·Ó

1/T

¡Òêѡ µÑÇÍÂèÒ§´éÇÂÍѵÃÒ¤ÇÒÁ¶Õ à·èÒ ¡Ñº

â´ÂÊÁÁØµÔ ÇèÒ Ç§¨Ã¡Òêѡ µÑÇÍÂèÒ§ÁÕ ¡ÒÃà¢éÒ ¨Ñ§ËÇÐẺ

ÊÁºÙóì (perfect synchronization) ¨Ò¡¹Ñ ¹ ÅӴѺ¢éÍÁÙÅ·Õ ä´é

{sk }

¨Ð¶Ù¡Êè§ä»ÂѧÍÕ¤ÇÍäÅà«ÍÃì à¾× Í

»ÃѺ ÃÙ»ÃèÒ§¢Í§ÊÑ­­Ò³ãËé à» ¹ 仵ÒÁ·ÒÃìà¡çµ ·Õ µéͧ¡Òà ¹Ñ ¹¤×Í ÍÕ¤ÇÍäÅà«ÍÃì ¨Ð¾ÂÒÂÒÁ·ÓãËé ÅӴѺ ¢éÍÁÙÅàÍÒµì¾Øµ

{yk }

·Õ ä´éÁÕ¤èÒã¡Åéà¤Õ§¡ÑºÅӴѺ¢éÍÁÙÅ·Õ µéͧ¡ÒÃ

µÃǨËÒÇÕà·ÍÃìºÔ¡ç¨Ð·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅ

{yk }

{dk }

ãËéÁÒ¡·Õ ÊØ´ ËÅѧ¨Ò¡¹Ñ ¹ ǧ¨Ã

à¾× ÍËÒÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{ak }

·Õ à» ¹ä»ä´éÁÒ¡·Õ ÊØ´

㹡Ò÷´Åͧ ¤èÒ SNR ¨Ð¹ÔÂÒÁµÒÁÊÁ¡Òà (3.27) ¹Ñ ¹¤×Í

à SNR = 10 log10

â´Â·Õ

Vp = g(∞) = 1

pulse) ³ àÇÅÒ

t

=

Vp2 σ2

! (dB)

¤×Í ¢¹Ò´¢Í§ÊÑ­­Ò³¾ÑÅÊì à»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È

áÅÐ

σ2

=

N0 /(2T )

¤×Í ¡ÓÅѧ¢Í§ÊÑ­­Ò³Ãº¡Ç¹

(isolated transition

n(t)

㹡Ò÷´Åͧ¹Õ ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì¨Ð¶Ù¡Í͡Ẻ·Õ ND = 2.5 áÅÐ SNR = 22 dB ³ ÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡ µèÒ§æ àªè¹

2, d2effmin , σw

áÅÐ

SNReff

σj

µÒÁ·Õ ¡Ó˹´ÁÒãËé Êèǹ¾ÒÃÒÁÔàµÍÃì

à» ¹µé¹ ¨Ð¶Ù¡¤Ó¹Ç³ËÒâ´Âãªé¢éÍÁÙÅà¾Õ§ 1 à«¡àµÍÃì (4096

ºÔµ) ÊÓËÃѺ áµèÅÐ SNR áÅÐ ND ã¹·Õ ¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒзÒÃìà¡çµ ·Õ ¶Ù¡ Í͡Ẻâ´Âà§× ͹䢺ѧ¤Ñº ẺâÁ¹Ô¡ (´ÙÃÒÂÅÐàÍÕ´ã¹ËÑÇ¢éÍ·Õ 3.2.1) áÅÐà¾× ÍãËé§èÒµèÍ¡ÒÃ͸ԺÒ ÊÑ­Åѡɳì GPRn ¨Ðãªé á·¹·ÒÃìà¡çµáººâÁ¹Ô¡·Õ Áըӹǹá·ç»à·èҡѺ

n á·ç» ¹Í¡¨Ò¡¹Õ

㹡Ò÷´Åͧ à¤Ã× Í§ËÁÒ − ¨Ð

ãªéá·¹¢éÍÁÙÅ −2, à¤Ã× Í§ËÁÒ + ¨Ðãªéá·¹¢éÍÁÙÅ +2, áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Ø¡µÑǨÐÁդسÊÁºÑµÔÊÁÁҵáѹ ¹Ñ ¹¤×Í

5.5.1

εa (D)

εa (D) = −εa (D)

¡ÒÃÇÔà¤ÃÒÐËìÃÐÂзҧ·Õ ¹éÍÂÊØ´

¨Ò¡áºº¨ÓÅͧáÅÐà§× Í¹ä¢·Õ ¡Ó˹´ãËé ¨Ðä´éÇèÒ ·ÒÃìà¡çµáºº GPR3 ·Õ ¶Ù¡Í͡ẺÊÓËÃѺÃкº¡Òà ºÑ¹·Ö¡áººá¹ÇµÑ §·Õ ND = 2.5 áÅÐ

σj = 0%

¤×Í

H(D) = 1 + 1.3022D + 0.6623D2

¨Ò¡¹Ñ ¹

ãËé ãªé ·ÒÃìà¡çµ ¹Õ 㹡Ò÷ӡÒèÓÅͧÃкº (system simulation) à¾× Í ¤Ó¹Ç³ËÒÃÐÂзҧÂؤÅÔ´ ¡ÓÅѧ Êͧ

d2 {εa (D)},

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ

d2eff {εa (D)},

áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

εa (D)

µèÒ§æ ·Õ à¡Ô´¢Ö ¹ ã¹Ãкº (·Õ ´éÒ¹¢ÒÍÍ¡¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ) ³ SNR = 22 dB «Ö §¨Ðä´éµÒÁ·Õ


5.5.

¼Å¡Ò÷´Åͧ

105

µÒÃÒ§·Õ 5.1: ÃÐÂзҧÂؤÅÔ´ ¡ÓÅѧ Êͧ áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

εa (D)

d2 {εa (D)},

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧ Êͧ

d2eff {εa (D)},

¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáºº GPR3 ·Õ ND = 2.5 áÅÐ SNR =

22 dB

¤ÇÒÁÂÒÇ (ºÔµ) ¢Í§

ÃÐÂзҧÂؤÅÔ´ (Euclidean distance)

d2

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å (E ective distance)

ÅӴѺ¢éͼԴ¾ÅÒ´

¨Ó¹Ç¹¤ÃÑ §

εa (D)

·Õ à¡Ô´¢Ö ¹

¢éͼԴ¾ÅÒ´ 1

12.5375

2

7.7579

+

3

8.2764

4

d2eff

ÅӴѺ¢éͼԴ¾ÅÒ´

¨Ó¹Ç¹¤ÃÑ §

εa (D)

·Õ à¡Ô´¢Ö ¹

0

17.0522

127

6.7243

+

2

10.0266

+

2

8.7950

+ +

1

11.2688

+ +

1

5

9.3135

+ +

0

7.0754

+ 0 +

6

9.8321

+ + +

0

9.5535

+ 0 +

1

7

10.3507

+ + +

0

10.6952

+ + 0 +

1

8

10.8692

+ + + +

0

7.8234

+ 0 + 0 +

4

Í× ¹æ

0

+

127

37

50

7

áÊ´§ã¹µÒÃÒ§·Õ 5.1 ¶éÒãªéËÅÑ¡¤ÇÒÁ¨ÃÔ§·Õ ÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÁÕ¤èÒÃÐÂзҧ·Õ ¹éÍÂÊØ´¨Ðà» ¹ÅӴѺ¢éͼԴ¾ÅÒ´ ÍÔ¹¾Øµ ·Õ à¡Ô´ ¢Ö ¹ ºèÍÂ·Õ ÊØ´ ´Ñ§¹Ñ ¹ ¨Ò¡µÒÃÒ§·Õ 5.1 ¶éÒ ãªé ÃÐÂзҧÂؤÅÔ´ ¡ÓÅѧ Êͧ ࡳ±ì 㹡ÒþԨÒÃ³Ò ¨Ðä´é ÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ à¡Ô´ ¢Ö ¹ ºèÍ ¤×Í

{

+ +} à¹× ͧ¨Ò¡ ÁÕ ¤èÒ

d2 {εa (D)}

ãªéÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍ ¤×Í

+},

{

à» ¹

+ }, áÅÐ

à·èÒ ¡Ñº 7.7579, 8.2764, áÅÐ 8.7950 µÒÁÅӴѺ áµè ¶éÒ

d2eff {εa (D)}

{ +}, {

{

d2 {εa (D)}

à» ¹à¡³±ì㹡ÒþԨÒÃ³Ò ¨Ðä´éÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´

+ 0 +}, áÅÐ

{

+ 0 + 0 +} à¹× ͧ¨Ò¡ ÁÕ¤èÒ

d2eff {εa (D)}

à·èҡѺ 6.7243, 7.0754, áÅÐ 7.8234 µÒÁÅӴѺ à¾× Íà» ¹¡ÒõÃǨÊͺ´ÙÇèÒ à¡³±ìã´ (ÃÐÂзҧÂؤÅÔ´ ËÃ×ÍÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å) ÁÕ¤ÇÒÁ¹èÒàª× Ͷ×Í ÁÒ¡¡Çèҡѹ ÊÓËÃѺãªé㹡Ò÷ӹÒÂÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍÂã¹Ãкº ¨Ö§ä´é·Ó¡ÒèÓÅͧ ÃкºÍÕ¡ ¤ÃÑ § Ë¹Ö § â´Â¡ÒÃÊè§ ¢éÍÁÙÅ ËÅÒÂæ à«¡àµÍÃì à¢éÒ ä»ã¹Ãкº (µÒÁẺ¨ÓÅͧã¹ÃÙ» ·Õ 3.2) ·Õ


106

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

SNR = 22 dB ¨¹¡ÃÐ·Ñ § ¢éͼԴ¾ÅÒ´ (error) ÊÐÊÁ·Õ à¡Ô´ ·Ñ §ËÁ´ÁÕ à» ¹ ¨Ó¹Ç¹ 500 ºÔµ ¨Ò¡¹Ñ ¹ ·Ó¡ÒÃÇÔà¤ÃÒÐËìÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾ØµáµèÅÐẺà¡Ô´¢Ö ¹·Ñ §ËÁ´¡Õ ¤ÃÑ § áÅÐàÁ× ÍÊÔ ¹ÊØ´¡Ò÷´Åͧ ¨Ð¾ºÇèÒ ÃкºÁÕ BER =

5.1945 × 10−4 ,

ºÔµ·Õ ¼Ô´¾ÅÒ´Áըӹǹ 500 ºÔµ, áÅÐÅӴѺ¢éͼԴ¾ÅÒ´

ÍÔ¹¾Øµ ·Ñ §ËÁ´ÁÕ ¨Ó¹Ç¹ 180 µÑÇ (ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ áµèÅÐẺ ÍÒ¨¨Ðà¡Ô´ ¢Ö ¹ ä´é ËÅÒ¤ÃÑ § ËÃ×Í ËÅÒµÑÇ) ¨Ò¡µÒÃÒ§·Õ 5.1 ¨Ð¾ºÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ à¡Ô´ ¢Ö ¹ ºèÍ ¤×Í +}, áÅÐ

{

+ 0 + 0 +} µÒÁÅӴѺ «Ö § ÊÍ´¤Åéͧ¡Ñº ¡ÒÃãªé ࡳ±ì

·Ó¹Ò¼Šà¾ÃÒÐ©Ð¹Ñ ¹

d2eff {εa (D)}

{

+},

{

d2eff {εa (D)}

+ 0

㹡ÒÃ

ÊÒÁÒö¹ÓÁÒãªé㹡Ò÷ӹÒÂÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµã´·Õ

à¡Ô´¢Ö ¹ºèÍÂã¹Ãкºä´é¶Ù¡µéͧÁÒ¡¡ÇèÒ¡ÒÃãªé

d2 {εa (D)}

ÍÂèÒ§äáçµÒÁ ¢éÍÊÃØ»¹Õ ¨ÐÁÕ¤ÇÒÁ¹èÒàª× Ͷ×Í

ÁÒ¡¢Ö ¹ ¡çµèÍàÁ× Í ã¹ÃкºÁÕÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ â´´à´è¹à¾Õ§µÑÇà´ÕÂÇ ¹Ñ ¹¤×Í àÁ× ÍÃкº·Ó§Ò¹ ·Õ ÃдѺ SNR ¤è͹¢éÒ§ÊÙ§ (ËÃ×Í·Õ ÃдѺ BER

< 10−4 )

㹷ӹͧà´ÕÂǡѹ µÒÃÒ§·Õ 5.2 áÊ´§¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´ÊÓËÃѺÃкº·Õ ãªé·ÒÃìà¡çµ Ẻ PR2,

H(D) = 1 + 2D + D2 ,

·Õ ND = 2.5 áÅÐ SNR = 22 dB «Ö §ã¹¡Ã³Õ¹Õ ¢éÍÁÙÅËÅÒÂæ

à«¡àµÍÃì¨Ð¶Ù¡Êè§à¢éÒä»ã¹Ãкº ¨¹¡ÃÐ·Ñ §ä´é BER =

1.4251 × 10−3 ,

ºÔµ·Õ ¼Ô´¾ÅÒ´Áըӹǹ 502

ºÔµ, áÅÐÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Ñ §ËÁ´ÁÕ ¨Ó¹Ç¹ 179 µÑÇ ¨ÐàËç¹ ä´é ÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ à¡Ô´ ¢Ö ¹ ºèÍÂã¹Ãкº¹Õ ¤×Í

d2eff {εa (D)}

{

+} áÅÐ

{

+ 0 +} µÒÁÅӴѺ «Ö § ÊÍ´¤Åéͧ¡Ñº ¡ÒÃãªé ࡳ±ì

㹡Ò÷ӹÒ¼Šà¾ÃÒÐ©Ð¹Ñ ¹ ¨Ò¡¼Å¡Ò÷´ÅͧÊÒÁÒöÊÃØ»ä´éÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´

ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍ ¤×Í ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÁÕ ÍÔ¹¾Øµ·Õ ÁÕ

d2 {εa (D)}

d2eff {εa (D)}

¹éÍÂ·Õ ÊØ´ (äÁèãªèÅӴѺ¢éͼԴ¾ÅÒ´

¹éÍÂ·Õ ÊØ´) ¹Í¡¨Ò¡¹Õ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÊÍ´¤Åéͧ¡Ñº

äÁè¨Óà» ¹¨Ðµéͧ໠¹µÑÇà´ÕÂǡѹ¡Ñº ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÊÍ´¤Åéͧ¡Ñº

d2eff {εa (D)}

d2 {εa (D)}

ã¹ÊèǹµèÍä»¹Õ ¨Ð·Ó¡ÒÃÇÔà¤ÃÒÐËì¼Å¡Ãзº¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡µèÍ¡ÒÃà¡Ô´ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµã¹Ãкº ãËé¾Ô¨ÒóÒÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §·Õ ND = 2.5 â´Âãªé·ÒÃìà¡çµ Ẻ GPR5 (¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§áµèÅÐá·ç» ¢Í§·ÒÃìà¡çµ ÊÓËÃѺ áµèÅÐ ã¹ÃÙ» ·Õ 3.7(b)) µÒÃÒ§·Õ 5.3 áÊ´§ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ àËÅèÒ¹Ñ ¹ ³ ¨Ø´ ·Õ ÃкºÁÕ BER = ¢Í§

σj /T

¹éÍ (0%

Ẻá¹ÇµÑ § ¤×Í

− 3%)

{2, −2}

10−4

εa (D)

σj /T

ÊÒÁÒö´Ù ä´é ¨Ò¡¢éÍÁÙÅ

áÅФÇÒÁ¶Õ 㹡ÒÃà¡Ô´ ÅӴѺ

¨Ò¡µÒÃÒ§·Õ 5.3 àÁ× Í ÃдѺ ¤ÇÒÁÃعáç

¨ÐàËç¹ä´éÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ â´´à´è¹ÊÓËÃѺÃкº¡Òúѹ·Ö¡

¹Í¡¨Ò¡¹Õ ¨Ó¹Ç¹¢Í§ÅӴѺ ¢éͼԴ¾ÅÒ´·Õ â´´à´è¹ ¨ÐÁÕ à¾Ô Á ¢Ö ¹ àÁ× Í ÃдѺ


5.5.

¼Å¡Ò÷´Åͧ

107

µÒÃÒ§·Õ 5.2: ÃÐÂзҧÂؤÅÔ´ ¡ÓÅѧ Êͧ áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ¤ÇÒÁÂÒÇ (ºÔµ) ¢Í§

εa (D)

d2 {εa (D)},

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧ Êͧ

¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáºº PR2 ·Õ ND=2.5 áÅÐ SNR=22 dB

ÃÐÂзҧÂؤÅÔ´ (Euclidean distance)

2

d

d2eff {εa (D)},

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å (E ective distance)

ÅӴѺ¢éͼԴ¾ÅÒ´

¨Ó¹Ç¹¤ÃÑ §

d2eff

ÅӴѺ¢éͼԴ¾ÅÒ´

¨Ó¹Ç¹¤ÃÑ §

εa (D)

·Õ à¡Ô´¢Ö ¹

{εa (D)}

εa (D)

·Õ à¡Ô´¢Ö ¹

0

34.0001

127

12.7191

+

¢éͼԴ¾ÅÒ´

{εa (D)}

1

24

2

16

+

3

16

+

9

16.1802

+

9

4

16

+ +

0

19.3624

+ +

0

5

16

+ +

2

13.6519

+ 0 +

6

16

+ + +

1

16.6625

+ 0 +

2

7

16

+ + +

0

16.8916

+ + +

0

8

16

+ + + +

0

15.5885

+ 0 + 0 +

2

Í× ¹æ

¤ÇÒÁÃعáç¢Í§

40

σj /T

0 127

26

13

ÁÒ¡¢Ö ¹ ÊÓËÃѺ ¡ÒÃÇÔà¤ÃÒÐËì à˵ءÒÃ³ì ¢éͼԴ¾ÅÒ´¢Í§Ãкº¡Òúѹ·Ö¡ Ẻ

á¹Ç¹Í¹ ¨Ð¾ºÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ â´´à´è¹ ¤×Í

{2, −2, 2}

[19]

»ÃÐ⪹ì·Õ ä´é¨Ò¡¡ÒÃÈÖ¡ÉÒ¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´ ¤×Í àÁ× Í·ÃÒºÇèÒÃкº·Õ ãªéÁÕÅӴѺ ¢éͼԴ¾ÅÒ´ ÍÔ¹¾Øµ ·Õ â´´à´è¹ Ẻ ã´ ¹Ñ¡ÇԨѠÊÒÁÒö ·Õ ¨Ð à¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾ ÃÇÁ ¢Í§ Ãкº ä´é â´Â ¡Òà Í͡ẺÃËÑÊ RLL (run length limited) [9] ËÃ×Íǧ¨Ãà¢éÒÃËÑÊ¡è͹ (precoder) [47] à¾× Íãªéà¢éÒÃËÑÊ ¢éÍÁÙÅ¢èÒÇÊÒáè͹·Õ ¨Ð·Ó¡ÒÃà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡ à¾× ÍËÅÕ¡àÅÕ Â§¡ÒÃà¡Ô´ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ â´´à´è¹àËÅèÒ¹Ñ ¹ ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

5.5.2

¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§

ã¹ËÑÇ¢éÍ¹Õ ¨ÐáÊ´§ãËéàËç¹ÇèÒ

SNReff

SNReff

ààÅÐ BER

áÅÐ BER ÁÕ¤ÇÒÁÊÑÁ¾Ñ¹¸ì¡Ñ¹¤è͹¢éÒ§ÁÒ¡ â´Â੾ÒÐÍÂèÒ§ÂÔ §

àÁ× Í¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡

σj /T

ÁÕ¤èÒ¹éÍ áÅÐÃкº·Ó§Ò¹·Õ ÃдѺ

SNR ÊÙ§à¾Õ§¾Í(¹Ñ ¹¤×Í àÁ× Íã¹ÃкºÁÕà˵ءÒóì¢éͼԴ¾ÅÒ´·Õ â´´à´è¹à¾Õ§µÑÇà´ÕÂÇ)


108

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

µÒÃÒ§·Õ 5.3: ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ÃкºÁÕ BER =

εa (D)

¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáÅÐ

σj /T

ẺµèÒ§æ ³ ¨Ø´·Õ

10−4

ÅӴѺ

PR2

¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

σj /T

+

GPR5

= 0%

σj /T

GPR5

= 0%

σj /T

GPR5

= 3%

σj /T

= 6%

σj /T

= 9%

4.90%

3.19%

3.36%

5.84%

41.53%

67.54%

83.25%

79.66%

35.21%

9.62%

+ +

5.79%

0.35%

1.98%

38.31%

21.53%

+ +

0.51%

0.58%

1.14%

6.66%

15.73%

+ + +

0.13%

0.23%

0.48%

2.66%

6.31%

+ 0 +

15.53%

8.75%

8.94%

3.03%

0.00%

+ 0 + 0 +

1.34%

0.73%

0.84%

0.22%

0.00%

Í× ¹æ

4.26%

2.72%

3.60%

8.06%

5.28%

+

ÃÙ»·Õ 5.5 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº·Õ ãªé·ÒÃìà¡çµáºº GPR5 ³ ÃдѺ ÃÙ»¢Í§ BER áÅÐ

SNReff

SNReff

σj /T

·Õ ND = 2.5 ¨ÐàËç¹ä´éÇèÒ »ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ»¢Í§ BER áÅÐ

¼ÅÅѾ¸ìà» ¹ä»µÒÁ·Õ ¤Ò´ËÇѧäÇé ¡ÅèÒǤ×Í àÁ× Í ¤èÒ

GPR5

σj /T

µèÒ§æ ã¹

SNReff

ÁÕ

ÁÕ¤èÒà¾Ô Á¢Ö ¹ ¤èÒ BER ¢Í§Ãкº¡ç¨Ðà¾Ô Á¢Ö ¹ áÅÐ

¢Í§Ãкº¡ç¨ÐÅ´¹éÍÂŧ ¹Í¡¨Ò¡¹Õ ÃÙ»·Õ 5.5(a) áÊ´§ãËéàËç¹ÇèÒ ³ ÃдѺ¤ÇÒÁÃعáç

¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡µ Ó ¤èÒ

SNReff

ÊÒÁÒö¹ÓÁÒãªé㹡Ò÷ӹÒ¤èÒ BER ä´é

¨Ò¡ÊÁ¡Òà (5.18) áÅÐ (5.20) ¨Ðä´éÇèÒ [19]

BER ≈ K2 Q â´Â·Õ

K2

¤×Í ¤èÒ¤§µÑÇ·Õ äÁè¢Ö ¹¡Ñº

2 σw

µ p ¶ 1 SNReff 2

µÑÇÍÂèÒ§àªè¹ àÁ× Í

σj /T = 0%

(5.22)

¤èÒ BER ·Õ ·Ó¹ÒÂä´é¨ÐáÊ´§

´éÇÂàÊé¹ Q(·) ã¹ÃÙ» «Ö §ÊÍ´¤Åéͧ¡Ñº¤èÒ BER ¨ÃÔ§·Õ ä´é¨Ò¡¡ÒèÓÅͧÃкº àÁ× Í ÃÙ»·Õ 5.6(a) áÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§ BER áÅРẺ㴡çµÒÁ ¶éÒÃкºÁÕ¤èÒ àÁ× Í

σj /T

SNReff

ÁÕ ¤èÒ ¹éÍ ´Ñ§¹Ñ ¹ ¤èÒ

SNReff

K2 = 2.3

¨ÐàËç¹ä´éÇèÒ äÁèÇèÒÃкº¨Ðãªé·ÒÃìà¡çµ

à·èҡѹáÅéÇ Ãкº¡ç¨ÐÁÕ¤èÒ BER ·Õ ã¡Åéà¤Õ§¡Ñ¹ â´Â੾ÒÐÍÂèÒ§ÂÔ §

SNReff

¨Ö§ ÊÒÁÒö¹ÓÁÒãªé 㹡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§

·ÒÃìà¡çµ ẺµèÒ§æ ä´é àªè¹à´ÕÂǡѺ ¡ÒÃãªé ¤èÒ BER ÍÂèÒ§äáçµÒÁ ¤ÇõÃÐ˹ѡ äÇé ÇèÒ ÃкºµèÒ§æ ¨Ð


5.5.

¼Å¡Ò÷´Åͧ

109

−1

10

−2

BER

10

−3

10

jitter = 0% jitter = 3% jitter = 6% jitter = 9% Q(⋅) with jitter = 0%

−4

10

−5

10

14

15

16

17

(a)

18

19

20

21

22

23

22

23

Electronics SNR (dB)

19

18

17

SNReff (dB)

16

15

14

13

jitter = 0% jitter = 3% jitter = 6% jitter = 9%

12

11

10

9 14

15

16

17

(b)

18

19

20

21

Electronics SNR (dB)

ÃÙ»·Õ 5.5: »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºã¹ÃÙ» (a) BER áÅÐ (b)

SNReff

¢Í§·ÒÃìà¡çµáºº GPR5

ãªé »ÃÔÁÒ³ SNR ·Õ µèÒ§¡Ñ¹ 㹡Ò÷ÓãËé à¡Ô´ ¤èÒ BER áÅФèÒ

SNReff

·Õ à·èÒ ¡Ñ¹ ´Ñ§·Õ áÊ´§ã¹ÃÙ» ·Õ

5.6(b)


110

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

−1

10

−2

BER

10

−3

10

PR2 [1 2 1], jitter=0% GPR5, jitter=0% GPR5, jitter=3% GPR5, jitter=6%

−4

10

−5

10

8

10

12

14

(a)

16

18

20

Effective SNR (dB)

−1

10

−2

BER

10

−3

10

−4

10

PR2 [1 2 1], jitter=0% GPR5, jitter=0% GPR5, jitter=3% GPR5, jitter=6%

−5

10

14

15

16

17

(b)

ÃÙ»·Õ 5.6:

(a) ¡ÃÒ¿ BER áÅÐ

SNReff ,

18

19

20

21

22

23

Electronics SNR (dB)

(b) ¡ÃÒ¿ BER áÅÐ SNR ¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ Ẻ

µèÒ§æ ³ ÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡·Õ à¢Õ¹ÇèÒ jitter µèÒ§æ ·Õ ND = 2.5


5.6.

ÊÃØ»·éÒº·

5.6

111

ÊÃØ»·éÒº·

㹡ÒÃÍ͡Ẻ·ÒÃìà¡çµ µÒÁà§× ͹䢺ѧ¤Ñº ẺâÁ¹Ô¡ (monic constraint) â´Âãªé Ẻ¨ÓÅͧã¹ÃÙ» ·Õ 3.2 ÅӴѺ ¢éͼԴ¾ÅÒ´

wk

ÊÒÁÒö·Õ ¨Ð¶Ù¡ ¾Ô¨ÒóÒä´é ÇèÒ à» ¹ ÊÑ­­Ò³Ãº¡Ç¹ã¹áºº¨ÓÅͧªèͧ

ÊÑ­­Ò³ GPR ẺÊÁÁÙÅ µÒÁÃÙ»·Õ 5.2 ¶éÒÊÑ­­Ò³Ãº¡Ç¹

wk

à» ¹ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺ

ºÇ¡áÅéÇ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº¨Ð¢Ö ¹ÍÂÙè¡Ñº ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ÁÕ¤èÒÃÐÂзҧÂؤÅÔ´·Õ ¹éÍÂÊØ´

dmin

ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ

wk

ÁÑ¡¨ÐÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ â´Â੾ÒÐÍÂèÒ§ÂÔ §

·Õ ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ìÊÙ§æ ´Ñ§¹Ñ ¹ ã¹¡Ã³Õ ¹Õ »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº¨Ð¢Ö ¹ ÍÂÙè ¡Ñº ÅӴѺ ¢éͼԴ¾ÅÒ´·Õ ÁÕ¤èÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å·Õ ¹éÍÂÊØ´

deffmin

¼Å¡Ò÷´ÅͧáÊ´§ãËéàËç¹ÇèÒ ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍ ¤×Í ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ ÁÕ

d2eff {εa (D)}

·Õ ¹éÍÂÊØ´ (äÁè ãªè ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ ÁÕ

d2 {εa (D)}

·Õ ¹éÍÂÊØ´) ¹Í¡¨Ò¡¹Õ

¨Ó¹Ç¹¢Í§ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ â´´à´è¹¨ÐÁÕà¾Ô Á¢Ö ¹ àÁ× ÍÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹ ¨Ôµ àµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ à´ÕÂÇ (àÁ× Í ÃкºÁÕ

σj /T

σj /T

ÁÒ¡¢Ö ¹ ã¹¡Ã³Õ ·Õ ÃкºÁÕ ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ â´´à´è¹ à¾Õ§µÑÇ

¹éÍ áÅзӧҹ·Õ ÃдѺ SNR ÊÙ§ à¾Õ§¾Í) ¹Ñ¡ÇԨѠÊÒÁÒöãªé»ÃÐ⪹ì

¨Ò¡¢éÍÁÙŢͧ¡ÒÃÇÔà¤ÃÒÐËìà˵ءÒóì¢éͼԴ¾ÅÒ´¹Õ ÁÒãªé㹡ÒÃÍ͡ẺÃËÑÊ RLL ËÃ×Íǧ¨Ãà¢éÒÃËÑÊ ¡è͹ à¾× Íãªéà¢éÒÃËÑÊ¢éÍÁÙÅ¢èÒÇÊÒáè͹·Õ ¨Ð·Ó¡ÒÃà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡ à¾× ÍËÅÕ¡àÅÕ Â§¡ÒÃà¡Ô´ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ â´´à´è¹ àËÅèÒ¹Ñ ¹ ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ «Ö §¨ÐªèÇ·ÓãËé»ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкº´ÕÁÒ¡¢Ö ¹ ÊØ´·éÒÂä´éáÊ´§ãËéàËç¹ÇèÒ

SNReff

ãªé㹡ÒÃÇÑ´»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкºä´é àªè¹à´ÕÂǡѺ¡ÒÃãªé BER ¹Ñ ¹¤×Í Ãкº·Õ ÁÕ

ÊÒÁÒö¹ÓÁÒ

SNReff

¤èÒ BER µ Ó áÅж֧áÁéÇèÒ ÃкºáµèÅÐÃкº¨Ðãªé·ÒÃìà¡çµµèÒ§¡Ñ¹ áµè¶éÒÃкºàËÅèÒ¹Ñ ¹ÁÕ

ÊÙ§ ¡ç¨ÐÁÕ

SNReff

à·èÒ

¡Ñ¹ Ãкº¡ç¨ÐÁÕ BER ã¡Åéà¤Õ§¡Ñ¹´éÇÂ

5.7

à຺½ ¡ËÑ´·éÒº·

1. ¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ PR2,

H(D) = 1 + 2D + D2 ,

ÍÔ¹¾ØµÁÕ¤ÇÒÁÂÒÇà·èҡѺ 25 ºÔµ â´Â·Õ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ·Õ Ë¹Ö §

−1,

1, 1,

−1, −1,

1, 1,

−1,

1,

−1,

1, 1,

−1,

1,

−1,

1, 1,

¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ

A1 (D) −1,

¤×Í

1, 1,

{1, −1, −1}

1, 1,

áÅÐÅӴѺ


112

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¢éÍÁÙÅÍÔ¹¾Øµ·Õ Êͧ 1, 1,

−1,

1, 1,

A2 (D)

−1, −1,

¤×Í

{1,

1, 1,

1,

−1}

−1,

1,

−1,

1,

−1,

−1,

1,

1, 1,

−1, −1,

1,

−1,

¨§ËÒÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ·Ñ §ËÁ´·Õ ¾ºã¹

Ãкº¹Õ

2. ¡Ó˹´ãËé ÊÁÒªÔ¡ áµèÅеÑÇ ¢Í§ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ ÁÕ ¤èÒ ·Õ à» ¹ ä»ä´é ¤×Í ºÑ¹·Ö¡ Ẻá¹Ç¹Í¹

(longitudinal

recording)

{−1,

1} ã¹Ãкº¡ÒÃ

ÅӴѺ ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ ·Õ à¡Ô´ ¢Ö ¹ ºèÍÂ

{2, −2, 2}

ã¹ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¤×Í

¨§¤Ó¹Ç³

ËÒà˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµ´Ñ§µèÍ仹Õ

2.1)

H(D) = 1 − D

2.2)

H(D) = 1 + D − D2 − D3

2.3)

H(D) = 1 + 2D − 2D3 − D4

2.4)

H(D) = 1 − 0.04D − 0.64D2

2.5)

H(D) = 1 + 0.22D − 0.65D2 − 0.36D3

2.6)

H(D) = 1 + 0.24D − 0.50D2 − 0.40D3 − 0.21D4

3. ¡Ó˹´ãËé ÊÁÒªÔ¡ áµèÅеÑÇ ¢Í§ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ ÁÕ ¤èÒ ·Õ à» ¹ ä»ä´é ¤×Í ºÑ¹·Ö¡áººá¹ÇµÑ §

(perpendicular

recording)

{−1,

1} ã¹Ãкº¡ÒÃ

ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ à¡Ô´¢Ö ¹ºèÍÂã¹

ÃÐËÇèÒ§¡Ãкǹ¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

¤×Í

à˵ءÒóì¢éͼԴ¾ÅÒ´¢Í§Ãкº¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµ´Ñ§µèÍ仹Õ

3.1)

H(D) = 1 + D

3.2)

H(D) = 1 + 3D + 3D2 + D3

3.3)

H(D) = 1 + 4D + 6D2 + 4D3 + D4

3.4)

H(D) = 1 + 1.30D + 0.66D2

3.5)

H(D) = 1 + 1.19D + 0.60D2 + 0.12D3

3.6)

H(D) = 1 + 1.21D + 0.62D2 + 0.16D3 + 0.01D4

{2, −2}

¨§¤Ó¹Ç³ËÒ


5.7.

à຺½ ¡ËÑ´·éÒº·

113

4. ¨§¾ÔÊÙ¨¹ì¤èÒÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ

d2 {εa (D)}

·Õ áÊ´§ã¹µÒÃÒ§·Õ 5.1

5. ¨§¾ÔÊÙ¨¹ì¤èÒÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ

d2 {εa (D)}

·Õ áÊ´§ã¹µÒÃÒ§·Õ 5.2

6. ¾Ô¨ÒóÒẺ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµã¹ÃÙ»·Õ 3.2 ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ ·Õ

ak ∈ {−1, 1}

ND = 2 áÅÐ SNR = 22 dB â´Â·Õ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{wk } = {−1.95,

1.66, 0.36,

·Õ à¡Ô´¢Ö ¹ºèÍÂÃкº ¤×Í

−0.63,

{2, −2, 2}

¶éÒÅӴѺ¢éͼԴ¾ÅÒ´

1.90, 0.10, 2.12} áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

¨§¤Ó¹Ç³ËÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ

¢Í§ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµ

6.1)

H(D) = 1 − D

6.2)

H(D) = 1 − D2

6.3)

H(D) = 1 + D − D2 − D3

H(D)

εa (D)

d2eff {εa (D)}

´Ñ§µèÍ仹Õ

7. ¾Ô¨ÒóÒẺ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ ã¹ÃÙ» ·Õ 3.2 ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § ·Õ ND = 2 áÅÐ SNR = 22 dB â´Â·Õ ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{wk }

=

εa (D)

{−5.37, −4.65,

0.56, 5.60,

·Õ à¡Ô´ ¢Ö ¹ ºèÍÂÃкº ¤×Í

−2.40, −8.26,

{2, −2}

H(D) = 1 + D

7.2)

H(D) = 1 + 2D + D2

7.3)

H(D) = 1 + 3D + 3D2 + D3

¶éÒÅӴѺ¢éͼԴ¾ÅÒ´

4.85} áÅÐÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

¨§ËÒÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧ Êͧ

¢Í§ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ¹Õ àÁ× ÍÃкºãªé·ÒÃìà¡çµ

7.1)

ak ∈ {−1, 1}

H(D)

´Ñ§µèÍ仹Õ

d2eff {εa (D)}


114

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ


º··Õ 6

ǧ¨ÃµÃǨËÒ NPML

㹺·¹Õ ¨Ð͸ԺÒÂËÅÑ¡¡Ò÷ӧҹáÅлÃÐâª¹ì ¢Í§ ǧ¨ÃµÃǨËÒ NPML (noise predictive maxi mum likelihood) [51, 52] «Ö §à» ¹Ç§¨ÃµÃǨËÒ·Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒǧ¨ÃµÃǨËÒ PRML (par tial response maximum likelihood) â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í Ãкº·Ó§Ò¹·Õ ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÊÙ§ ã¹·Ò§»¯ÔºÑµÔ ǧ¨ÃµÃǨËÒ NPML »ÃÐÂØ¡µì ÁҨҡǧ¨ÃµÃǨËÒ PRML â´Â¡ÒÃ¹Ó ¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹ (noise prediction process) ὧà¢éÒ ä»ÍÂÙè ã¹áµèÅÐàÊé¹ ÊÒ¢Ò (branch) ¢Í§á¼¹ÀÒ¾à·ÃÅÅÔÊ (trellis diagram) ´Ñ§·Õ ¨Ð͸ԺÒµèÍä»ã¹º·¹Õ ¾ÃéÍÁ·Ñ §áÊ´§ ¼Å¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾ÃÐËÇèҧǧ¨ÃµÃǨËÒ PRML áÅÐǧ¨ÃµÃǨËÒ NPML

6.1

º·¹Ó

à·¤¹Ô¤ PRML ¤×Í ¡ÒÃãªé §Ò¹ÃèÇÁ¡Ñ¹ ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃì Ẻ PR (partial response) áÅÐǧ¨Ã µÃǨËÒÇÕà·ÍÃìºÔ (Viterbi detector) «Ö § à» ¹ ·Õ ¹ÔÂÁãªé §Ò¹ÁÒ¡ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì µÒÁ·Õ ͸ԺÒÂ㹺··Õ 4 ãËé ¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁè µèÍà¹× ͧ·Ò§àÇÅÒ áººÊÁÁÙÅ ã¹â´àÁ¹

D

µÒÁÃÙ» ·Õ 6.1 àÁ× Í

A(D)

N (D) ¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺºÇ¡ ¤×Í ·ÒÃìà¡çµ (target),

¤×Í ¢éÍÁÙÅ ºÔµ ÍÔ¹¾Øµ,

(AWGN),

C(D)

¤×Í ªèͧÊÑ­­Ò³,

F (D) ¤×Í ÍÕ¤ÇÍäÅà«ÍÃìẺ

PR,

H(D)

Y (D) ¤×Í ¢éÍÁÙÅ·Õ ¨Ð¶Ù¡Êè§à¢éÒä»·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ, 115


116

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

N(D)

F(D)

channel

H ( D) Y(D) C ( D)

A(D) C(D)

Aˆ ( D )

Viterbi detector

ÃÙ»·Õ 6.1: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ

áÅÐ

Â(D)

¤×Í ¤èÒ»ÃÐÁÒ³¢Í§¢éÍÁÙźԵÍÔ¹¾Øµ

A(D)

¨Ò¡ÃÙ» ·Õ 6.1 ¢éÍÁÙÅ àÍÒµì¾Øµ ¢Í§ÍÕ¤ÇÍäÅà«ÍÃì

Y (D)

ÊÒÁÒöà¢Õ¹໠¹ ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃì

ä´éµÒÁÊÁ¡Òà (4.6) ¹Ñ ¹¤×Í

H(D) Y (D) = A(D)H(D) + N (D) | {z } C(D) | {z } wanted signal

(6.1)

W (D)

â´Â·Õ

W (D)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹·Õ ¨Ðà¢éÒä»ã¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ (µÒÁ·Õ ͸ԺÒÂã¹ËÑÇ¢éÍ·Õ 4.2.2)

â´Â·Ñ Çä» Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡·Õ ÊØ´ ¡çµèÍàÁ× Í

W (D)

ÁÕÅѡɳÐ

à» ¹ ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ áÅéÇ (â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í Ãкº·Ó§Ò¹·Õ ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ìÊÙ§æ) ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ (tap) ¹éÍ ãËéÁռŵͺʹͧàËÁ×͹¡ÑºªèͧÊÑ­­Ò³ 1

¨ÐÁÕ ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ

C(D)

H(D)

·Õ ÁÕ ¨Ó¹Ç¹á·ç»

·Óä´éÂÒ¡ÁÒ¡ ´Ñ§¹Ñ ¹â´Â·Ñ Çä»

W (D)

(colored noise) [25] «Ö § ¨ÐÊè§ ¼Å·ÓãËé »ÃÐÊÔ·¸ÔÀÒ¾¡ÒÃ

·Ó§Ò¹¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ Ŵŧ à¾ÃÒÐ©Ð¹Ñ ¹ ¶éÒ µéͧ¡ÒÃãËé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ÊÒÁÒö·Õ ¨Ð ·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÊÙ§ÊØ´àËÁ×͹à´ÔÁ ¹Ñ¡ÇԨѨеéͧËÒÇÔ¸Õ¡ÒÃã´ÇÔ¸Õ¡ÒÃË¹Ö §ã¹¡Ò÷ÓãËéͧ¤ì »ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹ã¹¢éÍÁÙÅ

Y (D)

(¹Ñ ¹¤×Í

W (D))

ÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì

ÊÕ¢ÒÇẺºÇ¡ ¡è͹·Õ ¨ÐÊ觼ÅÅѾ¸ì·Õ ä´éà¢éÒä»·Ó¡ÒöʹÃËÑÊ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ´Ñ§¹Ñ ¹ÍÒ¨¨Ð¡ÅèÒÇä´éÇèÒ à·¤¹Ô¤ NPML [51, 52] ¤×Í ¡ÒÃãªé§Ò¹ÃèÇÁ¡Ñ¹ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃìẺ PR áÅÐǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ·Õ ÁÕ¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹ ËÃ×Í ¡Ãкǹ¡ÒÃ㹡Òà ·ÓãËé ÊÑ­­Ò³Ãº¡Ç¹à» ¹ ÊÕ ¢ÒÇ (noise whitening process) ὧÍÂÙè ¢éÒ§ã¹ÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ´Ñ§ 1

¢éÍÁÙÅá«Á໠ŢͧÊÑ­­Ò³Ãº¡Ç¹

W (D)

áµèÅÐá«Á໠ŨÐÁÕÊËÊÑÁ¾Ñ¹¸ì (correlation) ¡Ñ¹


6.2.

¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹

117

wk target

ak

H(D)

rk

yk

âk Viterbi algorithm

predictor

P(D) ÃÙ»·Õ 6.2: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³¾ÃéÍÁǧ¨ÃµÃǨËÒ NPML

áÊ´§ã¹ÃÙ»·Õ 6.2 à¾ÃÒÐ©Ð¹Ñ ¹ àÁ× ÍÃкº·Ó§Ò¹·Õ ¤ÇÒÁ¨Ø¢éÍÁÙŢͧÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÙ§æ ǧ¨ÃµÃǨËÒ NPML ¨Ö§ ¤ÇÃ·Õ ¨Ð¶Ù¡ ¹ÓÁÒãªé §Ò¹ÁÒ¡¡ÇèÒ ¡ÒÃãªé ǧ¨ÃµÃǨËÒ PRML à¾× Í·Õ ¨Ðä´é »ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ ¢Í§Ãкº·Õ ´Õ¡ÇèÒ

6.2

¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹

ǧ¨Ã¡Ãͧ·Ó¹Ò (predictor lter) ·Õ ãªé§Ò¹·Ñ Çä»ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³´Ô¨Ô·ÑŨÐÁÕÅѡɳР·Ñ §·Õ à» ¹ Ẻ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì ¨Ó¡Ñ´ (FIR: nite impulse response) áÅÐẺ¼ÅµÍºÊ¹Í§ ÍÔÁ¾ÑÅÊìäÁè¨Ó¡Ñ´ (IIR: in nite impulse response) ¹Í¡¨Ò¡¹Õ Õ ¤Ø³ÊÁºÑµÔ·Ñ Ç仢ͧǧ¨Ã¡Ãͧ·Ó¹Ò ¤×Í ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ (prediction error) ¨Ð¤èÍÂæ Ŵŧ àÁ× Í¨Ó¹Ç¹á·ç»¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò à¾Ô Á¢Ö ¹ ã¹Ë¹Ñ§Ê×ÍàÅèÁ¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒÐǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ FIR à·èÒ¹Ñ ¹ ¾Ô¨ÒóÒẺ¨ÓÅͧ¡Ãкǹ¡ÒÃ㹡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ ã¹ÃÙ»·Õ 6.3 àÁ× Í ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ, áÅÐ

P (D)

ŵk

¤×Í ¤èÒ»ÃÐÁÒ³¢Í§

wk , ek = wk − ŵk

¤×Í ¿ §¡ìªÑ¹¶èÒÂâ͹¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂã¹â´àÁ¹

P (D) =

N X k=1

D

wk

¤×Í

¤×Í ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ,

«Ö §ÁÕÃÙ»ÊÁ¡Òà ¤×Í

pk Dk = p1 D + p2 D2 + . . . + pN DN

(6.2)


118

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

wk

ek predictor

wˆ k

P(D)

ÃÙ»·Õ 6.3: Ẻ¨ÓÅͧ¡Ãкǹ¡ÒÃ㹡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ

â´Â·Õ

pk

¤×Í ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô µÑÇ·Õ

¡Ãͧ·Ó¹Ò ´Ñ§¹Ñ ¹

ŵk

k

¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò áÅÐ

N

¤×Í ¨Ó¹Ç¹á·ç»·Ñ §ËÁ´¢Í§Ç§¨Ã

ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìä´é ¤×Í

ŵk =

N X

pi wk−i

(6.3)

i=1 ÊÁ¡Òà (6.3) ¨ÐÃÙé¨Ñ¡¡Ñ¹ã¹ª× ͧ͢ ǧ¨Ã¡Ãͧ·Ó¹ÒÂË¹Ö §¢Ñ ¹áººàªÔ§àÊé¹ (linear one step predic tor) àªè¹à´ÕÂǡѹ ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ

ek

ÊÒÁÒö¨Ñ´ ãËé ÍÂÙè ã¹ÃÙ» ¢Í§ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃì ä´é

´Ñ§¹Õ

ek = wk − ŵk = wk −

N X

pi wk−i

(6.4)

i=1 ËÃ×Íà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§â´àÁ¹

D

ä´é ¤×Í

E(D) = [1 − P (D)]W (D) â´Â·Õ ¾¨¹ì

[1 − P (D)]

(6.5)

¨ÐàÃÕ¡¡Ñ¹ ·Ñ Çä»ÇèÒ Ç§¨Ã¡Ãͧ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ (prediction error

lter)

6.3

¡ÒÃËÒ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ

¨Ø´»ÃÐʧ¤ì 㹡ÒÃÍ͡Ẻǧ¨Ã¡Ãͧ·Ó¹Ò ·Ó¹ÒÂ) ¤×Í ¡Ò÷ÓãËé¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ

P (D) ek

(¹Ñ ¹¤×Í ¡ÒÃËÒ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ

ÁÕ¤èÒ¹éÍÂ·Õ ÊØ´ ËÃ×ÍÍÕ¡¹ÑÂË¹Ö §¡ç¤×Í ¡Ò÷ÓãËé

ek

ÁÕ


6.3.

¡ÒÃËÒ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ

119

ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹ÊÕ ¢ÒÇãËé ÁÒ¡·Õ ÊØ´ ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ¢éÍÁÙÅ

ek

¶×Í ÇèÒ à» ¹ ͧ¤ì»ÃСͺ

¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ËŧàËÅ×Í ÍÂÙè ã¹¢éÍÁÙÅ ·Õ ¨ÐÊè§ à¢éÒ ä»·Ó¡ÒöʹÃËÑÊ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ´Ñ§¹Ñ ¹ ÇÔ¸Õ¡ÒÃËÒ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ·Õ ´Õ·Õ ÊØ´ ¡ç¤×Í ¡Ò÷ÓãËé¤èÒ¢éͼԴ¾ÅÒ´¡ÓÅѧÊͧ à©ÅÕ Â

(MSE: mean squared error) [51, 52]

£ ¤ £ ¤ E e2k = E (wk − ŵk )2 ÁÕ ¤èÒ ¹éÍÂ·Õ ÊØ´ â´Â·Õ

E[·]

(6.6)

¤×Í µÑÇ´Óà¹Ô¹¡ÒäèÒ ¤Ò´ËÁÒ «Ö § ÊÒÁÒö·Óä´é â´Â¡ÒÃËÒ͹ؾѹ¸ì ¢Í§

ÊÁ¡Òà (6.6) à·Õº¡Ñº¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ

pi

áµèÅеÑÇ áÅéÇãËé¼ÅÅѾ¸ì·Õ ä´éÁÕ¤èÒà·èÒ

¡Ñº¤èÒÈÙ¹Âì ¨Ò¡¹Ñ ¹ ·Ó¡ÒÃá¡éÃкºÊÁ¡ÒÃàªÔ§àÊ鹡ç¨Ðä´é¤ÓµÍºÍÍ¡ÁÒ ËÃ×Í ÍÒ¨¨ÐÍÒÈÑ ËÅÑ¡¡Òà àªÔ§µÑ §©Ò¡

(orthogonality principle) ·Õ ÇèÒ

E [(wk − ŵk )wm ] = 0 ÊÓËÃѺ

m = 1, 2, . . . , N

(6.7)

â´Â¡ÒÃá¡éÊÁ¡Òà (6.7) ¨Ðä´éÇèÒ

E[wk wm ] −

N X

pi E[wk−i wm ] = 0

i=1

E[wk wm ] = Rww (k − m) =

N X i=1 N X

pi E[wk−i wm ] pi Rww (k − i − m)

(6.8)

i=1 àÁ× Í

Rww (i)

¤×Í ¤èÒÍѵÊËÊÑÁ¾Ñ¹¸ì (auto correlation) ÅӴѺ·Õ

â´Â

Rww (i) = E[wk+i wk ] = E

"S−1 X

i

¢Í§ÊÑ­­Ò³Ãº¡Ç¹

wk

«Ö §¹ÔÂÒÁ

# wk+i wk

(6.9)

k=0 â´Â·Õ

S

¤×Í ¤ÇÒÁÂÒÇËÃ×ͨӹǹºÔµ¢Í§ÅӴѺ¢éÍÁÙÅ

{wk }

¶éÒá·¹¤èÒ

k−m=j

ã¹ÊÁ¡Òà (6.8)

¨Ðä´é¼ÅÅѾ¸ìà» ç¹ ÊÁ¡ÒùÍÃìÁÍÅ (normal equation) ¹Ñ ¹¤×Í

Rww (j) =

N X i=1

pi Rww (j − i)

(6.10)


120

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊÓËÃѺ

j = 1, 2, . . . , N 

 Rww (1)

   Rww (2)   . .  .  Rww (N ) | {z



      =       }

«Ö §ÊÒÁÒö¨Ñ´ãËéÍÂÙèã¹ÃÙ»¢Í§àÁ·ÃÔ¡«ìä´é ¤×Í

|

Rww (0)

Rww (1)

Rww (1)

Rww (0)

. . .

. . .

Rww (N − 1)

···

r

···

Rww (N − 1)

 p1

  Rww (N − 2)   p2    .. . . . .  . . .  Rww (1) Rww (0) pN {z } | {z ···

      

(6.11)

}

p

R

ËÃ×Í

r = Rp à¹× ͧ¨Ò¡

R

(6.12)

à» ¹àÁ·ÃÔ¡«ì¨ÑµØÃÑÊ (square matrix) ´Ñ§¹Ñ ¹ ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ

p

ÊÒÁÒöËÒä´é¨Ò¡¡ÒÃá¡éÊÁ¡Òà (6.12) ¹Ñ ¹¤×Í

p = R−1 r áÅФèÒ ¢éͼԴ¾ÅÒ´¡ÓÅѧ Êͧà©ÅÕ Â ·Õ ¹éÍÂÊØ´

(6.13)

(MMSE: minimum mean squared error) ¢Í§Ç§¨Ã

¡Ãͧ·Ó¹Ò¨ÐÁÕ¤èÒà·èҡѺ [52]

N X £ 2¤ pi Rww (i) E ek = Rww (0) −

(6.14)

i=1

6.4

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML

à¾× ÍãËé§èÒµèÍ¡ÒÃ͸ԺÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML ãËé¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ PR4 µÒÁÃÙ»·Õ 6.4 â´Â·Õ ¢éÍÁÙÅ·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒ NPML ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»ÊÁ¡Òà ·Ò§¤³ÔµÈÒʵÃìä´é ¤×Í

yk = rk + wk = ak − ak−2 + wk àÁ× Í

rk = ak − ak−2

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³ áÅÐ

wk

(6.15)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ


6.4.

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML

121

wk ak

PR4 target

1− D

yk

rk

2

NPML detector

aˆk

ÃÙ»·Õ 6.4: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ PR4

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML ¨ÐµèÒ§¨Ò¡ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ã¹ àÃ× Í§¢Í§¡ÒäӹdzàÁµÃÔ¡ÊÒ¢Ò (branch metric) ¡ÅèÒǤ×Í àÁµÃÔ¡ÊҢҢͧǧ¨ÃµÃǨËÒ NPML ¨Ð ÁÕ¾¨¹ì·Õ à» ¹¤èÒ·Ó¹Ò¢ͧÊÑ­­Ò³Ãº¡Ç¹

ŵk

à¢éÒÁÒÃèÇÁ´éÇ ¹Ñ ¹¤×Í

λk (u, q) = |yk − r̂k (u, q) − ŵk |2 àÁ× Í

r̂k (u, q)

¤×Í ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³·Õ äÁèÁÕÊÑ­­Ò³Ãº¡Ç¹ (noiseless channel output) ·Õ

ÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡Ê¶Ò¹Ð ³ àÇÅÒ·Õ

k

(6.16)

u

ä»Ê¶Ò¹Ð

q

¶éÒ¡Ó˹´ãËé

·Õ ÊÍ´¤Åéͧ¡ÑºàÊé¹·Ò§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡Ê¶Ò¹Ð

u

ak (q)

ä»Ê¶Ò¹Ð

¤×Í ¢éÍÁÙźԵÍÔ¹¾Øµ

q

´Ñ§¹Ñ ¹ ÊÓËÃѺªèͧ

ÊÑ­­Ò³ PR4 ¨Ðä´éÇèÒ

r̂k (u, q) = ak (q) − ak−2 (q) ¹Í¡¨Ò¡¹Õ ¤èÒ ·Ó¹Ò¢ͧÊÑ­­Ò³Ãº¡Ç¹

ŵk

¤×Í

ŵk =

(6.17)

ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè ã¹ÃÙ» ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃì ä´é

N X

pi wk−i

(6.18)

i=1 á·¹¤èÒ

wk = yk − ak + ak−2

¨Ò¡ÊÁ¡Òà (6.15) ŧã¹ÊÁ¡Òà (6.18) ¨Ðä´é

ŵk =

N X

pi (yk−i − ak−i + ak−i−2 )

(6.19)

i=1 á·¹¤èÒ

r̂k (u, q)

¨Ò¡ÊÁ¡Òà (6.17) áÅÐ

ŵk

¨Ò¡ÊÁ¡Òà (6.19) ŧã¹ÊÁ¡Òà (6.16) ¨Ðä´éà» ¹

¯2 ¯ N ¯ ¯ X ¯ ¯ pi (yk−i − âk−i (q) + âk−i−2 (q))¯ λk (u, q) = ¯yk − ak (q) + ak−2 (q) − ¯ ¯ i=1

(6.20)


122

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

â´Â·Õ

âk (q)

¤×Í ¤èÒ»ÃÐÁÒ³¢Í§¢éÍÁÙźԵÍÔ¹¾Øµ ³ àÇÅÒ·Õ

(survivor path) ·Õ ÁҶ֧ʶҹÐ

k

·Õ ÊÍ´¤Åéͧ¡ÑºàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè

q

àÁµÃÔ¡ÊÒ¢Òã¹ÊÁ¡Òà (6.20) äÁèàËÁÒÐÊÓËÃѺ¡ÒùÓÁÒãªé¡Ñº§Ò¹»ÃÐÂØ¡µì (application) ·Õ µéͧ ¡ÒäÇÒÁàÃçÇ ã¹¡ÒûÃÐÁÇżÅÊÙ§ àªè¹ ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì à¹× ͧ¨Ò¡ ¡ÒäӹdzàÁµÃÔ¡ ÊÒ¢ÒÁÕ ¿ §¡ìªÑ¹ ¡Òäٳ (ÊÓËÃѺ¡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹) à¾Ô Á¢Ö ¹ÁÒ á·¹·Õ ¨ÐÁÕ੾Òп §¡ìªÑ¹¡Òúǡ ¡ÒÃà»ÃÕº à·Õº ¡ÒÃàÅ×Í¡ (ACS: add compare select) àËÁ×͹¡Ñº·Õ ãªéã¹ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔẺ¸ÃÃÁ´Ò ´Ñ§¹Ñ ¹ à¾× ÍãËéǧ¨ÃµÃǨËÒ NPML ÊÒÁÒö¹ÓÁÒãªé¡Ñº§Ò¹»ÃÐÂØ¡µì·Õ µéͧ¡ÒäÇÒÁàÃçÇ㹡ÒûÃÐÁÇżÅÊÙ§ ä´é ÊÁ¡Òà (6.20) ¨Ðµéͧ¶Ù¡¨Ñ´ÃÙ»ãËÁèãËéà» ¹

¯ ¯2 N +2 K ¯ ¯ X X ¯ ¯ λk (u, q) = ¯zk − ak−i (q)gi − ak (q)¯ âk−i (q)gi + ¯ ¯ i=1

i=K+1

àÁ× Í

K

(6.21)

¤×Í ¾ÒÃÒàµÍÃì·Õ ãªé㹡ÒûÃйջÃйÍÁÃÐËÇèÒ§¤ÇÒÁ«Ñº«é͹ (complexity) áÅлÃÐÊÔ·¸ÔÀÒ¾

¢Í§Ç§¨ÃµÃǨËÒ NPML ¡ÅèÒǤ×Í ¶éÒ

K

ÁÕ¤èÒÁÒ¡ ¤ÇÒÁ«Ñº«é͹¡ç¨ÐÁÒ¡ áµè»ÃÐÊÔ·¸ÔÀÒ¾·Õ ä´é¡ç¨Ð´Õ

(áÅÐã¹·Ò§µÃ§¡Ñ¹¢éÒÁ),

zk = yk −

N X

yk−i pi

(6.22)

i=1 ¤×Í ¢éÍÁÙÅàÍÒµì¾Øµ¢Í§Ç§¨Ã¡Ãͧ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ ¤×Í ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§ ·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

[1 − P (D)]

´Ñ§áÊ´§ã¹ÃÙ»·Õ 6.5, áÅÐ

(e ective target) ã¹â´àÁ¹

D

gi

«Ö §¹ÔÂÒÁâ´Â

Heff (D) = 1 − g1 D − g2 D2 − . . . − gN +2 DN +2 = (1 − D2 )[1 − P (D)]

(6.23)

â´ÂÊÃØ»áÅéÇ ¡ÒÃÊÃéҧǧ¨ÃµÃǨËÒ NPML ã¹·Ò§»¯ÔºÑµÔ ·Óä´é´Ñ§µèÍ仹Õ

1) ¤Ó¹Ç³ËÒǧ¨Ã¡Ãͧ·Ó¹ÒÂ

2) ¤Ó¹Ç³ËÒÅӴѺ ¢éÍÁÙÅ

{yk }

{zk }

·Õ ÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ

P (D)]

µÒÁÃÙ»·Õ 6.5

P (D)

â´ÂãªéÊÁ¡Òà (6.13)

¨Ò¡ÊÁ¡Òà (6.22) â´Â¹ÓÅӴѺ ¢éÍÁÙÅ àÍÒµì¾Øµ ¢Í§ÍÕ¤ÇÍäÅà«ÍÃì

H(D)

·Õ µéͧ¡Òà ÁÒ¼èҹǧ¨Ã¡Ãͧ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ

[1 −


6.4.

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML

123

yk

âk

zk Viterbi algorithm predictor

P(D) G(D)

ÃÙ»·Õ

6.5:

â¤Ã§ÊÃéÒ§ ¢Í§ ǧ¨Ã µÃǨËÒ

NPML

·Õ ãªé ¡Ñº §Ò¹ »ÃÐÂØ¡µì ·Õ µéͧ¡Òà ¤ÇÒÁ àÃçÇ ã¹ ¡ÒÃ

»ÃÐÁÇżÅÊÙ§

3) ¹ÓÅӴѺ ¢éÍÁÙÅ

{zk }

·Õ ä´é ä»·Ó¡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ Ẻ¸ÃÃÁ´Ò áµè

á¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ãªé 㹡ÒäӹdzµÒÁÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ (µÒÁ·Õ ͸ԺÒÂã¹ËÑÇ¢éÍ ·Õ 4.3.3) ¨ÐµéͧÊÃéÒ§¨Ò¡·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

Heff (D)

«Ö §ËÒä´é¨Ò¡

Heff (D) = H(D)[1 − P (D)] àÁ× Í

H(D)

(6.24)

¤×Í ·ÒÃìà¡çµ·Õ ÊÍ´¤Åéͧ¡Ñº ÍÕ¤ÇÍäÅà«ÍÃì ·Õ ãªé ã¹Ãкº ¨ÐàËç¹ä´éÇèÒ ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§

·ÒÃìà¡çµ »ÃÐÊÔ·¸Ô¼Å¨Ðà» ¹ àÅ¢¨Ó¹Ç¹¨ÃÔ§ ¡ÅèÒǤ×Í ¶Ö§áÁéÇèÒ ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§ ¨Ó¹Ç¹àµçÁ áµè ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§

P (D)

H(D)

à» ¹ àÅ¢ ¨Ó¹Ç¹¨ÃÔ§ ´Ñ§¹Ñ ¹ ¼ÅÅѾ¸ì ·Õ ä´é

¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô à» ¹ àÅ¢¨Ó¹Ç¹¨ÃÔ§ à¾ÃÒÐ©Ð¹Ñ ¹ ÍÒ¨¨Ð¡ÅèÒÇä´é ÇèÒ ·ÒÃìà¡çµ »ÃÐÊÔ·¸Ô¼Å

¨Ðà» ¹ àÅ¢

Heff (D) Heff (D)

¨ÐÁÕ ¤×Í

·ÒÃìà¡çµáºº GPR áººË¹Ö §¡çä´é

µÑÇÍÂèÒ§·Õ 6.1

¨Ò¡¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì µÒÁẺ¨ÓÅͧã¹ÃÙ»·Õ 3.2 ÊÓËÃѺÃкº

¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ ·Õ ND = 2 áÅÐ SNR = 15 dB â´Â¡Ó˹´ãËé ·ÒÃìà¡çµ ·Õ µéͧ¡Òà ¤×Í

H(D) = 1 − D −0.66}

»ÃÒ¡®ÇèÒ

ÅӴѺ¢éͼԴ¾ÅÒ´

{wk }

·Õ ä´é ¤×Í

{0.86, −0.26, −0.13, −0.14, 0.35,

¨§¤Ó¹Ç³ËÒ

¡) ǧ¨Ã¡Ãͧ·Ó¹ÒÂ

P (D)

Ẻ 2 á·ç» áÅзÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

Heff (D)


124

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¢) ǧ¨Ã¡Ãͧ·Ó¹ÒÂ

ÇÔ¸Õ·Ó

P (D)

¨Ò¡ÅӴѺ¢éͼԴ¾ÅÒ´

Ẻ 4 á·ç» áÅзÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

{wk }

Heff (D)

·Õ ¡Ó˹´ãËé ¤èÒÍѵÊËÊÑÁ¾Ñ¹¸ì¢Í§

wk

ÊÒÁÒöËÒä´é¨Ò¡ÊÁ¡ÒÃ

(6.9) ´Ñ§¹Õ

Rww (i) = {0.2336, −0.0903, −0.0071, −0.0419, 0.2363, −0.5676} ÊÓËÃѺ

i = 0, 1, 2, . . . , 5

µÒÁÅӴѺ

¡) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» ÊÒÁÒöËÒä´é¨Ò¡¡ÒÃá¡éÊÁ¡Òà (6.11) ¹Ñ ¹¤×Í

 

 −0.0903

=

−0.0071

 0.2336

−0.0903

−0.0903

0.2336



 p1

p2

¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò ÁÕ¤èÒà·èҡѺ

 

 p1 p2

 = 

−1  0.2336

−0.0903

 −0.0903

0.2336 −0.0071   5.0323 1.9454 −0.0903   =  1.9454 5.0323 −0.0071   −0.4684  =  −0.2116 

−0.0903

´Ñ§¹Ñ ¹ ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» ¤×Í

P (D) = −0.4684D − 0.2116D2 áÅзÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å·Õ ÊÍ´¤Åéͧ¡ÑºÇ§¨Ã¡Ãͧ·Ó¹ÒÂ¹Õ ¤×Í

£ ¤ Heff (D) = (1 − D) 1 − (−0.4684D − 0.2116D2 ) = 1 − 0.5316D − 0.2568D2 − 0.2116D3 â´Â¨Ó¹Ç¹Ê¶Ò¹Ðã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ ãªéã¹Ç§¨ÃµÃǨËÒ NPML ¨ÐÁÕ·Ñ §ËÁ´

23 = 8

ʶҹÐ


6.4.

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML

125

¢) 㹷ӹͧà´ÕÂǡѹ ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 4 á·ç» ¡çÊÒÁÒöËÒä´é¨Ò¡¡ÒÃá¡éÊÁ¡Òà (6.11) ¹Ñ ¹¤×Í

 −0.0903

 0.2336

−0.0903 −0.0071 −0.0419

 p1

           −0.0071   −0.0903 0.2336 −0.0903 −0.0071   p2   =         −0.0419   −0.0071 −0.0903 0.2336 −0.0903   p3       0.2363 −0.0419 −0.0071 −0.0903 0.2336 p4 ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò ÁÕ¤èÒà·èҡѺ

 p1

−1  0.2336

−0.0903 −0.0071 −0.0419

           p2   −0.0903 0.2336 −0.0903 −0.0071     =          p3   −0.0071 −0.0903 0.2336 −0.0903        p4 −0.0419 −0.0071 −0.0903 0.2336   5.9512 3.2140 2.2035 2.0163 −0.0903      3.2140 7.0038 3.6575 2.2035   −0.0071   =    2.2035 3.6575 7.0038 3.2140   −0.0419   2.0163 2.2035 3.2140 5.9512 0.2363   −0.1762      0.0274    =    0.2412    1.0739

 −0.0903

  −0.0071    −0.0419   0.2363        

´Ñ§¹Ñ ¹ ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 4 á·ç» ¤×Í

P (D) = −0.1762D + 0.0274D2 + 0.2412D3 + 1.0739D4 áÅзÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å·Õ ÊÍ´¤Åéͧ¡ÑºÇ§¨Ã¡Ãͧ·Ó¹ÒÂ¹Õ ¤×Í

£ ¤ Heff (D) = (1 − D) 1 − (−0.1762D + 0.0274D2 + 0.2412D3 + 1.0739D4 ) = 1 − 0.8238D − 0.2036D2 − 0.2138D3 − 0.8327D4 + 1.0739D5


126

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

wk

target

ak

H(D)

yk

rk

NPML

zk

Viterbi

âk

G(D)

P(D) PRML

âk

Viterbi H(D)

ÃÙ»·Õ 6.6: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³áººÊÁÁÙÅ ¾ÃéÍÁ·Ñ §Ç§¨ÃµÃǨËÒ NPML áÅÐ PRML

â´Â¨Ó¹Ç¹Ê¶Ò¹Ðã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ ãªéã¹Ç§¨ÃµÃǨËÒ NPML ¨ÐÁÕ·Ñ §ËÁ´

25 = 32

ʶҹÐ

Êѧࡵ¨Ð¾ºÇèÒ Ç§¨ÃµÃǨËÒ NPML ¨ÐÁÕ ¤ÇÒÁ«Ñº«é͹ÁÒ¡¡ÇèÒ Ç§¨ÃµÃǨËÒ PRML ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ·ÒÃìà¡çµ »ÃÐÊÔ·¸Ô¼Å

Heff (D)

¨ÐÁÕ ¨Ó¹Ç¹á·ç» ÁÒ¡¢Ö ¹ ¡ÇèÒ ·ÒÃìà¡çµ »¡µÔ

H(D)

«Ö § à» ¹ ¼Å

ÁÒ¨Ò¡¨Ó¹Ç¹á·ç» ¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò ËÃ×Í ÍÒ¨¨Ð¡ÅèÒÇä´é ÇèÒ ¨Ó¹Ç¹Ê¶Ò¹Ðã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ãªéèã¹Ç§¨ÃµÃǨËÒ NPML Áըӹǹà·èҡѺ

¨Ó¹Ç¹Ê¶Ò¹Ð =

àÁ× Í »¡µÔ

|A|

|A|ν+N

á·¹¨Ó¹Ç¹¢éÍÁÙÅ ºÔµ ÍÔ¹¾Øµ ·Õ à» ¹ ä»ä´é ·Ñ §ËÁ´,

H(D),

áÅÐ

N

ν

¤×Í ¨Ó¹Ç¹Ë¹èǤÇÒÁ¨Ó¢Í§·ÒÃìà¡çµ

¤×Í ¨Ó¹Ç¹á·ç» ¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò ÍÂèÒ§äáçµÒÁ ¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨Ã

µÃǨËÒ NPML ÊÒÁÒö·ÓãËéŴŧä´éµÒÁ·Õ àʹÍã¹ [53]

µÑÇÍÂèÒ§·Õ 6.2

Ẻ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ áÅÐÍÕ¤ÇÍäÅà«ÍÃì ã¹ÃÙ» ·Õ 3.2 ÊÓËÃѺ Ãкº¡ÒÃ

ºÑ¹·Ö¡áººá¹Ç¹Í¹ ·Õ ND = 2 áÅÐ SNR = 15 dB ÊÒÁÒöŴÃÙ»ä´éà» ¹ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ ẺÊÁÁÙÅ µÒÁÃÙ» ·Õ 6.6 àÁ× Í ·ÒÃìà¡çµ ·Õ µéͧ¡Òà ¤×Í

H(D) = 1 − D

¶éÒ ¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ


6.4.

ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ NPML

ÍÔ¹¾Øµ

{ak } = {−1, 1, 1, 1}

¨§¶Í´ÃËÑÊ¢éÍÁÙÅ

yk

127

áÅÐÊÑ­­Ò³Ãº¡Ç¹

{wk } = {0.46, −1.20, 1.02, −0.59, −0.98}

´éÇÂ

¡) ǧ¨ÃµÃǨËÒ PRML

¢) ǧ¨ÃµÃǨËÒ NPML àÁ× Íãªéǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 1 á·ç» ¤×Í

ÇÔ¸Õ·Ó

¨Ò¡ÃÙ»·Õ 6.6 ¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

rk

P (D) = −0.753D

ËÒä´é¨Ò¡

rk = ak ∗ hk = {−1, 2, 0, 0, −1} ´Ñ§¹Ñ ¹ ¢éÍÁÙÅ·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒ PRML áÅÐ NPML ¤×Í

yk = rk + wk = {−0.54, 0.80, 1.02, −0.59, −1.98} ¡)

ÊÓËÃѺ Ãкº PRML ǧ¨Ã µÃǨËÒ ÇÕà·ÍÃìºÔ ¨Ð ·Ó ¡Òà ¶Í´ÃËÑÊ ¢éÍÁÙÅ

à·ÃÅÅÔÊ ·Õ ÊÃéÒ§¨Ò¡·ÒÃìà¡çµ ¢éÍÁÙÅ

{yk }

H(D) = 1 − D

{yk }

â´Â ãªé á¼¹ÀÒ¾

µÒÁ·Õ áÊ´§ã¹ÃÙ» ·Õ 6.7(a) «Ö § ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ

ÊÒÁÒöÊÃØ»ä´é µÒÁÃÙ»·Õ 6.8 â´Â·Õ µÑÇàÅ¢·Õ áÊ´§ÍÂÙ躹¨Ø´µèÍ (node) áµèÅШش ¤×Í ¤èÒ

àÁµÃÔ¡àÊé¹·Ò§·Õ ÁÒ¶Ö§ ³ ¨Ø´µèÍ¹Ñ ¹ áÅеÑÇàÅ¢·Õ áÊ´§ÍÂÙ躹àÊé¹ÊÒ¢ÒáµèÅÐàÊé¹ ¤×Í ¤èÒàÁµÃÔ¡ÊÒ¢Ò ¢Í§áµèÅÐàÊé¹ ÊÒ¢Ò·Õ ´Õ ·Õ ÊØ´ ·Õ ÁÒ¶Ö§ ·Õ ¨Ø´µèÍ ¹Ñ ¹æ à¾ÃÒÐ©Ð¹Ñ ¹ ¨Ò¡ÃÙ» ·Õ 6.8 ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§·Õ ¹éÍ ·Õ ÊØ´ ¤×Í ¤èÒ 2.24 ´Ñ§¹Ñ ¹Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð¶Í´ÃËÑÊ¢éÍÁÙÅâ´Â¡ÒÃÁͧÂé͹¡ÅѺ仵ÒÁàÊé¹·Ò§ ·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè (survivor path) ·Õ ÁÒ¶Ö§ ³ ¨Ø´µèÍ ·Õ ÁÕ ¤èÒ àÁµÃÔ¡ àÊé¹·Ò§à·èÒ ¡Ñº 2.24 «Ö § ¨Ð¾ºÇèÒ ¤èÒ »ÃÐÁÒ³¢Í§ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{âk }

·Õ ÊÍ´¤Åéͧ¡ÑºàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè¹Õ ¤×Í

{âk } = {â0 , â1 , â2 , â3 } = {−1, −1, 1, 1} «Ö §ÁÕ¤èÒäÁèµÃ§¡ÑºÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{ak } = {−1, 1, 1, 1}

·Õ Êè§ÁÒ¨Ò¡µé¹·Ò§ ´Ñ§¹Ñ ¹ ¡ÒöʹÃËÑÊ

´éÇÂǧ¨ÃµÃǨËÒ PRML ã¹¡Ã³Õ¹Õ ÁÕ¢éͼԴ¾ÅÒ´à¡Ô´¢Ö ¹à» ¹¨Ó¹Ç¹ 1 ºÔµ

¢)

ÊÓËÃѺÃкº NPML ÅӴѺ¢éÍÁÙÅ

{yk }

¨Ð¶Ù¡Ê觼èÒ¹à¢éÒä»ã¹Ç§¨Ã¡Ãͧ㹡Ò÷ÓãËéÊÑ­­Ò³

ú¡Ç¹à» ¹ÊÕ¢ÒÇ (noise whitening lter) â´Â¨Ðä´é¼ÅÅѾ¸ìÍÍ¡ÁÒà» ¹ ËÃ×ÍáÊ´§à» ¹ÅӴѺ¢éÍÁÙÅ

{zk }

Z(D) = Y (D)[1 − P (D)]

ã¹â´àÁ¹àÇÅÒ ä´é´Ñ§¹Õ

{zk } = {−0.5400, 0.3934, 1.6224, 0.1781, −2.4243, −1.4910}


128

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0

-1

0

-1 -1

2

2 -2

1

-1.5061

1 -1 0

0.4939

(a)

-0.4939

-1 1

ak = -1

1.5061

-2

ak = 1

1 1

0

(b) ÃÙ»·Õ 6.7: á¼¹ÀÒ¾à·ÃÅÅÔʢͧ (a) ·ÒÃìà¡çµ =

1 − 0.247D − 0.753D2

-1

0

0.29

0.29

H(D)

=

1−D

áÅÐ (b) ·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

·Õ ãªé㹡ÒöʹÃËÑÊ¢éÍÁÙŢͧÃкº PRML áÅÐ NPML µÒÁÅӴѺ

0.64

0.93

1.04

1.97

0.35

2.32

0.96

1

0

0.29

0.29

Heff (D)

0.64

0.93

2.24 0

1.89

0.35

2.24

3.92

6.16

ÃÙ»·Õ 6.8: á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÓËÃѺÃкº PRML

¨Ò¡¹Ñ ¹ ÅӴѺ ¢éÍÁÙÅ

{zk }

¨Ð¶Ù¡ ¶Í´ÃËÑÊ ¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Õ ·Ó§Ò¹â´Âãªé á¼¹ÀÒ¾

à·ÃÅÅÔÊ·Õ ÊÃéÒ§¨Ò¡·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

Heff (D) = H(D)[1 − P (D)] = 1 − 0.247D − 0.753D2

µÒÁ·Õ áÊ´§ã¹ÃÙ» ·Õ 6.6(b) â´Â·Õ ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ ¢éÍÁÙÅ

{zk }

ÊÒÁÒöÊÃØ» ä´é µÒÁÃÙ» ·Õ 6.9


6.5.

-1 -1

¼Å¡Ò÷´Åͧ

0

0.29

129

0.29

0.15

0.45

2.63

3.08

0.03

3.11

0.24

1.88

0.14 0.84

1 -1

0

1.07 0

0.59

0.01 0.79

4.59

9.56

0

4.18

0.45 4.48

-1 1

0 1.07

0

1.86

0.01

11

0

0.29

0.29

0.15

0.45

6.19 1.04

0.01 0.03

0.03

3.94

8.52

4.49 0.10

0.06

0.18 5.88

0.24

5.93

0.26

2.22

8.16

ÃÙ»·Õ 6.9: á¼¹ÀÒ¾ÊÃØ»¢Ñ ¹µÍ¹¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔÊÓËÃѺÃкº NPML

à¹× ͧ¨Ò¡ ¤èÒàÁµÃÔ¡ àÊé¹·Ò§·Õ ¹éÍÂ·Õ ÊØ´ ¤×Í ¤èÒ 0.24 à¾ÃÒÐ©Ð¹Ñ ¹ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨Ð¶Í´ÃËÑÊ ¢éÍÁÙÅâ´Â¡ÒÃÁͧÂé͹¡ÅѺ仵ÒÁàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè·Õ ÁÒ¶Ö§ ³ ¨Ø´µèÍ·Õ ÁÕ¤èÒàÁµÃÔ¡àÊé¹·Ò§à·èҡѺ 0.24 «Ö §¨Ð¾ºÇèÒ ¤èÒ»ÃÐÁÒ³¢Í§ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{âk }

·Õ ÊÍ´¤Åéͧ¡ÑºàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè¹Õ ¤×Í

{âk } = {â0 , â1 , â2 , â3 } = {−1, 1, 1, 1} «Ö §ÁÕ¤èҵç¡ÑºÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

{ak } = {−1, 1, 1, 1}

·Õ Êè§ÁÒ¨Ò¡µé¹·Ò§ ´Ñ§¹Ñ ¹ ¡ÒöʹÃËÑÊ´éÇÂ

ǧ¨ÃµÃǨËÒ NPML ã¹µÑÇÍÂèÒ§¢éÍ¹Õ ¨Ö§äÁèÁÕ¢éͼԴ¾ÅÒ´à¡Ô´¢Ö ¹

6.5

¼Å¡Ò÷´Åͧ

ã¹Êèǹ¹Õ ¨Ð·Ó¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨ÃµÃǨËÒ PRML áÅÐ NPML â´ÂãªéẺ¨ÓÅͧ ªèͧÊÑ­­Ò³¢Í§Ãкº¡Òúѹ·Ö¡ Ẻá¹Ç¹Í¹ (longitudinal recording) µÒÁÃÙ» ·Õ 6.10 â´Â·Õ


130

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

n(t)

ak

1− D {±1} 2

bk

p(t)

g(t)

sk

s(t) LPF

yk

equalizer

detector

âk

t k = kT ÃÙ»·Õ 6.10: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³¢Í§Ãкº¡Òúѹ·Ö¡áÁèàËÅç¡

ÊÑ­­Ò³ read back ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Òä³ÔµÈÒʵÃìä´é ¤×Í

p(t) =

S−1 X

bk g(t − kT ) + n(t)

(6.25)

k=0

bk = (ak − ak−1 )/2

àÁ× Í

ºÇ¡ËÃ×Íź áÅÐ ·Ñ §ËÁ´

bk = 0

S = 4096

(1.1), áÅÐ

n(t)

´éÒ¹à·èҡѺ

N0 /2

¤×Í ºÔµà»ÅÕ Â¹Ê¶Ò¹Ð (bk

= ±1

ÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹Ð

ËÁÒ¶֧ äÁèÁÕ¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹Ð),

ºÔµ ËÃ×Í 1 à«¡àµÍÃì (sector),

g(t)

ak ∈ ±1

¤×Í ºÔµÍÔ¹¾Øµ·Õ Áըӹǹ

¤×Í ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð µÒÁÊÁ¡ÒÃ

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺºÇ¡·Õ ÁÕ¤ÇÒÁ˹Òá¹è¹Ê໡µÃÑÁ¡ÓÅѧẺÊͧ

ÊÑ­­Ò³ read back

p(t)

¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧǧ¨Ã¡Ãͧ¼èÒ¹µ Ӻѵà·ÍÃìàÇÔÃìµÍѹ´Ñº·Õ 7 áÅж١·Ó

¡ÒêѡµÑÇÍÂèÒ§´éǤÇÒÁ¶Õ ¡ÒêѡµÑÇÍÂèÒ§à·èҡѺ

1/T

â´ÂÊÁÁصÔÇèÒ ¡Ãкǹ¡ÒÃ㹡ÒêѡµÑÇÍÂèÒ§ÁÕ

¡ÒÃà¢éҨѧËÇÐÃÐËÇèÒ§ÊÑ­­Ò³ read back áÅÐǧ¨ÃªÑ¡µÑÇÍÂèҧẺÊÁºÙóì (perfect synchroniza tion) ¨Ò¡¹Ñ ¹ ÅӴѺ¢éÍÁÙÅàÍÒµì¾Øµ

{sk }¨Ð¶Ù¡» ͹ä»ÂѧÍÕ¤ÇÍäÅà«ÍÃì (equalizer) à¾× Í»ÃѺÃÙ»ÃèÒ§¢Í§

ÊÑ­­Ò³ãËéà» ¹ä»µÒÁ·ÒÃìà¡çµ·Õ µéͧ¡Òà áÅéÇ¡çÊè§ÅӴѺ¢éÍÁÙÅàÍÒµì¾Øµ

{yk }

·Õ ä´é ä»·Ó¡ÒöʹÃËÑÊ

¢éÍÁÙÅ ´éÇÂǧ¨ÃµÃǨËÒ (detector) à¾× Í ËÒ¤èÒ »ÃÐÁÒ³¢Í§ÅӴѺ ¢éÍÁÙÅ ÍÔ¹¾Øµ

{ak }

·Õ à» ¹ ä»ä´é ÁÒ¡

·Õ ÊØ´ ã¹·Õ ¹Õ ¤èÒ SNR ·Õ ãªé¨Ð¹ÔÂÒÁâ´Â

µ SNR = 10 log10 àÁ× Í

Vp = 1

áÅÐ

σ2

=

Vp 2 σ2

¶ (dB)

¤×Í ¢¹Ò´ÊÙ§ÊØ´¢Í§ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È

N0 /(2T )

¤×Í ¡ÓÅѧ¢Í§ÊÑ­­Ò³Ãº¡Ç¹

n(t)

(6.26)

(isolated transition pulse)

¹Í¡¨Ò¡¹Õ áµèÅШش¢Í§ÍѵÃÒ¢éͼԴ¾ÅÒ´


6.5.

¼Å¡Ò÷´Åͧ

131

−1

10

−2

BER

10

−3

10

−4

10

PRML: 22 states NPML (2−tap predictor): 24 states NPML (4−tap predictor): 26 states

−5

10

10

11

12

13

14

15

16

SNR (dB)

ÃÙ»·Õ 6.11: »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2

ºÔµ (BER) ¨Ð¶Ù¡ ¤Ó¹Ç³â´Âãªé ¢éÍÁÙÅ ËÅÒÂæ à«¡àµÍÃì ¨¹¡ÇèÒ ¨Ðä´é ¢éͼԴ¾ÅÒ´ºÔµ ÁÒ¡¡ÇèÒ ËÃ×Í à·èÒ ¡Ñº 1000 ºÔµ ÃÙ» ·Õ 6.11 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨ÃµÃǨËÒ PRML áÅÐ NPML ·Õ ND = 2 àÁ× Í ¡Ó˹´ãËé·Ø¡Ãкºãªé·ÒÃìà¡çµáºº PR4,

H(D) = 1 − D2

ã¹á¼¹ÀÒ¾à·ÃÅÅÔʢͧǧ¨ÃµÃǨËÒ PRML ¤×Í

22 = 4

·Õ ãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔʢͧǧ¨ÃµÃǨËÒ NPML ¤×Í áºº 2 á·ç») áÅÐ

22+4 = 64

à¾ÃÒÐ©Ð¹Ñ ¹ ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´·Õ ãªé ʶҹР㹢³Ð·Õ ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´

22+2 = 16

ʶҹР(ÊÓËÃѺǧ¨Ã¡Ãͧ·Ó¹ÒÂ

ʶҹР(ÊÓËÃѺǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 4 á·ç») ´Ñ§¹Ñ ¹¨ÐàËç¹ä´éÇèÒ

ǧ¨ÃµÃǨËÒ NPML ÁÕ ¤ÇÒÁ«Ñº«é͹ÁÒ¡¡ÇèÒ Ç§¨ÃµÃǨËÒ PRML áµè ¨Ò¡¼Å¡Ò÷´ÅͧµÒÁÃÙ» ·Õ 6.11 ¾ºÇèÒ Ç§¨ÃµÃǨËÒ NPML ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡¡ÇèÒǧ¨ÃµÃǨËÒ PRML ÍÂèÒ§àËç¹ä´éªÑ´ ËÃ×Í ÍÒ¨¨Ð¡ÅèÒÇä´éè ÇèÒ ³ ÃдѺ BER =

10−4

ǧ¨ÃµÃǨËÒ NPML ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾´Õ ¡ÇèÒ Ç§¨ÃµÃǨËÒ

PRML »ÃÐÁÒ³ 2 dB ¹Í¡¨Ò¡¹Õ ¨Ò¡¼Å¡Ò÷´ÅͧÂѧ¾ºÇèÒ Ç§¨ÃµÃǨËÒ NPML ·Õ ãªéǧ¨Ã¡Ãͧ ·Ó¹Ò 2 á·ç» ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ã¡Åéà¤Õ§¡ÑºÇ§¨Ã¡Ãͧ·Ó¹Ò 4 á·ç» ´Ñ§¹Ñ ¹ ÊÓËÃѺÃкº·Õ ¾Ô¨ÒóҹÕ


132

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

0.078 0.076

Predictor MMSE

0.074 0.072 0.07 0.068 0.066 0.064 0.062

1

2

3

4

5

6

7

8

9

10

Number of predictor taps

ÃÙ»·Õ 6.12: »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ·Õ ãªé¨Ó¹Ç¹á·ç»µèÒ§¡Ñ¹ ·Õ SNR = 17 dB

ǧ¨ÃµÃǨËÒ NPML ÊÒÁÒöãªéǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» ¡çà¾Õ§¾ÍµèÍ¡ÒÃãªé§Ò¹áÅéÇ à¹× ͧ¨Ò¡ ãËé»ÃÐÊÔ·¸ÔÀÒ¾·Õ ã¡Åéà¤Õ§¡Ñº¡ÒÃãªéǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 4 á·ç» áµè¤ÇÒÁ«Ñº«é͹¨Ð¹éÍ¡ÇèÒÁÒ¡ ÃÙ»·Õ 6.12 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂ·Õ ãªé¨Ó¹Ç¹á·ç»µèÒ§¡Ñ¹ ·Õ SNR = 17 dB ¨ÐàËç¹ä´éÇèÒ Ç§¨Ã¡Ãͧ·Ó¹Ò 2 á·ç» ¡çÁÕ»ÃÐÊÔ·¸ÔÀÒ¾à¾Õ§¾ÍÊÓËÃѺ¡ÒÃãªé§Ò¹áÅéÇ à¹× ͧ¨Ò¡ ¶Ö§áÁéÇèÒ¨Ðà¾Ô Á¨Ó¹Ç¹á·ç»ÁÒ¡¢Ö ¹ »ÃÐÊÔ·¸ÔÀÒ¾·Õ ä´é¡çà¾Ô Á¢Ö ¹¹éÍÂÁÒ¡ «Ö §äÁè¤ØéÁ¤èҡѺ¤ÇÒÁ«Ñº«é͹·Õ ä´éÃѺ 㹷ӹͧà´ÕÂǡѹÃÙ»·Õ 6.13 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨ÃµÃǨËÒ PRML áÅÐ NPM L ·Õ ND = 2.5 â´Âãªé·ÒÃìà¡çµáºº PR4 àËÁ×͹à´ÔÁ áµè¤ÃÒÇ¹Õ ¨Ð¾ºÇèÒ Ç§¨ÃµÃǨËÒ NPML ãËé »ÃÐÊÔ·¸ÔÀÒ¾·Õ ´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRML ¤è͹¢éÒ§ÁÒ¡ àÁ× Íà·Õº¡Ñº¡Ò÷ӧҹ·Õ ND = 2 â´ÂÍÒ¨¨Ð ¡ÅèÒÇä´éèÇèÒ ³ ÃдѺ BER =

10−4

ǧ¨ÃµÃǨËÒ NPML ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRML

»ÃÐÁÒ³ 3 dB ·Ñ §¹Õ à» ¹à¾ÃÒÐÇèÒ ·Õ ¤ÇÒÁ¨Ø¢éÍÁÙŢͧÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÙ§ (ËÃ×Í ND ÊÙ§) ͧ¤ì»ÃСͺ ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙè ã¹¢éÍÁÙÅ ·Õ ¨Ð·Ó¡ÒöʹÃËÑÊ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨ÐÁÕ ÅÑ¡É³Ð à» ç¹ ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ ÁÒ¡¢Ö ¹ ¨Ö§ ·ÓãËé¡ÒÃãªé §Ò¹Ç§¨ÃµÃǨËÒ NPML ä´é »ÃÐÊÔ·¸ÔÀÒ¾·Õ ´Õ ¡ÇèÒ


6.5.

¼Å¡Ò÷´Åͧ

133

−1

10

−2

BER

10

−3

10

−4

10

PRML: 22 states NPML (2−tap predictor): 24 states −5

10

12

13

14

15

16

17

SNR (dB)

ÃÙ»·Õ 6.13: »ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2.5

¡ÒÃãªé§Ò¹Ç§¨ÃµÃǨËÒ PRML ÁÒ¡ à¾× ÍãËé ¡ÒÃà»ÃÕºà·Õº໠¹ ä»ÍÂèÒ§ÂصԸÃÃÁã¹àÃ× Í§¢Í§¤ÇÒÁ«Ñº«é͹¢Í§Ãкº ¨Ð·Ó¡Ò÷´Åͧ à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº 3 Ãкº ·Õ ND = 2.5 ´Ñ§µèÍ仹Õ

1) ǧ¨ÃµÃǨËÒ NPML ·Õ ãªé·ÒÃìà¡çµáºº PR4

H(D) = 1 − D2

áÅÐǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2

á·ç»

2) ǧ¨ÃµÃǨËÒ PRML ·Õ ãªé·ÒÃìà¡çµáºº PR 2

3) ǧ¨ÃµÃǨËÒ GPRML

H(D) = 1 + 2D − 2D3 − D4

·Õ ãªé ·ÒÃìà¡çµ Ẻ GPR «Ö § Í͡Ẻâ´Âà§× ͹䢺ѧ¤Ñº ẺâÁ¹Ô¡ (´Ù

ÃÒÂÅÐàÍÕ´ã¹ËÑÇ¢éÍ·Õ 3.2.1) â´Â·Õ ·ÒÃìà¡çµáºº GPR ¹Õ ¨Ð¶Ù¡Í͡ẺÊÓËÃѺáµèÅÐ SNR

µÒÁÃÙ» ·Õ 6.14 àÁ× Íá¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ãªé 㹡ÒöʹÃËÑÊ ¢éÍÁÙÅ ´éÇÂÍÑÅ¡ÍÃÔ·ÖÁ ÇÕà·ÍÃìºÔ ¢Í§·Ø¡ Ãкº 2

ǧ¨ÃµÃǨËÒ GPRML ¤×Í Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ·Õ ãªé ·ÒÃìà¡çµ Ẻ GPR ã¹¢³Ð·Õ ǧ¨ÃµÃǨËÒ PRML ¤×Í Ç§¨Ã

µÃǨËÒÇÕà·ÍÃìºÔ·Õ ãªé·ÒÃìà¡çµáºº PR


134

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

−1

10

2+2

NPML: 2 = 16 states PRML: [1 2 0 −2 −1] GPRML: (5−tap GPR) −2

BER

10

−3

10

−4

10

−5

10

12

13

14

15

16

17

SNR (dB)

ÃÙ»·Õ 6.14: ¼Å¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃкºµèÒ§æ ·Õ ND = 2.5

¨ÐÁÕ ¨Ó¹Ç¹Ê¶Ò¹Ðà·èÒ ¡Ñ¹ ¤×Í 16 ʶҹРáÅÐÍÕ¤ÇÍäÅà«ÍÃì ·Õ ãªé ¢Í§áµèÅÐÃкº¨Ð¶Ù¡ Í͡ẺãËé àËÁÒÐÊÁ¡Ñº·ÒÃìà¡çµ

H(D)

·Õ ¡Ó˹´ ¨Ò¡ÃÙ»¨ÐàËç¹ä´éÇèÒ Ç§¨ÃµÃǨËÒ NPML ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ÁÒ¡

¡ÇèÒ Ç§¨ÃµÃǨËÒ PRML ·Õ ãªé ·ÒÃìà¡çµ Ẻ PR áµè ÁÕ »ÃÐÊÔ·¸Ô ã¡Åéà¤Õ§¡Ñº ǧ¨ÃµÃǨËÒ GPRML ·Ñ §¹Õ à» ¹ à¾ÃÒÐÇèÒ ·ÒÃìà¡çµ »ÃÐÊÔ·¸Ô¼Å

Heff (D) = H(D)[1 − P (D)]

·Õ ãªé 㹡ÒÃÊÃéҧἹÀÒ¾

à·ÃÅÅÔʢͧǧ¨ÃµÃǨËÒ NPML ÊÒÁÒö·Õ ¨Ð¶Ù¡¾Ô¨ÒóÒä´éÇèÒà» ¹·ÒÃìà¡çµáºº GPR áººË¹Ö §ä´é à¹× ͧ¨Ò¡ ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Åà» ¹àÅ¢¨Ó¹Ç¹¨ÃÔ§

6.6

ÊÃØ»·éÒº·

àÁ× Í Ãкº·Ó§Ò¹·Õ ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÊÙ§ ͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ¨ÐÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ (colored noise) ÁÒ¡¢Ö ¹ ã¹¡Ã³Õ ¹Õ ǧ¨ÃµÃǨËÒ PRML äÁè ÊÒÁÒö·Ó§Ò¹ä´é ÍÂèÒ§ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ ´Ñ§¹Ñ ¹Ç§¨ÃµÃǨËÒ NPML ¨Ö§


6.7.

à຺½ ¡ËÑ´·éÒº·

135

ä´é ¶Ù¡ ¹ÓÁÒãªé à¾× Í à¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº ·Ñ §¹Õ à¹× ͧÁÒ¨Ò¡ÇèÒǧ¨ÃµÃǨËÒ NPML ¨ÐÁÕ ¡ÒÃãªé ǧ¨Ã¡Ãͧ·Ó¹Ò (㹡ÒÃ·Õ ¨Ð·ÓãËéÊÑ­­Ò³Ãº¡Ç¹áººÊÕ¡ÅÒÂà» ¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇ) ÃèÇÁ¡Ñº ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ㹡ÒöʹÃËÑÊ ¢éÍÁÙÅ ¹Í¡¨Ò¡¹Õ Âѧ ¾ºÇèÒ·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å·Õ ãªé 㹡ÒÃÊÃéÒ§ á¼¹ÀÒ¾à·ÃÅÅÔÊ ¢Í§Ç§¨ÃµÃǨËÒ NPML ÊÒÁÒö·Õ ¨Ð¶Ù¡ ¾Ô¨ÒóÒä´é ÇèÒ à» ¹ ·ÒÃìà¡çµ Ẻ GPR ¨Ö§ à» ¹à˵ؼŢéÍË¹Ö §ÇèÒ·ÓäÁǧ¨ÃµÃǨËÒ NPML ¨Ö§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRML à¾ÃÒÐÇèÒ ·ÒÃìà¡çµáºº GPR ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒ·ÒÃìà¡çµáºº PR µÒÁ·Õ ͸ԺÒÂ㹺··Õ 3 ¶Ö§áÁéÇèÒǧ¨ÃµÃǨËÒ NPML ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRML áµèǧ¨ÃµÃǨËÒ NPML ¨ÐÁÕ¤ÇÒÁ«Ñº«é͹ÁÒ¡¡ÇèÒ à¾ÃÒÐ©Ð¹Ñ ¹ 㹡ÒõѴÊÔ¹ã¨ÇèҨйÓǧ¨ÃµÃǨËÒ NPML ÁÒãªé§Ò¹ËÃ×Í äÁè ãËé¾Ô¨ÒóÒÇèÒ »ÃÐÊÔ·¸ÔÀÒ¾·Õ ¨Ðä´éÃѺà¾Ô Á¢Ö ¹¨Ð¤ØéÁ¤èҡѺ¤ÇÒÁ«Ñº«é͹·Õ µÒÁÁÒËÃ×ÍäÁè

6.7

à຺½ ¡ËÑ´·éÒº·

1. ¨§Í¸ÔºÒÂ·Õ ÁҢͧá¹Ç¤Ô´¢Í§Ç§¨ÃµÃǨËÒ NPML

2. ¨§Í¸ÔºÒ¤ÇÒÁᵡµèÒ§¢Í§Ç§¨ÃµÃǨËÒ PRML áÅÐǧ¨ÃµÃǨËÒ NPML

3. ¨Ò¡¡ÒÃÍ͡Ẻ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì µÒÁẺ¨ÓÅͧã¹ÃÙ»·Õ 3.2 ÊÓËÃѺÃкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § (perpendicular recording) ·Õ ND = 2.5 áÅÐ SNR = 20 dB â´Â¡Ó˹´ ãËé ·ÒÃìà¡çµ ·Õ µéͧ¡Òà ¤×Í

H(D) = 1 + D

{1.56, 0.35, −0.66, −0.69, 0.81, 0.20} ·Ó¹ÒÂ

P (D),

·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å

3.1) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 1 á·ç» 3.2) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» 3.3) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 3 á·ç» 3.4) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 4 á·ç»

â´Â ·Õ

Heff (D),

¢éÍÁÙÅ ¢Í§Ç§¨ÃµÃǨËÒ NPML ·Õ ãªé

»ÃÒ¡®ÇèÒ ÅӴѺ ¢éͼԴ¾ÅÒ´

ak ∈ {−1, 1}

{wk }

·Õ ä´é ¤×Í

¨§ ¤Ó¹Ç³ ËÒ Ç§¨Ã ¡Ãͧ

áÅÐáÊ´§á¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ ãªé㹡ÒöʹÃËÑÊ


136

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

4. ·ÓàËÁ×͹㹢éÍ 3 áµè ¡Ó˹´ãËé ·ÒÃìà¡çµ ·Õ µéͧ¡Òà ¤×Í ¢éͼԴ¾ÅÒ´

H(D) = 1 + 2D + D2

áÅÐÅӴѺ

{wk } = {−0.56, −1.65, −0.21, 0.49, −0.98, −0.09}

5. Ẻ¨ÓÅͧ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ áÅÐÍÕ¤ÇÍäÅà«ÍÃì ã¹ÃÙ» ·Õ 3.2 ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ Ẻ á¹ÇµÑ § ·Õ ND = 2.5 áÅÐ SNR = 20 dB ÊÒÁÒö·Õ ¨ÐÅ´ÃÙ»ä´éà» ¹áºº¨ÓÅͧªèͧÊÑ­­Ò³ ẺÊÁÁÙÅ µÒÁÃÙ»·Õ 6.6 àÁ× Í ·ÒÃìà¡çµ·Õ µéͧ¡Òà ¤×Í ÍÔ¹¾Øµ

H(D) = 1+D ¶éÒ¡Ó˹´ãËé ÅӴѺ¢éÍÁÙÅ

{ak } = {1, −1, −1, 1} áÅÐÊÑ­­Ò³Ãº¡Ç¹ {wk } = {−0.41, −0.33, 0.41, −0.59,

−1.29}

¨§¶Í´ÃËÑÊ¢éÍÁÙÅ

yk

´éÇÂǧ¨ÃµÃǨËÒ NPML àÁ× Íãªé

5.1) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 1 á·ç» ¤×Í

P (D) = −0.2613D

5.2) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» ¤×Í

P (D) = −0.3929D − 0.5035D2

5.3) ǧ¨Ã¡Ãͧ·Ó¹ÒÂẺ 2 á·ç» ¤×Í

P (D) = −0.3443D − 0.4656D2 + 0.0964D3

6. ·ÓàËÁ×͹㹢éÍ 5 áµè¡Ó˹´ãËé·ÒÃìà¡çµ·Õ µéͧ¡Òà ¤×Í ÍÔ¹¾Øµ

{ak } = {−1, 1, 1, −1}

−1.12, −3.18}

áÅÐÅӴѺ¢éͼԴ¾ÅÒ´

H(D) = 1 + 2D + D2 ,

ÅӴѺ¢éÍÁÙÅ

{wk } = {0.47, 0.25, −0.38, −0.09,


º··Õ 7

ǧ¨ÃµÃǨËÒ PDNP

㹺·¹Õ ¨Ð¡ÅèÒǶ֧·Õ ÁÒ, ËÅÑ¡¡Ò÷ӧҹ, áÅлÃÐ⪹ì¢Í§ ǧ¨ÃµÃǨËÒ PDNP (pattern dependent noise predictive) [54] «Ö § à» ¹ ǧ¨ÃµÃǨËÒ·Õ ¶Ù¡ Í͡ẺÁÒà¾× Í ¨Ñ´¡ÒáѺ ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ (media jitter noise) «Ö § ¾ººèÍÂã¹Ãкº¡Òúѹ·Ö¡ áÁèàËÅç¡ ·Õ ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ÊÙ§æ ´Ñ§·Õ ¨Ð͸ԺÒµèÍä»ã¹º·¹Õ ¾ÃéÍÁ·Ñ § áÊ´§¼Å¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾ÃÐËÇèҧǧ¨ÃµÃǨËÒ PDNP áÅÐǧ¨ÃµÃǨËÒ PRML

7.1

º·¹Ó

¨Ò¡·Õ ä´é͸ԺÒÂä»ã¹º··Õ 4 à·¤¹Ô¤ PRML ¤×Í à·¤¹Ô¤¡ÒÃãªé§Ò¹ÃèÇÁ¡Ñ¹ÃÐËÇèÒ§ÍÕ¤ÇÍäÅà«ÍÃìẺ PR (partial response equalizer) áÅÐǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ (Viterbi detector) «Ö §à» ¹·Õ ¹ÔÂÁãªé§Ò¹ ¡Ñ¹ ÁÒ¡ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì [27] à·¤¹Ô¤ PRML ¨Ð·Ó§Ò¹à» ¹ 2 ¢Ñ ¹µÍ¹ ¤×Í ¢Ñ ¹µÍ¹áá¨Ð·Ó¡ÒûÃѺ ÃÙ»ÃèÒ§¢Í§ÊÑ­­Ò³·Õ ä´é ÃѺ ãËé à» ¹ 仵ÒÁÃÙ»ÃèÒ§¢Í§·ÒÃìà¡çµ (target) ·Õ µéͧ¡Òà áÅÐ¢Ñ ¹µÍ¹·Õ Êͧ¨Ð·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅâ´Âǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ·Õ ÊÃéÒ§¢Ö ¹¨Ò¡ ·ÒÃìà¡çµ·Õ ¡Ó˹´äÇé ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¶×Í ÇèÒ à» ¹ ǧ¨ÃµÃǨËÒÅӴѺ àËÁÒÐ·Õ ÊØ´ (optimal sequence detector) ¡çµèÍàÁ× Í Í§¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙèã¹ÊÑ­­Ò³·Õ ¨Ð·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅÁÕÅѡɳР137


138

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

à» ¹ ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ [15] ÍÂèÒ§äáçµÒÁã¹·Ò§»¯ÔºÑµÔ ¡ÒÃãªé §Ò¹ÍÕ¤ÇÍäÅà«ÍÃì Ẻ PR ¨ÐÊ觼ŷÓãËéͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔÁդس ÅѡɳÐà» ¹ ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ (colored noise) â´Â੾ÒÐÍÂèÒ§ÂÔ § àÁ× Í ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÊÙ§ (ND ÊÙ§) «Ö § ã¹¡Ã³Õ ¹Õ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨ÐäÁè ¶×Í ÇèÒ à» ¹ ǧ¨ÃµÃǨËÒÅӴѺ àËÁÒÐ ·Õ ÊØ´ÍÕ¡µèÍä» ´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒ NPML (noise predictive maximum likelihood) [51, 52] ¨Ö§ ä´é¶Ù¡¹ÓÁÒãªéà¾× Íà¾Ô Á»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ãкº â´Â·Õ ǧ¨ÃµÃǨËÒ NPML ¨ÐÁÕ¡Ãкǹ¡ÒÃ㹡ÒÃ·Ó ãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ ¡è͹·Õ ¨ÐÊ觼ÅÅѾ¸ì·Õ ä´éä»·Ó¡ÒöʹÃËÑÊ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ã¹ º··Õ 6 áÊ´§ãËéàËç¹ÇèÒ Ç§¨ÃµÃǨËÒ NPML ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRML â´Â੾ÒÐ ÍÂèÒ§ÂÔ §·Õ ¤ÇÒÁ¨Ø¢éÍÁÙŢͧÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÙ§ ¹Í¡¨Ò¡¹Õ ·Õ ND ÊÙ§æ ͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ Âѧ¨ÐÁÕÅÑ¡É³Ð¢Ö ¹ÍÂÙè¡Ñº Ẻ¢éÍÁÙÅ (data pattern) µÑÇÍÂèÒ§àªè¹ ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× Í ºÑ¹·Ö¡ ¶×Íä´éÇèÒà» ¹ ÊÑ­­Ò³Ãº¡Ç¹·Õ ¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ (pattern dependent noise) ¡ÅèÒǤ×Í ÃдѺ ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ ¨Ð¢Ö ¹ ÍÂÙè ¡Ñº Ẻ¢éÍÁÙÅ ·Õ à¢Õ¹ŧä»ã¹ Ê× ÍºÑ¹·Ö¡ à·¤¹Ô¤µèÒ§æ ä´é¶Ù¡¹ÓàʹÍà¾× Í·ÓãËéÊÑ­­Ò³Ãº¡Ç¹»ÃÐàÀ·¹Õ [54, 55] ÁÕÅѡɳСÅÒ ໠¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇ (white noise) ¡è͹·Õ ¨Ð·Õ ·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ã¹ÊèǹµèÍä»¹Õ ¨Ð͸ԺÒ¶֧ ËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ PDNP [54] áÅÐà·¤¹Ô¤ ¡ÒÃÅ´¤ÇÒÁ «Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ PDNP

7.2

¡ÒÃ¢Ö ¹ÍÂÙè¡Ñºà຺¢éÍÁÙŢͧÊÑ­­Ò³Ãº¡Ç¹

¾Ô¨ÒóÒẺ¨ÓÅͧªèͧÊÑ­­Ò³ã¹ÃÙ»·Õ 7.1 ¡Ó˹´ãËé ¢Í§ ªèͧ ÊÑ­­Ò³ ã¹ â´àÁ¹ µÍºÊ¹Í§ÃÇÁã¹â´àÁ¹

D

D, H(D)

C(D) = (1 − D)G(D)

¤×Í ¼Å µÍºÊ¹Í§ ·ÒÃìà¡çµã¹ â´àÁ¹

¤×Í ¼ÅµÍºÊ¹Í§

D, Q(D)

¤×Í ¼Å

¢Í§Ç§¨Ã¡Ãͧ¼èÒ¹µ Ó áÅÐÍÕ¤ÇÍäÅà«ÍÃì, áÅÐà¾× ÍãËé §èÒµèÍ ¡ÒÃ͸ԺÒÂ

ÍÕ¤ÇÍäÅà«ÍÃìẺ zero forcing [2, 16] ¨Ð¶Ù¡¹ÓÁÒãªéã¹Êèǹ¹Õ ¹Ñ ¹¤×Í

Q(D) = H(D)/C(D)

ã¹

·Ò§»¯ÔºÑµÔ ·ÒÃìà¡çµáÅÐÍÕ¤ÇÍäÅà«ÍÃì«Ö §¶Ù¡ÊÃéÒ§â´Âǧ¨Ã¡ÃͧàªÔ§àÊé¹áºº¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¨Ó¡Ñ´ (FIR: nite impulse response) ¨ÐÁÕ¿ §¡ìªÑ¹¶èÒÂâ͹·Õ ᵡµèҧ仨ҡ ã¹·Ò§·ÄÉ®Õ à¾ÃÒÐ©Ð¹Ñ ¹ ¶éÒ ¡Ó˹´ãËé

H 0 (D)

áÅÐ

H(D)

áÅÐ

Q0 (D) = H 0 (D)/C(D)

Q(D)

·Õ µéͧ¡ÒÃ

¤×Í ¿ §¡ìªÑ¹ ¶èÒÂ


7.2.

¡ÒÃ¢Ö ¹ÍÂÙè¡Ñºà຺¢éÍÁÙŢͧÊÑ­­Ò³Ãº¡Ç¹

139

n(t)

ak {0,1}

1–D

bk

g(t)

sk

s(t)

p(t)

∆t k

LPF

tk = kT Q(D)

C(D)

yk

equalizer

âk

detector

H(D) ÃÙ»·Õ 7.1: Ẻ¨ÓÅͧªèͧÊÑ­­Ò³

â͹¨ÃÔ§¢Í§·ÒÃìà¡çµáÅТͧÍÕ¤ÇÍäÅà«ÍÃì µÒÁÅӴѺ â´Â·Õ

H 0 (D) 6= H(D)

áÅÐ

Q0 (D) 6= Q(D)

´Ñ§¹Ñ ¹ ¢éÍÁÙÅàÍÒµì¾Øµ¢Í§ÍÕ¤ÇÍäÅà«ÍÃìÊÒÁÒöà¢Õ¹໠¹ÊÁ¡Ò÷ҧ¤³ÔµÈÒʵÃìä´é ¤×Í

Y (D) = A(D)C(D)Q0 (D) + N (D)Q0 (D) = A(D)H 0 (D) + N (D)Q0 (D) = A(D)H 0 (D) + N (D)Q0 (D) + [A(D)H(D) − A(D)H(D)] = A(D)H(D) + A(D)[H 0 (D) − H(D)] + N (D)Q0 (D) | {z }

(7.1)

W (D) àÁ× Í

N (D)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊì ÊÕ ¢ÒÇẺºÇ¡ (AWGN) áÅÐ

W (D)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹

W (D)

ÃÇÁ·Õ ä´é ³ ´éÒ¹¢Òà¢éҢͧǧ¨ÃµÃǨËÒÅӴѺ µÒÁÊÁ¡Òà (7.1) ¨ÐàËç¹ä´éÇèÒ ÊÑ­­Ò³Ãº¡Ç¹ ÁÕÅÑ¡É³Ð¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ (data dependent) à¹× ͧ¨Ò¡ ÁÕ¾¨¹ì

A(D)

ÍÂÙè

ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡ à» ¹¼ÅÁÒ¨Ò¡ ¡ÒÃàÅ× Í¹µÓá˹觢ͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð áººÊØèÁ (random transition shift) ÃÐËÇèÒ§¡Ãкǹ¡Ò÷ÓãËé Ê× Í ºÑ¹·Ö¡ ÁÕ ÊÀÒ¾¤ÇÒÁà» ¹ áÁèàËÅç¡ (magnetization) µÒÁ·Õ µéͧ¡Òà â´Â¨ÐÁÕ ¿ §¡ìªÑ¹ ¤ÇÒÁ˹Òá¹è¹ ¤ÇÒÁ¹èÒ¨Ðà» ¹ Ẻà¡ÒÊì à«Õ¹·Õ ÁÕ ¤èÒ à©ÅÕ Â à·èÒ ¡Ñº ¤èÒ ÈÙ¹Âì áÅФèÒ ¤ÇÒÁá»Ã»Ãǹà·èÒ ¡Ñº ¨Ó¡Ñ´ãËéÁÕ¤èÒäÁèà¡Ô¹ ¤èÒÊÑÁºÙóì¢Í§

bk

T /2)

â´Â·Õ

σj

|bk |σj2

(¹Ñ ¹¤×Í

∆tk ∼ N (0, |bk |σj2 ))

¨Ð¶Ù¡¡Ó˹´à» ¹¨Ó¹Ç¹à»ÍÃìà«ç¹µì¢Í§ºÔµà«ÅÅì

T

à¹× ͧ¨Ò¡ µÓá˹觡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¶Ù¡¡Ó˹´â´Â¢éÍÁÙźԵÍÔ¹¾Øµ

¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ ¨Ö§ ¢Ö ¹ ÍÂÙè ¡Ñº Ẻ¢éÍÁÙÅ ¢Í§

áÅÐ

áÅж١

|bk |

{ak }

{ak }

¤×Í

´Ñ§¹Ñ ¹

ÃÙ» ·Õ 7.2


140

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

5 4.5

Noise power (dB)

4 3.5 3 2.5 2 1.5 1 001

100

010

110

001

101

011

111

Data pattern

ÃÙ»·Õ 7.2: ¡ÓÅѧ ÊÑ­­Ò³Ãº¡Ç¹·Õ ¢Ö ¹ ¡Ñº Ẻ¢éÍÁÙÅ ³ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃì ·Õ ¶Ù¡ Í͡Ẻ ÊÓËÃѺ·ÒÃìà¡çµ EEPR2 [1 4 6 4 1] ¢Í§Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ·Õ ND = 2.5, SNR = 30 dB, áÅÐ

σj /T = 10%

áÊ´§¡ÓÅѧÊÑ­­Ò³Ãº¡Ç¹ (noise power) ·Õ ¢Ö ¹¡Ñºáºº¢éÍÁÙÅ ³ ´éÒ¹¢ÒÍÍ¡¢Í§ÍÕ¤ÇÍäÅà«ÍÃìẺ 21 á·ç» (tap) ·Õ ¶Ù¡Í͡ẺãËéÊÍ´¤Åéͧ¡Ñº·ÒÃìà¡çµ EEPR2,

H(D) = 1 + 4D + 6D2 + 4D3 +

D4 ,

¢Í§Ãкº¡Òúѹ·Ö¡ Ẻá¹ÇµÑ § (perpendicular recording) ·Õ ND = 2.5, SNR = 30 dB,

áÅÐ

σj /T = 10%

¨ÐàËç¹ä´éªÑ´à¨¹ÇèÒ ¡ÓÅѧÊÑ­­Ò³Ãº¡Ç¹¨ÐÁÕ¤èÒÊÙ§ àÁ× ÍÅӴѺ¢éÍÁÙÅÁÕ¡ÒÃà»ÅÕ Â¹

ʶҹÐËÅÒ¤ÃÑ § àªè¹ Ẻ¢éÍÁÙÅ 010 áÅÐ 101 à» ¹µé¹

7.3

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP

¾Ô¨ÒóÒǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ [15] â´Â·Õ àÁµÃÔ¡ÊÒ¢Ò (branch metric) ¤Ó¹Ç³ä´é¨Ò¡

λk (u, v) = |yk −

ν X i=0

|

hi ak−1 |2 {z

r̂k (u,v)

}

(7.2)


7.3.

àÁ× Í

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP

(u, v)

gram),

yk

141

á·¹¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡Ê¶Ò¹Ð

u

ä»ÂѧʶҹÐ

v

ã¹á¼¹ÀÒ¾à·ÃÅÅÔÊ (trellis dia

¤×Í ¢éÍÁÙÅ·Õ ¨Ð·Ó¡ÒöʹÃËÑÊ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ,

ÊÑ­­Ò³·Õ äÁè ÁÕ ÊÑ­­Ò³Ãº¡Ç¹·Õ ÊÍ´¤Åéͧ¡Ñº

(u, v)

r̂k (u, v)

¤×Í ¢éÍÁÙÅàÍÒµì¾Øµªèͧ

(¹Ñ ¹¤×Í ¤èÒ ·Õ áÊ´§ÍÂÙè ã¹áµèÅÐàÊé¹ ÊҢҢͧ

á¼¹ÀÒ¾à·ÃÅÅÔÊ àªè¹ µÒÁ·Õ áÊ´§ã¹ÃÙ»·Õ 4.6) «Ö §ËÒä´é¨Ò¡

rk = ak ∗ hk =

ν X

ak−i hi

(7.3)

i=0 àÁ× Í

¤×Í µÑÇ´Óà¹Ô¹¡Òä͹âÇÅ٪ѹ (convolution operator),

µéͧ¡ÒÃ,

hk

¤×Í ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô µÑÇ·Õ

k

¢Í§·ÒÃìà¡çµ, áÅÐ

ν

H(D)

=

k=0 hk D

k ¤×Í ·ÒÃìà¡çµ·Õ

¤×Í Ë¹èǤÇÒÁ¨Ó¢Í§·ÒÃìà¡çµ

à¹× ͧ¨Ò¡ ͧ¤ì»ÃСͺ¢Í§ÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙèã¹¢éÍÁÙÅ

yk

«Ö §ËÒä´é¨Ò¡ (´ÙẺ¨ÓÅͧ¡ÒÃ

Í͡Ẻ·ÒÃìà¡çµã¹ÃÙ»·Õ 3.2)

wk = yk − rk D

â´Â¼Å¡ÒÃá»Å§

¢éÍÁÙÅ ¶éÒ Êè§ ¢éÍÁÙÅ

¤×Í

yk

W (D)

(7.4)

µÒÁÊÁ¡Òà (7.1) ¨ÐÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹·Õ ¢Ö ¹ÍÂÙè¡Ñºáºº

à¢éÒ ä»·Ó¡ÒöʹÃËÑÊ ´éÇÂǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ »ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» ¢Í§ÍѵÃÒ

¢éͼԴ¾ÅÒ´ºÔµ (BER) ·Õ ä´é ÃѺ ¨ÐäÁè ´Õ à¾ÃÒÐ©Ð¹Ñ ¹ 㹡ÒÃ·Õ ¨Ðà¾Ô Á »ÃÐÊÔ·¸ÔÀÒ¾ÃÇÁ¢Í§Ãкº ¨Ð µéͧÁÕ ¡ÒùӡÃкǹ¡ÒÃ㹡Ò÷ÓãËé ÊÑ­­Ò³Ãº¡Ç¹à» ¹ ÊÕ ¢ÒÇ ·ÓãËé¢éÍÁÙÅ

wk

(noise whitening process) à¾× Í

ÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇ ¡è͹·Õ ¨ÐÊ觼ÅÅѾ¸ì·Õ ä´éä»·Ó¡ÒöʹÃËÑÊ´éÇÂ

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ´Ñ§¹Ñ ¹ 㹡ÒÃãªé à·¤¹Ô¤ ¡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹ (noise prediction) ÃèÇÁ ¡ÑºÇ§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ð·ÓãËéàÁµÃÔ¡ÊÒ¢Òã¹ÊÁ¡Òà (7.2) µéͧ¶Ù¡´Ñ´á»Å§à» ¹

λk (u, v) = |yk − r̂k (u, v) − ŵk |2 àÁ× Í

ŵk

(7.5)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹·Õ ¶Ù¡·Ó¹Ò «Ö §ËÒä´é¨Ò¡

ŵk =

L X

pi wk−i

(7.6)

i=1 â´Â·Õ á·ç»,

P (D) = pi

PL

i=1 pi D

i ¤×Í Ç§¨Ã¡Ãͧ·Ó¹Ò (prediction lter) ÊÑ­­Ò³Ãº¡Ç¹áºº

¤×Í ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô µÑÇ·Õ

i

L

¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹, áÅТéͼԴ¾ÅÒ´·Õ à¡Ô´¨Ò¡


142

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡Ò÷ӹÒÂ

ek

¤×Í

ek = wk − ŵk ËÃ×Í

wk =

L X

(7.7)

pi wk−i + ek

(7.8)

i=1 â´Âǧ¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹

P (D)

·Õ ´Õ¨Ðµéͧ·ÓãËé¢éͼԴ¾ÅÒ´·Õ à¡Ô´¨Ò¡¡Ò÷ӹÒÂ

ek

ÁÕ

ÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇãËéÁÒ¡·Õ ÊØ´ àÁ× ÍÃкº·Ó§Ò¹·Õ ¤ÇÒÁ¨Ø¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ìÊÙ§æ ¨Ð·ÓãËéÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹ ¨Ð¢Ö ¹ ÍÂÙè ¡Ñº Ẻ¢éÍÁÙÅ ´Ñ§¹Ñ ¹ ǧ¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹·Õ àËÁÒÐÊØ´ ÍÂÙè ¡Ñº Ẻ¢éÍÁÙÅ ´éÇÂàªè¹¡Ñ¹ à¾ÃÒÐ©Ð¹Ñ ¹ 㹡ÒÃËÒ¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§

P (D)

P (D)

¡ç ¤ÇÃ·Õ ¨Ð¢Ö ¹

¨Ðµéͧ¨Ñ´ ÃÙ» ÊÁ¡ÒÃ

(7.8) ãËÁè ´Ñ§¹Õ

wk (a) =

L X

pi (a)wk−i (a) + ek (a)

(7.9)

i=1 àÁ× Í

a

ãªé᷹Ẻ¢éÍÁÙŵèÒ§æ ·Õ à» ¹ä»ä´é ÊÁ¡Òà (7.9) ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§àÁ·ÃÔ¡«ìä´é ¤×Í

wk (a) = p(a)T w(a) + ek (a) â´Â·Õ

p(a)

·Ó¹ÒÂ, áÅÐ

=

[p1 (a), p2 (a), . . . , pL (a)]T

w(a)

=

¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§ àÇ¡àµÍÃì

w(a)T

(7.10)

¤×Í àÇ¡àµÍÃì á¹ÇµÑ § ¢Í§¤èÒ ÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ

[wk−1 (a), wk−2 (a), . . . , wk−L (a)]T P (D)

ÊÒÁÒö¤Ó¹Ç³ËÒä´é â´Â¡Òäٳ ·Ñ § Êͧ¢éÒ§¢Í§ÊÁ¡Òà (7.10) ´éÇÂ

áÅéÇãÊèµÑÇ´Óà¹Ô¹¡ÒäèÒ¤Ò´ËÁÒ (expectation operator) «Ö §¨Ðä´é¼ÅÅѾ¸ìà» ¹

£ ¤ £ ¤ E wk (a)w(a)T = p(a)T E w(a)w(a)T = p(a)T R(a) àÁ× Í

£ ¤ E w(a)w(a)T ¤×Í àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì (auto correlation matrix) ¢Í§ÊÑ­­Ò³ £ ¤ wk (a), áÅÐ E ek (a)w(a)T = 0 µÒÁËÅÑ¡¡ÒÃàªÔ§µÑ §©Ò¡ (orthogonality principle)

R(a)

ú¡Ç¹

(7.11)

=

[25] á¡éÊÁ¡Òà (7.11) ¨Ðä´é¼ÅÅѾ¸ìà» ¹

p(a)T = E[wk (a)w(a)T ] R−1 (a)

(7.12)


7.3.

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP

143

áÅФèÒ¤ÇÒÁá»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ (predictor error variance) ¤×Í [54]

£ ¤ £ ¤ £ ¤T σp2 (a) = E wk (a)2 − E wk (a)w(a)T R−1 (a)E wk (a)w(a)T

(7.13)

ÊÁ¡Òà (7.12) áÅÐ (7.13) áÊ´§ãËéàËç¹ÇèÒ ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂáÅФèÒ¤ÇÒÁá»Ã»Ãǹ ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ¨Ð¢Ö ¹ÍÂÙè¡ÑºªèǧàÇÅÒ

a

k

·Õ ¾Ô¨ÒÃ³Ò (³ ªèǧàÇÅÒ

k

Ë¹Ö §æ ¡çÍÒ¨¨ÐÁÕẺ¢éÍÁÙÅ

k

·Õ µèÒ§¡Ñ¹ ä´é) ÍÂèÒ§äáçµÒÁ ¤èÒ ·Ñ § 2 ¤èÒ ¹Õ ÊÒÁÒö·Õ ¨Ð·ÓãËé äÁè ¢Ö ¹ ¡Ñº ªèǧàÇÅÒ

¡çä´é ¶éÒ ÊÁÁØµÔ ãËé

Ãкºà» ¹áººÊ൪ѹà¹ÃÕ (stationary) [10, 26] ËÃ×ͶéÒà» ¹Ãкº·Õ äÁèà» ¹áººÊ൪ѹà¹ÃÕ ÍÑÅ¡ÍÃÔ·ÖÁ Ẻ»ÃѺµÑÇ (adaptive algorithm) [4, 10, 16] ÊÒÁÒö·Õ ¨Ð¶Ù¡¹ÓÁÒãªé㹡ÒûÃѺ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§ ǧ¨Ã¡Ãͧ·Ó¹Ò áÅФèÒ¤ÇÒÁá»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂä´é à¹× ͧ¨Ò¡ ¤èÒ ¤ÇÒÁá»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ («Ö § à» ¹ ¼ÅÁÒ¨Ò¡¡Ãкǹ¡ÒÃ㹡Ò÷ÓãËé ÊÑ­­Ò³Ãº¡Ç¹à» ¹ ÊÕ ¢ÒÇ) ¢Ö ¹ ÍÂÙè ¡Ñº Ẻ¢éÍÁÙÅ µÒÁ·Õ áÊ´§ã¹ÊÁ¡Òà (7.13) ¤èÒ àÁµÃÔ¡ ÊҢҢͧ ǧ¨ÃµÃǨËÒ PDNP ¨Ðµéͧ¤Ó¹Ö§¶Ö§¡ÒÃ¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ´éÇ à¾ÃÒÐ©Ð¹Ñ ¹ àÁµÃÔ¡ÊҢҢͧǧ¨Ã µÃǨËÒ PDNP ÊÒÁÒöà¢Õ¹ä´éà» ¹

λk (u, v) = log (σp (u, v)) + àÁ× Í

σp2 (u, v)

à¡Õ ÂÇ¢éͧ¡Ñº

|yk − r̂k (u, v) − ŵk (u, v)|2 2σp2 (u, v)

¤×Í ¤èÒ¤ÇÒÁá»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ·Õ ÊÍ´¤Åéͧ¡Ñº

(u, v),

¢éÍÁÙÅ·Õ à¡Õ ÂÇ¢éͧ¡Ñº

áÅÐ

ŵk (u, v)

(u, v)

(7.14)

(u, v)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹·Õ ¶Ù¡·Ó¹ÒÂ·Õ ÊÍ´¤Åéͧ¡Ñº

áÅÐẺ¢éÍÁÙÅ·Õ

(u, v)

áÅÐẺ

«Ö §ËÒä´é¨Ò¡

ŵk (u, v) =

L X

pi (u, v){yk−i − r̂k−i (u, v)}

(7.15)

i=1 ¨Ò¡ÊÁ¡Òà (7.12) áÅÐ (7.13) ¨Ð¾ºÇèÒ Ç§¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³ á»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ

P (D)

áÅÐ

σp2

σp2

P (D) =

PL i

pi D i

áÅФèÒ¤ÇÒÁ

¨Ð¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅã¹áµèÅÐàÊé¹ÊÒ¢Ò ´Ñ§¹Ñ ¹ ¤èÒ¾ÒÃÒÁÔàµÍÃì

·Õ ãªé㹡ÒäӹdzàÁµÃÔ¡ÊÒ¢Òã¹áµèÅÐàÊé¹ÊҢҢͧἹÀÒ¾à·ÃÅÅÔʨÐÁÕ¤èÒµèÒ§¡Ñ¹

µÒÁẺ¢éÍÁÙÅ·Õ ÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

(u, v)

¹Ñ ¹æ

ã¹·Ò§»¯ÔºÑµÔáÅéÇ ¤èÒÊÑÁ»ÃÐÊÔ·¸Ô ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹

p(a)

¨Ð¢Ö ¹ÍÂÙè¡Ñº¢éÍÁÙÅ

ºÔµÀÒÂã¹Ë¹éÒµèÒ§àÅ× Í¹áºº¨Ó¡Ñ´ ( nite sliding window) [56] «Ö §à¢Õ¹᷹´éÇÂÊÑ­Åѡɳì

aW k−M


144

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊÓËÃѺ¤èҨӹǹàµçÁºÇ¡ ʶҹÐà·èҡѺ

M

áÅÐ

2max(ν+L, M )+W

W

´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒ PDNP ¨Ðãªéá¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ Áըӹǹ

ʶҹРÍÂèÒ§äáçµÒÁ à¾× ÍãËé§èÒµèÍ¡ÒÃ͸ԺÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP ã¹Êèǹ¹Õ ¨Ð¾Ô¨ÒóÒ੾ÒСóշÕ

W =0

Ê¶Ò¹Ð·Ñ §ËÁ´·Õ ãªéã¹á¼¹ÀÒ¾à·ÃÅÅÔÊÁըӹǹà·èҡѺ

7.4

áÅÐ

M < ν +L

à¾ÃÒÐ©Ð¹Ñ ¹ ¨Ó¹Ç¹

2ν+L

ÍÑÅ¡ÍÃÔ·ÖÁ PS PDNP

ÍÑÅ¡ÍÃÔ·ÖÁ PDNP ·Õ ͸ԺÒÂä»ã¹ËÑÇ¢éÍ ·Õ 7.3 µéͧ¡ÒÃá¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ÁÕ ¨Ó¹Ç¹Ê¶Ò¹Ð·Ñ §ËÁ´

2ν+L

ÍÂèÒ§äáçµÒÁ ¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ PDNP ÊÒÁÒö·ÓãËéŴŧä´éâ´Âãªéá¹Ç¤Ô´ ¡ÒÃ

» ͹¡ÅѺ¤èҵѴÊԹ㨠(decision feedback) [53] áÅÐà¾× ÍãËéÃкºÁÕ»ÃÐÊÔ·¸ÔÀÒ¾·Õ ÂÍÁÃѺä´é á¹Ç¤Ô´ ¹Õ µéͧ¡ÒäÇÒÁÅÖ¡ (depth) ¡Òû ͹¡ÅѺ ¤èÒ µÑ´ÊÔ¹ã¨·Õ ¤è͹¢éÒ§¹éÍ àÁ× Í à·Õº¡Ñº ¤ÇÒÁÂÒÇ (ËÃ×Í ¨Ó¹Ç¹á·ç») ¢Í§Ç§¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹ «Ö §ËÁÒ¤ÇÒÁÇèÒ á¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ ãªéÂѧ¨Óà» ¹ µéͧ¶Ù¡ ¢ÂÒÂãËé ãË­è ¢Ö ¹ (trellis expansion) ¹Ñ ¹¤×Í ÁÕըӹǹʶҹÐÁÒ¡¢Ö ¹ ¡ÇèÒ Ãкº·Õ äÁè ãªé ǧ¨Ã ¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹ ã¹Êèǹ¹Õ ¨Ð͸ԺÒ ÇÔ¸Õ¡ÒÃ·Õ àʹÍã¹ [53]

«Ö §¤ÅéÒ¡Ѻá¹Ç¤Ô´

¡ÒûÃÐÁÇżÅẺà¾Íà«ÍÃì

äÇàÇÍÃì (PSP: per survivor processing) [57] ÊÓËÃѺŴ¤ÇÒÁ«Ñº«é͹¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ PDNP â´Â ÍÑÅ¡ÍÃÔ·ÖÁãËÁè·Õ ä´é ã¹Ë¹Ñ§Ê×Í¹Õ ¨ÐàÃÕ¡ÇèÒ ÍÑÅ¡ÍÃÔ·ÖÁ PS PDNP (per survivor PDNP algorith m) ǧ¨ÃµÃǨËÒ PS PDNP ¨Ð·Ó§Ò¹µÒÁÍÑÅ¡ÍÃÔ·ÖÁ PDNP º¹¾× ¹°Ò¹¢Í§á¹Ç¤Ô´ PSP «Ö § ¨Ð ·ÓãËé á¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ãªé ã¹Ç§¨ÃµÃǨËÒ PS PDNP ÁÕ ¨Ó¹Ç¹Ê¶Ò¹Ðà·èÒ à´ÔÁ (ËÃ×Í à·èÒ ¡Ñº Ãкº·Õ äÁèãªèǧ¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹) ¡ÅèÒǤ×Í á·¹·Õ ¨Ð·Ó¡ÒâÂÒÂá¼¹ÀÒ¾à·ÃÅÅÔÊãËéãË­è¢Ö ¹ ǧ¨ÃµÃǨËÒ PS PDNP ¨Ð·Ó¡ÒÃÁͧÂé͹¡ÅѺ 仵ÒÁàÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè µÒÁá¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ÁÒ¶Ö§ ³ ¨Ø´µèÍ (node) ·Õ ¡ÓÅѧ ¾Ô¨ÒÃ³Ò áÅéÇ ãªé ¢éÍÁÙÅ µèÒ§æ ·Õ ÊÍ´¤Åéͧ¡Ñº àÊé¹·Ò§·Õ Âѧ ÁÕ ªÕÇÔµ ÍÂÙè ¹Ñ ¹ 㹡ÒäӹdzËÒ¤èÒÊÑ­­Ò³Ãº¡Ç¹·Õ ¶Ù¡·Ó¹Ò ËÃ×ÍÍÒ¨¨Ð¡ÅèÒÇä´éÇèÒ Ç§¨ÃµÃǨËÒ PS PDNP Âѧ¤§ãªéÊÁ¡Òà (7.14) 㹡ÒäӹdzËÒ¤èÒàÁµÃÔ¡ÊÒ¢Ò Â¡àÇé¹áµè ÊÑ­­Ò³Ãº¡Ç¹·Õ ¶Ù¡·Ó¹Ò¨ж١ ¤Ó¹Ç³¨Ò¡

ŵk (u, v) =

L X i=1

pi (u, v)ẑk−i (u, v)

(7.16)


7.4.

ÍÑÅ¡ÍÃÔ·ÖÁ PS PDNP

time k (0) -1 -1

0

145

zˆk −2 ( A )

k+1

zˆk −1 ( A ) time k

zˆk ( 0,1)

2

k+1

A (1) 1 -1

(1)

0

zˆk −2 ( B )

2 -2

(2) -1 1

0 -2

ak = 1

0

ak = -1

(3) 1 1

B

zˆk ( 2,1) zˆk −1 ( B )

yk

yk −1 = 1

yk − 2 = 0.5

(a) Trellis diagram: PR4

yk = 0.15

(b) Decoding procedure

ÃÙ»·Õ 7.3: (a) á¼¹ÀÒ¾à·ÃÅÅÔÊÊÓËÃѺ·ÒÃìà¡çµáºº PR4 áÅÐ (b) ¢Ñ ¹µÍ¹¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇ ἹÀÒ¾à·ÃÅÅÔÊ

àÁ× Í

ẑk−i (u, v)

¤×Í ÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙè ã¹¢éÍÁÙÅ

yk

·Õ à» ¹ ¼Å·ÓãËé à¡Ô´ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

ÊÍ´¤Åéͧ¡ÑºàÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè·Õ à» ¹¼Å·ÓãËéà¡Ô´¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

(u, v)

«Ö §¹ÔÂÒÁâ´Â

ẑk (u, v) = yk (u, v) − r̂k (u, v)

(7.17)

´Ñ§¹Ñ ¹¨ÐàËç¹ä´éÇèÒ Ç§¨ÃµÃǨËÒ PS PDNP ¨Ðãªéá¼¹ÀÒ¾à·ÃÅÅÔÊ·Õ ÁըӹǹʶҹÐà·èҡѺ áµè¨ÐµéͧÁÕ¢Ñ ¹µÍ¹à¾Ô ÁàµÔÁ㹡ÒÃà¡çº¤èҢͧ

{ẑk−1 , ẑk−2 , . . . , ẑk−L }

ʶҹÐ

ÊÓËÃѺ·Ø¡àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµ

ÍÂÙè

µÑÇÍÂèÒ§·Õ 7.1

ÃÙ»·Õ 7.3 áÊ´§µÑÇÍÂèÒ§¡ÒÃËÒ¤èÒÊÑ­­Ò³Ãº¡Ç¹·Õ ὧÍÂÙèã¹¢éÍÁÙÅ

ãËé à¡Ô´ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐÊÍ´¤Åéͧ¡Ñº

H(D) = 1 − D2 , Áըӹǹ

L=2

(u, v)

yk

·Õ à» ¹¼Å·Ó

µÒÁÊÁ¡Òà (7.17) ¢Í§Ãкº·Õ ãªé ·ÒÃìà¡çµ Ẻ PR4,

«Ö §ÁÕá¼¹ÀÒ¾à·ÃÅÅÔʵÒÁÃÙ»·Õ 7.3(a) àÁ× Íǧ¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹·Õ ãªé

á·ç»


146

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÇÔ¸Õ·Ó

ãËé¾Ô¨ÒóÒʶҹР(1) ³ àÇÅÒ

k+1

µÒÁ·Õ áÊ´§ã¹ÃÙ»·Õ 7.3(b) ¨ÐàËç¹ä´éÇèÒÁÕàÊé¹·Ò§·Õ ÇÔ §

à¢éÒÁÒËÒʶҹР(1) à» ¹¨Ó¹Ç¹ 2 àÊé¹·Ò§ ¤×Í àÊé¹·Ò§ A áÅÐ B à¾ÃÒÐ©Ð¹Ñ ¹ ¤èÒ ÊÁ¡Òà (7.16) ÊÓËÃѺ

i

ẑk−i (u, v)

ã¹

= 1 áÅÐ 2 ËÒä´é´Ñ§µèÍä»¹Õ ÊÓËÃѺàÊé¹·Ò§ A ¨Ðä´éÇèÒ

ẑk (0, 1) = yk − r̂k (0, 1) = 0.15 − 2 = −1.85 ẑk−1 (A) = ẑk−1 (0, 1) = yk−1 − r̂k−1 (0, 0) = 1 − 0 = 1 ẑk−2 (A) = ẑk−2 (0, 1) = yk−2 − r̂k−2 (0, 0) = 0.5 − 0 = 0.5 áÅÐÊÓËÃѺàÊé¹·Ò§ B ¨Ðä´éÇèÒ

ẑk (2, 1) = yk − r̂k (2, 1) = 0.15 − 0 = 0.15 ẑk−1 (B) = ẑk−1 (2, 1) = yk−1 − r̂k−1 (3, 2) = 1 − (−2) = 3 ẑk−2 (B) = ẑk−2 (2, 1) = yk−2 − r̂k−2 (1, 3) = 0.5 − 2 = −1.5

7.5

¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ PDNP

㹡ÒÃà»ÃÕºà·Õº¤ÇÒÁ«Ñº«é͹ (complexity) ¢Í§Ç§¨ÃµÃǨËÒ ¨Ð¾Ô¨ÒóҨҡ¨Ó¹Ç¹µÑÇ´Óà¹Ô¹¡Òà (operator) ÊÓËÃѺ¡Òúǡ (addition) áÅСÒäٳ (multiplication) ·Õ µéͧãªé㹡Ò÷ӧҹ¢Í§áµèÅРǧ¨ÃµÃǨËÒ µÒÃÒ§·Õ 7.1 à»ÃÕºà·Õº¤ÇÒÁ«Ñº«é͹¢Í§Ç§¨ÃµÃǨËÒ PDNP áÅÐ PS PDNP àÁ× Í

Np

¤×Í ¨Ó¹Ç¹áºº¢éÍÁÙÅ·Õ ãªéã¹Ç§¨ÃµÃǨËÒ PDNP áÅÐ PS PDNP ¨Ò¡µÒÃÒ§¨ÐàËç¹ä´éÇèÒ Ç§¨Ã

µÃǨËÒ PS PDNP ÁÕ¤ÇÒÁ«Ñº«é͹áÅеéͧ¡ÒÃãªé˹èǤÇÒÁ¨Ó (memory requirement) ¹éÍ¡ÇèÒ Ç§¨ÃµÃǨËÒ PDNP ÁÒ¡

7.6

¼Å¡Ò÷´Åͧ

ã¹Êèǹ¹Õ ¨Ð·Ó¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§Ç§¨ÃµÃǨËÒ PRML áÅÐ PDNP â´ÂãªéẺ¨ÓÅͧ ªèͧÊÑ­­Ò³ µÒÁÃÙ»·Õ 7.1 àÁ× ÍÊÑ­­Ò³ read back ÊÒÁÒöà¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§ÊÁ¡Òä³ÔµÈÒʵÃì


7.6.

¼Å¡Ò÷´Åͧ

147

µÒÃÒ§·Õ 7.1: ¨Ó¹Ç¹¢Í§µÑÇ´Óà¹Ô¹¡ÒÃ·Õ ãªéµèÍ¢éÍÁÙÅ 1 ºÔµ ¢Í§Ç§¨ÃµÃǨËÒ PDNP áÅÐ PS PDNP

ǧ¨ÃµÃǨËÒ

¨Ó¹Ç¹¢Í§µÑÇ´Óà¹Ô¹¡ÒÃ·Õ ãªéµèÍ¢éÍÁÙÅ 1 ºÔµ

(detector) ǧ¨ÃµÃǨËÒ PDNP

˹èǤÇÒÁ¨Ó·Õ µéͧ¡ÒÃ

¡Òúǡ

¡Òäٳ

(memory requirement)

(4L + 7)2ν+L

(2L + 8)2ν+L

(2L + 4)2ν+L + Np L + 2

(2L + 8)2ν

(2L + 8)2ν

(2L + 8)2ν + Np L

ǧ¨ÃµÃǨËÒ PS PDNP

ä´é¤×Í

p(t) =

S−1 X

bk g(t − kT + ∆tk ) + n(t)

(7.18)

k=0

ak ∈ {0, 1}

àÁ× Í

¤×Í ¢éÍÁÙÅ ºÔµ ÍÔ¹¾Øµ ·Õ ÁÕ ¨Ó¹Ç¹·Ñ §ËÁ´

ºÔµà»ÅÕ Â¹Ê¶Ò¹Ð (bk

= ±1

ÁÕ¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹Ð),

S = 4096

ºÔµ,

bk = (ak − ak−1 )

ÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹á»Å§Ê¶Ò¹ÐºÇ¡ËÃ×Íź áÅÐ

g(t)

bk = 0

n(t)

¤×Í ÊÑ­­Ò³

ú¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺºÇ¡ (AWGN) ·Õ ÁÕ¤ÇÒÁ˹Òá¹è¹Ê໡µÃÑÁ¡ÓÅѧẺÊͧ´éÒ¹à·èҡѺ

∆tk

¤×ÍäÁè

¤×Í ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹ÐµÒÁÊÁ¡Òà (1.1) ÊÓËÃѺÃкº¡ÒÃ

ºÑ¹·Ö¡áººá¹Ç¹Í¹ áÅеÒÁÊÁ¡Òà (1.2) ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §,

áÅÐ

¤×Í

N0 /2,

¤×Í ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡ (media jitter noise) ·Õ ÁÕ¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹

¤ÇÒÁ¹èÒ¨Ðà» ¹ Ẻà¡ÒÊì à«Õ¹ (Gaussian probability density function) â´ÂÁÕ ¤èÒà©ÅÕ Â à·èÒ ¡Ñº ¤èÒ ÈÙ¹Âì áÅФèÒ ¤ÇÒÁá»Ã»Ãǹà·èÒ ¡Ñº à¡Ô¹

T /2)

àÁ× Í

σj

|bk |σj2

(¹Ñ ¹¤×Í

∆tk ∼ N (0, |bk |σj2 ))

¨Ð¶Ù¡ ¡Ó˹´à» ¹ ¨Ó¹Ç¹à»ÍÃìà«ç¹µì ¢Í§ºÔµ à«ÅÅì

(absolute value) ¢Í§

T

áÅж١ ¨Ó¡Ñ´ ãËé ÁÕ ¤èÒ äÁè

áÅÐ

|bk |

¤×Í ¤èÒÊÑÁºÙóì

bk

ÊÑ­­Ò³ read back

p(t)

¨Ð¶Ù¡Ê觼èÒ¹ä»Âѧǧ¨Ã¡Ãͧ¼èÒ¹µ Ó (LPF: low pass lter) ºÑµà·ÍÃì

àÇÔÃìµÍѹ´Ñº·Õ 7 áÅж١·Ó¡ÒêѡµÑÇÍÂèÒ§´éǤÇÒÁ¶Õ ¡ÒêѡµÑÇÍÂèÒ§à·èҡѺ

1/T

â´ÂÊÁÁصÔÇèÒ¡Ãкǹ

¡ÒÃ㹡Òêѡ µÑÇÍÂèÒ§ÁÕ ¡ÒÃà¢éÒ ¨Ñ§ËÇÐÃÐËÇèÒ§ÊÑ­­Ò³ read back áÅÐǧ¨ÃªÑ¡ µÑÇÍÂèҧẺÊÁºÙóì (perfect synchronization) ¨Ò¡¹Ñ ¹ ÅӴѺ¢éÍÁÙÅàÍÒµì¾Øµ

{sk } ¨Ð¶Ù¡» ͹ä»ÂѧÍÕ¤ÇÍäÅà«ÍÃì à¾× Í»ÃѺ

¤Ø³ÅѡɳТͧÊÑ­­Ò³ãËéà» ¹ä»µÒÁ·ÒÃìà¡çµ·Õ µéͧ¡Òà áÅéÇ¡çÊè§ÅӴѺ¢éÍÁÙÅàÍÒµì¾Øµ

{yk }

·Ó¡ÒöʹÃËÑÊ¢éÍÁÙÅ´éÇÂǧ¨ÃµÃǨËÒ (detector) à¾× ÍËÒ¤èÒ»ÃÐÁÒ³¢Í§ÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ

·Õ ä´éä»

{ak }

·Õ


148

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

à» ¹ä»ä´éÁÒ¡·Õ ÊØ´ 㹺·¹Õ ¤èÒ SNR ¨Ð¹ÔÂÒÁâ´Â

µ SNR = 10 log10 àÁ× Í

Ei

Ei N0

¶ (7.19)

1

¤×Í ¾Åѧ§Ò¹¢Í§¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¢Í§ªèͧÊÑ­­Ò³

¹Í¡¨Ò¡¹Õ áµèÅШش ¢Í§ BER

¨Ð¶Ù¡ ¤Ó¹Ç³â´Âãªé ¢éÍÁÙÅ ËÅÒÂæ à«¡àµÍÃì (sector) ¨¹¡ÇèÒ ¨Ðä´é ¢éͼԴ¾ÅÒ´ºÔµ ÁÒ¡¡ÇèÒ ËÃ×Í à·èÒ ¡Ñº 1000 ºÔµ ǧ¨ÃµÃǨËÒ 3 Ẻ ¤×Í Ç§¨ÃµÃǨËÒ PRML, ǧ¨ÃµÃǨËÒ PDNP, áÅÐǧ¨ÃµÃǨËÒ PS PDNP ¨Ð¶Ù¡·Ó¡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾ ÊÓËÃѺÃкº·Õ ãªé·ÒÃìà¡çµáºº GPR3 (·ÒÃìà¡çµáºº 3 á·ç» ·Õ ¶Ù¡Í͡ẺµÒÁà§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡ [19]) áÅÐÍÕ¤ÇÍäÅà«ÍÃìẺ 21 á·ç» â´Â·Õ ·ÒÃìà¡çµ Ẻ GPR3 ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ ¤×Í Ãкº¡Òúѹ·Ö¡áººá¹ÇµÑ § ¤×Í

H(D) = 1 + 0.05D − 0.65D2

H(D) = 1+1.25D +0.62D2

áÅÐÊÓËÃѺ

ÃÙ»·Õ 7.4 à»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾

ã¹ÃÙ»¢Í§ BER ¢Í§Ç§¨ÃµÃǨËÒ·Ñ § 3 Ẻ ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹áÅÐẺá¹ÇµÑ § ·Õ

N D = 2.5

áÅÐÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡

σj /T = 10%

¨Ò¡ÃÙ»·Õ 7.4(a) ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹Ç¹Í¹ ǧ¨ÃµÃǨËÒ PS PDNP ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ ã¡Åéà¤Õ§¡ÑºÇ§¨ÃµÃǨËÒ PDNP áµèǧ¨ÃµÃǨËÒ·Ñ §Êͧ¹Õ ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PRM L â´Â੾ÒÐÍÂèÒ§ÂÔ § ·Õ SNR ÊÙ§ (¹Ñ ¹¤×Í àÁ× Í ÊÑ­­Ò³Ãº¡Ç¹ËÅÑ¡ ã¹Ãкº ¤×Í ÊÑ­­Ò³Ãº¡Ç¹ ¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡) ã¹¢³Ð·Õ ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ § (´ÙÃÙ»·Õ 7.4(b)) ǧ¨ÃµÃǨËÒ PRML ´ÙàËÁ×͹¨ÐÁÕ»ÃÐÊÔ·¸ÔÀÒ¾´Õ¡ÇèÒǧ¨ÃµÃǨËÒ PS PDNP áÅÐǧ¨ÃµÃǨËÒ PDNP àÅ硹éÍ ·Õ SNR µ Ó ·Ñ §¹Õ ÍÒ¨¨Ðà» ¹à¾ÃÒÐÇèÒ ·Õ SNR µ Ó ÊÑ­­Ò³Ãº¡Ç¹ËÅÑ¡ã¹ÃкºäÁèãªèÊÑ­­Ò³Ãº¡Ç¹ ¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡ ´Ñ§¹Ñ ¹ ǧ¨ÃµÃǨËÒ PS PDNP áÅÐǧ¨ÃµÃǨËÒ PDNP «Ö §¶Ù¡Í͡ẺÁÒãËé ¨Ñ´¡ÒáѺ ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ ¨Ö§ äÁè ÊÒÁÒö·Ó§Ò¹ä´é ÍÂèÒ§ÁÕ »ÃÐÊÔ·¸ÔÀÒ¾ ÍÂèÒ§äà ¡çµÒÁ ·Õ SNR ÊÙ§ ǧ¨ÃµÃǨËÒ PS PDNP áÅÐǧ¨ÃµÃǨËÒ PDNP ¨ÐÁÕ »ÃÐÊÔ·¸ÔÀÒ¾´Õ ¡ÇèÒ Ç§¨Ã µÃǨËÒ PRML ÁÒ¡ ¨Ò¡¡Òüŷ´ÅͧáÊ´§ãËéàËç¹ä´éÇèÒ Ç§¨ÃµÃǨËÒ PS PDNP ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ ã¡Åéà¤Õ§¡Ñº ǧ¨ÃµÃǨËÒ PDNP à¾ÃÒÐ©Ð¹Ñ ¹ ã¹·Ò§»¯ÔºÑµÔ ǧ¨ÃµÃǨËÒ PS PDNP ÍÒ¨¨Ð¶Ù¡ ¹ÓÁÒãªé᷹ǧ¨ÃµÃǨËÒ PDNP à¹× ͧ¨Ò¡ ÁÕ¤ÇÒÁ«Ñº«é͹¹éÍ¡ÇèÒÁÒ¡ 1

¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¢Í§ªèͧÊÑ­­Ò³ ÁÕ¤èÒà·èҡѺ ͹ؾѹ¸ì¢Í§¼ÅµÍºÊ¹Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

¡Òúѹ·Ö¡áººá¹Ç¹Í¹ áÅÐÁÕ¤èÒà·èҡѺ

0

g (t)/2

ÊÓËÃѺÃкº¡Òúѹ·Ö¡áººá¹ÇµÑ §

g 0 (t)

ÊÓËÃѺÃкº


7.7.

ÊÃØ»·éÒº·

149

−1

10

PRML (4 states) PS − PDNP (4 states) PDNP (32 states) −2

BER

10

−3

10

−4

10

−5

10

18

20

22

24

26

28

30

32

(a) SNR (dB) 0

10

PRML (4 states) PS − PDNP (4 states) PDNP (32 states) −1

BER

10

−2

10

−3

10

−4

10

18

20

22

24

26

28

30

32

34

36

(b) SNR (dB)

ÃÙ»·Õ 7.4:

»ÃÐÊÔ·¸ÔÀÒ¾ã¹ÃÙ» ¢Í§ BER ¢Í§Ç§¨ÃµÃǨËÒµèÒ§æ ÊÓËÃѺ Ãкº¡Òúѹ·Ö¡ (a) Ẻ

á¹Ç¹Í¹ áÅÐ (b) Ẻá¹ÇµÑ § ·Õ ND = 2.5 áÅÐ

7.7

σj /T = 10%

ÊÃØ»·éÒº·

àÁ× Í ¤ÇÒÁ¨Ø ¢éÍÁÙÅ ¢Í§ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì ÊÙ§ ¢Ö ¹ ÊÑ­­Ò³Ãº¡Ç¹·Õ ´éÒ¹¢Òà¢éÒ ¢Í§Ç§¨ÃµÃǨËÒÊÑ­ÅÑ¡É³ì ¹Í¡¨Ò¡¨ÐÁÕÅѡɳÐà» ¹ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ (colored noise) áÅéÇ Âѧ¨ÐÁÕÅÑ¡É³Ð¢Ö ¹ÍÂÙè¡Ñºáºº


150

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¢éÍÁÙÅ (data pattern) ´éÇ àªè¹ ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì ¢Í§Ê× Í ºÑ¹·Ö¡ (media jitter noise) â´Â ·Õ ÃдѺ¤ÇÒÁÃعáç¢Í§ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡¨Ð¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ·Õ à¢Õ¹ŧä»ã¹ Ê× Í ºÑ¹·Ö¡ ǧ¨ÃµÃǨËÒ PDNP ¨Ö§ ä´é ¶Ù¡ Í͡Ẻ¢Ö ¹ ÁÒ à¾× Í ¨Ñ´¡ÒáѺ ÊÑ­­Ò³Ãº¡Ç¹àËÅèÒ¹Õ â´Â ¨Ðà» ¹¡Ò÷ӧҹÃèÇÁ¡Ñ¹ÃÐËÇèҧǧ¨Ã¡Ãͧ·Ó¹ÒÂÊÑ­­Ò³Ãº¡Ç¹¡ÑºÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ ¶Ö§áÁéÇèÒ Ç§¨ÃµÃǨËÒ PDNP ¨ÐãËé »ÃÐÊÔ·¸ÔÀÒ¾·Õ ´Õ ¡ÇèÒ Ç§¨ÃµÃǨËÒ PRML ÁÒ¡ áµè ǧ¨Ã µÃǨËÒ PDNP ÁÕ ¤ÇÒÁ«Ñº«é͹ÁÒ¡¡ÇèÒ Ç§¨ÃµÃǨËÒ PRML à¾ÃÒÐÇèÒ á¼¹ÀÒ¾à·ÃÅÅÔÊ ·Õ ãªé 㹠ǧ¨ÃµÃǨËÒ PDNP ¨ÐÁÕ ¨Ó¹Ç¹Ê¶Ò¹Ðà¾Ô Á ÁÒ¡¢Ö ¹ à¹× ͧÁÒ¨Ò¡¨Ó¹Ç¹á·ç» ¢Í§Ç§¨Ã¡Ãͧ·Ó¹Ò ÊÑ­­Ò³Ãº¡Ç¹ » ­ËÒ¹Õ ÊÒÁÒö·Õ ¨Ðá¡éä¢ä´é â´Â ¡ÒûÃÐÂØ¡µì ãªé ÍÑÅ¡ÍÃÔ·ÖÁ PDNP µÒÁá¹Ç¤Ô´ ¢Í§ PSP «Ö §¨Ðä´é¼ÅÅѾ¸ìà» ¹ ǧ¨ÃµÃǨËÒ PS PDNP «Ö §ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ã¡Åéà¤Õ§¡ÑºÇ§¨ÃµÃǨËÒ PDNP áµèÁÕ¤ÇÒÁ«Ñº«é͹¹éÍ¡ÇèÒÁÒ¡

7.8

à຺½ ¡ËÑ´·éÒº·

1. ¨§Í¸ÔºÒÂ·Õ ÁҢͧá¹Ç¤Ô´¢Í§ÍÑÅ¡ÍÃÔ·ÖÁ PDNP

2. ¨§¾ÔÊÙ¨¹ì¤èÒ¤ÇÒÁá»Ã»Ãǹ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ µÒÁÊÁ¡Òà (7.13)

3. ¨§¾ÔÊÙ¨¹ì¤èÒàÁµÃÔ¡ÊÒ¢Ò·Õ ãªéã¹Ç§¨ÃµÃǨËÒ PDNP µÒÁÊÁ¡Òà (7.14)

4. ¨§Í¸ÔºÒ¤ÇÒÁᵡµèÒ§¢Í§Ç§¨ÃµÃǨËÒ PRML, ǧ¨ÃµÃǨËÒ NPML, áÅÐǧ¨ÃµÃǨËÒ PDNP

5. ¨§Í¸ÔºÒÂËÅÑ¡¡Ò÷ӧҹ¢Í§Ç§¨ÃµÃǨËÒ PS PDNP

6. ¨§¾ÔÊÙ¨¹ì¨Ó¹Ç¹¢Í§µÑÇ´Óà¹Ô¹¡Òà (¡ÒúǡáÅСÒäٳ) ·Õ ãªéµèÍ¢éÍÁÙÅ 1 ºÔµ µÒÁ·Õ áÊ´§ã¹ µÒÃÒ§·Õ 7.1

7. ¨§¾ÔÊÙ¨¹ì¨Ó¹Ç¹Ë¹èǤÇÒÁ¨Ó·Õ µéͧ¡Òà µÒÁ·Õ áÊ´§ã¹µÒÃÒ§·Õ 7.1


º··Õ 8

¡ÒÃÍÍ¡à຺ÃËÑÊ RLL

㹺·¹Õ ¨Ð͸ԺÒ¶֧˹éÒ·Õ áÅÐËÅÑ¡¡Ò÷ӧҹ¢Í§ÃËÑÊ RLL (run length limited) [9] «Ö §à» ¹·Õ ¹ÔÂÁ ãªé§Ò¹ÁÒ¡ã¹Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ¾ÃéÍÁ·Ñ §áÊ´§¢Ñ ¹µÍ¹¡ÒÃÍ͡Ẻ ÃËÑÊ RLL ÍÂèÒ§§èÒ à¾× Íãªé㹡ÒÃà¢éÒáÅжʹÃËÑÊ¢éÍÁÙÅ

8.1

º·¹Ó

ÃËÑÊ RLL ¤×Í ÃËÑÊÁÍ´ÙàŪѹ (modulation code) »ÃÐàÀ·Ë¹Ö §·Õ ¹ÔÂÁãªéÁÒ¡ã¹ÍØ»¡Ã³ìÎÒÃì´´ÔÊ¡ì ä´Ã¿ì â´Â¨Ð·Ó˹éÒ·Õ ã¹¡ÒáÓ˹´¨Ó¹Ç¹¢Í§ºÔµ 0 áÅкԵ 1 (µÒÁÃٻẺ¢Í§ NRZI) ·Õ àÃÕ§ µÔ´ ¡Ñ¹ ã¹ÅӴѺ ¢éÍÁÙÅ ·Õ µéͧ¡ÒèÐà¢Õ¹ŧä»ã¹Ê× Í ºÑ¹·Ö¡ â´Â·Ñ Çä» ÃËÑÊ RLL ¨Ð¶Ù¡ ¡Ó˹´´éÇ ¾ÒÃÒÁÔàµÍÃì 4 µÑÇ ¤×Í

1)

m

2)

n

m, n, d,

áÅÐ

k

â´Â¨ÐÍÂÙèã¹ÃÙ»¢Í§ÃËÑÊ

m/n (d, k)

àÁ× Í

¤×Í ¨Ó¹Ç¹¢éÍÁÙźԵÍÔ¹¾Øµ (µèÍ¡ÒÃà¢éÒÃËÑÊË¹Ö §¤ÃÑ §) ·Õ ¨Ð·Ó¡ÒÃà¢éÒÃËÑÊ RLL

¤×Í ¨Ó¹Ç¹¢éÍÁÙźԵàÍÒµì¾Øµ (µèÍ¡ÒÃà¢éÒÃËÑÊË¹Ö §¤ÃÑ §) ·Õ ä´é¨Ò¡¡ÒÃà¢éÒÃËÑÊ RLL â´Â·Ñ Çä»

n≥m

àÊÁÍ

3)

d

¤×Í àÅ¢¨Ó¹Ç¹àµçÁ·Õ ¡Ó˹´¨Ó¹Ç¹·Õ¹ éÍÂ·Õ ÊØ´¢Í§ºÔµ 0 ·Õ ÍÂÙèÃÐËÇèÒ§ºÔµ 1

4)

k

¤×Í àÅ¢¨Ó¹Ç¹àµçÁ·Õ ¡Ó˹´¨Ó¹Ç¹·ÕÁ Ò¡·Õ ÊØ´¢Í§ºÔµ 0 ·Õ ÍÂÙèÃÐËÇèÒ§ºÔµ 1 151


152

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

# input bits

RLL code

(m bits)

# output bits (n bits)

ÃÙ»·Õ 8.1: Ẻ¨ÓÅͧ¡ÒÃà¢éÒÃËÑÊ RLL

àÁ× Í ¢éÍÁÙźԵ 1 ¨ÐÊÍ´¤Åéͧ¡Ñº¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð (transition) ¢Í§¡ÃÐáÊä¿¿ Òà¢Õ¹ (write cur rent) ·Õ ¨Ð» ͹à¢éÒä»ã¹ËÑÇà¢Õ¹ (write head) à¾× Í·ÓãËéÊ× ÍºÑ¹·Ö¡ ³ ºÃÔàdz·Õ µéͧ¡ÒèÐà¢Õ¹¢éÍÁÙŠŧä»ÁÕÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ (magnetization) µÒÁ·Õ µéͧ¡Òà Êèǹ¢éÍÁÙźԵ 0 ËÁÒ¶֧ äÁèÁÕ¡Òà à»ÅÕ Â¹Ê¶Ò¹Ð¢Í§¡ÃÐáÊä¿¿ Òà¢Õ¹ à¾ÃÒÐ©Ð¹Ñ ¹ ¾ÒÃÒÁÔàµÍÃì

d

¨ÐªèÇ·ÓãËé ºÔµ 1 ÊͧºÔµ ÍÂÙèËèÒ§

¡Ñ¹ «Ö § ¨ÐªèÇÂÅ´¼Å¡Ãзº¢Í§¡ÒÃá·Ã¡ÊÍ´ÃÐËÇèÒ§ÊÑ­Åѡɳì (ISI: intersymbol interference) Êèǹ¾ÒÃÒÁÔàµÍÃì

k

¨ÐªèÇÂÃѺ»ÃСѹÇèÒ ÅӴѺ¢éÍÁÙÅ·Õ ¨Ðà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡¨ÐÁÕºÔµà»ÅÕ Â¹Ê¶Ò¹Ð

à¡Ô´¢Ö ¹ÊÁ ÓàÊÁÍà¾Õ§¾Í à¾× Í·Õ ¨Ð·ÓãËéÃкºä·ÁÁÔ §ÃԤѿàÇÍÐÃÕ (timing recovery) ÊÒÁÒö·Ó§Ò¹ä´é ÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ µÑÇÍÂèÒ§àªè¹ ¶éÒÅӴѺ¢éÍÁÙÅ·Õ ¨Ðà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡ ¤×Í

···11111 ··· 11111··· ÅӴѺ ¢éÍÁÙÅ ¹Õ ¶×Í ÇèÒ à» ¹ ÅӴѺ ¢éÍÁÙÅ ·Õ äÁè ´Õ à¹× ͧ¨Ò¡ ¨Ð·ÓãËé à¡Ô´ » ­ËÒ ISI ÍÂèÒ§Ãعáç ã¹·Ò§ µÃ§¡Ñ¹¢éÒÁ ¶éÒÅӴѺ¢éÍÁÙÅ·Õ ¨Ðà¢Õ¹ŧä»ã¹Ê× ÍºÑ¹·Ö¡ ¤×Í

···00000 ··· 00000··· ¡ç ¨Ð¶×Í ÇèÒ à» ¹ ÅӴѺ ¢éÍÁÙÅ ·Õ äÁè ´Õ àªè¹¡Ñ¹ à¹× ͧ¨Ò¡ ¨Ð·ÓãËé à¡Ô´ » ­ËÒàÃ× Í§¡ÒÃà¢éÒ ¨Ñ§ËÇÐ (synchro nization) ¢Í§Ãкºä·ÁÁÔ § ÃÔ ¤Ñ¿àÇÍÐÃÕ à¾ÃÒÐ©Ð¹Ñ ¹ à¾× Í ËÅÕ¡àÅÕ Â§ÅӴѺ ¢éÍÁÙÅ ·Ñ § 2 Ẻ¹Õ ¨Ö§ ÁÕ ¤ÇÒÁ¨Óà» ¹ ·Õ ¨Ðµéͧà¢éÒ ÃËÑÊ ÅӴѺ ¢éÍÁÙÅ ´éÇÂÃËÑÊ RLL «Ö § ã¹·Ò§»¯ÔºÑµÔ áÅéÇ ¡ÒÃà¢éÒ áÅжʹÃËÑÊ ´éÇÂÃËÑÊ RLL ÊÒÁÒö·Óä´é§èÒÂâ´Â¡ÒÃãªé µÒÃÒ§¤é¹ËÒ (look up table) 㹡ÒÃà¢éÒáÅжʹÃËÑÊ ¢éÍÁÙÅ ÃÙ»·Õ 8.1 áÊ´§áºº¨ÓÅͧ¡ÒÃà¢éÒÃËÑÊ RLL â´Â·Õ ÍѵÃÒÃËÑÊ (code rate) ¨Ð¹ÔÂÒÁâ´Â ¨Ó¹Ç¹ ºÔµÍÔ¹¾Øµ

m

ËÒôéǨӹǹºÔµàÍÒµì¾Øµ

n

¹Ñ ¹¤×Í

R=

m ≤1 n

(8.1)


8.2.

¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

à¹× ͧ¨Ò¡ ¨Ó¹Ç¹ºÔµàÍÒµì¾Øµ

m

n

(D, K)

153

·Õ ä´é¨Ò¡¡ÒÃà¢éÒÃËÑʨÐÁըӹǹÁÒ¡¡ÇèÒËÃ×Íà·èҡѺ¨Ó¹Ç¹ºÔµÍÔ¹¾Øµ

àÊÁÍ ´Ñ§¹Ñ ¹ ¢éÍàÊÕ¢ͧ¡ÒÃà¢éÒÃËÑÊ RLL ·Õ àËç¹ä´éªÑ´à¨¹¡ç¤×Í ¨Ð·ÓãËéà¡Ô´ ºÔµÊèǹà¡Ô¹ (redun

dant bit) «Ö § ¨Ð·ÓãËé ÊÙ­àÊÕ à¹× Í·Õ ¡ÒèѴ à¡çº ¢éÍÁÙÅ ·Õ µéͧ¡ÒÃã¹ÎÒÃì´ ´ÔÊ¡ì ä´Ã¿ì 仺ҧÊèǹ ´Ñ§¹Ñ ¹ 㹡ÒÃàÅ×Í¡ÃËÑÊ RLL ã´ÁÒãªé §Ò¹ ¡ç ¤ÇÃ·Õ ¨ÐàÅ×Í¡ãªé ÃËÑÊ RLL ·Õ ÁÕ ÍѵÃÒÃËÑÊ

R

à¢éèÒã¡Åé ¤èÒ 1

ãËé ÁÒ¡·Õ ÊØ´ à¾× Í Å´¡ÒÃÊÙ­àÊÕ à¹× Í·Õ ¡ÒèѴ à¡çº ¢éÍÁÙÅ ·Õ µéͧ¡Òà ÊÓËÃѺ à¹× ÍËÒ㹺·¹Õ ¨Ð͸ԺÒ¶֧ ¢Ñ ¹µÍ¹¡ÒÃÍ͡ẺÃËÑÊ RLL ÍÂèÒ§§èÒ àÁ× Í¡Ó˹´¾ÒÃÒÁÔàµÍÃì

8.2

m, n, d,

áÅÐ

¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

¡Ó˹´ãËé ÅӴѺ ¢éÍÁÙÅ ÁÕ ¤ÇÒÁÂÒÇ·Ñ §ÊÔ ¹ ºÑ§¤Ñº (constraint)

(d, k)

L

k

ÁÒãËé

(d, k)

ºÔµ ¨Ó¹Ç¹ÅӴѺ ¢éÍÁÙÅ ·Ñ §ËÁ´ ·Õ ÊÍ´¤Åéͧ¡Ñº à§× ͹ä¢

ÊÒÁÒöËÒä´é¨Ò¡ÊÁ¡ÒõèÍä»¹Õ [9]

N (L) = L + 1, 1 ≤ L ≤ d + 1

(8.2)

N (L) = N (L − 1) + N (L − d − 1), d + 1 ≤ L ≤ k

(8.3)

N (L) = d + k + 1 − L +

k X

N (L − i − 1), k < L ≤ d + k

(8.4)

i=d

N (L) =

k X

N (L − i − 1), L > d + k

(8.5)

i=d

àÁ× Í

N (L) = 0

ÊÓËÃѺ

L<0

ã¹¡Ã³Õ ·Õ ¾ÒÃÒÁÔàµÍÃì

áÅÐ

k =∞

N (0) = 1

¨Ó¹Ç¹ÅӴѺ ¢éÍÁÙÅ ·Ñ §ËÁ´·Õ ÊÍ´¤Åéͧ¡Ñº à§× ͹䢺ѧ¤Ñº

(d, ∞)

¨ÐÊÒÁÒöËÒä´é¨Ò¡ÊÁ¡ÒõèÍ仹Õ

àÁ× Í

Nd (L) = 0

·Ñ §ËÁ´

Nd (L)

Nd (L) = L + 1, 1 ≤ L ≤ d + 1

(8.6)

Nd (L) = Nd (L − 1) + Nd (L − d − 1), L > d + 1

(8.7)

ÊÓËÃѺ

L<0

·Õ ÁÕ¤ÇÒÁÂÒÇ

L

áÅÐ

Nd (0) = 1

µÒÃÒ§·Õ 8.1 áÊ´§µÑÇÍÂèÒ§¨Ó¹Ç¹ÅӴѺ ¢éÍÁÙÅ

·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

(d, ∞)


154

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

µÒÃÒ§·Õ 8.1: ºÑ§¤Ñº

µÑÇÍÂèÒ§¨Ó¹Ç¹ÅӴѺ ¢éÍÁÙÅ ·Ñ §ËÁ´

Nd (L)

·Õ ÁÕ ¤ÇÒÁÂÒÇ

L

·Õ ÊÍ´¤Åéͧ¡Ñº à§× ͹ä¢

(d, ∞) d

L=4

L=5

L=6

L=7

L=8

L=9

L = 10

1

8

13

21

34

55

89

144

2

6

9

13

19

28

41

60

3

5

7

10

14

19

26

36

4

5

6

8

11

15

20

26

5

5

6

7

9

12

16

21

àÁ× Í·ÃÒº¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´

K

·Ñ §ËÁ´

N (L)

·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

(d, k)

·Õ ÊÒÁÒö¹ÓÁÒãªéá·¹ÅӴѺ¢éÍÁÙÅáµèÅÐÅӴѺ¨ÐÁÕ¤èÒà·èҡѺ

K = d log2 {N (L)} e (bits) àÁ× Í

dxe

áÅéÇ ¨Ó¹Ç¹ºÔµ

(8.8)

x

á·¹¨Ó¹Ç¹àµçÁ ºÇ¡·Õ ¹éÍÂ·Õ ÊØ´ ·Õ ÁÕ ¤èÒ ÁÒ¡¡ÇèÒ ËÃ×Í à·èÒ ¡Ñº ¤èÒ

ÊÒÁÒöãªé¢éÍÁÙźԵ¨Ó¹Ç¹ 3 ºÔµ

{000,

àªè¹ ¶éÒ

N4 (6) = 8

¡ç

001, 010, 011, 100, 101, 110, 111} 㹡ÒÃá·¹ÅӴѺ

¢éÍÁÙÅáµèÅÐẺ

8.3

¤ÇÒÁ¨Ø¢Í§ÃËÑÊ RLL à຺

(d, k)

ã¹·ÄɯբͧÃкºÊ× ÍÊÒà ¤ÇÒÁ¨Ø (capacity) ËÁÒ¶֧ ¤èÒÊÙ§ÊØ´¢Í§ÍѵÃÒÃËÑÊ

R

·Õ ÊÒÁÒö·ÓãËé

ÊÑÁÄ·¸Ô¼Åä´é «Ö §¨Ð¹ÔÂÒÁâ´Â [9]

1 log2 {N (L)} L→∞ n

C(d, k) = lim àÁ× Í

N (L)

C(d, k)

¤×Í ¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

(8.9)

(d, k)

¹Í¡¨Ò¡¹Õ ¤èÒ¤ÇÒÁ¨Ø

Âѧ à» ¹ ¾ÒÃÒÁÔàµÍÃì ·Õ ºè§ºÍ¡¶Ö§ ¤ÇÒÁÊÒÁÒö㹡ÒèѴ à¡çº ¢éÍÁÙÅ ¢èÒÇÊÒâͧ¼Ùéãªé ·Õ µéͧ¡ÒÃ

¨Ðà¢Õ¹ŧä»ã¹Ê× Í ºÑ¹·Ö¡ (äÁè ¹Ñº ºÔµ Êèǹà¡Ô¹) ¹Ñ ¹¤×Í ¶éÒ ¤èÒ ÊÒÁÒö¨Ñ´à¡çº¢éÍÁÙÅ¢èÒÇÊÒâͧ¼Ùéãªéä´éÁÒ¡àªè¹¡Ñ¹

C(d, k)

ÂÔ § ÁÒ¡ ¡ç áÊ´§ÇèÒ Ãкº


8.3.

¤ÇÒÁ¨Ø¢Í§ÃËÑÊ RLL à຺

(D, K)

155

¹Í¡¨Ò¡¹Õ 㹡ÒÃà»ÃÕºà·Õº»ÃÐÊÔ·¸ÔÀÒ¾¢Í§ÃËÑÊ RLL ẺµèÒ§æ ÊÒÁÒö·Óä´éâ´Â¡ÒþԨÒÃ³Ò ·Õ ¤èÒ »ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ (code e ciency) «Ö §¹ÔÂÒÁâ´Â

η=

R C(d, k)

(8.10)

¡ÅèÒǤ×Í ÃËÑÊ RLL ·Õ ÁÕ¤èÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊÁÒ¡ ¡çáÊ´§ÇèÒÁÕ»ÃÐÊÔ·¸ÔÀҾ㹡ÒÃãªé§Ò¹ÊÙ§

8.3.1

(d, k)

ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ¢Í§ÃËÑÊ RLL à຺

à¹× ͧ¨Ò¡ ¡ÒäӹdzËÒ¤èÒ ¤ÇÒÁ¨Ø

C(d, k)

´Ñ§¹Ñ ¹ â´Â·Ñ Çä» ¨Ö§¹ÔÂÁ¤Ó¹Ç³ËÒ¤èÒ

L → ∞

ã¹ÊÁ¡Òà (8.9) àÁ× Í

C(d, k)

·Óä´é ¤è͹¢éÒ§ÅÓºÒ¡

¨Ò¡ ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ (asymptotic infor

mation rate) «Ö §¹ÔÂÒÁâ´Â [58]

C(d, k) = log2 {λmax } àÁ× Í

λmax

¤×Í ÃÒ¡¨Ó¹Ç¹¨ÃÔ§ (real root) ·Õ ÁÕ¤èÒÁÒ¡ÊØ´ ¢Í§ÊÁ¡ÒÃ

xk+2 − xk+1 − xk−d+1 + 1 = 0,

k<∞

(8.12)

xd+1 − xd − 1 = 0,

k=∞

(8.13)

µÒÃÒ§·Õ 8.2 áÊ´§ÍѵÃÒ¢èÒÇÊÒÃàªÔ§ àÊé¹ ¡Ó¡Ñº

C(d, k)

¡ÒÃá¡éÊÁ¡Òà (8.11) (8.13) áÅÐàÁ× Í¾Ô¨ÒóҤèÒ ãªé·Ñ Ç仨ÐÁÕ¤èÒ¾ÒÃÒÁÔàµÍÃì

8.3.2

(8.11)

d≤2

¢Í§ÃËÑÊ RLL Ẻ

C(d, k)

(d, k)

µèÒ§æ ·Õ ä´é ¨Ò¡

ã¹µÒÃÒ§·Õ 8.2 ¨Ð¾ºÇèÒ ÃËÑÊ RLL ·Õ

àÊÁÍ à¾× Í·Õ ¨ÐÃѺ»ÃСѹä´éÇèÒÍѵÃÒÃËÑÊ

R ≥ 1/2

ÍѵÃÒ¤ÇÒÁ˹Òàà¹è¹

ÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR (density ratio) ËÃ×Í ¤ÇÒÁ˹Òá¹è¹ ¡ÒúÃèØ

(packing density) à» ¹

¾ÒÃÒÁÔàµÍÃì ·Õ ºè§ºÍ¡¶Ö§ ÃÐÂзҧ·Ò§¡ÒÂÀÒ¾ (physical distance) ÃÐËÇèÒ§µÓáË¹è§ ¢Í§¡ÒÃà»ÅÕ Â¹ Ê¶Ò¹Ð·Õ µÔ´¡Ñ¹ 2 µÓáË¹è§ ¢Í§ÅӴѺ¢éÍÁÙÅ·Õ à¢éÒÃËÑÊ RLL «Ö §¹ÔÂÒÁâ´Â

DR = (1 + d)R

(8.14)


156

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

µÒÃÒ§·Õ 8.2: ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ¢Í§ÃËÑÊ RLL Ẻ

â´Â·Õ

R

d=1

d=2

d=3

d=4

(d, k)

k

d=0

1

0.6942

2

0.8792

0.4057

3

0.9468

0.5515

0.2878

4

0.9752

0.6174

0.4057

0.2232

5

0.9881

0.6509

0.4650

0.3218

0.1823

10

0.9997

0.6909

0.5418

0.4460

0.3746

0.3158

15

0.9999

0.6939

0.5501

0.4615

0.3991

0.3513

1.0000

0.6942

0.5515

0.4650

0.4057

0.3620

µèÒ§æ

d=5

¤×Í ÍѵÃÒÃËÑÊ

µÒÃÒ§·Õ 8.3 áÊ´§¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¤ÇÒÁ¨Ø àÁ× Í¤ÇÒÁ¨Ø

C(d, k)

C(d, k) áÅÐÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR ¨ÐàËç¹ä´éÇèÒ

Ŵŧ ¤èÒÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR ¡ç¨Ðà¾Ô Á¢Ö ¹ ¤ÇÒÁÊÑÁ¾Ñ¹¸ì¹Õ ÊÒÁÒö͸ԺÒÂä´é

´Ñ§µèÍä»¹Õ ¨Ò¡ÊÁ¡Òà (8.14) àÁ× Í¾ÒÃÒÁÔàµÍÃì

d

à¾Ô Á¢Ö ¹ ¤èÒ DR ¡ç¨Ðà¾Ô Á¢Ö ¹ áµè¾ÒÃÒÁÔàµÍÃì

d

¢Ö ¹ ¹Õ ÁÕ ¤ÇÒÁËÁÒÂÇèÒ ¢éÍÁÙÅ ·Õ ¶Ù¡ à¢éÒ ÃËÑÊ ¨ÐÁÕ ºÔµ Êèǹà¡Ô¹ à¾Ô Á ÁÒ¡¢Ö ¹ (à¾ÃÒÐÇèÒ ¾ÒÃÒÁÔàµÍÃì

·Õ à¾Ô Á

d

¤×Í

¨Ó¹Ç¹ºÔµ 0 ¹éÍÂÊØ´ ·Õ ÍÂÙè ÃÐËÇèÒ§ºÔµ 1) ¶éÒ ¾Ô¨ÒóÒÇèÒ Ê× Í ºÑ¹·Ö¡ ÁÕ à¹× Í·Õ ã¹¡ÒèѴ à¡çº ¢éÍÁÙÅ ·Õ ¨Ó¡Ñ´ ´Ñ§¹Ñ ¹ ÃкºÊÒÁÒö·Õ ¨Ð¨Ñ´à¡çº¢éÍÁÙÅ¢èÒÇÊÒâͧ¼Ùéãªéä´é¹éÍÂŧ à¹× ͧ¨Ò¡ µéͧàËÅ×Íà¹× Í·Õ ºÒ§ÊèǹäÇé ÊÓËÃѺ¨Ñ´à¡çººÔµÊèǹà¡Ô¹ à¾ÃÒÐ©Ð¹Ñ ¹ ¨Ö§Ê觼ŷÓãËé¤èÒ¤ÇÒÁ¨Ø

8.4

C(d, k)

·Õ ¤Ó¹Ç³ä´éÁÕ¤èÒ¹éÍÂŧ

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ (FSM: nite state machine) ¢Í§ÃËÑÊ RLL ¨ÐáÊ´§ãËéàË繶֧ ¡ÒÃà»ÅÕ Â¹á»Å§ ¢Í§Ê¶Ò¹Ðã¹ÃËÑÊ RLL µÒÁà§× ͹䢺ѧ¤Ñº ¢Í§¾ÒÃÒÁÔàµÍÃì

(d, k)

¨ÐÁÕà¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´µÒÁÃÙ»·Õ 8.2 àÁ× Í

Si

(d, k)

¤×Í Ê¶Ò¹Ð

i

µÑÇÍÂèÒ§àªè¹ ÃËÑÊ RLL Ẻ

áÅеÑÇàÅ¢·Õ áÊ´§ÍÂÙèµÒÁàÊé¹ÅÙ¡ÈÃ

¤×Í ¢éÍÁÙźԵàÍÒµì¾Øµ·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº¢Í§¾ÒÃÒÁÔàµÍÃì

(d, k) ¨Ò¡ÃÙ»·Õ 8.2 ʶҹÐàÃÔ Áµé¹


8.4.

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL

157

µÒÃÒ§·Õ 8.3: ¤ÇÒÁÊÑÁ¾Ñ¹¸ìÃÐËÇèÒ§¤ÇÒÁ¨Ø

0

S1

S2

0

C(d, k)

d

C(d, ∞)

1

0.6942

1.3884

2

0.5515

1.6545

3

0.4650

1.8600

4

0.4057

2.0285

5

0.3620

2.1720

DR =

0

Sd

áÅÐÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR

(1 + d)C(d, ∞)

S d +1

0

Sk

1

1

ÃÙ»·Õ 8.2: à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

¨ÐÍÂÙè·Õ ʶҹÐ

S1

S k +1

1

(d, k)

«Ö §ãËé¶×ÍÇèÒà» è¹à˵ءÒóì·Õ à¨ÍºÔµ 1 µÑÇááã¹ÅӴѺ¢éÍÁÙÅ à¾ÃÒÐ©Ð¹Ñ ¹ ºÔµµèÍä»

¨Ðµéͧ໠¹ºÔµ 0 à» ¹¨Ó¹Ç¹ÍÂèÒ§¹éÍ ä»Âѧ ʶҹÐ

0

Sd+1 )

d

µÑǵԴµè͡ѹ (¹Ñ ¹¤×Í Ê¶Ò¹Ð

¾ÍÅӴѺ ¢éÍÁÙÅ ÁÕ ºÔµ 0 ¤Ãº

d

S1

¡ç¨Ðà´Ô¹·Ò§à» ¹àÊ鹵ç

µÑÇ áÅéÇ ¨Ò¡à§× ͹䢺ѧ¤Ñº

(d, k)

áÊ´§ÇèÒ ºÔµ

µÑǶѴä»ÊÒÁÒö໠¹ä´é·Ñ §ºÔµ 0 ËÃ×ͺԵ 1 «Ö §¶éÒà» ¹ºÔµ 1 àÁ× Íã´ Ãкº¡ç¨ÐµéͧÇÔ §¡ÅѺä»àÃÔ Áµé¹·Õ ʶҹÐ

S1

ãËÁè áµè¶éÒà» ¹ºÔµ 0 ¡ç¨ÐÁÕºÔµ 0 ä´éÍÕ¡äÁèà¡Ô¹

k−d

µÑÇ áÅÐàÁ× ÍÁÕºÔµ 0 µÔ´µè͡ѹ¤Ãº

µÑÇáÅéÇ ºÔµµÑǶѴ仨еéͧ໠¹ºÔµ 1 à·èÒ¹Ñ ¹ ¹Ñ ¹¤×Í Ãкº¨Ð¶Ù¡ºÑ§¤ÑºãËé¡ÅѺä»àÃÔ Áµé¹·Õ ʶҹÐ

k

S1

ãËÁèâ´ÂÍѵâ¹ÁѵÔ

µÑÇÍÂèÒ§·Õ 8.1

¨§áÊ´§á¼¹ÀÒ¾à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL µÒÁà§× ͹䢺ѧ¤Ñº¢Í§¾ÒÃÒÁÔàµÍÃì

(d, k) = (1, 3) ÇÔ¸Õ·Ó

¾ÒÃÒÁÔàµÍÃì

(1, 3)

ËÁÒ¶֧ ÅӴѺ¢éÍÁÙŨÐÁÕºÔµ 0 ÍÂèÒ§¹éÍÂË¹Ö §µÑÇ ËÃ×ÍÍÂèÒ§ÁÒ¡ÊÒÁµÑÇ


158

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

S1

0

0

S2

0

S3

1

S4

1

1

ÃÙ»·Õ 8.3: à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(1, 3)

µÔ´µè͡ѹ ·Õ ÍÂÙèÃÐËÇèÒ§ºÔµ 1 «Ö §ÊÒÁÒöà¢Õ¹໠¹à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ä´é µÒÁÃÙ»·Õ 8.3

8.5

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ ¢Í§ÃËÑÊ RLL Ẻ

(d, k)

ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè ã¹ÃÙ» ¢Í§ àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹

ʶҹР(state transition matrix) ä´é «Ö §¹ÔÂÒÁâ´Â àÁ·ÃÔ¡«ì á¹ÇµÑ § â´Â·Õ ÊÁÒªÔ¡¢Í§àÁ·ÃÔ¡«ì

D(i, j)

¹Ñ ¹¤×Í á¶Ç·Õ

i

D

·Õ ÁÕ¢¹Ò´

áÅÐá¹ÇµÑ §·Õ

(k + 1) j

á¶Ç áÅÐ

¨Ð¶Ù¡¡Ó˹´â´Â

D(i, 1) = 1, i ≥ d + 1   1, j = i + 1 D(i, j) =  0, else µÑÇÍÂèÒ§àªè¹ à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ðä´é´Ñ§¹Õ

(1, 3)

(k + 1)

(8.15)

µÒÁÃÙ»·Õ 8.3 ÊÒÁÒöà¢Õ¹໠¹àÁ·ÃÔ¡«ì

 0 1 0 0

     1 0 1 0   D=    1 0 0 1    1 0 0 0 àÁ·ÃÔ¡«ì

D

(8.16)

ã¹ÊÁ¡Òà (8.16) ÊÒÁÒöÊÃéÒ§ä´é ´Ñ§µèÍä»¹Õ ¶éÒ ¡Ó˹´ãËé áµèÅÐá¶Ç᷹ʶҹÐ

áµèÅРʶҹР¡ÅèÒǤ×Í á¶Ç ·Õ Ë¹Ö § ãªé á·¹ ʶҹÐ

S1

áÅÐ á¶Ç ·Õ Êͧ ãªé á·¹ ʶҹÐ

S2

àªè¹à´ÕÂǡѹãËéáµèÅÐá¹ÇµÑ §á·¹Ê¶Ò¹ÐáµèÅÐʶҹР¡ÅèÒǤ×Í á¹ÇµÑ §·Õ Ë¹Ö §ãªé᷹ʶҹÐ

à» ¹µé¹

S1

áÅÐ


8.5.

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

159

S2

à» ¹µé¹ ´Ñ§¹Ñ ¹ 㹡ÒÃÊÃéÒ§àÁ·ÃÔ¡«ì ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¨Ò¡à¤Ã× Í§

á¹ÇµÑ § ·Õ Êͧãªé ᷹ʶҹÐ

ʶҹШӡѴ¢Í§ÃËÑÊ RLL Ẻ (ʶҹÐ

S1 )

(1, 3)

µÒÁÃÙ»·Õ 8.3 ãËé¾Ô¨ÒóҷÕÅÐá¹ÇµÑ § àªè¹ ã¹á¹ÇµÑ §·Õ Ë¹Ö §

Si

ãËé ´Ù ÇèÒ ÁÕ àÊé¹ ÅÙ¡ÈèҡʶҹÐ

ÇèÒ ÁÕàÊé¹ÅÙ¡ÈèҡʶҹÐ

S2 , S3 ,

áÅÐ

S4

ã´ºéÒ§·Õ ÇÔ § à¢éÒ ÁÒ·Õ Ê¶Ò¹Ð

S1

¨Ò¡ÃÙ» ·Õ 8.3 ¨Ð¾º

´Ñ§¹Ñ ¹ ã¹á¹ÇµÑ §·Õ Ë¹Ö §¹Õ ¤èÒ 1 ¨Ð¶Ù¡ãÊèà¢éÒä»ã¹á¶Ç·Õ

Êͧ, á¶Ç·Õ ÊÒÁ, áÅÐá¶Ç·Õ ÊÕ Êèǹá¶Ç·Õ Ë¹Ö §¨ÐãËéà» ¹¤èÒ 0 㹷ӹͧà´ÕÂǡѹ ¶éÒ¾Ô¨ÒÃ³Ò·Õ á¹ÇµÑ § ·Õ Êͧ (ʶҹÐ

S2 )

ãËé´ÙÇèÒÁÕàÊé¹ÅÙ¡ÈèҡʶҹÐ

¾ºÇèÒ ÁÕàÊé¹ÅÙ¡ÈèҡʶҹÐ

S1

Si

ã´ºéÒ§·Õ ÇÔ §à¢éÒÁÒ·Õ Ê¶Ò¹Ð

àÊé¹à´ÕÂÇ·Õ ÇÔ §ÁÒ·Õ Ê¶Ò¹Ð

S2

S2

¨Ò¡ÃÙ»·Õ 8.3 ¨Ð

´Ñ§¹Ñ ¹ ¤èÒ 1 ¨Ð¶Ù¡ãÊèà¢éÒä»ã¹á¶Ç·Õ

Ë¹Ö § Êèǹá¶ÇÍ× ¹æ ¨ÐÁÕ¤èÒà» ¹¤èÒ 0 à» ¹µé¹ ÊÓËÃÑºã¹¡Ã³Õ·Õ ¾ÒÃÒÁÔàµÍÃì

k=∞

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

à¢Õ¹ãËéÍÂÙèã¹ÃÙ»¢Í§àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ·Õ ÊÁÒªÔ¡¢Í§àÁ·ÃÔ¡«ì

D(i, j)

D

·Õ ÁÕ¢¹Ò´

(d + 1)

á¶Ç áÅÐ

(d, ∞)

(d + 1)

ÊÒÁÒö

á¹ÇµÑ § â´Â

¨Ð¶Ù¡¡Ó˹´â´Â

D(i, j) = 1, j = i + 1 D(d + 1, 1) = D(d + 1, d + 1) = 1 D(i, j) = 0, else

µÑÇÍÂèÒ§·Õ 8.2

¨§áÊ´§á¼¹ÀÒ¾à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ áÅÐàÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ¢Í§ÃËÑÊ RLL

µÒÁà§× ͹䢺ѧ¤Ñº¢Í§¾ÒÃÒÁÔàµÍÃì

ÇÔ¸Õ ·Ó

(8.17)

¾ÒÃÒÁÔàµÍÃì

(0, 3)

(d, k) = (0, 3)

ËÁÒ¶֧ ÅӴѺ ¢éÍÁÙÅ ¨ÐÁÕ ºÔµ 0 ÍÂèÒ§¹éÍÂË¹Ö § µÑÇ ËÃ×Í ÍÂèÒ§ÁÒ¡ÊÒÁ

µÑÇ µÔ´µèÍ ¡Ñ¹ ·Õ ÍÂÙè ÃÐËÇèÒ§ ºÔµ 1 «Ö § ÊÒÁÒö à¢Õ¹ à» ¹ à¤Ã× Í§ ʶҹР¨Ó¡Ñ´ ä´é µÒÁ ÃÙ» ·Õ 8.4 â´Â ·Õ àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ÊÍ´¤Åéͧ¡Ñºà¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¹Õ ¤×Í

 1 1 0 0

     1 0 1 0   D=    1 0 0 1    1 0 0 0


160

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

1

0

S1

S2

0

0

S3

1

S4

1

1

ÃÙ»·Õ 8.4: à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

8.5.1

(0, 3)

¡ÒÃËÒÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

D

ÊÒÁÒö¹ÓÁÒãªé㹡ÒäӹdzËÒÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ

C(d, k)

¨Ò¡ÊÙµÃã¹ÊÁ¡Òà (8.11) ¹Ñ ¹¤×Í

© ª C(d, k) = log2 λD max â´Â·Õ

D λmax

ʶҹÐ

D

(8.18)

¤×Í ¤èÒÅѡɳÐ੾ÒШӹǹ¨ÃÔ§ (real eigenvalue) ·Õ ÁÕ¤èÒÁÒ¡ÊØ´¢Í§àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹

«Ö §ËÒä´é¨Ò¡¡ÒÃá¡éÊÁ¡ÒÃ

det(D − λI) = 0 àÁ× Í det(·) ¤×Í ¡ÒÃËÒ¤èÒ´Õà·ÍÃìÁÔá¹¹µì (determinant), ·Õ ÁÕ¢¹Ò´à·èҡѺàÁ·ÃÔ¡«ì

8.5.2

D,

áÅÐ

λ

¤×Í àÁ·ÃÔ¡«ìàÍ¡Åѡɳì (identity matrix)

¤×Í ¤èÒÅѡɳÐ੾ÒÐ

ÅӴѺ¢éÍÁÙÅ·Õ ÊÍ´¤Åéͧ¡Ñºà¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL à຺

¨Ò¡à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(d, k)

¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Õ à» ¹ä»ä´é·Ñ §ËÁ´·Õ ÁÕ¤ÇÒÁÂÒÇ

Sj

I

(8.19)

¨ÐÁÕ¤èÒà·èҡѺ ¤èҢͧÊÁÒªÔ¡á¶Ç·Õ

µÑÇÍÂèÒ§·Õ 8.3

i

·Õ áÊ´§ã¹ÃÙ»·Õ 8.2 àÁ× Í

L ºÔµ ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð Si

áÅÐá¹ÇµÑ §·Õ

j

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

¢éÍÁÙÅ·Õ à» ¹ä»ä´é·Ñ §ËÁ´·Õ ÁÕ¤ÇÒÁÂÒÇ

2

Si

¢Í§àÁ·ÃÔ¡«ì

(0, 2)

ºÔµ ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

DL

(d, k)

¤×Í Ê¶Ò¹Ð

i

´Ñ§¹Ñ ¹

áÅéÇä»ÊÔ ¹ÊØ´·Õ ʶҹÐ

¹Ñ ¹¤×Í

DL (i, j)

áÊ´§ã¹ÃÙ»·Õ 8.5 ¨§ËҨӹǹÅӴѺ

Si

áÅéÇä»ÊÔ ¹ÊØ´·Õ ʶҹÐ

Sj

ÊÓËÃѺ


8.5.

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

161

1 S1

0

0

S2

1

S3

1

ÃÙ»·Õ 8.5: à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(0, 2)

0 ≤ i, j ≤ 3 ÇÔ¸Õ ·Ó

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ ¢Í§ÃËÑÊ RLL Ẻ

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

D

(0, 2)

ã¹ÃÙ» ·Õ 8.5 ÊÒÁÒöà¢Õ¹ãËé ÍÂÙè ã¹ÃÙ» ¢Í§

ä´é ¤×Í

 1 1 0

  D= 1 0 1  1 0 0

   

2 ºÔµ   2 1 1     D2 =  2 1 0    1 1 0

´Ñ§¹Ñ ¹ ¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ·Õ à» ¹ä»ä´é·Ñ §ËÁ´·Õ ÁÕ¤ÇÒÁÂÒÇ

(8.20)

ËÒä´é¨Ò¡

D2

«Ö §ÁÕ¤èÒà·èҡѺ

(8.21)

ÊÁ¡Òà (8.21) ºÍ¡ãËé·ÃÒºÇèÒ

ÅӴѺ¢éÍÁÙÅ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

S1 → S1

Áըӹǹà·èҡѺ

D(1, 1) = 2

µÑÇ ¤×Í

{01,

11}

ÅӴѺ¢éÍÁÙÅ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

S2 → S1

Áըӹǹà·èҡѺ

D(2, 1) = 2

µÑÇ ¤×Í

{01,

11}

ÅӴѺ¢éÍÁÙÅ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

S3 → S1

Áըӹǹà·èҡѺ

D(3, 1) = 1

µÑÇ ¤×Í

{11}

ÅӴѺ¢éÍÁÙÅ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

S1 → S2

Áըӹǹà·èҡѺ

D(1, 2) = 2

µÑÇ ¤×Í

{10}


162

8.6

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¢Ñ ¹µÍ¹¡ÒÃÍÍ¡à຺ÃËÑÊ RLL

ã¹Êèǹ¹Õ ¨ÐáÊ´§¢Ñ ¹µÍ¹¡ÒÃÍ͡ẺµÒÃÒ§¤é¹ËÒ à¾× Íãªé㹡ÒÃà¢éÒáÅжʹÃËÑÊ RLL â´ÂãªéµÑÇÍÂèÒ§ ´Ñ§µèÍ仹Õ

µÑÇÍÂèÒ§·Õ 8.4

¨§ÊÃéÒ§µÒÃÒ§¤é¹ËÒÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ RLL Ẻ

¤ÃÑ §·Õ ¼èÒ¹¡ÒÃà¢éÒÃËÑÊáÅéǨÐÁÕ¤ÇÒÁÂÒÇà·èҡѺ

ÇÔ¸Õ·Ó

n=3

(0, 2)

â´Â·Õ ¢éÍÁÙÅáµèÅÐ

¾ÃéÍÁ·Ñ §ËÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

η

¡ÒÃÊÃéÒ§µÒÃÒ§¤é¹ËÒÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ RLL ÊÒÁÒöáºè§ÍÍ¡à» ¹ 4 ¢Ñ ¹µÍ¹ ´Ñ§¹Õ

¢Ñ ¹µÍ¹·Õ 1:

ãËé¾Ô¨ÒóҴÙÇèÒ ¢éÍÁÙÅ 3 ºÔµÁÕ·Ñ §ËÁ´¡Õ Ẻ «Ö §¨Ðä´éÇèÒ ÁÕ·Ñ §ËÁ´ 8 Ẻ ¤×Í

{000, 001, 010, 011, 100, 101, 110, 111} ¢Ñ ¹µÍ¹·Õ 2:

ãËé¾Ô¨ÒóҴÙÇèÒ ÁÕ¢éÍÁÙŪشä˹ºéÒ§·Õ äÁèÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

ÇèÒ ÁÕ¢éÍÁÙÅ·Ñ §ËÁ´ 7 Ẻ·Õ ¼èÒ¹à§× ͹䢺ѧ¤Ñº

(0, 2)

(0, 2)

«Ö §¨Ð¾º

¹Ñ ¹¤×Í

{001, 010, 011, 100, 101, 110, 111} ¢Ñ ¹µÍ¹·Õ 3:

ãËéÅͧ¹Ó¢éÍÁÙÅ·Õ ä´é¨Ò¡¢Ñ ¹µÍ¹·Õ 2 áµèÅеÑÇ ÁÒ·Ó¡Òõè͡ѹ·Ñ §·Ò§«éÒÂáÅзҧ¢ÇÒ

áÅéÇ´ÙÇèÒ ÁÕ¢éÍÁÙŵÑÇä˹ºéÒ§·Õ àÁ× Í¹ÓÁÒµè͡ѹáÅéÇ ¨ÐäÁèÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº ¾ºÇèÒ ¢éÍÁÙÅ 001 áÅÐ 100 àÁ× Í¹Ó仵è͡Ѻ¢éÍÁÙŵÑÇÍ× ¹¨Ð·ÓãËéà§× ͹䢺ѧ¤Ñº

(0, 2)

(0, 2)

ã¹·Õ ¹Õ ¨Ð

¼Ô´ä» ´Ñ§¹Ñ ¹

¢éÍÁÙÅ 001 áÅÐ 100 ¨Ðµéͧ¶Ù¡ µÑ´ ·Ô § ä» ·ÓãËé ¢éÍÁÙÅ ·Õ ËŧàËÅ×Í ÍÂÙè ÁÕ à¾Õ§ 5 Ẻ ·Õ ¼èÒ¹à§× Í¹ä¢ ºÑ§¤Ñº

(0, 2)

¹Ñ ¹¤×Í

{010, 011, 101, 110, 111} «Ö §¢éÍÁÙÅàËÅèÒ¹Õ ¡ç¤×Í ¢éÍÁÙÅ·Õ ÊÒÁÒö¹ÓÁÒãªéà» ¹¢éÍÁÙÅ·Õ ¼èÒ¹¡ÒÃà¢éÒÃËÑÊ RLL áÅéÇ ¢Ñ ¹µÍ¹·Õ 4:

¨Ò¡¢éÍÁÙÅ·Ñ § 5 Ẻ·Õ ä´éã¹¢Ñ ¹µÍ¹·Õ 3 ãËéàÅ×Í¡ÁÒ 4 Ẻ (Ẻ㴡çä´é) à¾× Íãªéã¹

¡ÒÃÊÃéÒ§µÒÃÒ§¤é¹ËÒ ÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ¢éÍÁÙÅÍÔ¹¾Øµ·ÕÅÐ 2 ºÔµ ¹Ñ ¹¤×Í 11} «Ö §¨Ðä´é µÒÁµÒÃÒ§·Õ 8.4 áÅШÐä´éÇèÒ ÍѵÃÒÃËÑÊ ¤×Í

{00,

01,

10,


8.6.

¢Ñ ¹µÍ¹¡ÒÃÍÍ¡à຺ÃËÑÊ RLL

163

µÒÃÒ§·Õ 8.4: µÒÃÒ§¤é¹ËÒÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ RLL Ẻ ¢éÍÁÙÅÍÔ¹¾Øµ

¢éÍÁÙÅàÍÒµì¾Øµ

00

010

01

011

10

101

11

110

R=

(0, 2)

2 3

㹷ӹͧà´ÕÂǡѹ ¡ÒÃËÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ RLL ÊÒÁÒöËÒä´é µÒÁ¢Ñ ¹µÍ¹µèÍä»¹Õ àÃÔ Áµé¹ ãËéÊÃéÒ§à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´¢Í§ÃËÑÊ RLL Ẻ

(0, 2)

«Ö §¨Ðä´éµÒÁÃÙ»·Õ 8.5 ¨Ò¡¹Ñ ¹ãËéÊÃéÒ§àÁ·ÃÔ¡«ì

¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð·Õ ÊÍ´¤Åéͧ¡Ñº à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ ã¹ÃÙ» ·Õ 8.5 «Ö § ¨Ðä´é ¼ÅÅѾ¸ì µÒÁÊÁ¡Òà (8.20) ¹Ñ ¹¤×Í

 1 1 0

  D= 1 0 1  1 0 0 ¢Ñ ¹µÍ¹µèÍÁÒ ¤×Í ¡ÒÃËÒ¤èÒ ÅѡɳÐ੾ÒТͧàÁ·ÃÔ¡«ì

   

D

«Ö § ÊÒÁÒöËÒä´é ¨Ò¡á¡é ÊÁ¡Òà (8.19)

¹Ñ ¹¤×Í



 1 1 0

 1 0 0

        det  1 0 1  − λ  0 1 0  = 0     1 0 0 0 0 1   1−λ 1 0     det  1 −λ 1  = 0   1 0 −λ −λ3 + λ2 + λ + 1 = 0

(8.22)


164

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

â´Â¡ÒÃá¡éÊÁ¡Òà (8.22) àÃÒ¨Ðä´éÇèÒ

λ = 1.8393, −0.4196 + 0.6063i, −0.4196 − 0.6063i ´Ñ§¹Ñ ¹

λD max = 1.8393

¨Ð¶Ù¡¹ÓÁÒãªé㹡ÒäӹdzËÒ¤ÇÒÁ¨Ø

C(d, k)

µÒÁÊÁ¡Òà (8.18) ¹Ñ ¹¤×Í

C(d, k) = log2 {λmax } = log2 {1.8393} = 0.87916 áÅлÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

η

ËÒä´é¨Ò¡ÊÁ¡Òà (8.10) «Ö §¨Ðä´éÇèÒ

η=

8.7

R 2/3 = = 0.7583 C(d, k) 0.87916

µÑÇÍÂèÒ§ÃËÑÊ RLL à຺µèÒ§æ

ÃËÑÊ RLL ÁÕËÅÒÂẺ¢Ö ¹ÍÂÙè¡Ñº¾ÒÃÒÁÔàµÍÃì

(d, k)

áÅÐÍѵÃÒÃËÑÊ

R

·Õ ãªé ã¹ÂؤàÃÔ Áµé¹¢Í§ÍØ»¡Ã³ì

ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ÃËÑÊ RLL ·Õ ãªé¨ÐÁÕª× ÍÇèÒ ÃËÑÊ FM (frequency modulation) â´ÂÁÕµÒÃÒ§¤é¹ËÒ ÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ µÒÁÃÙ»·Õ 8.6(a) â´ÂÃËÑÊ FM ¹Õ ¨Ðãªé§Ò¹ÃèÇÁ¡ÑºÇ§¨ÃµÃǨËҨشÊÙ§ÊØ´ (peak detector) áÅÐÁÕÍѵÃÒÃËÑÊ

R = 1/2 «Ö §¨Ð·ÓãËéµéͧÊÙ­àÊÕÂ¾× ¹·Õ ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ìä»»ÃÐÁÒ³

50% à¾× Íà¡çº¢éÍÁÙźԵÊèǹà¡Ô¹ µÑÇÍÂèÒ§¡ÒÃà¢éÒÃËÑÊàªè¹ ¶éÒÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ ¤×Í ¢éÍÁÙÅ·Õ ä´é¨Ò¡¡ÒÃà¢éÒÃËÑÊ FM ¤×Í

{110000}

ÅӴѺ

{11 11 01 01 01 01}

Êѧࡵ¨Ð¾ºÇèÒ ÃËÑÊ FM ÂÍÁãËéÅӴѺ¢éÍÁÙÅ·Õ à¢éÒÃËÑÊáÅéÇÁÕºÔµ 1 µÔ´¡Ñ¹ä´é «Ö §¨Ð¡èÍãËéà¡Ô´» ­ËÒ àÃ× Í§ ISI ´Ñ§¹Ñ ¹ ¨Ö§ÁÕ¡ÒþѲ¹ÒÃËÑÊãËÁè·Õ àÃÕ¡ÇèÒ ÃËÑÊ MFM (modi ed frequency modulation) ËÃ×ͺҧ¤ÃÑ §àÃÕ¡ÇèÒ ÃËÑÊÁÔÅàÅÍÃì (Miller code) «Ö §ÁÕµÒÃÒ§¤é¹ËÒÊÓËÃѺ¡ÒÃà¢éÒáÅжʹÃËÑÊ µÒÁ ÃÙ» ·Õ 8.6(b) â´Â·Õ

x = 0

¶×Í ÇèÒ à» ¹ ÃËÑÊ RLL Ẻ

0.5/0.5515 = 0.9066

¶éÒ ºÔµ ¡è͹˹éÒ ºÔµ

(1, 3)

x

ÁÕ ¤èÒ à» ¹ ºÔµ 1 ¹Í¡¹Ñ ¹

áÅÐÁÕ ÍѵÃÒÃËÑÊ

R = 1/2

{01 01 00 10 10 01 01}

ÃËÑÊ MFM ¹Õ

â´Â¨ÐÁÕ »ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

µÑÇÍÂèÒ§¡ÒÃà¢éÒÃËÑÊ àªè¹ ¶éÒÅӴѺ¢éÍÁÙÅÍÔ¹¾Øµ ¤×Í

·Õ ä´é¨Ò¡¡ÒÃà¢éÒÃËÑÊ MFM ¤×Í

x = 1

{1100011}

η =

ÅӴѺ¢éÍÁÙÅ

à» ¹µé¹ ¹Í¡¨Ò¡¹Õ ÃÙ»·Õ 8.6(c) áÅÐ 8.6(d)

áÊ´§µÑÇÍÂèÒ§ÃËÑÊ RLL ẺµèÒ§æ ·Õ ãªéã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ¨Ò¡ÇÔÇѲ¹Ò¡Òâͧ¡ÒþѲ¹ÒÃËÑÊ RLL


8.7.

µÑÇÍÂèÒ§ÃËÑÊ RLL à຺µèÒ§æ

165

user bits

coded bits

user bits

coded bits

0 1

01 11

0 1

x0 01

(a) FM code

(b) MFM code

user bits

coded bits

user bits

coded bits

00 01 10 11 0000 0001 1000 1001

101 100 001 010 101000 100000 001000 010000

10 11 000 010 011 0010 0011

0100 1000 000100 100100 001000 00100100 00001000

(c) 2/3 (1,7) RLL code

(d) 1/2 (1,7) RLL code

ÃÙ»·Õ 8.6: µÑÇÍÂèÒ§ÃËÑÊ RLL ẺµèÒ§æ ·Õ ãªéã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

¨Ð¾ºÇèÒ ¾ÒÃÒÁÔàµÍÃì

d

·Õ ãªéã¹ÃËÑÊ RLL ¨Ð¤èÍÂæ Ŵŧ à¾× ÍÅ´¨Ó¹Ç¹ºÔµÊèǹà¡Ô¹ ·ÓãËéÊÒÁÒö

¨Ñ´à¡çº¢éÍÁÙÅ·Õ µéͧ¡ÒÃã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ìä´éÁÒ¡¢Ö ¹ 㹡ÒõѴÊÔ¹ã¨ÇèÒ¨ÐàÍÒÃËÑÊ RLL ã´ÁÒãªé§Ò¹ã¹Ãкº¨Ð¢Ö ¹ÍÂÙè¡Ñº» ¨¨ÑÂËÅÒÂæ ÍÂèÒ§ ´Ñ§¹Õ 1) ¾ÒÃÒÁÔàµÍÃì

2) ÍѵÃÒÃËÑÊ

3) ¤ÇÒÁ¨Ø

(d, k)

R = m/n

C(d, k)

4) »ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

η

5) ÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR


166

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

«Ö §â´Â·Ñ Çä»áÅéÇ ¨Óà» ¹µéͧ»ÃйջÃйÍÁ» ¨¨ÑÂ·Ñ §ËÁ´ãËéàËÁÒÐÊÁ¡ÑºÊÀÒ¾áÇ´ÅéÍÁ㹡Ò÷ӧҹ ¢Í§Ãкº ÃËÑÊ RLL ·Õ à¤Âãªéã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì àªè¹ ÃËÑÊ RLL Ẻ

4/5 (0, 2),

áÅÐÃËÑÊ RLL Ẻ

8/9 (0, 3)

1/2 (2, 7),

ÃËÑÊ RLL Ẻ

à» ¹µé¹ Êѧࡵ¨Ð¾ºÇèÒ ÃËÑÊ RLL ·Õ ãªé¨ÐÁÕÍѵÃÒÃËÑÊà¢éÒ

ã¡Åé¤èÒ 1 àÃ× ÍÂæ à¾× ÍÅ´¨Ó¹Ç¹ºÔµÊèǹà¡Ô¹ áÅоÒÃÒÁÔàµÍÃì

d

·Õ ãªé¡ç¨ÐÁÕ¤èÒà» ¹¤èÒ 0 ¹Ñ ¹¤×Í ÂÍÁãËé

ÁÕºÔµ 1 µÔ´¡Ñ¹ä´é «Ö §¶Ö§áÁéÇèҨСèÍãËéà¡Ô´» ­ËÒàÃ× Í§ ISI áµèÃкº¡çÊÒÁÒö¨Ñ´¡ÒáѺ ISI ¹Õ ä´é´éÇ ෤¹Ô¤ PRML µÒÁ·Õ ͸ԺÒÂ㹺··Õ 4

8.8

ÃËÑÊ

(0, G/I)

ÊÓËÃѺªèͧÊÑ­­Ò³ PRML

ÊÓËÃѺªèͧÊÑ­­Ò³ PRML ·Õ ãªé·ÒÃìà¡çµáºº PR4, (·Õ äÁèÁÕÊÑ­­Ò³Ãº¡Ç¹) ³ àÇÅÒ

k

H(D) = 1 − D2 ,

¢éÍÁÙÅàÍÒµì¾ØµªèͧÊÑ­­Ò³

ÁÕ¤èÒà·èҡѺ¼ÅµèÒ§ÃÐËÇèÒ§¢éÍÁÙÅÍÔ¹¾Øµ 2 µÑÇ ³ àÇÅÒ

k

áÅÐ

k−2

´Ñ§¹Ñ ¹ ªèͧÊÑ­­Ò³¹Õ ¨ÐÁÕ ¤Ø³ÊÁºÑµÔ ¾ÔàÈÉ·Õ ÇèÒ ÅӴѺ ¢éÍÁÙÅ ÂèÍÂàÅ¢¤Õ (odd subsequence) ¨Ðà» ¹ ÍÔÊÃШҡÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Ùè (even subsequence) ´Ñ§¹Ñ ¹ à¾× Í·Õ ¨ÐŴ˹èǤÇÒÁ¨ÓàÊé¹·Ò§ (path memory) ¢Í§Ç§¨ÃµÃǨËÒÇÕà·ÍÃìºÔ ¨Ó¹Ç¹¢Í§ºÔµ 0 ·Õ àÃÕ§µÔ´µèÍ ¡Ñ¹ ¢Í§áµèÅÐÅӴѺ ¢éÍÁÙÅ ÂèÍ¨РµéͧÁÕä´éäèÁèà¡Ô¹

I

µÑÇ áÅÐà¾× ͪèÇ·ÓãËéÃкºä·ÁÁÔ §ÃԤѿàÇÍÐÃÕÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾

áÅШӹǹ¢Í§ºÔµ 0 ·Õ àÃÕ§µÔ´µè͡ѹã¹ÅӴѺ¢éÍÁÙÅÃÇÁ (ÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Õ ÁÒÃÇÁ¡ÑºÅӴѺ¢éÍÁÙÅ ÂèÍÂàÅ¢¤Ùè) ¨ÐµéͧÁÕä´éäèÁèà¡Ô¹ 0 ã¹ÃËÑÊ

(0, G/I)

G

µÑÇ àËÁ×͹¡Ñº¾ÒÃÒÁÔàµÍÃì

k

ã¹ÃËÑÊ RLL Ẻ

(d, k)

ÊÓËÃѺ¤èÒ

¹Ñ ¹¨ÐËÁÒ¶֧ ÃкºÍ¹Ø­ÒµãËéÅӴѺ¢éÍÁÙÅÃÇÁÊÒÁÒöÁÕºÔµ 1 àÃÕ§µÔ´µè͡ѹ

ä´é àËÁ×͹¡Ñº¾ÒÃÒÁÔàµÍÃì

d

ã¹ÃËÑÊ RLL Ẻ

(d, k)

[9, 59, 60] à¾ÃÒÐÇèÒ Ç§¨ÃµÃǨËÒ PRML

ÁÕ¤ÇÒÁÊÒÁÒö㹨Ѵ¡ÒáѺ ISI ·Õ à¡Ô´¢Ö ¹ä´é ¶éÒ¡Ó˹´ãËé

γ = {γ1 , γ2 , . . . , γn } à» ¹ÅӴѺ¢éÍÁÙÅẺ亹ÒÃÕ·Õ ÁÕ¤ÇÒÁÂÒÇ n ºÔµ à¾ÃÒÐ©Ð¹Ñ ¹

ÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Õ

γo

áÅÐÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Ùè

γe

¨Ð¹ÔÂÒÁâ´Â [60]

γ o = {γ1 , γ3 , γ5 , . . . , γ2dn/2e−1 } γ e = {γ2 , γ4 , γ6 , . . . , γ2bn/2c } àÁ× Í

bxc

ÂèÍÂàÅ¢¤Õ

á·¹¨Ó¹Ç¹àµçÁºÇ¡·Õ ÁÒ¡·Õ ÊØ´ ·Õ ÁÕ¤èÒ¹éÍ¡ÇèÒËÃ×Íà·èҡѺ¤èÒ

γo

áÅÐÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Ùè

γe

¨ÐÁÕ¤ÇÒÁÂÒÇà·èҡѺ

dn/2e

x

´Ñ§¹Ñ ¹ ¨Ðä´éÇèÒ ÅӴѺ¢éÍÁÙÅ

áÅÐ

bn/2c

µÒÁÅӴѺ ÅӴѺ


8.9.

ÊÃØ»·éÒº·

¢éÍÁÙÅ

γ

ÃÇÁ

γ

¨Ð¶Ù¡àÃÕ¡ÇèÒà» ¹ÅӴѺ¢éÍÁÙÅ·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

ÁÕºÔµ 0 àÃÕ§µÔ´µè͡ѹä´éäÁèà¡Ô¹

¡Ñ¹ä´éäÁèà¡Ô¹

I

µÑÇÍÂèÒ§·Õ 8.5 RLL Ẻ

ÇÔ¸Õ·Ó

167

µÑÇ àÁ× Í

G

áÅÐ

¡Ó˹´ãËé

(0, G/I)

I

G

γ

γ = {110100101000110011} G

áÅÐ

·Õ ¡Ó˹´ãËéÁÒ ¨Ðä´éÇèÒ

I

¢Í§

G=3

áÅÐ

áÅÐ

γe

ÁÕºÔµ 0 àÃÕ§µÔ´µèÍ

I=4

¤×Í ÅӴѺ ¢éÍÁÙÅ ·Õ ¶Ù¡ à¢éÒ ÃËÑÊ ´éÇÂÃËÑÊ

γ

γ o = {100110101}

à¾ÃÒÐ©Ð¹Ñ ¹ ¨Ó¹Ç¹ºÔµ 0 ·Õ àÃÕ§µÔ´¡Ñ¹ ÁÒ¡·Õ ÊØ´ ¢Í§ ´Ñ§¹Ñ ¹ ¨Ð¾ºÇèÒ

γo

¡çµèÍàÁ× Í ÅӴѺ¢éÍÁÙÅ

à» ¹àÅ¢¨Ó¹Ç¹àµçÁºÇ¡

¨§ËÒ¤èÒ¾ÒÃÒÁÔàµÍÃì

¨Ò¡ÅӴѺ¢éÍÁÙÅ

µÑÇ áÅÐÅӴѺ¢éÍÁÙÅÂèÍÂ

(0, G/I)

γ, γo,

¹Ñ ¹¤×Í ÅӴѺ¢éÍÁÙÅ

γ

áÅÐ

γe

áÅÐ

γ e = {110000101}

¤×Í 3, 2, áÅÐ 4 µÒÁÅӴѺ

¶Ù¡à¢éÒÃËÑÊ´éÇÂÃËÑÊ RLL Ẻ

¨Ò¡µÑÇÍÂèÒ§·Õ 8.5 ¨ÐÊѧࡵàËç¹ä´éÇèÒ ÅӴѺ ¢éÍÁÙÅ ·Õ ¶Ù¡ à¢éÒ ÃËÑÊ ´éÇÂÃËÑÊ RLL Ẻ ·Õ ¶Ù¡µéͧ¹Ñ ¹ ¤èÒ ¾ÒÃÒÁÔàµÍÃì

G ≤ 2I

ÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ì àªè¹ ÃËÑÊ

8.9

àÊÁÍ µÑÇÍÂèÒ§ÃËÑÊ

8/9 (0, 4/4)

(0, G/I)

áÅÐÃËÑÊ

(0, 3/4)

(0, G/I)

·Õ ãªé ã¹Ãкº¡ÒûÃÐÁÇżÅ

16/17 (0, 6/6)

à» ¹µé¹

ÊÃØ»·éÒº·

â´Â·Ñ Çä» ÃËÑÊ RLL ÁÕ ËÅÒÂÃٻẺ 㹺·¹Õ ä´é ͸ԺÒ¶֧ ËÅÑ¡¡Ò÷ӧҹáÅÐ¢Ñ ¹µÍ¹¡ÒÃÍ͡Ẻ ÃËÑÊ RLL (run length limited) µÒÁà§× ͹䢺ѧ¤Ñº

(d, k)

â´Â·Õ ¾ÒÃÒÁÔàµÍÃì

d

¨Ðà» ¹ µÑÇ¡Ó˹´

¨Ó¹Ç¹·Õ ¹éÍÂ·Õ ÊØ´ ¢Í§ºÔµ 0 ·Õ ÍÂÙè ÃÐËÇèÒ§ºÔµ 1 (µÒÁÃٻẺ NRZI) áÅоÒÃÒÁÔàµÍÃì

k

¨Ðà» ¹

µÑÇ¡Ó˹´ÁÒ¡·Õ ÊØ´¢Í§ºÔµ 0 ·Õ ÍÂÙèÃÐËÇèÒ§ºÔµ 1 à¹× ͧ¨Ò¡ ºÔµ 1 á·¹¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð¢Í§¡ÃÐáÊ ä¿¿ Ò à¢Õ¹ ´Ñ§¹Ñ ¹ ¾ÒÃÒÁÔàµÍÃì

d

¨Ö§ ªèÇÂÅ´¼Å¡Ãзº·Õ à¡Ô´ ¨Ò¡ ISI ã¹¢³Ð·Õ ¾ÒÃÒÁÔàµÍÃì

k

¨Ð

ÃѺ»ÃСѹÇèÒ ÅӴѺ¢éÍÁÙÅ·Õ à¢Õ¹ŧã¹Ê× ÍºÑ¹·Ö¡¨ÐÁÕ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ðà» ¹ÃÐÂÐæ à¾× ͪèÇ·ÓãËéÃкº ä·ÁÁÔ §ÃԤѿàÇÍÐÃÕÊÒÁÒö·Ó§Ò¹ä´éÍÂèÒ§ÁÕ»ÃÐÊÔ·¸ÔÀÒ¾ ÍÂèÒ§äáçµÒÁ ã¹» ¨¨ØºÑ¹¹Õ ÃËÑÊ RLL ·Õ ¹ÔÂÁ ãªéÁÑ¡¨ÐÍÂÙèã¹ÃÙ»¢Í§ÃËÑÊ ´ÔÊ¡ìä´Ã¿ì

(0, G/I)

«Ö §à» ¹ÃËÑÊ·Õ ¶Ù¡Í͡ẺÁÒãËéãªé§Ò¹¡ÑºÃкº PRML ¢Í§ÎÒÃì´


168

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

㹡ÒÃàÅ×Í¡ÃËÑÊ RLL ÁÒãªé§Ò¹¨Ðµéͧ¤Ó¹Ö§¶Ö§» ¨¨ÑµèÒ§æ ä´éá¡è ¾ÒÃÒÁÔàµÍÃì

R,

¤ÇÒÁ¨Ø

C(d, k),

»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

η,

(d, k),

ÍѵÃÒÃËÑÊ

áÅÐ ÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ DR à» ¹µé¹ à¾× ÍãËéä´éÃËÑÊ

RLL ·Õ ´ÕÊØ´ÊÓËÃѺ§Ò¹»ÃÐÂØ¡µì¹Ñ ¹æ ¢éÍÊѧࡵ·Õ ¾º¢Í§ÃËÑÊ RLL ·Õ ¹ÓÁÒãªé§Ò¹ã¹ÎÒÃì´´ÔÊ¡ìä´Ã¿ì ã¹» ¨¨ØºÑ¹ ¤×Í ¾ÒÃÒÁÔàµÍÃì

d

¨Ð¤èÍÂæ ŴŧÁÒà» ¹¤èÒ 0 áÅÐÍѵÃÒÃËÑÊ

R

¢Í§ÃËÑÊ RLL ·Õ ãªé ¡çÁÕ

¤èÒà¢éÒã¡Åé¤èÒ 1 ÁÒ¡¢Ö ¹àÃ× ÍÂæ ·Ñ §¹Õ à¾× Íà» ¹¡ÒÃÅ´¨Ó¹Ç¹ºÔµÊèǹà¡Ô¹·Õ à¡Ô´¢Ö ¹ã¹Ãкº ·ÓãËéÊÒÁÒö ¨Ñ´à¡çº¢éÍÁÙÅ¢èÒÇÊÒâͧ¼Ùéãªéä´éÁÒ¡¢Ö ¹

8.10

à຺½ ¡ËÑ´·éÒº·

1. ¨§¤Ó¹Ç³ËҨӹǹÅӴѺ¢éÍÁÙÅ·Ñ §ËÁ´·Õ ÁÕ¤ÇÒÁÂÒÇ·Ñ §ÊÔ ¹

(d, k)

L ºÔµ ·Õ ÊÍ´¤Åéͧ¡Ñºà§× ͹䢺ѧ¤Ñº

àÁ× Í

1.1)

d = 0, k = 2,

áÅÐ

L=5

1.2)

d = 1, k = 3,

áÅÐ

L=6

1.3)

d = 1, k = 7,

áÅÐ

L = 10

1.4)

d = 2, k = 7,

áÅÐ

L = 10

1.5)

d = 2, k = ∞,

áÅÐ

L = 10

2. ¨§áÊ´§á¼¹ÀÒ¾à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´ áÅФӹdzËÒàÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ¢Í§ÃËÑÊ RLL µÒÁà§× ͹䢺ѧ¤Ñº¢Í§¾ÒÃÒÁÔàµÍÃì

2.1)

d=1

áÅÐ

k=5

2.2)

d=1

áÅÐ

k=7

2.3)

d=2

áÅÐ

k=5

2.4)

d=2

áÅÐ

k=7

(d, k)

3. ¨Ò¡â¨·Âìã¹¢éÍ·Õ 2 ¨§¤Ó¹Ç³ËÒ¤ÇÒÁ¨Ø ¼ÅÅѾ¸ì·Õ äè´é¡Ñº

C(d, k)

àÁ× Í

C(d, k)

â´ÂãªéÊÁ¡Òà (8.12) ¾ÃéÍÁ·Ñ §à»ÃÕºà·Õº

·Õ ¤Ó¹Ç³ä´é¨Ò¡ÊÁ¡Òà (8.18)


8.10.

à຺½ ¡ËÑ´·éÒº·

169

4. ¨§ËҨӹǹÅӴѺ ¢éÍÁÙÅ ·Õ à» ¹ ä»ä´é ·Ñ §ËÁ´·Õ ÁÕ ¤ÇÒÁÂÒÇ ÊÔ ¹ÊØ´·Õ ʶҹÐ

Sj

d = 0, k = 2,

áÅÐ

L=5

4.2)

d = 1, k = 3,

áÅÐ

L=5

4.3)

d = 1, k = 7,

áÅÐ

L = 10

4.4)

d = 2, k = 7,

áÅÐ

L = 10

5. ¡Ó˹´ãËéÅӴѺ¢éÍÁÙÅ·Õ ¶Ù¡à¢éÒÃËÑÊ´éÇÂÃËÑÊ RLL Ẻ ¢Í§

γ

ºÔµ ·Õ ÍÍ¡¨Ò¡Ê¶Ò¹Ð

¢Í§ÃËÑÊ RLL µÒÁà§× ͹䢺ѧ¤Ñº¢Í§¾ÒÃÒÁÔàµÍÃì

4.1)

I

L

(0, G/I)

(d, k)

Si

áÅéÇ ä»

àÁ× Í

¨§ËÒ¤èÒ¾ÒÃÒÁÔàµÍÃì

G

áÅÐ

´Ñ§µèÍ仹Õ

5.1)

γ = {10101000101001100101}

5.2)

γ = {10010100001101001100111101}

5.3)

γ = {11100011001010001001100101}

6. ¨§ÊÃéÒ§µÒÃÒ§¤é¹ËÒÊÓËÃѺ ¡ÒÃà¢éÒ áÅжʹÃËÑÊ RLL Ẻ ¼èÒ¹¡ÒÃà¢éÒÃËÑÊáÅéǨÐÁÕ¤ÇÒÁÂÒÇà·èҡѺ

L=2

L=3

L=4

(1, 2)

â´Â·Õ ¢éÍÁÙÅ áµèÅФÃÑ § ·Õ

¾ÃéÍÁ·Ñ §ËÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

8. ¨§ÊÃéÒ§µÒÃÒ§¤é¹ËÒÊÓËÃѺ ¡ÒÃà¢éÒ áÅжʹÃËÑÊ RLL Ẻ ¼èÒ¹¡ÒÃà¢éÒÃËÑÊáÅéǨÐÁÕ¤ÇÒÁÂÒÇà·èҡѺ

â´Â·Õ ¢éÍÁÙÅ áµèÅФÃÑ § ·Õ

¾ÃéÍÁ·Ñ §ËÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

7. ¨§ÊÃéÒ§µÒÃÒ§¤é¹ËÒÊÓËÃѺ ¡ÒÃà¢éÒ áÅжʹÃËÑÊ RLL Ẻ ¼èÒ¹¡ÒÃà¢éÒÃËÑÊáÅéǨÐÁÕ¤ÇÒÁÂÒÇà·èҡѺ

(0, 1)

(1, 3)

â´Â·Õ ¢éÍÁÙÅ áµèÅФÃÑ § ·Õ

¾ÃéÍÁ·Ñ §ËÒ»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ


170

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ


ÀÒ¤¼¹Ç¡ ¡

µÒÃÒ§¿ §¡ìªÑ¹

¿ §¡ìªÑ¹

Q(x)

Q

à» ¹¿ §¡ìªÑ¹·Õ ÊÒÁÒö¨Ñ´ãËéÍÂÙèã¹ÃÙ»¢Í§¿ §¡ìªÑ¹¡ÒÃᨡᨧÊÐÊÁ¢Í§µÑÇá»ÃÊØèÁẺ

à¡ÒÊìà«Õ¹ä´é«Ö §à» ¹·Õ ¹ÔÂÁãªé§Ò¹·Ò§´éҹʶԵÔÈÒʵÃìáÅдéÒ¹ÇÔÈÇ¡ÃÃÁÈÒʵÃì â´Â¿ §¡ìªÑ¹ ¹ÔÂÒÁ´Ñ§¹Õ

1 Q(x) = √ 2π

Z

x

Q(x)

½ 2¾ y dy exp − 2

(¡.1)

«Ö §à» ¹¡ÒÃËÒ¤èÒ»ÃԾѹ¸ìÊèǹËÒ§¢Í§¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹àÁ× Í ¤×Í ¿ §¡ìªÑ¹àÅ¢ªÕ ¡ÓÅѧ (exponential function) â´Â·Ñ Çä» ¤èҢͧ¿ §¡ìªÑ¹ ÊÒÁÒöËÒä´é¨Ò¡µÒÃÒ§¤é¹ËÒ (look up table) áµèã¹¡Ã³Õ·Õ ¤èÒä´é´Ñ§¹Õ

xÀ3

½ 2¾ 1 x Q(x) ≈ √ exp − 2 x 2π

µÒÃÒ§µèÍä»¹Õ ¨ÐáÊ´§¤èҢͧ¿ §¡ìªÑ¹

Q(x)

ÊÓËÃѺ

0 ≤ x ≤ 3.59

171

¨Ð

Q(x)

¿ §¡ìªÑ¹

ÊÓËÃѺ¤èÒ

Q(x)

exp{·}

x

µèÒ§æ

ÊÒÁÒö»ÃÐÁÒ³

(¡.2)


172

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

x

Q(x)

0

0.50000

0.36

x

Q(x) 0.35942

0.72

x

Q(x) 0.23576

1.08

x

Q(x) 0.14007

1.44

x

Q(x) 0.0749340

0.01

0.49601

0.37

0.35569

0.73

0.23270

1.09

0.13786

1.45

0.0735290

0.02

0.49202

0.38

0.35197

0.74

0.22965

1.10

0.13567

1.46

0.0721450

0.03

0.48803

0.39

0.34827

0.75

0.22663

1.11

0.133500

1.47

0.070781

0.04

0.48405

0.40

0.34458

0.76

0.22363

1.12

0.131360

1.48

0.069437

0.05

0.48006

0.41

0.34090

0.77

0.22065

1.13

0.129240

1.49

0.068112

0.06

0.47608

0.42

0.33724

0.78

0.21770

1.14

0.127140

1.50

0.066807

0.07

0.47210

0.43

0.33360

0.79

0.21476

1.15

0.125070

1.51

0.065522

0.08

0.46812

0.44

0.32997

0.80

0.21186

1.16

0.123020

1.52

0.064255

0.09

0.46414

0.45

0.32636

0.81

0.20897

1.17

0.121000

1.53

0.063008

0.10

0.46017

0.46

0.32276

0.82

0.20611

1.18

0.119000

1.54

0.061780

0.11

0.45620

0.47

0.31918

0.83

0.20327

1.19

0.117020

1.55

0.060571

0.12

0.45224

0.48

0.31561

0.84

0.20045

1.20

0.115070

1.56

0.059380

0.13

0.44828

0.49

0.31207

0.85

0.19766

1.21

0.113140

1.57

0.058208

0.14

0.44433

0.50

0.30854

0.86

0.19489

1.22

0.111230

1.58

0.057053

0.15

0.44038

0.51

0.30503

0.87

0.19215

1.23

0.109350

1.59

0.055917

0.16

0.43644

0.52

0.30153

0.88

0.18943

1.24

0.107490

1.60

0.054799

0.17

0.43251

0.53

0.29806

0.89

0.18673

1.25

0.105650

1.61

0.053699

0.18

0.42858

0.54

0.29460

0.90

0.18406

1.26

0.103830

1.62

0.052616

0.19

0.42465

0.55

0.29116

0.91

0.18141

1.27

0.102040

1.63

0.051551

0.20

0.42074

0.56

0.28774

0.92

0.17879

1.28

0.100270

1.64

0.050503

0.21

0.41683

0.57

0.28434

0.93

0.17619

1.29

0.098525

1.65

0.049471

0.22

0.41294

0.58

0.28096

0.94

0.17361

1.30

0.096800

1.66

0.048457

0.23

0.40905

0.59

0.27760

0.95

0.17106

1.31

0.095098

1.67

0.047460

0.24

0.40517

0.60

0.27425

0.96

0.16853

1.32

0.093418

1.68

0.046479

0.25

0.40129

0.61

0.27093

0.97

0.16602

1.33

0.091759

1.69

0.045514

0.26

0.39743

0.62

0.26763

0.98

0.16354

1.34

0.090123

1.70

0.044565

0.27

0.39358

0.63

0.26435

0.99

0.16109

1.35

0.088508

1.71

0.043633

0.28

0.38974

0.64

0.26109

1.00

0.15866

1.36

0.086915

1.72

0.042716

0.29

0.38591

0.65

0.25785

1.01

0.15625

1.37

0.085343

1.73

0.041815

0.30

0.38209

0.66

0.25463

1.02

0.15386

1.38

0.083793

1.74

0.040930

0.31

0.37828

0.67

0.25143

1.03

0.15151

1.39

0.082264

1.75

0.040059

0.32

0.37448

0.68

0.24825

1.04

0.14917

1.40

0.080757

1.76

0.039204

0.33

0.37070

0.69

0.24510

1.05

0.14686

1.41

0.079270

1.77

0.038364

0.34

0.36693

0.70

0.24196

1.06

0.14457

1.42

0.077804

1.78

0.037538

0.35

0.36317

0.71

0.23885

1.07

0.14231

1.43

0.076359

1.79

0.036727


173

x

Q(x)

x

Q(x)

x

Q(x)

x

Q(x)

x

Q(x)

1.80

0.035930

2.16

0.0153860

2.52

0.0058677

2.88

0.00198840

3.24

0.00059765

1.81

0.035148

2.17

0.0150030

2.53

0.0057031

2.89

0.00192620

3.25

0.00057703

1.82

0.034380

2.18

0.0146290

2.54

0.0055426

2.90

0.00186580

3.26

0.00055706

1.83

0.033625

2.19

0.0142620

2.55

0.0053861

2.91

0.00180710

3.27

0.00053774

1.84

0.032884

2.20

0.0139030

2.56

0.0052336

2.92

0.00175020

3.28

0.00051904

1.85

0.032157

2.21

0.0135530

2.57

0.0050849

2.93

0.00169480

3.29

0.00050094

1.86

0.031443

2.22

0.0132090

2.58

0.0049400

2.94

0.00164110

3.30

0.00048342

1.87

0.030742

2.23

0.0128740

2.59

0.0047988

2.95

0.00158890

3.31

0.00046648

1.88

0.030054

2.24

0.0125450

2.60

0.0046612

2.96

0.00153820

3.32

0.00045009

1.89

0.029379

2.25

0.0122240

2.61

0.0045271

2.97

0.00148900

3.33

0.00043423

1.90

0.028717

2.26

0.0119110

2.62

0.0043965

2.98

0.00144120

3.34

0.00041889

1.91

0.028067

2.27

0.0116040

2.63

0.0042692

2.99

0.00139490

3.35

0.00040406

1.92

0.027429

2.28

0.0113040

2.64

0.0041453

3.00

0.00134990

3.36

0.00038971

1.93

0.026803

2.29

0.0110110

2.65

0.0040246

3.01

0.00130620

3.37

0.00037584

1.94

0.026190

2.30

0.0107240

2.66

0.0039070

3.02

0.00126390

3.38

0.00036243

1.95

0.025588

2.31

0.0104440

2.67

0.0037926

3.03

0.00122280

3.39

0.00034946

1.96

0.024998

2.32

0.0101700

2.68

0.0036811

3.04

0.00118290

3.40

0.00033693

1.97

0.024419

2.33

0.0099031

2.69

0.0035726

3.05

0.00114420

3.41

0.00032481

1.98

0.023852

2.34

0.0096419

2.70

0.0034670

3.06

0.00110670

3.42

0.00031311

1.99

0.023295

2.35

0.0093867

2.71

0.0033642

3.07

0.00107030

3.43

0.00030179

2.00

0.022750

2.36

0.0091375

2.72

0.0032641

3.08

0.00103500

3.44

0.00029086

2.01

0.022216

2.37

0.0088940

2.73

0.0031667

3.09

0.00100080

3.45

0.00028029

2.02

0.021692

2.38

0.0086563

2.74

0.0030720

3.10

0.00096760

3.46

0.00027009

2.03

0.021178

2.39

0.0084242

2.75

0.0029798

3.11

0.00093544

3.47

0.00026023

2.04

0.020675

2.40

0.0081975

2.76

0.0028901

3.12

0.00090426

3.48

0.00025071

2.05

0.020182

2.41

0.0079763

2.77

0.0028028

3.13

0.00087403

3.49

0.00024151

2.06

0.019699

2.42

0.0077603

2.78

0.0027179

3.14

0.00084474

3.50

0.00023263

2.07

0.019226

2.43

0.0075494

2.79

0.0026354

3.15

0.00081635

3.51

0.00022405

2.08

0.018763

2.44

0.0073436

2.80

0.0025551

3.16

0.00078885

3.52

0.00021577

2.09

0.018309

2.45

0.0071428

2.81

0.0024771

3.17

0.00076219

3.53

0.00020778

2.10

0.017864

2.46

0.0069469

2.82

0.0024012

3.18

0.00073638

3.54

0.00020006

2.11

0.017429

2.47

0.0067557

2.83

0.0023274

3.19

0.00071136

3.55

0.00019262

2.12

0.017003

2.48

0.0065691

2.84

0.0022557

3.20

0.00068714

3.56

0.00018543

2.13

0.016586

2.49

0.0063872

2.85

0.0021860

3.21

0.00066367

3.57

0.00017849

2.14

0.016177

2.50

0.0062097

2.86

0.0021182

3.22

0.00064095

3.58

0.00017180

2.15

0.015778

2.51

0.0060366

2.87

0.0020524

3.23

0.00061895

3.59

0.00016534


174

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ


ÀÒ¤¼¹Ç¡ ¢

Êٵä³ÔµÈÒʵÃì·Õ ÊӤѭ

ÀÒ¤¼¹Ç¡¹Õ ¨ÐáÊ´§Êٵä³ÔµÈÒʵÃì ·Õ ãªéºèÍÂÊÓËÃѺ ¡ÒÃÇÔà¤ÃÒÐËì Ãкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ ÎÒÃì´´ÔÊ¡ìä´Ã¿ì¢Í§Ë¹Ñ§Ê×ÍàÅèÁ¹Õ

¢.1

µÃÕ⡳ÁÔµÔ (Trigonometric)

sin(−α) = − sin(α) cos(−α) = cos(α) sin(α) = cos(α − π/2) sin2 (α) + cos2 (α) = 1 sin(α ± β) = sin(α) cos(β) ± cos(α) sin(β) cos(α ± β) = cos(α) cos(β) ∓ sin(α) sin(β) sin(α) sin(β) =

1 2

cos(α − β) − 12 cos(α + β)

sin(α) cos(β) =

1 2

sin(α + β) + 12 sin(α − β)

cos(α) cos(β) =

1 2

cos(α − β) + 21 cos(α + β)

cos(α) sin(β) =

1 2

sin(α + β) − 12 sin(α − β) 175


176

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

sin(2α) = 2 sin(α) cos(α) cos(2α) = cos2 (α) − sin2 (α) = 1 − 2 sin2 (α) = 2 cos2 (α) − 1 sin2 (α) = 21 {1 − cos(2α)} cos2 (α) = 21 {1 + cos(2α)} ejα = cos(α) + j sin(α) sin(α) = (ejα − e−jα )/(2j) cos(α) = (ejα + e−jα )/2

¢.2

R R R R R R R R R R R R R

»ÃԾѹ¸ìäÁè¨Ó¡Ñ´à¢µ (Inde nite Integral)

u dv = uv −

R

v du

àÁ× Í

xn dx = xn+1 /(n + 1)

u

áÅÐ

àÁ× Í

v

à» ¹¿ §¡ìªÑ¹¢Í§

n 6= −1

x−1 dx = ln(x) eax dx = eax /a ln(x) dx = x ln(x) − x xeax dx = eax (ax − 1)/a2 x2 eax dx = eax (a2 x2 − 2ax + 2)/a3 sin(ax) dx = −(1/a) cos(ax) cos(ax) dx = (1/a) sin(ax) sin2 (ax) dx = x/2 − sin(2ax)/4a x sin(ax) dx = (1/a2 ){sin(ax) − ax cos(ax)} cos2 (ax) dx = x/2 + sin(2ax)/4a x cos(ax) dx = (1/a2 ){cos(ax) + ax sin(ax)}

x


ÀÒ¤¼¹Ç¡ ¤

¡ÒÃËÒ͹ؾѹ¸ì¢Í§àÇ¡àµÍÃìààÅÐàÁ·ÃÔ¡«ì

ÀÒ¤¼¹Ç¡¹Õ ¨ÐÊÃØ»ÊٵáÒÃËÒ͹ؾѹ¸ì (di erentiation) ¢Í§àÇ¡àµÍÃìáÅÐàÁ·ÃÔ¡«ì«Ö §¨Ðà» ¹»ÃÐâª¹ì µèÍ¡ÒÃÇÔà¤ÃÒÐËìÃкº¡ÒûÃÐÁÇżÅÊÑ­­Ò³¢Í§ÎÒÃì´´ÔÊ¡ìä´Ã¿ìã¹Ë¹Ñ§Ê×ÍàÅèÁ¹Õ ¡Ó˹´ãËé

k

u

áÅÐ

x

à» ¹àÇ¡àµÍÃìá¹ÇµÑ § (column vector) ¢¹Ò´

á¶Ç áÅÐ 1 á¹ÇµÑ §) áÅÐãËé

A

k×1

à» ¹àÁ·ÃÔ¡«ì¨ÑµØÃÑÊ (square matrix) ¢¹Ò´

(¹Ñ ¹¤×ÍÁըӹǹ

k×k

´Ñ§¹Ñ ¹

¨Ò¡¡ÒÃËÒ͹ؾѹ¸ì¢Í§àÇ¡àµÍÃìààÅÐàÁ·ÃÔ¡«ì¨Ðä´é¤ÇÒÁÊÑÁ¾Ñ¹¸ì´Ñ§¹Õ

xT u = uT x ¡ ¢ ∂ xT u =x ∂u ∂ (Au) = AT ∂u ∂ (Au) =A ∂uT áÅÐ

¶éÒàÁ·ÃÔ¡«ì

(¤.1)

(¤.2)

(¤.3)

(¤.4)

¡ ¢ ¡ ¢ ∂ uT Au = A + AT u ∂u A

à» ¹àÁ·ÃÔ¡«ì·Õ ÊÁÁҵà (symmetric) ¹Ñ ¹¤×Í

A = AT

¡ ¢ ∂ uT Au = 2Au ∂u 177

(¤.5) ¨Ðä´éÇèÒ

(¤.6)


178

áÅÐ

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡ ¢ ∂ 2 uT Au = 2A ∂u ∂uT

(¤.7)


ÀÒ¤¼¹Ç¡ §

¤ÓÈѾ·ìà·¤¹Ô¤

¡Ãкǹ¡ÒÃà¢Õ¹

write process minimization process

¡Ãкǹ¡Ò÷ÓãËéÁÕ¤èÒ¹éÍÂÊØ´

¡Ãкǹ¡ÒÃ㹡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹

noise prediction process

¡Ãкǹ¡ÒÃ㹡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ ¡Ãкǹ¡ÒÃÍèÒ¹

read process write current

¡ÃÐáÊä¿¿ Òà¢Õ¹ ¡ÅØèÁ¢éÍÁÙÅ

data packet

¡ÒáŠÓÃËÑʾÑÅÊì

PCM (pulse code modulation) modulation

¡ÒáŠÓÊÑ­­Ò³ (¡ÒÃÁÍ´Ùàŵ) ¡ÒâÂÒÂÊÑ­­Ò³Ãº¡Ç¹

noise enhancement

synchronization

¡ÒÃà¢éҨѧËÇÐ

perfect synchronization

¡ÒÃà¢éҨѧËÇÐÍÂèÒ§ÊÁºÙÃ³ì ¡ÒäҴËÁÒÂ, ¤èÒ¤Ò´ËÁÒ ¡ÒèѴà¡çº¢éÍÁÙÅ

sampling

¡ÒÃà´Ô¹áººÊØèÁ

random walk

acquisition

¡ÒõÃǨËÒÅӴѺ ¡ÒõԴµÒÁ

digital data storage

projection

¡ÒêѡµÑÇÍÂèÒ§

¡ÒÃä´éÁÒ

expectation

data storage

¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ ¡ÒéÒÂ

noise whitening process

sequence detection

tracking

179


180

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¡Ò÷ӹÒÂÊÑ­­Ò³Ãº¡Ç¹ ¡ÒÃá·Ã¡ÊÍ´

noise prediction

interference

¡ÒÃá·Ã¡ÊÍ´ÃÐËÇèÒ§ÊÑ­Åѡɳì

ISI (intersymbol interference)

¡Òúǡ ¡ÒÃà»ÃÕºà·Õº ¡ÒÃàÅ×Í¡ ¡Òúѹ·Ö¡

ACS (add compare select)

recording

¡Òúѹ·Ö¡áººá¹ÇµÑ §

perpendicular recording

¡Òúѹ·Ö¡áººá¹Ç¹Í¹

longitudinal recording binary recording

¡Òúѹ·Ö¡áººäº¹ÒÃÕ ¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡

magnetic recording

¡ÒûÃÐÁÇżÅẺà¾Íà«ÍÃìäÇàÇÍÃì

¡ÒûÃѺ¤èÒ·Ò§àÇÅÒ ¡Òû ͹¡ÅѺ

timing adjustment

feedback

¡Òû ͹¡ÅѺ¤èҵѴÊԹ㨠¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

decision feedback

transition

¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È

Z

¡ÒÃá»Å§«Õ

isolated transition

transform

¡ÒÃá»Å§¿ÙàÃÕÂÃì

Fourier transform

¡ÒÃá»Å§¿ÙàÃÕÂÃì·Õ µèÍà¹× ͧ·Ò§àÇÅÒ ¡ÒÃá¾Ãè¡ÃШÒ¢ͧ¢éͼԴ¾ÅÒ´ ¡ÒÃú¡Ç¹

PSP (per survivor processing)

digital signal processing

¡ÒûÃÐÁÇżÅÊÑ­­Ò³´Ô¨Ô·ÑÅ

continuous time Fourier transform error propagation

disturbance

¡ÒÃźÅéÒ§ÊÀÒ¾áÁèàËÅç¡

demagnetization

¡ÒÃàÅ× Í¹µÓá˹觢ͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð

transition shift

¡ÒÃàÅ× Í¹µÓá˹觢ͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐẺÊØèÁ ¡ÒÃ˹èǧàÇÅÒ¡ÒõѴÊԹ㨠¡ÒÃËÒ͹ؾѹ¸ì ¡ÓÅѧ

decision delay

di erentiation

power

à¡ÒÊìà«Õ¹ ¢¹Ò´

random transition shift

Gaussian magnitude

¢éͼԴ¾ÅÒ´

error prediction error

¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ ¢éͼԴ¾ÅÒ´¡ÓÅѧÊͧà©ÅÕ Â

MSE (mean squared error)

¢éͼԴ¾ÅÒ´¡ÓÅѧÊͧà©ÅÕ Â·Õ ¹éÍÂÊØ´ ¢éͼԴ¾ÅÒ´·Ò§¤ÇÒÁ¶Õ ¢éͼԴ¾ÅÒ´·Ò§à¿Ê

MMSE (minimum mean squared error)

frequency error phase error


181

¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ

timing error

¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ·Õ ËŧàËÅ×ÍÍÂÙè ¢éÍÁÙÅ亹ÒÃÕ

binary data

¢éÍÁÙÅÇÔÂص (á«Áà» Å) ¤ÇèÐà» ¹ÁÒ¡ÊØ´ ¤ÇÒÁ¨Ø

residual timing error

discrete data ML (maximum likelihood)

capacity

¤ÇÒÁ¨ØªèͧÊÑ­­Ò³

channel capacity

complexity

¤ÇÒÁ«Ñº«é͹

¤ÇÒÁ¶Õ ¡ÒêѡµÑÇÍÂèÒ§ ¤ÇÒÁ¶Õ µÑ´

sampling rate

cut o frequency normalized frequency

¤ÇÒÁ¶Õ Ẻ¹ÍÃìÁÍÅäÅ«ì ¤ÇÒÁ¶Õ 乤ÇÔµÊì

Nyquist frequency

¤ÇÒÁ¹èÒ¨Ðà» ¹

probability

¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´

probability of error

¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´ÅӴѺ ¤ÇÒÁá»Ã»Ãǹ

probability of sequence error

variance

¤ÇÒÁäÁèà» ¹àªÔ§àÊé¹

nonlinearity

¤ÇÒÁ˹Òá¹è¹¡ÒúÃèØ

packing density

¤ÇÒÁ˹Òá¹è¹¢Í§¡Òúѹ·Ö¡áºº¹ÍÃìÁÍÅäÅ«ì ¤ÇÒÁ˹Òá¹è¹àªÔ§¾× ¹·Õ

areal density

¤ÇÒÁ˹Òá¹è¹Ê໡µÃÑÁ¡ÓÅѧ ¤Í¹âÇÅ٪ѹ

power spectral density

convolution

¤èÒ¡ÓÅѧÊͧà©ÅÕ Â ¤èÒà©ÅÕ Â

ND (normalized recording density)

mean square

mean

¤èҵѴÊÔ¹ã¨

decision

¤èҵѴÊԹ㨢³ÐË¹Ö §áººá¢ç§ ¤Òº (àÇÅÒ)

instantaneous hard decision

period

¤Òº¡ÒêѡµÑÇÍÂèÒ§ ¤ÒºàÇÅҢͧºÔµ ¤èÒàÁµÃÔ¡ÊÒ¢Ò

sampling period bit period branch metric

¤èÒàÁµÃÔ¡àÊé¹·Ò§

path metric

¤èÒÅѡɳÐ੾ÒÐ

eigenvalue

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´

FSM ( nite state machine)

à¤Ã× Í§ËÁÒÂà¢éҨѧËÇÐ §Ò¹»ÃÐÂØ¡µì (á;ÅÔपѹ)

sync mark application


182

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

à§× ͹䢺ѧ¤Ñº

constraint

à§× ͹䢺ѧ¤ÑºáººâÁ¹Ô¡ ¨Ò¹ (¨Ò¹ºÑ¹·Ö¡)

disk

¨Ò¹ºÑ¹·Ö¡áÁèàËÅç¡ ¨ÔµàµÍÃì

monic constraint

magnetic disk

jitter

¨ÔµàµÍÃì·Ò§àÇÅÒ

timing jitter

node

¨Ø´µèÍ ¨Ø´ÊÁ´ØÅ

equilibrium point

¨Ø´àʶÕÂÃÀÒ¾

stable point

ªèͧÊÑ­­Ò³

channel

ªèͧÊÑ­­Ò³ÇÔÂص

discrete channel

ªèͧÊÑ­­Ò³Ê× ÍÊÒÃ

communication channel

ªèͧÊÑ­­Ò³ÍèÒ¹

read channel

ªÔ»ªèͧÊÑ­­Ò³ÍèÒ¹ àªÔ§àÊé¹

read channel chip

linear

«é͹àËÅ× ÍÁ

overlap superparamagnetic

«Ùà»ÍÃì¾ÒÃÒáÁ¡à¹µÔ¡ à«¡àµÍÃì

sector

á«Áà» Å, µÑÇÍÂèÒ§ ä«à¤ÔÅÊÅÔ» ´Ô¨Ô·ÑÅ

sample

cycle slip digital

´Õà·ÍÃìÁÔá¹¹µì à´«ÔàºÅ

determinant

dB (decibel)

â´àÁ¹

domain

â´àÁ¹

D

D

â´àÁ¹

Z

Z

â´àÁ¹¤ÇÒÁ¶Õ

domain frequency domain

â´àÁ¹àÇÅÒ

time domain

µÑǤٳÅÒ¡ÃÒ¹¨ì µÑÇªÕ ºÍ¡

domain

Lagrange multiplier

indicator

µÑÇ´Óà¹Ô¹¡Òä͹âÇÅ٪ѹ µÑÇ´Óà¹Ô¹¡ÒäèÒ¤Ò´ËÁÒ µÑÇ´Óà¹Ô¹¡ÒÃ˹èǧàÇÅÒ µÑǹÓ˹éÒ µÒÃÒ§¤é¹ËÒ

convolution operator expectation operator delay operator

predecessor look up table


183

·ÒÃìà¡çµ

target

·ÒÃìà¡çµ·Õ àËÁÒÐ·Õ ÊØ´

optimal target

·ÒÃìà¡çµáºº GPR

generalized partial response (GPR) target

·ÒÃìà¡çµáºº PR

partial response (PR) target

·ÓãËéà» ¹ºÃ÷Ѵ°Ò¹

normalize

à·¤¹Ô¤¡ÒûÃÐÁÒ³¤èÒ㹪èǧ à·ÃÐ亵ì á·ç»

interpolation technique

TB (terabyte) tap

ä·ÁÁÔ §¿ §¡ìªÑ¹

timing function

ä·ÁÁÔ §¿ §¡ìªÑ¹áºº¹ÍÃìÁÍÅäÅ«ì

normalized timing function

timing recovery

ä·ÁÁÔ §ÃԤѿàÇÍÃÕ

ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Õ ãªé¡Ñ¹·Ñ Çä» ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ¹ÔùÑÂ

conventional timing recovery

deductive timing recovery

ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ»ÃÐÁÒ³¤èÒ㹪èǧ ä·ÁÁÔ §ÃԤѿàÇÍÃÕẺÍØ»¹Ñ ä·ÁÁÔ §ÅÙ» ºÔµ

interpolated timing recovery

inductive timing recovery

timing loop

bit

ºÔµ¢èÒÇÊÒÃ

message bit

ºÔµà«ÅÅì (¤Òº¢Í§àÇÅÒã¹Ë¹Ö §ºÔµ) ºÔµà»ÅÕ Â¹Ê¶Ò¹Ð ºÔµÊèǹà¡Ô¹ ẹ´ìÇÔ´·ì

bit cell

transition bit

redundant bit bandwidth

ẹ´ìÇÔ´·ìà¡Ô¹à» ¹ÈÙ¹Âì ẹ´ìÇÔ´·ì¢Í§ÅÙ»

zero excess bandwidth

loop bandwidth

Ẻ¢éÍÁÙÅ

data pattern

Ẻ¨ÓÅͧ

model

Ẻ¨ÓÅͧ¡Ò÷ӧҹ¢Í§Ç§¨Ãà¿ÊÅçÍ¡ÅٻẺàªÔ§àÊé¹ áºº¨ÓÅͧªèͧÊÑ­­Ò³·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙŠẺ¨ÓÅͧªèͧÊÑ­­Ò³àÊÁ×͹¨ÃÔ§ Ẻ¨ÓÅͧªèͧÊÑ­­Ò³ÍØ´Á¤µÔ

亹ÒÃÕ (°Ò¹Êͧ)

ideal channel model

binary

»ÃÐÊÔ·¸Ô¼Å¢Í§ÃËÑÊ

equivalent discrete time channel model

realistic channel model

byte

亵ì (1 亵ì = 8 ºÔµ)

linearized phase locked loop (PLL) model

code e ciency

»ÃѺ¤èÒ (ãËéà» ¹» ¨¨ØºÑ¹)

update

»ÃÔÁҳ˹èǧàÇÅÒã¹ÅÙ»

loop delay


184

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

¼ÅµÍºÊ¹Í§

response

¼ÅµÍºÊ¹Í§¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ¼ÅµÍºÊ¹Í§¢Í§Ãкº

system response

¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ ¼ÅµÍºÊ¹Í§ä´ºÔµ

transition response

frequency response dibit response

¼ÅµÍºÊ¹Í§·ÒÃìà¡çµ

target response

¼ÅµÍºÊ¹Í§ºÒ§Êèǹ

PR (partial response) PRML (partial response maximum likelihood)

¼ÅµÍºÊ¹Í§ºÒ§Êèǹ¤ÇèÐà» ¹ÁÒ¡ÊØ´ ¼ÅµÍºÊ¹Í§ºÒ§ÊèǹẺ·Ñ Çä» ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì

impulse response FIR ( nite impulse response)

¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊì¨Ó¡Ñ´ ¼ÅµÍºÊ¹Í§ÍÔÁ¾ÑÅÊìäÁè¨Ó¡Ñ´ á¼è¹«Õ´Õ

IIR (in nite impulse response)

CD (compact disc)

á¼è¹´ÕÇÕ´Õ

DVD (digital versatile disc)

á¼è¹ºÑ¹·Ö¡áÁèàËÅç¡ á¼¹ÀÒ¾

magnetic oppy disk

diagram

á¼¹ÀÒ¾à·ÃÅÅÔÊ ¾Åѧ§Ò¹

GPR (generalized partial response)

trellis diagram

energy

¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇÂ

unit energy

¾ËعÒÁ (â¾ÅÔâ¹àÁÕÂÅ)

polynomial

à¾Íà«ÍÃìäÇàÇÍÃìä·ÁÁÔ §ÃԤѿàÇÍÃÕ

per survivor timing recovery

à¾Íà«ÍÃìäÇàÇÍÃìä·ÁÁÔ §ÃԤѿàÇÍÃÕẺ·Ó§Ò¹« Ó â¾Å ¿ÅÑ¡«ì ¿ §¡ìªÑ¹

per survivor iterative timing recovery

pole ux function

¿ §¡ìªÑ¹¢Ñ ¹ºÑ¹ä´

step function

¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹

probability density function

¿ §¡ìªÑ¹¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹ ¿ §¡ìªÑ¹â¤Ã๤à¡ÍÃìà´ÅµÒ ¿ §¡ìªÑ¹ä´àäà´ÅµÒ

Kronecker delta function

Dirac delta function

¿ §¡ìªÑ¹¶èÒÂâ͹

transfer function

¿ §¡ìªÑ¹ÍÔÁ¾ÑÅÊì

impulse function

ä¿¿ Ò¡ÃÐáʵç ÀÒÇÐ

d.c. (direct current)

mode

ÀÒÇСÒÃä´éÁÒ

Gaussian probability density function

acquisition mode


185

ÀÒÇСÒÃä´éÁÒẺÊÁºÙóì ÀÒÇСÒõԴµÒÁ

perfect acquisition

tracking mode

ÀÒÇСÒý ¡ÍºÃÁ

training mode

àÁµÃÔ¡ (µÑÇÇÑ´)

metric

àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð àÁ·ÃÔ¡«ìÊËÊÑÁ¾Ñ¹¸ì¢éÒÁ

cross correlation matrix

àÁ·ÃÔ¡«ìÍѵÊËÊÑÁ¾Ñ¹¸ì

auto correlation matrix identity matrix

àÁ·ÃÔ¡«ìàÍ¡Åѡɳì äÁèà¢éҨѧËÇÐ

state transition matrix

asynchronous

äÁèÁÕÊËÊÑÁ¾Ñ¹¸ì¡Ñ¹

uncorrelated ECC (error correction code)

ÃËÑÊá¡é䢢éͼԴ¾ÅÒ´ ÃËÑʤ͹âÇÅ٪ѹ ÃËÑÊÁÍ´ÙàŪѹ

convolutional code modulation code

ÃдѺ¢Õ´àÃÔ Áà»ÅÕ Â¹

threshold level

Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ

data storage system

Ãкº¡Òúѹ·Ö¡áÁèàËÅç¡ Ãкº·Õ ¶Ù¡à¢éÒÃËÑÊ

magnetic recording system

coded system

Ãкº·Õ ÁÕᶺ¤ÇÒÁ¶Õ ¨Ó¡Ñ´ Ãкº·Õ äÁèä´é¶Ù¡à¢éÒÃËÑÊ ÃкºÊ× ÍÊÒôԨԷÑÅ ÃÐÂÐ

band limited system uncoded system

digital communication system

stage

ÃÐÂзҧ¡ÓÅѧÊͧà©ÅÕ Â

MSD (mean squared distance)

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å

e ective distance

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ ÃÐÂзҧÂؤÅÔ´

Euclidean distance

ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ ÃÒ¡¡ÓÅѧÊͧà©ÅÕ Â ÃÙ»¤Å× ¹

squared e ective distance

squared Euclidean distance

RMS (root mean square)

waveform

àèÔÊàµÍÃìẺàÅ× Í¹

shift register

ÅÍ¡ÒÃÔ·ÖÁ¸ÃÃÁªÒµÔ

natural logarithm

Åͧ¼Ô´Åͧ¶Ù¡ ÅӴѺ

trial and error

sequence

ÅӴѺ¢éͼԴ¾ÅÒ´

error sequence

ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ

input error sequence

ÅӴѺ¢éͼԴ¾ÅÒ´ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ

valid input error sequence


186

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÅӴѺ¢éÍÁÙÅ

data sequence

ÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Õ

odd subsequence

ÅӴѺ¢éÍÁÙÅÂèÍÂàÅ¢¤Ùè

even subsequence

ǧ¨Ã VCO ǧ¨Ã¡Ãͧ

VCO (voltage controlled oscillator) lter interpolation lter

ǧ¨Ã¡Ãͧ¡ÒûÃÐÁÒ³¤èÒ㹪èǧ ǧ¨Ã¡Ãͧ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒ ǧ¨Ã¡Ãͧ·Ó¹ÒÂ

prediction error lter

predictor lter

ǧ¨Ã¡Ãͧ㹡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ Ç§¨Ã¡ÃͧẺàªÔ§àÊé¹

linear lter LPF (low pass lter)

ǧ¨Ã¡Ãͧ¼èÒ¹µ Ó

ǧ¨Ã¡Ãͧ¼èÒ¹µ ÓÍØ´Á¤µÔ

ideal low pass lter

loop lter

ǧ¨Ã¡ÃͧÅÙ» ǧ¨Ãà¢éÒÃËÑÊ

encoder

ǧ¨Ãà¢éÒÃËÑÊá¡é䢢éͼԴ¾ÅÒ´ ǧ¨Ãà¢éÒÃËÑÊÁÍ´ÙàŪѹ ǧ¨ÃªÑ¡µÑÇÍÂèÒ§ ǧ¨ÃµÃǨËÒ

noise whitening lter

error correction code (ECC) encoder

modulation encoder

sampler detector

ǧ¨ÃµÃǨËÒ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ

TED (timing error detector)

ǧ¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹·Õ äÁèÁÕ˹èǤÇÒÁ¨Ó

memoryless threshold detector

ǧ¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹·Õ äÁèÁÕ˹èǤÇÒÁ¨ÓẺËÅÒÂÃдѺ

multi level memoryless threshold de

tector ǧ¨ÃµÃǨËÒ¢Õ´àÃÔ Áà»ÅÕ Â¹áººËÅÒÂÃдѺ ǧ¨ÃµÃǨËҨشÊÙ§ÊØ´ ǧ¨ÃµÃǨËÒÅӴѺ

peak detector sequence detector

ǧ¨ÃµÃǨËÒÅӴѺ·Õ ¤ÇèÐà» ¹ÁÒ¡ÊØ´ ǧ¨ÃµÃǨËÒÅӴѺ·Õ àËÁÒÐ·Õ ÊØ´ ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ

MLSD (maximum likelihood sequence detector)

optimal sequence detector

Viterbi detector

ǧ¨ÃµÃǨËÒÊÑ­Åѡɳì ǧ¨ÃµÃǨËÒ·Õ àËÁÒÐ·Õ ÊØ´ ǧ¨Ã¶Í´ÃËÑÊ

multi level threshold detector

symbol detector opimal detector

decoder

ǧ¨Ã¶Í´ÃËÑÊá¡é䢢éͼԴ¾ÅÒ´ ǧ¨Ã¶Í´ÃËÑÊÁÍ´ÙàŪѹ

error correction code (ECC) decoder

modulation decoder

ǧ¨Ã¡Ãͧ·Ó¹ÒÂË¹Ö §¢Ñ ¹áººàªÔ§àÊé¹

linear one step predictor

ǧ¨Ãà»ÅÕ Â¹ÊÑ­­Ò³á͹ÐÅçÍ¡à» ¹ÊÑ­­Ò³´Ô¨Ô·ÑÅ

ADC (analog to digital converter)


187

ǧ¨Ãà¿ÊÅçÍ¡ÅÙ»

PLL (phase locked loop)

ǧ¨ÃÀÒ¤ÃѺ

receiver

ǧ¨ÃÀÒ¤Êè§

transmitter

ǧ¨ÃÁÍ´ÙàÅàµÍÃì

modulator

ǧ¨ÃËÒ͹ؾѹ¸ì

di erentiator

àÇ¡àµÍÃìÅѡɳÐ੾ÒÐ

eigenvector

àÇ¡àµÍÃìÅѡɳÐ੾ÒÐẺ¹ÍÃìÁÍÅäÅ«ì stationary

Ê൪ѹà¹ÃÕ Ê¶Ò¹Ð

normalized eigenvector

state

ʶҹеèÍä»

next state start state

ʶҹÐàÃÔ Áµé¹

ʶһ µÂ¡ÃÃÁªèͧÊÑ­­Ò³ÍèÒ¹ Ê໡µÃÑÁ¤èÒÈÙ¹Âì

read channel architecture

spectral null

ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡ (¡Ò÷ÓãËéà» ¹áÁèàËÅç¡) ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅ硢ͧÊ× ÍºÑ¹·Ö¡ ÊÀҾźÅéÒ§áÁèàËÅç¡ ÊÀÒ¾ãËé«ÖÁ¼èÒ¹ä´é

medium magnetization

coercivity permeability

ÊÁ¡ÒùÍÃìÁÍÅ

normal equation

ÊÁ¡ÒûÃѺ¤èÒ

update equation

ÊÅѺà»ÅÕ Â¹ (·ÃÒ¹Êâ¾Ê) Êèǹ»ÃСͺ

transpose

component

ÊËÊÑÁ¾Ñ¹¸ì

correlation

ÊËÊÑÁ¾Ñ¹¸ì¢éÒÁ ÊÑ­­Ò³

magnetization

cross correlation

signal

ÊÑ­­Ò³äº¹ÒÃÕ (ÊÑ­­Ò³ÊͧÃдѺ) ÊÑ­­Ò³¾ÑÅÊìä´ºÔµ

dibit pulse

ÊÑ­­Ò³¾ÑÅÊì乤ÇÔµÊìÍØ´Á¤µÔ ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð

ideal Nyquist pulse transition pulse

ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹ÐàÍ¡à·È ÊÑ­­Ò³Ãº¡Ç¹

binary signal

isolated transition pulse

noise

ÊÑ­­Ò³Ãº¡Ç¹¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇ

white Gaussian noise

ÊÑ­­Ò³Ãº¡Ç¹à¡ÒÊìÊÕ¢ÒÇẺºÇ¡ ÊÑ­­Ò³Ãº¡Ç¹¤ÇÒÁÃé͹

transition noise

AWGN (additive white Gaussian noise)

thermal noise

ÊÑ­­Ò³Ãº¡Ç¹¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡

media jitter noise


188

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

ÊÑ­­Ò³Ãº¡Ç¹·Õ ¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ

pattern dependent noise

ÊÑ­­Ò³Ãº¡Ç¹·Õ ÍÂÙè¹Í¡á¶º¤ÇÒÁ¶Õ ÊÑ­­Ò³Ãº¡Ç¹áººÊÕ

out of band noise

colored noise

ÊÑ­­Ò³Ãº¡Ç¹ÊÕ¢ÒÇ

white noise

ÊÑ­­Ò³Ãº¡Ç¹Ê× ÍºÑ¹·Ö¡

media noise

ÊÑ­­Ò³á͹ÐÅçÍ¡·Ò§ä¿¿ Ò·Õ ä´é¨Ò¡ËÑÇÍèÒ¹ (ÊÑ­­Ò³ read back) coe cient

ÊÑÁ»ÃÐÊÔ·¸Ô

Ê× ÍºÑ¹·Ö¡ (¨Ò¹áÁèàËÅç¡) àʶÕÂÃÀÒ¾

media (or medium)

stable

àÊé¹â¤é§ÃÙ»µÑÇàÍÊ

S curve

àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè àÊé¹ÊÒ¢Ò

survivor path

branch

˹èǧàÇÅÒ

delay

˹èǤÇÒÁ¨Ó¢Í§ªèͧÊÑ­­Ò³ ˹èǤÇÒÁ¨ÓàÊé¹·Ò§

path memory

ËÅÑ¡¡ÒÃàªÔ§µÑ §©Ò¡

orthogonality principle

write head

ËÑÇà¢Õ¹ ËÑÇÍèÒ¹

channel memory

read head

à˵ءÒóì

event

à˵ءÒóì¢éͼԴ¾ÅÒ´

error event

à˵ءÒóì¢éͼԴ¾ÅÒ´·Õ â´´à´è¹ áËÅ觵鹷ҧ

dominant error event

source

áËÅ觻ÅÒ·ҧ

destination

ÍͿ૵·Ò§¤ÇÒÁ¶Õ

frequency o set

ÍͿ૵·Ò§à¿Ê

phase o set

ÍͿ૵·Ò§à¿Ê¢Í§¡ÒêѡµÑÇÍÂèÒ§ ÍͿ૵·Ò§àÇÅÒ ÍѵÃÒ¡ÒâÂÒÂ

timing o set gain

ÍѵÃÒ¡ÒêѡµÑÇÍÂèÒ§

sampling rate

ÍѵÃÒ¡ÒêѡµÑÇÍÂèҧẺà¡Ô¹¨ÃÔ§ ÍѵÃÒ¡ÒÃÅÙèà¢éÒ

oversampling rate

convergence rate

ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ

BER (bit error rate)

ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ ÍѵÃÒ¤ÇÒÁ˹Òá¹è¹ ÍѵÃÒºÔµ

sampling phase o set

bit rate

asymptotic information rate

density ratio

read back signal


189

ÍѵÃÒÃËÑÊ

code rate

ÍѵÃÒÊ觢éÍÁÙÅ

data rate

ÍѵÃÒÊèǹ¤èÒ¡ÓÅѧà©ÅÕ Â¢Í§ÊÑ­­Ò³·Õ µéͧ¡ÒõèͤèÒ¡ÓÅѧà©ÅÕ Â¢Í§ÊÑ­­Ò³Ãº¡Ç¹ ratio) ÍѵÊËÊÑÁ¾Ñ¹¸ì

auto correlation

ÍÑÅ¡ÍÃÔ·ÖÁ (¢Ñ ¹µÍ¹ÇÔ¸Õ) ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ

algorithm

Viterbi algorithm inductive

ÍÔ¹´Ñ¡·Õ¿ (»ÃÐàÀ·¢Í§ËÑÇÍèÒ¹) ÍÔ¹¾Øµ (ÃѺà¢éÒ, ¹Óà¢éÒ) ÍÔ ÁµÑÇ

input

saturated

ÍÕ¤ÇÍäÅ૪ѹẺÊÁºÙóì ÍÕ¤ÇÍäÅà«ÍÃì

perfect equalization

equalizer

ÍÕ¤ÇÍäÅà«ÍÃìẺ PR

partial response (PR) equalizer

ÍÕ¤ÇÍäÅà«ÍÃìẺ¼ÅµÍºÊ¹Í§àµçÁ àÍÒµì¾Øµ (Êè§ÍÍ¡, ¹ÓÍÍ¡) àÍÒµì¾Øµ¢Í§ªèͧÊÑ­­Ò³ á͹ÐÅçÍ¡

àÎÔõ«ì

output channel output

analog

ÎÒÃì´´ÔÊ¡ìä´Ã¿ì

hard disk drive

Hz (hertz)

full response equalizer

SNR (signal to noise


190

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ


ºÃóҹءÃÁ

[1] S. X. Wang and A. M. Taratorin, Magnetic information storage technology. San Diego: Academic Press, 1999.

[2] B. Sklar, Digital communications:

fundamentals and applications:

Prentice Hall,

2nd edition, 2001.

[3] A. M. Taratorin, Magnetic recording systems and measurements:

Guzik Technical

Enterprises, 2004.

[4] J. W. M. Bergmans, Digital baseband transmission and recording. Boston/London/ Dordrecht: Kluwer Academic Publishers, 1996.

[5] T. Suzuki, Perpendicular magnetic recording: Its basics and potential for the future, IEEE Trans. on Magnetics, vol. MAG 20, no. 5, pp. 675 680, September 1984.

[6] J. Moon, The role of signal processing in data storage, IEEE Signal Processing Magazine, pp. 54 72, July 1998.

[7] S. B. Wicker, Error control systems for digital communication and storage. New Jersey: Printice Hall International, 1995.

[8] B. Vasic and E. M. Kurtas, Coding and signal processing for magnetic recording systems. New York: CRC Press, 2005. 191


[9] K. A. S Immink, Runlength limited sequences, in Proc. of the IEEE, vol. 78, no. 11, pp. 1745 1759, November 1990.

[10] » ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, ¡ÒûÃÐÁÇżÅÊÑ­­Ò³ÊÓËÃѺ¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ àÅèÁ 1: ¾× ¹°Ò¹ ªèͧÊÑ­­Ò³ÍèÒ¹ à¢Õ¹,

ÈÙ¹Âìà·¤â¹âÅÂÕÍÔàÅç¡·Ã͹ԡÊìààÅФÍÁ¾ÔÇàµÍÃìààË觪ҵÔ

(๤෤),

2550.

[11] T. A. Roscamp, E. D. Boerner, and G. J. Parker, Three dimensional modeling of perpendicular recording with soft underlayer, J. of Applied Physics, vol. 91, no. 10, May 2002.

[12] A. V. Oppenheim, A. S. Willsky, and S. H. nawab, Signals and systems. New Jersey: Prentice Hall, 2nd edition, 1997.

[13] Available online at http://www.mathworks.com

[14] » ÂÐ â¤ÇÔ¹·ì·ÇÕÇѲ¹ì, ¤ÙèÁ×Íâ»Ãá¡ÃÁÀÒÉÒ SCILAB ÊÓËÃѺ¼ÙéàÃÔ Áµé¹ (¾ÔÁ¾ì¤ÃÑ §·Õ 2), ÈÙ¹Âì¼ÅÔµ µÓÃÒàÃÕ¹, ʶҺѹ෤â¹âÅÂÕ¾ÃШÍÁà¡ÅéÒ¾Ãй¤Ãà˹×Í, 2549.

[15] G. D. Forney, Maximum likelihood sequence estimation of digital sequences in the presence of intersymbol interference, IEEE Trans. on Information Theory, vol. IT 18, no. 3, pp. 363 378, May 1972.

[16] J. R. Barry, E. A. Lee, and D. G. Messerschmitt, Digital communication. Boston: Kluwer Academic Publishers, 3nd edition, 2003.

[17] H. K. Thapar and A. M. Patel, A class of partial response systems for increasing storage density in magnetic recording, IEEE Trans. on Magnetics, vol. 23, no. 5, pp. 3666 3668, September 1987.


[18] P. Kovintavewat, I. Ozgunes, E. Kurtas, J. R. Barry, and S. W. McLaughlin, General ized partial response targets for perpendicular recording with jitter noise, IEEE Trans. on Magnetics, vol. 38, no. 5, pp. 2340 2342, September 2002.

[19] J. Moon and W. Zeng, Equalization for maximum likelihood detector, IEEE Trans. on Magnetics, vol. 31, no. 2, pp. 1083 1088, March 1995.

[20] S. Raghaven and H. K. Thapar, Feed forward timing recovery for digital magnetic recording, in Proc. of ICC 91, vol. 2, pp. 794 798, June 1991.

[21] A. N. D Andrea, U. Mengali, and G. M. Vitetta, Approximate ML decoding of coded PSK with no explicit carrier phase reference, IEEE Trans. on Communication, vol. 42, no. 2/3/4, pp. 1033 1039, Feb/Mar/Apr 1994.

[22] U. Mengali and A. N. D Andrea, Synchronization techniques for digital receivers, New York: Plenum Press, 1997.

[23] J. Moon and L. R. Carley, Performance comparison of detection methods in magnetic recording, IEEE Trans. on Magnetics, vol. 26, pp. 3155 3172, November 1990. [24] K. H. Mueller and M. Müller, Timing recovery in digital synchronous data receivers, IEEE Trans. on Communication, vol. COM 24, no. 5, pp. 516 531, May 1976.

[25] M. H. Hayes, Statistical digital signal processing and modeling. John Wiley & Sons Inc., New York, 1996.

[26] A. Leon Garcia, Probability and random processes for electrical engineering. New York: Addison Wesley Inc., 2nd edition, 1994.

[27] R. D. Cideciyan, F. Dolivo, R. Hermann, W. Hirt, and W. Schott, A PRML system for digital magnetic recording, IEEE J. Selected Areas Commun., vol. 10, no. 1, pp. 38 56, January 1992.


[28] H. Sha ee, Timing recovery for sampling detectors in digital magnetic recording, in Proc. of ICC 96, vol. 1, pp. 577 581, January 1996.

[29] A. R. Nayak, Iterative timing recovery for magnetic recording channels with low signal to noise ratio, Ph.D. dissertation, Georgia Institute of Technology, Georgia, June 2004.

[30] P. Kovintavewat, Timing recovery based on per survivor processing, Ph.D. disserta tion, Georgia Institute of Technology, Georgia, October 2004.

[31] J. R. Barry, A. Kavc̆ić, S. W. McLaughlin, and A. R. Nayak, Iterative timing recovery, IEEE Signal Processing Magazine, vol. 21, pp. 89 102, January 2004.

[32] X. Jin and A. Kavc̆ić, Cycle slip detection using soft output information, in Proc. of ICC 01, vol. 9, pp. 2706 2710, June 2001.

[33] A. R. Nayak, J. R. Barry, and S. W. McLaughlin, Joint timing recovery and turbo equalization for coded partial response channels, IEEE Trans. on Magnetics, vol. 38, no. 5, pp. 2295 2297, September 2003.

[34] F. M. Gardner, Interpolation in digital modems Part I: Fundamentals, IEEE Trans. on Communication, vol. 41, no. 3, pp. 501 507, March 1993.

[35] L. Erup, F. M. Gardner, and R. A. Harris, Interpolation in digital modems Part II: Implementation and performance, IEEE Trans. on Communication, vol. 41, no. 6, pp. 998 1008, June 1993.

[36] Z. Wu, J. M. Cio , and K. D. Fisher A MMSE interpolated timing recovery scheme for the magnetic recording channel, in Proc. of ICC 97, vol. 3, pp. 1625 1629, 1997.


[37] M. Spurbeck and R. T. Behrens, Interpolated timing recovery for hard disk drive read channels, in Proc. of ICC 97, vol. 3, pp. 1618 1624, 1997.

[38] P. Kovintavewat, J. R. Barry, F. M. Erden, and E. M. Kurtas, Per survivor timing recovery for uncoded partial response channels, in Proc. of ICC 04, Paris, France, vol. 5, pp. 2715 2719, June 20 24, 2004.

[39] P. Kovintavewat, J. R. Barry, F. M. Erden, and E. M. Kurtas, Per survivor iterative timing recovery for coded partial response channels, in Proc. of Globecom 04, Texas, USA, vol. 4, pp. 2604 2608, Nov 29 Dec 3, 2004.

[40] P. Kovintavewat, J. R. Barry, M. F. Erden, and E. M. Kurtas, Robustness of Per Survivor Iterative Timing Recovery in Perpendicular Recording Channels, IEEE In ternational Conference on Magnetics (INTERMAG 2005), Nagoya, Japan, pp. 1613 1614, April 4 8, 2005.

[41] D. G. Messerschmitt, Design of nite impulse response for the Viterbi algorithm and decision feedback equalizer, in Proc. of ICC 74, pp. 37D 1 5, June 1974.

[42] J. Fitzpatrick, J. K. Wolf, and L. Barbosa, New equalizer targets for sampled magnetic recording system, in Proc. of the 25th Asilomar Conference on Signals Systems and Computers, pp. 30 34, November 1991.

[43] T. Oenning and J. Moon, Partial response maximum likelihood detection for perpen dicular recording, IEEE International Conference on Magnetics (INTERMAG 2000), p. HT 08, 2000.

[44] I. Lee, C. Modlin, and J. M. Cio , Equalized maximum likelihood receiver in a magnetic recording channel, in Proc. of ICC 93, pp. 1970 1973, November 1993.


[45] C. T. Beare, The choice of the desired impulse response in combined linear Viterbi algorithm equalizers, IEEE Trans. on Communication, vol. COM 26, no. 8, pp. 1301 1307, August 1978.

[46] I. Lee and J. M. Cio , Equalized maximum likelihood receiver with a unit energy constraint, IEEE Trans. on Magnetics, vol. 33, no. 1, pp. 855 862, January 1997.

[47] A. Ghrayeb and W. E. Ryan, Precoder design for concatenating convolutional codes with generalized partial response channels, in Proc. of Globecom 00, San Francisco, CA, 2000, pp. 1859 1864.

[48] L. C. Barbosa, Maximum likelihood sequence estimators: a geometric view, IEEE Trans. on Information Theory, vol. 35, no. 2, pp. 419 427, March 1989.

[49] J. Caroselli and J.

K. Wolf, Error event characterization of partial response systems

in magnetic recording systems with medium noise, in Proc. of Globecom 98, vol. 5, pp. 2724 2728.

[50] B. E. Moision, P. H. Siegel, and E. Soljanin, Distance enhancing for digital recording, IEEE Trans. on Magnetics, vol. 34, no. 1, pp. 68 74, January 1998.

[51] P. R. Chevillat, E. Eleftheriou, and D. Maiwald, Noise predictive partial response equalizers and applications, in Proc. of ICC 92, vol. 2, pp. 942 947, June 1992.

[52] E. Eleftheriou and W. Hirt, Noise predictive maximum likelihood (NPML) detection for the magnetic recording channel, in Proc. of ICC 96, vol. 1, pp. 556 560, June 1996.

[53] S. Altekar, Detection and coding techniques for magnetic recording channels, Ph.D. dissertation, Univ. Calif. San Diego, June 1997.


[54] J. Moon and J. Park, “Pattern dependent noise prediction in signal dependent noise,” IEEE J. Selected Areas Comm., vol. 19, no. 4, pp. 730 – 743, June 2001.

[55] J. Caroselli, S. A. Altekar, P. McEwen, and J. K. Wolf, “Improved detection for magnetic recording systems with media noise,” IEEE Trans. on Magnetics, vol. 33, no. 5, pp. 2779 – 2781, 1997.

[56] X. Yang and E. Kurtas, “Noise predictive equalization and detection,” Internal report, Channels Department, Seagate Technology, Pittsburgh, PA, May 2003.

[57] R. Raheli, A. Polydoros, and C. K. Tzou, “Per survivor processing: a general approach to MLSE in uncertain environments,” IEEE Trans. on Commun., vol. 43, no. 2, pp. 354 – 364, Feb./Mar./Apr. 1995.

[58] C. E. Shannon, “A mathematical theory of communication,” The Bell system technical journal, vol. 27, pp. 379 – 423, 623 – 656, July, October, 1948.

[59] P. H. Siegel and J. K. Wolf, “Modulation and coding for information storage,” IEEE Communications Magazine, pp. 68 – 86, December 1991.

[60] K. A. S. Abdel Gha ar and J. H. Weber, “Constrained block codes for class IV partial response channels with maximum likelihood sequence estimation,” IEEE Trans. on Information Theory, vol. 42, no. 5, pp. 1405 – 1424, September 1996.



´Ãê¹Õ PW50 ,

6

ËÅÑ¡¡ÒÃàªÔ§µÑ §©Ò¡, 142

¡Òúѹ·Ö¡áººá¹ÇµÑ §, 6

ÍÑÅ¡ÍÃÔ·ÖÁ, 140 144

¡Òúѹ·Ö¡áººá¹Ç¹Í¹, 6

à¾Íà«ÍÃìäÇàÇÍÃì (PSP), 144 àÁµÃÔ¡ÊÒ¢Ò, 140, 141

GPRML, 133

PLL, 18, 21, 67 NPML, 115 135, 138

¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì, 23 30

¡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ, 117

Íѹ´Ñº·Õ Êͧ

¢éͼԴ¾ÅÒ´¡ÓÅѧÊͧà©ÅÕ Â, 119

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊé¹, 28 30

·Õ ¹éÍÂÊØ´, 120

ÊÁ¡ÒûÃѺ¤èÒ, 22

¤ÇÒÁ«Ñº«é͹, 126

Íѹ´Ñº·Õ Ë¹Ö §

·ÒÃìà¡çµ»ÃÐÊÔ·¸Ô¼Å, 122, 123

¡ÒÃÇÔà¤ÃÒÐËìàªÔ§àÊé¹, 23 28

»ÃÐÊÔ·¸ÔÀÒ¾, 129

ÊÁ¡ÒûÃѺ¤èÒ, 22

ǧ¨Ã¡Ãͧ·Ó¹ÒÂ, 117, 118 ËÅÑ¡¡Ò÷ӧҹ, 120 129 ËÅÑ¡¡ÒÃàªÔ§µÑ §©Ò¡, 119

àʶÕÂÃÀÒ¾, 24, 28 preamble, 22, 34 PRML, 44, 65 88, 115, 126, 137

àÁµÃÔ¡ÊÒ¢Ò, 121, 122

ÃËÑÊ

(0, G/I),

166

ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ, 69

PDNP, 137 150 ¡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ, 141

ÍÑÅ¡ÍÃÔ·ÖÁÇÕà·ÍÃìºÔ, 75

¤ÇÒÁ«Ñº«é͹, 146

ÍÕ¤ÇÍäÅà«ÍÃì, 66

»ÃÐÊÔ·¸ÔÀÒ¾, 146

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, 71

ǧ¨Ã¡Ãͧ·Ó¹ÒÂ, 141

á¼¹ÀÒ¾à·ÃÅÅÔÊ, 72 199


SCILAB, 8, 54

ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð, 6

SNR, 21, 104, 130

à˵ءÒóì¢éͼԴ¾ÅÒ´, 96

Electronics, 55

¡Òúѹ·Ö¡áººäº¹ÒÃÕ, 5

¢Í§à˵ءÒóì¢éͼԴ¾ÅÒ´, 100

¤ÇÒÁäÁèà» ¹àªÔ§àÊé¹, 5

µèͺԵ (Eb /N0 ), 31

¡ÒûÃѺ¤èÒ·Ò§àÇÅÒ, 37

»ÃÐÊÔ·¸Ô¼Å, 47, 101, 103

¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ, 47 54 Ẻ

¡®¢Í§¿ÒÃÒà´Âì, 5

Ẻ¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇÂ, 53

¡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ, 116, 138,

ÍèÒ¹, 5 8 à¢Õ¹, 5 ¡ÃÐáÊä¿¿ Òà¢Õ¹, 4, 5, 152 ¡ÒáŠÓÃËÑʾÑÅÊì (PCM), 2 ¡ÒêѡµÑÇÍÂèÒ§, 11, 18, 21, 36 ¡Òúǡ ¡ÒÃà»ÃÕºà·Õº ¡ÒÃàÅ×Í¡, 122 ¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡, 1 Ẻ¨ÓÅͧªèͧÊÑ­­Ò³, 9 ¡Òúѹ·Ö¡áººá¹ÇµÑ §, 3, 44, 54, 96, 103, 147

ẺâÁ¹Ô¡ (h0

¡Òúѹ·Ö¡áººá¹Ç¹Í¹, 3, 96, 129, 147

49

ẺÊÁºÙóì, 55, 104, 130, 147 ¡ÒÃà´Ô¹áººÊØèÁ, 20 ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð, 4, 11, 33, 76, 139, 152 ·Õ ´Õ·Õ ÊØ´, 77 ¼ÅµÍºÊ¹Í§, 6, 45 ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡, 5 àÁ·ÃÔ¡«ì, 158 àÍ¡à·È, 6, 55, 104, 130 ¡ÒÃàÅ× Í¹µÓá˹觢ͧ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹ÐẺÊØèÁ, 55, 139 ¡ÒÃá·Ã¡ÊÍ´ÃÐËÇèÒ§ÊÑ­Åѡɳì (ISI), 6, 65,

ÊÑ­­Ò³¾ÑÅÊìà»ÅÕ Â¹Ê¶Ò¹Ð, 6 à˵ءÒóì¢éͼԴ¾ÅÒ´, 97

= 1),

¡ÒÃà¢éҨѧËÇÐ, 18, 22, 34, 67

·ÒÃìà¡çµáºº PR, 13, 44 ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ , 8

52

Ẻ·ÒÃìà¡çµà©¾ÒÐ, 53

¡Ãкǹ¡ÒÃ

141

h1 = 1,

69, 152 ¡ÒÃá»Å§

Z,

24, 28

¡ÒÃá»Å§¿ÙàÃÕÂÃì·Õ µèÍà¹× ͧ·Ò§àÇÅÒ, 8

·ÒÃìà¡çµáºº PR, 12, 44 ¼ÅµÍºÊ¹Í§àªÔ§¤ÇÒÁ¶Õ , 8

¢Õ´¨Ó¡Ñ´«Ùà»ÍÃì¾ÒÃÒáÁ¡à¹µÔ¡, 3


¢éͼԴ¾ÅÒ´

¡Òúѹ·Ö¡áººá¹Ç¹Í¹, 44

¡ÓÅѧÊͧà©ÅÕ Â (MSE), 47, 119 ºÔµà«ÅÅì, 6, 24, 55, 103, 147 ¡ÓÅѧÊͧà©ÅÕ Â·Õ ¹éÍÂÊØ´ (MMSE), 47, 120 ¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ, 47 54 ·Ò§àÇÅÒ, 21, 24, 30, 38 ẺÃÒ¡¡ÓÅѧÊͧà©ÅÕ Â (RMS), 34

¼ÅµÍºÊ¹Í§, 116 ¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð, 6, 45 ¢Ñ ¹ºÑ¹ä´, 25 ¢éͼԴ¾ÅÒ´, 26

¤ÇèÐà» ¹ÁÒ¡ÊØ´ (ML), 44 ¤ÇÒÁ¶Õ Ẻ¹ÍÃìÁÍÅäÅ«ì, 8 ¤ÇÒÁ¶Õ 乤ÇÔµÊì, 44

·ÒÃìà¡çµ, 12, 45, 68, 138 ºÒ§Êèǹ (PR), 12, 43, 68 ÍÔÁ¾ÑÅÊì, 20, 30, 148

¤ÇÒÁ˹Òá¹è¹

ÍÔÁ¾ÑÅÊì¨Ó¡Ñ´ (FIR), 67, 117, 138

¡ÒúÃèØ, 155

ÍÔÁ¾ÑÅÊìäÁè¨Ó¡Ñ´ (IIR), 117

¢Í§¡Òúѹ·Ö¡¢éÍÁÙÅ, 6 ¢Í§¡Òúѹ·Ö¡áºº¹ÍÃìÁÍÅäÅ«ì (ND), 6 Ê໡µÃÑÁ¡ÓÅѧ, 20, 55, 103, 130, 147

ªèͧÊÑ­­Ò³ÍèÒ¹, 4

àªÔ§¤ÇÒÁ¶Õ , 8, 43 àµçÁ, 67 ä´ºÔµ, 8, 45

¿ §¡ìªÑ¹ ¢Ñ ¹ºÑ¹ä´, 23

·ÒÃìà¡çµ, 53

¢éͼԴ¾ÅÒ´, 6

·Õ àËÁÒÐ·Õ ÊØ´, 47 »ÃÐÊÔ·¸Ô¼Å, 122

¤ÇÒÁ˹Òá¹è¹¤ÇÒÁ¹èÒ¨Ðà» ¹áººà¡ÒÊìà«Õ¹, 55, 78, 103, 139, 147

Ẻ GPR, 13, 45 Ẻ

h1 = 1,

52

«Ô§¡ì, 20 ¶èÒÂâ͹, 13, 24, 46, 117, 138

Ẻ·ÒÃìà¡çµà©¾ÒÐ, 53 àÅ¢ªÕ ¡ÓÅѧ, 8 Ẻ¾Åѧ§Ò¹Ë¹Ö §Ë¹èÇÂ, 53 â¤Ã๤à¡ÍÃìà´ÅµÒ, 81 ẺâÁ¹Ô¡, 49, 104, 133, 148 Ẻ PR, 12 ¡Òúѹ·Ö¡áººá¹ÇµÑ §, 44

ÀÒÇÐ, 22 ¡ÒõԴµÒÁ, 22


Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ, 3

¡ÒÃä´éÁÒ, 22 ¡ÒÃä´éÁÒẺÊÁºÙóì, 34

Ẻ¨ÓÅͧ, 4 ÃкºÊ× ÍÊÒôԨԷÑÅ, 17

ÃËÑÊ

(0, G/I),

166 167

ÃٻẺ NRZI, 4, 151

ÃËÑÊ RLL, 3, 91, 151 167 ÅӴѺ¢éͼԴ¾ÅÒ´, 93

ISI, 152 ¡ÒÃÍ͡Ẻ, 162 164 ¤ÇÒÁ˹Òá¹è¹¡ÒúÃèØ, 155 ¤ÇÒÁ¨Ø, 154 155 µÒÃÒ§¤é¹ËÒ, 152, 164

ÍÔ¹¾Øµ·Õ ¶Ù¡µéͧ, 94

ǧ¨Ã VCO, 18 ǧ¨Ã¡Ãͧ ¢éͼԴ¾ÅÒ´¡Ò÷ӹÒÂ, 118

ºÔµÊèǹà¡Ô¹, 153

·Ó¹ÒÂ, 117

¾ÒÃÒÁÔàµÍÃì, 151

·Ó¹ÒÂË¹Ö §¢Ñ ¹áººàªÔ§àÊé¹, 118

ÃËÑÊ

(0, G/I),

166

ÃËÑÊ FM, 164 ÃËÑÊ MFM, 164

¼èÒ¹µ Ó, 4 ¤ÇÒÁ¶Õ µÑ´, 20 ÅÙ», 18

ÃËÑÊ Miller, 164

ǧ¨ÃªÑ¡µÑÇÍÂèÒ§, 4, 18

ÍѵÃÒ¢èÒÇÊÒÃàªÔ§àÊ鹡ӡѺ, 155, 160

ǧ¨ÃµÃǨËÒ, 4

ÍѵÃÒ¤ÇÒÁ˹Òá¹è¹, 155 156

¢Õ´àÃÔ Áà»ÅÕ Â¹

ÍѵÃÒÃËÑÊ, 152

·Õ äÁèÁÕ˹èǤÇÒÁ¨Ó, 31

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, 156 158

·Õ äÁèÁÕ˹èǤÇÒÁ¨ÓẺËÅÒÂÃдѺ, 34

à§× ͹䢺ѧ¤Ñº

ẺËÅÒÂÃдѺ, 21

¨Ó¹Ç¹ÅӴѺ¢éÍÁÙÅ, 153 154 àÁ·ÃÔ¡«ì¡ÒÃà»ÅÕ Â¹Ê¶Ò¹Ð, 158 159 ä·ÁÁÔ §ÃԤѿàÇÍÐÃÕ, 152

ẺÍè͹, 38 Ẻá¢ç§, 38 ¢éͼԴ¾ÅÒ´·Ò§àÇÅÒ, 18

ÃËÑÊ RS, 3

·Õ àËÁÒÐ·Õ ÊØ´, 69

ÃËÑʤ͹âÇÅ٪ѹ, 69

ÅӴѺ·Õ ¤ÇèÐà» ¹ÁÒ¡ÊØ´ (MLSD), 77

ÃËÑÊÁÍ´ÙàŪѹ, 151

ÅӴѺàËÁÒÐ·Õ ÊØ´, 137


ÇÕà·ÍÃìºÔ, 12, 21, 69, 115 ¤ÇÒÁ«Ñº«é͹, 80

¡Òúѹ·Ö¡áººá¹Ç¹Í¹, 6 àÍ¡à·È, 55, 104, 130

¤ÇÒÁÅÖ¡¡ÒöʹÃËÑÊ, 80

ä´ºÔµ, 8

¤ÇÒÁ¹èÒ¨Ðà» ¹¢Í§¢éͼԴ¾ÅÒ´ÅӴѺ, 88

乤ÇÔµÊì, 20

»ÃÐÊÔ·¸ÔÀÒ¾, 87

乤ÇÔµÊìÍØ´Á¤µÔ, 12

ǧ¨ÃµÃǨËÒÅӴѺ, 69

ÊÑ­­Ò³Ãº¡Ç¹, 11, 20, 21, 45, 68, 93

ÍÑÅ¡ÍÃÔ·ÖÁ, 69, 75 80

¡ÒâÂÒÂ, 45, 66

à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, 70 72

¡ÒÃ¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ, 138 140

á¼¹ÀÒ¾à·ÃÅÅÔÊ, 70, 72 75

¡Ò÷ӹÒÂ, 117, 141

ÊÑ­Åѡɳì, 21

¨ÔµàµÍÃì¢Í§Ê× ÍºÑ¹·Ö¡, 55, 61, 103, 107, 138, 139

ǧ¨Ã¶Í´ÃËÑÊ ÁÍ´ÙàŪѹ, 4

·Õ ¢Ö ¹ÍÂÙè¡Ñºáºº¢éÍÁÙÅ, 138

á¡é䢢éͼԴ¾ÅÒ´, 4

·Õ ÍÂÙè¹Í¡á¶º¤ÇÒÁ¶Õ , 11, 20

ǧ¨ÃÁÍ´ÙàÅàµÍÃì, 4, 5

ÀÒÂã¹Ç§¨ÃªÑ¡µÑÇÍÂèÒ§, 18

ǧ¨Ãà¢éÒÃËÑÊ

ÊÕ¢ÒÇ, 60, 119

¡è͹, 91 ÁÍ´ÙàŪѹ, 3

à¡ÒÊìÊÕ¢ÒÇ, 69 à¡ÒÊìÊÕ ¢ÒÇẺºÇ¡, 20, 55, 66, 75, 99, 103, 115, 138

á¡é䢢éͼԴ¾ÅÒ´, 3

ẺÊÕ, 93, 99, 116, 138 ʶһ µÂ¡ÃÃÁªèͧÊÑ­­Ò³ÍèÒ¹, 11

Ê໡µÃÑÁ¤èÒÈÙ¹Âì, 8, 44, 68

ÊÀÒ¾¤ÇÒÁà» ¹áÁèàËÅç¡, 1, 139, 152 ÊÀҾźÅéÒ§áÁèàËÅç¡, 2, 5 ÊÀÒ¾ãËé«ÖÁ¼èÒ¹ä´é, 3

ËÅÑ¡¡ÒÃàªÔ§µÑ §©Ò¡, 119, 142

ÍͿ૵

ÊÁ¡ÒùÍÃìÁÍÅ, 119

·Ò§¤ÇÒÁ¶Õ , 22, 35

ÊÑ­­Ò³¾ÑÅÊì

·Ò§à¿Ê, 21, 22, 33

à»ÅÕ Â¹Ê¶Ò¹Ð, 6, 55, 103, 130, 147 ¡Òúѹ·Ö¡áººá¹ÇµÑ §, 6

ÊÁ¡ÒûÃѺ¤èÒ, 24, 28 ·Ò§àÇÅÒ, 20, 33


ÍѵÃÒ¡ÒêѡµÑÇÍÂèÒ§, 37 Ẻà¡Ô¹¨ÃÔ§, 37

¨Ø´àʶÕÂÃÀÒ¾, 32 à˵ءÒóì¢éͼԴ¾ÅÒ´, 91

ÍѵÃÒ¡ÒÃÅÙèà¢éÒ, 22 25, 28, 38

¡ÒÃÇÔà¤ÃÒÐËì, 91 111

ÍѵÃÒ¢éͼԴ¾ÅÒ´ºÔµ (BER), 47

¡ÒÃÇÔà¤ÃÒÐËìÃÐÂзҧ·Õ ¹éÍÂÊØ´, 104 107

ÍѵÃÒÃËÑÊ, 69, 152, 154, 156

¤ÇÒÁÊÑÁ¾Ñ¹¸ì ÃÐËÇèÒ§

ÍѵÊËÊÑÁ¾Ñ¹¸ì, 49, 60, 100, 119, 142 ÍÑÅ¡ÍÃÔ·ÖÁ

SNReff

áÅÐ BER,

107 109 ¤ÇÒÁËÁÒÂ, 93

PDNP, 140 144

¹ÔÂÒÁ, 95

PS PDNP, 144 146

ÃËÑÊ RLL, 91

ÇÕà·ÍÃìºÔ, 69 80, 122

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å, 99 103

Ẻ»ÃѺµÑÇ, 143

ÃÐÂзҧ»ÃÐÊÔ·¸Ô¼Å¡ÓÅѧÊͧ, 101

ÍÕ¤ÇÍäÅà«ÍÃì, 4, 12, 47, 53, 66, 104

·Õ ¹éÍÂÊØ´, 101

ÍѵÃÒ¡ÒÃÅÙèà¢éÒ, 67

ÃÐÂзҧÂؤÅÔ´, 97 99

Ẻ zero forcing, 138

ÃÐÂзҧÂؤÅÔ´¡ÓÅѧÊͧ, 97

Ẻ¼ÅµÍºÊ¹Í§ºÒ§Êèǹ (PR), 43, 68, 115

·Õ ¹éÍÂÊØ´, 99 ǧ¨Ãà¢éÒÃËÑÊ¡è͹, 91

Ẻ¼ÅµÍºÊ¹Í§àµçÁ, 67 ẹ´ìÇÔ´·ì à¤Ã× Í§Ê¶Ò¹Ð¨Ó¡Ñ´, 70, 71 ÃËÑÊ RLL, 156 ǧ¨ÃµÃǨËÒÇÕà·ÍÃìºÔ, 71 à·¤¹Ô¤¡ÒûÃÐÁÒ³¤èÒ㹪èǧ, 37 à¾Íà«ÍÃìäÇàÇÍÃìä·ÁÁÔ §ÃԤѿàÇÍÃÕ, 38 Ẻ·Ó§Ò¹« Ó, 40

¢Í§ÅÙ», 22, 28 à¡Ô¹à» ¹ÈÙ¹Âì, 20 Ẻ¢éÍÁÙÅ, 138 Ẻ¨ÓÅͧ ¡Ò÷ÓãËéÊÑ­­Ò³Ãº¡Ç¹à» ¹ÊÕ¢ÒÇ, 117 ¡Ò÷ӧҹ¢Í§Ç§¨Ã PLL, 23

àÊé¹·Ò§·Õ ÂѧÁÕªÕÇÔµÍÂÙè, 80, 92

¡ÒÃÍ͡Ẻ·ÒÃìà¡çµ, 47

àÊé¹â¤é§ÃÙ»µÑÇàÍÊ, 21, 30 34

ªèͧÊÑ­­Ò³

¨Ø´ÊÁ´ØÅ, 33

GPR ẺÊÁÁÙÅ, 93


¡Òúѹ·Ö¡ÃкºáÁèàËÅç¡, 9 ·Õ äÁèµèÍà¹× ͧ·Ò§àÇÅÒẺÊÁÁÙÅ, 66 ÍØ´Á¤µÔ, 12 13, 19 àÊÁ×͹¨ÃÔ§, 11 12 Ãкº¡ÒèѴà¡çº¢éÍÁÙÅ´Ô¨Ô·ÑÅ, 3 á¼¹ÀÒ¾à·ÃÅÅÔÊ, 70, 72, 91, 115, 141

ä«à¤ÔÅÊÅÔ», 33 ä·ÁÁÔ §¨ÔµàµÍÃì, 18, 20, 39 ä·ÁÁÔ §¿ §¡ìªÑ¹, 30 Ẻ¹ÍÃìÁÍÅäÅ«ì, 32 ä·ÁÁÔ §ÃԤѿàÇÍÃÕ, 12, 17 41 ¡ÒÃÍ͡Ẻ¤èÒ¾ÒÃÒÁÔàµÍÃì, 23 34 ÀÒÇСÒõԴµÒÁ, 22 ÀÒÇСÒÃä´éÁÒ, 22 Ẻ´Ô¨Ô·ÑÅ, 37 Ẻ·Õ ãªé¡Ñ¹·Ñ Çä», 18 23 »ÃÐÊÔ·¸ÔÀÒ¾, 34 37 Ẻ¹ÔùÑÂ, 18 Ẻ»ÃÐÁÒ³¤èÒ㹪èǧ, 37 ẺÍØ»¹ÑÂ, 18


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.