5194

4864

كارشناسي ارشد

مخابرات سيستم

اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر

1388

[هشت]،75ص. :مصور،جدول،نمودار

ص.ع.به فارسي و انگليسي

محمد علي خسروي فرد

2/3/89

علي محمد دوست حسيني، حسين سعيدي

1396/09/29

كتابنامه

مهندسي برق و كامپيوتر

ID4864

به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي

Properties of the Random LT Code Fatemeh Arbabjolfaei f arbabjolfaee@ec iut ac ir March 2010 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Language Persian Degree M Sc st M A Khosravifard khosravi@cc iut ac irAbstractOne of the widely used models for communication channels is the Binary Erasure Channel BEC An important application of the BEC is to model data transmission over theInternet Mostly the reliability of transmission of data over the Internet is guaranteed bythe use of appropriate protocols These protocols are inefficient in many cases such astransmission of data from one server to multiple receivers or transmission of data overpoor wireless or satellite links Another solution to make the data transmission reliable is touse linear erasure correcting codes An estimation of the erasure rate of the channel isrequired in designing these codes However in situations such as satellite or wirelesstransmission where receivers experience sudden abrupt changes in their receptioncharacteristics it is not possible to track the erasure rate of the channel correctly In orderto provide reliability and efficiency in data transmission a new class of codes calledFountain codes Rateless codes have been proposed Given a set of input symbols message symbols a Fountain code produces apotentially limitless stream of output symbols linear combinations of the input symbols Each receiver tries to recover the original input symbols from the set of received outputsymbols For BEC decoding of a Fountain code is equivalent to solving a consistentsystem of linear equations and its computational complexity is deeply influenced by thecoefficient matrix of the system Computational complexity of decoding and the number ofoutput symbols required to achieve a high probability of successful decoding are the mainconcerns in designing Fountain codes With the increase in the processing power of theCPUs it seems that the number of required output symbols will become the mostimportant factor In this thesis two important Fountain codes designed for the ML decoding aredescribed one of them has been selected for MBMS download delivery within thirdgeneration partnership project 3GPP Then we try to find the optimal Fountain code withML decoding algorithm in the cases that the complexity of decoding is not of importance We define the quasi random LT code and introduce it as the unique optimal Fountain code

محمد علي خسروي فرد

علي محمد دوست حسيني، حسين سعيدي