Rateless Rate Adaption Based On Erasure Codes

Signal attenuation, noise, multipath fading, and interference all make channelconditions varying timely which make it challenge to achieve high throughput in thewireless communication system. Rateless based on erasure codes compared with adaptivemodulation and coding technology that has some characteristics of high precisionadaptive, low complexity of PHY and backward compatibility with the protocol. Intime-varying conditions,Rateless based on erasure codes can narrow gap between theactual spectral efficiency of the system and the Shannon limit.In this thesis，we built a rate adaptive communication system based on erasurecodes which mainly consists of raptor codes, rate adaptation strategies and BP decoding.Raptor codes is the core of the whole system and also a key part of the wholecommunication system spectral efficiency. Adaptive strategy involved in deciding whichbits take part in decoding. Adaptive strategy’s performance decides the rate-adaptivewhether could exact match varying channel conditions, so we will study puncturingadaptive strategy and incremental redundancy strategies that impact on systemperformance. Since BP decoding has some shortcomings, the paper also proposeddecoding algorithm based on the pilot codes and restrained brief oscillate algorithm toimprove BP decoding performance. Pilot codes of decoding algorithm based on full useof known information to help BP decoding, thus BP decoding can correct more bit errors,through experimental tests, pilot codes decoding algorithm can effectively improve theaverage brief of node, reduce decoding complexity, improve performance of erasurecodes. There are some loops in Tanner graph which will affect convergence of the decoder,restrained brief oscillate algorithm can effectively inhibit the Tanner graph loop brings theinevitable shocks. Soft value zero algorithm improve the performance of erasure codes.This paper studies rate adaptive technology based on erasure codes which canimprove the spectral efficiency of the system and reduce system implementationcomplexity, what’s more which can improve system decoding performance in practice.