一种改进的基于LDPC码的信源信道联合译码方法  

A Modified LDPC-based Joint Source-channel Decoding Scheme

作　　者：孙新梅[1] 张晓林[1] 刘荣科[1] 高洁[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100083

出　　处：《遥测遥控》2006年第3期25-29,共5页Journal of Telemetry,Tracking and Command

基　　金：国防预研基金(51421010905HK0125)

摘　　要：提出一种改进的将LDPC迭代译码和隐马尔可夫估计相融合的迭代译码算法。一方面通过隐马尔可夫信源估计获取信源残留冗余,作为外信息提供给LDPC迭代译码;另一方面,通过LDPC迭代译码获得更精确的信源估计模型参数,加快迭代收敛速度,获得更好的迭代译码性能。其优点是无需知道任何关于信源的先验信息,计算复杂度低,并行操作性强,尤其适用于宽带多媒体传输。In this paper,we develop a modified LDPC-based joint source-channel decoding scheme.This iterative decoding scheme combines the hidden Markov source estimation(HMSE) with the low density parity check(LDPC) codes HMSE is used to extract the source redundancy accurately without any prior information about the source and offers external information to the LDPC decoder so as to helps to progress the performnce.Moreover,the decoding procedure may be implemented in parallel,resulting in good performance with low complexity and delay.Simulation results show that the proposed scheme may achieve better performance than that of standard decoding scheme over BIAWGN channel.Especially,it may work very well in multimedia communications.

关 键 词：联合信源信道译码 低密度奇偶校验码 迭代译码 隐马尔可夫模型 隐马尔可夫信源估计 

分 类 号：TN911.2[电子电信—通信与信息系统]