LDPC调制系统中迭代解调解码信息的线性优化  

Linear Optimization of Iterative Demapping/Decoding LLR for LDPC Coded BICM Systems

作　　者：黄平[1] 李岳衡[1] 

机构地区：[1]河海大学计算机与信息学院,南京211100

出　　处：《电讯技术》2015年第12期1395-1400,共6页Telecommunication Engineering

基　　金：河海大学中央高校基本科研业务费专项资金项目(2010B06514)~~

摘　　要：在中短码长低密度奇偶校验（LDPC）码编码的比特交织编码调制系统（BICM）中,由二分图短圈导致的置信传播（BP）解码器输出外附信息之间的相关性是降低迭代解调/解码性能的主要原因,因此有必要对解码器输出外附对数似然比（LLR）信息进行优化.针对该问题,提出了一种简单有效的LLR线性优化算法.该算法首先基于匹配LLR值应满足的连续性条件,并结合BP解码器输出外附LLR信息的统计特性,对解码器输出LLR信息的条件概率密度函数（PDF）的计算进行简化,推导出了乘性优化因子,据此对解码器输出外附信息进行线性优化.仿真结果表明,与未优化方案相比,对于（504,252）LDPC码编码的BICM系统,所提优化算法可获得0.6-0.8dB的信噪比增益;和推广互信息（GMI）优化方法相比,两者的误码性能接近,所提算法仅差0.1dB,且具有更小的运算复杂度.In the medium and short length Low-Density Parity-Check( LDPC) coded Bit Interleave Coded Modulation( BICM) systems, due to the short circle in the bipartite graph, the dependency between Log Likelihood Ratio( LLR) of the variable nodes of Belief Propagation( BP) decoding is the major factor to the degradation of iterative demapping/decoding performance. It is necessary to optimize the output extrinsic LLR from BP decoder. In this paper, a linear LLR optimization algorithm for BP decoder output is pro-posed. Based on the consistency condition which matched LLR should satisfy,combining with the statistic characteristic of BP decoding output LLR,the proposed algorithm simplifies the probability density function ( PDF) calculation of BP decoding output LLR and derives a simple optimization factor which is used to the BP decoding output LLR for optimization. Simulation results show that the proposed method achieves 0. 6~0 . 8 dB Signal-to-Noise Ratio ( SNR ) gains compared with the traditional BICM Iterative Decoding (BICM-ID) systems with(504,252) LDPC codes. Moreover,the performance of the proposed method is close to that of the Generalized Mutual Information(GMI) optimization by 0. 1 dB with minor complexity.

关 键 词：低密度奇偶校验码 比特交织编码调制 迭代解调/解码 线性优化 

分 类 号：TN911.22[电子电信—通信与信息系统]