一种多进制LDPC码加权符号翻转译码算法  

Weighted Symbol-Flipping Decoding Algorithm for Nonbinary LDPC Codes

作　　者：张用宇[1] 

机构地区：[1]中国人民解放军91469部队

出　　处：《通信技术》2015年第11期1222-1227,共6页Communications Technology

摘　　要：提出了一种低复杂度基于翻转规则的多进制低密度奇偶校验(Low-Density Parity-Check,LDPC)码符号翻转译码算法。为寻求有效码字,该算法在符号向量空间迭代地更新硬判决的接收符号向量。每一次迭代只改变一个符号,其符号翻转函数综合考虑了不满足校验式的个数和接收比特和计算出符号的可靠性度量。在高阶伽罗华域中采用一种无限环路规避和翻转符号选取方法,同时提出了翻转规则设计方法,该设计决定了计算复杂度和差错性能。仿真结果表明,该符号翻转算法在帧长为150符号的16进制LDPC码中取得了纠错性能和计算复杂度的有效权衡。A low-complexity symbol-flipping algorithm for nonbinary LDPC （Low-Density Parity-Check） codes based on flipping rules is proposed. In searching of valid codeword, the proposed algorithm iterative- ly updates the received symbol vector of hard-decision in symbol vector space. Only one symbol is flipped in each iteration, and symbol flipping function comprehensively considers the number of failed checks and the reliability of the received bits and calculated symbols. A scheme to avoid infinite loops and select flip- ping symbol in high-order Galois field is adopted and meanwhile, the design of flipping rules is also pro- posed, which determines the computational complexity and error performance. Simulation results indicate that this algorithm could achieve an effective tradeoff of between error-correcting performance and compu- tational complexity for the 16-ary （150,76） LDPC code.

关 键 词：多进制 低密度奇偶校验码 迭代译码 翻转规则 

分 类 号：TN911.22[电子电信—通信与信息系统]