摘要
对短LDPC码的分阶统计译码(OSD)算法进行了分析,研究了BP和OSD的级联算法及对数似然比累积算法,考虑到译码复杂度和性能的折中,提出了一种改进的级联算法,用最小和算法替代BP算法,然后与对数似然比累计算法进行级联。仿真结果表明,和原始的BP算法相比,译码性能有了很大的提升;和BP-OSD级联算法相比,译码复杂度降低,性能几乎一致。
This paper analyses the ordered statistic decoding (OSD) algorithm of short LDPC codes and investigates BPOSD concatenation algorithm and LLR accumulation algorithm. Considering the tradeoff between the complexity and performance of decoding, the min-sum algorithm was used instead of the BP algorithm, and it was concatenated with LLR ac- cumulation algorithm. Simulation results show that compared with the original BP algorithm, the performance of the im- proved concatenation algorithm has been greatly raised, and compared with the BP-OSD concatenation algorithm, its decoding complexity has been reduced and the performance is nearly the same.
出处
《重庆邮电大学学报(自然科学版)》
北大核心
2010年第1期50-53,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
江苏省高校自然科学研究计划项目(04KJB510094)
江苏省高校研究生创新计划(xm04-32)