北京邮电大学学报

  • EI核心期刊

北京邮电大学学报

• •    

基于信道信息的动态加权多元LDPC译码算法

陈海强1,冼文烙1,李昱霖1,李庆年2,黎相成1   

  1. 1. 广西大学
    2. 南宁学院
  • 收稿日期:2023-10-09 修回日期:2024-01-17 发布日期:2024-06-25
  • 通讯作者: 李庆年
  • 基金资助:
    国家自然科学基金;国家自然科学基金;国家自然科学基金;广西重点研发计划

Dynamic weighted non-binary LDPC decoding algorithm based on channel information

  • Received:2023-10-09 Revised:2024-01-17 Published:2024-06-25

摘要: 对于多元LDPC码,基于信道信息的动态加权多元LDPC译码算法(CIW-AlgB)在可靠性计算中考虑来自信道的初始消息与外信息之间的相关可靠性代替硬判决信息与外信息之间的二进制汉明距离,并提出基于信道信息的加权修正系数,可以根据信道信息对译码算法进行动态加权。与多元加权符号翻转算法相比,本文算法在译码性能和译码复杂度之间进行了有效的折中。仿真结果表明,与加权符号翻转译码算法相比,所提算法能够获得显著的性能提升且具有更快的译码收敛速度。

关键词: LDPC码, 符号翻转, 信道, 动态加权

Abstract: For multi-component LDPC codes, the channel information based dynamic weighted multi-component LDPC decoding algorithm (CIW-AlgB) replaces the binary Hamming distance between hard decision information and external information in reliability calculation by considering the correlation reliability between the initial message from the channel and external information, and proposes a weighted correction coefficient based on channel information. The decoding algorithm can be dynamically weighted according to the channel information. Compared with the non-binary weighted symbol flipping algorithm, the proposed algorithm makes an effective compromise between decoding performance and decoding complexity. The simulation results show that compared with the weighted symbol flipping decoding algorithm, the proposed algorithm achieves significant performance improvement and has faster decoding convergence speed.

Key words: LDPC code, symbol flipping, channel, dynamic weighted

中图分类号: