基于改进的非二进制LDPC解码算法

摘要

由于传统的LDPC码位翻转解码算法在计算复杂度方面无法得到提高等缺点,为降低传统非二进制低密度奇偶校验码解码算法的复杂度和误码率,提出一种改进的基于NB-LDPC解码算法.对现有的并行符号翻转解码算法(parallel symbol flipping decoding algorithm,PSFD)进行优化改进,采用多次投票方法,使每个不满意的校验节点对相应的变量节点进行多次投票,并结合误差估计方法.考虑到额外的复杂性,两级投票是提出优化算法的最佳选择.大量实验结果表明,相比原有的LDPC码位翻转解码算法,该优化算法避免使用码相关的投票阈值,在BER等于10-5时,误码率低约0.4 dB,在中低数值的SNR上,具有最优参数的优化算法比原有算法收敛得更快.%Because the computation complexity of the traditional LDPC decoding algorithm cannot be improved, to reduce the complexity and error rate of the decoding algorithm of non-binary low density parity check codes, a NB-LDPC decoding algorithm was proposed.The existing algorithm of parallel symbol flipping decoding (PSFD) was optimized and improved, the multiple voting method was used, making every unsatisfied check node vote for corresponding variable nodes, and the error estimation method was combined.Considering the extra complex factors, tow-level voting method was the optimal strategy for decoding algorithm.The experimental results show that the proposed algorithm avoids the use of the code related voting threshold.When the BER is equal to 10-5, the bit error rate is about 0.4 dB lower.In addition, for the low and middle value SNR, the optimization algorithm with the optimal parameters converges faster than the original algorithm.