基于遗传思想的多元LDPC码拓展最小和改进算法

Improved extended min-sum algorithm for non-binary LDPC codes based on genetic algorithm

针对多元LDPC码的EMS译码计算量的问题,采用遗传算法的思想对其进行了改进,选择高度可靠的变量节点,针对这些特殊节点增加附加运算,即采用修正参数放大消息判决位对应符号置信值的方法,而其他置信值相对变小。这种方法通过加快收敛速度,减少了译码过程迭代的次数,最终能够减少译码计算量,节省计算资源。通过仿真结果及计算量分析,证明了改进算法能够在不影响译码纠错性能的情况下降低计算量。

The paper studied EMS decoding algorithm of non-binary low density parity check codes and the thought of the genetic algorithm was adopted to reduce computation complexity. In decoding, highly reliable variable nodes were selected, for these special nodes, attached operation was added, i.e. correction parameter was used to enlarge the belief value of the symbols corresponding to the judgment location, while other belief values relatively were decreased. This method can speed up convergence rate and reduce the number of decoding iterations, finally reduce decoding computation complexity and save computation resources. The simulation and computational analysis show that the improved algorithm can reduce computation complexity without affecting the error correction performance.