基于自纠正最小和的LDPC译码改进算法

Improved algorithm based on self-corrected min-sum decoding for LDPC codes

针对低密度奇偶校验码(LDPC)中非零元素含量较低,导致在译码过程中存在节点信息可靠度判定不够准确的问题,本文提出了一种基于自纠正最小和的LDPC译码改进算法。分析置信传播(BP)译码算法的迭代规律,利用最小和算法的第一最小值和第二最小值设置变量节点信息的修正阈值,改进自纠正最小和算法。采用次序统计量理论获取与两个最小值对应的归一化因子,阻止不可靠变量节点信息在迭代译码过程中的传递扩散。仿真结果表明:在误比特率为10^(-5)时,该算法可获得约0.2 dB的译码性能增益,平均迭代次数最多可降低18.2%,证明所提算法可有效提高译码性能和迭代收敛性能。

Low density parity check code is a kind of linear block code.Due to the low content of non-zero elements,there is a problem that the node information reliability determination is not accurate enough in the decoding process.Therefore,an improved LDPC decoding algorithm based on self correcting minimum sum is proposed.The iteration rule of BP decoding algorithm is analyzed.The first minimum and the second minimum of the minimum sum algorithm are used to set the correction threshold of variable node information,and the self-corrected min-sum algorithm is improved.The order statistics theory is used to obtain the normalization factor corresponding to the two minimum values to prevent the transmission and diffusion of unreliable variable node information in the iterative decoding process.Simulation results show that when the bit error rate is 10^(-5),the proposed algorithm can obtain about 0.2 dB of decoding performance gain,and the average number of iterations can be reduced by 18.2%at most,which proves that the proposed algorithm can effectively improve the decoding performance and iterative convergence performance.