多元LDPC码Min-max自适应译码算法

Adaptive Min-max algorithm for non-binary LDPC decoding

为了降低多元低密度奇偶校验(Low-density parity check,LDPC)码Min-max译码算法的运算量,提出一种自适应Min-max(Adaptive Min-max,AMM)译码算法。该方法以Min-max算法为基础,以每次迭代后的校验节点错误率(Checknode Error Rate,CER)为调节参数,采用自适应算法对变量节点的向量长度进行截短,去除置信度较低的分量,仅对置信度较高的分量进行更新。当CER降低到一定程度时,对校验节点个数进行自适应截短,仅对不满足校验方程的校验节点进行消息迭代更新,进一步降低AMM算法的复杂度。仿真结果表明,在相同误码性能条件下,AMM算法运算量较固定长度截短的Min-max算法减少20%。

In this paper, an adaptive Min-max（AMM） algorithm for non-binary LDPC decoding is proposed to reduce the computational complexity of the Min-max decoding algorithm. Using the error rate of check-nodes as an adjusting parameter to truncate the message vector in decoding iteration adaptively, the computation complexity of AMM algorithm is greatly reduced with little per- formance loss. When the error rate of check-nodes decreases to a certain small value, stop updating the messages of the correct checked nodes. By reducing the number of nodes participated in updating adaptively, the computation complexity of AMM algo- rithm is further reduced. The simulation results show that the computation complexity of AMM algorithm is reduced by 20% com- pared with the fixed message truncation Min-max algorithm given the same bit error rate.