一种改进的LDPC码译码算法

An Improved Decoding Algorithm for LDPC

指出了由于短LDPC(低密度奇偶校验)码的Tanner图中会出现环路,使得变量节点之间的信息不再相互独立,从而导致对数域BP译码算法的性能下降.从平衡迭代译码性能与硬件实现复杂度的角度出发,提出了一种改进的BP算法,通过变量消息的修正来补偿校验消息简化的损失.仿真结果表明:改进的算法加快了算法的收敛速度,降低了迭代译码次数,减小了译码延时,是一种兼顾性能与复杂度的折中算法.

Loops usually exist in the Tanner graph of short LDPC codes, and the information among the variable nodes is not independent, and consequently the performance of LLR BP decoding algorithm will decrease. For balancing the performance and complexity, a new modified decoding algorithm is proposed based on BP algorithm, which normalizes the variable message to compensate the simplified check message. The simulation results show that the proposed algorithm can improve the convergence property and reduce the computation complexity.