基于分组混合策略的LDPC置信传播译码算法

LDPC BP-based decoding algorithm combined with group shuffle schedule

置信传播(BP-Based)译码算法通过对校验节点消息的简化处理,并在传递的变量信息之间引进相关性,可大大降低算法的复杂度。为了减少BP-Based算法的迭代次数和性能损失,采用分组混合策略对BP-Based算法进行改进,加快了信息更新速度,减少了错误信息从停止集传播到Tanner图中其他节点的概率。MATLAB仿真表明,在几乎不增加复杂度的情况下,改进算法可以提高译码性能。

The belief propagation （BP-based） decoding algorithm can reduce the computational complexity by simplifying the information update in the check nodes and introducing the correlation among the transmitting information. In order to decrease the performance loss and the iteration numbers of the simplified BP-based decoding algorithm, the improved decoding algorithm combined the simplified BP-based decoding algorithm with group shuffle schedule was presented in this paper. The information update in the check nodes was speeded up and the information error probability from the stop set to the nodes in Tanner graph was significantly lowered in such improved algorithm. The MATLAB simulation results show that the improved algorithm can effectively enhance the decoding performance.