基于子图的边缘串行消息传递算法

Edge Serial Message Passing Algorithm Based on Subgraph

稀疏码分多址接入(Sparse Code Multiple Access,SCMA)作为一种重要的码域非正交多址技术,能够满足5G通信的海量机器接入需求,但是接收端采用的消息传递算法(Message Passing Algorithm,MPA)存在收敛速度慢、计算复杂度较高的问题.本文提出一种基于子图的边缘串行消息传递算法(SE-MPA),将原始因子图划分为多个不同子图,在子图的基础上优化消息更新顺序,利用边缘串行更新序列集,被更新的用户节点消息立刻用来更新相连的资源节点.仿真结果表明,该算法在少量系统BER性能损失条件下,能够有效加快消息收敛速度,降低计算复杂度.

As an important code domain non orthogonal multiple access technology,sparse code division multiple ac⁃cess(SCMA)can meet the requirements of 5G communication for massive machine access.However,the message passing algorithm(MPA)adopted by the receiver has the problems of slow convergence speed and high computational complexity.This paper proposes an edge serial message passing algorithm based on subgraph(SE-MPA).The original factor graph is di⁃vided into several different subgraphs.Based on the subgraph,the message update sequence is optimized,and the edge seri⁃al update sequence set is adopted.The updated user node messages are used to update the connected resource nodes immedi⁃ately.Simulation results show that the algorithm can effectively speed up the message convergence and reduce the computa⁃tional complexity in the case of a small amount of BER performance loss.