基于临近信息传递算法的大规模MIMO信号检测  

Signal detection for massive MIMO systems using approximate message passing

作　　者：陈建新 方勇[1] 孙彦赞[1] Chen Jianxin;Fang Yong;Sun Yanzan.(School of Communication and Information Engineering, Shanghai University, Shanghai 200444, Chin)

机构地区：[1]上海大学通信与信息工程学院

出　　处：《电子测量技术》2018年第3期17-21,共5页Electronic Measurement Technology

基　　金：国家自然科学基金(61271213,61673253,61501289);教育部博士点基金(20133108110014)项目资助

摘　　要：为了应对大规模MIMO系统由于天线数量庞大带来信号检测计算复杂度高的问题，提出了改进的临近信息传递算法。首先通过在因子图上建立大规模MIMO概率系统模型实现符号检测；然后通过对干扰信号与噪声进行高斯临近能够降低信息计算的数量与复杂度；最后为了使临近信息传递算法应用在高阶调制系统，采用高斯树临近方法临近信息。通过仿真发现当系统调制方式为BPSK时，两种方法随着天线数量的增加其误码率均逼近最大似然检测器，当系统调制方式为高阶调制（如16QAM）时，高斯树临近方法优于高斯临近。In this paper, an improved approximate message passing algorithm （AMP） is proposed to deal with the problem of high computation of signal detection caused by the large number of antennas in massive MIMO system. We first achieve symbol detection by establishing a massive MIMO system model on the factor graph. Then, we approximate the interference signal and noise through Gaussion approximation to reduce complexity and the number of information calculation. Finally, a Gaussian tree approximation is adopt to apply the approximation message passing algorithm to the high-order modulation system. Simulation results show that when the system modulation mode is BPSK, the bit error rate of Gaussian and Gaussian tree approximation towards the maximum likelihood detector with the number of antennas increasing. When high-order modulation is used, such as 16QAM, the Gaussian tree approximation is superior to Gaussian approximation.

关 键 词：大规模MIMO 信号检测 因子图 临近信息传递 高斯临近 

分 类 号：TN929.5[电子电信—通信与信息系统]