| 摘要: |
| 由于天线间距较近和一些非白高斯干扰，大规模多输入多输出（Multiple-Input Multiple-Output,MIMO）系统的信道矩阵存在如列相关、低秩和非零均值等“病态”特征，给基于近似消息传递的迭代接收机算法造成了严重的性能损失。为此，提出了一种基于酉变换近似消息传递(Unitary Transformation Approximate Message Passing,Ut-AMP)的接收机算法。首先针对均值非零、列相关和条件数过大三种常见的病态特征，分析了酉变换的必要性；其次，利用变量间的马尔科夫特性将大规模MIMO系统进行因式分解和因子图建模，推导了联合置信传播、平均场和Ut-AMP的消息传递算法，并与现有算法的复杂度进行了对比；最后建立大规模MIMO接收机模型，对所有相关算法进行数值仿真。仿真结果表明，所提接收机算法能够以复杂度略有提升为代价，在病态信道矩阵环境中大幅提升接收机的鲁棒性，具有很高的研究和应用价值。 |
| 关键词: 大规模MIMO接收机 酉变换 病态矩阵 近似消息传递 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目(61801434)；中国博士后科学基金面上项目(2019M652576,2018M642784)；河南省科技公关项目(202102210313，202102210172)；河南省高等学校重点课题(20B510005) |
|
| Massive MIMO Receiver Design Using Unitary Approximate Message Passing |
| WANG Youshun,GAO Tongdi,YUAN Zhengdao,WANG Jiabin,ZHU Zhengyu |
| (1.Artificial Intelligence Engineering Research Center,Zhengzhou Vocational University of Information and Technology,Zhengzhou 450001,China;
2.School of Information Engineering,Zhengzhou University,Zhengzhou 450001,China;;3.The 713th Research Institute of China Shipbuilding Industry Corporation,Zhengzhou 450001,China) |
| Abstract: |
| Due to the close antenna distance and some non-white Gaussian interference,the channel matrix of massive multiple-input multiple-output(MIMO) system has some ill characteristics such as column correlation,low rank and non-zero mean,which may cause serious performance losses to the message passing based iterative receiver.Based on unitary transformation approximate message passing(Ut-AMP),an iterative algorithm for massive MIMO is proposed.Firstly,the necessity of the unitary transformation is analyzed according to three common ill characteristics,such as non-zero mean,column correlation and large conditional number.Secondly,factorization and modeling of massive MIMO system are carried out using Markov characteristics between variables,and the message passing algorithms of joint belief propagation,mean field and Ut-AMP are derived.Finally,a massive MIMO receiver environment is established and the proposed algorithm and existing algorithms are simulated and compared.Simulation results show that the proposed algorithm can greatly improve the robustness in ill-conditioned channel matrix environment with only a little increase in complexity,which is of high research and application value. |
| Key words: massive MIMO receiver unitary transformation ill matrix approximate message passing |
《电讯技术》编辑部 