A Framework for Hardware Impairments-Aware Multi-Antenna Transceiver Design in IoT Systems via Majorization-Minimization

作     者：Gong, Shiqi Wang, Jintao Zhao, Xin Ma, Shaodan Xing, Chengwen 

作者机构：Univ Macau State Key Lab Internet Things Smart City Macau Peoples R China Beijing Inst Technol Sch Cyberspace Sci & Technol Beijing 100081 Peoples R China Univ Macau Dept Elect & Comp Engn Macau Peoples R China Beijing Inst Technol Sch Informat & Elect Beijing 100081 Peoples R China 

出 版 物：《IEEE INTERNET OF THINGS JOURNAL》 (IEEE Internet Things J.)

年 卷 期：2023年第10卷第1期

页      面：417-433页

核心收录：

学科分类：0810[工学-信息与通信工程] 0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Natural Science Foundation of China Science and Technology Development Fund, Macau [0051/2022/A1, 0036/2019/A1, SKL-IOTSC(UM)-2021-2023] Research Committee of University of Macau [MYRG2020-00095-FST] 

主　　题：Hardware MIMO communication Transceivers Receivers Internet of Things Optimization Minimization Average total mean square error (MSE) hardware impairments low-complexity scheme majorization-minimization (MM)-based iterative algorithm MSE floor effect stochastic channel state information (CSI) errors 

摘      要：In view of the nonideality of communication links in the Internet of Things (IoT) originating from transceiver hardware impairments, in this article, we introduce a general framework for hardware impairments-aware multiantenna transceiver design, which considers different availabilities of CSI at the transmitter (CSIT) and the receiver (CSIR). The well-known Kronecker model is applied to characterize stochastic channel state information (CSI) errors. For each case, we aim to minimize the (average) total mean square error (MSE) of all data streams subject to the practical per-antenna power constraints. To address the nonconvexity of the formulated problem, we propose an efficient majorization-minimization (MM)-based iterative algorithm to transform the original problem into a series of convex subproblems with semiclosed-form optimal solutions. For low-complexity implementation, we also develop an alternative scheme for directly finding a high-quality suboptimal solution by considering both worst case hardware impairments and worst case CSI errors. In particular, since an explicit expression of the average total MSE for the perfect CSIR and imperfect CSIT case is hard to derive, we instead optimize its effective upper and lower bounds. The prospective applications of our work in the two currently popular multiple-input-multiple-output (MIMO) IoT scenarios are then discussed. Furthermore, we fundamentally reveal the MSE floor effect caused by both hardware distortion and CSI imperfection in the high-SNR regime. Numerical results illustrate the excellent average total MSE and average bit error rate (BER) performance of our proposed algorithms over the adopted benchmark schemes.