Transmission Design for Intelligent Reflecting Surface-Aided MIMO SWIPT Systems With Finite-Alphabet Inputs

• Published in: IEEE transactions on communications Vol. 72; no. 5; pp. 3095 - 3109
• Main Authors: Zhu, Xiaodong, Liu, Zhen, Shi, Yuzhong, Di, Haoyang, Meng, Zhonglou, Tu, Xiaodong
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.05.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Approximation algorithms; Codes; Convexity; Energy harvesting; finite-alphabet inputs; Intelligent reflecting surface (IRS); Maximization; MIMO communication; multiple-input multiple-output (MIMO); Optimization; Phase shift keying; Power transfer; Quadrature amplitude modulation; Receivers; Reconfigurable intelligent surfaces; simultaneous wireless information and power transfer (SWIPT); transmission design; Wireless communication
• ISSN: 0090-6778; 1558-0857
• DOI: 10.1109/TCOMM.2024.3356445

This paper investigates the transmission design for an intelligent reflecting surface (IRS)-aided simultaneous wireless information and power transfer (SWIPT) system, where the base station (BS), information decoding receiver (IDR), and energy harvesting receiver (EHR) are equipped with multiple antennas. The IRS can facilitate the transmission of information and the acquisition of energy by regulating its reflection phase shifts. By jointly optimizing the transmit precoder at the BS and the phase shifts of the IRS, our goal is to maximize the mutual information (MI) between the BS and the IDR subject to the constraints of the transmit power at the BS and the harvested energy at the EHR. Unlike existing works, where ideal Gaussian signals are assumed to be the inputs, this paper considers the actual scenario of finite-alphabet signals, such as phase shift keying (PSK) and quadrature amplitude modulation (QAM), as the input signals. To facilitate the solution of this optimization problem, we use the channel cutoff rate (CR) as a surrogate for MI. By doing this, the original problem of MI maximization is transformed into an approximate equivalent problem of CR maximization. Two cases, where perfect or imperfect channel state information (CSI) is available at the BS, are both considered. The formulated problems in the two cases are intractable due to the non-convexity and variable coupling. However, by analyzing the structure of the problems, for each CSI case, we develop a corresponding solution algorithm based on alternating optimization. For the perfect CSI case, the algorithm combines successive convex approximation (SCA) with penalty-based manifold optimization, while, for the imperfect CSI case, the algorithm is involved with SCA, S-Procedure, and semidefinite relaxation (SDR). Both the solution algorithms can be utilized to effectively optimize the precoder and the phase shifts, whose superiority is well verified by simulation results.