Robust Beamforming Design for Self-Sustainable IRS-Aided MIMO SWIPT Communication

Abstract

This paper investigates a novel self-sustainable intelligent reflecting surface (IRS)-enhanced multi-input multi-output simultaneous wireless information and power transfer (SWIPT) system under imperfect channel state information. Equipped with an energy harvesting module, the IRS harvests energy from received signals to meet its operational needs. In particular, we focus on maximizing the weighted sum rate (WSR) of all information users (IUs) by jointly optimizing the access point (AP) precoding matrices and IRS reflection coefficient matrix. Meanwhile, the WSR maximization is constrained by the maximum transmit power of the AP and by the harvested power requirements of both IRS and energy users (EUs). To address the highly coupled and non-convex problem, we first adopt the minimum mean square error method to recast the WSR maximization problem into a simpler equivalent problem and split the transformed problem into two sub-problems, which can be tackled alternatively. In addition, we employ the successive convex approximation (SCA) technique to approximate the non-convex sub-problems. Furthermore, we adopt the Lagrangian dual transformation, majorization-minimization method and penalty-based technique to obtain the closed-form solutions. The simulation results demonstrate the effectiveness of the proposed algorithm and show that, with proper IRS deployment, the proposed self-sustainable IRS can achieve 96%–99% of the WSR obtained by the externally powered IRS under the default simulation setup, while eliminating the need for an external power supply.