Resource allocation scheme for multi-cluster NOMA-SWIPT systems with multiple IRSs

Abstract

To address the challenge of providing seamless cellular connectivity for a massive number of Internet-of-Things (IoT) devices within limited resources, this paper integrates non-orthogonal multiple access (NOMA), simultaneous wireless information and power transfer (SWIPT), and intelligent reflecting surface (IRS) technologies. Considering practical factors such as imperfect channel state information (CSI), non-ideal successive interference cancellation (SIC), and non-linear energy harvesting (EH), this paper constructs a multi-cluster NOMA-SWIPT transmission model assisted by multiple IRSs. For this model, with the aim of maximizing system energy efficiency (EE) under constraints such as maximum base station (BS) transmit power, IRS reflection phase shifts, minimum transmission rate, and minimum energy harvesting (EH), a non-convex resource allocation problem is formulated. The solution to this problem requires the joint optimization of the BS transmit beamforming vectors, IRS reflection phase shifts, and power splitting (PS) ratios. To address this challenge, the original problem is initially decomposed into three non-convex sub-problems. Subsequently, by employing Schur’s complement, the S-procedure, General sign-definiteness, and successive convex approximation (SCA), these non-convex sub-problems are transformed into solvable convex optimization sub-problems. Finally, an alternating iterative method is proposed to solve these sub-problems, thereby addressing the original resource allocation problem. Simulation results validate not only the convergence of the robust resource allocation scheme based on alternating iteration but also demonstrate that leveraging the close-range coverage of IRSs can significantly enhance system energy efficiency, even under non-ideal SIC and imperfect CSI conditions.