Throughput and harvested energy optimization in RIS-assisted multi-user OFDMA-SWIPT systems

Abstract

Reconfigurable Intelligent Surface (RIS) is a promising technology that enables control of a wireless environment to boost system performance. It has attracted attention in the context of beyond 5G (B5G) wireless communication. The demand for power and data in the Internet-of-Things (IoT) applications is envisioned to be fulfilled using RIS with the Simultaneous Wireless Information and Power Transfer (SWIPT) technique. In this paper, we explore RIS-assisted multi-user (MU) orthogonal frequency division multiple access (OFDMA), where users are capable of enabling SWIPT by utilizing either Time Switching (TS) or Power Splitting (PS) protocols of SWIPT. We intend to optimize the performance of RIS-assisted SWIPT-IoT systems, with focus on maximization of two distinct objectives, viz., Sum-Rate and sum-Energy Harvesting (EH), while adhering to Quality-of-Service (QoS) constraints by optimizing the PS/TS factor, sub-carrier assignment matrix, and power allocation to sub-carriers. The proposed alternating optimization-based algorithmic solutions are compared with the asymptotic solution based on the Block Coordinate Descent (BCD) technique while considering the Practical Discrete Phase Shift (PDPS) Model of RIS. It is clear from the results of our simulation that a greater number of RIS elements and sub-carriers is required to achieve a higher throughput and sum-EH. © 2026 Elsevier B.V.