Reinforcement Learning and Deep Learning-Assisted Spectrum Management for RIS-SWIPT-Enabled 6G Systems

Abstract

6G wireless communication networks may satisfy all the requirements that cannot be realized with 5G networks. In this vein, reconfigurable intelligent surfaces (RISs) have received extensive research attention recently due to their spectral and energy efficiency, and ease of deployment. Correspondingly, this chapter first provides an overview of RIS technology with its specifications and characteristics. Furthermore, the chapter proposes a RIS-aided simultaneous wireless information and power transfer (SWIPT) mechanism for vehicle-to-everything networks that can utilize both the advantages of RISs and SWIPT mechanisms to enhance the network's performance. A case study presents a proposed model designed to minimize transmitted power while maximizing the total harvested energy and overall rate. Next, a summary of the existing works and deep reinforcement learning (DRL) framework for RIS-SWIPT-enabled 6G systems is presented. Finally, the chapter highlights some key research challenges and opportunities for RIS-SWIPT networks for 6G systems that can be addressed by using DRL methods. © 2025 The Institute of Electrical and Electronics Engineers, Inc.