Deep Reinforcement Learning Based Energy Consumption Minimization for Intelligent Reflecting Surfaces Assisted D2D Users Underlaying UAV Network

Device-to-device communication (D2D-C) is a leading-edge technique in 5G and forthcoming 6G networks due to its benefits for enhanced spectrum efficiency and energy-efficiency (EE). Despite these potential advantages, co-channel interference, cross-channel interference, and massive connectivity are the major issues in D2D-C. In this study, we explore the uplink RIS-assisted communication system in the presence of device-to-device pairs (D2DPs). Reconfigurable intelligent surface (RIS) is introduced into wireless communication networks with the assistance of an unmanned aerial vehicle (UAV) in order to solve the issues of co-channel and cross-channel interference. In order to estimate the UAV's trajectory and RIS's phase shift while ensuring that every user's data requirement is met, a centralized-declining deep-Q network (C-DDQN) method is proposed. In the C-DDQN method, a central controller acts as an agent. The UAV's trajectory and the RIS's phase shift are controlled by a central controller. Numerical outcomes showed that the suggested scheme overcomes state-of-the-art techniques in terms of results.