Minimizing Age of Synchronization for Aerial RIS-Assisted Data Collection with Wireless Power Transfer: A Deep Reinforcement Learning Approach

This paper investigates an age of synchronization (AoS) minimization framework in aerial reconfigurable intelligent surface (RIS)-assisted data collection, where a unmanned aerial vehicle (UAV) equipped with RIS assists the communication between base station and Internet of things devices (IoTDs), while the other UAV charges IoTDs with wireless power transfer technology. An optimization problem is formulated to jointly optimize UAVs’ trajectories, RIS phase-shift, device scheduling, and charging power allocation, with the objective of minimizing the system’s average AoS. To solve the dynamic, high-dimensional, and non-convex optimization problem, it is firstly reformulated as a Markov decision process, and then a deep reinforcement learning framework based on multi-agent deep deterministic policy gradient (MADDPG) is proposed. Finally, simulation results demonstrate the proposed framework’s effectiveness and reliability, verifying that it outperforms the benchmark schemes in terms of the average AoS.