AI-based Navigation and Communication Control for a Team of UAVs with Reconfigurable Intelligent Surfaces Supporting Mobile Internet of Vehicles

Unmanned aerial vehicles (UAVs) are employed in wireless communication networks (WCNs) to improve coverage and quality. The applications of UAVs become problematic in the millimeters wave fifth-generation (5G) and beyond in the optical 6G WCNs because of two reasons: 1) higher path loss which means UAVs should fly at lower altitudes to be closer to the user’s equipment; 2) complexities associated with a multi-input multi-output antenna to be incorporated in the UAV as an active aerial base station. We propose equipping UAVs with a (passive) reconfigurable intelligent surface (RIS) to resolve the issues with UAV-enabled wireless communication in 5G/6G. In this paper, the trajectory planning of the RIS-equipped UAV (RISeUAV) that renders aerial LoS service (ALoSS) is elaborated. The ALoSS facilitates vehicle-to-vehicle (V2V) and vehicle-to-everything (V2X) communication in obstructed dense urban environments for Internet-of-vehicles. (IoVs). To handle the nonconvexity and computation hardness of the optimization problem we use AI-based deep reinforcement learning to effectively solve the optimality and time complexity issues. Numerical simulation results assess the efficacy of the proposed method.