Unmanned aerial vehicles (UAVs) aid vehicular networks (V2X) in ensuring efficient edge computation and communication. IRS cited its ability to intelligently regulate wireless channels and low-cost advantages as potential technologies to improve performance in UAV-aided V2X networks. In this paper, a reparameterization-based reinforcement learning scheme is designed jointly to minimize the system energy consumption for continuous UAV trajectory optimization, beam design, edge computation resource allocation, and discrete UAV-vehicle-IRS association. Simulation results verified the efficiency and properties of the proposed learning-based scheme and outlined the advantages of IRS.
Intelligent Edge Computation and Trajectory Optimization in IRS-Enhanced UAV-Aided Vehicular Wireless Networks
2024-10-25
599462 byte
Conference paper
Electronic Resource
English