Joint Trajectory Planning and Power Control for Energy Efficiency Optimization in UAV-Assisted IoV Networks

The Internet of Vehicles (IoV) is significantly challenged by network congestion in dense urban areas and coverage gaps in remote regions, especially under the growing demands of autonomous driving and smart mobility. Unmanned aerial vehicles (UAVs) offer a promising solution as aerial base stations under flexible and rapid deployment. However, the UAV-assisted IoV systems face critical bottlenecks such as complex connection with vehicles, resource management in dynamic networks, and limited adaptability to real-time traffic fluctuation. In view of these issues, we propose an adaptive strategy for UAV trajectory planning and power control in densely distributed vehicular networks. We formulate a joint optimization problem that maximizes the energy efficiency subject to the constraints on UAV velocity, transmission power, and data rate requirement. To reduce the solution complexity, the vehicles are divided into multiple clusters via the Gaussian mixture model (GMM), and the UAVs are pre-deployed at cluster centers and associated with the vehicles of the corresponding clusters. Subsequently, a multi-agent twin delayed deep deterministic policy gradient (MATD3) algorithm is employed to design the UAV trajectory and power control, enabling decentralized and cooperative decision-making of the UAVs. Extensive simulations demonstrate that the proposed approach outperforms the baseline schemes in terms of communication rate, energy efficiency, and training stability.