Quantity Allocation, Trajectory Planning, and Association Optimization for Efficient Relay Transmissions in UAV-Aided Vehicular Networks

Unmanned aerial vehicles (UAVs) have emerged as promising enablers for enhancing vehicular networks, particularly by serving as aerial relays to address challenges of dynamic topology and frequent disconnections in highly mobile environments. This paper proposes a novel UAV-aided relay transmission scheme, where multiple UAVs operate as airborne relays to establish reliable base station-to-UAV (B2U) and UAV-to-vehicle (U2V) communication links. To quantitatively evaluate system performance, we define and derive the upper-bound delay under joint consideration of UAV quantity allocation, trajectory planning, and UAV-vehicle-channel association. We further formulate the upper-bound delay minimization problem and prove it NP-hard. To address the computational complexity, we develop a decoupling framework through constraint relaxation, which divides the problem into two tractable subproblems. The vehicle distribution based UAV horizontal deployment subproblem is solved by a K-means based limited condition vehicular clustering approach, while the mixed altitude deployment-channel association subproblem addressed by a hybrid proximal policy optimization based approach. Simulation results demonstrate that the proposed solution significantly reduces delay and enhances the efficacy of the scheme in UAV-aided vehicular networks.