UAV-Enabled Adaptive Beamforming for ISAC Vehicular Networks

This paper proposes a novel integrated sensing and communication (ISAC) scheme for unmanned aerial vehicle (UAV)-enabled vehicular networks, where vehicles are considered as extended targets with multiple resolvable scatters. Based on reflected signals, the UAV predicts the vehicle’s state and generates tailored beams to track and communicate with it. To address the asymmetric sensing and communication requirements, we propose a three-stage ISAC scheme with dynamic sensing duration and frequency based on real-time performance. Specifically, the initial vehicle state is estimated in the first stage, followed by using wide beams in the second stage for coverage, employing an extended Kalman filtering approach for vehicular state prediction and tracking. In the third stage, the UAV selectively transmits either an ISAC beam or a communication-only beam based on the monitored sensing/communication performance. Simulation results demonstrate the effectiveness and adaptability of the proposed scheme.