Integrated Sensing and Communication Task Optimization for Cellular-Connected UAV

We investigate an Integrated Sensing and Communication (ISAC) system, where the ground base stations (BSs) provide communication and sensing services to the Uncrewed Aerial Vehicle (UAV). A joint optimization problem is proposed, which aims to optimize the UAV flight trajectory, the ground BSs beamforming angle and the time slot allocation strategy simultaneously, and minimize the completion time of the UAV while maximizing the information rate under the limitation of the average radar estimation rate. We propose a hybrid Dueling Deep Q Network and Twin Delayed Deep Deterministic Policy Gradient (D2QN-TD3) algorithm based on multi-step learning to solve the optimization problem. Simulation results are provided to validate the effectiveness of the proposed algorithm.