Joint Beamforming Design and Trajectory Optimization for STAR-RIS Assisted UAV-enabled ISAC System

Recognizing the coverage limitations of traditional reconfigurable intelligent surface (RIS), the simultaneously transmitting and reflecting RIS (STAR-RIS) with the ability to provide coverage on both sides of the surface, has been regarded as a promising technology especially for the integrated indooroutdoor communication environment. In this paper, we investigate a STAR-RIS assisted unmanned aerial vehicle (UAV)-enabled integrated sensing and communication (ISAC) system in the presence of imperfect channel state information (CSI). Specifically, the UAV serves as an aerial base station to provide communication services to outdoor users, as well as positioning for indoor users. To maximize the communication sum-rate while satisfying the sensing beampattern gain requirement, the UAV’s trajectory, beamforming, and the STAR-RIS transmission/reflection coefficients are jointly optimized. Meanwhile, the UAV’s flight safety, the maximum flight duration, and the minimum communication rate for each user are also considered. An online decision-making framework that utilizes deep reinforcement learning (DRL) is proposed to address the sum-rate maximization problem with imperfect CSI. Furthermore, in order to decouple the continuous optimization variables and improve the applicability of the system, we introduce a twin-twin-delayed deep deterministic policy gradient algorithm based on self-attention mechanism (SA-TTD3), which utilizes dual agents to decouple and optimize variables, and incorporates attention mechanisms to improve applicability. The numerical results indicate that the proposed SA-TTD3 algorithm significantly improves the performance of the ISAC system compared with the baseline schemes.