Beamforming Design for Near-Field FIM-Aided ISAC Systems: A DRL Approach

This article explores multi-target detection and multi-user communication in a near-field (NF) integrated sensing and communication (ISAC) system, where a flexible intelligent metasurface (FIM) is deployed at the base station (BS). Specifically, the BS operates in downlink mode to communicate with users and sense information from targets utilizing the FIM. The goal is to design optimal beamforming at the BS to meet both communication and sensing performance requirements. To achieve this goal, a multi-objective optimization problem is formulated, to minimize the Cramér-Rao bound while maximizing the sum rate, by jointly optimizing FIM's surface shape and transmit beamforming. To solve the resulting non-convex optimization problem, we develop a reinforcement learning (RL)-based algorithm. Simulation results demonstrate that employing FIM results in a 21% increase in sum rate.