Deep Reinforcement Learning-Driven Secure ISAC Optimization Using STAR-RIS in 6G Networks

The integration of sensing and communication in next-generation wireless networks—known as Integrated Sensing and Communication (ISAC)—has become a promising approach to meet the demands of 6G networks. This paper investigates the optimization of secure, multi-user ISAC systems enhanced by simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS). To address the complex, dynamic, and high-dimensional nature of the joint optimization problem involving beamforming, sensing, and secure transmission, we propose a Deep Reinforcement Learning (DRL)-based framework. Specifically, a Soft Actor-Critic (SAC) algorithm is employed to adaptively control both active and passive beamforming while ensuring security against eavesdroppers and maximizing sensing accuracy. Simulation results demonstrate that the proposed DRL framework significantly outperforms traditional optimization-based methods in terms of communication throughput, sensing performance, and resilience to eavesdropping. Additionally, the system exhibits strong adaptability to environmental changes and user mobility. This study offers a scalable and intelligent approach for real-time ISAC resource distribution in 6G networks, paving the way for secure, high-efficiency wireless systems.

Keywords: 6G networks, ISAC, STAR-RIS, Deep Reinforcement Learning, Soft Actor-Critic, secure communication, intelligent surfaces, beamforming optimization, multi-user systems, wireless sensing