Active RIS-Assisted Secure Wireless Communication Based on DRL

Since the Reconfigurable Intelligent Surface (RIS) can precisely regulate electromagnetic waves, it is widely used in the field of physical layer security to enhance the security of transmission. However, in contrast to active RIS, passive RIS can only attain a limited secrecy benefit due to the "multiplicative fading" effect in the reflecting channel link between the transmitter and receiver. In this paper, a secure communication scheme is proposed for active RIS-assisted wireless networks. The problem is formulated by jointly optimizing the phase shift of the active RIS and the transmission beamforming vectors of the base station (BS). Given the complexity of the problem, the optimization of secure beamforming is cast as a deep reinforcement learning (DRL) problem and the optimal solution is sought using the Deep Deterministic Policy Gradient (DDPG) algorithm. The simulation results indicate that the active RIS demonstrates a substantial enhancement in the system's secrecy rate (SR) and alleviates the impact of the "multiplicative fading" phenomenon.