Optimizing Secrecy Energy Efficiency in RIS-assisted MISO systems using Deep Reinforcement Learning

This article investigates the maximization of secrecy energy efficiency (SEE) in B5G mobile systems where a suite of reconfigurable intelligent surface (RIS) modules is incorporated. Taking into account the location information of legitimate users and eavesdroppers, we formulate the problem as a joint optimization of the phase shifts, physical orientations, and locations of the RIS modules, as well as resource allocation at the base station (BS). The problem is then solved by leveraging a deep reinforcement learning (DRL) approach proposed in this paper. The case study results demonstrate the effectiveness of the proposed scheme in improving the secrecy energy efficiency of communication systems using RIS.