Secrecy Rate Maximization for Active Reconfigurable Intelligent Surface Assisted MIMO Systems

Reconfigurable intelligent surface (RIS) is a promising technology for future 6G communications and has been used to enhance secrecy performance. However, the performance improvement is restricted by the “double fading” effect of the reflection channel link. To address this issue, we introduce an active RIS design, where the reflecting elements of RIS not only adjust the phase shift but also amplify the reflected signal through the amplifier integrated into its elements. To obtain a satisfactory solution to the non-convex problem resulting from this design, the penalty dual decomposition based alternating gradient projection (PDDAPG) method is proposed. We compare the proposed algorithm with decoupling-fraction-based alternating optimization (DFAO). Specifically, the complexity of the proposed algorithm grows linearly with the number of reflective units in the RIS, while the complexity of the benchmark algorithm increases with the power of 4.5 times the number of reflective units. To further address the quality of service (QoS) constraints regarding the information rate requirements of users, we apply the fractional programming (FP) method and the successive convex approximation (SCA) method to optimize the precoder of the base station (BS) and the active beamformers of the phase shifts. The simulation results have demonstrated the effectiveness of the proposed PDDAPG method. Moreover, the active RIS can effectively mitigate the influence of “double fading” effects and achieve higher energy efficiency (EE) compared to passive RIS.