Secure Communication for Spatially Correlated RIS-Aided Multiuser Massive MIMO Systems: Analysis and Optimization

TL;DR

This paper analyzes and optimizes secure communication in RIS-aided multiuser massive MIMO systems with artificial noise, deriving closed-form expressions for ergodic secrecy rate and proposing algorithms for power and phase shift optimization based on statistical CSI.

Contribution

It provides the first closed-form ergodic secrecy rate expression for spatially correlated RIS-aided MIMO systems and develops a gradient-based optimization method using only statistical CSI.

Findings

Spatial correlation at RIS affects secrecy performance.

Optimized RIS phase shifts improve security.

High spatial correlation degrades secrecy performance.

Abstract

This letter investigates the secure communication in a reconfigurable intelligent surface (RIS)-aided multiuser massive multiple-input multiple-output (MIMO) system exploiting artificial noise (AN). We first derive a closed-form expression of the ergodic secrecy rate under spatially correlated MIMO channels. By using this derived result, we further optimize the power fraction of AN in closed form and the RIS phase shifts by developing a gradient-based algorithm, which requires only statistical channel state information (CSI). Our analysis shows that spatial correlation at the RIS provides an additional dimension for optimizing the RIS phase shifts. Numerical simulations validate the analytical results which show the insightful interplay among the system parameters and the degradation of secrecy performance due to high spatial correlation at the RIS.