Constant Modulus IRS-Assisted Secure Beamforming for Massive MIMO

The physical layer (PHY) security is a key technology for constant modulus (CM) massive MIMO systems. In this paper, we consider using an intelligent reflecting surface (IRS) to provide the system with additional degrees of freedom, thus enhancing its performance. We approach the problem from an optimization perspective, aiming to maximize the security rate. The optimization problem presents significant challenges due to highly coupled variables, non-convex constraints and the non-concave objective function. To tackle these challenges, we propose an efficient alternating optimization (AO) method. This method guarantees convergence to a Karush-Kuhn-Tucker (KKT) point and separately optimizes the CM beamforming and phase shift coefficients at the IRS as two sub-problems. Both sub-problems are solved using similar frameworks of the Dinkelbach method and the alternating direction method of multipliers (ADMM). Simulation results show that the proposed method outperforms many other related methods.