Cramér–Rao Bound Optimization With Security Constraints in IRS-Enabled MU-ISAC Systems

In recent years, theoretical research on intelligent reflecting surfaces (IRS)-enabled Integrated Sensing and Communication (ISAC) systems has developed rapidly. Since the sensing and communication signals are superimposed, a sensing target could be a potential eavesdropper. Therefore, the physical layer security (PLS) of IRS-enabled ISAC systems has been studied. In this paper, we propose an optimization problem that minimizes the Cramér-Rao Bound (CRB) of the sensing target angle estimation in the multi-user, single sensing target scenario, where both the sensing and communication links are blocked. The transmit beamforming, the IRS phase shift matrix, and the sensing signals covariance are jointly optimized. The constraints include the minimum Signal-to-Interference-plus-Noise-Ratio (SINR) for communication users (CUs), the maximum SINR for the potential eavesdropper, the maximum transmit power, and the modulus-1 constraint on the IRS elements. To solve this non-convex problem, an alternating optimization (AO) framework is developed. Specially, we propose a Riemannian manifold algorithm to optimize the IRS phase shifts and semi-definite programming (SDP) to optimize both the transmit beamforming and the sensing signals. Simulation results demonstrate the effectiveness and convergence of the proposed AO framework, with comparisons to four baselines highlighting the advantages of optimizing IRS phases and communication beamforming, as well as the importance of incorporating security constraints in ISAC systems.