IRS-Assisted Secure Radar Communication Systems With Malicious Target

The article studies the intelligent reflecting surfaces (IRS) assisted secure radar communication systems where dual-functional radar-communication base station (BS) communicates with multiple legitimate users and senses a malicious radar target. Concretely, three optimization problems, eavesdropper achievable rate minimization, minimum communication signal-to-interference-plus-noise ratio maximization, and communication sum-rate maximization, are considered, where the optimization variables include beamforming vector and artificial noise covariance at the BS and phase shift matrix at the IRS. Due to the coupling of these optimization variables, all the three optimization problems are non-convex. To address these non-convex problems, we first reformulate them by the fractional transformation and Lagrangian dual transformation, and then utilize the alternating optimization (AO) framework to optimize the variables alternately. Moreover, some advanced techniques are utilized to solve the resulted AO sub-problems, e.g., the penalty-based method to solve the rank-one constraint and the Riemannian conjugate gradient to design phase shift matrix. Finally, these proposed optimizations are verified by computer simulations.