Improving detection performance in DFRC systems using symbol-level precoding and IRS with continuous phase shifters

This paper investigates the security issues related to symbol-level precoding (SLP) in dual-function radar-communication (DFRC) systems employing intelligent reflective surfaces (IRS) with continuous phase shifters (CPS). It focuses on the joint optimization of the DFRC base station's (BS's) receive beamformer, the IRS's phase shifts, and the transmit waveform. The objective of the optimization is to maximize the radar received signal-to-interference-plus-noise ratio (SINR) for eavesdroppers (Eves), while adhering to constraints on BS transmission power, ensuring constructive interference (CI) to legitimate users (LUs) and destructive interference (DI) to Eves. The challenge arises from multiple coupled optimization variables and a non-convex objective function. To tackle this problem, we propose an alternating optimization algorithm that integrates Minimum Variance Distortionless Response (MVDR) beamforming and Fractional Programming (FP) techniques to achieve a suboptimal solution. Simulation results indicate that our proposed algorithm significantly enhances radar detection capabilities and communication security. Compared to IRS equipped with discrete phase shifters, an IRS with CPS shows superior detection performance. Although SLP has lower computational efficiency than traditional block-level precoding (BLP) methods, it provides improved security rate performance. These findings offer substantial support for the design and application of practical systems.