Joint Constellation Shaping and Waveform Reconstruction for ISRJ Suppression in ISAC Systems

Abstract

To mitigate interrupted sampling repeater jamming (ISRJ) in integrated sensing and communications (ISAC) systems, this letter proposes a constellation shaping method to enhance the reliability of radar waveform reconstruction under jamming environments. By analyzing the maximum a posteriori (MAP)-based reconstruction process, we theoretically derive an exponential Bhattacharyya upper bound that characterizes the reconstruction reliability. A joint geometric-probabilistic constellation shaping (GPCS) optimization problem is formulated to maximize the achievable information rate (AIR) while ensuring a minimum pairwise Bhattacharyya distance to enhance anti-jamming robustness. The resulting non-convex optimization problem is efficiently solved via an alternating optimization (AO) framework combining semidefinite relaxation (SDR) and quasi-Newton methods. Simulation results demonstrate that the proposed GPCS scheme significantly improves waveform reconstruction reliability under ISRJ, suppresses false targets, and enhances radar detection performance.