Intelligent Reflecting Surface Aided Radar Spoofing under Imperfect Geometric and Channel Knowledge

Abstract

Intelligent reflecting surface (IRS) technology has emerged as a promising paradigm in electronic countermeasures (ECM). By passively manipulating electromagnetic waves, it addresses the hardware and detectability limitations of traditional active jamming to enable reliable radar spoofing. However, as practical airborne deployments operate in dynamic environments, the spoofing performance of existing designs degrades significantly under imperfect geometric and channel knowledge such as sensing uncertainties (e.g., angle and channel estimation errors) and hardware constraints (e.g., phase quantization)-especially when employing massive arrays. To address these critical vulnerabilities, we propose the robust successive convex approximation (R-SCA) algorithm. Our framework integrates principal component analysis (PCA) feature extraction with diagonal loading to ensure mathematical stability, an adaptive Nyquist grid to mitigate constraint leakage, and asymmetric dual-variable weighting to execute a robust peak shift under overlapping angular uncertainty distributions. Furthermore, finite-bit quantization is isolated as a post-processing operation to preserve gradient stability. Simulation results validate that, compared to standard idealized baseline designs, the proposed R-SCA algorithm significantly reduces peak angle deviation and suppresses stealth constraint leakage, thereby sustaining highfidelity radar spoofing and electromagnetic stealth under severe system uncertainties.