Secure ISAC Systems Empowered by Compound Reconfigurable Antenna Arrays

Abstract: In integrated sensing and communication (ISAC) systems, the use of dual-functional signals inherently exposes confidential communication information during radar sensing, particularly when the sensing target itself acts as an eavesdropper. Conventional physical-layer security solutions rely on directional beamforming or artificial noise injection implemented via signal processing in the baseband (BB) domain. However, these BB-domain approaches are constrained by insufficient spatial resolution and the adverse effects of spatially correlated channels. To overcome these limitations, this paper proposes a novel secure ISAC framework empowered by compound reconfigurable antenna (CRA) arrays, which offer simultaneous reconfigurability of radiation patterns and polarization states in the electromagnetic (EM) domain. Specifically, we develop a comprehensive channel model incorporating virtual angular domain, spatial domain, and depolarization effects, and formulate a mixed-integer nonlinear programming (MINLP) problem to jointly design EM-domain and BB-domain precoders and combiners. To efficiently solve this complex optimization problem, we propose an iterative decomposition-based algorithm leveraging fractional programming (FP), majorization-minimization (MM), second-order cone programming (SOCP), and penalty methods. Extensive simulation results demonstrate that the CRA array architecture with proposed joint EM-and BB-domain design achieves significant performance improvements in secure ISAC systems. In particular, radar sensing gains of up to 12dB are observed over conventional beamforming, while robust communication security is maintained. These results highlight the considerable benefits attainable by jointly leveraging additional degrees of freedom (DoFs) in the EM domain for secure ISAC system design.