Secure-IRS: Defending Against Adversarial Physical-Layer Sensing in ISAC System

Ziyu Chen, Denny Vishnu Puri Landika, Alvin Yang, Yizhu Wen, Haofan Cai, Yao Zheng, Hanqing Guo

Abstract: Integrated Sensing and Communication (ISAC) systems have emerged as a key technology in 5G and 6G networks, enabling the simultaneous use of sensing and communication within the same frequency band. A major advancement in this field is the incorporation of Intelligent Reflecting Surfaces (IRS), which enhances signal coverage and strength by adjusting the reflection angle and strength of beamforming signals. While IRS technology improves wireless propagation environments and sensing accuracy, it also introduces new security challenges, particularly in adversarial wireless sensing scenarios. In this paper, we propose a novel countermeasure against adversarial sensing on the physical layer by introducing a randomized phase increment in the IRS placement, disrupting sensing accuracy. We systematically design this countermeasure, provide theoretical validation, and conduct real-world experiments with 8 groups of settings and 80 trials to demonstrate its effectiveness. Our results show that our countermeasure can greatly reduce privacy leakage by reducing 100% attack success rate, making the adversary to obtain the real indicator of the user in the ISAC scenario.

External IDs:dblp:conf/iccnc/ChenLYWCZG25