Interference Exploitation in ISAC Systems: Secure Precoding Schemes Against Smart Data Eavesdroppers

Abstract

This paper investigates the data security enhancement of interference-exploitation precoding schemes in integrated sensing and communication (ISAC) systems, where the sensing target may act as a potential data eavesdropper. We first analyze the data security risks of constructive interference (CI)-based ISAC precoding schemes and existing destructive interference (DI)-based secure ISAC precoding schemes under the worst-case security scenario, where the data eavesdropper and the legitimate communication user share the same azimuth angle. Specifically, we show that the existing DI-based secure precoding schemes render the distribution of the signals received by the eavesdropper predictable. By exploiting this distribution, a smart data eavesdropper can employ effective symbol detection methods to significantly reduce its symbol error rate (SER), thereby posing severe data security risks. To this end, we propose a randomized DI-based secure ISAC precoding scheme and introduce a refined version to further enhance data security. The optimization problems of the aforementioned schemes are formulated in a unified framework, and the dual problem is derived to provide further insights. Furthermore, we analyze the relationship between the angular offset of the eavesdropper’s azimuth and that of its received signal, and we extend secure ISAC precoding schemes from the case with precisely known target location to that with uncertain target location. Moreover, a rotational randomized DI-based scheme is proposed for the uncertain target location scenario to enhance the radar illumination power while guaranteeing system security. Simulation results validate both the ISAC performance and the data security of the proposed schemes.