IEEE Transactions on Wireless Communications
Secrecy Capacity and Achievable Secrecy Rate of ISAC Systems Over Multi-Antenna Wiretap Channels Under Sensing SCNR Constraint
Zengbao Zhu, Qimei Cui, Yuming Huang, Xiaofang Zhao, Guoshun Nan, Na Li, Xiaofeng Tao, Ping Zhang
Abstract
This paper investigates the fundamental secrecy capacity and achievable secrecy rate of multi-antenna integrated sensing and communication (ISAC) systems. A multi-antenna transmitter simultaneously delivers confidential messages to a single-antenna communication receiver and emits signals for target detection by a multi-antenna sensing receiver in signal-dependent clutter. The signal-to-clutter-plus-noise ratio (SCNR) is adopted as a general sensing performance metric. We model the system as a multi-antenna wiretap channel with the channel input subject to both SCNR and total power constraints. Secrecy capacity is derived based on the generalized secrecy capacity under the compact input covariance constraint and Stewart’s perturbation theory. Efficient iterative methods are developed to maximize the secrecy capacity and the achievable secrecy rate of the unified sensing and communication (S&C) signals. We prove for the first time that a dual-functional sensing-communication signal superposed with a deterministic sensing signal achieves capacity, where the communication receiver entities employ interference cancellation to eliminate interference from the deterministic sensing signal. Numerical results validate the effectiveness of our methods and analysis. The capacity-achieving signaling improves the secrecy rate by 10.47%, 34.55%, and 51.74%, compared to the DFAN scheme, the NOMA-inspired signaling, and the unified S&C signaling, respectively.