Secure Cell-free ISAC Systems with Limited Fronthaul Capacity: Centralized versus Distributed Processing

This paper investigates secure cell-free integrated sensing and communication (ISAC) systems with limited-capacity fronthaul. To address the multi-static sensing challenges under fronthaul constraints, two processing architectures, namely centralized and distributed, are proposed and analyzed. For both architectures, the maximum a-posteriori ratio test (MAPRT) detection statistics and their distributions are derived, where the sensing signals are also served as artificial noise (AN) for interfering with potential eavesdroppers. In the centralized case, the central processing unit (CPU) jointly designs beamforming to maximize the detection probability under security constraints; however, since the sensing echoes must be transmitted to CPU via fronthaul links, the performance is limited by the fronthaul capacity. In contrast, the distributed approach performs local target detection and secure beamforming design at each AP, which avoids fronthaul limitations. Simulation results reveal the respective operating regions of the centralized and distributed approaches, providing insights into future design of secure cell-free ISAC systems under fronthaul limitations.