Probability-Based Stealthy Attacks for Cyber-Physical Systems with Resource Constraint

This paper focuses on designing resource-constrained stealthy attacks in linear Cyber-physical systems (CPS). The attacker’s primary objective is to maximize the disruption of the system’s performance while remaining undetected by detectors. In light of attack resource constraints, the attacker must also deliberate on the selection of target sensors to maximize the effectiveness of attacks. This paper presents the corresponding attack strategy, granting the attacker the flexibility to choose the specific sensors. Moreover, we formulate the optimization problem to determine the optimal attack strategy. Leveraging the Schur-Horn theorem, we successfully derive the optimal attack strategy and outline principles for the attacker in selecting sensor channels. Finally, the theoretical findings are validated through simulation experiments.