Active Defense Strategy in Cyber-Physical Systems: Misleading Unauthorized Observers

Securing Cyber-Physical Systems (CPS) is a critical concern, extensively explored in the literature. Traditional approaches involve developing detectors to identify cyber-attacks or implementing passive defense strategies such as encrypting communication channels and using firewalls. Differently, in this paper, we address the challenge of securing CPS against unauthorized observers, which aims to estimate the system states by eavesdropping on system input and output. An active defense strategy, called Misleading Unauthorized Observer (MUO), that modifies input-output signals in order to mislead the unauthorized observer, is introduced. These modifications are additional signals undetectable by innovation or residual-based detectors and designed with the aim of arbitrarily increasing the estimation error at the unauthorized observer side. For this purpose, system properties required for guaranteeing the existence of such a defense strategy are stated. Then the design of these additional signals is formulated as an optimization problem with undetectability constraints. The effectiveness of the proposed approach is validated through simulations of a moving body in the XY plane, demonstrating how the active defending system successfully misleads the unauthorized observer.