Learning-based attacks in Cyber-Physical Systems: Exploration, Detection, and Control Cost trade-offs
This research addresses the critical problem of securing cyber-physical systems against sophisticated learning-based attacks, which is important for system operators and security researchers. The work provides theoretical trade-offs for detection and control costs.
This paper investigates learning-based attacks in linear cyber-physical systems where an attacker learns system dynamics to hijack control signals and deceive the controller. The authors provide tight upper and lower bounds on the expected \u03b5-deception time for the controller and a probabilistic lower bound on the attacker's learning time required for a given deception time. They also establish a lower bound on the energy expenditure needed for detection.
We study the problem of learning-based attacks in linear systems, where the communication channel between the controller and the plant can be hijacked by a malicious attacker. We assume the attacker learns the dynamics of the system from observations, then overrides the controller's actuation signal, while mimicking legitimate operation by providing fictitious sensor readings to the controller. On the other hand, the controller is on a lookout to detect the presence of the attacker and tries to enhance the detection performance by carefully crafting its control signals. We study the trade-offs between the information acquired by the attacker from observations, the detection capabilities of the controller, and the control cost. Specifically, we provide tight upper and lower bounds on the expected $ε$-deception time, namely the time required by the controller to make a decision regarding the presence of an attacker with confidence at least $(1-ε\log(1/ε))$. We then show a probabilistic lower bound on the time that must be spent by the attacker learning the system, in order for the controller to have a given expected $ε$-deception time. We show that this bound is also order optimal, in the sense that if the attacker satisfies it, then there exists a learning algorithm with the given order expected deception time. Finally, we show a lower bound on the expected energy expenditure required to guarantee detection with confidence at least $1-ε\log(1/ε)$.