Optimal Eavesdropping for Cyber-Physical Systems via a Energy Switching Mechanism

This paper proposes optimal attack strategies for an energy-constrained eavesdropper in cyber-physical systems (CPSs). During data transmission from the sensor to the user terminal equipped with the detector, an eavesdropper may attempt to compromise the process by passively monitoring information or actively intercepting information and disrupting the channel. As the energy supply for eavesdropping device may be insufficient in some cases, the eavesdropper needs the use of energy switching to launch attacks in order to improve energy efficiency. Therefore, the intention of this paper is to design the optimal eavesdropping attack strategy under energy constraints. Firstly, a threshold is utilized as the basis for eavesdropper energy switching. We then introduce a Lagrange multiplier to transform the constrained Markov decision process (MDP) into nonconstrained MDP. Finally, this paper presents the sufficient condition for the existence of stationary policies and deduces an optimal deterministic stationary strategy with a threshold structure. A numerical example is provided to illustrate this.