Cyber-physical risk modeling and mitigation for the smart grid using a game-theoretic approach

Traditional probabilistic risk assessment approaches do not capture `threats' in the risk modeling. In this paper, we propose a game-theoretic approach for cyber-physical risk modeling and mitigation that allows modeling of `threats' by including attacker behavior, and can be adapted for dynamic attack scenarios. We also introduce a cyber-physical cost modeling framework that captures attacker actions in the cyber layer, attack impacts on the physical layer, and defender actions both in cyber and physical layer. We provide some insights into the benefits of applying game theory for cyber-physical risk modeling and mitigation through a simple, intuitive case study on a 3-bus power system. Finally, we identify some practical challenges and limitations for applying game theory to large systems and conclude the paper with some directions for future work.