Cyber-Attacks on Remote State Estimation in Industrial Control System:A Game-Based Framework

摘要

Recently, the security of Industrial Control Systems (ICSs) is widely concerned all over the world. In this paper, the security issues in remote state estimation process of ICSs will be considered. The communication channel between sensor nodes and the remote state estimator may be maliciously interfaced and manipulated by the internal or external attacker. With resources constraints for both the sensor and the attacker side, the interactive decision making process of whether to send or receive data packets or not for estimation process and whether to launch an attack on some data packets or not for an attacker are studied in this paper. A game theory based framework is formulated in the paper and it has been proved that a Nash equilibrium of the final pay-off arbitration game is existed. For the practical computation convenience, an on-line updating algorithm is proposed. What's more, the simulation of the game-based framework described in this paper is demonstrated to verify the validity and efficiency of this framework. The experimental results have shown that the game-based framework could improve performance of the decision making and estimation process and mitigate the impact of the attack. It may provide a novel and feasible approach to protect the state estimation process and improve the intrusion tolerance in ICSs.