Stability Analysis of Cyber-Physical Systems Based on Predictive Control under Denial of Service Attacks with Incomplete Information

We studied the denial of service (DoS) attacks on a cyber-physical system (CPS) under incomplete strategy information and propose stability conditions of the CPS with prediction compensation strategy. Firstly, combining actual conditions with existing related modeling methods, we presented a new model of DoS attacks based on the attack frequency and the duration. Since mutual unknown strategy information exists for both attackers and intrusion detection systems in actual conditions, we exploited the Nash equilibrium game to determine the interactions between intrusion detection systems and the attackers. We approached the problem on Nash equilibrium of the attack and defense game under incomplete strategy information, and, then, we compared complete and incomplete information of the Nash equilibrium solution. Secondly, considering the CPSs under DoS attacks, we determined the stability conditions by applying a predictive compensation strategy with a switched system method. Finally, numerical simulation results showed the correctness of the proposed methods by a ball and beam system.