Security control for two-time-scale networked systems under unreliable Markov networks subject to hybrid attacks

Abstract

This work addresses the security control problem of two-time-scale networked systems operating on unreliable Markov networks under hybrid attacks. Markov jump processes are employed to model the different types of network attacks that the system may encounter, thereby defining three operational modes: normal operation, deception attacks, and scaling attacks. A non-fragile control strategy is proposed to mitigate gain fluctuations resulting from unexpected errors and disturbances, which can enhance the robustness of the designed controller. A Lyapunov function with singular perturbation parameter is constructed to analyze both the system stability and $H_{\infty }$ performance index, where the singular perturbation parameter characterize the dynamics associated with fast and slow time-scale variations, leading to a set of sufficient stability conditions. Finally, the validity of the proposed control method is verified via numerical simulations.