Resilient Control of Stochastic Cyber–Physical Systems Against Stealthy Attacks: Complex Dynamical Networks Encryption Strategy

Jun-Lan Wang,Xiao-Jian Li
DOI: https://doi.org/10.1109/tsmc.2024.3493859
2024-01-01
Abstract:This article studies the resilient control issue of the Markovian jump cyber–physical systems (CPSs) under stealthy integrity attacks. In order to enhance the security of the Markovian jumping CPSs, an encryption scheme based on complex dynamical networks (CDNs) is presented. Note that the existing encryption mechanism based on single-node chaotic systems cannot detect attacks when part of the encrypted information is eavesdropped by the attacker. However, the method presented here plays the advantage of complexity of CDNs in the encryption link and can still effectively identify attacks. In addition, the communication delay of data in the communication network is considered. It is shown that compared with the existing results, the encryption scheme proposed in this article does not involve the control link when realizing CDNs synchronization, thus widening the allowable range of delay. Furthermore, the synchronization of drive–response complex chaotic networks guarantees the Markovian jumping CPSs nominal performance without attacks and the stochastic input-to-state stability with attacks. In the end, two examples are given to describe that the proposed security architecture can detect attacks both in theory and simulation.
What problem does this paper attempt to address?