Secure Event-Triggered Consensus Control of Linear Multiagent Systems Subject to Sequential Scaling Attacks

• Published in: IEEE transactions on cybernetics Vol. 52; no. 10; pp. 10314 - 10327
• Main Authors: He, Wangli, Mo, Zekun
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.10.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Additives; Control systems; Denial-of-service attack; Dynamic event-triggered scheme; Eigenvalues and eigenfunctions; Linear systems; Multiagent systems; multiagent systems (MASs); Parameters; Probabilistic logic; Probabilistic models; Protocols; Scaling factors; secure consensus; sequential attack; System performance; Vehicle dynamics
• ISSN: 2168-2267; 2168-2275; 2168-2275
• DOI: 10.1109/TCYB.2021.3070356

This article investigates secure consensus of linear multiagent systems under event-triggered control subject to a scaling deception attack. Different from probabilistic models, a sequential scaling attack is considered, in which specific attack properties, such as the attack duration and frequency, are defined. Moreover, to alleviate the utilization of communication resources, distributed static and dynamic event-triggered control protocols are proposed and analyzed, respectively. This article aims at providing a resilient event-triggered framework to defend a kind of sequential scaling attack by exploring the relationship among the attack duration and frequency, and event-triggered parameters. First, the static event-triggered control is studied, and sufficient consensus conditions are derived, which impose constraints on the attack duration and frequency. Second, a state-based auxiliary variable is introduced in the dynamic event-triggered scheme. Under the proposed dynamic event-triggered control, consensus criteria involving triggering parameters, attack constraints, and system matrices are obtained. It proves that the Zeno behavior can be excluded. Moreover, the impacts of the scaling factor, triggering parameters, and attack properties are discussed. Finally, the effectiveness of the proposed event-triggered control mechanisms is validated by two examples.