Distributed Dual-Terminal Adaptive Triggered Consensus Tracking of Descriptor Multi-Agent Systems Against Asynchronous DoS Attacks With Markov Switching Topologies
J.J. Zhu, Guangming Zhuang, Xiangpeng Xie, Jianwei Xia
Abstract
This article investigates the issue of distributed dual-terminal adaptive event-triggered admissible consensus tracking for linear leader-following descriptor multi-agent systems (DMASs) against asynchronous denial-of-service (DoS) attacks. Due to the various off/on moments and durations of consecutive asynchronous DoS attacks on different edges, the communication topologies of DMASs are considered to be randomly switched according to different types of DoS attacks, and the switched topologies are governed by Markov chains. To mitigate dispensable information transmission, a novel dual-terminal adaptive event-triggered protocol is put forward, which excludes the Zeno behavior in addition to flexibly adjusting the inter-event interval. Then, a local estimator based on state reconstruction is designed in a triggered way to exactly estimate agents’ states between two successive triggered moments. Additionally, by constructing a modified mode-dependent Lyapunov-Krasovskii functional and employing singular value decomposition technique, the frequencies and durations of DoS attacks are analyzed and the admissible consensus tracking conditions are acquired for DMASs when frequencies and durations of DoS attacks satisfy bounded conditions. Lastly, multiple transistor electrical circuit systems are employed to demonstrate the validity of the proposed strategy.