Event-Based Bipartite Consensus of Multi-Agent Systems Subject to DoS Attacks
Mengping Xing, Jianquan Lu, Yang Liu, Xiangyong Chen
Abstract
This paper investigates the event-triggered bipartite consensus control of multi-agent systems subject to DoS attack. Existing results of relevant issues are usually confined by an implicit assumption that the real-time information of DoS attack is available for the construction of triggering strategy. However, some practical systems may not be equipped with elaborated attack detection mechanism, which means the trigger may not be notified the emission and suspension instants of attack timely. To settle with this scenario, we develop a novel distributed hybrid event-driven strategy which schedules the ordinary and attack-affected triggering laws by judging whether the previous transmission attempt of corresponding follower is successful, rather than utilizing detailed attack information directly. Furthermore, taking antagonistic interactions among agents into consideration, a resilient event-based control scheme is developed over signed digraph. Then, based on the analysis of hybrid Lyapunov functions exhibiting discontinuous property at practical triggering instants, a set of sufficient conditions which can ensure the exponential leader-following bipartite consensus of the investigated systems are presented. We also prove that Zeno behavior is excluded for triggering attempt sequences. Besides, the developed results are further extended to a self-triggered case. Finally, simulation results are provided to verify the effectiveness of the proposed control scheme.