Fully distributed event-triggered secure consensus of general linear multi-agent systems under sequential scaling attacks

In this article, secure consensus problem is studied for a general linear multi-agent system, where a fully distributed event-triggered scheme is introduced to increase the feasibility and improve the scalability of the control protocol. Firstly, an edge-based sequential scaling attack is formulated with constraints on the attack duration and the frequency, which includes multiplicative deception attacks and DoS attacks as a particular situation when the scaling factor is equal to 0. Then a fully distributed event-triggered control protocol is designed, in which global information is no longer needed. Sufficient conditions related to triggering parameters, the scaling factor, the attack duration and attack frequency are derived ensuring the asymptotic consensus of MAS. Furthermore, the impact of the triggering parameters and the scaling factor on the attack duration and the attack frequency are also discussed. The Zeno behavior is proved not to happen. Finally, two simulation examples based on unmanned intelligent vehicles are conducted to verify the results. • An edge-based sequential scaling attack model is proposed with constraints on the attack frequency and the attack duration, which includes multiplicative deception attacks and DoS attacks as a particular situation when the scaling factor is equal to 0. Compared with existing attack models Xu et al. (2020), Tang et al. (2021), Xu et al. (2019), Cetinkaya et al. (2019), Li et al. (2020), He et al. (2020), Feng et al. (2017), Feng and Hu (2020), He et al. (2018), the sequential scaling attack is more general and more suitable for describing the actual attacks. • A novel fully distributed event-triggered control protocol which can resist attacks is designed. By adding an exponential term to the adaptive coupling weights, sufficient conditions for reaching asymptotic consensus for general linear MASs are obtained. Compared with the other existing fully distributed event-triggered mechanism (Cheng and Li, 2019 [12,13]; Li et al., 2020; Cheng and Li, 2019; Xu et al., 2020; Xu et al., 2022), secure asymptotic consensus is firstly obtained in this paper. • An algorithm for obtaining the feedback gain matrix is proposed, which is only related to the system matrices and the scaling factor. Compared with existing algorithms for obtaining the feedback gain matrix He and Mo (2021), the influence of scaling factor on system stability is obtained without using any global information.