‘Input disturbance’-based distributed event-triggered secure control for MAS under DoS attacks: Switching mode strategy

Abstract

This paper investigates the event-triggered consensus control of deterministic nonlinear multi-agent systems (MASs) within the context of Denial-of-Service (DoS) attacks. In such adversarial scenarios, malicious attackers possess the potential to pilfer sensitive data by effectively distracting the security apparatus via DoS attacks. Although the consensus problem of MASs under DoS attacks has been the subject of extensive academic scrutiny, previous research endeavors predominantly presume that the overarching aim of DoS attacker is to precipitate a system-wide collapse. Nevertheless, little research has been dedicated to the scenario where attackers might be motivated by the objective of data extraction. To overcome this obstacle, this paper proposes a distributed event-triggered control mechanism predicated on the’input-disturbance’ methodology for well-defined nonlinear MASs, further employing a switching-mode strategy to enhance the flexibility of the mechanism. Moreover, in cognizance of the latency inherent in the system’s response dynamics, the proposed mechanism accommodates a bounded time-delay during mode-transition upon the detection of the initiation or cessation of a DoS attack. Through the application of Lyapunov methodologies and matrix theoretic constructs, a suite of sufficient conditions is deduced to ensure the consensus of MASs suffered from DoS attacks. Ultimately, a comprehensive simulation example and comparative analysis are presented to verify the validity and efficiency of the proposed mechanism.