Intermittent Event-Triggered Predefined-Time Resilient Consensus for Second-Order MASs Under Asynchronous DoS Attacks

Abstract

This paper investigates the problem of intermittent event-triggered and self-triggered practical predefined-time resilient consensus for second-order multi-agent systems (MASs) under asynchronous denial-of-service (DoS) attacks. Unlike the existing studies that typically assume synchronized attacks affecting all communication links, a more general scenario is considered where DoS attacks independently target controller-actuator channels and communication channels, with each channel experiencing distinct attack patterns. To tackle the challenges posed by such attacks, we introduce the concept of an average effective attack rate and adopt an aperiodical intermittent control framework. Based on this framework, we further develop enhanced intermittent dynamic event-triggered control and intermittent self-triggered control mechanisms for second-order MASs, which aim to optimize communication efficiency while preserving secure consensus under asynchronous dual-channel DoS attacks. Moreover, a new bounded time-varying function is proposed for designing predefined-time controllers, ensuring that the settling time is independent of control parameters, initial states, and topology information. Finally, an example of single-link manipulators is considered to demonstrate the effectiveness of the proposed algorithm. Note to Practitioners—This work is motivated by the coordination control of single-link manipulators, yet is applicable to general second-order resource-constrained MASs subject to cyber threats. Most cooperative tasks rely on communication networks, which inevitably renders the system vulnerable to DoS attacks. Unlike existing works that assume synchronized DoS attacks over all links, our work considers a more challenging scenario where each individual channel, regardless of whether it is a controller-actuator channel or a communication channel, experiences a unique and mutually asynchronous attack pattern. To counter such asynchronous attacks, we propose an intermittent event-triggered control scheme that achieves secure consensus while reducing controller update frequency. An intermittent self-triggered control mechanism is further developed that avoids the need for continuous monitoring, wherein the next triggering instant is computed based on the current state information during control intervals. Additionally, considering the practical demand for coordination efficiency, we design an intermittent predefined-time controller that guarantees practical predefined-time consensus under asynchronous DoS attacks, with a convergence time upper bound that is independent of global information. The proposed algorithm has several potential applications, including intelligent transportation systems, and smart grids.