Distributed Optimization for Heterogenous Multi-Agent Systems Under DoS Attacks: Resilient Double-Layer Sampling Filtering Framework

This paper investigates the distributed optimization for heterogenous multi-agent systems (MASs) under zero-topology Denial-of-Service (DoS) attacks, where the attacks are launched within the communication channels. A novel filtering attack framework is proposed based on the periodic sampling and event-triggering mechanism to counteract the impact of the attacks. Under this framework, the consumption of communication resources is reduced, and the minimum lower boundary of triggering time can be obtained to avoid the occurrence of Zeno behavior. Moreover, the relationship between the equilibrium point and optimal solution of heterogenous MASs under DoS attacks is revealed by employing a proportional-integral strategy. Furthermore, constructing an auxiliary system guarantees the output consensus and global exponential convergence of the MAS. Finally, a simulation example is provided to show that the proposed approach has a faster convergence rate and better robust performance against DoS attacks.