Node-Wise Asynchronous Consensus of Switched Multi-Agent Systems With Intermittent Communication Losses

Abstract

This paper studies fixed-gain state-feedback leader-following consensus for switched multi-agent systems, in which the physical mode is common to all agents, whereas each follower implements its feedback gain according to a locally delayed controller-mode signal. The neighbor-based online controller uses only current state-feedback information and the follower’s local implemented mode. Intermittent communication losses, including but not limited to denial-of-service (DoS) attacks, are modeled as per-node binary signals with per-node duration constraints. The stability analysis is formulated as a centralized offline certification problem using a graph-weighted multiple Lyapunov certificate combined with a persistent dwell-time (PDT) partition, node-wise mismatch measures, and per-node loss bounds. The obtained stability criterion reveals tradeoffs among node-wise delays, loss durations, and control parameters. Finally, several simulations of agent dynamics under communication losses confirm the predicted exponential convergence.