Node-to-Node Fault-Tolerant Control of Layered Multi-Agent Systems Under Deception Attack

The node-to-node bipartite consensus problem of multi-agent systems with a two-layer network consisting of general linear dynamics, is investigated. A novel observer technique composed of two state and fault observers is introduced in the presence of deception attack and sensor faults. Then, an observer-based fault-tolerant control framework is proposed to deal with joint effects of sensor faults and deception attack obeying the Bernoulli distribution launched from malicious adversaries. By exploiting matrix analysis and Lyapunov stability theory, sufficient conditions for achieving node-to-node mean-square bounded bipartite consensus are obtained. Numerical examples illustrate the effectiveness of the proposed approach.