Resilient consensus for multi-agent systems under distributed denial-of-service attacks: a graph-based approach
Lang Zou, Xiangbin Liu, Xiaoyu Zhang, Hongye Su · Journal of Control and Decision · 2024
This paper considers the leader-following consensus problem of general linear multi-agent systems (MASs) under high-intensity distributed denial-of-service (DDoS) attacks, which are full-coverage on the communication network from the perspective of the whole system. Motivated by a simulation example of double-integrator MASs, a distributed leader-following consensus scheme based on directed acyclic graph (DAG) is proposed to improve the system robustness against DDoS attacks. Through Lyapunov stability analysis, it is proved that the consensus of the MASs with DAG can be achieved under DDoS attacks of nearly arbitrary intensity, as long as there is no channel completely blocked by the attacks. The proposed consensus scheme only needs simple requirement for the directed acyclic topology graph of the MASs, making it easy to implement in practice. Finally, a numerical simulation of multi-robot systems under DDoS attacks is carried out to illustrate the effectiveness and superiority of the proposed DAG-based consensus scheme.