HMM-Based Secure Control of Multiagent Systems Under DoS Attacks

This note presents an asynchronous distributed active secure control (ADASC) strategy with a safe mode to address the consensus issue for a leader-follower multiagent system (MAS) under denial-of-service (DoS) attacks. The communication network of the leader-follower MAS is exposed to stochastic DoS attacks, which result in different communication topologies. Based on whether the connectivity of the network topology is disrupted, different DoS attack modes are classified into two types: successful DoS attacks and failed DoS attacks. To actively mitigate the impact of DoS attacks on the system, the system is equipped with two modes of operation: normal mode and safe mode. The open-loop divergence rate for each agent running in safe mode is lower than that in normal mode. Upon detecting a successful DoS attack, the system will actively switch to safe mode. Acknowledging the potential for error in DoS attack detection, a hidden Markov model (HMM) is employed to solve the asynchronous problem between the actual DoS attack modes and the estimated DoS attack types. By means of the Lyapunov function approach, some sufficient conditions are obtained to guarantee the mean-square consensus of the leader-follower MAS. Finally, an illustrative example is provided to justify the results of this note.