An Improved Protocol to Consensus of Delayed MASs with UNMS and Aperiodic DoS Cyber-Attacks

Abstract

The present research focuses on the issue of leader-following consensus for multi-agent systems (MASs) suffered from uncertain nonhomogeneous Markov switching (UNMS), time-varying delay, and denial of service (DoS) cyber-attacks. Firstly, in contrast with the existing results on MASs, the communication topology is governed by a UNMS jump process in which the transition rates (TRs) of UNMS may be partially known or completely unknown. Secondly, the changes of communication topologies due to frequently aperiodic DoS cyber attacks are taken into consideration, which may destroy the chains of communication and lead to network paralysis in MASs. By introducing the concepts of average DoS ratio, novel mean-square leader-following consensus conditions which establish the relationship between control gains and sleeping, and active periods of DoS signal for MASs with the UNMS are proposed. It shows that there is no direct restriction on the time-varying delay of the system and the intervals between the sleeping period and the active period of the DoS jamming signal. Finally, the present theoretical result is validated by a numerical example.