Distributed Model Predictive Consensus of MASs Against False Data Injection Attacks and Denial-of-Service Attacks

Abstract

This article provides a secure distributed output feedback model predictive control (DOFMPC) solution for the leader-following consensus problems of homogeneous linear disturbed multiagent systems against multiple cyber attacks. The false data injection (FDI) attacks on the sensor-controller communication channel and denial-of-service (DoS) attacks on the controller-actuator communication channel occur simultaneously. To defend against dual-channel multiple attacks, we propose a secure DOFMPC scheme consisting of three modules. Firstly, a robust multivariate observer is built to separate FDI attacks from uncompromised states. Secondly, a distributed output feedback model predictive controller is designed to generate effective control sequences. Thirdly, a buffered actuator is added to realize adequate compensation to defend against DoS attacks. The proposed secure DOFMPC scheme ensures the recursive feasibility of the formulated optimization control problem and achieves the leader-following consensus of the multiagent system. Finally, an illustrative example is presented to demonstrate the secure performance of the proposed DOFMPC scheme.