Security control of multi-agent systems under false data injection attacks

This paper studies the consensus problem of multi-agent systems (MASs) with false data injection (FDI) attacks and noises. Based on the graphical method and Kalman filter framework, distributed state estimators are designed for the agents to estimate the states of neighbors. A threshold comparison approach is proposed to obtain the secure states. Combining the state estimation approach with the threshold scheme, security mechanism is put forward to ensure the mean square consensus of MASs subject to FDI attacks. Finally, numerical simulations are provided to illustrate the theoretical results.