False Data Injection Attacks Against Networked Predictive Output Tracking Control Systems

Abstract

This chapter discusses the design problem of stealthy FDI attacks against NCSs from the viewpoint of malicious attackers. A Kalman filter-based networked predictive output tracking control method is proposed for networked MIMO stochastic linear systems, where random network-induced delays, packet disorders, and packet dropouts in the feedback and forward channels are considered. To destroy the stability of the resulting closed-loop networked system without being detected by a detector based on an output residual, stealthy FDI attacks are designed against the measurement data in the feedback channel and the control commands in the forward channel, respectively, and the corresponding attack conditions are derived. Both simulation and experimental results are given to show the effectiveness of the proposed methods.