Investigation of control theoretic cyber attacks on controllers

Abstract

National critical infrastructures like power plants, power grids, industrial process plants, water distribution systems employ a hierarchy of controllers designed to control the physical/chemical processes safely and optimally. They deploy sophisticated control algorithms implemented in software and exchange control commands and information over networks. Various researches have examined attack scenarios in such embedded control systems from control theoretic perspectives. This paper revisits the theoretical aspects of these attacks and postulates that such attacks can be detected by statistical techniques like sequential probability ratio test SPRT, cumulative sum CUSUM, multiple model Kalman filter etc. The proposed techniques are studied closely for their effectiveness by extensive simulations.