A data-driven cyber-physical detection and defense strategy against data integrity attacks in smart grid systems

I consider a cyber-physical perspective to the problem of detecting and defending against data integrity attacks in smart grid systems. In this paper, the problem of transient stability with distributed control using real-time data from geographically distributed Phasor Measurement Units (PMUs) is studied via a flocking-based modeling paradigm. I demonstrate how cyber corruption can be identified through the effective use of telltale physical couplings within the power system. I also develop a data-driven real-time cyber-physical detection and defense strategy with distributed control using real-time data from PMUs. The proposed strategy leverages the physical coherence to probe and detect PMU data corruption and estimate the true information values for attack mitigation.