SPARSE ERROR CORRECTION FOR PMU DATA UNDER GPS SPOOFING ATTACKS

Abstract

Time-synchronized phasor measurements from phasor measurement units (PMUs) are valuable for real time monitoring and control. However, their reliance on civilian GPS signals makes them vulnerable to GPS signal spoofing attacks which can be launched by an adversary to falsify PMU data entries. In this paper, we consider the problem of correcting PMU data to mitigate the impact of GPS spoofing attacks. We exploit the unique structure of how GPS spoofing attacks affect PMU data as well as the sparse nature of attacked PMUs. We formulate PMU data correction as a sparse optimization problem that incorporates the unique structure of GPS spoofing attacks. We develop an iterative greedy algorithm to recover tampered PMU measurements and demonstrate its efficacy using simulations with the IEEE RTS-96 test case under both observable and unobservable power network settings.