Detection and mitigation of GPS Spoofing Attacks on Phasor Measurement Units using deep learning

Abstract

The widespread use of Phasor Measurement Units (PMUs) is considered one of the most important contributions to the quality of power system monitoring. PMUs are digital devices that provide synchronized phasor measurements time-referenced by the Global Positioning System (GPS). The lack of encryption in conventional GPS receivers used in PMUs makes them vulnerable to GPS Spoofing Attacks (GSAs) that cause the GPS receiver to lose track, resulting in a phase shift of all measurements of the affected PMU. Even small phase shifts can downgrade the performance of the state estimator, indicating their ability to propagate throughout the entire system. Therefore, ensuring the integrity of PMU measurements is of utmost importance. State-of-the-art statistical methods are insensitive to small phase shifts, and they become computationally intensive as the system size and the number of simultaneous attacks increase. This paper presents the first application of deep learning to simultaneously detect and mitigate the effects of GSAs against PMUs. The performance of the proposed method is evaluated on IEEE 14-, 57-, and 118-bus systems and compared to state-of-the-art statistical and deep learning methods.