2020/04/15 by Martin Higgins, Teng Fei, Higgins, Martin +3 · 1 citation
Computer Science · Engineering · #Adversarial Robustness in Machine Learning #FOS: Electrical engineering #Internet Traffic Analysis and Secure E-voting #Smart Grid Security and Resilience #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2004.07004
openalex publication_date 2020/04/15 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
This paper examines how moving target defences (MTD) implemented in power\nsystems can be countered by unsupervised learning-based false data injection\n(FDI) attack and how MTD can be combined with physical watermarking to enhance\nthe system resilience. A novel intelligent attack, which incorporates\ndensity-based spatial clustering and dimensionality reduction, is developed and\nshown to be effective in maintaining stealth in the presence of traditional MTD\nstrategies. In resisting this new type of attack, a novel implementation of MTD\ncombining with physical watermarking is proposed by adding Gaussian watermark\ninto physical plant parameters to drive detection of traditional and\nintelligent FDI attacks, while remaining hidden to the attackers and limiting\nthe impact on system operation and stability.\n