Abstract:

This paper presents a cross-layer strategy for detecting a variety of potential cyber-attacks on the Smart Grid. While most literature focus on False Data Injection attacks on the measurements taken throughout the Smart Grid, there are many ways in which an attacker can affect power system real-time operation. Namely, an attacker can focus on the communication layer of the Smart Grid, intercepting messages for the sake of snooping, flooding the communication channels, or changing measurement values. This paper acknowledges these different types of attacks and analyzes not only measurement values, but also inter-arrival time, transmission delay, and packet count via machine learning techniques to detect several different types of cyber-attacks on the Smart Grid. When analyzed alongside the measurement data in real-time, the data-driven Cross-Layer Ensemble CorrDet with Adaptive Statistics (CECD-AS) strategy presented in this paper shows vast improvement in detecting cyber-attacks over the traditional physics-based State Estimation as well as the Ensemble CorrDet with Adaptive Statistics strategy that only uses Smart Grid measurement data. The proposed cross-layered framework is validated on the IEEE 118 bus system.

Links:

Citation:

A. Starke, K. Nagaraj, C. Ruben, N. Aljohani, S. Zou, A. Bretas, J. McNair and A. Zare,"Cross-Layered Distributed Data-Driven Framework for Enhanced Smart Grid Cyber-Physical Security." Under Review.

@article{starke2022cross,
title={Cross-layered distributed data-driven framework for enhanced smart grid cyber-physical security},
author={Starke, Allen and Nagaraj, Keerthiraj and Ruben, Cody and Aljohani, Nader and Zou, Sheng and Bretas, Arturo and McNair, Janise and Zare, Alina},
journal={IET Smart Grid},
year={2022},
publisher={Wiley Online Library}
}