Real-Time Cyber-Physical False Data Attack Detection in Smart Grids Using Neural Networks
Abstract
Attacks on cyber-physical systems have recently increased in frequency, impact, and publicity. In this paper, a cyber-physical false data attack detection mechanism is proposed to protect the operation of power transmission and distribution systems by automatically inferring underlying physical relationships using cross-sensor analytics in order to detect sensor failures, replay attacks, and other data integrity issues in real-time. We investigate a neural network based mechanism acting on voltage and current readings resulting from a wide variety of load conditions on the IEEE 30-bus power system standard and compare its performance with a support vector machine based mechanism. Experiments showed that 99% detection accuracy of replay attacks was achieved using the proposed neural network mechanism. More importantly, we showed that the best approach was to not create physics-based features using what we knew about the system, but rather to use a neural network to automatically learn the laws, and then use the outputs of that to build a classifier to identify whether and where data spoofing occurs. Thus, it is preferable to infer and exploit the physics using a single machine learning solution
- Authors:
-
- ORNL
- Publication Date:
- Research Org.:
- Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
- Sponsoring Org.:
- USDOE National Nuclear Security Administration (NNSA)
- OSTI Identifier:
- 1426580
- DOE Contract Number:
- AC05-00OR22725
- Resource Type:
- Conference
- Resource Relation:
- Conference: Computational Science and Computational Intelligence - Las Vegas, Nevada, United States of America - 12/14/2017 5:00:00 AM-12/16/2017 5:00:00 AM
- Country of Publication:
- United States
- Language:
- English
Citation Formats
Ferragut, Erik M., Laska, Jason A., Olama, Mohammed M., and Ozmen, Ozgur. Real-Time Cyber-Physical False Data Attack Detection in Smart Grids Using Neural Networks. United States: N. p., 2017.
Web. doi:10.1109/CSCI.2017.1.
Ferragut, Erik M., Laska, Jason A., Olama, Mohammed M., & Ozmen, Ozgur. Real-Time Cyber-Physical False Data Attack Detection in Smart Grids Using Neural Networks. United States. https://doi.org/10.1109/CSCI.2017.1
Ferragut, Erik M., Laska, Jason A., Olama, Mohammed M., and Ozmen, Ozgur. 2017.
"Real-Time Cyber-Physical False Data Attack Detection in Smart Grids Using Neural Networks". United States. https://doi.org/10.1109/CSCI.2017.1. https://www.osti.gov/servlets/purl/1426580.
@article{osti_1426580,
title = {Real-Time Cyber-Physical False Data Attack Detection in Smart Grids Using Neural Networks},
author = {Ferragut, Erik M. and Laska, Jason A. and Olama, Mohammed M. and Ozmen, Ozgur},
abstractNote = {Attacks on cyber-physical systems have recently increased in frequency, impact, and publicity. In this paper, a cyber-physical false data attack detection mechanism is proposed to protect the operation of power transmission and distribution systems by automatically inferring underlying physical relationships using cross-sensor analytics in order to detect sensor failures, replay attacks, and other data integrity issues in real-time. We investigate a neural network based mechanism acting on voltage and current readings resulting from a wide variety of load conditions on the IEEE 30-bus power system standard and compare its performance with a support vector machine based mechanism. Experiments showed that 99% detection accuracy of replay attacks was achieved using the proposed neural network mechanism. More importantly, we showed that the best approach was to not create physics-based features using what we knew about the system, but rather to use a neural network to automatically learn the laws, and then use the outputs of that to build a classifier to identify whether and where data spoofing occurs. Thus, it is preferable to infer and exploit the physics using a single machine learning solution},
doi = {10.1109/CSCI.2017.1},
url = {https://www.osti.gov/biblio/1426580},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Fri Dec 01 00:00:00 EST 2017},
month = {Fri Dec 01 00:00:00 EST 2017}
}