Operational moving target defences for improved power system cyber-physical security
File(s)
Author(s)
Higgins, Martin
Type
Thesis
Abstract
In this work, we examine how Moving Target Defences (MTDs) can be enhanced to circumvent intelligent false data injection (FDI) attacks against power systems. Initially, we show how, by implementing state-of-the-art topology learning techniques, we can commit full-knowledge-equivalent FDI attacks against static power systems with no prior system knowledge. We go on to explore how naive applications of topology change, as MTDs, can be countered by unsupervised learning-based FDI attacks and how MTDs can be combined with physical watermarking to enhance system resilience. A novel intelligent attack, which incorporates dimensionality reduction and density-based spatial clustering, is developed and shown to be effective in maintaining stealth in the presence of traditional MTD strategies. In resisting this new type of attack, a novel implementation of MTD is suggested. The implementation uses physical watermarking to drive detection of traditional and intelligent FDI attacks while remaining hidden to the attackers. Following this, we outline a cyber-physical authentication strategy for use against FDI attacks. An event-triggered MTD protocol is proposed at the physical layer to complement cyber-side enhancements. This protocol applies a distributed anomaly detection scheme based on Holt-Winters seasonal forecasting in combination with MTD implemented via inductance perturbation. To conclude, we developed a cyber-physical risk assessment framework for FDI attacks. Our assessment criteria combines a weighted graph model of the networks cyber vulnerabilities with a centralised residual-based assessment of the physical system with respect to MTD. This combined approach provides a cyber-physical assessment of FDI attacks which incorporates both the likelihood of intrusion and the prospect of an attacker making stealthy change once intruded.
Version
Open Access
Date Issued
2021-10
Date Awarded
2022-03
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Teng, Fei
Parisini, Thomas
Sponsor
Engineering and Physical Sciences Research Council
Grant Number
EP/L015471/1
Publisher Department
Electrical and Electronic Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)