DeeBBAA: A Benchmark Deep Black-Box Adversarial Attack Against Cyber-Physical Power Systems

Cyber-physical infrastructure faces threats from evasive false data injection attacks that can significantly impact their security and performance. Adversarial attacks are a popular evasive false data injection threat model that generally targets AI/ML models employed in cyber-physical systems (CPSs...

Full description

Saved in:
Bibliographic Details
Published in:IEEE internet of things journal 2024-01, Vol.11 (24), p.40670-40688
Main Authors: Bhattacharjee, Arnab, Bai, Guangdong, Tushar, Wayes, Verma, Ashu, Mishra, Sukumar, Saha, Tapan K.
Format: Article
Language:English
Subjects:
AC state estimation
adversarial attack against
Black boxes
Cyber-physical systems
cyber-security
Cyberattack
Data collection
Data models
deep learning
Detectors
Infrastructure
Machine learning
Power measurement
Power systems
regression
State estimation
stealthy false data injection attack (SFDIA)
Topology
Voltage measurement
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Cyber-physical infrastructure faces threats from evasive false data injection attacks that can significantly impact their security and performance. Adversarial attacks are a popular evasive false data injection threat model that generally targets AI/ML models employed in cyber-physical systems (CPSs). However, AI/ML models are often not necessary to carry out several key functions of critical CPSs. In this study, we explore the potential for designing effective adversarial attacks targeting critical functions within cyber-physical infrastructure, that do not rely on AI systems, while retaining their adversarial nature against related black-box AI and non-AI functionalities. Specifically, we introduce an evasive deep black-box adversarial attack (DeeBBAA) designed to disrupt the nonlinear state estimation process of an unidentified power network. The attack is engineered to operate under practical constraints, without prior knowledge of the network's topology, and by targeting fewer than twenty-five percent of the network's measurements during data collection (or eavesdropping) and the subsequent online attack injection phases. The study demonstrates that even within these restrictive parameters, DeeBBAA exhibits significant evasiveness against a broad spectrum of conventional, statistical, and machine learning-based cyberattack detection techniques, resulting in substantial deviations in the estimated network states.
ISSN:2327-4662
2327-4662
DOI:10.1109/JIOT.2024.3454257