Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids

Recent research has shown promise in using machine learning for cyberattack detection in power systems. However, current studies face limitations: a) dependence on either physical or cyber features, overlooking multi-modal cyber-physical (CP) correlations; b) unrealistic full observability assumptio...

Full description

Saved in:
Bibliographic Details
Main Authors: Jacob Sweeten, Amr Elshazly, Abdulrahman Takiddin, Muhammad Ismail, Shady S. Refaat, Rachad Atat
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Access Journal of Power and Energy
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11105456/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849239089857953792
author Jacob Sweeten
Amr Elshazly
Abdulrahman Takiddin
Muhammad Ismail
Shady S. Refaat
Rachad Atat
author_facet Jacob Sweeten
Amr Elshazly
Abdulrahman Takiddin
Muhammad Ismail
Shady S. Refaat
Rachad Atat
author_sort Jacob Sweeten
collection DOAJ
description Recent research has shown promise in using machine learning for cyberattack detection in power systems. However, current studies face limitations: a) dependence on either physical or cyber features, overlooking multi-modal cyber-physical (CP) correlations; b) unrealistic full observability assumptions; c) focus on detecting basic attacks instead of advanced threats such as ransomware (RW); and d) use of deep learning (DL) models built for 2D data, despite the graph-structured nature of power systems. To address these gaps, we develop a CP testbed using OPAL-RT and a cyber range to simulate both physical and cyber layers under full and partial observability. The testbed produces a realistic multi-modal dataset covering normal operations and various cyberattacks, including RW, brute force, false data injection, reverse shell, and backdoor. Using this dataset, we design graph neural network (GNN)-based multi-modal intrusion detection systems (IDSs) that fuse CP features and capture spatio-temporal dependencies. Results show that CP fusion improves detection rates (DRs) by up to 16% compared to single-modal inputs. The proposed GNN-based IDSs outperform benchmarks by up to 26% in DR, remain effective under partial observability, and demonstrate up to 6% improvement in scalability when applied to larger system topologies.
format Article
id doaj-art-451e88098fb945d58a8577be98ff6b50
institution Kabale University
issn 2687-7910
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Access Journal of Power and Energy
spelling doaj-art-451e88098fb945d58a8577be98ff6b502025-08-20T04:01:15ZengIEEEIEEE Open Access Journal of Power and Energy2687-79102025-01-011251552810.1109/OAJPE.2025.359462511105456Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power GridsJacob Sweeten0Amr Elshazly1https://orcid.org/0009-0005-5224-0721Abdulrahman Takiddin2https://orcid.org/0000-0003-4793-003XMuhammad Ismail3https://orcid.org/0000-0002-8051-9747Shady S. Refaat4https://orcid.org/0000-0001-9392-6141Rachad Atat5https://orcid.org/0000-0001-8075-6243Cybersecurity Education, Research, and Outreach Center (CEROC), Tennessee Technological University, Cookeville, TN, USACybersecurity Education, Research, and Outreach Center (CEROC), Tennessee Technological University, Cookeville, TN, USADepartment of Electrical and Computer Engineering, FAMU-FSU College of Engineering, Florida State University, Tallahassee, FL, USACybersecurity Education, Research, and Outreach Center (CEROC), Tennessee Technological University, Cookeville, TN, USASchool of Physics, Engineering, and Computer Science, University of Hertfordshire, Hatfield, U.K.Department of Computer Science and Mathematics, Lebanese American University, Beirut, LebanonRecent research has shown promise in using machine learning for cyberattack detection in power systems. However, current studies face limitations: a) dependence on either physical or cyber features, overlooking multi-modal cyber-physical (CP) correlations; b) unrealistic full observability assumptions; c) focus on detecting basic attacks instead of advanced threats such as ransomware (RW); and d) use of deep learning (DL) models built for 2D data, despite the graph-structured nature of power systems. To address these gaps, we develop a CP testbed using OPAL-RT and a cyber range to simulate both physical and cyber layers under full and partial observability. The testbed produces a realistic multi-modal dataset covering normal operations and various cyberattacks, including RW, brute force, false data injection, reverse shell, and backdoor. Using this dataset, we design graph neural network (GNN)-based multi-modal intrusion detection systems (IDSs) that fuse CP features and capture spatio-temporal dependencies. Results show that CP fusion improves detection rates (DRs) by up to 16% compared to single-modal inputs. The proposed GNN-based IDSs outperform benchmarks by up to 26% in DR, remain effective under partial observability, and demonstrate up to 6% improvement in scalability when applied to larger system topologies.https://ieeexplore.ieee.org/document/11105456/Cyber-physicalcyberattack detectionpower systemsgraph neural networksmachine learningsmart grids
spellingShingle Jacob Sweeten
Amr Elshazly
Abdulrahman Takiddin
Muhammad Ismail
Shady S. Refaat
Rachad Atat
Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
IEEE Open Access Journal of Power and Energy
Cyber-physical
cyberattack detection
power systems
graph neural networks
machine learning
smart grids
title Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
title_full Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
title_fullStr Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
title_full_unstemmed Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
title_short Cyber-Physical Fusion for GNN-Based Attack Detection in Smart Power Grids
title_sort cyber physical fusion for gnn based attack detection in smart power grids
topic Cyber-physical
cyberattack detection
power systems
graph neural networks
machine learning
smart grids
url https://ieeexplore.ieee.org/document/11105456/
work_keys_str_mv AT jacobsweeten cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids
AT amrelshazly cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids
AT abdulrahmantakiddin cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids
AT muhammadismail cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids
AT shadysrefaat cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids
AT rachadatat cyberphysicalfusionforgnnbasedattackdetectioninsmartpowergrids