Graph Neural Network-Based Multi-Sensor Correlation Framework for Fine-Grained Detection and Localisation of Spoofing and Replay Attacks in Energy Systems

Suruthi G; Madhumita K

doi:10.71097/IJSAT.v16.i4.8731

Graph Neural Network-Based Multi-Sensor Correlation Framework for Fine-Grained Detection and Localisation of Spoofing and Replay Attacks in Energy Systems

Author(s)	Ms. Suruthi G, Prof. Dr. Madhumita K
Country	India
Abstract	The digitization of modern energy systems has greatly extended the reliance on distributed sensors and IoT-based monitoring equipment to ensure stability, efficiency, and resilience. However, the resulting dependence also leaves systems vulnerable to sophisticated cyber attacks like spoofing and replay attacks, where attackers inject fake or time-shifted sensor readings into the critical processes. These attacks are especially compelling because they are nearly indistinguishable to automated control loops, operators, and may induce cascading blackouts, equipment malfunctions, and wide-ranging instability. Traditional intrusion detection systems and anomaly-based machine learning provide only system-level alerts, indicating that something is anomalous but without actionable information such as the physical location of these compromised sensors or the type of attack that occurred. This coarseness slows down the operator and erodes confidence in detection results. To tackle these issues, in this paper we propose a Graph Neural Network (GNN)-based multi-sensor correlation framework for fine-grained cyberattack detection and localization in smart energy systems. The framework models the energy network as a dynamic graph where the nodes are sensors and edges denote their dependencies, and then allows the model to learn both spatial and temporal correlations across distributed measurements. The framework uses graph attention mechanisms to identify suspicious nodes and distinguish between different types of attacks (i.e., spoofing vs. replay), while at the same time offering interpretable outputs, which boost the operator's confidence. A hybrid GNN-LSTM model is proposed to provide a scalable framework for learning over extensive sensor networks and to model long-distance dependencies. Simulation on IEEE benchmark bus systems shows that the proposed method outperforms the conventional IDS models in terms of detection effectiveness and false positive rate while also improving localization precision by 55%. The obtained results validate the usefulness of the proposed framework not only for enhancing the resilience against real-time cyberattacks in smart grids but also for enabling the interpretability of the actions and the scalability of defence mechanisms of the utmost importance for safe operations of future energy infrastructures.
Keywords	Graph Neural Networks (GNNs), Cyberattack Detection, Smart Energy Systems, Sensor Data Correlation, Spoofing and Replay Attacks.
Field	Computer > Network / Security
Published In	Volume 16, Issue 4, October-December 2025
Published On	2025-10-25
DOI	https://doi.org/10.71097/IJSAT.v16.i4.8731
Short DOI	https://doi.org/g98ndh

View / Download PDF File

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJSAT DOI prefix is
10.71097/IJSAT

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 16 Isu 3 Cover Page Vol 16 Isu 2 Cover Page Vol 16 Isu 1

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJSAT Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijsat.org

International Journal on Science and Technology

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Graph Neural Network-Based Multi-Sensor Correlation Framework for Fine-Grained Detection and Localisation of Spoofing and Replay Attacks in Energy Systems

Share this