A Cyber-Physical Transformer For Intrusion Detection In Water Supply Networks

Abstract

Cyber-attacks pose significant risks to critical infrastructure systems, including water supply networks, with potential consequences ranging from data theft to infrastructure damage and public health hazards. Traditional approaches to identifying cyber-attacks lag behind those for physical attacks due to the unique challenges posed by cyber-attacks’ rapid propagation and large-scale impact. While previous studies have explored cyber-attack detection using model-based analysis and Generative Adversarial Networks (GANs), This research aims to advance the field by integrating hydraulic modeling with network traffic modeling, generating multidimensional data from two distinct modeling domains that can be correlated and combined into a single complex time series. Utilizing this data, we leverage novel Transformer-based architectures to enhance the robustness and accuracy of intrusion and anomaly detection in water supply networks.

Stav Cohen

PhD Candidate

I’m a DataScience PhD candidate at the Technion – Israel Institute of Technology