Model-based approach for cyber-physical attack detection in water distribution systems

Area of Science:

• Cybersecurity in critical infrastructure
• Water resource management systems
• Data science and anomaly detection

Background:

• Modern water distribution systems (WDSs) rely on Supervisory Control and Data Acquisition (SCADA) and Programmable Logic Controllers (PLCs).
• The increasing integration of cyber components in WDS operations heightens vulnerability to cyberattacks.
• Existing detection methods may not identify complex, hydraulically subtle cyber threats.

Purpose of the Study:

• To develop a model-based methodology for detecting sophisticated cyberattacks in WDSs.
• To combine hydraulic principles with advanced anomaly detection for enhanced threat identification.
• To evaluate the proposed algorithm's performance against established benchmarks.

Main Methods:

• A model-based approach integrating detailed hydraulic understanding of WDSs.
• Implementation of an anomaly detection algorithm for identifying cyber intrusions.
• Testing and validation using data from the BATtle of the Attack Detection ALgorithms (BATADAL) competition.

Main Results:

• The proposed algorithm demonstrated superior performance in identifying complex cyberattacks.
• The methodology effectively detected threats that traditional hydraulically based rules could not.
• Achieved state-of-the-art results on the publicly available BATADAL dataset.

Conclusions:

• The combined hydraulic and anomaly detection approach offers a robust solution for WDS cybersecurity.
• This methodology enhances the ability to detect sophisticated cyberattacks in critical water infrastructure.
• The findings provide a valuable tool for securing WDS operations against evolving cyber threats.