Secure IIoT architecture with blockchain-enabled anomaly detection for water distribution cyber-physical systems

Area of Science:

• Cyber-Physical Systems Security
• Industrial Internet of Things (IIoT) Applications
• Blockchain Technology in Infrastructure

Background:

• Industrial Internet of Things (IIoT) adoption in water distribution cyber-physical systems (CPS) increases efficiency but expands the attack surface.
• Cyber-attacks on water systems can manipulate sensor data and control logic, leading to severe hydraulic disturbances.

Purpose of the Study:

• To propose a secure IIoT architecture for water-distribution CPS integrating blockchain and anomaly detection.
• To enhance data integrity, validate device identity, and enable real-time attack detection.

Main Methods:

• A permissioned blockchain for tamper-resistant data recording and smart contract-based identity validation.
• A hybrid anomaly-detection model using LSTM autoencoders with cyber-physical feature fusion.
• Evaluation on BATADAL, WADI, and SWaT datasets, simulated networks, and industrial testbeds.

Main Results:

• The proposed framework significantly outperforms classical and deep-learning baselines in accuracy, recall, and AUC.
• Demonstrated cross-dataset transferability in heterogeneous water and industrial CPS environments.
• Performance is influenced by sensor density, process complexity, and attack distribution.

Conclusions:

• The developed framework offers a robust solution for securing IIoT in water distribution CPS.
• Highlights the need for domain adaptation methods in highly heterogeneous deployments for optimal performance.