Cyber-Physical System Architecture for Intrusion Detection in Industrial IoT Networks

The proliferation of Industrial Internet of Things (IIoT) devices has dramatically expanded the attack surface of critical infrastructure systems, necessitating robust and adaptive intrusion detection mechanisms. This paper presents a novel cyber-physical system (CPS) architecture specifically designed for intrusion detection in industrial IoT networks. Our proposed framework integrates a multi-layered detection strategy combining signature-based and anomaly-based detection techniques, leveraging deep learning models — specifically a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) architecture — to identify both known and zero-day threats. We evaluate our system on the CICIDS-2018 dataset augmented with IIoT-specific traffic profiles, achieving a detection accuracy of 97.6%, a false positive rate of 1.2%, and a false negative rate of 1.4%, outperforming state-of-the-art methods. The proposed architecture accounts for the resource-constrained nature of IIoT devices through edge-based preprocessing and hierarchical detection, and it complies with IEC 62443 and NIST SP 800-82 cybersecurity standards. Our results demonstrate that the proposed system provides a practical, scalable, and effective solution for securing industrial environments including manufacturing, energy, and water treatment systems.

Cited via DOI — this article is permanently archived on Zenodo.