PhD Thesis Appendix B Implementation

This repository contains the implementation of the multi-level detection framework using the Factory I/O simulation-based testbed.

Purpose

This repository supports reproduction of the experiments presented in Appendix B of the PhD thesis, including:

• Micro-level detection (Process / OT): OT process anomaly detection using LSTM-FCN, including raw input data, preprocessing scripts, experimental results, and Jupyter notebooks.
• Meso-level detection (Network / OT): OPC UA network anomaly detection using Isolation Forest, including raw input data, feature extraction scripts, experimental results, and notebooks.
• Meso-level detection (Host / IT): Host anomaly detection using Wazuh logs (Debian-based environment), including rule-based scoring scripts, processed datasets, and notebooks.
• Macro-level detection: Implementation of the fusion logic, experimental evaluation, and visualization scripts.
• Fused anomaly log: Exported anomaly logs from micro-, meso-, and macro-level detection, including experimental LLM-based summarisation outputs.

This repository is provided for research transparency and reproducibility. It is not intended as a production-ready system.