Railway transportation systems are critical infrastructures, but they also face vulnerabilities from both cyber and physical threats. This research proposes a comprehensive approach using Bayesian Networks to model and analyze these threats. The model classifies, identifies, and assesses risks to railway networks by analyzing conditional dependencies. This research proposes a comprehensive approach to modeling and analyzing these threats using Bayesian Networks. Bayesian networks are graphical models that depict complex relationships among variables in a probabilistic framework. The proposed method involves modeling the interactions between cyber threats, physical components, and their potential consequences by identifying key system elements, defining possible cyber threats and physical impacts, constructing a Bayesian network structure, and estimating probability tables with conditions. The model seeks to classify, identify, and assess the risks to railway networks by analyzing the conditional dependencies among system vulnerabilities and potential threat vectors. Using the ML.NET library and C# programming language, a probabilistic framework was developed for real-time threat detection and risk assessment. The model demonstrated high accuracy, precision, and recall, offering a practical and scalable solution for detecting and mitigating cyber-physical threats.

Authors: O. Y. Ogunlola, O. Owolafe, A. O. Oronti, B.K. Alese

Published in: International Conference for Internet Technology and Secured Transactions (ICITST-2024)

• Date of Conference: 4-6 November 2024
• DOI: 10.20533/ICITST.2024.0016
• ISBN: 978-1-913572-76-1
• Conference Location: St Anne’s College, Oxford University, UK