Modelling and Detection of Security Threats to Submarine Optical Cables

Research Project, 2025 – 2030

Submarine optical fiber cables are vital to global communications, but face growing security threats, from accidental damages to deliberate attacks. Current security monitoring methods rely on reactive detection, lacking a robust theoretical foundation to distinguish threats from benign disturbances. This project develops a comprehensive physics-based framework to model and detect security threats in submarine optical fiber networks. By integrating optical signal propagation models, threat interaction models, and physics-informed machine learning, we aim to link external disturbances—such as mechanical stress or tampering—to measurable optical effects. The research follows three interconnected tasks: (i) theoretical modeling of security-related interactions, (ii) experimental validation using lab facilities and a dedicated underwater test site, and (iii) development of advanced detection strategies. The resulting models will improve anomaly detection and go beyond existing sensing techniques by adapting state of the art algorithms to the marine-fiber environment.  This project establishes a rigorous, model-driven approach to submarine cable security, with broad applications in network resilience, environmental sensing, and early-warning systems. The project results will guide future security-aware infrastructure planning and installation.