From Communication to Sensing: Deep Learning for Fiber Perturbation Detection and Classification Using Streaming Telemetry
Paper i proceeding, 2026

Optical fiber networks are increasingly explored as sensing platforms for monitoring environmental and physical perturbations without requiring dedicated sensing infrastructure. However, commercially available coherent optical transceivers do not readily expose all raw parameters typically used for physical-layer sensing. Rather, they produce telemetry data that represents a latent combination of the transmission and the measuring systems. In this work, we propose a deep learning framework for accurate detection and classification of pressure and temperature induced perturbations in optical fibers using coherent transceiver telemetry data. By leveraging telemetry information already available in operational coherent optical networks, the proposed approach enables low-cost and real-time monitoring using existing communication infrastructure. Experimental evaluation on a dataset containing both temperature and pressure perturbations demonstrates an average classification accuracy of 98%, highlighting the potential of coherent transceiver telemetry for intelligent fiber monitoring and anomaly awareness applications.

Fiber optic sensing

deep learning

real-time monitoring

Författare

Fehmida Usmani

Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk

Jan Bukovsky

Vysoke uceni technicke v Brne

Petr Dejdar

Vysoke uceni technicke v Brne

Wadood Ur Rehman Ranjha

National University of Sciences and Technology (NUST)

Petr Munster

Vysoke uceni technicke v Brne

Tomas Horvath

Vysoke uceni technicke v Brne

Adrian Tomasov

Vysoke uceni technicke v Brne

Elaine Wong

University of Melbourne

Marija Furdek Prekratic

Chalmers, Elektroteknik, Kommunikation, Antenner och Optiska Nätverk

2026 26th International Conference on Transparent Optical Networks Icton 2026


9798319544216 (ISBN)

26th International Conference on Transparent Optical Networks, ICTON 2026
Prague, Czech Republic,

Ämneskategorier (SSIF 2025)

Kommunikationssystem

Telekommunikation

DOI

10.1109/ICTON71309.2026.11705158

Mer information

Senast uppdaterat

2026-11-02