Polarized optical signals traveling through optical networks are sensitive to environmental changes, making them suitable for capturing valuable data about their surroundings. This enables the entire optical infrastructure to function as a sensing and localization grid for earthquake early detection. In this paper, we experimentally validate the co-existence of a sensing detector with WDM traffic generators, making the proposed approach both practical and robust for real-world applications. Offline data post-processing and fine-tuning of a pretrained machine learning method are employed to identify the pattern of polarization changes caused by the primary earthquake wave, which precedes the destructive wave by tens of seconds. To assess the system's effectiveness, we replicate, in a laboratory setup, the dynamics of a real M4.3 earthquake as a function of strain, propagating along 38 km of deployed fiber in a ring topology. The system captures two different earthquakeinduced polarization datasets, showcasing that the machine learning model, pretrained on one modality, successfully adapts to another. The model still achieves high accuracy in detecting the primary earthquake wave despite propagation challenges, demonstrating the potential for real-world integration of optical fiber networks as both high-speed communication channels and smart environmental sensors.

Experimental Validation on Using Optical Network-as-a-Sensor for Earthquake Early Warning

Awad, Hasan;Straullu, Stefano;Usmani, Fehmida;Virgillito, Emanuele;Proietti, Roberto;Marianna, Hovsepyan;Curri, Vittorio
2025

Abstract

Polarized optical signals traveling through optical networks are sensitive to environmental changes, making them suitable for capturing valuable data about their surroundings. This enables the entire optical infrastructure to function as a sensing and localization grid for earthquake early detection. In this paper, we experimentally validate the co-existence of a sensing detector with WDM traffic generators, making the proposed approach both practical and robust for real-world applications. Offline data post-processing and fine-tuning of a pretrained machine learning method are employed to identify the pattern of polarization changes caused by the primary earthquake wave, which precedes the destructive wave by tens of seconds. To assess the system's effectiveness, we replicate, in a laboratory setup, the dynamics of a real M4.3 earthquake as a function of strain, propagating along 38 km of deployed fiber in a ring topology. The system captures two different earthquakeinduced polarization datasets, showcasing that the machine learning model, pretrained on one modality, successfully adapts to another. The model still achieves high accuracy in detecting the primary earthquake wave despite propagation challenges, demonstrating the potential for real-world integration of optical fiber networks as both high-speed communication channels and smart environmental sensors.
2025
979-8-3315-0150-1
File in questo prodotto:
File Dimensione Formato  
Experimental_Validation_on_Using_Optical_Network-as-a-Sensor_for_Earthquake_Early_Warning.pdf

accesso riservato

Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 1.31 MB
Formato Adobe PDF
1.31 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/3002534