Resilient Anomaly Detection in Fiber-Optic Networks - A Machine Learning Framework for Multi-Threat Identification Using State-of-Polarization Monitoring

Aug 5, 2026

A comprehensive machine-learning framework is presented for robust anomaly identification in optical fiber networks using real-time state-of-polarization (SOP) monitoring. SOP data are exploited under three threat scenarios: malicious or critical vibration events, overlapping mechanical disturbances, and malicious fiber tapping for eavesdropping. Supervised models including k-nearest neighbor, random forest, XGBoost, and decision trees classify vibration events, while resilience is evaluated under background interference by superimposed sinusoidal noise at varied frequencies, clarifying detection fidelity under ambient vibrations and informing noise-mitigation needs.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages22

Reading Time55 min

AccessFree to download

Abstract

Introduction

Optical networks and reliability needs

Environmental monitoring with pre-installed fiber infrastructures

Fiber sensitivity and sensing overview

Related methods and framework design

POLITECNICO DI TORINO Repository ISTITUZIONALE Resilient Anomaly Detection in Fiber-Optic Networks: A Machine Learning Framework for Multi-Threat Identification Using State-of-Polarization Monitoring Original Resilient Anomaly Detection in Fiber-Optic Networks: A Machine Learning Framework for Multi-Threat Identification Using State-of-Polarization Monitoring / Malik, Gulmina; Dipto, Imran Chowdhury; Masood, Muhammad Umar; Mohamed, Mashboob Cheruvakkadu; Straullu, Stefano; Bhyri, Sai Kishore; Galimberti, Gabriele Maria; Napoli, Antonio; Pedro, João; Wakim, Walid; Curri, Vittorio. -In: AI. -ISSN 2673-2688. -6:7(2025) . [10 .3390/ai6070131] Availability: This version is available at: 11583/3001170 since: 2025-06-20T13:45:06Z Publisher: MDPI Published DOI:10.3390/ai6070131 Terms of use: This article is made available under terms and conditions as specified in the corresponding bibliographic description in the repository Publisher copyright (Article begins on next page) 04 October 2025 AI Article Resilient Anomaly Detection in Fiber-Optic Networks: A Machine Learning Framework for Multi-Threat Identification Using State-of-Polarization Monitoring Gulmina Malik 1, *, Imran Chowdhury Dipto 1, Muhammad Umar Masood 1, Mashboob Cheruvakkadu Mohamed 1, Stefano Straullu 2, Sai Kishore Bhyri 3, Gabriele Maria Galimberti 4, Antonio Napoli 5, João Pedro 6, Walid Wakim 7 and Vittorio Curri 1 Academic Editor: Yibeltal Chanie Manie Received: 16 May 2025 Revised: 9 June 2025 Accepted: 13 June 2025 Published: 20 June 2025 Citation: Malik , G.; Dipto, I.C.; Masood, M.U.; Mohamed, M.C.; Straullu, S.; Bhyri, S.K.; Galimberti, G.M.; Napoli, A.; Pedro, J.; Wakim, W.; et al. Resilient Anomaly Detection in Fiber-Optic Networks: A Machine Learning Framework for Multi-Threat Identification Using State-ofPolarization Monitoring. AI 2025, 6, 131. [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)ai6070131 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://creativecommons.org/](https://creativecommons.org/)[ ](https://creativecommons.org/)[licenses/by/4.0/](licenses/by/4.0/)) . 1 Department of Electronics and Telecommunications, Polytechnic University of Turin, 10129 Turin, Italy; imran.dipto@polito.it (I.C.D.); muhammad.masood@polito.it (M.U.M.); mashboob.cheruvakkadu@polito.it (M.C.M.); vittorio.curri@polito.it (V.C.) 2 LINKS Foundation, 10129 Turin, Italy; [stefano.straullu@linksfoundation.com](stefano.straullu@linksfoundation.com) 3 Optical Networks, Nokia, Bangalore 560045, India; [sai.bhyri@nokia.com](sai.bhyri@nokia.com) 4 Optical Networks, Nokia, 20060 Milan, Italy; [gabriele.galimberti@nokia.com](gabriele.galimberti@nokia.com) 5 Optical Networks, Nokia, 81541 Munich, Germany; [antonio.napoli@nokia.com](antonio.napoli@nokia.com) 6 Optical Networks, Nokia, 2720-092 Carnaxide, Portugal; [joao.pedro@nokia.com](joao.pedro@nokia.com) 7 Optical Networks, Nokia, Naperville, IL 60563, USA; [walid.wakim@nokia.com](walid.wakim@nokia.com) * Correspondence: gulmina.malik@polito.it Abstract We present a thorough machine-learning framework based on real-time state-of-polarization (SOP) monitoring for robust anomaly identification in optical fiber networks. We exploit SOP data under three different threat scenarios: (i) malicious or critical vibration events,(ii) overlapping mechanical disturbances, and (iii) malicious fiber tapping (eavesdropping) . We used various supervised machine learning techniques like k-Nearest Neighbor (k-NN), random forest, extreme gradient boosting (XGBoost), and decision trees to classify different vibration events. We also assessed the framework’s resilience to background interference by superimposing sinusoidal noise at different frequencies and examining its effect

FAQs

What monitoring signal does the framework use for anomaly detection in fiber networks?

It uses real-time state-of-polarization (SOP) monitoring data to derive polarization signatures for detecting anomalies.

Which threats and interference types are considered in the study?

Three threat scenarios are used: malicious/critical vibrations, overlapping mechanical disturbances, and malicious fiber tapping. Resilience is tested by superimposing sinusoidal noise at different frequencies to model background interference.

How does the system identify and classify vibration events?

It applies supervised machine learning classifiers such as k-NN, random forest, XGBoost, and decision trees to classify the vibration events based on SOP-derived features.

Aug 5, 2026

Upload by Seraphina