Leveraging Shared Data and Models for ML-Based QoT Estimation: Toward Standardized and Generalizable Models
This work presents the first study of a machine learning-based quality-of-transmission (QoT) estimator using synthetic and experimental datasets from four different organizations, including live network data. It investigates the feasibility of a unified model, enabling vendor-ready solutions with minimal customization across diverse optical network scenarios.
Authors
- Johannes Fischer (ORCID: https://orcid.org/0000-0002-1886-6633)
- Angela Mitrovska
- Jasper Müller (ORCID: https://orcid.org/0000-0002-7021-5416)
- Stephan Pachnicke (ORCID: https://orcid.org/0000-0001-7321-7938)
- Pooyan Safari (ORCID: https://orcid.org/0000-0002-5916-8575)
- Ronald Freund (ORCID: https://orcid.org/0000-0001-9427-3437)
- Xiao Ma (ORCID: https://orcid.org/0009-0003-8904-3110)
- Behnam Shariati (ORCID: https://orcid.org/0000-0003-1195-3862)
Institutions
- Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute (DE)
- Allgemeiner Deutscher Automobil Club (DE)
- Hochschule für Angewandte Wissenschaften Kiel (DE)
- Christian-Albrechts-Universität zu Kiel (DE)
- Clinical Research Center Kiel (DE)
Publication Details
- Published
- 2025-09-28
- DOI
- https://doi.org/10.1109/ecoc66593.2025.11263378
- Primary Topic
- Optical Network Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00