Policy-driven Conformal Prediction for Trustworthy QoT Estimation
Paper in proceeding, 2025
We propose Conformal QoT, a policy-driven framework that combines statistically guaranteed QoT estimation with operational decision policies, enabling reliable lightpath-feasibility predictions under domain shift and improving accuracy from 92% to 99.6% on open datasets.
Optical networks
Quality of transmission
Artificial intelligence
Machine learning
Author
Kiarash Rezaei
Chalmers, Electrical Engineering, Communication, Antennas and Optical Networks
Omran Ayoub
University of Applied Sciences and Arts of Italian Switzerland (SUPSI)
Paolo Monti
Chalmers, Electrical Engineering, Communication, Antennas and Optical Networks
Carlos Natalino Da Silva
Chalmers, Electrical Engineering, Communication, Antennas and Optical Networks
Proceedings of the Optical Fiber Communication Conference (OFC) 2026
M4A.X
Optical Fiber Communication Conference (OFC) 2026
Los Angeles, CA, USA,
Los Angeles, CA, USA,
Efficient Confluent Edge Networks (ECO-eNET)
European Commission (EC) (EC/HE/101139133), 2024-01-01 -- 2028-12-31.
Areas of Advance
Information and Communication Technology
Subject Categories (SSIF 2025)
Computer Sciences
Telecommunications
Information Systems
Infrastructure
Chalmers e-Commons (incl. C3SE, 2020-)
Related datasets
Datasets for QoT estimation in SDM networks [dataset]
DOI: 10.1364/JOCN.558452