Machine Learning for Ultrawide Bandwidth Amplifier Configuration

Ultrawide bandwidth optical amplifier designs based on multiple pump wavelengths additionally benefit from gain control per sub-bands, at the cost of an increased configuration complexity. In such systems it becomes possible to adapt system parameters such as spectral tilt and optical output power p...

Full description

Saved in:
Bibliographic Details
Published in2019 21st International Conference on Transparent Optical Networks (ICTON) pp. 1 - 4
Main Author Ionescu, Maria
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.07.2019
Subjects
Artificial neural networks
Machine learning
Optical amplifiers
Optical fiber amplifiers
Optical fiber communication
Raman/EDFA optical amplifiers
Stimulated emission
ultra-wideband systems
Online AccessGet full text

Cover

More Information
Summary:Ultrawide bandwidth optical amplifier designs based on multiple pump wavelengths additionally benefit from gain control per sub-bands, at the cost of an increased configuration complexity. In such systems it becomes possible to adapt system parameters such as spectral tilt and optical output power per amplifier, in response to changing conditions in the network, such as wavelength re-configuration and repairs. Machine learning gives a high accuracy performance in configuring and monitoring amplifiers easing the complexity incurred by human-effort.
ISSN:2161-2064
DOI:10.1109/ICTON.2019.8840453