Machine Learning for Ultrawide Bandwidth Amplifier Configuration

• Published in: 2019 21st International Conference on Transparent Optical Networks (ICTON) pp. 1 - 4
• Main Author: Ionescu, Maria
• Format: Conference Proceeding
• Language: English
• 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

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...