Self-Normalizing Neural Network, Enabling One Shot Transfer Learning for Modeling EDFA Wavelength Dependent Gain
This addresses the problem of efficiently modeling optical amplifier gain for network operators, though it appears incremental as it builds on existing neural network and transfer learning concepts.
The paper tackles modeling wavelength-dependent gain for multiple EDFAs by introducing a semi-supervised, self-normalizing neural network framework that enables one-shot transfer learning, achieving high-accuracy transfer across different amplifier types in experiments on 22 EDFAs in Open Ireland and COSMOS testbeds.
We present a novel ML framework for modeling the wavelength-dependent gain of multiple EDFAs, based on semi-supervised, self-normalizing neural networks, enabling one-shot transfer learning. Our experiments on 22 EDFAs in Open Ireland and COSMOS testbeds show high-accuracy transfer-learning even when operated across different amplifier types.