SNR Optimization of Multi-Span Fiber Optic Communication Systems Employing EDFAs With Non-Flat Gain And Noise Figure

Abstract

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Users may download and print one copy of any publication from the public portal for the purpose of private study or research.  You may not further distribute the material or use it for any profit-making activity or commercial gain  You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Spectral Power Profile Optimization of Field-Deployed WDM Network by Remote Link Modeling

“…At the receiver, the data channels are extracted with two optical filters and an EDFA, followed by a coherent frontend and a digital storage oscilloscope (DSO). The DSP is pilot-based as reported in our previous work [8] . For the experimental demonstration, two links of the network are chosen.…”

“…The methods proposed extend the methodology in our previous work [7], [8] performed in a lab environment. The overall goal is to predict the signal power and the accumulated noise power for all WDM channels of a link as a function of its input power profile.…”

“…In contrast to our previous work [8] , where the EDFA is entirely modeled by neural networks (NNs), in this work a combination of a physical model and NNs is used. More specifically, the physical EDFA model presented in Meseguer et al [10] is modified, replacing both look up tables for the total power dependent gain profile and noise figure functions with NNs, ensuring that the resulting EDFA model is differentiable.…”

“…The model is then used to optimize operation specific properties of the network [13]- [15] . In particular, the spectral power per WDM channel is optimized with a target of equalizing the signal to noise ratio (SNR) across all channels without requiring feedback on the quality of transmission as in Garbhapu et al [15] and without requiring gain flattening filters after the EDFAs [8] .…”

Fiber-Agnostic Machine Learning-Based Raman Amplifier Models

“…Moreover, such numerical models are not differentiable for the backward pumping scheme (the preferred one due to its robustness to pump power fluctuations). This prevents their applications in system optimizations based on gradient descent routines [11]. Therefore, machine learning (ML) has been recently applied to provide fast and highly-accurate gain and noise predictions for erbium-DFAs (EDFAs) [12]- [15] and RAs [16]- [21].…”