Supply-Power-Constrained Cable Capacity Maximization Using Deep Neural Networks

TL;DR

This paper demonstrates a method to significantly increase cable capacity efficiency by using deep neural networks to optimize launch powers, eliminating the need for gain flattening filters in fiber optic links.

Contribution

It introduces a machine learning approach with deep neural networks to optimize cable capacity under supply power constraints, achieving notable efficiency improvements.

Findings

19% capacity gain per Watt achieved

Elimination of gain flattening filters

Optimization of launch powers using deep learning

Abstract

We experimentally achieve a 19% capacity gain per Watt of electrical supply power in a 12-span link by eliminating gain flattening filters and optimizing launch powers using machine learning by deep neural networks in a massively parallel fiber context.