Data-Driven Modeling of Directly-Modulated Lasers

TL;DR

This paper develops differentiable machine learning models for directly-modulated lasers to enable end-to-end optimized communication links, addressing the challenge of analytically differentiable channels.

Contribution

It introduces and compares machine learning-based differentiable models for directly-modulated lasers, facilitating improved link optimization.

Findings

Machine learning models can effectively simulate laser behavior.

Differentiable laser models enable end-to-end link optimization.

The approach improves the design and analysis of laser-based communication systems.

Abstract

The end-to-end optimization of links based on directly-modulated lasers may require an analytically differentiable channel. We overcome this problem by developing and comparing differentiable laser models based on machine learning techniques.