Achievable Information Rates for Nonlinear Fiber Communication via End-to-end Autoencoder Learning

TL;DR

This paper introduces an end-to-end machine learning approach to compute achievable information rates for nonlinear fiber channels by jointly optimizing input and auxiliary channel distributions without explicit channel models.

Contribution

It presents a novel autoencoder-based method for estimating AIRs in fiber communications, bypassing the need for explicit channel knowledge.

Findings

Effective joint optimization of input and auxiliary distributions

Accurate estimation of AIRs for nonlinear fiber channels

Potential for improved communication system design

Abstract

Machine learning is used to compute achievable information rates (AIRs) for a simplified fiber channel. The approach jointly optimizes the input distribution (constellation shaping) and the auxiliary channel distribution to compute AIRs without explicit channel knowledge in an end-to-end fashion.