Achievable Information Rates for Nonlinear Fiber Communication via End-to-end Autoencoder Learning
This work addresses the challenge of characterizing communication limits in nonlinear fiber systems for researchers in optical communications, but it is incremental as it applies known autoencoder techniques to a specific domain.
The authors tackled the problem of computing achievable information rates for nonlinear fiber communication by using an end-to-end autoencoder to jointly optimize input and auxiliary distributions without explicit channel knowledge, resulting in a method that provides numerical AIR estimates for this simplified channel.
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.