Neural Network Equalizers and Successive Interference Cancellation for Bandlimited Channels with a Nonlinearity

TL;DR

This paper introduces neural network-based equalizers combined with successive interference cancellation to efficiently approach the performance of joint detection and decoding in nonlinear bandlimited channels, demonstrated through optical fiber simulations.

Contribution

It proposes a novel neural network equalizer architecture integrated with SIC, reducing complexity while maintaining high information rates in nonlinear channels.

Findings

Neural network equalizers outperform traditional methods in nonlinear channels.

Combining NN equalizers with SIC approaches the performance of joint detection and decoding.

Simulation results show significant gains in optical fiber links.

Abstract

Neural networks (NNs) inspired by the forward-backward algorithm (FBA) are used as equalizers for bandlimited channels with a memoryless nonlinearity. The NN-equalizers are combined with successive interference cancellation (SIC) to approach the information rates of joint detection and decoding (JDD) with considerably less complexity than JDD and other existing equalizers. Simulations for short-haul optical fiber links with square-law detection illustrate the gains.