Abstract

Conventional soft demappers designed for AWGN channels suffer from performance loss under realistic channels. We propose a neural network soft demapper and show a gain of 0.35dB in an 800Gb/s coherent transmission experiment using DP-32QAM.

© 2020 The Author(s)

PDF Article
More Like This
Recurrent Neural Network Soft Demapping for Mitigation of Fiber Nonlinearities and ISI

Maximilian Schaedler, Fabio Pittalà, Stefano Calabrò, Georg Böcherer, Christian Bluemm, and Stephan Pachnicke
M5F.4 Optical Fiber Communication Conference (OFC) 2021

Inter-channel Nonlinear Crosstalk Mitigation Based on Neural Network

Tao Xu, Xiatao Huang, Taowei Jin, Shaohua Hu, Jing Zhang, Bo Xu, Xingwen Yi, and Kun Qiu
JS2B.17 Optoelectronics and Communications Conference (OECC) 2021

Nonlinear Equalizer by Feature Engineering Based-Deep Neural Network for Coherent Optical Communication System

Xinyu Liu, Yongjun Wang, and Chao Li
M4A.275 Asia Communications and Photonics Conference (ACPC) 2020

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription