Multi-Task Learning Aided Neural Networks for Equalization in Coherent Optical System

In this paper, we propose an NN-based multi-symbol equalization scheme inspired by multi-task learning. Compared with the traditional single-output NN-based equalizers, the computational complexity (CC) can be significantly reduced. We conduct an experiment to transmit a 128-Gb/s DP-16QAM signal tra...

Full description

Saved in:
Bibliographic Details
Published in2022 IEEE/CIC International Conference on Communications in China (ICCC) pp. 874 - 877
Main Authors Huang, Fangfang, Huang, Xiatao, Zhang, Jing, Xu, Bo, Yi, Xingwen, Zhang, Qianwu, Qiu, Kun
Format Conference Proceeding
LanguageEnglish
Published IEEE 11.08.2022
Subjects
Artificial neural networks
computational complexity
equalization
Equalizers
Integrated optics
multi-task learning
Optical computing
Optical fiber communications
Optical fiber networks
Optical fibers
Symbols
Online AccessGet full text

Cover

More Information
Summary:In this paper, we propose an NN-based multi-symbol equalization scheme inspired by multi-task learning. Compared with the traditional single-output NN-based equalizers, the computational complexity (CC) can be significantly reduced. We conduct an experiment to transmit a 128-Gb/s DP-16QAM signal transmission over 600-km standard single mode fiber. The experimental results show that a maximum CC reduction of 97% can be achieved with Bi-LSTM-based multi-symbol equalization compared with the corresponding single-output NNs.
DOI:10.1109/ICCC55456.2022.9880731