Computing continuous nonlinear Fourier spectrum of optical signal with artificial neural networks

Egor Sedov; Yaroslav Prylepskiy; Igor Chekhovskoy; Sergei Turitsyn; Michael E. Zelinski; Tarek M. Taha; Jonathan Howe

doi:10.1117/12.2594127

Computing continuous nonlinear Fourier spectrum of optical signal with artificial neural networks

Egor Sedov, Yaroslav Prylepskiy, Igor Chekhovskoy, Sergei Turitsyn, Michael E. Zelinski (Editor), Tarek M. Taha (Editor), Jonathan Howe (Editor)

Research output: Contribution to journal › Conference article › peer-review

Abstract

In this work, we demonstrate that the high-accuracy computation of the continuous nonlinear spectrum can be performed by using artificial neural networks. We propose the artificial neural network (NN) architecture that can efficiently perform the nonlinear Fourier (NF) optical signal processing. The NN consists of sequential convolution layers and fully connected output layers. This NN predicts only one component of the continuous NF spectrum, such that two identical NNs have to be used to predict the real and imaginary parts of the reflection coefficient. To train the NN, we precomputed 94035 optical signals. 9403 signals were used for validation and excluded from training. The final value of the relative error for the entire validation dataset was less than 0.3%. Our findings highlight the fundamental possibility of using the NNs to analyze and process complex optical signals when the conventional algorithms can fail to deliver an acceptable result.

Original language	English
Article number	118430J
Number of pages	22
Journal	Proceedings of SPIE - International Society for Optical Engineering
Volume	11843
DOIs	https://doi.org/10.1117/12.2594127
Publication status	Published - 1 Aug 2021
Event	Applications of Machine Learning 2021 - San Diego, United States Duration: 1 Aug 2021 → 5 Aug 2021

Bibliographical note

Copyright 2021 SPIE. One print or electronic copy may be made for personal use only. Systematic reproduction, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.

Funding: E.S. and I.C. acknowledges the support from Russian Science Foundation under Grant 17-72-30006 and the support by the grant of the President of the Russian Federation (MK-677.2020.9). E.S. and S.T. are supported by the EPSRC programme grant TRANSNET, EP/R035342/1. S.T. and J.P. acknowledge the support of Leverhulme Trust project RPG-2018-063.

Keywords

Nonlinear Fourier transform
Neural network
Optical communication
Signal processing

Access to Document

10.1117/12.2594127

Computing continuous nonlinear Fourier spectrum
Copyright 2021 SPIE. One print or electronic copy may be made for personal use only. Systematic reproduction, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.
Final published version, 328 KB

Cite this

@article{0d717f5778464094897aa11e6ff099e6,

title = "Computing continuous nonlinear Fourier spectrum of optical signal with artificial neural networks",

abstract = "In this work, we demonstrate that the high-accuracy computation of the continuous nonlinear spectrum can be performed by using artificial neural networks. We propose the artificial neural network (NN) architecture that can efficiently perform the nonlinear Fourier (NF) optical signal processing. The NN consists of sequential convolution layers and fully connected output layers. This NN predicts only one component of the continuous NF spectrum, such that two identical NNs have to be used to predict the real and imaginary parts of the reflection coefficient. To train the NN, we precomputed 94035 optical signals. 9403 signals were used for validation and excluded from training. The final value of the relative error for the entire validation dataset was less than 0.3%. Our findings highlight the fundamental possibility of using the NNs to analyze and process complex optical signals when the conventional algorithms can fail to deliver an acceptable result.",

keywords = "Nonlinear Fourier transform, Neural network, Optical communication, Signal processing",

author = "Egor Sedov and Yaroslav Prylepskiy and Igor Chekhovskoy and Sergei Turitsyn and Zelinski, {Michael E.} and Taha, {Tarek M.} and Jonathan Howe",

note = "Copyright 2021 SPIE. One print or electronic copy may be made for personal use only. Systematic reproduction, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited. Funding: E.S. and I.C. acknowledges the support from Russian Science Foundation under Grant 17-72-30006 and the support by the grant of the President of the Russian Federation (MK-677.2020.9). E.S. and S.T. are supported by the EPSRC programme grant TRANSNET, EP/R035342/1. S.T. and J.P. acknowledge the support of Leverhulme Trust project RPG-2018-063. ; Applications of Machine Learning 2021 ; Conference date: 01-08-2021 Through 05-08-2021",

year = "2021",

month = aug,

day = "1",

doi = "10.1117/12.2594127",

language = "English",

volume = "11843",

journal = "Proceedings of SPIE - International Society for Optical Engineering",

issn = "0277-786X",

publisher = "SPIE",

}

TY - JOUR

T1 - Computing continuous nonlinear Fourier spectrum of optical signal with artificial neural networks

AU - Sedov, Egor

AU - Prylepskiy, Yaroslav

AU - Chekhovskoy, Igor

AU - Turitsyn, Sergei

A2 - Zelinski, Michael E.

A2 - Taha, Tarek M.

A2 - Howe, Jonathan

N1 - Copyright 2021 SPIE. One print or electronic copy may be made for personal use only. Systematic reproduction, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited. Funding: E.S. and I.C. acknowledges the support from Russian Science Foundation under Grant 17-72-30006 and the support by the grant of the President of the Russian Federation (MK-677.2020.9). E.S. and S.T. are supported by the EPSRC programme grant TRANSNET, EP/R035342/1. S.T. and J.P. acknowledge the support of Leverhulme Trust project RPG-2018-063.

PY - 2021/8/1

Y1 - 2021/8/1

N2 - In this work, we demonstrate that the high-accuracy computation of the continuous nonlinear spectrum can be performed by using artificial neural networks. We propose the artificial neural network (NN) architecture that can efficiently perform the nonlinear Fourier (NF) optical signal processing. The NN consists of sequential convolution layers and fully connected output layers. This NN predicts only one component of the continuous NF spectrum, such that two identical NNs have to be used to predict the real and imaginary parts of the reflection coefficient. To train the NN, we precomputed 94035 optical signals. 9403 signals were used for validation and excluded from training. The final value of the relative error for the entire validation dataset was less than 0.3%. Our findings highlight the fundamental possibility of using the NNs to analyze and process complex optical signals when the conventional algorithms can fail to deliver an acceptable result.

AB - In this work, we demonstrate that the high-accuracy computation of the continuous nonlinear spectrum can be performed by using artificial neural networks. We propose the artificial neural network (NN) architecture that can efficiently perform the nonlinear Fourier (NF) optical signal processing. The NN consists of sequential convolution layers and fully connected output layers. This NN predicts only one component of the continuous NF spectrum, such that two identical NNs have to be used to predict the real and imaginary parts of the reflection coefficient. To train the NN, we precomputed 94035 optical signals. 9403 signals were used for validation and excluded from training. The final value of the relative error for the entire validation dataset was less than 0.3%. Our findings highlight the fundamental possibility of using the NNs to analyze and process complex optical signals when the conventional algorithms can fail to deliver an acceptable result.

KW - Nonlinear Fourier transform

KW - Neural network

KW - Optical communication

KW - Signal processing

UR - https://www.spiedigitallibrary.org/conference-proceedings-of-spie/11843/2594127/Computing-continuous-nonlinear-Fourier-spectrum-of-optical-signal-with-artificial/10.1117/12.2594127.full

U2 - 10.1117/12.2594127

DO - 10.1117/12.2594127

M3 - Conference article

SN - 0277-786X

VL - 11843

JO - Proceedings of SPIE - International Society for Optical Engineering

JF - Proceedings of SPIE - International Society for Optical Engineering

M1 - 118430J

T2 - Applications of Machine Learning 2021

Y2 - 1 August 2021 through 5 August 2021

ER -

Computing continuous nonlinear Fourier spectrum of optical signal with artificial neural networks

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this