Predicting nonlinear reshaping of periodic signals in optical fibre with a neural network

Sonia Boscolo; John M. Dudley; Christophe Finot

doi:10.1016/j.optcom.2023.129563

Predicting nonlinear reshaping of periodic signals in optical fibre with a neural network

Sonia Boscolo, John M. Dudley, Christophe Finot^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

Abstract

We deploy a supervised machine-learning model based on a neural network to predict the temporal and spectral reshaping of a simple sinusoidal modulation into a pulse train having a comb structure in the frequency domain, which occurs upon nonlinear propagation in an optical fibre. Both normal and anomalous second-order dispersion regimes of the fibre are studied, and the speed of the neural network is leveraged to probe the space of input parameters for the generation of custom combs or the occurrence of significant temporal or spectral focusing.

Original language	English
Article number	129563
Journal	Optics Communications
Volume	542
Early online date	5 May 2023
DOIs	https://doi.org/10.1016/j.optcom.2023.129563 https://doi.org/10.1016/j.optcom.2023.129563
Publication status	Published - 1 Sept 2023

Bibliographical note

Keywords

Neural networks
Nonlinear propagation
Optical fibres
Pulse shaping

Access to Document

OC - Finot - 2023_vf_SB
Copyright © 2023. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/
Accepted author manuscript, 1.18 MBLicence: CC BY-NC-ND 4.0

Cite this

@article{cc3cb2c545b24765ba22bd34cbb2ced9,

title = "Predicting nonlinear reshaping of periodic signals in optical fibre with a neural network",

abstract = "We deploy a supervised machine-learning model based on a neural network to predict the temporal and spectral reshaping of a simple sinusoidal modulation into a pulse train having a comb structure in the frequency domain, which occurs upon nonlinear propagation in an optical fibre. Both normal and anomalous second-order dispersion regimes of the fibre are studied, and the speed of the neural network is leveraged to probe the space of input parameters for the generation of custom combs or the occurrence of significant temporal or spectral focusing.",

keywords = "Neural networks, Nonlinear propagation, Optical fibres, Pulse shaping",

author = "Sonia Boscolo and Dudley, {John M.} and Christophe Finot",

note = "Copyright {\textcopyright} 2023. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/",

year = "2023",

month = sep,

day = "1",

doi = "10.1016/j.optcom.2023.129563",

language = "English",

volume = "542",

journal = "Optics Communications",

issn = "0030-4018",

publisher = "Elsevier",

}

TY - JOUR

T1 - Predicting nonlinear reshaping of periodic signals in optical fibre with a neural network

AU - Boscolo, Sonia

AU - Dudley, John M.

AU - Finot, Christophe

PY - 2023/9/1

Y1 - 2023/9/1

N2 - We deploy a supervised machine-learning model based on a neural network to predict the temporal and spectral reshaping of a simple sinusoidal modulation into a pulse train having a comb structure in the frequency domain, which occurs upon nonlinear propagation in an optical fibre. Both normal and anomalous second-order dispersion regimes of the fibre are studied, and the speed of the neural network is leveraged to probe the space of input parameters for the generation of custom combs or the occurrence of significant temporal or spectral focusing.

AB - We deploy a supervised machine-learning model based on a neural network to predict the temporal and spectral reshaping of a simple sinusoidal modulation into a pulse train having a comb structure in the frequency domain, which occurs upon nonlinear propagation in an optical fibre. Both normal and anomalous second-order dispersion regimes of the fibre are studied, and the speed of the neural network is leveraged to probe the space of input parameters for the generation of custom combs or the occurrence of significant temporal or spectral focusing.

KW - Neural networks

KW - Nonlinear propagation

KW - Optical fibres

KW - Pulse shaping

UR - https://www.sciencedirect.com/science/article/pii/S0030401823003103?via%3Dihub

U2 - 10.1016/j.optcom.2023.129563

DO - 10.1016/j.optcom.2023.129563

M3 - Article

SN - 0030-4018

VL - 542

JO - Optics Communications

JF - Optics Communications

M1 - 129563

ER -

Predicting nonlinear reshaping of periodic signals in optical fibre with a neural network

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this