Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization

Elias Giacoumidis; Jinlong Wei; Sofien Mhatli; Marc F. Stephens; Nick J. Doran; Andrew  Ellis; Benjamin  Eggleton

doi:10.1364/OFC.2017.W3J.2

Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization

Elias Giacoumidis
, Jinlong Wei
, Sofien Mhatli
, Marc F. Stephens
, Nick J. Doran
, Andrew Ellis
, Benjamin Eggleton

Aston Institute of Photonic Technologies (AiPT)

University of Sydney
EPT Université de Carthage
Huawei Technologies Duesseldorf GmbH

Research output: Chapter in Book/Published conference output › Conference publication

10 Citations (SciVal)

Abstract

We experimentally demonstrate a Newton support vector machine (N-SVM)-based nonlinear equalizer (NLE) of reduced classifier complexity for 40-Gb/s 16-QAM CO-OFDM. At 2000-km N-SVM extends the launched optical power by 2-dB compared to Volterra-based NLE.

Original language	English
Title of host publication	Optical Fiber Communication Conference 2017
DOIs	https://doi.org/10.1364/OFC.2017.W3J.2
Publication status	Published - 23 Mar 2017
Event	Optical Fiber Communication Conference, OFC 2017 - Los Angeles, United States Duration: 19 Mar 2017 → 23 Mar 2017

Conference

Conference	Optical Fiber Communication Conference, OFC 2017
Country/Territory	United States
City	Los Angeles
Period	19/03/17 → 23/03/17

Access to Document

10.1364/OFC.2017.W3J.2

Cite this

@inproceedings{b9c22e34c00747b9b8b558077d03f32a,

title = "Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization",

abstract = "We experimentally demonstrate a Newton support vector machine (N-SVM)-based nonlinear equalizer (NLE) of reduced classifier complexity for 40-Gb/s 16-QAM CO-OFDM. At 2000-km N-SVM extends the launched optical power by 2-dB compared to Volterra-based NLE.",

author = "Elias Giacoumidis and Jinlong Wei and Sofien Mhatli and Stephens, \{Marc F.\} and Doran, \{Nick J.\} and Andrew Ellis and Benjamin Eggleton",

year = "2017",

month = mar,

day = "23",

doi = "10.1364/OFC.2017.W3J.2",

language = "English",

isbn = "978-1-943580-23-1",

booktitle = "Optical Fiber Communication Conference 2017",

note = "Optical Fiber Communication Conference, OFC 2017 ; Conference date: 19-03-2017 Through 23-03-2017",

}

Giacoumidis, E, Wei, J, Mhatli, S, Stephens, MF, Doran, NJ , Ellis, A & Eggleton, B 2017, Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization. in Optical Fiber Communication Conference 2017., paper W3J.2, Optical Fiber Communication Conference, OFC 2017, Los Angeles, United States, 19/03/17. https://doi.org/10.1364/OFC.2017.W3J.2

TY - GEN

T1 - Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization

AU - Giacoumidis, Elias

AU - Wei, Jinlong

AU - Mhatli, Sofien

AU - Stephens, Marc F.

AU - Doran, Nick J.

AU - Ellis, Andrew

AU - Eggleton, Benjamin

PY - 2017/3/23

Y1 - 2017/3/23

N2 - We experimentally demonstrate a Newton support vector machine (N-SVM)-based nonlinear equalizer (NLE) of reduced classifier complexity for 40-Gb/s 16-QAM CO-OFDM. At 2000-km N-SVM extends the launched optical power by 2-dB compared to Volterra-based NLE.

AB - We experimentally demonstrate a Newton support vector machine (N-SVM)-based nonlinear equalizer (NLE) of reduced classifier complexity for 40-Gb/s 16-QAM CO-OFDM. At 2000-km N-SVM extends the launched optical power by 2-dB compared to Volterra-based NLE.

UR - https://www.osapublishing.org/abstract.cfm?URI=OFC-2017-W3J.2

UR - https://www.scopus.com/pages/publications/85019962776

U2 - 10.1364/OFC.2017.W3J.2

DO - 10.1364/OFC.2017.W3J.2

M3 - Conference publication

AN - SCOPUS:85025824366

SN - 978-1-943580-23-1

BT - Optical Fiber Communication Conference 2017

T2 - Optical Fiber Communication Conference, OFC 2017

Y2 - 19 March 2017 through 23 March 2017

ER -

Nonlinear inter-subcarrier intermixing reduction in coherent optical OFDM using fast machine learning equalization

Abstract

Conference

Access to Document

Other files and links

Fingerprint

20dB net-gain fiber optical parametric amplification of 18×120Gb/s polarization-multiplexed signals

Experimental investigation of nonlinear signal distortions in ultra-wideband transmission systems

Femtosecond laser inscribed axial long-period fiber gratings in two-mode fiber for efficient optical angular momentum generation

Cite this