Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization

Elias Giacoumidis; Son T. Le; Mohammad Ghanbarisabagh; Mary McCarthy; Ivan Aldaya; Sofien Mhatli; Mutsam A. Jarajreh; Paul A. Haigh; Nick J. Doran; Andrew D. Ellis; Benjamin J. Eggleton

doi:10.1364/OL.40.005113

Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization

Elias Giacoumidis^*, Son T. Le, Mohammad Ghanbarisabagh, Mary McCarthy, Ivan Aldaya, Sofien Mhatli, Mutsam A. Jarajreh, Paul A. Haigh, Nick J. Doran, Andrew D. Ellis, Benjamin J. Eggleton

^*Corresponding author for this work

Aston Institute of Photonic Technologies (AiPT)

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

Abstract

We experimentally demonstrate ∼2 dB quality (Q)-factor enhancement in terms of fiber nonlinearity compensation of 40 Gb/s 16 quadrature amplitude modulation coherent optical orthogonal frequency-division multiplexing at 2000 km, using a nonlinear equalizer (NLE) based on artificial neural networks (ANN). Nonlinearity alleviation depends on escalation of the ANN training overhead and the signal bit rate, reporting ∼4 dB Q-factor enhancement at 70 Gb/s, whereas a reduction of the number of ANN neurons annihilates the NLE performance. An enhanced performance by up to ∼2 dB in Q-factor compared to the inverse Volterra-series transfer function NLE leads to a breakthrough in the efficiency of ANN.

Original language	English
Pages (from-to)	5113-5116
Number of pages	4
Journal	Optics Letters
Volume	40
Issue number	21
Early online date	30 Sept 2015
DOIs	https://doi.org/10.1364/OL.40.005113
Publication status	Published - 30 Oct 2015

Access to Document

10.1364/OL.40.005113

Cite this

@article{0f968725822c4f2c9a4e36573a65571f,

title = "Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization",

abstract = "We experimentally demonstrate ∼2 dB quality (Q)-factor enhancement in terms of fiber nonlinearity compensation of 40 Gb/s 16 quadrature amplitude modulation coherent optical orthogonal frequency-division multiplexing at 2000 km, using a nonlinear equalizer (NLE) based on artificial neural networks (ANN). Nonlinearity alleviation depends on escalation of the ANN training overhead and the signal bit rate, reporting ∼4 dB Q-factor enhancement at 70 Gb/s, whereas a reduction of the number of ANN neurons annihilates the NLE performance. An enhanced performance by up to ∼2 dB in Q-factor compared to the inverse Volterra-series transfer function NLE leads to a breakthrough in the efficiency of ANN.",

author = "Elias Giacoumidis and Le, {Son T.} and Mohammad Ghanbarisabagh and Mary McCarthy and Ivan Aldaya and Sofien Mhatli and Jarajreh, {Mutsam A.} and Haigh, {Paul A.} and Doran, {Nick J.} and Ellis, {Andrew D.} and Eggleton, {Benjamin J.}",

year = "2015",

month = oct,

day = "30",

doi = "10.1364/OL.40.005113",

language = "English",

volume = "40",

pages = "5113--5116",

journal = "Optics Letters",

issn = "0146-9592",

publisher = "OSA Publishing",

number = "21",

}

TY - JOUR

T1 - Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization

AU - Giacoumidis, Elias

AU - Le, Son T.

AU - Ghanbarisabagh, Mohammad

AU - McCarthy, Mary

AU - Aldaya, Ivan

AU - Mhatli, Sofien

AU - Jarajreh, Mutsam A.

AU - Haigh, Paul A.

AU - Doran, Nick J.

AU - Ellis, Andrew D.

AU - Eggleton, Benjamin J.

PY - 2015/10/30

Y1 - 2015/10/30

N2 - We experimentally demonstrate ∼2 dB quality (Q)-factor enhancement in terms of fiber nonlinearity compensation of 40 Gb/s 16 quadrature amplitude modulation coherent optical orthogonal frequency-division multiplexing at 2000 km, using a nonlinear equalizer (NLE) based on artificial neural networks (ANN). Nonlinearity alleviation depends on escalation of the ANN training overhead and the signal bit rate, reporting ∼4 dB Q-factor enhancement at 70 Gb/s, whereas a reduction of the number of ANN neurons annihilates the NLE performance. An enhanced performance by up to ∼2 dB in Q-factor compared to the inverse Volterra-series transfer function NLE leads to a breakthrough in the efficiency of ANN.

AB - We experimentally demonstrate ∼2 dB quality (Q)-factor enhancement in terms of fiber nonlinearity compensation of 40 Gb/s 16 quadrature amplitude modulation coherent optical orthogonal frequency-division multiplexing at 2000 km, using a nonlinear equalizer (NLE) based on artificial neural networks (ANN). Nonlinearity alleviation depends on escalation of the ANN training overhead and the signal bit rate, reporting ∼4 dB Q-factor enhancement at 70 Gb/s, whereas a reduction of the number of ANN neurons annihilates the NLE performance. An enhanced performance by up to ∼2 dB in Q-factor compared to the inverse Volterra-series transfer function NLE leads to a breakthrough in the efficiency of ANN.

UR - http://www.scopus.com/inward/record.url?scp=84957539304&partnerID=8YFLogxK

U2 - 10.1364/OL.40.005113

DO - 10.1364/OL.40.005113

M3 - Article

AN - SCOPUS:84957539304

SN - 0146-9592

VL - 40

SP - 5113

EP - 5116

JO - Optics Letters

JF - Optics Letters

IS - 21

ER -

Fiber nonlinearity-induced penalty reduction in CO-OFDM by ANN-based nonlinear equalization

Abstract

Access to Document

Other files and links

Fingerprint

Cite this