DocumentCode
3222
Title
Approximations for the Nonlinear Self-Channel Interference of Channels With Rectangular Spectra
Author
Savory, Seb J.
Author_Institution
UCL Electron. & Electr. Eng., Univ. Coll. London, London, UK
Volume
25
Issue
10
fYear
2013
fDate
15-May-13
Firstpage
961
Lastpage
964
Abstract
The Gaussian noise (GN) model, in which the fiber nonlinearity is modeled as an additive GN process, has been recently shown in the literature to be accurate for uncompensated coherent systems. Nevertheless, it does not have an exact analytical solution requiring analytical approximations to be made. Herein, we propose a new means of approximating the nonlinear self-channel interference (SCI) in the GN model, for the case of ideal Nyquist WDM channels that have rectangular spectra, bandlimited to the Nyquist bandwidth. We begin by introducing the method to estimate the peak power spectral density of the nonlinear interference before applying it to calculating the total SCI noise of a channel. The analytical solution is compared with the previously reported approximation and the exact numerical solution, to quantify the approximation error. The proposed approximation is accurate to within 0.3 dB of the GN model as the symbol rate is varied from 10 to 100 GBd. Finally, we demonstrate that for a superchannel, the total nonlinear interference for the central channel can be approximated to within 0.3 dB for three or more channels.
Keywords
Gaussian noise; channel estimation; light interference; nonlinear optics; optical fibre communication; wavelength division multiplexing; Gaussian noise model; additive GN process; analytical solution; approximation error; bandlimited Nyquist bandwidth; central channel; fiber nonlinearity; ideal Nyquist WDM channels; nonlinear self-channel interference; peak power spectral density; rectangular spectra; superchannel; symbol rate; total SCI noise; Approximation methods; GN model; optical fiber communication;
fLanguage
English
Journal_Title
Photonics Technology Letters, IEEE
Publisher
ieee
ISSN
1041-1135
Type
jour
DOI
10.1109/LPT.2013.2255869
Filename
6491442
Link To Document