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 :
بازگشت