Perturbation-Assisted DBP for Nonlinear Compensation in Polarization Multiplexed Systems

We investigate an improved digital back propagation (DBP) method based on a multi-stage perturbation theory to compensate for nonlinear impairments in fiber optic links. In the standard DBP, the nonlinear operation only calculates the phase shift based on the power of the current signal sample. In t...

Full description

Saved in:
Bibliographic Details
Published in:IEEE photonics technology letters Vol. 29; no. 21; pp. 1812 - 1815
Main Authors: Xiaojun Liang, Downie, John D., Hurley, Jason E.
Format: Journal Article
Language:English
Published: IEEE 01.11.2017
Subjects:
Coherent optical communications
digital back propagation
nonlinear compensation
Nonlinear distortion
Optical distortion
Optical mixing
Optical polarization
perturbation theory
Q-factor
ISSN:1041-1135, 1941-0174
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We investigate an improved digital back propagation (DBP) method based on a multi-stage perturbation theory to compensate for nonlinear impairments in fiber optic links. In the standard DBP, the nonlinear operation only calculates the phase shift based on the power of the current signal sample. In the proposed perturbation-assisted DBP (P-DBP) method, the nonlinear operation calculates nonlinear phase shifts and complex distortions taking into account the correlation among neighboring optical pulses, which significantly improves the accuracy of nonlinear calculation. We obtain analytical expressions to calculate the nonlinear phase shifts due to the self-phase modulation and intra-channel cross-phase modulation effects and the complex distortions due to the intra-channel four wave mixing effect. We experimentally demonstrate that in a 256-Gb/s polarization multiplexed system over a 2,000-km fiber optic link, the P-DBP method using 4 steps shows the same compensation performance as standard DBP using 40 steps. We perform algorithm complexity analysis and show that the P-DBP method reduces the computational complexity up to 82% as compared with the standard DBP for the same gain.
ISSN:1041-1135
1941-0174
DOI:10.1109/LPT.2017.2752360