University of British Columbia
Interplay between fiber nonlinearity and probabilistic amplitude shaping
Abstract
dc:descriptionThe maximum information rate of a communication channel, often referred to as the Shannon capacity, is achieved by optimizing the distribution of the transmitted symbols. Probabilistic constellation shaping is one principled approach to perform this optimization. Its fairly recently developed variant known as Probabilistic Amplitude Shaping (PAS) is of particular interest as it combines constellation shaping and error-correction coding. Conventional shaping methods, including PAS, are typically designed for the Additive White Gaussian Noise (AWGN) channel, or more generally, for linear channels. However, in optical fiber communications, the main bottleneck for increasing data rate is the so-called Kerr nonlinear effect, establishing an overall nonlinear channel. Therefore, the proper understanding of the nonlinear behavior of the optical channel and the development of nonlinearity tolerant PAS have been active areas of research. This thesis revisits the interplay between channel nonlinearity and PAS. We derive an effective linear channel model for the interaction of transmitted symbol energy sequences and the nonlinear distortion. Based on this, we introduce a visualization tool for nonlinearity analysis which permits us to explain several phenomenological observations reported in previous works. Furthermore, we develop a new nonlinearity tolerant PAS scheme, which outperforms state-of-the-art methods by a significant margin.
Degree
thesis:*- Name thesis:degree_name
- Master of Applied Science - MASc
- Level thesis:degree_level
- master's
- Discipline thesis:degree_discipline
- Electrical and Computer Engineering
- Grantor dc:publisher
- University of British Columbia
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Askari, Mohammad Taha
- Contributors dc:contributor
-
- Lampe, Lutz
Rights
dc:rights- Statement dc:rights
-
- Attribution-NonCommercial-NoDerivatives 4.0 International
- Language dc:language
- eng
Identifiers
dc:identifier.*- Handle dc:identifier
- http://hdl.handle.net/2429/82624