Methods for Nonlinear Impairments Compensation in Fiber-Optic Communication Systems

Abstract

Fiber optic links are the backbone of the current high-speed tele- and data communi- cation networks. Two major factors which cause degradation in the performance of this type of communication systems are ber optic link nonlinear impairments, self-phase and cross-phase modulation, and linear additive noise. To tackle with linear noise, the average energy of constellation points should be increased and this increment also increases the nonlinear impairments. This research explores some possible methods of tackling the is- sues caused by nonlinear impairments without changing the average energy of constellation points.

In chapter 1, we will review mathematical models introduced in literature for modelling nonlinearities in ber optic links. In Chapter 2, we will introduce a simpli ed model for modelling nonlinearities which can help us with compensating procedure by reducing computational complexity and providing feasibility for parallel computations. In Chapter 3, a novel method called Masking Data Samples will be introduced which can reduce the power of self-phase modulation by sampling from that, which is a random variable, and choosing the best sample in sense of SPM noise between all obtained ones. In Chapter 4, a modi ed version of known tree-structure codes will be introduced which can reduce the power of SPM noise by a new sorting method in the phase of producing look-up table for codes. At the last chapter, a method for exploiting the statistical characteristics and memory of the samples of cross-phase modulation noise will be introduced which can help us with detecting data symbols more accurately.