Candidate: Hadi NekoeiQachkanloo
Title: Nonlinear Impairment Compensation in Fiber-Optic Systems Using Iterative Decoding
Date: August 15, 2019
Place: EIT 3142
Supervisor(s): Khandani, Amir K.
Optical communication systems are vital for high rate telecommunication. Fiber-optic communication system is an excellent choice due to its low loss, high bandwidth, and robustness to electromagnetic interference. However, fiber-optic links suffer from linear and nonlinear impairments which limit their performance. Digital signal processing techniques can be used to compensate for the linear impairments. On the other hand, nonlinear impairment is much harder to tackle.
There exist two main nonlinear noises which are caused by Kerr effect. Each channel in the fiber-optic link has two poles namely Xpole and Ypole. In a single channel case, transmitted signal over each pole generates intensity-dependent noise on both poles which is called Self Phase Modulation (SPM) noise. On the other hand, when multiple signal channels co-propagate in a single fiber, the power fluctuations of one signal channel cause a phase shift to another channel, which is due to the Cross Phase Modulation (XPM) effect.
Through this thesis, our main contribution is using LDPC-Coded Modulation with Iterative Damping and Decoding to improve bit probabilities. The decoding system can be modified to accept a priori information which helps us to remove nonlinear noise using demapping. In addition, we propose a joint detection method to compensate for SPM noise. In this method, we exploit two main statistical characteristics of noise samples which are space domain and time domain correlations to improve naive minimum distance detection. Finally, we report the experimental results.
200 University Avenue West
Kitchener, ON N2L 3G1