Loading...
Optical Communication Impairments Mitigation Using Interpretable Machine Learning
Karami, Mahyar | 2023
97
Viewed
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 56337 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Pakravan, Mohammad Reza; Hadi, Mohammad
- Abstract:
- Optical impairments, as a challenging obstacle in optical communication systems, have been under deep attention of optical engineers in recent years. These impairments can be divided into two categories of linear and nonlinear impairments. Dispersion is a significant linear impairment leads to broadening of the optical signal and interference between transmitted symbols during detection. Nonlinear impairments such as cross-phase modulation (XPM) and self-phase modulation (SPM) also distort the transmitted signal through the fiber, causing interference at the receiver and degrading the system's performance. An appropriate solution for compensating optical fiber impairments is to utilize machine learning-based compensators. A common aspect of most research conducted in this field is the use of machine learning models with non-interpretable parameters, which makes the adjustment of the compensator untractable and complex. Considering the analytical complexity of the optical fiber impairments, a complex high dimensional artificial neural network is required to compensate fiber impairments with an acceptable performance. In this research, the method of deep unfolding is employed to include interpretability into the compensator structure. In this regard, two methods, called IDUIM and N-IDUIM, are proposed, which are dedicated to mitigating linear impairments and simultaneous mitigation of linear and nonlinear impairments, respectively. The proposed methods are general enough to be used for impairment compensation in various optical communication systems. To assess this, the performance of the proposed methods for mitigating linear impairments is validated on for optical and microwave communication systems. We also evaluate the proposed methods for simultaneous mitigation of linear and nonlinear impairments in a nonlinear long-haul optical communication system. The simulations performed to evaluate proposed models demonstrate their better performance compared to conventional methods that solely rely on analytical models or black-box machine learning models
- Keywords:
- Interpretable Machine Learning ; Optical Networks ; Impairment Mitigation ; Optical Communication Networks ; Self-Phase Modulation