- 48 views
- 63 downloads
Signal Processing for Impairment Control in Fiber Optic Communication
-
- Author / Creator
- Abolfathimomtaz, Abbas
-
Fiber optics is the only infrastructure capable of handling the large data traffic in the backend of many modern communication systems. However, the higher growth rate of data traffic compared to that of fiber system capacity raises significant concerns about meeting future demands. Recognizing this potential shortage, many researchers have recently directed their efforts toward increasing fiber capacity.
Fiber capacity is primarily limited by system impairments, which are rooted either in Kerr nonlinearity or the suboptimality of the transmitter/receiver. In the first half of our work, we focus on controlling Kerr nonlinearity, while the second half is dedicated to addressing another impairment occurring due to the suboptimal receiver structure.
Kerr nonlinearity is a power-dependent impairment that introduces nonlinear interference noise (NLIN). Based on existing fiber models, Kerr nonlinearity depends on the statistical properties of the power launched into the system; therefore, by controlling the launched power, one can reduce NLIN. One of the main factors impacting signal power behavior in a communication system is the constellation from which symbols are drawn. In the first completed work, we formulate the impact of power fluctuation on NLIN with respect to the launched symbols. We then demonstrate that by properly pairing the transmitted symbols in the wavelength division multiplexing (WDM) channels, we can minimize power fluctuation and consequently reduce fiber NLIN.
In the next completed work, we design the signal power spectral density (PSD) to minimize NLIN. Fiber models express NLIN power based on the PSD of the signal launched into the fiber. Thus, a rational question is: what is the optimal PSD that minimizes NLIN? By formulating the problem as an optimization problem, we find the optimal PSD that minimizes system NLIN, which differs from that achieved by the commonly used raised cosine pulse.
The suboptimal structure of the existing transmitter/receiver also limits fiber capacity. One of these suboptimalities is the order of the phase noise and dispersion compensators. While an optimal receiver should compensate for receiver laser phase noise before dispersion compensation, the lack of phase information before the dispersion compensator forces the incorrect compensation order. This suboptimal order gives rise to equalization-enhanced phase noise (EEPN), which restricts system performance at high baud rates and over long distances. To address the problem of EEPN, in the third completed work, we introduce a new formulation of EEPN that enables its compensation. We then provide two digital signal processing (DSP)-ready implementations of this compensator applicable to different receiver structures.
In the final completed work, we design a carrier phase estimator (CPE) that can extract receiver laser phase noise information before dispersion compensation. Our design leverages the fact that fiber chromatic dispersion causes the positive and negative excess bandwidths of the pulse to propagate at different velocities. This velocity mismatch, along with the fact that both positive and negative excess bandwidths are modulated with the same information, enables us to observe the receiver laser phase variation over time after proper signal processing. Our proposed CPE prevents the generation of EEPN by allowing the optimal sequence of dispersion and receiver laser phase noise compensations.
-
- Graduation date
- Fall 2024
-
- Type of Item
- Thesis
-
- Degree
- Doctor of Philosophy
-
- License
- This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.