<p> Space-division multiplexing (SDM) has been extensively proposed to overcome the next capacity crunch with ever-increasing data and video traffic. Among several SDM approaches, mode-division-multiplexing (MDM) in few-mode fiber (FMF) is the most auspicious technology. One key challenge in FMF transmission systems is random mode coupling among different fiber modes, which can cause severe inter-modal crosstalk. Moreover, large accumulated differential mode group delay (DMGD) can induce significant inter-symbol interference (ISI). </p><p> The approach of adaptive multi-input multi-output (MIMO) digital signal processing (DSP) has been proposed and demonstrated to untangle the crosstalk between the spatial modes and compensate the DMGD. In FMF systems, compared with time-domain adaptive MIMO signal processing, the implementation of frequency domain method achieves much lower hardware complexity. In this dissertation, a single-stage adaptive MIMO equalizer is proposed to compensate both DMGD and chromatic dispersion (CD) simultaneously in order to further reduce the hardware complexity. </p><p> Except for hardware complexity, the convergence rate of adaptive MIMO equalizer is another essential concern. The adaptive MIMO equalizer with slower convergence speed requires longer training symbols, thus decreasing the system overall efficiency. In the dissertation, two advanced step size control methods are presented to increase the convergence rate of the conventional FD-LMS algorithm. The first approach is the signal power spectral density (PSD) dependent method, which adopts the step size for each frequency bin inverse to its power level in order to converge the estimated equalization error to zero, thus it is the optimal solution in the systems with noise-free channel. The other method is the noise PSD directed method, which adopts the frequency bin-wise step size to render the estimated error converge to the channel background noise, thus it is the optimum solution in the systems with additive white Gaussian noise channel.</p>
| Identifer | oai:union.ndltd.org:PROQUEST/oai:pqdtoai.proquest.com:3687684 |
| Date | 07 April 2015 |
| Creators | He, Xuan |
| Publisher | University of Louisiana at Lafayette |
| Source Sets | ProQuest.com |
| Language | English |
| Detected Language | English |
| Type | thesis |
Page generated in 0.0095 seconds