Research On Kalman Filter Based Polarization-state And Carrier Recovery Of Optical Signal

The combination of advanced modulation formats, polarization division multiplexing(PDM), coherent detection and digital signal processing has attracted a lot attentions due to their potential of improving spectral efficiency and transmission reach in next-generation long-haul optical fiber communication networks. The researchers' enthusiasm for the study of the digital signal processing algorithm has never waned since high performance signal processing algorithm can greatly improve the optical receiver's BER performance. Based on present research hot spot, we focus our research topic on the kalman filter technology which can be used for polarization-state tracking, frequency offset estimation and carrier phase recovery in the digital signal processing of coherent receiver. The performance of digital signal processing based Kalman filter in coherent receiver is numerically or experimentally studied in above three application environment.Considering polarization-state restoration technique, first, we extended the Kalman measurement equation from one, which is used for polarization and carrier phase tracking of polarization division multiplexing QPSK,into two which can be utilized for 16QAM's polarization and carrier phase tracking. Additionally, we discussed the influence of polarization-state rotation, phase noise, residual frequency offset, residual chromatic dispersion, polarization mode dispersion and tuning parameters to the extended Kalman filter. Secondly, in order to solve the problem of frequency offset dependence of above extended Kalman filter, this paper proposed a radius directed linear kalman filter for polarization-state tracking. The single-measurement equation filter is used for PDM-QPSK, and the dual-measurement equation filter is used for PDM-16 QAM, which is parallel with above extended Kalman filter scheme. It's found that the method also has fast convergence rate, excellent tracking capability and the advantages of easier implementation. Finally, with the deepening of the research, we found that the linear kalman filter polarization tracking method for PDM-QPSK can be directly applied in PDM-16 QAM, PDM-64 QAM and PDM time domain QPSK/16 QAM signal by adjusting the tuning parameters without changing the filter structure. That means the algorithm is “completely blind track” and has a less computing complexity.For frequency offset estimation, we proposed and numerically compared the “angle approach” and “point approach” Kalman filter based frequency offset estimation in the case of training sequence method and M-th power method to remove the modulation phase, respectively. The traditional Viterbi & Viterbi algorithm and Morelli frequency peak analysis algorithm are also given into account as comparison.In view of carrier phase recovery, the Kalman filter based QPSK and 16 QAM signal carrier phase recovery structure are studied. It turns out that the double-measurement equation Kalman filter scheme has batter estimation accuracy than the single-measurement equation scheme for 16 QAM signal. Finally, we examine the estimation accuracy and laser linewidth tolerance of Kalman filter.Results in this thesis promote Kalman filter's research progress and have practical reference significance in the area of digital signal processing of coherent optical communication. The proposed kinds of Kalman filter are especially suitable to deal with rapidly polarization-state time-varying signals, accurate metrology of polarization-state, frequency offset and phase noise, and burst-mode coherent receiver that only samples short segments of received data for signal recovery or monitoring.