Polarization Demultiplexing Based On Kalman Filter In Coherent Optical Communication Systems

Polarization division multiplexing(PDM)technology with high-order modulation format has been widely used in coherent optical communication systems.This technology can double the capacity of the systems without much additional costs.In order to recover the original information,polarization demultiplexing(PolDemux)technology should be applied.PolDemux algorithms based on Stokes Space(SS)are transparent to modulation formats,which makes it easier to applied to higher-order modulation formats.Meanwhile,Kalman filter(KF)based algorithms are robust to noise.In this work,a PolDemux method based on adaptive Kalman filter(AKF)in SS is proposed,in which the process noise covariance has been estimated adaptively.By using AKF,tedious and time consuming parameter-by-parameter tuning method can be avoided.Compared with conventional EKF,faster convergence and better estimation accuracy of the proposed method are confirmed by both numerical simulations and experiments.The main works are listed as following:Firstly,the basic theories of PDM coherent optical communication system are briefly introduced.Meanwhile,some impairments in PDM systems,such as chromatic dispersion(CD),birefringence effect and polarization mode dispersion(PMD)are analyzed as well.Subsequently,a PolDemux method based on AKF in SS for coherent optical communication system is proposed.Then,a 28 GBaud PDM-QPSK/16QAM coherent optical communication system is built based on the VPItransmissionMaker~TMM software platform for verifying the feasibility of the method.Subsequently,the PolDemux performance of four PolDemux methods(i.e.CMA/LS-PF/EKF/AKF)are analyzed.Meanwhile,the effects of residual CD and PMD are also investigated.Finally,base on the optical communication experimental platform,a 16 Gbaud PDM-16QAM coherent optical communication system is set up.The performance of conventional EKF and proposed AKF methods is tested and verified over different transmission distance(i.e.back-to-back and 560 km).The experimental results show that the proposed method may significantly improve the adaptive capability of EKF by adaptively estimating the process noise covariance.By using AKF,tedious and time consuming parameter-by-parameter tuning method can be avoided.The convergence is also found to be about 4 times quicker than that of conventional EKF.Meanwhile,the polarization tracking stability of the proposed method is also experimentally confirmed.