Noise Adaptive Kalman Filter for Joint Polarization Tracking and Channel Equalization Using Cascaded Covariance Matching

We propose a noise adaptive Kalman filter for joint polarization tracking and channel equalization using cascaded covariance matching. With the process noise covariance (Q) and the measurement noise covariance (R) estimated in a cascaded manner, the proposed Kalman filter scheme can be implemented in adaptive mode. The experimental results demonstrate that in wide range of OSNR and polarization rotation, the scheme can adaptively configure noise covariance parameters and consequently can achieve optimal performance even in long transmission cases. Besides, the scheme is more robust against the initial errors in Q and R compared to nonadaptive filter and R /Q-adaptive algorithms with the allowable initial values of Q and R within 6 and 9 magnitudes, respectively.