Multitap complex Kalman filter for handling large PMD and fast RSOP

Kalman filter, as an efficient recursive algorithm, can effectively estimate the system state by combining prediction and measurement information, and has achieved good results in the field of fiber optic communication. We propose a multi tap complex Kalman filter algorithm based on the characteristics of RSOP and PMD acting in the time and frequency domains, respectively. By complicating the state vector, it can simultaneously handle joint damage of up to 20Mrad/s highspeed RSOP and DGD=100ps large PMD; And by diagonalizing the Kalman gain matrix during the observation vector iteration process, the goal of simplification is achieved, reducing its complexity compared to the original multi tap complex Kalman.