Return-difference and spectral factorisation relationship for the discrete-time Kalman filter (Q1057024)

In this paper a number of covariance factorisation/return difference results are derived for optimal discrete-time time-varying filtering problems. The results presented are interesting extensions of similar results known to hold in the time-invariant systems case. Using these results a time-varying polynomial operator approach is proposed for calculating the Kalman gain matrix for a time-varying ARMA model. The methods and results of this paper are of potential interest for both control and estimation applications.