We study the optimal filtering problem for time-varying systems with correlated noises and uncertain observations.
We design an optimal Kalman filtering algorithm in the sense of minimum mean square error (MMSE).
We propose a new Kalman filtering algorithm which is suitable for online applications.