A decoupling Kalman filtering technique for optimal estimation of Markov chains

摘要

An efficient decoupling Kalman filtering technique developed by K. Chui and G. Chen (1987) is applied to certain Markov chains with finite-dimensional stationary state-transition matrices. For optimal estimation of a Markov chain with an n*n stationary state-transition matrix, the resultant computational algorithm consists of only n-1 simple one-dimensional recursive formulas.