An innovations approach to least squares estimation--Part IV: Recursive estimation given lumped covariance functions
T. Kailath, R. Geesey · IEEE Transactions on Automatic Control · 1971
We show how to recursively compute linear least squares filtered and smoothed estimates for a lumped signal process in additive white noise. However, unlike the Kalman-Bucy problem, here only the covariance function of the signal process is known and not a specific state-variable model. The solutions are based on the innovations representation for the observation process.