Abstract

Prediction error filtering has been widely used for deconvolution. The mean squared error in prediction is a monotonically nonincreasing function of operator length, and the value of the error is readily available from the Wiener-Levinson algorithm. In general, the value of this error for the infinitely long operator is not known a priori. It is shown that the final value of the error can be obtained by considering the Kolmogorov spectrum factorization. Simple criteria can then be established for operator effectiveness and length.

First Page Preview

First page PDF preview
You do not currently have access to this article.