Minimization of the mean square error of estimating the power spectrum of stochastic process
S. N. Kirillov and O. E. Shustikov
Ryazan
Pages: 82–88 Subsection: ANALYSIS AND SYNTHESIS OF SIGNALS AND IMAGES
Abstract
The problem of optimal estimating the power spectrum of stochastic process in the base of discrete exponential functions is considered. An optimal weight function is synthesized. The function minimizes the mean square error of estimation under a given limitation on the bias error. An adaptive procedure of modifying the power spectrum estimator under mutual cancellation of bias error components is developed. The procedure will increase the accuracy of spectral estimation.
