zbMATH — the first resource for mathematics

An algorithm for eliminating microerrors of noise in the solution of statistical dynamics problems. (English. Russian original) Zbl 1079.93525
Autom. Remote Control 59, No. 5, Part 2, 679-688 (1998); translation from Avtom. Telemekh. 1998, No. 5, 82-94 (1998).
Summary: To improve the estimates of autocorrelation and mutual correlation functions and the conditionality of correlation matrices, an algorithm is designed for determining microerrors introduced by noise into the products of readouts of random signals. It is shown that the sum of microerrors with due regard for their signs forms a common error of obtained estimates. A formula is put forward that makes it possible to exclude the value of the variance of noise from the common error of estimates of correlation functions both in the case where the correlation between noise and a useful signal is equal to zero and in the case where it is different from zero.

93E10 Estimation and detection in stochastic control theory