Truncated estimating parameters of additive mixture of radio signal and kurtosis non-Gaussian noise

Abstract

Introduction. Classical methods based on the use of Gaussian random signal model has its advantages and disadvantages. Therefore, Maximum Likelihood Method have not found wide implementation due to the high computational complexity. Method of Moments does not have the properties of asymptotic optimality, although it leads to a relatively simple calculations. In general, the methods do not consider more complex structure of real noise. Therefore the accuracy of signal processing algorithms may be insufficient. The aims and objectives of research. The aim of the paper is to adapt Methods of Polynomial Maximization (MPM) and Truncated Stochastic Polynomial Maximization (MTSPM) for joint estimation of radiosignal and kurtosis non-Gaussian noise parameters. The Objectives of research is to develop effective methods of statistical data processing, which would allow increasing the accuracy and speed of signal processing. Construction of the Polynomial algorithms for joint estimating. To find joint estimates the systems of equations are constructed. To estimate the radiosignal frequency is used MPM and the noise variance – MTSPM. Statistical properties of the radiosignal frequency estimates. To study the statistical properties of radiosignal parameter estimates the asymptotic dispersions of estimates are calculated. Comparison of the asymptotic dispersion of radiosignal frequency estimates and a graphical representation of the results. The efficiency of polynomial estimation algorithms increases with the stochastic polynomial degree and as the values of coefficients of kurtosis approach the tolerance range limit. Conclusion. The effective methods of signal processing to enhance the accuracy and speed of non-Gaussian signals processing are developed. The results can be used to improve the estimation accuracy of radiosignal parameters in radiolocation, radio navigation and other areas, where the accuracy of signal processing algorithms plays an important role.