UDC 621.37:51-74
PARAMETRIC SPECTRAL ANALYSIS OF UNIMODAL SPECTRUM FOR NOISY SIGNALS
V. G. Andreyev, PhD (technical sciences), professor, RSREU, Ryazan, This email address is being protected from spambots. You need JavaScript enabled to view it.
H. L. Tran, post-graduate student, RSREU, Ryazan, This email address is being protected from spambots. You need JavaScript enabled to view it.
We proposed and investigated a method of restoring autocorrelation coefficients of discrete autocorrelation function of random signals with unimodal spectrum and an unknown form of power spectral density. The aim is to develop methods for improvement of accuracy of spectral estimation of signals with a priori uncertainty of their shapes of unimodal spectrum. The method is based on finding the optimum value of α weighting factor, which characterizes Gaussian shape and (1−α) for resonance shape of unimodal spectrum envelope. Experiments show that the proposed approach makes it possible to reduce 4...10 times the discrepancy between control and model spectrums as compared to the known methods of spectral analysis, in particular, AR method. Increasing the adequacy of spectral estimation makes it possible to reduce 4 ... 5 times the length of the time sample while maintaining achievable accuracy of spectral estimation by other known parametric methods. Increasing of adequacy is achieved through the use of a priori information about spectral properties of signals.
Key words: spectrum, spectral estimation, restoration of autocorrelation coefficients, autoregressive model, autoregression, power spectral density, MUSIC.