ABSTRACT
This chapter considers the parametric approach that is based on model parameters rather than on autocovariance function. These models include the autoregressive (AR) model, the moving average (MA) model, and the autoregressive-moving average (ARMA) model. It is assumed that the data are the result of the output of one of these systems with input white noise having finite variance. The basic ideas are to assume that a signal is the result of an AR process and find extra values of the signal using the extrapolation method. If the Eigenvectors of a matrix are linearly independent then the eigenvalues are distinct. The chapter also considers discrete spectra embedded in white noise. The motivation for studying parametric models for spectrums estimation is based on the ability to achieve better power spectral density estimation, assuming that incorporate the appropriate model. The parametric approach has been devised to produce better spectral resolution and better spectra estimation.
