ABSTRACT
Estimating the power spectrum, given a long random sequence {x(n)}, can be accomplished by, for example, creating its autocorrelation function and then taking the Fourier transform (FT) of it. However, several problems appear in establishing power spectra densities. First, the sequence may not be long enough and some of the times can be very short. Second, the spectra characteristics may change with time. Third, data very often are corrupted with noise. Therefore, the spectrum estimation is a problem that involves estimating S x ( e j ω ) https://s3-euw1-ap-pe-df-pch-content-public-u.s3.eu-west-1.amazonaws.com/9781315112855/573894e1-85cd-4b1a-afda-92257f939f4b/content/inline8_1.tif"/> from a finite noisy measurements of a sequence, {x(n)}.
