Abstract
Techniques for reliably estimating the power spectral density function for both small and large samples of a stationary stochastic process are described. These techniques have been particularly successful in cases where the range of the spectrum is l...