On the estimation of some functionals of the spectral density function of Gaussian random processes

The paper deals with the estimation of some nonlinear functionals of the spectral density function of a Gaussian stationary discrete time random process. The spectral density function being smooth enough, the mean square error of the estimates proposed here is shown to be $O(n^{-1})$, as $n\to\infty$, where $n$ is the sample size.