On Sampling Stationary Stochastic Processes

Let \(\{X(t),t\in\mathbb{R}\}\) be a mean square continuous real stationary process with finite second moments and vanishing mean. If the spectral measure \(\mu\) of the process is supported in \([-\pi,\pi]\) we say that the process is bandlimited to \([-\pi,\pi]\). The authors show that for essentially bounded spectral densities of a bandlimited process there is an isometry between a weighted \(L^2\)-space and the Hilbert space spanned by the process. They prove some versions of Whittaker-Shannon-Kotelnikov sampling theorem, e.g.~that under conditions mentioned above the expansion \(X(t)= \sum_{-\infty}^{\infty} X(n) [\sin\pi(t-n)]/[\pi(t-n)]\) holds in the \(L^2\)-sense for each \(t\in\mathbb{R}\). Generally speaking, expansions for the \(e^{it\omega}\) in the \(L^2\)-space are transferred into sampling expansions of the process \(X(t)\) converging in the mean square sense.

Recommendations

Cites work

Cited in

(13)

This page was built for publication: On Sampling Stationary Stochastic Processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2729466)