scientific article; zbMATH DE number 1762648
Let \(X_{\Delta}^{(1)}=\{X_{\Delta}^{(1)}(t),\;t\in\mathbb{R}\}\), \(X_{\Delta}^{(2)}=\{X_{\Delta}^{(2)}(t),\;t\in \mathbb{R}\},\;\Delta>0\), be two independent, stationary zero-mean Gaussian processes with common spectral density \(f_{\Delta}=\{f_{\Delta}(\lambda),\;\lambda\in \mathbb{R}\}\). Let \(H\in L_2(\mathbb{R}^2)\) be a response function and let \[ Y_{\Delta}=\int_{\infty}^{\infty}\int_{\infty}^{\infty} H(s_1,s_2)X_{\Delta}^{(1)}(t-s_1)X_{\Delta}^{(2)}(t-s_2)ds_1ds_2,\quad t\in\mathbb{R}, \] be the response of the corresponding bilinear system to the input signal \(\vec{X}_{\Delta}=(X_{\Delta}^{(1)},X_{\Delta}^{(2)})\). Consider the random variable \[ \hat{H}_{\Delta,h,H}(\tau_1,\tau_2)=(c^2N)^{-1}\sum_{n=1}^N Y_{\Delta}(nh)X_{\Delta}^{(1)}(nh-\tau_1)X_{\Delta}^{(2)}(nh-\tau_2) \] as an estimator of \(H(\tau_1,\tau_2)\). The authors prove that \(\hat{H}_{\Delta,h,H}(\tau_1,\tau_2)\) is an asymptotically unbiased estimator of \(H(\tau_1,\tau_2)\) as \(\Delta\to\infty\) at the fixed point \((\tau_1,\tau_2)\). An upper bound for the absolute value of the bias is proposed and the mean square consistency of the estimator is proved. Properties of the random field \[ \hat{Z}_{\Delta,h.N}(\vec{\tau})=\sqrt{Nh/\|f_{\Delta}\|_1} \left[\hat{H}_{\Delta,N}(\vec{\tau})-E\hat{H}_{\Delta,N}(\vec{\tau})\right] \] are investigated. For more details see \textit{Valerii V. Buldygin} and \textit{Fu Li}, Theory Stoch. Process. 3(19), No. 1-2, 35-54 (1997; Zbl 0927.62095).
- On cross-correlogram IRF’s estimators of two-output systems in spaces of continuous functions
- ESTIMATION FOR THE FIRST-ORDER DIAGONAL BILINEAR TIME SERIES MODEL
- scientific article; zbMATH DE number 1366685
- On asymptotic behaviour of cross-correlogram estimators of response functions in linear Volterra systems
- Estimation of impulse response functions in two-output systems
- scientific article; zbMATH DE number 758210
- Integrals involving cyclic products of kernels and asymptotic normality of cross-correlogram estimates of the response function
- On asymptotic behaviour of the error term in cross-correlogram estimation of response functions in linear systems
- Asymptotic normality of cross-correlogram estimates of the response function
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4538799)
