Choice of weighting for averaged non-parametric transfer function estimates
Research output: Contribution to conference › Paper › peer-review
Standard Standard
2005. 833-838 Paper presented at 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601), Nassau, Bahamas.
Research output: Contribution to conference › Paper › peer-review
HarvardHarvard
APA
CBE
MLA
VancouverVancouver
Author
RIS
TY - CONF
T1 - Choice of weighting for averaged non-parametric transfer function estimates
AU - Heath, W.P.
PY - 2005/5/16
Y1 - 2005/5/16
N2 - We consider the statistical properties of nonparametric transfer function estimates formed via the weighted sum of empirical transfer function estimates. In the case that the system input is considered deterministic, we show that the weighting corresponding to the classical averaged periodogram estimate is optimal. A similar asymptotic result is derived for the case where the input is considered stochastic.
AB - We consider the statistical properties of nonparametric transfer function estimates formed via the weighted sum of empirical transfer function estimates. In the case that the system input is considered deterministic, we show that the weighting corresponding to the classical averaged periodogram estimate is optimal. A similar asymptotic result is derived for the case where the input is considered stochastic.
U2 - 10.1109/CDC.2004.1428777
DO - 10.1109/CDC.2004.1428777
M3 - Papur
SP - 833
EP - 838
T2 - 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601)
Y2 - 14 December 2004 through 17 December 2004
ER -