2aSP9. Short-time Fourier transform: Error estimation with nonstationary signals.

Session: Tuesday Morning, December 3

Time: 11:00

Author: John C. Burgess
Location: Dept. of Mech. Eng., Univ. of Hawaii, 2540 Dole St., Honolulu, HI 96822


When used with nonstationary signals, a discrete Fourier transform (DFT or FFT) can provide only approximate values for the spectrum of the true, underlying signal. The purpose of this presentation is to provide some insight into the magnitudes of errors to be expected with increasing nonstationarity and noise, with some hints about how to ``manage'' the errors. [Work supported by the National Research Council (Canada) and the Lanikai Foundation.]

ASA 132nd meeting - Hawaii, December 1996