论文信息 - Spectral moments from the ACF or directly from the signal

Spectral moments from the ACF or directly from the signal

Abstract The moments of a bandlimited power spectrum can be estimated from samples of the autocorrelation function without computing the spectrum. Further, short-time estimates of the moments can be continually formed directly from the signal, circumventing ACF computation. Such fast moment computation is of interest in EEG analysis.

Hans Werner Strube

[1] Kenneth S. Miller,et al. A covariance approach to spectral moment estimation , 1972, IEEE Trans. Inf. Theory.

[2] D Michael,et al. Automatic EEG analysis: a segmentation procedure based on the autocorrelation function. , 1979, Electroencephalography and clinical neurophysiology.

[3] Georg Lindgren,et al. Spectral moment estimation by means of level crossings , 1974 .

[4] G. Lindgren,et al. Frequency estimation from crossings of an unknown level , 1980 .

[5] K. S. Miller,et al. Estimation of spectral moments of time series , 1970 .

[6] Peter M. Schultheiss,et al. Short‐Time Frequency Measurement of Narrow‐Band Random Signals in the Presence of Wide‐Band Noise , 1954 .

[7] B Saltzberg,et al. Moments of the power spectral density estimated from samples of the autocorrelation function (a robust procedure for monitoring changes in the statistical properties of lengthy non-stationary time series such as the EEG. , 1985, Electroencephalography and clinical neurophysiology.