Title :
Uncertainties in adaptive maximum entropy frequency estimators
Author :
Keeler, R. Jeffrey
Author_Institution :
NOAA Wave Propagation Lab., Boulder, CO, USA
fDate :
10/1/1978 12:00:00 AM
Abstract :
For a real sinusoid in white noise, the weight vector noise associated with an adaptive maximum entropy (AME) frequency estimator causes the spectrum peak to shift away from the input frequency. For a long adaptive filter, approximate expressions for this estimator´s peak value and frequency variance are derived. For low SNR, the probability of obtaining a spurious peak is evaluated.
Keywords :
Adaptive filters; Approximation algorithms; Autocorrelation; Entropy; Frequency estimation; Least squares approximation; Random variables; Signal to noise ratio; Uncertainty; White noise;
Journal_Title :
Acoustics, Speech and Signal Processing, IEEE Transactions on
DOI :
10.1109/TASSP.1978.1163128