The Modified Basis Function Method for Time-Varying Frequency Estimation

Main Article Content

Chukiet Sodsri

Abstract

A modified version of the basis function method, called the modified basis function method, is proposed for estimating time-varying frequency of nonstationary signals. The modification is accomplished by adjusting a TVAR process, used as a linear predictor, and applying a combination of both the forward and the backward predictors for calculating TVAR parameters. The time-varying frequency estimate was extracted from location of the closet pole to the unit circle in the complex z-plane. Two nonstationary signals, one is chirp and another is of sinusoidally time-


varying frequency, are used as examples. From our results, the proposed approach yielded better accuracy in estimating the time-varying frequency in either noisy or noise-free situation than using the traditional basis function method.


Keywords: Time-varying autoregressive, Modified Basic function method, Nonstationary signal, Frequency estimation


Corresponding author: E-mail: sodsri@su.ac.th

Article Details

Section
Original Research Articles

References

[1] Sodsri, C., 2005 TVAR Modelling and Time-Varying Frequency Estimation of Nonstationary Signals, Accepted to present in The 2nd International Symposium On Mathematical, Statistical and Computer Sciences, Bangkok, Thailand.
[2] Eom, K.B., 1999 Analysis of Acoustic Signatures from Moving Vehicles Using Time-Varying Autogressive Models, Multisignal Systems and Signal Processing, 10, pp 357-378.
[3] Hall, M.G., Oppenheim, A.V. and Willsky, A.S., 1983 Time Varying Parametric Modeling of Speech, Signal Processing, 5, pp 267-285.
[4] Beex, A.A and Shan, P., 1999 A time-varying Prony Method for Instantaneous Frequency Estimation at Low Frequency, Proceeding of the 1999 IEEE International Symphosium on Circuits and Systems, 3, pp. 5-8.
[5] Neidzwiecki, M., 2000 Identification of Time-varying Processes, John Wiley & Sons, Chicester, England.