ifestar2

Purpose

Instantaneous frequency estimation using AR2 modelisation.

Synopsis

[fnorm,t2,ratio] = ifestar2(x)
[fnorm,t2,ratio] = ifestar2(x,t)

Description

ifestar2 computes an estimation of the instantaneous frequency of the real signal x at time instant(s) t using an auto-regressive model of order 2. The result fnorm lies between 0.0 and 0.5. This estimate is based only on the 4 last signal points, and has therefore an approximate delay of 2.5 points.

Name	Description	Default value
x	real signal to be analyzed
t	time instants (must be greater than 4)	(4:length(x))
fnorm	output (normalized) instantaneous frequency
t2	time instants coresponding to fnorm. Since the algorithm do not systematically give a value, t2 is different from t in general
ratio	proportion of instants where the algorithm yields an estimation

This estimator is the causal version of the estimator called "4 points Prony estimator" in article [1].

Example

Here is a comparison between the instantaneous frequency estimated by ifestar2 and the exact instantaneous frequency law, obtained on a sinusoidal frequency modulation :

         [x,if]=fmsin(100,0.1,0.4); x=real(x); 
         [if2,t]=ifestar2(x);
         plot(t,if(t),t,if2);

The estimation follows quite correctly the right law, but with a small bias and with some weak oscillations.

See Also

instfreq, kaytth, sgrpdlay.

Reference

[1] Prony "Instantaneous frequency estimation using linear prediction with comparisons to the dESAs", IEEE Signal Processing Letters, Vol 3, No 2, p 54-56, February 1996.

© GdR ISIS - Contact