Help filtering data with unequal/uneven sampling

Hello,

I am trying to compute the autocorrelation of data sampled at uneven time intervals. My measurements correspond to the axial velocity of a flow, and I record a data point at every time a tracer particle is recognized, which can occur at any time.

I have a method for computing the autocorrelation function which has been tested previously; however the present data has a non-zero assymptote (i.e., the autocorrelation levels off at about 0.1). I believe this is due to noise in the system, and would like to try and filter my data, however I am not sure the best way to do this.

Is matlab's built in fdesign filters intended to work with unequal sampling times? I can determine an average sampling rate, but I don't know if that's appropriate. Is there another method for filtering this kind of data?

Thank you very much for your help,

- Tim Wabel

[dpb]

How irregular is the sample rate? You might have better luck interpolating to a constant rate first, depending...

[Tim Wabel]

It varies between roughly 10 to 200 microseconds. Below are the first 20 'inter arrival times' separating the bursts:

 "Inter Arrival Time Ch. 1 (usec)"
63.779995
222.680008
174.700027
95.779953
47.880051
19.719999
13.479963
19.319996
118.480003
5.980022
83.940048
12.22
38.819969
51.459995
114.180031
76.379974
25.679939
4.82006

I've played around a little with interpolating to a fixed rate, but I'm a little concerned with how that would be changing the correlation between bursts.

[dpb]

I don't know what to think to tell you w/ that kind of variability...not sure what it means not knowing much about the situation. Looks difficult, though...