How to correctly divide long-time domain signal for wavelet transform?

Viewed 20

I have a system that samples 5 different signals, X,Y, Z,W,P . We sampling each signal in 100HZ for 24 Hs.

We want to apply wavelet transformation for each signal, the questions are

  1. What is the recommended window size for wavelet processing in our case?
  2. May we need to use a window with overlapping (if so, what is the theoretical reason)?
  3. Using CWT OR DWT ?

P.S we are working with the python pywavelet package

Thanks

0 Answers
Related