Signal Analysis Primer

A Signal Analysis Primer for the MATLAB Naïve: v1.3

An introduction to basic concepts in signal processing and their implementation in MATLAB.  Updates are ongoing (most recent: 9/13/2014; update includes minor edits and explanation of multitaper methods) and I welcome any and all suggestions for its improvement!  Thanks go to Matthijs van der Meer and Rachel Samson for feedback and edits.

Note: I don’t recommend reading this PDF on the Firefox viewer, because, for simplicity, I use default MATLAB figure settings and the line width on many of the figures is too narrow.

0. About MATLAB 
1. Fourier transformations and signal generation 
   MATLAB Exercises
   1.1) Generating signals that oscillate 
   1.2) Using the Fourier transformation to convert the signals 
        into power spectra 
   1.3) Zero-padding a signal 
2. Signal filtering and more signal generation 
   MATLAB Exercises
   2.1) Filter random noise to 6-10 Hz 
   2.2) Decompose filtered signal with a Fourier transform 
   Chapter 2 Appendix: function filt_LFP 
3. Time-frequency analysis, wavelets, and the Hilbert transform 
   MATLAB Exercises
   3.1) Generate a signal with a 8 to 10 Hz segment followed by a 
        16 to 20 Hz segment 
   3.2) Apply a sliding-window Fourier analysis 
   3.3) Apply a Morlet and Mexican hat wavelet analysis 
   3.4) Compute the instantaneous phase and amplitude of a signal 
        using the Hilbert 
4. Coherence and phase synchrony 
   MATLAB Exercises
   4.1) Generate two simple signals 
   4.2) Compute the magnitude squared coherence between the two 
   4.3) Compute the phase synchrony between the two signals 
   Chapter 4 Appendix: Function LFPSynch


One Reply to “Signal Analysis Primer”

  1. It is a good article for a beginner like me. I want to know more on wavelet analysis. If you dont mind, would you please email me some tutorial using MATLAB on wavelett and Hilbert-Huang Transform on some signals. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *