Notes to the work at seminars:

• When you work in CTU FEE classrooms, all below referred functions are available in directory K:\VYUKA\DSP\m. This directory can be added into search path in MATLAB (From Menu 'File -> Set Path'). This setup can be saved for future sessions by the control button 'Save' in opened dialog window.
• Similarly, all referred signals are available in directory K:\VYUKA\DSP\signaly in FEE classrooms. When path is set to this directory too, they can be directly loaded into MATLAB.
• When you work at your PC without the access to disc K: you must download all referred functions and signals localy!

Guidelines to seminar:

• Generation of low-frequency color noise.
  • Generate white noise with Gaussian distribution, zero mean, and variance equal to 1.
  • Determine transfer function of the 1st order AR system for generation of low-frequency (LF)color noise with bandwidth B = 300 Hz when sampling frequency is supposed to be fs = 8000 Hz.
  • Observe in MATLAB frequency response of given system (AR model).
  • Generate required color noise by filtering of white noise by designed AR system (signal nc1). Observe smoothed PSD estimation of generated noise, its periodogram (i.e. unsmoothed short-time power spectrum), and spectrogram.

• Generation of other color noises using other 1st order AR or MA models.
  • Repeat the previous task for the case of generation high-frequency (HF) noise using AR model of the 1st order with the same parameters (signal nc2).
  • Generare LF and HF color noise using MA models with equivalent parameters, i.e. the width of attenuated band should be B = 300 Hz when sampling frequency is fs = 8000 Hz (signal nc1ma LF noise, nc2ma for HF noise).
  • Observe always frequency response of given MA model, smoothed PSD estimation of generated noise as well as its periodogram (i.e. unsmoothed short-time power spectrum) and spectrogram.

• Generation of band color noise using 2nd-order AR model.
  • Determine transfer function of 2nd order AR system for generation of band color noise when the poles are equal to p1|2 = 0.9 * exp ( +/- j * 0.9 ). Gain of AR system is equal to 1.
  • Determine central frequency and width of frequency band of this AR system. Observe again frequency response of given system.
  • Generate again the noise by the filtering with white noise excitation and observe again its PSD, periodogram, and spectrogram.