how can i filtering white gaussian noise using shaping filter H(f)

i want filtering white gaussian noise using shaping filter h(f). what is the function of shaping filter in matlab

Answer 1

[Star Strider]

Define the filter as a designfilt object (if you use designfilt), as numerator and denominator polynomials, or preferably a second-order-section and gain matrix, then use the filtfilt function to filter it, as you would any other signal. Your filter will pass the selected random-amplitude values in the frequency spectrum you specify for the filter. You can see the results if you do a fft (link) on the output of the filter.

[Good mind]

i dont understand,i have H(f)=(f_h^4 f^2)/((f^2+f_l^2 ) (f^2+f_h^2 )^2 ) where f_h=120hz and f_l=60hz. i need the filtered wg noise to add it to ecg signal. i m beginer .thank you for your help

[Star Strider]

Your ‘H(f)’ makes no sense to me.

Please consult a textbook on discrete or digital signal processing, and documentation for the MATLAB cheyb2 or butter and other related functions.

The passband you have chosen for your EKG filter will eliminate essentially everything of interest in your EKGs.

See my discussion of EKG signal processing in Detecting QRS complex in ECG signal (link) and other related posts for help in understanding how to do what you want with your EKG record.

[Good mind]

thank you sir for your precious time. i dont need how can i denoise ecg signal because in my case i will use wavelet transform and wiener filter. my problem is before adding wg noise to ecg signal, i should filtering this wg noise with shaping filter

[Star Strider]

My pleasure.

I understand that you want to add band-limited white noise to your signal. The filter design procedure is the same.

To design a bandpass filter with a lower passband frequency of 60 Hz and an upper passband frequency of 120 Hz, you can use this design. Your EKG signal has to have a sampling frequency of at least 250 Hz for this filter to work. Be sure to use the sampling frequency of your signal, if it is not 250 Hz as this filter was designed for:

Fs = 250;                                               % Sampling Frequency
Fn = Fs/2;                                              % Nyquist Frequency
Wp = [60  120]/Fn;                                      % Passband Frequencies (Normalized)
Ws = [59  121]/Fn;                                      % Stopband Frequencies (Normalized)
Rp = 10;                                                % Passband Ripple (dB)
Rs = 50;                                                % Stopband Ripple (dB)
[n,Ws] = cheb2ord(Wp,Ws,Rp,Rs);                         % Filter Order
[c,b,a] = cheby2(n,Rs,Ws);                              % Filter Design
[sosbp,gbp] = zp2sos(c,b,a);                            % Convert To Second-Order-Section For Stability

figure(1)
freqz(sosbp, 2^16, Fs)                                  % Bode Plot Of Filter

Use the filtfilt function to do the actual filtering.

The freqz call is not necessary for the code. It shows you what the filter passband and stopband will look like in the frequency domain, so you can be certain it is stable and gives you the result you want. Other designs such as the Butterworth designs are also possible with this code, although you have to increase the transition regions (the frequency differences between the stopbands and passbands) to use other designs to get stable filters.

[Good mind]

thank you very very much sir.

[Star Strider]

My pleasure.