noisecg

Purpose

Analytic complex gaussian noise (white or colored).

Synopsis

noise = noisecg(N)
noise = noisecg(N,a1)
noise = noisecg(N,a1,a2)

Description

noisecg computes an analytic complex gaussian noise of length N with mean 0 and variance 1.0.

Name	Description	Default value
N	length of the output vector
a1	first coefficient of the auto-regressive filter used to color the noise	0
a2	second coefficient of the auto-regressive filter used to color the noise	0
noise	output vector containing the noise samples

noise=noisecg(N) yields a complex white gaussian noise.

noise=noisecg(N,a1) yields a complex colored gaussian noise obtained by filtering a white gaussian noise through a first order filter whose impulse response is

$$H(z)\ =\ \frac{\sqrt{1-a_1^2}}{1-a_1\ z^{-1}}.$$

noise=noisecg(N,a1,a2) yields a complex colored gaussian noise obtained by filtering a white gaussian noise through a second order filter whose impulse response is

$$H(z)\ =\ \frac{\sqrt{1-a_1^2-a_2^2}}{1-a_1\ z^{-1}-a_2\ z^{-2}}.$$

Example

         N=500; noise=noisecg(N);
         [abs(mean(noise)),std(noise).^2]
         ans = 
               0.0152    0.9680

         subplot(211); plot(real(noise)); axis([1 N -3 3]);
         subplot(212); f=linspace(-0.5,0.5,N); 
         plot(f,abs(fftshift(fft(noise))).^2);

See Also

rand, randn, noisecu.

© GdR ISIS - Contact