Zero-mean Gaussian white-noise process with known power spectral density (PSD)

Hello everybody

How can I generate zero-mean Gaussian white-noise process with known power spectral density (PSD)which is a constant ? (I want to add this noise to some acceleration data to model an accelerometer sensor)

Thanks in advance.

Answer 1

[Wayne King]

Just use randn() and multiply by the square root of the variance.

The theoretical PSD of WGN is \sigma^2 (\Delta t) where \sigma^2 is the variance and \Delta t is the sampling interval.

For example:

   sigma2 = 2;
   x = sqrt(sigma2)*randn(100,1);

[Noel]

I question this answer. pcov(2*randn(10000,1)) = 1 pcov(10*randn(10000,1)) = 15 pcov(100*randn(10000,1) = 35 How does this fit into "The theoretical PSD of WGN is \sigma^2 (\Delta t) where \sigma^2 is the variance and \Delta t is the sampling interval."