Random noise vs. White/Gaussian noise

I've been using this as my white noise so far,

s = (rand(N,1)-0.5)*2;

If I used one of MATLAB's noise functions what would be the difference?

The reason I ask is because I have to design and implement a basic adaptive filter using noise as the input source, but also having background noise as the noise that needs to be removed. It seemed counter-intuitive to have them as exactly the same types of noise.

Answer 1

[Wayne King]

rand() is a MATLAB random number generator. It generates random variables that follow a uniform probability distribution. randn() generates random numbers that follow a Gaussian distribution.

In the Statistics Toolbox, you have the ability to generate a wide variety of "noise" distributions.

[Tom]

many thanks wayne

[Honglei Chen]

I would suggest you to at least use randn. Uniform is just not a very practical model for noises.