crandn

Generate m sequences of n Gaussian random numbers with a specified autocorrelation function

[cg, psg] = crandn(rgau,m)
Generate correlated Gaussian sequences by Fourier synthesis.

Input parameters:
rgau = correlation function - length n/2
m = number of realisations

Output:
cg = m x n matrix containing m sequences of n correlated variates from
a zero mean, unit variance normal distribution
psg = input power spectrum (Fourier transform of correlation function)

Note: Since this uses the fast Fourier transform, it will be fastest if n is a power of 2.

crandndemo.m is a demonstration. Note that one plot requires the function plotcdfkuiper (File Exchange ID #21280)