Can't you just call the random number generator twice
x = randn(1e3,1);
y = randn(1e3,1);
If you look at the cross-correlation sequence of those two sequences you can show they are uncorrelated.
For example, let's compute the cross-correlation sequence and place approximate 95%-confidence intervals.
[xc,lags] = xcorr(x,y,20,'coeff');
lconf = -1.96/sqrt(length(x));
upconf = 1.96/sqrt(length(x));
stem(lags,xc,'markerfacecolor',[0 0 1]);
set(gca,'ylim',[lconf-0.03 1.05]);
hold on;
line(lags,lconf*ones(size(lags)),'color','r','linewidth',2);
line(lags,upconf*ones(size(lags)),'color','r','linewidth',2);
They look uncorrelated to me.
There are more advanced things you can do, but do you need that?
