Plotting of Power Spectral Density of Noise

Dear all, I have desire to plot the PSD for noise underwater water. The equations are as shown in the image below

The desired output is as shown in the image below

But the obtained output is as below

So what can be done in order to get the desired output?

[Rick Rosson]

Please post your MATLAB code.

Answer 1

[Star Strider]

You are seeing classical thermal or ‘1/f’ noise. So to me, your plot appears to be correct, unless you are using the same data used to produce the plot in the publication. In that event, you would need to post (attach) those data so we can experiment with them to see what the problem may be.

[Star Strider]

You have some fundamental errors. The LHS of your equations are given as 10*log(...) (and we don’t know what base the logarithm is, but I’m assuming 10 because it fits the context), so you have to take the antilog of the RHS to calculate them correctly. Then plot 10 times the log10 of the product:

f = linspace(1,1E+6, 250);
s = 0;
w = 1;
y = 50 - (18*log10(f));
nt = 17- (30*log10(f));
NT = 10.^(nt/10);
ns = ( (40)+(20*(s-0.5)) +(26*log10(f)) -(60*log10(f+0.03))  );
NS = 10.^(ns/10);
nw = ( (50)+(7.5*sqrt(w)) +(20*log10(f)) -(40*log10(f+0.4))  );
NW = 10.^(nw/10);
nth = -15 + (20*log10(f));
NTH = 10.^(nth/10);
hop = nt+ns+nw+nth;
Tot = prod([NT; NS; NW; NTH]);
Toto = 10*log10(Tot);

semilogx(f,Toto);

This still doesn’t reproduce the figure in the book. I suspect there is more to this than you’ve told us. This code is (as I read the page you posted) at least now mathematically correct. I have no idea what ‘y’ is doing, or if it’s important.