4-PAM simulation

Hello, I am working on a simulation of reading an audio signal into a 4-PAM modulator and demodulating in on the other side. I have Matlab and the Signal Processing toolbox that has the PAM modulator in it. I am having issues trying to format the data into the correct context to get the modulation to work. My current script is shown below:

fileIn = 'dog_dare2.wav';
info = audioinfo(fileIn);
nyRate = info.SampleRate*2;
%extract the information

%set the sampling frequency to twice that of the file

%Extract the data and the sample rate
[origY,origFs] = audioread(fileIn);

%Play the sound back
%sound(origY,origFs)
figure
ax1 = subplot(2,1,1);
ax2 = subplot(2,1,2);

%Plot the input signal and the digitized version
t1=[1/origFs:1/origFs:length(origY)/origFs];
plot(ax1,t1,origY)

%upsample the original date to nyquist rate
upY = upsample(y,2);
t2=[1/nyRate:1/nyRate:length(upY)/nyRate];
%add a plot
plot(ax2,t2,upY)

So I am analyzing the input signal for the sampling rate and then switching to the nyquist rate to double the sampling rate of the original. This is where I would think I would be able to put the information into the modulator but the modulator only accepts values from [0,M-1 (or 3 in this case)]. So I need to translate or scale the information into an acceptable input for the PAM modulator to work. Please help me out when you get a chance.

[Greg Girard]

Additional help.

So I now have the PAM modulated envelope and I am trying to combine that with an AM carrier wave but I am having a hard time combining them because of the size and matrices.

if true
  m = min(max(-2, floor(upY/0.5)),1)+2;
z = pammod(m,4);
subplot(2,3,4)
plot(t2,z)
title('PAM modulation envelope')
axis([0 2 -1 1])
%Am carrier frequency is between 535-1605kHz
carrier = 1605000;
cWave = cos(carrier*t2);
subplot(2,3,5)
plot(t2,cWave)
title('Carrier wave')
%axis([0 10*(1/1605000) -1 1])
subPam = z(1:1000,:);
subCarrier = cWave(:,1:1000);
t3=[1/nyRate:1/nyRate:length(upY)/nyRate];
modWave = subPam*subCarrier;
subplot(2,3,6)
plot(t2,modWave); 
end

So I can't just mix them together because they're both 46584x1 matrices (so the result is too large). I tried to make a sub sample of the first 1000 samples to try to modulate some at a time but I can't get the time to match up and they must be the same length. The modWave length is 1000x1000. Little help....still not super great at matlab.

Answer 1

[Walter Roberson]

You need to quantize the signal to get a result 0 to 3 and put that result through the modulator.

[Greg Girard]

So I attempted to do that on my own code that is shown below:

m = zeros(size(upY));
for idx = 1:size(upY)
   if upY(idx) < -0.5
       m(idx) = 0;
   elseif upY(idx) >= -0.5 && upY(idx) < 0
       m(idx) = 1;
   elseif upY(idx) >= 0 && upY(idx) < 0.5
       m(idx) = 2;
   else
       m(idx) = 3;
   end
end

Is there a built in function to do this better or would this work? I wasn't sure where to put the boundaries to include 0 or not since the upsampled signal has all those 0's in it. When I did this and do pammod I do get a warning that says "Warning: Imaginary parts of complex X and/or Y arguments ignored." Is that normal or did I not use it correctly?

[Walter Roberson]

m = min(max(-2, floor(upY / 0.5)), 1) + 2;

[Greg Girard]

Thank you for the help Walter. One more question for you (or anyone). So with these values, do I convert them to binary values first (dec2Bin) or do I just put these right into the pammod to get the correct output?

[Walter Roberson]

You pass them directly, like is documented for pammod

The message signal must consist of integers between 0 and M-1.