Can you help me out on making the BER vs SNR plot ?

Hello

I am trying to make the plot BER (Bit Error Correction ) versus SNR (Signal To Noise Ratio) via the following code:

clc;clear;close all

N=100;

% Number of Bytes

x(1:2:N)=-3;

x(2:2:N)=3;

% Noise Generation

n=randn(1,N)

r=x+n;

hist(r,100)

% Calculation Of Signal

% Calculation Of Noise

% Calculation of SNR

Px=mypower(x)

Pn=mypower(n)

SNR=10*log10(Px./Pn)

for ii=1:length(SNR)

    % noise addition

    y = x + 10^(SNR(ii)/20)*n; % additive white gaussian noise

    ipHat = real(y)>0;

     nErr(ii) = size(find([n- ipHat]),2);

end

BER=nErr/N;

figure (1)

plot(SNR,BER,'r*')

mypower function is: function P=mypower(x)

N=length(x);

P=(1/N)*sum(x.^2);

end

Okay , so If I will execute the code I will get only one star at the graph . Can you please tell me if there is anything anything I can change in order to get the graph with all the possible points/stars ?

Thank you .

[Tamir Suliman]

what is the function mypower(x) is not defined correctly ?

Answer 1

[Daniel kiracofe]

Not 100% sure of the final answer here, but based on your statement that you are only getting 1 star on the graph, I assume that you are expecting SNR and BER to be arrays. Right now they are just scalars, so that is why you are only getting 1 star. The way you have written it, SNR will have the same size as Px. And px has the size of whatever is returned from mypower(). And mypower returns a scalar, because of the sum() function in there. So if you want SNR to be an array, you have two options, depending on what you are trying to do: 1) modify mypower to return an array instead of a scalar 2) add another loop around the entire code (e.g. for j=1:something), vary some parameter of interest with every iteration of j, and then compute SNR(j) = 10*log10(Px./Pn) on each iteration (hope that makes sense).

[Altin Guberi]

How to make arrays function ?