8 QAM symbol error and bit error

[joserse46]

Hello, so I am simulating 8 QAM in matlab and have some questions about symbol error probability as my results are not making sense.
d1, d2, d3 are bits(vectors) sent
d1hat, d2hat, d3hat are the received (over AWGN channel)

n is the number of symbols
N=n*3 is the number of bits sent

C1 = xor(d1,d_hat1);
C2 = xor(d2,d_hat2);
C3 = xor(d3,d_hat3);

%bit error
Pb_simulated(l) = (sum(C1)+sum(C2) + sum(C3))*(1/N);%overall error for each EB/NO

%symbol error
C4=or(C1,C2);
C5 = or(C4,C3);
Ps_simulated(l) = (sum(C5))*(1/n)

the problem is for the first coulple of eb/no values i have a number greater than 1


any suggestions?

 

[zorro]

Hi joserse46,

The code seems to be ok.
Are you sore that the length of the vectors is n?
In other words,
do you get the same results with the following?

Pb_simulated(l) = (sum(C1)+sum(C2) + sum(C3)) / (3*length(C1));
Ps_simulated(l) = sum(C5) / length(C5);

Regards

Z