16 QAM and rayleigh channel with matlab

[leony]

qam to bit mapping in matlab

I am trying to simulate a 16 QAM modulation in rayleigh channel. However the BER is always around 0.5 no matter the Eb/No is.
The channel model I am using is:

ch=1/sqrt(2)*(randn(N,1)+sqrt(-1)*randn(N,1)); where N is the number of bits.

noise=1/sqrt(2)*(randn(N,1)+sqrt(-1)*randn(N,1)); is the AWGN.

s_hat=sqrt(snr(k)).*abs(ch).*X + noise; is the symbols after channel and AWGN.

(snr(k)) is the SNR or Eb/No loop.

I do not use shaping filters. Just bits to symbol mapping, the channel and noise after than symbol to bit mapping.

However the BER is always around 0.5. What can be the problem here?

Thanks in advance....

 

[bee]

qam+rayleigh+matlab

u will always get BER 0.5 if u didnt do any thing to remove the effect of channel at the RX(i am not saying noise).
Use some equaliztion tech to cancel out the effect of channel and u will get the correct result.

 

[leony]

Thanks for the quick reply. But I am using the same structure for 4-QAM without equalizer and BER is the same as the theoretical values. I tested this again with BPSK and it is ok as well. But when simulating 16-QAM, BER is always 0.5. What would be the reason of this?

 

[bee]

can u post the code u are using for BPSK and QPSK ?