implementing m-psk modulation and demodulation with convolutional encoder and viterbi decoder

I have implemented M-PSK modulation and demodulation system using AWGN,convolutional encoder and viterbi decoder.for the 8,16,32 and 256 psk, it give me a correct constellation points but when i used 64 and 128 psk there are some error at the comvolutional encoder and m-psk modulator baseband and it cannot display the constellation diagram

here are the link to download the circuit so that it will be more easier for you to answer my question.

http://www.mediafire.com/?ff4qeoj6blak16t

http://www.mediafire.com/?0e1r9rj9472xfi1

Answer 1

[Mike Mulligan]

From a quick look at your model, here are some things to try: 1) Make the frame size a multiple of the number of bits per modulation symbol (6 for 64psk and 7 for 128psk); 2) your demodulator is set to produce hard decisions, so your Viterbi decoder should be set to work on hard decisions too; 3) your error rate calculation block should have a receive delay equal to the traceback depth in your Viterbi decoder; 4) your phase offset should be the same in the modulator and demodulator. Once you adjust these settings you will still need to set the AWGN block appropriately to get reasonable error rates.