Hi Philip,
It’s method interesting. But i have some questions. Firstly, did you have leading zeros and trailing zeros in BER test set?
Secondly, why did you disable “Dispersion compensation” in block DSP for QPSK? what is difference between block “Decision” and “M-ary threshold detector”?
And last question, in my opinion, it’s curiously for optical system BER=0. There is ber within the limits of 10^-10 ….10^-50.