hi all,
i'm dng my project on adaptive channel equalisation using Higher Order Neural network(HONN).i ve some doubts. i have taken x(n) as the original signal; and r(n) as the noise signal. r(n) is input to HONN. Output of HONN is y(n). Error signal is found out by e(n)=x(n)-y(n). Until e(n) reduces to a small value the training is performed.
My question is: In real time application, such as if applying the adaptive channel equalisation in modems, we wont know x(n) value i.e original signal. then how could e(n) value to be found.
thks in advance.