Why not Just use Adaptive Equalization on the Receiver?
From Fourier analysis we know that if we have the correct Amplitude and Phase response of a signal it will be perfectly characterized and can be perfectly recreated. This is true in a system with no noise present. If we were to assume the same Amplitude and Delay corruptions of the signal F above and that these corruptions occurred at the transmitter, then equalization at the receive side will allow the original square waveform to be recreated with one problem. The noise at F will have been increased by 3dB. This occurs because the link noise is added in after the signal was corrupted. Thus, to recreate the frequency spectrum a re-boost at the frequency component of F is done, which also increases the noise at that frequency accordingly. Equalizing at the incorrect point will improve the signal spectrum but will not improve the Signal to Noise ratio; in fact, it will hurt the Signal-to-Noise ratio. The only way to properly optimize link/system performance is to make sure the link/syste