Adaptive pre-equalization using neural-like algorithm

For severe intersymbol-interference (ISI) channels, a linear post-equalizer at the receiver causes noise enhancement which degrades the performance. To avoid such a problem we propose the use of adaptive pre-equalization at the transmitter. Based on the back-propagation algorithm used for multi-layer neural networks, we derive an adaptive algorithm to train the pre-equalizer. The simulation example presented shows that substantial power gain can be achieved with this adaptation technique. It is also shown how to extend the training algorithm in order to adapt the pre-equalizer and post-equalizer simultaneously