Achieving Maximum Possible Download Speed on ADSL Systems

Recently, in [1] we developed a theoretical frame-work for achieving the maximum possible speed on a constrained digital channel with a finite alphabet. We obtained the capacity of the channel numerically, using a constrained Blahut-Arimoto algorithm which incorporated an average power constraint P at the transmitter. Our simulations showed that under certain conditions the capacity approached very closely, the Shannon bound. We also showed the maximizing input distributions. In this paper, the theoretical framework developed in [1] is applied to a practical example, the downstream channel of an asymmetric digital subscriber loop (ADSL), connection where the inputs to the channel are quantized and the outputs are real. We test how closely we approach the minimum configuration bound for this Channel under a practical noise environment.