Estimation of the Parameters of a Bernoulli–Gaussian Communications Channel

Abstract

A model is presented for a digital subscriber line in the form of a Bernoulli–Gaussian communications channel that is indirectly described in the form of a likelihood function. An expression is presented for the accuracy of the standard estimate of the pulse characteristics based on traditional averaging. A lower bound is formalized for the dispersion of the estimate of the pulse characteristics of a communications channel, demonstrating the possibility of improving the accuracy of the effective estimate.