Blind equalization using direct channel estimation

Abstract

In performing blind equalization, we propose a direct channel estimation method based on entropy-maximization of input signal with its known probability density function. That is, the proposed method estimates filter coefficients of the channel instead of equalizing filter coefficients which most of equalization methods try to estimate. Because the channel usually has a much shorter length than the equalizing filter, this method requires much smaller parameters to be estimated, and the channel can be equalized with much less computational demands. In addition, simulation results show that the proposed method can recover signals with a much smaller error than conventional methods.