IIR adaptive equalizer using channel estimator

Abstract

The authors propose an infinite impulse response (IIR) adaptive equalizer using a channel estimator. This equalizer is able to equalize the channel characteristics with the least mean square (LMS) algorithm even under the ill condition where the channel is distorted severely. It has a cascade structure of an all pole filter which has the inverse characteristics of the channel and a finite impulse response (FIR) adaptive filter updated by the LMS algorithm. In the case where only the FIR adaptive filter is used as the equalizer, the convergence property of the LMS algorithm changes for the worse under the ill condition. Because the use of the all pole filter, however, reduces such effects of the ill condition, the convergence property of the LMS algorithm of the FIR adaptive filter at the second stage is improved. Although the proposed equalizer is an IIR system, its stability is guaranteed by a simple operation. The performance of the proposed equalizer is shown by computer simulations.<>