Performance analysis of the parametric channel estimators for MLSE equalization in multipath channels with AWGN

The parametric channel estimator is used in the parametric maximum-likelihood sequential estimation algorithms used to equalize the multipath wireless channels. We examine the performance of the parametric channel estimators against additive white Gaussian noise and the nonparametric channel model. We first show that the parametric channel estimators eliminate all the noise outside the signal subspace, and thus significantly reduce the performance degradation caused by the noise. Next, we show that the conventional channel estimators are special cases of the parametric channel estimator subject to some deteriorating constraints regardless of the channel model.