Abstract
Communication channels are, in general, not deterministic but stochastic.Hence, modeling the stochastic channels explicitly could make equalization more efficient.We regard the channel parameters and the transmitted data in an equalization problem with stochastic channels as unknown parameters and missing data, respectively.The EM algorithm gives an estimate of the posterior distribution of the missing data in such a problem,that enables the Bayes estimation.The proposed equalizer shows a better performance than the conventional one in computer simulations.
