Aiming at the equalization problem of sparse channels in single-input multiple-output (SIMO) systems

a new iterative frequency-domain equalization algorithm is proposed based on the maximum likelihood (ML) criterion. The equalization of multiple signals is first modeled as the maximum likelihood frequency-domain signal sequence estimation from incomplete observations

and approximately solved by means of the expectation-maximization (EM) algorithm in an iterative manner. Analytic expression of the equalization output is finally obtained in the form of weighted summation of each discrete-frequency signals. In each iteration

the proposed scheme alternates between equalization output update and channel posterior distribution update. During the later step

the inherent sparse nature of the channels is exploited by employing sparse promoting prior distributions. Then

the sparse Bayesian learning iterative inference method is applied to the proposed model in order to obtain joint conditional posterior distribution of the channel parameters. Simulation results show that the proposed scheme has a good convergence and steady-state performance

and approaches the steady-state symbol error rate (SER) with known channel parameters

at moderate and high signal-to-noise ratio (SNR) values.