Semi-blind MIMO channel identification based on error adjustment

Multiple input multiple output (MIMO) systems have received considerable interest over the last decade because of high system capacity. In this paper, we propose an error-adjustment scheme for semiblind channel estimation. The algorithm consists of a two-step iteration to achieve even better performance. The initial channel estimates are computed using a variant of cyclic prefix (CP): the know symbol padding only technique. Then the estimate error of the received symbol is used to adjust the initial channel estimates. Computer simulations show that the proposed algorithm offers good estimation behavior for MIMO systems.