In this paper, a very efficient semiblind approach that uses a training-based least square criterion along with a blind constraint is proposed for multiple-input–multiple-output–orthogonal frequency-division multiplexing (MIMO–OFDM) channel estimation. The blind constraint is derived from the linear prediction of the received MIMO–OFDM signal and is used in conjunction with a weighting factor in the semiblind cost function. An appealing scheme for the determination of the weighting factor is presented as a part of the proposed approach. A perturbation analysis of the proposed method is conducted to justify the superiority of the semiblind solution and to obtain a closed-form expression for the mean square error (MSE) of the blind constraint, further facilitating the calculation of the weighting factor. The proposed method is validated through computer simulation-based experimentations, showing a very high estimation accuracy of the semiblind solution in terms of the MSE of the channel estimate.
展开▼
