Weighted LS multiuser channel estimation for LTE

This paper addresses the channel estimation based on tentative data estimates in multicarrier systems, such as the LTE. We propose an iterative estimation approach based on solving in every iteration a specific weighted optimization problem. The weights are derived for two cases: known and partly known (or unknown) variances of the data estimation errors, assumed Gaussian; this assumption holds for the LTE uplink with the SC-FDMA modulation. We compare by simulation the proposed reweighted estimation approach with the conventional iterative estimation based on solving the unweighted optimization problem in a multiuser LTE uplink scenario. The comparison shows that the proposed weighted approach significantly outperforms the conventional unweighted estimation.