Efficient Channel Estimation Method for MIMO Antenna Selection Systems Exploiting Temporal Correlation of Channel

This paper addresses the issue of MIMO channel estimation with the aid of a rough estimate of temporal correlation statistics where MIMO antenna selection system is employed. Under the temporally correlated channel, proposed method controls allocation of length of training symbols for each antenna element so that the elements which are likely to be selected are estimated more precisely than the other elements. In order to utilize this estimation scheme effectively, this paper also proposes the antenna selection method which takes account of a difference of channel estimation accuracy among antenna elements. When the ML channel estimation was adopted, the proposed selection method is with almost the same computational cost as the conventional selection methods. The proposed method does not rely on spatial correlation statistics which is not always stable in real propagation scenarios. The numerical simulation using 3GPP-SCM revealed that the proposed method works effectively under shorter training symbols employed.