Skip to Main Content
A method to improve the performance of multiple-input-multiple-output systems is to employ a large number of antennas and select the optimal subset depending on the specific channel realization. A simple antenna-selection criterion is to choose the antenna subset that maximizes the mutual information. However, when the receiver has finite complexity decoders, this criterion does not necessarily minimize the error rate (ER). Therefore, different selection criteria should be tailored to the specific receiver implementation. In this paper, we develop new antenna-selection criteria to minimize the ER in spatial multiplexing systems with lattice-reduction-aided receivers. We also adapt other known selection criteria, such as maximum mutual information, to this specific receiver. Moreover, we consider adaptive antenna-selection algorithms when the channel is not perfectly known at the receiver but can only be estimated. We present simulation examples to show the ER of the different selection criteria and the convergence of the adaptive algorithms. We also discuss the difference in complexity and performance among them.