Transmit/receive processing with partial or full feedback information for multiple-input multiple-output systems

Abstract

This dissertation investigates two kinds of processing methods for multiple-input mul tiple-output (MIMO) downlink systems. The first part of the investigation focuses on transmit-processing (preprocessing) for single-user (SU)-MIMO systems. The second part of the investigation focuses on transmit- and receive-processing (pre- and post-processing) for multi-user (MU)-MIMO systems. First, the SU-MIMO systems are considered with partial feedback information. There are many SU-MIMO preprocessing methods. The simple method employed for this study is double space-time transmit diversity (DSTTD) with antenna grouping (or shuffling). An efficient antenna grouping method for DSTTD is developed based on the minimum mean square error (MMSE) criterion. In order to reduce the computational complexity of the 4-by-2 DSTTD receiver, a simple criterion which is equivalent to MMSE criterion, is derived. Compared to the existing signal-to-noise ratio (SNR)-based method, the proposed method involves considerably reduced computational complexity in finding the antenna grouping matrix while showing comparable performance. Furthermore, the new preprocessing methods, for the space-time coded systems including a DSTTD system, are proposed with transmit antenna selection and grouping (ASG). This new preprocessing yields higher postprocessed signal-to-interference-plus-noise ratio (SINR) than the conventional preprocessing. And the peak-to-average power ratio (PAPR) reduction techniques are studied for applying the proposed ASG into the broadband systems such as an orthogonal frequency division multiplexing (OFDM) systems. Also, the optimal criterion for transmit antenna selection minimizing the pairwise error probability (PEP) is examined for antenna selection system when optimal ML receiver is employed. In the second part, MU-MIMO systems are considered under the available full channel state information at the transmitter (CSIT). A regularized channel diagonalization method i...