Efficient transmit antenna selection for correlated MIMO channels

Abstract

Multiple-input multiple-output (MIMO) wireless communications system can achieve increasing capacity or reliablity without extra bandwidth and power. Thus, there are many researches for MIMO systems. However, as increasing the number of antennas along to the development of wireless communications, increased multiple RF chains gives burden to the system because of complexity. Therefore, antenna selection is introduced as a technique to reduce the burden. To select optimal antenna subset, each post-processing signal-to-noise ratio (SNR) or minimum singular value of all the candidates is needed. As increasing the number of antennas, the number of candidates also increases rapidly. Since the complexity of computing SNR or singular value is burden of system, large number of antenna elements is also obstacle of antenna selection technique. Therefore, there are several efforts to reduce the complexity of exhaustive search for antenna selection. The efficient and fast antenna selection algorithms from the past have a common feature that they reduce the size of candidates or channel matrix. That is, these methods reduce the number of computing SNRs or singular values by abbreviating the scale of candidates. However, these methods use exhaustive search for small scale of candidates. This thesis tries to improve this exhaustive search method. This thesis proposes two methods to improve the conventional antenna selection. Firstly, this proposed method does not compute all the post-processing SNRs or singular values of the channel matrix for antenna subset. Instead, this method decides whether each candidate is selected or not. After a antenna subset is selected, this proposed method stops searching so that the post-processing SNRs or the minimum singular values of rest candidates does not have to be computed. Thus, the complexity can be reduced. Whether a antenna subset is selected or not is realized by setting a threshold. An antenna subset whose minimum singular ...