Abstract

Employing multiple antennas in multiple-input multiple-output (MIMO) communication systems has many advantages over systems employing single-transmit and single-receive antennas. The most important of these advantages is the tremendous increase in channel capacity. However, employing multiple antennas results in a significant increase in the system complexity, and hence cost, since each employed antenna requires a separate radio frequency (RF) chain. Antenna selection has been introduced recently as a means to alleviate this complexity. The main idea behind antenna selection is to select at the transmitter and/or receiver a good subset of the available antennas. By this, the number of required RF chains reduces to as few as the number of selected antennas, thereby reducing the system complexity and its cost. Antenna selection has been considered for both space-time trellis codes (STTCs) and orthogonal space-time block codes (STTCs) with favorable results. However, all works in this area considered a particular channel model, namely, quasi-static Rayleigh fading. In this thesis, we consider receive antenna selection for STTCs for generalized Rayleigh fading channels. Specifically, we derive explicit upper and lower bounds on the performance of STTCs with antenna selection over fast and block Rayleigh fading channels. The latter channel model is a good model for most wireless communication channels because it encompasses a wide range of mobility and data rates. For both channel models, we show that the diversity order deteriorates with antenna selection and it becomes proportional to the number of selected antennas and not the number of available antennas. However, having more antennas available, while keeping the number of selected antennas fixed, increases the coding gain, which naturally comes as a result of an increased average SNR due to antenna selection. This is unlike the case for quasi-static fading where the diversity order is maintained with antenna selection provided that the underlying STTC is full rank. As for STBCs, it appears that the channel model has no impact on the resulting diversity order with antenna selection. It has been shown that the diversity order of STBCs with antenna selection is the same as that of the full complexity system. In light of the results of this thesis and existing results in this area, the conclusion is that, when antenna selection is considered, it is recommended to use STBCs when the channel is fast fading or block fading with high mobility (or low data rate). On the other hand, it is recommended to use STTCs with antenna selection when the channel is modeled by quasi-static fading or block fading with low mobility (or high data rate)