隨著無線通訊技術的應用增加,高傳輸率己經變成一個重要的研究課題。多輸入多輸出型態是其中一種達成方法。首先,我們研究室內環境的通道特性來設計一個更好的多輸入多輸出系統。我們觀察到因為接收到的信號雜訊比變小了,所以阻擋直接波無法提昇系統容量。另外,我們證明了在高信號雜訊比時,多輸入多輸出技術在增加通道容量的表現上會比相位陣列天線表現的還要好。之後,我們基於奇異值分解來建立了三種四輸入四輸出正交分頻多工系統的架構。這三種架構可以分類為:(1)四個子通道全利用 (2)只利用三個子通道 (3)使用可適應位元與功率分配的演算法來傳送資料。我們證明了去除有最小奇異值(singular value)的子通道而只利用三個子通道來傳輸資料是一種簡單的改進效能方法。另外,使用可適應位元與功率分配可以有效地增加傳輸率並且使不同接受位置的位元錯誤率更穩定。
As the increase of applications of wireless communications, high transmit rate becomes an important research topic. The Multiple-Input-Multiple-Output architecture is one of the solutions. In order to design a better MIMO system in the indoor environments, we study the characteristics of channels. We observe that blocking LOS doesn’t increase the capacity because the receiver SNR is reduced. We also demonstrate that MIMO techniques are more effective in increase channel capacity than phased array techniques at high SNR. Then, we build up three architectures of a 4x4 MIMO OFDM system which are categorized as (1) using all four sub-channels, (2) using only three sub-channels and (3) using adaptive bit and power loading algorithms for data transmission. We demonstrate that using only three sub-channels by discarding the sub-channel with the smallest singular value is a simple method to improve the performance. Using the adaptive bit and power loading algorithm can effectively increase the data rate and make BER stable at different Rx locations.