A Generalized Procedure for Evaluating MIMO Channel Capacity

在本論文中，我們提出一套通用的程序來評估多輸入多輸出系統的通道容量，此程序藉由傳送和接收天線的複數向量場型、去向波分佈和來向波分佈及其極化，便可產生通道矩陣而進一步求得通道容量，此外，天線元素間的相關係數也可經由此程序求得。 利用半波長雙極天線作為多天線之元素，可研究不同因素對於天線間相關係數大小和通道容量之影響，比如來向波分佈、雙極天線的間距、雙極天線之擺放方向，本論文亦著重於探討互耦效應造成的影響。根據模擬結果分析，我們發現來向波分佈是影響相關係數的主要因素，即使天線的間距超過一個波長的距離或雙極天線正交擺放，角擴展越小或是直式波能量很強仍會造成天線間較高的相關度。此外，互耦效應會降低天線間的相關係數而提升通道容量，亦會造成輸入阻抗的不匹配降低信號能量導致通道容量的衰減，此兩因素在天線擺放非常近的時候，通道容量衰減的程度會比所提升的來的大。 我們亦實際進行室內環境下的通道量測，探討天線的間距對於通道容量的影響，經由數據分析得到相似的結果，即是天線元素的間距較小時，會得到較低的通道容量，藉此也驗證所提出的通用程序。此外也對極化分集做了探討，發現在有直接波的室內環境下，若傳送端和接收端的天線組態同時採用正交極化，對於多輸入多輸出系統會有較好的分集增益和通道容量。In this thesis, a generalized procedure is proposed to evaluate the channel capacity of a MIMO communication system. According to the MIMO antenna pattern and the information about the radio waves, we can generate the channel matrix to obtain the channel capacity. In addition, the correlation coefficient between antenna elements can also be computed through this procedure. Applying half-wavelength dipoles as the MIMO antenna elements, effect of AOA distributions, antenna spacings, and antenna configurations on the correlation and the channel capacity can be examined. We find that the AOA distribution is the dominant factor impacting the correlation. Though the antenna spacing is large or the orthogonal antenna configuration is used, correlation is still high with small angular spread or a strong LOS component. Besides, this thesis also focuses on mutual coupling effect. Results show that mutual coupling will lower the correlation thus increasing the capacity, however, it yields impedance mismatch which degrades the SNR. With very small antenna spacing, the amount of decreased capacity is more than the increased. Furthermore, MIMO channel measurements were conducted in indoor environment, by which we can investigate the effect of antenna spacing on the capacity as well. From the results of data analysis, similar results are obtained thus validating the proposed procedure. Besides, polarization diversity was also studied through the channel measurement. We found that orthogonal polarization at both communication ends will benefit the capacity and diversity gain