巨量均勻矩形陣列之有效到達方向估測

巨量(massive)多輸入多輸出(multi-input multi-output, MIMO)系統已經成為滿足不斷增長的數據消費需求的關鍵技術，因為它具有更多的通道容量、更高的頻譜效率和可靠性，到達方向(direction-of-arrival, DOA)估測的問題對巨量MIMO系統容量和頻譜效率具有顯著影響。本論文係於巨量多輸入多輸出(MIMO)系統處理具有低計算負荷的二維(two dimensional, 2D)到達方向(DOA)估測，廣為人知的2D多重訊號分類(multiple signal classification, MUSIC)方法可用於均勻矩形陣列的DOA估測，並具有良好的角度估測精度，但是它需要高計算複雜度。首先，通過重新排列接收訊號矩陣，本論文說明了方位角度和俯仰角度可以被分離的估測。與2D MUSIC方法相比，分離的MUSIC方法可以有效地避免方位角度和俯仰角度估測之間的干擾；且經過配對程序，可以獲得估測的方位角和俯仰角。但是，對於傳統的基於頻譜搜尋的MUSIC方法來說，它的計算成本仍然是很高。此外，複雜性和估測精準度嚴格的依賴於搜尋區間使用的格柵尺寸大小，這是耗時的且適當之搜尋格柵尺寸的選擇也不清楚。因此，期望能提供更好的角度估測性能和具有較低的計算複雜度，而由於求根版本的估測器能提供的解析臨界性能優於基於搜尋的估測器，故本論文亦採用傳統的多項式求根方法如root-MUSIC來取代傳統的頻譜搜尋估測。同時，本論文於2D DOA估測亦提出一種無需配對的快速root-MUSIC方法，與傳統的root-MUSIC方法相比較，它實現起來更簡單並且具有更低的計算負荷。最後，使用各種類型的天線配置之DOA估測性能亦被評估，還提供了電腦模擬結果用於說明所提出方法的有效性，特別是當低訊號雜訊比和/或小的空間取樣數目時。

關鍵字

巨量MIMO系統；均勻矩形陣列；到達方向；多重訊號分類；多項式求根；快速root-MUSIC

並列摘要

Massive multiple input multiple output (MIMO) system has become a crucial technology meeting the ever-increasing data consumption demand, for it has more channel capacity, higher frequency spectrum efficiency, and reliability. The problem of direction of arrival (DOA) estimation has a marked impact on massive MIMO system capacity and frequency spectrum efficiency. This thesis deals with two-dimensional (2D) DOA estimation with low computational load for MIMO systems. It is well known that the 2D multiple signal classification (MUSIC) method can be used for DOA estimation in uniform rectangular array and has a good angle estimation accuracy; however, it requires high computation complexity. First, through rearranging the received signal matrix, this thesis illustrates that the azimuth and elevation angles can be separately estimated. Compared with the 2D MUSIC method, the separate MUSIC method can avoid the interference between azimuth and elevation estimations effectively. By the pairing procedure, the estimated azimuth and elevation angles can be obtained. But, it is also computationally expensive for the conventional spectrum searching-based MUSIC method. Moreover, the complexity and estimation accuracy strictly depend on the grid size used during the search. It is time consuming and the search grid is not clear. It is expected to give a better angle estimation performance and have a much lower computational complexity. Due to the resolution threshold performance of root-version estimator is better than the searching-based estimators. Therefore, this thesis also uses the conventional polynomial rooting method, root-MUSIC, to replace the conventional spectrum searching method. Meanwhile, a fast root-MUSIC method is proposed for 2D DOA estimation without pairing. In comparison with conventional root-MUSIC method, it is much simpler to implement and has a lower computational load. Finally, the DOA estimation performance has been evaluated with various types of antenna configuration. Computer simulation results are also provided for illustrating the effectiveness of the presented methods, particularly when the signal-to-noise ratio is low and/or the number of snapshots is small.