次世代行動通訊系統之信號入射方向估測與波束合成演算法研究

Abstract

在本論文中基於Matrix pencil演算法使用二維天線陣列進行訊號方位估測。和最常見的MUSIC演算法相比此種演算法可以明確的知道入射訊號的目標個數，並且在入射訊號為相關性相當高的訊號時也可使用。在本論文中，為了考量真實情況下演算法之表現，特別使用了3D全波電磁模擬軟體來模擬二維天線陣列在接收多個均勻平面波訊號入射時的接收訊號，使用此種模擬方法可以模擬出接收天線單元之間的耦合對演算法之影響，也可以模擬出在接收角度較大時，天線場型對演算法之影響。除此之外，本論文還提出可結合波束成形法和訊號估測演算法之方法，可使用在Matrix pencil演算法中解特徵值時所對應之特徵向量來造出一天線陣列因子用以對應入射訊號之方向，並且降低其他來波方向之天線場型大小。此種演算法可進行目標的偵測與追蹤並可應用於次世代無線通訊系統，提升訊號的訊號雜訊比進而達到提升訊號通道容量的效果。

In this thesis, the direction of arrival (DOA) estimation based on the Matrix Pencil method is developed using a two-dimensional array antenna. In comparison with the commonly used Multiple Signal Classification (MUSIC) method, such an algorithm can allow us to know the number of incoming signals and directly resolve the correlated signals without preprocessing the correlated signals. Specifically, the received signals obtained by the full-wave simulation of the 2D array antenna incident by plane waves from multiple directions were employed for the DOA estimation; the two important factors including antenna coupling and directional receiving pattern were taken into account herein. Additionally, the beam-forming can also be achieved by directly employing the eigenvectors associated with the eigenvalues corresponding to the directions of arrival signals. The algorithm developed in this work can be applied for detecting and tracking the direction of arrival signals transmitted by mobile users in the next generation wireless communication systems. Significantly, the beam-forming technology can generate directional pattern to increase the receiver signal-to-noise ratio (SNR) and channel capacity.