第五代巨量天線徑分多重接取蜂巢式系統中通道估計方法之研究

Abstract

隨著通信科技日新月異，第五代行動通訊標準即將在2010年正式定案。目前，學術界及產業界無不致力於第五代行動通訊系統的發展以作為滿足未來嚴苛的資料傳輸要求之解決方案。為了因應這個問題，我們研究第五代通訊系統可能的技術之一，稱為巨量天線徑分多重接取(MDMA)蜂巢式行動通訊系統。本論文中探討可用於符合此系統運作模式下，基於通道特性的兩個面向而發展出相對應的兩種方法來減少通道估計誤差。其中一個面相是利用通道間時間上的相關性並結合典型卡曼濾波器與來作通道預測與估計，另一個面相則是根據佈建的巨量天線間對於通道路徑延遲的相關資訊來做為輔助路徑篩選之方法，稱為大數決策輔助路徑篩選。這兩種方法在此通訊系統下之效能表現亦呈現於模擬結果中，我們且針對此兩種方法對系統效能的影響作討論與分析。

With the ever-changing nature of telecommunication technologies, the fifth generation (5G) mobile communication standards will be finalized by 2020. Currently, both the academia and industry are dedicated to the development of possible solutions that meet the challenging demands of the 5G communication systems. In response to this, we research on one of the possible techniques known as the massive antenna multipath division multiple access (MDMA) cellular system. In this thesis, methods for improving channel estimation under this mobile radio communication system are investigated. Two methods for reducing channel estimation error are proposed by exloring channel characteristics from two different aspects. One is the application of the Kalman filter with temporal channel correlations. The other one is the majority-vote-aided path selection with similar multipath channel delay information over receiver antennas. Simulation results are given and discussed on the performance of these two methods.