基於波束成型技術下之空間預編碼頻分雙工大規模多輸入多輸出系統

Abstract

在這篇論文中，我們提出在頻分雙工大規模多輸入多輸出系統下，基於波束成型技術下之空間預編碼來達到減少下行通道資料狀態估測的開銷以及通道狀態資料的回傳開銷的目的。在大規模多輸入多輸出系統中，基地台上的傳送天線數量會變得非常多，下行通道資料狀態估測的開銷以及通道狀態資料的回傳開銷會變得巨大。為了提供頻率平緩的通道，結合正交分頻多工與大規模大規模多輸入多輸出系統的狀況下，如何解決這樣的開銷難題會變得更加關鍵。我們所提出的方法可以減少下行通道資料狀態估測的開銷以及通道狀態資料的回傳開銷。從模擬結果中我們可以觀察到，我們所提出的方法可以減少下行通道資料狀態估測的開銷，同時對於通道狀態資料的回傳開銷也非常顯著低降低。此外，我們所提出的方法跟前人提出的方法相比，我們所提出的方法同時擁有較低的運算複雜度。

In this thesis, we proposed a beamforming-based spatial precoding scheme in frequency division duplex (FDD) massive multiple-input-multiple-output (MIMO) systems in order to reduce the downlink training overhead and channel state information (CSI) feedback overhead. When the number of transmit antennas on a base station (BS) is growing large in massive MIMO systems, the downlink training overhead for CSI estimation and the CSI feedback overhead is growing tremendously. This problem would be more critical when we combine orthogonal frequency-division multiplexing (OFDM) with massive MIMO for applying frequency-flat channel in each subcarrier. The proposed method reduces the downlink training overhead and the CSI feedback overhead, thus enhances spectral efficiency. Simulation results show that the proposed method can estimate CSI without heavy downlink training overhead and the CSI feedback overhead is significantly reduced. Furthermore, the proposed method also applies lower computation complexity comparing with other algorithms in recent literature.