隨著無線通訊技術發展與多媒體應用服務成長，系統營運商與用戶端皆期許以提供或擁有?高傳輸速?及?好服務品質，傳統點對點傳輸（Point-to-point communication）技術受限於通道衰減、遮蔽效應與多?徑通道效應影響，用戶於蜂巢細胞覆蓋範圍邊際或處於重?遮蔽環境下訊號通訊品質時常低?。中繼站網?（Relay Networks）為近?廣泛討?與研究的新興通訊概?，網?發射端或稱為源節點（Source node）可透過其附近中繼站資源?轉遞訊號至接收端或稱目標節點（Destination node），?用空間分集效?（Spatial Diversity Gain）增加通道多樣性以大幅改善傳統點對點傳輸通訊品質?佳問題或增加無線通訊網?細胞覆蓋範圍。 為了更好的利用中繼網路的優點，可獲得的通道狀態資訊量（Channel State Information）是決定效能的主要因素，目前研究文獻大?假設中繼端及接收端可獲得所有的通道資訊，包含中繼網?每條?徑或每次跳躍的通道資訊。 在本論文中，我們考慮在放大轉遞（Amplify-and-Forward）中繼網路下進行通道估測，並提出兩種基於領航訊號的估測演算法：最大概似估測（maximum likelihood estimation）和最小均方誤差估測（minimum mean square error estimation）。最大概似估測的優點是易於實現，因為不需要知道通道統計特性（channel statistics），降低其計算複雜度，但在效能會較差。相反的，最小均方誤差估計儘管需要更高的計算複雜度，但由於它利用通道統計特性有效的增進系統效能。 接著我們討論不同的通道估測方法，可概分為兩類：?接式通道估測（Integrated Channel Estimation）及分?式通道估測（Disintegrated Channel Estimation），串接式通道估測方法在發射端傳送領航訊號（pilot signal），中繼站接收後將其功率放大，同時引入另一領航訊號與接收訊號一同轉傳至接收端。接收端可藉以估測「中繼端-接收端」通道及「發射端-中繼端-接收端」?接式通道；分離式通道估測方法則是在接收端先估測「中繼端-接收端」通道，再利用估測到的「中繼端-接收端」通道協助估測「發射端-中繼端」通道。模擬結果顯示在分離式通道估測方法下有較佳的效能。 於傳統點對點通道估測方法中，時域通道估測方法可?用功率延遲概觀（Power delay profile）特性強化通道估測效能，因此其效能遠優於頻域通道估測方法。因此我們也在中繼網路使用時域通道估測方法來增進效能。 因為中繼端的高斯雜訊對於「發射端-中繼端-接收端」?接式通道估測干擾，我們藉由雜訊濾除矩陣（Denoising Matrix）抑制雜訊。分別對串接式通道估測和分離式通道估測使用，模擬結果顯示有不錯的效能增進。 最後會將串接式通道估測和分離式通道估測在不同領航訊號數目與不同權重矩陣下的效能比較。 Recently, wireless communication technology and media application service grows faster than ever, providers and users all expect higher transmission rate and better Quality of Service (QoS). Because the conventional point-to-point communication is susceptible to channel fading, shadowing effect, and multipath channel effect, communication quality at cell edge or severe shadowing environments is low. Relay networks have been emerged as a promising technology to improve the link quality in severe channel fading and shadowing environments or increase the cell coverage by spatial diversity gain. In most of the existing literature, it is assumed that the perfect channel state information (CSI) of each link are available at the destination. However, the successful implementation of the relay networks relies on the accurate CSIs of each link. In this thesis, we propose time-domain channel estimators for relay networks in multipath channels. Based on the maximum likelihood (ML) and minimum mean square error (MMSE) criteria, two kinds of channel estimation methods, integrated and disintegrated, are investigated for estimating relay-to-destination (RD), source-to-relay (SR), or their cascaded channel links. A denoise-and-forward (DF) estimator is proposed to further improve the performance by imposing a denoising matrix to effectively suppress the noise enhancement effect at the relay node. Computer simulation is used to demonstrate the effectiveness of the proposed time-domain channel estimation methods.
