此篇論文提供一個存在強干擾信號的正交分頻多工系統中估測信號多重路徑到達角度的方法，此方法主要是以特徵分解的方式來建構信號空間給估測角度演算法來偵測角度，然後我們將此角度估測結果應用於廣義天線旁瓣消除器(GSC)接收器的設計。在此所提供的接收器與最小均方錯誤接收器比較，可得其並不需要知道使用者的訓練序列(training sequence)，如此可以增加資料輸出量，這是因為廣義天線旁瓣消除器接收器用到達方位角估測的資訊來取代訓練序列。而且我們使用多級威能濾波器(MSWF)運用在廣義天線旁瓣消除器上來增加追蹤能力和減少運算複雜度，並且在信號進入廣義天線旁瓣消除器接收器前提供一個信號處理的機制來增加接收效能。我們可從電腦模擬結果得知上述的方法所得到的效能結果。 This thesis proposed an eigenanalysis-based method for constructing signal space to estimate Angle-Of-Arrival in OFDM communication systems. After DOA estimation, we use the application as the constraints of the GSC receiver. The proposed receiver doesn't need user's training sequence compared to MMSE receiver that decrease data throughput. This is because that the GSC receiver utilizes the DOA estimation to replace the information of the training data. And then we use the MSWF which is one of the reduced rank filters to the GSC to increase tracking ability and reduce the computation complexity. Moreover, processing the signal before applying the signal vector to GSC is superior to none of pre-processing scheme. The performance results indicate that proposed method outperforms the Generalized Sidelobe Cancelers (GSC) with classical Weiner filter under less observation data. We can see the performance results by simulations.