由於目前使用的頻段愈來愈擁擠，5G 系統為了要達到更高的傳輸 速率，故開始物色頻率更高而且頻寬更寬的毫米波。所以近幾年來， 毫米波系統的設計開始逐漸受到重視。 本篇論文提出一個毫米波系統受到強烈干擾的情形下，如何消除掉 干擾對毫米波系統的影響。由於毫米波傳輸特性上的限制，必須使用 巨量天線才能達到一般訊號的傳輸距離。為了要減少硬體設計的複雜 度和能量消耗，所以要使用類比和數位混合的波束成形。接收到的訊 號會先經過類比端，在類比端就可以先做第一步處理，把一部份干擾 對系統的影響消除掉，處理過後的訊號會再進入數位端，數位端就可 以有效的把殘餘的干擾去除掉。廣義旁瓣消除器是一個可以應用於各 種干擾和雜訊消除問題的方法，我們將其應用於數位波束成形再結合 多級維納濾波器，在有限樣本之條件下，多級維納濾波器的降維運算 法可以減少計算複雜度並且提供可靠的效能。;Since the frequency bands currently used are becoming more crowded, in order to achieve higher transmission rates, the 5G system has begun to look for millimeter waves with wider bandwidth. Therefore, in recent years, the design of millimeter wave systems has gradually received attention. This thesis proposes schemes about how to cancel the effect of interference on the millimeter wave system, while the millimeter wave system suffers from strong interference. Due to the limitation of millimeter wave transmission characteristics, a huge number of antennas must be used to achieve a longer transmission distance. In order to reduce the complexity and the energy consumption of the hardware design, the hybrid beamforming technique that combines analog and digital beamforming must be used. The received signal will go through the analog terminal first, and the first step can be performed on the analog terminal to cancel part of the effect of the interference on the system. The processed signal will then enter the digital terminal, and the digital terminal can effectively suppress the remaining interference. Generalized sidelobe canceller is a commonly used method for interference and noise cancellation, and thus we apply it in the digital beamforming combined with the multistage Wiener filter. Under the condition of limited samples, the dimensionality reduction algorithm of the multistage Wiener filter can reduce the computational complexity and effectively provide better performance.
