用於扇形分區分碼多重進接系統之波束空-時干擾消除犁耙接收器

Abstract

在分碼多重進接之蜂巢式通信系統，所有用戶都使用共同之頻段，干擾隨著用戶數增多而增加。傳統犁耙接收器無法完善消除多重進接干擾，於是能提供另一空間自由度的陣列天線，和充分利用干擾資訊的多用戶偵測技術就漸漸受到重視。扇形分區也被廣泛的用來增加系統容量，卻因常需做區域間切換，而增加系統額外運算量，但若改成由選擇分及波束方式來做動態區域的合成。就可減少系統切換的次數，並可平衡各區域用戶負載數量。在本論文中，將會探討各種空-時二維犁耙接收器之性能，雖然空-時聯合處理器能提供優越之性能，但卻因處理自由度太大，需要很複雜的運算量，極不利即時處理之應用，為了避免這問題，我們採用波束空-時方法來降低訊號處理維度，波束空-時處理機制，包括兩組處理器，一組為具適應性波束合成處理器，另一組為具適應性相關處理器，波束合成處理器用來有效消除區域外干擾，和接收指定空間區域內多重路徑信號，而相關處理器則用來消除區域內多重進接強干擾，和接收指定時域內多重路徑信號，最後這些處理得到的資料，則透過一個以最大信號雜訊比方法，建設性的捕捉多重路徑信號。這兩種處理器都是透過一種旁波消除方式來實現，並可由通道估測協助和部分適應性處理結合，來降低系統複雜度和改善系統收斂情形。

In the cellular code division multiple access (CDMA) communication systems, each user communicate via the same radio channel such that interference increases with the number of subscribers. Traditional RAKE receivers cannot handle the multiple access interference (MAI) well, and this prompts the use of extra spatial degree of freedom (antenna array) and extra information of MAI (multiuser detection). Cell sectorization also has been widely used for improving system capacity. However, it causes extra computational load due to handoff. Through the dynamic sector synthesis by selecting of diversity beamformer, this allows to balance the traffic loading across sectors and to manage handoff overhead. In this thesis, we will discuss the performance of space-time 2-D RAKE receivers. In despite of the superior performance of the space-time joint processor, it needs more complex computations and the real-time processing is hard to achieve when the joint degree of freedom is large. To remedy this, we make use of the concept of reduced-dimension beamspace-time processing. The beamspace-time processing scheme involves two processors, a set of adaptive beamformers and a set of adaptive correlators. The beamformers are constructed to provide effective suppression of unwanted interference and reception of signals from a prescribed working sector. And the correlators are constructed to suppress the in-sector strong MAI and to receive the diversity paths from prescribed time region. Finally, the output data obtained by these processors are maximum ratio combined to capture the signal multipath components coherently. In particular, both two processors can be realized in the form of generalized sidelobe canceller (GSC) with the aid of channel estimation, and partially adaptivity is incorporated for reduce complexity processing and improved convergence.