在非理想環境下具有快速收斂與強健能力之波束成型技術

Abstract

可適性波束成型技術是透過調整各個天線陣列上的權重係數，來達到提取欲接收信號並消除干擾信號和雜訊的技術，主要可以分為兩種類型，第一種所需要的已知資訊為欲接收信號的入射角方向，此種類型稱作是指引式波束成型技術，例如linearly constrained minimum variance beamformer (LCMV)，而近年來發展出了另一種，不需要知道欲接收信號入射角方向，其運用信號的某些特徵，像是週期恆定性(cyclostationary)，來達到波束成型，此類型稱作是盲目式波束成型技術，而本論文會根據此兩種波束成型技術來發展出多種具快速收斂性且強健式的技術，以對抗非理想效應的影響。 本論文大致可分為兩個部分，第一部分為第三章，我們利用改良式的average cyclic correlation matrix algorithm (ACCM)，此方法是用迭代平均的方式(Iterative Averaging, IA)估計出在週期頻率誤差影響下最接近真實的週期共軛自相關矩陣，接著我們將Full Data-Dependent Loading方法引入改善此演算法的收斂性，如此便可發展出一套新的具快速收斂性及強健性的波束成型技術，並且我們以實驗模擬的方式證明了此方法的有效性。 第二部分為第四章以後，我們將原本僅用於單一接收信號的支持向量機-波束成型技術，推廣到多重接收信號也能使用，且由文獻中我們了解到支持向量機-波束成型技術的好處是他可以結合許多已知的強健式波束成型技術，幫助這些方法達到更好的收斂性(本論文中所提到的良好收斂性指的是使用少量資料點即可達到不錯的array output signal-to-interference-plus-noise ratio (Output SINR))甚至還會增強其效能(本論文所提到的效能是以Output SINR來評斷)，故我們的貢獻主要是將支持向量機-波束成型技術推廣為支持向量機-多重波束成型技術，並且結合了許多已知對抗非理想環境的強健式波束成型技術，如Cheng’s method、ACCM、Efficient Robust Array Beamforming(ERAB)、Compesaion method等等 ，發展在非理想環境下具有快速收斂性及強健式的波束成型技術。

Adaptive array beamforming which can extract signals of interest from specific angles while suppress interferences and noise by adjusting weights on the array elements. And they can be classified into two types. The first type we call it conventional beamforming. For conventional beamforming, the a priori information is the direction of the desired signal, and this kind of technique is also called steered-beam beamforming, for example linearly constrained minimum variance beamformer (LCMV). In recent years, another kind of adaptive array beamforming without the knowledge of the direction of the desired signal has been widely presented. It utilizes some characteristic of the signals to achieve beamforming, for example, signal cyclostationary and thus it is called “blind” beamformer. The purpose of this thesis is mainly to develop several fast convergent and robust techniques for both steered-beam beamformer and blind beamformer using cyclostationary in order to tackle the performance degradation under non-ideal enviroments. This thesis can be divided into two parts. Chapter 3 is the first part of this thesis. To deal with the cycle frequency error (CFE), we present an iterative averaging (IA) scheme in conjunction with average cyclic correlation matrix algorithm (ACCM) to estimate the actual cyclic correlation matrix. In addition, we apply Fully Data-Dependent method to improve the convergence of the proposed method. In this way, we would expect that the propsed method will be a fast convergent and robust beamforming techniques. We will present the simulation results to show the effectiveness of the proposed method. From the forth chapter, we present a new sheme of the suppert vectoe machines beamformer(SVM-beamformer) which can deal with multiple signals of interest (SOIs). The advantages of using SVM-beamformer (which is use to handle single SOI) are that it can combine with the existing robust array beamforming techniques and improve their convergence and performance. (In this thesis, the fast convergence means that we can use less data to get good performance, and the performance means the array output signal-to-interference-plus-noise ratio (SINR)). Besides, we note that it can also be applied to the environment of multiple SOIs. Therefore, we combine several robust array beamforming techiques with the SVM-beamformer to deal with the problem of array beamforming under multiple SOIs and non-ideal enviroments.