在同調環境下的可適性寬頻波束成型技術

Abstract

可適性寬頻天線陣列利用在天線陣列中的每一個感測元件及延遲元件後擺入一個可以調整的權重係數（weight），藉著調整這些係數，使得天線陣列的輸出對想要信號的入射方向有最好的增益，同時抑制干擾信號入射方向的增益或產生零增益(null)，令想要信號對干擾信號加雜訊功率比（Signal to Interference plus Noise Ratio, SINR）為最大。 在論文中我們將利用二階的auto-regressive model來產生我們所需的寬頻訊號，並研究波束成形技術在此寬頻訊號遭遇信號同調（coherence）的問題。同調信號干擾會造成信號抵消現象，使得波束成型器失敗，而在傳統上有Shan與Kailath的空間平均法(Spatial Smoothing technique)及其衍生的權重式空間平均法（Weighted Spatial Smoothing）與漸進式空間平均法(Progressive Spatial Smoothing)來解決這個問題。 本文首先使用Frost提出的LCMV架構並配合上述傳統的演算法，接著主要提出TBTAM (Toeplitz-Block-Toeplitz Approximation Method)與I-TBTAM(Iteration-TBTAM)配合AIC、MDL作為最佳解決信號同調干擾問題的方法。最後引入寬頻廣義旁辦消除器架構作為比較。

An adaptive broadband antenna array is a collection of sensor elements and tap delay line whose outputs are combined by an iteratively adjusted combining (or weight) vector so as to pass a desired signal on desired arrival directions with minimum distortion while rejecting interfering signals which makes the maximum output SINR. Here, we use second order auto-regressive model to generate broadband signals, and consider the beamforming technique under coherent environment that cause signal cancellation phenomena. The phenomena can cause signal loss in the case of narrowband signals, or cause significant signal distortion in the case of wideband signals. Traditionally, one method is known as spatial smoothing technique which was proposed by Shan and Kailath, and the derivational methods as weighted spatial smoothing and progressive spatial smoothing. In this paper, we use LCMV structure of Frost, and propose our method TBTAM (Toeplitz–Block-Toeplitz Approximation Method) and I-TBTAM (Iteration-TBTAM) combining with AIC,MDL method as the optimum beamforming technique under coherent environment. Finally, we present broadband general sideband canceller as compared with LCMV.