應用訊號方位估測演算法於雷達目標搜索

Abstract

本論文探討利用訊號方位估測(Direction of Arrival Estimation)方法在雷達目標偵測技術之應用，並以MUSIC演算法實現該具備訊號方位估測功能之雷達系統。於雷達系統中常利用目標物之反射波偵測目標位置，而此訊號通常為高度相關性訊號。本論文利用spatial smoothing的技巧解決傳統MUSIC演算法無法解析高度相關性訊號的問題。此外，在系統軟硬體架構方面，包含發射機電路、接收機電路及資料擷取 (DAQ) 裝置，並利用NI LabVIEW軟體實現訊號方位估測演算法及訊號處理。實驗結果證實本系統於無反射室及複雜環境中皆能準確判斷多目標之訊號方位，特別是具有高度相關特性之訊號，最大角度誤差可控制在5度內。

This thesis investigates the method of Direction of Arrival (DoA) Estimation for RADAR application. A DoA Estimation system is implemented based on (Multiple Signal Classification) MUSIC algorithm. Different from the transmitted signals having various codes in the wireless communication scenery, the waves scattered from multiple targets are highly correlated attributed to the same illuminator (transmitter), resulting in singular covariance matrix and degrading the performance of conventionally used MUSIC algorithm. To resolve this problem, forward-backward spatial smoothing technique is used for correlated signal classification. The overall system includes a transmitter (a horn antenna), one-dimensional receiving antennas array and data acquisition (DAQ) device. The software NI LabVIEW is exploited for signal processing and implementation of the algorithm. The measured results indicate that the system has excellent performance on multiple-target detection, particularly for coherent signals, and the maximum angle error is 5 degree.