| 摘要: | 近年來無人機（UnmannedAerial Vehicle, UAV）技術快速發
展，應用領域涵蓋環境監測、空拍地理資訊、多媒體製作與基礎
設施檢查等。然而，無人機的普及也帶來潛在的空域安全與隱私
風險，非法侵入禁區或受限區域的事件時有發生，因此即時且準
確的無人機方向估計技術成為重要課題。
本研究提出一套基於射頻訊號的方向估計方法，透過接收
UAV 所發射之無線訊號，利用陣列天線進行方向估計（Direction
of Arrival, DOA）。在演算法設計方面，採用計算複雜度低、估計
即時性的單脈衝演算法（Monopulse）為核心，並結合具有全向監
測能力之均勻圓形天線陣列（UniformCircularArray,UCA）進行和
波束與差波束之合成設計。UCA在幾何結構上具備方位對稱性與
360 度無遮蔽視野，相較於傳統均勻線性陣列（ULA）在覆蓋範圍
上更具優勢。
為進一步提升角度估計的穩定性與抗雜訊能力，系統中引入
無跡卡爾曼濾波器（UnscentedKalman Filter, UKF）對 Monopulse
演算法輸出的非線性量測進行濾波與平滑處理，有效降低雜訊影
響並提升追蹤精度。此外，本文亦針對UCA的波束成型特性（如
主瓣寬度、方向穩定性與對稱性）進行分析，並比較不同幾何參數
下對Monopulse 估計表現的影響。
最後，為驗證該系統於實務應用中的可行性與穩定性，本文
於軟體定義無線電（SoftwareDefinedRadio,SDR）平台上實作基於
UCA架構之方向估計系統，並與模擬結果進行比對，以驗證其在
即時UAV追蹤情境下的整體效能表現。此外，亦於模擬中與分時
L-Shaped 陣列架構進行比較，分析不同陣列配置對估計精度與角
度覆蓋範圍的影響，佐證本系統於UAV方向追蹤應用中之效能與
可擴展性。;In recent years, the rapid development of unmanned aerial vehicles (UAVs) has brought forth a wide range of applications, including environmental monitoring, aerial mapping, multimedia production, and in frastructure inspection. However, the increasing deployment of UAVs has also raised concerns regarding airspace safety and privacy violations.
UnauthorizedUAVintrusionsintorestrictedareaspresentpotentialthreats, making real-time and accurate direction-of-arrival (DOA) estimation a critical research topic.
This study proposes a radio frequency-based direction estimation system that receives UAV-emitted signals and performs DOA estimation using an antenna array. A low-complexity and high-accuracy monopulse algorithm is adopted as the core estimator, integrated with a uniform cir
cular array (UCA) capable of forming sum and difference beams. The symmetric structure of UCA enables 360-degree omnidirectional coverage, offering greater spatial flexibility compared to traditional uniform linear arrays (ULA).
To further enhance angular estimation robustness under noisy environments, the proposed system incorporates an Unscented Kalman Filter (UKF) to smooth and track the nonlinear angle estimates from the
monopulse algorithm, effectively reducing noise-induced fluctuations.
Additionally, the beamforming characteristics of UCA,including main lobe width, direction stability, and symmetry,are analyzed to evaluate their impact on estimation accuracy under different geometric configurations.
Finally, to validate the practical feasibility of the system, a proto type based on Software Defined Radio (SDR) hardware is implemented
using a UCA architecture. The experimental results are compared with simulation outcomes to verify the system＇s accuracy, low latency, and scalability in real-time UAVtrackingscenarios. Furthermore, simulation based comparisons with a time-division L-shaped array are conducted to analyze differences in angular coverage and estimation precision, demonstrating the effectiveness and versatility of the proposed system in UAV direction tracking applications. |
