滑坡是一種主要的自然災害，在全球範圍內造成大量的人員傷亡和財產損失，其發生頻率由於氣候變化和人類活動而不斷增加。傳統的監測方法在惡劣天氣和崎嶇地形中面臨挑戰，而地基合成孔徑雷達（GB-SAR）能夠提供不受天氣影響的連續高解析度監測。然而GB-SAR的性能可能會受到環境因素和系統組態的影響。本研究集成 GB-SAR 與向量網路分析儀（VNA），透過角反射器反射分析，確定本系統的最優參數。實驗在一個微波暗室中進行（尺寸為 6.0 米（長）× 3.0 米（寬）× 2.7 米（高）），通過自動掃描在 橫向移動57 個位置上(間距 2.5 cm )，並在不同距離（110 cm至 410 cm）放置角反射器，進行GB-SAR接收雷達回波，後續使用 MATLAB 進行信號處理，包括距離與方位壓縮、峰值檢測，以及基於 DBSCAN 與 KMeans 的聚類分析。該實驗重複十次以確保結果的一致性。系統設計還包括用於移動野外部署的準備工作。實驗結果提出目前最佳參數：頻率範圍為 5.0–5.6 GHz，頻寬為 200 kHz，發射功率為 +15 dBm，掃描長度為 1.4 米。雷達測量結果穩定且可重複，能夠準確檢測和定位角反射器。聚類分析有效地區分不同的反射模式，表明該系統具有精確監測地表位移的能力。綜合上述， GB-SAR 與 VNA 結合可提高雷達測量的準確性與可靠性。經過驗證的系統在複雜地形下具備即時滑坡監測與預警的潛力，有助於實現更安全的災害管理。;Landslides are a major natural hazard causing significant casualties and damage worldwide, with increasing frequency linked to climate change and human activities. Traditional monitoring methods face challenges in adverse weather and rough terrain, while Ground-Based Synthetic Aperture Radar (GB-SAR) provides continuous, high-resolution monitoring independent of weather. However, GB-SAR performance can be affected by environmental factors and system configuration. The research involved integrating GB-SAR with VNA (Vector Network Analyzer), investigating reflection characteristics of corner reflectors, and determining optimal system parameters. Experiments were performed in a microwave chamber (6.0 m (L) × 3.0 m (W) × 2.7 m (H)) using automated scanning across 57 positions at 2.5 cm intervals with corner reflectors placed at varying distances (110 cm – 410 cm). Data collection was followed by MATLAB-based signal processing including range and azimuth compression, peak detection, and clustering analysis using DBSCAN and KMeans. The procedure was repeated ten times to ensure consistency. The system design also included preparation for mobile field deployment. Results identified optimal operational parameters: 5.0–5.6 GHz frequency range, 200 kHz bandwidth, +15 dBm power, and a scan length of 1.4 m. The radar measurements were stable and repeatable, accurately detecting and localizing corner reflectors. Clustering analysis effectively distinguished reflection patterns, indicating the system’s capability for precise ground displacement monitoring. Results show, integrating GB-SAR with VNA improves radar measurement accuracy and reliability. The validated system shows promise for real-time landslide monitoring and early warning applications in complex terrain, facilitating safer disaster management.
