Environmental Boundary Estimation of Urban Areas Using FMCW Radar With Data Clustering for UAV Detection

Abstract The increasing presence of unmanned aerial vehicles (UAVs) in urban environments presents challenges for reliable detection due to clutter caused by buildings, infrastructure, and other static structures. Conventional Doppler‐based radar methods often struggle in such conditions, particularly when UAVs exhibit low radial velocities. This paper presents a non‐Doppler scattering‐point framework for UAV detection, based on a frequency‐modulated continuous wave (FMCW) radar system operating in the X‐band (9.950–10.026 GHz). The system utilizes a rotating platform for azimuth scanning and applies azimuth‐range mapping combined with adaptive DBSCAN clustering and β ‐expanded convex hull boundary estimation to reduce false detections from static clutter. Experimental validation was conducted in a controlled urban setting with multiple buildings and varied UAV trajectories. The method was evaluated across several elevation angles, demonstrating consistent detection performance and improved distinction between UAV detections and points caused by environmental clutter. These results support the use of FMCW radar and spatial clustering techniques as an effective alternative to Doppler‐independent methods for UAV monitoring in complex environments.