6-dimensional UAV tracking with dipole Green’s functions and ground reflections

Kurzfassung

Recent studies and surveys have shown the increasing use of drones (Unmanned Aerial Vehicles, UAVs) in many domains. However, the noise of UAVs is often perceived as annoying, prompting concerns regarding noise pollution and the need for novel acoustic certification guidelines. To measure UAV noise emissions in flight, their trajectory must be known. Given the integration of microphone arrays in existing sound measurement setups, acoustic tracking emerges as a reasonable method for UAV localization. This paper proposes a beamforming-based approach for short-range 6-dimensional UAV tracking with a dipole Green’s function and a mirror source model of the ground. The proposed method modifies the steering vector formulation by incorporating dipole directivity and coherent ground reflections directly into the acoustic transfer model. The dipole formulation is derived from the spatial derivative of the monopole Green’s function, while the reflected contribution is modeled using an image-source approach. This enables a more realistic representation of the measured UAV sound field, particularly in near-ground measurement scenarios. Conventional monopole and dipole beamforming are compared, with and without ground reflection. For evaluation, optical motion tracking results are serving as a “ground-truth”. The reflection-aware dipole formulation provides the most accurate localization, demonstrating that the inclusion of dipole directivity and coherent ground reflections significantly improves acoustic UAV tracking accuracy in both translational and rotational degrees of freedom.