Multiple-wave source localization using UAVs in NLOS environments

Abstract

Localization techniques for unknown radio wave sources are crucial from the perspective of efficient utilization of frequency resources. The authors have studied methods for localizing a single wave source using unmanned aerial vehicles (UAVs) in non-line-of-sight (NLOS) environments based on maximum likelihood estimation. In this study, we propose a localization method for multiple wave sources by extending the single-wave source localization method. In the proposed method, the direction of arrivals (DoAs) at UAVs is modeled with a mixture of von-Mises distributions, and the wave sources are estimated by superimposing the DoA distributions estimated at the UAVs. The proposed method is validated with a simple simulation experiment with two wave sources.