Impact of multipath on anchor selection strategies for UAV localization in mobile networks

Positioning is gaining increasing importance in mobile systems due to its potential integration with many 5G verticals. In non-terrestrial networks, positioning of aerial vehicles constitutes a crucial element of the communication, control, and applications of these systems. Unmanned aerial vehicles (UAVs) have become strategic actors in this technical and economic trend. A distinctive feature of UAV positioning in mobile networks is that signals from multiple base stations, or anchors, can typically be received and employed for localization, as propagation conditions, including the fundamental line of sight availability, improve with altitude. However, this advantage poses serious issues in terms of computational complexity, channel occupation, and latency unless an appropriate selection of available anchors is performed. In this work, we investigate and discuss several anchor selection strategies under realistic propagation scenarios, with an emphasis on the impact of multipath, according to the 3GPP channel models for aerial vehicles. The application of these strategies to channels affected by multipath using a standard least squares positioning technique reveals interesting properties and design directions for a feasible solution to the anchor selection problem. Notably, we show that even 3 or 4 well-selected anchors are often sufficient to achieve sub-meter 3D positioning accuracy.