Distributed UAV cluster fusion positioning method with heterogeneous navigation information

With the development of the Internet of Things, drones have gradually become the main equipment in fields such as road monitoring, transportation, relay signal stations, etc. Lightweight drone swarms, due to their wide coverage and high efficiency, are the trend of drone development. However, existing UAV navigation sources such as satellite navigation, inertial navigation, and radio navigation are unable to independently realize high-precision, high real-time navigation and positioning in all time and space. Also, navigation information fusion technology represented by Kalman filtering cannot meet the needs of drone swarm navigation. Therefore, developing distributed multi-type navigation information fusion positioning is a core issue for the development of drone swarms. To address these issues, this paper proposes a heterogeneous navigation source fusion method based on an information geometry framework. This method unifies various types of navigation information parameters into an information probability model, reducing the impact of discrepancies in time, output frequency, and information format. Secondly, a parallel distributed fusion filtering framework is designed to quickly merge distributed navigation sources for drone swarms. In the constructed drone swarm experiment, the method proposed in this paper was compared with existing collaborative positioning methods in scenarios such as random motion and signal mutation. The results showed that the method proposed in this paper has faster fusion positioning speed and error suppression capabilities.