UAV Swarm Localization via Information Geometry and Multidimensional Scaling

In recent years, with the wide application of drones and other real-life applications, the positioning and navigation technologies supporting them have developed rapidly. However, single navigation means still face bottlenecks in accuracy and stability in complex environments: for example, satellite navigation is susceptible to obstruction interference, inertial navigation accumulates serious errors, and radio and visual navigation are limited by environment and equipment performance, making it difficult to realize all-weather, all-geography high-precision positioning. In response to the above problems, multi-source fusion has made great progress in existing navigation technology, and now multi-source fusion has become an important direction to improve navigation performance, but the existing fusion algorithms are often difficult to handle heterogeneous, non-linear, high-dimensional data due to the lack of real-time and insufficient information integration. Especially in three-dimensional space, fusing multiple navigation data to achieve accurate and efficient localization is still a major challenge. In this paper, we propose a high dynamic UAV cluster localization method based on the information geometry theory to address the above problems, by unifying all kinds of navigation information into probability distributions, fusing the data based on the information geometry framework, and finally combining the geometric accuracy factor of each UAV in the cluster itself relative to the base station with the multidimensional scale transformation (MDS) to achieve high-precision localization, and the simulation results show that this algorithm has the ability of high precision with adapting high dynamic system.