A Review of Small UAV Navigation System Based on Multisource Sensor Fusion

In recent years, unmanned aircraft systems (UASs) have played an increasingly significant role in the military and civil fields. The flight control system, as the "hub" of an unmanned aerial vehicle (UAV), is responsible for the key function of autonomous flight, while a reliable and stable navigation system provides important information such as position status for flight control and represents the "sensory" function of the UAV. A highly autonomous and credible UAV requires a navigation system that meets specific requirements for accuracy, integrity, and continuity, resulting in a multitude of sensors on-board the UAV that are heterogeneous, redundant, and multisource, creating a highly complex navigation system. In this article, we review multisensor fusion (MSF) technology for small UAVs over the last 20 years and provide an overview of three typical multisource fusion architectures based on filtering, factor graph optimization, and data-driven, focusing on inductive identification of key technologies for multisource information fusion state estimation systems, including calibration techniques to improve data quality, observability analysis to provide theoretical support, additional model constraint correction using aircraft, and resilient fusion management techniques across all sources. Finally, we propose future directions for UAS navigation systems to address the limitations of the existing systems.