Measurement-Domain Cooperative Navigation for Multi-UAV Systems Augmented by Relative Positions

This paper proposes a novel cooperative navigation scheme to support multiple Unmanned Aerial Vehicle (UAV) formation flight. Comparing to standalone Global Navigation Satellite System (GNSS), the new approach takes advantage of inner communication capability among UAVs to achieve navigation information fusion. As a result, the navigation accuracy, robustness and continuity can be improved. This is of particular significance in urban environments, because the user receivers are subject to high multipath, non-line-of-sight signal reception and signal blockage. With the aid of relative sensing technologies, all the GNSS observations are firstly combined to estimate the position of a designated point (defined as virtual centroid). Then, the positions of each individual UAV are derived by fusing the position estimate of virtual centroid and the known relative positions. This method is implemented, validated, and analyzed through a series of simulations, and the results suggest significant accuracy improvement as compared to standalone GNSS approaches. Moreover, sensitivity analyses are carried out to address the impact of environmental changes on navigation performance.