High Integrity GNSS Navigation and Safe Separation Distance to Support Local-Area UAV Networks

Abstract

A local-area Unmanned Aerial Vehicle (UAVs) network can provide local-area differential corrections and optimized guidance, including the maintenance of safe separation distances, for precise and safe operation of the UAVs. In networked UAV operations, only the uncorrelated component of position error between two UAVs in the same network contributes to potential separation violations, because “in-network” UAVs share the same source of guidance and differential corrections. Therefore, models of uncorrelated errors need to be defined to establish safe separation distances between UAVs. This paper develops a methodology to estimate safe separation distance for UAVs which share the same source of guidance and local-area differential corrections. Using this methodology, ionospheric and tropospheric models for Navigation System Error (NSE) are developed theoretically. The airborne multipath error and Flight Technical Error (FTE) components which depend on hardware, environment, and operational conditions are also determined though UAV flight experiments. The standard deviation of FTE was estimated in both straight and curved flight segments. The results show that a lateral FTE error of 0.78 meters in curved segments is much greater than lateral FTE in straight segments and vertical FTE in both segments (all about 0.4 meters). This is due to the momentum of the UAV when it is taking turns and the limited controller response time of the UAV. The flight experiments also show that UAV multipath errors are reasonably well-bounded by the standard airborne multipath model developed for the Ground Based Augmentation System (GBAS). From the estimated error models, simulations of safe separation distances were conducted under the “in-network” UAVs scenario. Simulation results for a 24-satellite GPS constellation and the probabilities of safe separation suggested in prior work show that vertical separations between UAVs in the same network vary from about 3.0 – 6.5 meters, while horizontal separations vary between 2.2 and 3.5 meters. These values change with time according to the visible GPS satellite geometry and are known to the controller in real time.