Recently, as the use of Unmanned Aerial Vehicles (UAVs) is increasing, a Local-Area Differential Global Navigation Satellite System (LADGNSS) that provides users with error corrections and integrity information was proposed to help UAVs carry out their assigned missions. In order to operate UAVs supported by LADGNSS, the minimum separation distance has to be defined to prevent collisions. This paper proposes a methodology to estimate the desired separation distance. To calculate the separation distance, residual errors which cannot be completely corrected by LADGNSS and path-following errors which depend on the controller’s performance should be considered. Meanwhile, the multipath error included in residual errors cannot be expressed theoretically due to its unpredictable randomness, and path-following errors differ depending on which controller is used. Therefore, this paper focuses on devising a multipath error model and a path-following error model with each type of empirical data. Furthermore, specific values of the separation distance satisfying the probabilistic level to ensure the safety of UAV can be calculated with these models by applying risk assessment methodology.

Keywords: UAVs, LADGNSS, separation distance