Real-time kinematic positioning for unmanned air vehicles

Real-Time Kinematic (RTK) positioning with lowcost mass-market Global Navigation Satellite System (GNSS) receivers and antennas is attractive for precise landing of Unmanned Air Vehicles (UAV). Code multipath mitigation and a reliable resolution of the carrier phase integer ambiguities are two challenges. In this paper, we describe an RTK method which estimates a code multipath parameter for each double difference to fully exploit the temporal correlation of multipath and to prevent a mapping of the multipath into the baseline and ambiguities. The ambiguity fixing is performed in two phases: In the first phase, sets of integer candidate vectors are derived from the float solution at multiple epochs. We merge the sets of candidate vectors to increase the likelihood of including the correct candidate vector in the set of candidates. In the second phase, we track each candidate vector by determining a single epoch baseline estimate for each candidate vector. The respective measurement residuals are accumulated over time to increase the likelihood of selecting the correct candidate. The proposed method was applied to real measurements from two low-cost ANavS GPS modules and enabled a millimeterlevel positioning accuracy despite code multipath of up to 50 m.