Multi-GNSS RTK positioning with integer ambiguity resolution: from double-differenced to single-differenced

Abstract

The development of global navigation satellite systems (GNSS), especially BeiDou navigation satellite system with global coverage (BDS-3), has brought benefits for high-precision positioning. Real-time kinematic (RTK) positioning based on double-differenced (DD) observations has been widely used in high-precision positioning as common errors are eliminated. However, the biases at the receiver-end, which can be dynamically constrained, are also eliminated during the DD process. Therefore, it makes sense to turn RTK from DD to single-differenced (SD) as the advantages of dynamic constraints of the receiver biases can be exploited. In this contribution, we first present RTK models based on DD observations suitable for short, medium and long baselines. Then, based on SD observations, the full-rank RTK models are constructed with the S-system theory. Using observations from GPS, BDS-3 and Galileo, we first demonstrate the short-term stability of receiver-related biases. The SD RTK positioning performance with the stability of those receiver-related biases regarding integer ambiguity resolution success rate and positioning accuracy are analyzed. With those biases, RTK can achieve high performance, and this is more advantageous in multi-GNSS scenarios.