LAUSR

ABSTRACT Combining the Global Positioning System (GPS) and the developing Chinese BeiDou Navigation Satellite System (BDS) positioning has received increased attention. However, most research is based on the empirical stochastic model, and the stochastic model of the BDS is usually analogous to GPS. In this contribution, we first estimate the variances of observables for all GPS and BDS satellites and fit them with predefined exponential types of elevation-dependent functions. The established stochastic model is then used in instantaneous short-baseline real-time kinematic (RTK) positioning. The results show that the precisions of both GPS and BDS observables can fit the elevation-dependent function model well, except the BDS geostationary earth orbit (GEO) satellites; thus, we downweight BDS GEO observables. Our refined stochastic model can produce improved ambiguity resolution performance. We also find that the combined system can improve the ambiguity resolution performance and positioning accuracy, especially with high elevation cut-off angles.