Single-frequency BDS/GPS RTK with low-cost U-blox receivers

Technology of low-cost single-frequency RTK (real time kinematic) is booming in popular GNSS application, however, the poor data quality and hardware performance of low-cost receivers bring about great challenges to its positioning algorithm. In this contribution, a BDS/GPS single-frequency inexpensive navigation and positioning (SINAP) system is developed to achieve high-precise high-stability single-frequency RTK positioning with low-cost GNSS receivers. In data processing, we implement the Kalman filter estimation with constant velocity (CV) and constant acceleration (CA) dynamic models. Since it is not feasible to fix the whole integer ambiguities in real applications, we utilize the PAR (partial ambiguity resolution) method to acquire a subset of integer ambiguities effectively. Experiments with static and kinematic data were carried out by using low-cost single-frequency u-blox receivers. To emphasize the benefits of the proposed method, we compare the performance of SINAP to those of the open source software RTKLIB and the u-blox chip. It is shown that results of SINAP are competitive to the ones of u-blox and outperform the ones of RTKLIB. Moreover, the performance of GPS-only, BDS-only and combination of BDS/GPS are investigated. The results reveal that using BDS/GPS can achieve higher fix-rate and higher positioning accuracy than using GPS or BDS only. Additionally, with Kalman filter, we compare the CA and the CV dynamic models using kinematic data, which illustrates that both models can provide comparatively ideal results with no obvious difference.