Smart-RTK: Multi-GNSS Kinematic Positioning Approach on Android Smart Devices with Doppler-Smoothed-Code Filter and Constant Acceleration Model

Abstract Global Navigation Satellite System (GNSS), once dedicated to military and geodetic applications, is entering civilian life with the development of low-cost internal multi-GNSS chips in mass-market smart devices. The recently enabled Application Programming Interface (API) to GNSS raw measurement in Android Nougat operating system, make it possible to implement precise positioning technology on Android smart devices, such as Real-Time Kinematic Positioning (RTK) and Precise Point Positioning (PPP). An optimized kinematic positioning approach on Android smart devices with Doppler-Smoothed-Code (DSC) filter and Constant Acceleration (CA) model is assessed in this paper. In this optimized approach, DSC filter is used to reduce the code measurement noise, which is extremely high on smart devices and CA model is used to accurately predict the kinematic state of smart devices. The optimized approach is named Smart-RTK for its applicability to smart devices, respectively. The performance of the Smart-RTK approach is validated by two Google/HTC Nexus 9 tablets separately under stationary, walking, and vehicular condition. The numerical experiments show the significant improvement on positioning accuracy and continuity. The positioning Root Mean Square Error (RMSE) in horizontal component reaches about 0.3 ∼ 0.6 m in stationary condition and 0.4 ∼ 0.7 m in walking condition, improved by about 85% compared with that of chipset original solutions. In the subsequent vehicular experiment, the horizontal positioning RMSE is about 0.85 m, 50% better than that of chipset solutions.