Characteristics of raw multi-GNSS measurement error from Google Android smart devices

Abstract

Developers targeting Android platforms can obtain raw GNSS measurements, which can achieve submeter or even decimeter-level positioning accuracy. An accurate receiver measurement error model is an important prerequisite for precise positioning with smart devices. Therefore, we analyzed the measurement error characteristics of raw GNSS data from smart devices using both embedded and external antennas. We find that the GNSS signals produced by smart devices have non-uniform signal strengths, rapid C/N0 variations, and low C/N0 at high elevations. The pseudorange noise is about 10 times larger than that from geodetic receivers; the carrier phase noise of Nexus 9 is 3–5 times larger than that of geodetic receivers, and unexpectedly is half of that of μ-blox. We provide theoretical parameters for the noise versus C/N0 models of the GNSS chipset for different smart devices. Unfortunately, the carrier phase tracking of Samsung Galaxy S8 and Huawei Honor v8 are discontinuous due to the duty-cycle issue, which results in greater noise and carrier phase unavailability. Moreover, we found two unique error characteristics of the carrier phase available from Nexus 9 anomalous “jagged” distribution and random initial phase biases, which is evident in the controlled environment test. Finally, we obtained promising positioning results: the horizontal and vertical RMS of pseudorange single-point positioning are about 10–20 m; the static carrier phase relative positioning (CRP) solutions of Nexus 9 can achieve centimeter-level precision, whereas both horizontal and vertical STDs are about 1 cm or better but with decimeter-level biases. When using an external antenna, the resulting biases are as small as a few centimeters. Encouragingly, the actual vehicle test results showed that the STD of the Nexus 9 kinematic CRP 3D-distance error is 0.169 m, and the percentages of errors falling into ± 0.1 m and ± 0.5 m are 63.59% and 100%, respectively. Furthermore, multi-GNSS is able to provide more reliable position services in GNSS-adverse environments.