Global navigation satellite system (GNSS) raw data were made available in the application programming interface (API) starting from version 7.0 of the Android operating system. This opens possibilities for precise positioning with Android devices, as externally generated GNSS corrections can now be included in the positioning estimation in a convenient way. The Nexus 9 tablet is a good candidate for an early assessment of the raw GNSS observables and the corresponding derived precise positions, as it also supports many of the optional features and observation types presented by the API, including carrier phase observations which play an important role in many precise positioning techniques. It is known from the previous studies that poor handling of multipath in smartphones and tablets is a big challenge when it comes to precise GNSS positioning with these kinds of devices. Hence, this study assesses the raw GNSS observations and the calculated precise positions of the Nexus 9 tablet in two experimental setups with different multipath impacts. In addition, various biases of the observations are determined, some of which is not present on high-grade geodetic receivers. The analysis is done for GPS and GLONASS, which are supported by the Nexus 9 tablet. The study shows that multipath plays an important role for the expected accuracy of the calculated precise positions, both due to the induced error on the measurements, and due to loss of lock of the GNSS signals, which significantly affects precise positioning from carrier-phase measurements. Position accuracy ranges from just below 1m to a few decimeters between the experimental setups with moderate and low levels of multipath respectively, for positioning based on carrier-phase observations. It is, furthermore, demonstrated that consideration of code inter-system biases and code inter-frequency biases of the Nexus 9 tablet are crucial for differential GNSS with multiple GNSS systems and when GLONASS observations are used in the positioning solutions. Besides these expected biases, also other less expected behaviors were discovered in the Nexus 9 GNSS observations, including various drifts for the code and phase observations. This study also proposes some strategies to handle these.
