Exploring GNSS Crowdsourcing Feasibility: Combinations of Measurements for Modeling Smartphone and Higher End GNSS Receiver Performance

Abstract

GNSS receiver data crowdsourcing is of interest for multiple applications, e.g., weather monitoring. The bottleneck in this technology is the quality of the GNSS receivers. Therefore, we lay out in an introductory manner the steps to estimate the performance of an arbitrary GNSS receiver via the measurement errors related to its instrumentation. Specifically, we do not need to know the position of the receiver antenna, which allows also for the assessment of smartphone GNSS receivers having integrated antennas. Moreover, the method is independent of atmospheric errors so that no ionospheric or tropospheric correction services provided by base stations are needed. Error models for performance evaluation can be calculated from receiver RINEX (receiver independent exchange format)data using only ephemeris corrections. For the results, we present the quality of different receiver grades through parametrized error models that are likely to be helpful in stochastic modeling, e.g., for Kalman filters, and in assessing GNSS receiver qualities for crowdsourcing applications. Currently, the typical positioning precision for the latest smartphone receivers is around the decimeter level, while for a professional-grade receiver, it is within a few millimeters.

Analyzing Android GNSS Raw Measurements Flags Detection Mechanisms for Collaborative Positioning in Urban Environment

“…Multipath flags tend to be inconsistent whereas cycle slip flags were proven to be coherent despite their high false alarm activation frequency. Previous studies conducted in static scenarios [7] [8] concluded that collaborative smartphone positioning could be achieved and that GNSS measurements can be modeled. However, Android multipath and cycle slip indicators might not be used as reference parameters to qualitatively assess smartphone positioning performance.…”

Why It Makes Sense to Use High Cost Sensors to Do Low Cost Sensor Research

“…High-and low-cost GNSS receivers may be investigated experimentally in order to obtain appropriate error models and error parameters for simulations [Lehtola et al, 2019b]. In the experimental campaign, see Figure 2, low-cost receivers (smartphones) performed more poorly than the high-cost ones in several regards.…”

“…Figure 2: Experimental measurement campaign with four different GNSS receivers, one with two antenna configurations. Reproduced from [Lehtola et al, 2019b].…”

Towards tropospheric delay estimation using GNSS smartphone receiver network

“…Furthermore, as we are considering crowdsourcing, this uncertainty needs also to include the errors originating from the measurement devices, i.e., GNSS receivers (see Tables 1-3). These uncertainties have been studied in Lehtola et al (2019) and the references there-in.…”