Previous Abstract Return to Session A2 Next Abstract

Session A2: Applications of GNSS Measurements from Smartphones

RTK-GNSS with Smartphone in Moving Vehicles Using GNSS Repeater
Nobuaki Kubo, Tomohiro Ozeki, Kaito Kobayashi, Tokyo University if Marine Science and Technology
Date/Time: Wednesday, Sep. 13, 4:00 p.m.

Research and development using smartphone observation data has progressed rapidly since ION GNSS+ smartphone sessions started 5–6 years ago, and there has been an increasing number of presentations on implementing RTK-GNSS in smartphones. It has been reported that there are some issues with the GNSS antennas of smartphones. The signal level of the GNSS antenna of the smartphone was lower than that of the standard GNSS patch antenna. Hence, tracking the carrier phase was difficult for most satellites, resulting in an insufficient number of satellites for the RTK-GNSS. The instability of the phase center of the GNSS antenna of the smartphone was also noted.
Therefore, we came up with the idea of using the GNSS repeater to address the issues mentioned above.
While training using a GNSS repeater (GPSS Operations Enabled) in a university classroom, we noticed that the signal level of the smartphone (Google Pixel 5a) was reasonably stable. This is because the signal level and phase center problems of carrier phase tracking as described above were resolved to some extent by receiving GNSS signals from the GNSS repeater. Variations in the phase center of the GNSS antenna of the smartphone are less likely to occur because the GNSS repeater has a single antenna and the signal comes from one direction. Furthermore, the signal strength is sufficiently high and stable as long as the GNSS repeater and smartphone are properly set up. Using the GNSS observation data at both the base station and in the smartphone in hand, we performed post-processing of RTK-GNSS data and obtained FIX solutions.
Then, we changed the test location from the room to inside a car. The standard GNSS antenna was mounted on the rooftop of the car and connected to a GNSS repeater inside the car. Before conducting the test, we confirmed that the strength of the signal was within the range of weak radio waves permitted by the Japanese Radio Act. Specifically, the electric-field strength must be within 35 µV/m at 3 m. We simulated the radio wave strength of the GNSS repeater to ensure that it was within the specified range. The GNSS repeater and smartphone were placed in a steel box inside the car to further reduce leakage. During the experiment, we verified data from a commercial GNSS receiver at all four sides of the car and confirmed that turning the GNSS repeater on and off had no effect on the signal level of the commercial GNSS receiver.
The experiment using the car was conducted at the Etchujima campus of the Tokyo University of Marine Science and Technology. A standard GNSS antenna was installed on the roof of the car. The antenna was connected to a commercially available u-blox F9P receiver and a GNSS repeater (GPSS Operations Enabled). The smartphone (Google Pixel 5a) received signals from the GNSS repeater. The GNSS observation data were obtained using rinex ON software. The car was stopped or driven at a low speed in the campus. GNSS observations at the base station on campus were obtained simultaneously using the u-blox F9P receiver. Although the surrounding environment consisted of buildings and trees, the receiver was able to obtain almost 100% FIX solutions, and these positions were used to evaluate the results from post-processed RTK-GNSS data using the smartphone. This RTK engine was developed by modifying the RTKLIB. The GNSS observation data acquired on the smartphone were single-frequency GPS/QZSS/GALILEO/BDS/GLONASS. A few seconds were required to obtain the initial FIX solution. Once the FIX solution was prepared, we obtained stable FIX solutions even when the car moved at low speeds. The accuracy of pseudo-range measurements were enhanced using the velocity vector as its accuracy from the Doppler frequency was high. Through these car tests, we could obtain almost 100% FIX solutions based on the measurements from the smartphone. The accuracy was in centimeters compared to the data of the u-blox F9P receiver. This paper also discusses a few issues in the inter-system bias for positioning.

Previous Abstract Return to Session A2 Next Abstract