Geodetic monitoring experiment by low-cost GNSS receivers and gogps positioning engine

Transcription

1 Geodetic monitoring experiment by low-cost GNSS receivers and gogps positioning engine Stefano Caldera1, Eugenio Realini1, Daisuke Yoshida2 1 2 Geomatics Research & Development (GReD) srl, c/o ComoNExT, via Cavour 2, Lomazzo, Como, Italy Graduate School for Creative Cities, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka, Japan Abstract This paper describes an experiment of GNSS-based monitoring by means of different models and brands of low-cost receivers. Single-frequency observations were processed by gogps software, in relative positioning mode with respect to a virtual reference station and three permanent stations. The experiment was carried out in Osaka, Japan; the results here described cover a timespan of two months. The daily solutions show standard deviations ranging from about 1 mm to 1 cm, depending on the distance of the base station and its observation quality. Keywords GNSS, low-cost receivers, monitoring, gogps 1 Introduction Monitoring the deformations of critical infrastructures, such as large buildings, bridges, dams or high voltage electricity towers, is important to guarantee both the safety of the population and the functioning of national/regional services. In this regard, GNSS-based monitoring provides advantages over other techniques, since it is continuous in time, it has a global coverage, and it does not require the work of operators directly on the site to be monitored. Traditional GNSS monitoring is performed by means of geodetic dual-frequency receivers; however, such receivers have high costs that make the widespread adoption of GNSS monitoring impractical. Recently, low-cost GNSS receivers are being experimented by the geodetic community also for high-precision static positioning, with good results (Buchli et al., 2012; Benoit et al., 2014). 2 Experiment setup The GNSS monitoring equipment, composed of a low-cost patch GNSS antenna, three GNSS receivers and a splitter, was installed in the Media Center, Osaka City University, Japan. The Media Center is ten-story building, built in 1996, and it is one of the tallest building in the neighborhood (Figure 1, left), guaranteeing a good sky visibility for receiving GNSS satellite signals. The GNSS antenna, a 377

2 NV2410 by Technologies AG, was placed on its rooftop (Figure 1, center); the low-cost GNSS receivers, namely a, a and an NV08C-CSM, were connected to a 4-way splitter, which, in turn was connected to the antenna by a 30 meter long cable (Figure 1, right). The data logging was carried out by gogps Java programs on a server machine, storing binary data and creating RINEX files on-the-fly. Figure 1: Antenna position on the rooftop of the Osaka City University Media Center building (left); NV2410 GNSS antenna installed on the building rooftop (center); low-cost GNSS receivers and splitter (right). 3 Results The results reported in this paper refer to a period of 60 days, from 25 December 2014 to 22 February Code and phase observations by each of the three low-cost receivers was post-processed by the relative positioning algorithm of gogps, MATLAB version, which applies an extended Kalman filter (for this experiment, the dynamic model was set to "static") on doubledifferenced code and phase observations, estimating the antenna coordinates and float carrier phase ambiguities. The LAMBDA3 algorithm (Verhagen and Li, 2012) was applied epoch-by-epoch on the float ambuigity solutions. As base stations, a Virtual Reference Station (VRS), generated by the JENOBA positioning service, and three permanent stations belonging to GEONET (the nation-wide GNSS network of Japan) were used. The VRS was generated at a distance of about 60 m from the low-cost antenna. The three GEONET stations, with marker names 0336, 0337 and 0761 were located at about 9.5 km, 7.5 km and 11 km, respectively. In particular, station 0336 is located in a densely urbanized area of Osaka, and it is surrounded by tall buildings and nearby trees, degrading its observation quality. The other two GEONET stations have relatively similar surroundings, with low buildings and open sky visibility above the elevation cutoff used in this work (15 degrees). All the processing was done on GPS-only observations, although the receiver and all the base stations were in fact logging also GLONASS data; GLONASS-only and GPS+GLONASS solutions are going to be investigated in future tests. 378

3 3.1 Comparison between low-cost receivers Figure 2 shows the daily solutions obtained by processing the three low-cost receivers with respect to the VRS. The three solutions do not highlight significant differences in terms of repeatability, although the older u-blox model (the LEA-4T) shows a few more spurious oscillations. Figure 2: Daily solutions obtained by processing each low-cost receiver with respect to the VRS. 379

4 3.2 Comparison between base stations Figure 3 shows the daily solutions obtained by processing the u-blox LEA-6T with respect to the VRS and the three GEONET stations. The GEONET station 0336 confirms to provide the worst repeatability, due to the aforementioned bad quality of its observations. Figure 3: Daily solutions by processing the u-blox LEA-6T receiver with the four base stations. 4 Conclusions The results confirm that the observation quality of low-cost GNSS receivers, even with a low-cost patch antenna, are sufficient to perform precise daily positioning by the double-difference method with respect to a base station. The precision level, ranging from 1 mm to 1 cm in standard deviation in this experiment (Table 1) is influenced by the distance of the base station and its observation quality. The rightmost column of Table 1 reports a stability index expressed as SI = σ 2 (E)+ σ2 (N )+ σ 2 (U ). 380

5 Master station (approx. baseline length) VRS (60 m) 0337 (7.5 km) 0336 (9.5 km) 0761 (11 km) Low-cost receiver St.Dev. (EAST) St.Dev. (NORTH) St.Dev. (UP) SI Table 1: solution repeatability for all the combinations of base stations and lowcost receivers. Acknowledgments The authors would like to thank the JENOBA Co., Ltd service for their assistance with the VRS tests. This work was supported by JSPS KAKENHI Grant Number References Benoit, L., Briole, P., Martin, O., & Thom, C. (2014). Real-time deformation monitoring by a wireless network of low-cost GPS. Journal of Applied Geodesy, Buchli, B., Sutton, F., & Beutel, J. (2012). GPS-equipped wireless sensor network node for high-accuracy positioning applications. In Wireless Sensor Networks (pp ). Springer Berlin Heidelberg. Verhagen, S., Li, B. (2012). LAMBDA - Matlab implementation, version 3.0. Delft University of Technology and Curtin University. 381

6 382