Jun CHEN Differential GNSS positioning with low-cost receivers Duration of the Thesis: 6 months Completion: May 2013 Tutor: Prof. Dr. sc.-techn. Wolfgang Keller Dr. Maorong Ge (Potsdam-GFZ) Examiner: Prof. Dr. sc.-techn. Wolfgang Keller Background The Global Navigation Satellite System (GNSS) can provide geo-spatial positioning anytime and anywhere, it is widely used and GNSS receivers become the most popular location sensor nowadays. In most cases, geodetic GNSS receivers are used to do positioning and navigation. However, within the rapid development in GNSS receiver hardware technologies in recent years, low-cost GNSS receivers appear to meet people s needs. The motivation to work with low-cost receivers arises from the high cost of geodetic receivers. For saving costs, low-cost receivers become more and more widely used, especially in some small regions of interest such as monitoring for land sliding and so on. Objective: Our master thesis, at present, is aimed at studying and developing a data processing strategy or a software to do double differenced precise positioning with low-cost receivers. In this processing strategy, a geodetic receiver is set to be the reference receiver, and a low-cost receiver is set to be the roving receiver. We do not take real-time observations to do positioning, but we use the observations which have been observed for a period of time. This observation is transferred into RINEX format. This data processing strategy is used to decode the measurements from low-cost single frequency GPS receivers, collect instant RTK corrections from the geodetic dual frequency GPS receiver, and resolve integer ambiguities and output positioning solutions epoch by epoch. Methods: This data processing strategy mainly includes: 1
Read RINEX data Error model development Parameter estimation Quality control Ambiguity fixing *.obs (epoch by epoch) *.nav (broadcast ephemeris) Error budgets Double differential positioning Sequential least square estimation Static/Kinematic mode DIA-test principle Multi-outliers/cycle slips detection LAMBDA algorithm, Ratio Test Partial ambiguity fixing Estimated Result Precise coordinate Figure 1 flowchart of data processing Quality control. DIA-test principle is introduced to detect and adapt outlier/cycle slip in measurements. The improvement of multi-outliers/cycle slips detection is made in this part. Figure 2 flowchart of DIA-test principle Ambiguity fixing. To reach the best precision and reliable results, LAMBDA algorithm and ratio test are used for fixing ambiguity. An improvement is designed to shorten ambiguity fixing time in this part. "float" solution Estimate position and ambiguities Ambiguity fixing Estimate integer ambiguities "fixed" solution Estimated precise position Figure 3 flowchart of ambiguity resolution 2
Results The detail of test data is: Table 6.1 information of test data Data name Reference receiver Rover receiver Baseline Length Observation Time (Dual-frequency) (Single-frequency) Ublox short TRIMBLE NetR5 Ublox about 0.1m 24.08.2012 Ublox long TRIMBLE NetR5 Ublox about 3km 24.08.2012 The testing procedures are: - choose 2 hour length data. Test the ambiguity fixing efficiency and the fixed ENU (East, North and Up) precision; - choose 100 groups of data randomly. Each group has the same length of observation time. This procedure is aimed at testing the improved efficiency of partial ambiguity fixing, as well as the calculated ENU precision. In Static positioning mode, the test results are: Figure 4 ratio test and fixed ENU result of Ublox short (static, 2h) Figure 5 ratio test and fixed ENU result of Ublox long (static, 2h) 3
In kinematic positioning mode, the result can be shown as: Figure 6 ratio test and fixed ENU result Ublox short (Kinematic, 2h) Figure 7 ratio test and fixed ENU result of Ublox long (kinematic, 2h) Conclusion In this thesis, the following issues are addressed and solved: 1. Through setting different weight to the code and carrier phase observation, the large noise effect is reduced and precise result is available. 2. DIA-test principle is valid in outlier and cycle slips detection as well as reparation for low-cost GNSS receiver s data. Multi-outlier/cycle slips detection are applied successfully in the software, which can avoid error coupling in the observations. 3. Partial ambiguity fixing to shorten ambiguity fixing time is applied. From the test results, it is clear that partial ambiguity fixing method can improve the ambiguity fixing efficiency and shorten ambiguity fixing time significantly. 4. From the result of test data, it shows that the positioning precision in static mode is better than kinematic mode. In addition, the precision of result is much better when the baseline is shorter. 4
References Sebber, G. (2003): Satellite Geodesy. De Gruyter, Berlin, New York. Teunissen, P.J.G. (1994): A New Method for Fast Carrier Phase Ambiguity Estimation, Proceedings IEEE Position Location and Navigation Symposium PLAN94, Las Vegas, 11-15 April, 1994, pp. 562-573. 5