Abstract
Integer ambiguity validation is an indispensable and critical step in GNSS carrier phase positioning for precise and reliable positioning applications. The crucial problems associated with any ambiguity validation methods are as follows. 1) The fixed ambiguity vector can be separated from all other ambiguity candidates under certain tests (separability). 2) The probability of fixing to wrong ambiguity
combinations (mis-fixing) can be controlled to an acceptable level based on different application requirements. Traditional ambiguity validation methods,
such as the R-ratio and the difference tests which use one statistical test to control both separability and misfixing rate, are widely used due to easier computation. The performances of these methods are generally acceptable. However, experiments show that these tests with a fixed threshold can cause either a small percentage of mis-fixing or overly conservative with long observation time. In this paper, we propose a new Geometry Based Ambiguity Validation (GBAV) method which uses two statistical tests to control
geometry separability and mis-fixing probability separately. The thresholds for both tests can be strictly determined based on user requirements to control the
quality of ambiguity resolution. Three 24-hour GNSS (GPS, BDS) datasets (two short baselines and one middle-range baseline) are processed using the
proposed GBAV method, and compared with the popular R-ratio method. The results show that by giving proper control on the mis-fixing probability
(<0.01%), there is no mis-fixing case in all three datasets.
combinations (mis-fixing) can be controlled to an acceptable level based on different application requirements. Traditional ambiguity validation methods,
such as the R-ratio and the difference tests which use one statistical test to control both separability and misfixing rate, are widely used due to easier computation. The performances of these methods are generally acceptable. However, experiments show that these tests with a fixed threshold can cause either a small percentage of mis-fixing or overly conservative with long observation time. In this paper, we propose a new Geometry Based Ambiguity Validation (GBAV) method which uses two statistical tests to control
geometry separability and mis-fixing probability separately. The thresholds for both tests can be strictly determined based on user requirements to control the
quality of ambiguity resolution. Three 24-hour GNSS (GPS, BDS) datasets (two short baselines and one middle-range baseline) are processed using the
proposed GBAV method, and compared with the popular R-ratio method. The results show that by giving proper control on the mis-fixing probability
(<0.01%), there is no mis-fixing case in all three datasets.
Original language | English |
---|---|
Pages (from-to) | 28 |
Number of pages | 40 |
Journal | Journal of Global Positioning Systems |
Volume | 18 |
Issue number | 1 |
DOIs | |
Publication status | Published - Jan 2022 |