Multipath Error Fusion Modeling Methods for Multi-GNSS

The multipath error is considered to be the most limiting factor for high precision positioning applications. The sidereal filtering (SF) method can be used to mitigate the multipath error in the observation domain, and it has been successfully applied in the multipath mitigation in global positioning systems (GPS) and regional BeiDou navigation satellite systems (BDS2). However, there are few reports on the SF method in other systems. The performance of the SF method relies on the explicit orbit repeat periods of satellites in diverse systems or even different types of constellations. It is therefore inconvenient to utilize the SF method for multi-GNSS multipath error mitigation. Alternatively, a space domain multipath error reduction method, which establishes the multi-point hemispherical grid model (MHGM) using the residuals of the double-differenced carrier phase observations in the ambiguity-fixed period, has been modified. It is an integrated model for multi-GNSS, without considering the diversity of different systems and constellations. To compare the performance of MHGM and SF from a multi-GNSS point of view, the determination method of orbit repeat periods via the broadcast ephemerides is summarized, and the SF method is extended to the global BeiDou navigation satellite system (BDS3) and Galileo navigation satellite system. Further test results show that the performance of MHGM and SF are comparable from the perspective of root mean squares (RMS) and the power spectrum analysis of double-differenced residuals, as well as the static positioning results. This implies that the space domain MHGM can obtain similar correction effects as the SF method in the observation domain, but the former is more flexible for modeling with various systems’ data. In addition, the established MHGM using the data of multi orbit periods demonstrates a better performance compared with that of only one orbit period, and an average improvement of 13.1% in the RMS of the double-differenced residuals can be achieved.

Performance evaluation of low-cost GPS systems for static and kinematic applications

This thesis looks at improving positional accuracy of low-cost systems by investigating a method to isolate the multipath error based on wavelet analysis. Several sets of static and kinematic data were collected in different types of environment using a single-frequency GPS receiver. The code minus carrier combination of the GPS observables was exploited. After accounting for certain errors and resolving the ionospheric delay using ionospheric maps, the remaining terms were essentially multipath and noise. Wavelet analysis was then used to extract the multipath error. These approximations were utilized to identify and remove those satellites that were severely contaminated with multipath. Another approach investigated the subtraction of multipath approximations obtained by wavelet analysis from the corresponding code measurements. The positioning results of these two approaches were compared with those of the original data and assessed. For the static data sets, eliminating satellites contaminated with multipath proved to be most effective. For the kinematic sessions, neither of the two approaches displayed any improvement.

Performance evaluation of low-cost GPS systems for static and kinematic applications

This thesis looks at improving positional accuracy of low-cost systems by investigating a method to isolate the multipath error based on wavelet analysis. Several sets of static and kinematic data were collected in different types of environment using a single-frequency GPS receiver. The code minus carrier combination of the GPS observables was exploited. After accounting for certain errors and resolving the ionospheric delay using ionospheric maps, the remaining terms were essentially multipath and noise. Wavelet analysis was then used to extract the multipath error. These approximations were utilized to identify and remove those satellites that were severely contaminated with multipath. Another approach investigated the subtraction of multipath approximations obtained by wavelet analysis from the corresponding code measurements. The positioning results of these two approaches were compared with those of the original data and assessed. For the static data sets, eliminating satellites contaminated with multipath proved to be most effective. For the kinematic sessions, neither of the two approaches displayed any improvement.