Application Of Multipath Hemispherical Model For Short Baseline Multipath Effect Elimination

Global Positioning System (GPS) plays an increasingly important role in modern life with the development of spatial positioning technology and its accuracy is required to be improved continually. In Differential Global Positioning System (DGPS) technology, multipath error is the dominant measurement error, because multipath effect is significantly affected by environment and cannot be eliminated through differential method. For eliminating short baseline multipath effect, the study of this paper is divided into the following aspects:(1) Theoretical research of multipath effect. By analyzing the cause of multipath effect, discuss its amplitude and frequencies characteristics, repetition period. Analyzing the spatial correlation characteristics of multipath by experimentation and then introduce the temporal and spatial invariance of multipath effect.(2) Building Multipath Hemispherical Model (MHM). By using temporal and spatial invariance of multipath effect, we build MHM to eliminate the multipath effect in real time. This paper introduces the modeling principles, algorithms process of Multipath Hemispherical Model, and analyzes this model’s feasibility in dynamic application. This paper also introduces that this model can be used to eliminate dynamic short baseline multipath effect when antenna and reflection source are relative static.(3) Designing experiments to verify the feasibility and validity of this model. This paper designs static and dynamic experiments. Static experiment uses observation data in several consecutive days to build model and real-time corrects baseline multipath effect. Dynamic experiment, using onboard observation data in the Huangpu River, analyzes the modeling results of quasi-static and dynamic data. Through dynamic experiment, we verify the feasibility of using this model onboard and correcting dynamic data multipath effect by dynamic data modeling. Two sets of experimental data shows that this model can effectively eliminate low-frequency multipath effect and verify the model’s validity of eliminating multipath effect in static and dynamic short baseline cases.