Enhancement of Real-time Absolute GNSS Positioning Performance

Abstract

In recent years, the Global Navigation Satellite System (GNSS) community has experienced dramatic changes with the influence of global technological trends, such as digitalization, big data, artificial intelligence, unmanned aerial vehicles, autonomous cars, mobile and wearable technologies, etc. The requirement for instantaneous positioning solutions driven by mainly up-to-date technological trends has had considerable importance nowadays like never before. As a consequence, there has recently been growing attention to achieve higher positioning accuracy in real-time with more cost-effective GNSS solutions. In order to respond the increasing requirement, so many efforts have been made recently to enhance the existing positioning models. At this point, absolute positioning techniques have taken significant interest from the GNSS users for a long time since they eliminate the requirement of the simultaneous reference station and therefore provide cost-effectiveness and operational simplicity compared to the relative/differential GNSS positioning techniques. Furthermore, absolute positioning techniques are naturally compatible with the low-cost GNSS receivers, most of which are in mobile devices. Although the great majority of low-cost GNSS chipsets are able to provide only single-frequency observations, some chipset manufacturers have released new low-cost models which can record dual-frequency code and phase observations. To achieve higher positioning accuracy more cost-effectively, more complicated and enhanced approaches are required for real-time absolute GNSS positioning techniques. In this context, the main objective of this thesis is to provide enhanced positioning approaches for real-time absolute GNSS positioning techniques, taking single- and dual-frequency GNSS receivers into consideration. For this purpose, two fundamental positioning approaches were proposed in this thesis to be employed with three absolute GNSS positioning techniques, namely single-frequency code pseudorange positioning, single-frequency code-phase combination, and dual-frequency Precise Point Positioning (PPP) solutions. While the first approach was designed to work with real-time service (RTS) products of the International GNSS Service (IGS), ultra-rapid products are the fundamental orbit and clock source for the second approach. Both positioning approaches are compatible with the multi-GNSS solution that contains GPS, GLONASS, Galileo, and BDS satellites. On the other side, this thesis proposed a novel filtering method that integrates the robust Kalman filter and variance component estimation methods for real-time absolute GNSS positioning techniques. Besides, to perform the proposed positioning approaches and algorithms, a GNSS analysis software, which is named PPPH-RT, was developed as a part of this thesis. Several experimental tests were conducted to evaluate the performance of enhanced positioning approaches. All positioning processes were performed in kinematic mode as being compatible with real-time conditions. The results showed that the positioning performance of real-time absolute positioning techniques employing ultra-rapid products can be improved with the proposed positioning approach. Employment of WHU (Wuhan University) ultra-rapid products which have an update interval of 1 hour in the multi-GNSS solution provided better positioning performance for all real-time positioning techniques. The results also indicated that ultra-rapid products with the enhanced positioning approaches are an important alternative for real-time positioning solutions, especially considering that they can be employed without any additional connection. On the other hand, the results demonstrated that the enhanced positioning approach with IGS-RTS products provides better positioning performance for all real-time absolute GNSS positioning techniques when compared with ultra-rapid products. In addition, unlike ultra-rapid products, there was not any time-dependent deterioration in the positioning performance obtained from the IGS-RTS products. However, one drawback of the IGS-RTS products is the requirement of an external connection. Moreover, when the performance of the proposed filtering method was analyzed, it is understood that the proposed filtering method is able to handle stochastic properties of multi-GNSS observations better than conventional approaches and improves the positioning performance of real-time absolute GNSS positioning solutions considerably. Compared with the traditional filtering approach which contains a standard Kalman filter with the weighting scheme depending on higher variance ratios, the proposed filtering method improves the 3D positioning accuracy of real-time single-frequency code pseudorange positioning, single-frequency code-phase combination, and dual-frequency PPP solutions by 52.5%, 24.8%, and 43.9%, respectively.