Session D3: GNSS Augmentation and Robustness for Autonomous Navigation (GNSS Integrity Augmentation)

Integrity of PPP-RTK with Regional Slant Ionospheric Grid Model and Error Bounds
Tiantian Tang, Yan Xiang, Sijie Lyu, Ling Pei, Wenxian Yu, Shanghai Jiao Tong University
Date/Time: Thursday, Sep. 19, 9:43 a.m.

The development of global navigation satellite systems (GNSS) has greatly contributed to the development and application of navigation technology. In particular, the positioning and navigation functions of GNSS have become indispensable tools in the aerospace, transportation, geographic survey and military fields. However, accuracy and reliability are not provided in the GNSS system. To fulfill this need, integrity monitoring of navigation systems has become essential. Integrity is a key measure of the reliability of a satellite navigation system and is the ability of the system to alert the user in a timely manner when an anomaly occurs. Blanch et al. proposed an improved RAIM (Advanced RAIM, ARAIM) method based on Receiver Autonomous Integrity Monitoring (RAIM) in 2014. The protection level is designed to meet the overall navigation and positioning failure rate requirement. It is calculated on the basis of the known the risk of constellation and satellite failure and other prior information about risk assessment. This provides an important framework and basis for the integrity study.
With the continuous development of positioning technology, different requirements for integrity have been put forward. From single point positioning to precise point positioning (PPP), real-time kinematics (RTK), and PPP-RTK, from the original fault risk assessment of constellations and satellites, it has been extended to the integrity monitoring of more errors in the positioning process and the performance evaluation of the service-end products. This study focuses on the integrity of the PPP-RTK. In recent years, the PPP-RTK technology has evolved from the use of ionosphere-free combined observations to the use of undifferential and uncombined observation data. The undifferential and uncombined technique can further accelerate the convergence speed of PPP by constraining the ionospheric and tropospheric parameters. Compared with the traditional ionosphere-free combined PPP model, which amplifies the observation noise by nearly three times, the undifferential uncombined PPP avoids the noise amplification caused by the combination of observations. However, the regional high-precision positioning based on the undifferential and uncombined PPP-RTK technology needs to rely on the construction of regional ionospheric error and tropospheric error correction models. When working on the integrity of PPP-RTK, it is necessary to pay attention to the influence of the unmodeled error and uncertainty of the product on the integrity calculation in high-precision positioning. In this study, we focus on the error model and uncertainty quantification of the ionosphere.
However, at equatorial/low latitudes, ionospheric activity is more frequent, and its characteristics are more complex. This is mainly due to the effects of solar radiation and the structure of the Earth's magnetic field, which lead to the instability and irregularity of the ionosphere. As a result, ionospheric error correction models cannot fully capable of performing well on all occasions as a means of eliminating the effect on positioning errors. The characteristics of the ionosphere at low latitudes include frequent activity, complex spatial and temporal variations, and the frequent occurrence of ionospheric scintillation. These characteristics also bring challenges to the integrity of the navigation system, and the traditional integrity methods lack the study of establishing a clear error model for the influence of ionospheric characteristics in the integrity calculation and also lack a clear statement on the method of quantifying the ionospheric uncertainty.
This study has the following innovations and contributions. First, this study analyzes the ionospheric characteristics and proposes an ionospheric error model and uncertainty quantification index for high-precision positioning integrity. The hourly ionospheric delay exhibits characteristics more in line with the Laplace distribution. At the same time, the characteristics of variation with time and region are obvious. The analysis focuses on the consideration of the ionospheric spatio-temporal correlations and their influence on the undifferential and uncombined PPP-RTK positioning errors. After analyzing and processing a large amount of data, which contains data at different times and latitudes, the ionospheric error model and uncertainty quantification index are proposed in this study.
Secondly, under the implementation of PPP-RTK service-end and user-end algorithms, on the basis of the SSR information, the service-end ionospheric products are evaluated. Ionospheric integrity parameters, including ionospheric failure probability (Psat), standard deviation, nominal envelope deviation (bnom) and other integrity-related information, are provided while synchronizing the output products. An explicit error model and uncertainty quantification of the server-end ionosphere product are completed. The above priori information is fed into the process of calculating protection levels on the user's end to provide integrity computational models that are more resilient to scenarios such as signal degradation (ionospheric scintillation) in active ionospheric regions.
Finally, in this study, 40 stations in China are modeled and processed for mid- and low-latitude regions, respectively, a large number of ionospheric data are analyzed, and our error model is validated. The experiments show that with our error model, the envelope effect of the protection level on the positioning error in the case of active ionosphere at low latitudes is improved by 15% with the undifferential and uncombined PPP-RTK technique, which significantly improves the accuracy of the calculation of the protection level. At the same time, in the calculation of the ionosphere-augmented PPP-RTK integrity, for the case of a calm ionosphere in the mid-latitude region, the accuracy of the protection level calculation is significantly improved in the both horizontal and vertical directions, the protection level can be actively lowered, which facilitates a more accurate binding of the positional error and effectively reduces the redundancy space.
Considering high integrity requirements, this study starts from the scenario of an active ionosphere at low latitudes, and analyzes the properties of the ionosphere and their impacts on the scale of positioning errors and convergence time. In terms of error modeling, this study uses actual data to analyze the evaluation of service-end products with different levels of ionospheric activity in the middle and low-latitude regions. In terms of integrity calculation, this study brings the service-end integrity parameters to the user end as a reference for protection level calculation. This study effectively solves the problems of over-envelope phenomenon in mid-latitude regions and the difficulty of traditional algorithms to envelope in the case of active ionosphere in low-latitude regions, providing a guarantee of high precision and high integrity for higher requirements.