Refinement of GNSS stochastic model combining elevation angle and SNR and its effect on RTK positioning performance

The accuracy of Global Satellite System Navigation (GNSS) observations will be affected by atmospheric delay, NLOS signal, multipath and other factors. However, the sensitivity of elevation angle or SNR model to different error sources is different. As a result, the traditional single random model based on elevation angle or signal noise ratio (SNR) cannot meet the high-precision positioning and navigation of the whole scene, coupled with the emergence of multi frequency and multi system, the accuracy of observations in different frequency bands of different GNSS systems and even the same system will also be different, which also brings a great challenge to the traditional model weighting. Based on the analysis of the advantages and disadvantages of elevation random model and SNR random model, a combined random model of elevation angle and SNR is proposed in this paper; The pseudo-range and phase noise of GPS, BDS, Galileo are extracted respectively through single difference between stations and cubic difference between epochs, and the combined random model of extracted and SNR is refined. Experiments show that the correct fixing rates of ambiguity of SNR model, elevation angle model and combined model are 92.42%, 95.85% and 97.69% respectively; Compared with the elevation angle model, the positioning accuracy of the combined model is improved by 50.0% in the horizontal direction and 37.1% in the elevation direction.