Abstract: With the continuous improvement of smartphone chip and antenna performance, high-precision positioning based on mobile phones has gradually attracted widespread attention from academia and industry. Compared with single-point positioning, real-time kinematic (RTK) positioning usually shows higher positioning performance. However, there are still relatively few studies on the stochastic model of mobile phone RTK positioning. Therefore, this study takes Xiaomi 8 as an example to focus on the stochastic model of smartphones in a multi-system RTK positioning scenario. The research results show that different systems There are significant differences in satellite pseudorange noise, while the difference in phase noise is smaller. Based on this, this paper proposes a stochastic model of base station signal-to-noise ratio combined with comprehensive weighting between satellite systems, which is more efficient than the traditional signal-to-noise ratio model. The weight distribution between different systems was accurately considered. Static and kinematic experiments were conducted under open and occluded conditions. The results show that compared with the traditional signal-to-noise ratio model, the new model is RTK positioning accuracy in the three-dimensional direction under static openness, static occlusion, kinematic openness and kinematic occlusion has increased by 18.7%, 18.3%, 3.3% and 4.6% respectively.