A Method for PPP Ambiguity Resolution Based on Bayesian Posterior Probability

After the corrected by the uncalibrated hardware delay fractional part (FCB) product, the integer characteristic of precise point positioning (PPP) ambiguity can be recovered, which can significantly shorten the convergence time. However, the incorrectly fixed ambiguity seriously affects the accuracy of the results. In the convergence stage of PPP, the traditional method based on ratio test is prone to miss detection and fixed ambiguity incorrectly, which brings an extremely deviation to the results at the user end. Therefore, we propose a step-by-step PPP ambiguity resolution method based on Bayesian posterior probability. The method is deduced from the rigorous mathematical theory, which is guaranteed by the confidence level and can be adjusted according different scenarios. 116 global MGEX stations are selected as the server end to generate two kinds of FCB products, i.e., ionosphere-free combined FCB and undifferenced and uncombined FCB products, and another 50 stations are selected as the user end to test the proposed method. Compared with the traditional method, the success rate of the new method increases from 95.0% to 99.5% in static condition, and the missing detection rate detection reduces from 6.5% to 0.2%; the success rate of the new method increases from 82.9% to 95.9% in kinematic condition, and the missing detection rate detection reduces from 13.1% to 1.9%.