An improved method to accelerate the convergence of PPP-RTK with sparse CORS stations’ augmentation

Abstract

Real-time kinematic Precise Point Positioning (PPP-RTK) with dense regional reference stations’ augmentation has been proved to be effective for rapid precise positioning. However, with increasing distance between the rover and the reference receivers, the decorrelation of the atmospheric errors will make it more and more difficult to fix the ambiguities quickly. Recent studies show that PPP using the raw observables by estimating the ionospheric and tropospheric delays not only can improve the convergence of PPP but also can overcome the re-initialization of PPP, however the correlations between the atmospheric delays and the ambiguities will cause the wrong fixing of the ambiguities, which will introduce biases in the coordinates. In this contribution, in order to achieve fast precise positioning augmented with sparse continuously operating reference stations (CORS), PPP with raw observables are used as basic observations, the L1/L2 ambiguities are estimated firstly, then the wide-lane and narrow-lane ambiguities are formed, and the wide-lane and narrow-lane ambiguities are fixed recursively. In this new method, the narrow-lane ambiguities are free from ionospheric delays, so the correct fixing can be guaranteed, meanwhile the re-initialization can also be overcame. To evaluate the proposed strategy, four rover stations with average distance of 33 km within a sparse reference network, the average distance of which is about 200 km, are chosen to test the positioning performance. The simulated results show that the wide-lane ambiguities can be fixed immediately, and the narrow-lane ambiguities can be fixed quickly, usually at 3–30 epochs, and after ambiguity fixing, the positioning accuracy can achieve at cm-level.