Precision Analysis of the BDS GEO／IGSO／MEO Observables and Stochastic Model Refining

With 10-days observation data of CUT0 station in MGEX observation network, we calculate the single-difference residuals sequence using the method of the zerobaseline single-difference model, and then deducethe noise of GEO／IGSO／MEO observables. The refined elevation stochastic model is established by the leastsquare method through the statisticsof observation noise with the changing of elevation. The results show that:there is slight difference in different kind satellite of BDS about observation precision, and the precision from high to low isMEO, IGSO, GEO; the observation precision of phase is about 129 times higher than pseudorangein B1, and it’s about 118 times in B2; the accuracy of BDS observation is slightly better than GPS. Further, the baseline test results indicate that the refined stochastic model can improvethe accuracy in epoch-by-epoch kinematic positioning. The average point accuracyis increased by 42.1％,and RMS values are increased by15.3％, 31.6％, 31.4％ in E, N, U direction respectively.