Abstract: Navigation precise point positioning (PPP) technology has been widely used in many fields because of its flexible operation, and the fix of ambiguity is the core problem of PPP. As the BeiDou-3 navigation satellite system (BDS-3) stations are mostly concentrated in the asia-pacific region, therefore, the narrow lane Fractional Cycle Bias (FCB) of the new system signal broadcast by BDS-3 is affected by the observation environment and observation conditions, and the estimated values of each time period are quite different, therefore, the estimation accuracy of FCB in narrow roadways seriously restricts the whole-cycle ambiguity of BDS-3 precise point positioning. In order to solve this problem, a new FCB estimation method based on minimum variance estimation and prior information is proposed. Firstly, gross error detection and isolation are performed on the observed data of each station, and minimum variance estimation is made according to prior variance and mean on the basis of processing"Clean" data. Finally, the data from 38 observatories in asia-pacific region are used for verification. The experimental results show that the monthly stability of FCB in the new system is less than 0.2 weeks, and the fixed rate of FCB in the new system can reach 95.1% under static condition when BDS-3 precise point positioning is performed, the root mean square errors (root mean square rms) of the positioning accuracy in the east, north and sky directions were 1.1, 1.3 and 3.9 cm.

Key words: narrow lane, prior Information, BDS-3, precise point positioning, ambiguity resolution, fractional cycle bias, gross error detection, variance estimation