Abstract:

Stable and reliable satellite wide-lane fractional cycle bias (FCB) products are the key to fix the wide-lane integer ambiguity correctly and then accomplish integer ambiguity resolution in undifferenced ionosphere-free Precise Point Positioning with the Integer-Recovery Clock method. This article discusses the approach of estimating undifferenced wide-lane FCB by sequential least squares proposed by Laurichesse et al, then points out that the daily wide-lane FCBs in satellites and receivers are assumed unchanged in advance, while lack of discussion and analysis of its time-varying property. Global GPS observations of ten days on 350 stations are processed and the time-varying property of wide-lane FCB in satellites and receivers is analyzed and discussed in detail from three aspects, undifferenced FCB between receiver and satellite and single-difference FCB between receivers. Results show that: (1)After smoothing, undifferenced wide-lane FCB is stable in a continuous arc without cycle slip, and can converge to 0.1 cycles by using more than 90 epochs(45 minutes) data. So we suggest the shortest arc to estimate wide-lane FCB should be longer than 45 minutes. (2)The rate of change disagrees over time in different days for wide-lane FCB in receivers, the maximum change in one day can reach up to 0.3 cycles. Besides, receiver reboot will make FCB reinitialize and destroy its continuity. Thus in this paper, the analysis of wide-lane FCB time-varying property in receivers indicates that the assumption of sequential least squares proposed by Laurichesse et al is not reliable. In order to obtain more precise undifferenced wide-lane FCB in satellites, the instability of daily wide-lane FCB in receivers needs to be considered. A new receiver FCB parameter should be added in the case of receiver reboot and the receiver FCB parameter can be estimated as a random walk process.

Key words: Wide-lane ambiguity, fractional cycle bias, Time-varying property, Precise Point Positioning, Ambiguity resolution