UNCERTAINTY ESTIMATION ON GPS TIME TRANSFER

摘要

The traditional GPS common view technique, using C/A code receivers, is the main timerntransfer method used by various timing laboratories over the world. Moreover, this methodrnis used to realize the TAI (Temps Atomique International) and the TA (F) (TempsrnAtomique Français). Clock offsets between laboratory clocks are determined according torna fixed procedure defined by the CCTF (Comité Consultatif du Temps et des Fréquences).rnUsing this procedure, one can perform on average 54 tracks per day (theoretically 90 tracksrnper day), providing clock offsets. Each of these clock offsets results from one 780’s trackrnand is obtained as a result of a quadratic regression, followed by various model-basedrncorrections and finally a linear regression. The clock offsets are then issued with theirrnstandard deviations. However, this simplified estimate does not take into account thernstatistical properties of the different types of noise present in the measurement. We proposernhere to rigorously estimate this time uncertainty for various types of noise that characterizernthe transmitted GPS time offset data. This is achieved by the calculation of the covariancernmatrix of time samples. This method provides us with the variances of the drift coefficientsrnand of the residuals, in the case of a linear drift model for 1-day sample sets, taking intornaccount the different types of noise. In this paper, we compare the obtained results withrnsimulated and real data over several days.