GPS frequency transfer using carrier phase measurements

GPS time transfer is usually carried out with C/A code measurements using the common-view technique. An uncertainty of a few nanoseconds can be obtained for an integration time of about 1000 seconds when ionospheric measurements and precise orbits are used. This limits the capability of frequency comparison to the level of 3 parts in 10(14) for an integration time of one day. Recently new frequency standards have been developed that claim an accuracy at or below 1x10(-14). For existing time transfer techniques it is a challenge to compare their frequency over a sufficiently short interval. Previous studies have already shown the potential of carrier-phase and P-code measurements from geodetic GPS receivers. This paper draws up tentative uncertainty budgets for clock frequency comparisons using phase and code measurements from commercially available dual-frequency and multi-channel GPS receivers. Although the level of noise involved in carrier phase measurements is two or three orders of magnitude lower than that of code measurements in the short term, other error sources such as tropospheric delay of GPS signals, multipath propagation and the phase ambiguity problem lead to a limitation in the accuracy of frequency transfer at the 10(-15) level over an averaging time of one day. Carrier phase measurements, however, provide an improvement over the performance expected from code measurements for all averaging times smaller than one day.