Determination of GPS-Galileo Time Offset to Support System Interoperability

Kurzfassung

The future European satellite navigation system Galileo will use an independent timescale - Galileo system time (GST) - as a reference for the determination of satellite orbit and clock parameters. The difference of the GPS and Galileo system timescales causes a slowly changing bias between GPS and Galileo pseudorange measurements. Thus, a combined GPS&Galileo user navigation solution will include an additional unknown which is undesirable in environments with limited satellite visibility. To overcome this problem, Galileo will broadcast the GPS-Galileo time offset (GGTO) in its navigation message. In the paper we discuss two potential options for the determination of the GPS-Galileo time offset: 1) with a GPS receiver installed at Galileo Precise Timing Facility (PTF) and connected to the physical realisation of GST; and 2) with a time transfer link between Galileo PTF and US Naval Observatory (USNO), which continuously determines the difference between its reference timescale and GPS Time. Since the capabilities of GGTO determination will depend on the accuracy of GST and GPS Time dissemination, we assessed this accuracy using Galileo simulations made with the DLR's GNSS simulation software and GPS observations from stations of the Permanent Tracking Network of EUREF. GPS data were processed 1) with broadcast satellite clocks and ephemeris, and 2) with IGS products, to check the accuracy of the broadcast parameters. Finally, selected algorithms for determination and prediction of GGTO were tested using GPS data collected by receivers at ORB, DLR and PTB time laboratories whose clocks emulated the Galileo system time. We compared the results obtained with the algorithms under test, which utilize only pseudorange measurements, with the results of processing of carrier-phase observations (the processing was made with “Bernese” software at ORB). These tests allowed us to assess a potential accuracy of the GGTO determination and prediction.