Research of Eliminating the Day-Boundary Discontinuities in GNSS Carrier Phase Time Transfer through Network Processing

Time and frequency transfer through global navigation satellite system (GNSS) precise point positioning (PPP) based on carrier-phase measurements has been widely used for clock comparisons in national timing laboratories. However, the time jumps up to one nanosecond at the day boundary epochs of adjacent daily batches lead to discontinuities in the time transfer results. Therefore, it is a major obstacle to achieve continuous carrier phase time transfer. The day-boundary discontinuities have been studied for many years, and they are believed to be caused by the long-term pseudorange noise during estimation of the clock offset in the daily batches and are nearly in accordance with a Gaussian curve. Several methods of eliminating the day-boundary discontinuity were proposed during the past fifteen years, such as shift and overlapping, longer batch processing, clock handover, and ambiguity stacking. Some errors and new noise limit the use of such methods in the long-term clock stability comparison. One of the effective methods is phase ambiguity fixing resolution in zero-differenced PPP, which is based on the precise products of wide-lane satellite bias (WSB) provided by the new international GNSS Service (IGS) Analysis Center of Centre National d’Etudes Spatiales (CNES) and Collecte Localisation Satellites (CLS). However, it is not suitable for new GNSS, such as the Beidou Satellite System (BDS), GALILEO, and QZSS. For overcoming the drawbacks above, Multi-GNSS Experiment (MGEX) observation data of 10 whole days from MJD 58624 to 58633have been network processed by batch least square resolution. These observations come from several ground receivers located in different national timing laboratories. Code and carrier phase ionosphere-free measurements of GPS and BDS satellites are used, and the time transfer results from network processing are compared with PPP results provided by Bureau International des Poids et Mesures (BIPM) and used for international atomic time (TAI) computation (TAIPPP) and universal time coordination (UTC). It is shown that the time offsets of three different time links are almost continuous and the day-boundary discontinuities are sharply eliminated by network processing, although a little extent of day-boundary discontinuities still exist in the results of UTC(USNO)-UTC(PTB). The accuracy of time transfer has been significantly improved, and the frequency stability of UTC(NTSC)-UTC(PTB) can be up to 6.8 × 10(−15) on average time of more than one day. Thus, it is suitable for continuous multi-GNSS time transfer, especially for long-term clock stability comparison.