摘要：Continuously operating GPS networks have been used for many years in support of: (a) geodetic goals such as the determination of crustal motion on a variety of spatial scales, and (b) to provide pseudo-range corrections for Wide Area DGPS (WADGPS) implementations. Recently, regional-scale GPS permanent networks have been developed for multi-functional uses, including to support centimetre-accuracy, medium-range, carrier phase-based GPS positioning for surveying or precise navigation applications. In such an implementation the generation of carrier phase correction messages in a manner analogous to WADGPS requires that the integer ambiguities between GPS reference stations be fixed in real-time. Although the ambiguities could be resolved at the beginning of operation the challenge remains: how to recover an integer ambiguity if a cycle slip or data gap occurs, or if a new satellite rises? In this paper, the linear data combination algorithm that has been used is described, and the issue of “ambiguity recovery” techniques for data correction generation purposes is addressed. Three strategies are suggested: (1) using an “ambiguity recovery” technique if the data gap is shorter than a minute or so; (2) re-determining the integer ambiguities using an ionospheric correction derived from the tracking to other satellites when a new satellite rises, or after a long period of data loss, and; (3) re-determining the integer ambiguities with the aid of data corrections generated on previous days. Several 7-day continuous data sets were used for algorithm testing. The corrections can be generated from multiple reference stations in the post-processing mode and then used for comparison purposes with the simulated real-time processing mode using the proposed algorithm. Results confirm that the proposed algorithm can provide reliable carrier phase data corrections for centimetre-accuracy, real-time GPS positioning.
