In various GNSS applications with high requirements for precision, integer ambiguity resolution (IAR) is of great significance for taking full advantage of precise carrier-phase observations. Until now, there are two approaches to achieving IAR in network solutions, i.e., to resolve double-difference (DD) integer ambiguities or to resolve zero-difference (ZD) integer ambiguities. In this paper, we will present an approach to resolving between-satellite single-difference (BSSD) integer ambiguities in network solutions. BSSD ambiguity fixing can be divided into two main steps: Firstly, WL satellite FCBs are estimated to help to fix WL BSSD ambiguities and then narrow-lane (NL) BSSD ambiguity resolution is performed in a bootstrapping mode: datum BSSD ambiguities are selected and compulsorily fixed to the nearest integers and then a standard sequential fixing procedure is employed for the remaining independent BSSD ambiguities. Network solutions with GPS data from about 130 stations were conducted to validate the proposed approach. Experimental results show that the quality of satellite orbits, station coordinates and satellite clocks obtained with the new approach was almost the same as that with the DD approach. It is also shown that the new approach enjoyed slightly higher fixing ratio for both WL and NL ambiguities and was superior in computation efficiency, e.g., an improvement of 60% on average was achieved in this study. As demonstrated by experiments of precise point positioning (PPP) with 30-min data, satellite clocks achieved with the new approach have the ability to support IAR in PPP, just like those achieved with ZD IAR approach.
from Latest Results for Journal of Geodesy http://link.springer.com/10.1007/s00190-019-01251-z