Optimal integration of single-baseline GPS solutions in network-based kinematic positioning

Although the demand for high accuracy kinematic positioning using multiple base stations has increased over the last several years, most commercial GPS processing software still only provides single-baseline solutions. Thus, methods were investigated to improve the accuracy of kinematic positioning by using network configuration on the basis of several single-baseline solutions. As discussed in this study, the positioning accuracy and network stability are improved by applying geodetic network adjustment theories to a kinematic positioning application. Four different methods to remove the rank-deficiency (RLESS, BLIMPBE, SCLESS, and PMINOLESS) are analyzed in this study. The 3D RMS (root mean squared) error has been improved from 4.4 cm (maximum) to 2.7 cm by using network-based kinematic positioning, and BLIMPBE is preferable over other methods in terms of the minimum bias in the rover position and the reproducing property of the reference stations.