In any Global Positioning System (GPS) application, we need to process code- and carrier-phases to obtain time-dependent positions. Based on a great amount of data measured by one ground sensor, our algorithm can remove nearly all of systematic ranging biases. The algorithm employs a least-squares technique to estimate measurement residuals in time series of immediately early days. The empirical mode decomposition of the Hilbert-Huang transform is used to analyze the residuals. Moreover, a grey relational analysis method leads to the correction for signal-in-space range biases. One of the experiments deals with the SPP0-SP3A baseline, data of which include: 18 m length, Leica SR530+AT504/AT502 sensors, 1-Hz sampling rate, 15 deg masking, local time 8:00-9:30 (hh:mm) in the morning on days of year 61????65 in 2008. When examining the daily adjusted residuals, high correlation can be caused by signal-in-space multipath effects. At the same time, the short baseline shows a height improvement of 53 percent on the conventional GPS solution. Logically, this raised accuracy will lead to higher sensitivity when used in detecting minuscule movement. The real-time kinematic positioning technique could be applied to finding out the trajectory of a moving vehicle.

Keywords: GPS phase residuals analysis, empirical mode decomposition, grey relation analysis