Performance analysis of an alternative technique for relative positioning by GPS

The performance of a method for post-mission relative positioning by GPS is evaluated. The method can be implemented as an alternative to the traditional double-differencing technique using a single baseline. Description, methodology and results of tests conducted on land, marine and airborne kinematic modes are presented. In this method, precise orbits and clock corrections are first utilized at the reference station to estimate the measurement atmospheric corrections. These corrections are extrapolated at the user location, and are next combined with the precise orbits and clock corrections to estimate the rover position using the precise point positioning (PPP) technique. The unknown vector in this case includes the unknown coordinates, a receiver clock offset, and the phase ambiguities. A residual ambiguity term for each satellite is added to the parameters to absorb the differential initial phase and inter-frequency biases. The discussed approach has significant operational advantages over the traditional single baseline method including: relatively longer distances can be considered, and positioning is not restricted to synchronization of the rover and the reference station measurements since only measurement corrections from the latter are used. The method also permits the use of more than one reference station in computing the atmospheric corrections, for instance when using archived raw data from RTK reference networks. In this case, the satellite initial phase bias can be estimated at the reference station and used at the rover. During testing, variable reference-to-user ranges were utilized. Positioning by means of the presented method and the traditional single baseline was compared and analyzed. In general, their discrepancies ranged from a few centimeters to a decimeter. The phase measurement ambiguities needed a longer period to be resolved in the PPP mode, ranging between 150 and 400 seconds.