A novel Differential Global Navigation Satellite System (DGNSS) Carrier Phase (CP) batch processor is developed in this dissertation. The batch processor estimates both the differential position solution and the ambiguities of the differential CP measurements. The design of the batch processor uses geometry change information over time to minimize the effects of time-varying measurement errors, for instance, ionospheric and/or tropospheric spatial and temporal decorrelation errors. Tropospheric errors that affect the CP batch processor are studied in detail using measurements from two GPS data collection sites. Three novel techniques are implemented in the CP batch processor: 1) A geometry correction that is applied to the single difference (SD) CP measurements and not to the geometry matrix; 2) Error bounding based on both covariance and deterministic error analyses; and 3) RAIM-based integrity analyses over time. The combination of the three techniques ensures the accuracy and integrity of the final position solution and ambiguity estimations.
This dissertation provides batch processor results for simulations over a 10-km static GPS baseline as well as results for live DGPS measurements over a 5-km baseline. To make up for the reduced observability due to the use of CP measurements only and still obtain deci-meter level accuracy, the batch processor requires at least seven satellites for convergence times on the order of tens of minutes. Significantly faster convergence times, on the order of a few minutes, are obtained with multiple constellations, such as a combined GPS/Galileo constellation, which results in a practical carrier phase-only batch processor.
Integrity analysis is performed for the batch processor through the injection of different types of constant and slowly varying errors into the differential CP measurements. Results show that GPS alone provides a limited failure detection/exclusion capability. However, the combined GPS/Galileo constellation provides a very high failure detection/exclusion capability.