Methods for enhancing carrier phase GNSS positioning and attitude determination performance.

Abstract

This thesis explores the methods for enhancing the performance of using low cost, single frequency Carrier phase Differential Global Navigation Satellite System (CDGNSS) in real-time, safety or liability critical applications. This is done by improving the integer ambiguity resolution performance and carrier phase error modeling. CDGNSS is considered for a broad range of real-time applications which require both a high precision relative positioning and attitude determination system. This is because of the drift-free nature of the GNSS measurement errors and the precise nature of the carrier phase measurement. The key to making full use of the precise nature of the carrier phase measurement is to fix the integer ambiguity quickly and reliably. This poses the biggest challenge for a low cost single frequency system. For the attitude determination problem, the precisely known baseline lengths can be used to improve the integer ambiguity resolution performance. Traditionally, the relative positioning problem was solved independently of the attitude determination problem and, thus could not leverage the precisely known baseline lengths of the attitude determination system. However, by integrating the two systems together, the precisely known baseline lengths can be used to improve the relative positioning system as well. The first part of the thesis develops an integration framework to improve the integer ambiguity resolution performance for the relative positioning system and the attitude determination system simultaneously. The second part of the thesis provides a GNSS antenna Phase Center Variation (PCV) error model development to improve the accuracy of the integrated system. It also examines the feasibility analysis of using the developed error model for a real-time dynamic application. The challenging of using this in the real time lies in the fact that PCV error magnitude is small (less than 2cm) and the developed error model is a function of unknown parameter such as attitude. A feasibility analysis of the developed model with a set of specific antennas is performed and assessed.