Estimation and analysis of multi GNSS differential code biases using a hardware signal simulator
[Thesis]
Ammar, Muhammad
The University of Nottingham (United Kingdom)
2019
p.
Ph.D.
The University of Nottingham (United Kingdom)
2019
The ionosphere has the largest contribution to the Global Navigation Satellite System (GNSS) error budget. Its background effect can be mostly modelled, but sharp gradients in its Total Electron Content (TEC) adversely affect differential GNSS due to error decorrelation. Furthermore, irregularities in the ionosphere cause signal fluctuations known as scintillation, which may lead to cycle slips, accuracy degradation and even loss of receiver lock on satellite. In the last few decades, specialized GNSS Ionospheric Scintillation Monitor Receivers (ISMRs) have been developed with a view to support continuous ionospheric monitoring and modelling by estimating TEC and different scintillation parameters, and to help develop future receivers with robust tracking under extreme ionospheric conditions. However, it is not a straight forward task to derive accurate TEC information from these specialized receivers because the recorded pseudorange measurements are contaminated by instrumental biases, the so-called Differential Code Biases (DCBs). The Nottingham Geospatial Institute (NGI) - former Institute of Engineering Surveying and Space Geodesy (IESSG) - pioneered and currently undertakes scintillation and TEC monitoring in Northern Europe using dual frequency GPS receivers, the NovAtel/AJ Systems GSV4004, and in the last decade the relatively new multi-frequency multi-constellation Septentrio PolaRxS Pro receivers. Considering the hardware delays existing within these scintillation monitors to be stable for reasonable periods of time, the recorded TEC measurements have been used quite successfully on a relative basis in a number of experiments. Yet, to enable the calculation of absolute TEC, either for ionospheric monitoring or to facilitate non-differential positioning techniques such as Precise Point Positioning (PPP), these (and indeed any other conventional multi-frequency) receivers must be calibrated to account for their respective DCBs. The research work presented in this thesis has been carried out in two main phases. The first phase involves estimating the DCB of a multi frequency, multi constellation GNSS receiver (such as the Septentrio PolaRxS Pro) using a hardware signal simulator, whereby the state of the ionosphere and other variables can be controlled. It has been shown that a hardware signal simulator such as the Spirent GSS8000 can be used effectively to estimate a consistent and more realistic set of DCBs between different signal pairs for any multi frequency, multi constellation receiver. The second phase replicates the procedure carried out by the International GNSS Service (IGS) or the Multi GNSS EXperiment (MGEX) to determine receiver and satellite DCBs in a global ionospheric analysis using the IGS/MGEX network. By including the calibrated receiver from the first phase in this network, it has been proved that for all practical purposes of ionospheric modelling, using the 'known' receiver DCB of that receiver as an external constraint, is a valid approach to resolving the rank deficiency problem that arises in DCB estimation for receiver/satellite networks. One final aspect that this research aims to address was the possible benefit of estimated DCBs in PPP processing. From initial investigations, no plausible benefit was observed and hence it was decided not to proceed further because of time constraints. The indications are that the effect of the estimated DCBs cannot be observed in PPP while working with the Ionospheric Free (IF) combination, whereas in the case of PPP based on uncombined raw observations, the estimated DCBs derived from relatively noisy pseudoranges in comparison to the carrier phase observations, are not of sufficient accuracy to be used for correcting the ionospheric delay. This needs to be further investigated in future.