Resumen de Robust carrier tracking techniques for gnss receivers affected by ionospheric scintillation
Global Navigation Satellite Systems (GNSS) have become an indispensable tool in different areas in our modern society for positioning purposes using radio-frequency ranging signals. Some application examples are the positioning and navigation in ground, maritime and aviation environments, as well as their use in agriculture, surveying and precise timing and synchronization in communication systems and finances. The tracking stage is one of the core tasks within a GNSS receiver to keep aligned with the satellites and, to date, most receivers equip conventional tracking techniques with ease of implementation that suffice to operate in environments with favorable working conditions. However, in the recent years, the success of GNSS in open-sky environments has led to the emergence of applications that expand toward scenarios with harsher conditions, such as urban canyons and soft-indoor environments. The trend is to provide user mobile terminals such as smartphones with positioning capabilities in scenarios where receivers face new technological challenges owing to the abounding propagation impairments. In this sense, the so-called ionospheric scintillation is one of the issues degrading the performance of GNSS receivers, particularly in equatorial regions and at high latitudes. It introduces rapid carrier phase and signal power variations, and has a detrimental effect particularly onto the tracking stage.
The objective of this thesis is to design and develop new techniques for the robust tracking of GNSS signals affected by ionospheric scintillation disturbances. The presented approach is based on the use of Kalman filtering techniques, and the main contributions of the thesis are three. First, the analysis of ionospheric scintillation and the tracking of carrier dynamics despite the presence of the former. We design a Kalman filter with a hybrid formulation that allows the robust monitoring of both contributions separately. This arises from carrying out a detailed analysis of ionospheric scintillation which concludes that scintillation phase variations can be characterized through autoregressive processes, and thus be dealt with within the Kalman filter in a natural manner. Second, the design of adaptive Kalman filter-based techniques that allow self-adjusting their loop bandwidth to the actual scintillation conditions, which are rather time-varying in practice. This part includes a scintillation detector, a real-time estimator of the autoregressive model parameters, and an implementation to address the problem of non-linear signal amplitude attenuation introduced by scintillation itself. The goodness of the proposed techniques is later validated by carrying out an extensive simulation campaign using both synthetic data and real scintillation time series, and the outperformance region with respect to conventional tracking techniques is quantified. Third, a novel method for the derivation of expressions for the termed Bayesian Cram\'er-Rao bound (BCRB), which allow characterizing the behavior of Kalman filters in a closed-form manner, thus becoming a contribution to the literature of practical usefulness to design Kalman filters for any kind of application.
