Kalman filter based robust GNSS signal tracking algorithm in presence of ionospheric scintillations
Résumé
Ionospheric scintillations are created by diffraction when the transmitted propagating waves encounter a medium made of irregular structures with variable refraction indexes. The recombination of the waves after propagation can be constructive or destructive and the resulting signal at output of the receiver antenna may present rapid variations of phase and amplitude. Scintillations are essentially observed in the regions located under the geomagnetic equator, where irregularities in the ionosphere like plasma bubbles occur more frequently. They also happen over Polar Regions and are associated with the penetration of charged particles along the magnetic field lines. Under the auroral oval region, scintillations can be local or spread over a great part of the oval. According to past observations scintillations occur mainly at the period of the equinox and the solstice. During the rest of the time there are almost no observations. After sunset, the modification in the ionosphere layers generates such variations. Scintillations cause very brutal and fast fades of the received signal, and once these phenomena occur they can last from 30 minutes up to several hours. However, continuous GNSS Carrier Phase Measurements are important observations needed for data demodulation and are increasingly used in GNSS receivers, both at user side (e.g. for precise positioning), and for ground segments to compute the navigation and integrity data. The carrier phase is traditionally tracked in the GNSS receivers using PLL, potentially aided by FLL. Carrier tracking loops may be optimized, depending on the application and environment, by selecting the appropriated loop order, integration time, and bandwidth (trade-off between accuracy and robustness). Phase Loops are however known to be less robust than code tracking loops, and the GNSS receivers may thus suffer from phase tracking loss, for example when tracking low C/N0 signals (attenuations), or fast varying signals such as signals affected by scintillations. This strongly impacts the positioning service availability, as well as the capability to demodulate the navigation message data, in situations where ionospheric scintillations affect the received signal. One thus has to implement innovative techniques and receiver architectures to provide robust carrier phase tracking, either by improving or optimizing the classical tracking loops (optimized parameters of the PLL, potentially dynamically) or by defining different architectures such as Kalman filter-based tracking loops (either scalar or vector architectures, taking benefits of the different tracking channels, constellations and frequencies which may be differently affected by the disturbance). Other techniques coming, for example, from the telecommunication domain can also be interesting to estimate the propagation channel parameters. It can also be taken advantage of the improved structure of the modern GNSS signals, providing in particular a pilot signal component. The improvement technique investigated in this paper consists in replacing the conventional phase lock loop filter by a Kalman Filter PLL (KFP), as inspired by the technique proposed by Psiaki et al.. Kalman loop filters provide the optimal filter gain when the statistical levels of uncertainty of the state and observation vectors are well known. So the Kalman filter is continuously adapting the filter bandwidth to the noise level. KFP tracking loop indeed show a better resistance to weak GNSS signal tracking compared to classical loop filters. We were therefore interested in analyzing the potential of such KFP variants for tracking the GPS L1 carrier phase in presence of scintillations, adapting in particular the feedback control signal. This was conducted during a project financed by CNES. The aim of this paper is therefore to present the development of a GPS L1 phase tracking technique based on a Kalman Filter improving the tracking robustness in presence of ionospheric scintillations, and to present results of its performance using simulations. The paper starts with a description of the phenomenon of ionospheric scintillation, including the possible models for signals affected by ionospheric scintillations, focusing on the model selected for this study which is GISM (Global Ionospheric Scintillation Model) developed by IEEA. The second section presents a review of the state of the art of ionospheric mitigation tracking techniques. In a third part, the proposed robust GPS L1 tracking technique proposed is described. Then the simulation environment is described, and simulations results are presented, showing improved performance of the proposed tracking technique.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...