T. Pinto Jayawardena, C. N. Mitchell, R. Boyles, J. Bruno, K. Bolmgren, A.M. Ali, G. Buesnel, B. Forte and R. Watson, Spirent Communications, UK

View Abstract Sign in for premium content


Objective This paper discusses a unique approach to modelling ionospheric scintillation effects on GPS signals observed at high latitudes within a Radio Frequency Constellation Simulator that represent real-world events. It highlights the advantages of implementing the method in a test framework for RPNT system developers and users. The paper follows on from previous work carried out on ionospheric scintillation modelling for low latitudes; providing a global scintillation model capable of simulating realistic scintillation conditions experienced by GPS signals. Large scale ExB plasma motions in the ionosphere over auroral latitudes form fast moving electron density structures that result in scintillation of radio signals propagating through their gradients. In GNSS signals, these density structures cause (predominantly) phase scintillation that result in the receiver being unable to track the fluctuating carrier phase of the signals. Safety-critical and high-precision industries in high latitudes (e.g. polar route flights and off-shore positioning) share growing concerns about effects such as degraded accuracy, precision and availability of GNSS-based systems caused by ionospheric scintillation. This has been brought to further attention by the increase in shipping routes across the polar regions that require navigation on the arctic oceans. Testing such systems to evaluate their robustness against scintillation prior to product launch is thus imperative, where it is necessary to ensure the reliable operation of those systems under scintillation conditions. However, as evidenced from previous literature, conventional test paradigms for GNSS devices are not adequate given the nature of scintillation threats to GNSS. There is therefore a current need to provide capability to simulate realistic ionospheric scintillation environments within test frameworks aimed at improving the robustness of GNSS-based PNT systems. In contrast to statistical models, the proposed method uses historical GPS L1 data to generate empirical models for phase and amplitude scintillation across a grid of latitude and local time for a given location. Scintillation strengths experienced by each satellite in view are then derived from the model, which is used as a guidance to implement real-world scintillating RF profiles to desired satellites within the Spirent simulator-based Robust PNT (RPNT) test framework. The RF scintillation profiles used within the framework are extracted from historical GPS L1 50 Hz raw data. Data collected between 2012 and 2014 from Tromsø, Norway are analyzed to obtain a series of profiles corresponding to different strengths of amplitude and phase scintillation. Extraction of scintillation signatures from raw data is achieved using established methods described in past literature: Phase scintillation is extracted from the received carrier phase by de-trending the signal using a 6th order high-pass Butterworth filter; while amplitude perturbations are derived by normalizing the scintillating signal intensity to the mean received intensity. Scintillation profiles for GPS L2 and L5 signals are derived from L1 data through mathematical relationships established in literature. Anticipated/Actual Results The software was validated by comparing select events from captured live-sky data with simulations of scintillation conditions that would be most representative of those events. Results show that the scintillation strength profiles of true events are well represented by the simulations, for satellite elevations above 20 degrees; below which signals generally experience a multipath rich environment. Conclusions There is a tangible need to incorporate testing against ionospheric scintillation conditions within frameworks that evaluate GNSS-based robust PNT systems. Existing test methodologies are generally based on statistical models and/or characteristic simulations that tend to under-represent the severity of the impact of scintillation on GNSS systems. The proposed method aims to re-create realistic scintillation events with the use of a constellation simulator and demonstrates its performance with simulated scintillation conditions for GPS L1 signals. The work presented here concludes the efforts on a global scintillation model for GPS signals developed for use within an RPNT test framework. Significance of Your Work The proposed technique enables receivers to be tested at any location and time, with the added flexibilities of selecting different combinations of scintillation profiles (for a given set of satellites) and the selection of satellites for which scintillation is implemented. The ability to replay real, observed phenomena as proposed in this paper is expected to present the best opportunity for developers, integrators and users of GNSS systems to build up a picture of their systems’ resilience to ionospheric scintillation, and is likely to result in a much more effective risk assessment tool.