engleski

Statistical learning TEC predictive model for GNSS ionospheric delay mitigation in self-adaptive environment-aware SDR GNSS position estimation algorithm

Complexity of space weather, geomagnetic, and ionospheric conditions render them the most prominent source of the Global Satellite Navigation System (GNSS) positioning performance disruptions and degradation. Mitigation of space weather, geomagnetic, and ionospheric effects on GNSS positioning performance involve utilisation of generalised/global GNSS ionospheric correction models, and costly bespoke augmentation infrastructure (systems for additional signal and information provision). GNSS operators have been required to ensure the quality of GNSS Positioning, Navigation, and Timing (PNT) service, despite the fact that the positioning environment that causes the GNSS positioning performance degradation is completely out of the operator’s control. Recent scientific and technology advancements allows for redrawing the concept of GNSS position estimation algorithm and its deployment. The Software-Defined Radio (SDR) concept in GNSS receiver design renders the GNSS position estimation algorithm transparent and open for modifications and advancements, including development and deployment of bespoke GNSS position estimation algorithms tailored for specific GNSS applications. GNSS receivers are embedded in the mass-market mobile computational and communication platforms, such as modern smartphones, equipped with sensors, which may be used for observations of positioning environment conditions. The SDR concept combines with statistical learning methods in development and deployment of tailored GNSS ionospheric error correction model based on the situation awareness of the immediate positioning environment (i. e. space weather, geomagnetic, and ionospheric) conditions. Here a novel concept for GNSS SDR position estimation algorithm is proposed, based on raw GNSS pseudorange observations corrected for Total Electron Content (TEC)- derived GNSS ionospheric effects through utilisation of a self-adaptive positioning environment-aware statistical learning correction model. The positioning environment awareness is accomplished either with the direct embedded sensor-based observations, or through mobile internet access to trusted third-party data. The TEC/GNSS ionospheric correction model adapts itself to the identified class of positioning environment conditions status, and produces correction based on its previous experience and the situation awareness of current positioning environment conditions. We demonstrated success of the proposed self-adaptive environment-aware TEC/GNSS ionospheric effect mitigation approach in the proof-of-principle applied in several scenarios of disturbed ionospheric conditions at different latitudes. Developed in the open-source R environment for statistical computing, the proof-of-principle self-adaptive environment aware TEC GNSS SDR ionospheric correction model is a development and a testing framework that will allow for further enhancement and improvement of the proof-of- principle. Its ultimate goal is to become a personalised GNSS SDR ionospheric effect correction model that generates correction based on the previous experience of the process and the actual positioning environment conditions, as a more affordable alternative to existing not flexible models and the expensive infrastructure in augmentation of the GNSS position estimation process.

GNSS ionospheric effects ; statistical learning ; positioning environment awareness

nije evidentirano