Every day, GNSS signals generate terabytes of data. This vast amount of GNSS data represents a unique scientific opportunity to expand the boundaries of human knowledge in domains like Earth Sciences, Fundamental Physics, Metrology or Space Science.

The GNSS Science Support Centre (GSSC), led by ESA’s Galileo Science Office at ESAC (Spain), aims to consolidate a world-wide GNSS Preservation and Exploitation platform through the provision of products and services to foster research in Europe, and notably the use of Galileo by the scientific community.

Announced at the Galileo Colloquium 2017 in Valencia, GSSC started operations hosting ESA’s IGS Global Data Centre, reaching full IGS GDC synchronization in January 2018. GSSC currently holds all IGS data and products available from other IGS Data Centers. The GSSC is also one of the original providers of GNSS data and products produced by ESA’s Navigation Office at ESOC (Germany).

Moreover, GSSC plays a key role in ESA efforts to provide a GNSS Science cyber-infrastructure for the myriad of ESA projects that have used or will use GNSS resources to perform science.

Building on this foundation, on-going GSSC development projects leverage on mainstream Big Data, Cloud, Virtualisation and Container technologies to introduce a paradigm shift characterised by the move of processing components close to the data. Hence, in 2019 GSSC will enhance and improve access to GNSS data and products in multiple ways.

GSSC will extend standard processing systems typically executed at Data Centres, like quality checks and product generation, providing capabilities for the scientific community to use, modify and deploy their own processing systems. This will allow scientists to carry out on-the-fly, interactive data analysis using Jupyter Notebooks technology, without the need to download the data to their own infrastructure.

GSSC will provide access to long term recording of GNSS Intermediate Frequency (Big) data using multiple GNSS Intermediate Frequency Recording Stations (GIFRES) to demonstrate the feasibility of a highly customisable, smart monitoring network.

Finally, GSSC focus on scientific community collaboration and open access to GNSS information will be driven by following activities:

+ Integration of additional GNSS information and processing resources.
+ GNSS advanced discovery services.
+ Improved monitoring of GNSS resources synchronization and utilization.
+ Dedicated areas for GNSS science teams to share and upload their own data and products.
+ Set-up of an experimental GSSC archive extension for GNSS Space Users Data.

This paper presents GSSC current progress and future plans for the provision of a GNSS computational environment that integrates innovative Data Mining, Machine Learning and Notebook techniques to enable novel research and collaboration opportunities.

The GNSS Science Support Centre (GSSC) is an initiative by ESA’s GNSS Navigation Science Office to foster the consolidation of a world-wide reference centre for the GNSS scientific community. The GSSC, hosted at the European Space Astronomy Centre near Madrid, integrates a wide range of GNSS assets including data, products and tools in a single environment promoting scientific collaboration and innovative research.

The GSSC Pilot Project is a first step towards the consolidation of the GNSS Science Exploitation and Preservation Platform. This project, carried out by GMV for ESA, aims at the early provision of GNSS assets in a single, state-of-the-art platform that unifies and builds on many different sources of GNSS-related information and tools, greatly expanding their visibility. In addition, this platform will provide added value services to the scientific communities, which include Earth Sciences, Space Science, Metrology or Fundamental Physics. The Pilot Project began in September 2018, building on an existing pre-operational site (https://gssc.esa.int) and is expected to be complete by the end of 2019.

The core of the GSSC Pilot Project is a large repository of GNSS data, including global providers such as the International GNSS Service (IGS), regional ground-based GNSS receiver networks, space-based GNSS receivers and other services such as laser ranging data from the International Laser Ranging Service (ILRS). This data shall be accessible via several interfaces, including FTP and HTTPS, while several public APIs shall allow tailoring applications to specific user needs. Several tools and utilities widely used by the GNSS community shall also be made available to execute either locally or in the cloud.

GSSC assets can be classified depending on their integration method into native assets (hosted by GSSC), federated assets (hosted externally but integrated in the GSSC processing systems) and external assets (referenced to external hosts). The aim is to host natively as many assets as possible, mitigating overload of existing services (for example, real-time NTRIP streams from other providers), while also providing lesser forms of integration when for various reasons it is not possible natively.

Not only GSSC users have at their disposal a large range of existing datasets, but there is also the possibility to upload assets to a private user area. This area allows developing and executing scientific tools in a cloud environment. It also allows the possibility to share these tools or data with the wider community enhancing collaboration opportunities.

This novel approach provided by the GSSC Pilot Project brings state-of-the-art technologies to GNSS science, while paving way for many future enhancements.

Abstract (alternative topic: Metrology M06 - Precise orbit determination):

The International GNSS service (IGS) has been providing precise reference products for the Global Navigation Satellite Systems (GNSS) GPS and (starting later) GLONASS since about 25 years. These orbit, clock correction, coordinate reference frame, troposphere, and ionosphere products are freely distributed and widely used by scientific, administrative, and commercial users from all over the world. The IGS facilities needed for data collection, product generation, product combination, as well as data and product dissemination, are well established. The Center for Orbit Determination in Europe (CODE) is one of the Analysis Centers (AC) of the IGS since the beginning. It generates the IGS products using the Bernese GNSS Software.

In the current decade new GNSS (European Galileo and Chinese BeiDou) and regional complementary systems to GPS (Japanese QZSS and Indian IRNSS) were deployed. The existing GNSS are constantly under modernization, offering, among others, more stable satellite clocks and new signals. The additional systems and signals open new opportunities, but do also pose new challenges. The exploitation of the new data and their integration into the existing IGS processing chains was the goal of the Multi-GNSS EXtension (MGEX) when it was initiated in 2012. CODE has been participating in the MGEX with its own orbit and clock solution from the beginning. Since 2014 CODE’s MGEX (COM) contribution considers five GNSS - namely GPS, GLONASS, Galileo, BeiDou2 (BDS2), and QZSS.

The validation of the COM products by satellite laser ranging (SLR), orbit misclosures, long-arc fits, and comparison with external products by the MGEX project itself, helped to identify many challenges and unsolved issues concerning the analysis of data from the new GNSS. In the recent years many of those issues have been successfully addressed, while others still need to be solved. We provide a summary of the recent developments of the COM solution with respect to processing strategy, orbit modelling, attitude modelling, antenna phase center modelling, handling of observation biases, and ambiguity resolution. The significant contribution of disclosed (Galileo and QZSS) satellite meta data to the aforementioned improvements shall be emphasized.

Finally, we plan to give an overview of the next development steps we plan for the COM solution in the near future. The possible transition of COM elements into legacy IGS products provided by CODE shall be discussed – with emphasis on Galileo.

In the current GNSS processing, in most cases the Continuously Operating Reference Stations (CORS) station recorded ‘raw’ GNSS information are referred as the baseband observables of pseudorange code and carrier phase. However, the digitized intermediate frequency (IF) data, after the receiver front end, is the first recordable and more fundamental measurement. It contains more signal related information (such as spectrum background) and also independent of the receiver processing strategy impact. Due to its gigantic data volume and the corresponding demands for high speed transmission interfaces as well as huge data storage, traditionally it is not considered to have the digital IF data archived continuously. Instead, after demodulation, correlation and integration, the base band observables in much lower data rate and smaller size are generated and stored. This results in unrecoverable loss of information in terms of data preservation. In addition, since the receiver processing algorithms are proprietary to each receiver manufacturer and are varying significantly, the measurements generated and stored from one specific GNSS receiver are not the “rawest” data and are influenced by the specific receiver design. At the moment, the common standard data stored by the receivers is the RINEX data and the re-processing algorithms are limited only at the PVT level. Although this has been serving the industry for quite a long time and proven to be effective and sufficient, more and more emerging applications in the scientific society demand preserving rawer data.

In the recent decades, benefited from the new Information and Communications technology emerging, cost dropping and capability increasing, rapid and innovative developments have been observed in the mega data storage, processing and transmission. These information technology (IT) evolutions, together with the advanced GNSS software defined receiver (SDR) technology, open up a new opportunity for systematic recording the GNSS IF data at affordable prices.

The systematic recording of digital IF would allow re-processing of the GNSS signal, using any signal processing technique (e.g. acquisition and tracking algorithms), including those yet to be developed. Comparing to the traditional baseband observables based GNSS information recording, the digital IF data recording will unlock the recovery of much richer information regarding the environment, background noise and the GNSS signal itself. In particular, it will provide valuable information for the ionosphere disturbance monitoring and the interference detection. The recorded IF data will be useful for GNSS monitoring, identification of vulnerabilities, and also innovative or future scientific applications. If continued permanent IF data are recorded at multiple stations systematically, a higher level of GNSS signal/environment details could be preserved, which could benefit the whole GNSS community for the future, enable the historical GNSS data analysis at the signal level, rather than the observable level.

In this paper, the objective and methodology of the proposed IF data recording CORS station will be introduced, the system structure design and implementation will be described, six identified user cases will be discussed, and the early test results after the first half of the test campaign will be shown and evaluated.

This presentation reports on the work by NSL, University of Nottingham and SixSq on the GNSS Big Data Project for ESA. The presentation focuses on the IT challenges and implementation.
Digitised intermediate frequency (IF) data is the first and most fundamental measurement available following antenna signal receipt. Due to its data rate, digital data cannot be stored consistently and is converted to lower density measurements such as pseudoranges, code- and carrier phase which generate much lower data rate. The algorithms to derive observables are however specific to each receiver and vendor. The conversion step from IF to observables therefore leads to an unrecoverable loss of information.
Therefore, to avoid this loss of information, the raw measurements have to be recorded and saved. Faced with a potential deluge of data (e.g. 21 TB daily per measurement location), we had to depart from standard IT architectures, since no reasonable network could cope with the data if it was to be processed in a cloud or a data centre.
Our approach to address this challenge was to build a platform with following key attributes.
• Constant collection and recording of GNSS IF high-fidelity raw data and observables at the edge (i.e. near the receiver)
• Transfer a subset of only valuable data to the cloud for permanent archiving, thus building a historical dataset of significant events
• Provide users with a unified access and ability to process these datasets of both historical data in the cloud and (near-)real-time data at the edge, to enable scientific and industrial discoveries
• Build a system that could scale to many stations, thus building potentially a Petascale GNSS data system.

From a technical aspect, the implementation includes cutting edge technologies. The following lists important components to this implementation:
• Scalable storage based on object storage technology, pioneered by Amazon Web Services, but made available on both edge and cloud.
• Meta-data of all the data objects persisted anywhere in the system is registered in a global meta-data catalogue, optimised for writes and search
• Ring buffers in both cloud and edge to ensure storage capacity and costs are controlled
• Data driven provisioning of user applications (i.e. near data processing)
• Management platform implemented using SixSq’s Nuvla software, including edge device monitoring, meta-data catalogue and user application orchestration.
• Edge device software built on top of SixSq’s NuvlaBox solution
• Edge hardware based on Hewlett Packard Enterprise Edgeline system
• Exoscale public cloud used via the GEANT network and commercial WAN networks.
In conclusion, SixSq developed with its partners an innovative edge-to-cloud architecture to solve the GNSS Big Data challenge. Looking forward, this solution is being industrialised, such that the resulting product can be offered to the market, at scale. Further, we believe this architecture is also applicable to a number of other scientific challenges, where data is being produced by a number of distributed sensors, being on Earth or in space.