SatcomWeather | ESA's ARTES Programmes

- Competitiveness & Growth
- Core Competitiveness
Status: Completed
Last updated: 2016-08-31 00:00:00
Activity Code:

Objectives

Very small aperture terminals (VSATs) access satellites in geosynchronous orbits. The link quality is very dependent on the tropospheric situation along the fixed slant path between the terminal and the satellite. However, there are a variety of other parameters too such as the orbital motion of the satellite, the pointing accuracy of the terminal, and thermal drifts of the transmitter/receiver that degrade the link quality. Satellite- and Satellite-Network-Operators are seeking for ways to optimize their service quality and their network operation, since their business is challenged by competition from terrestrial services and by the changing trend in TV consumption as a result of streaming solutions. The SatcomWeather project was initiated to support VSAT operators who have a commercial interest in optimizing VSAT networks and data throughput over time. Hundreds of millions of measured VSAT signals and throughput records will be investigated, with the objective being to classify stations with optimization potential, optimize thresholds for installations, and derive weather data from the VSAT-measurements. Within the project, algorithms to harmonize, access and analyse the data will be developed and a software information tool will be prototyped that guaranties high performance access to the datasets and adequate graphical visualisation.

Challenges

Development of a prototype application to determine how different exchangeable models for multi-dimensional analyses are integrated into a SW-framework to allow high performance analyses of 'big-data'.
Development of a design of the optimal solution for storing the data.
Design of an optimal and high performance mechanism to display results and maps.
Validate the existing models to calculate and optimise the influence of weather data variables on attenuation.
Develop an attenuation prediction model of radio signals, possibly similar to a weather service.

Benefits

Statistic Station Classification (Classification of network problems and filter on alarms and events by analysing the long term behaviour of links) by enabling SatOps to apply physical models on 'big data' implemented in a software to verify (or classify) VSAT measurements.
Near Real Time Station Classification (to classify network problems and filter out problems in near real time)
Station Attenuation Forecast (to optimise ACM for Ka-band and higher frequencies).

System Architecture

The planned architecture of the SW is along a high modular system approach. The application operation and data production is triggered via an application programming interface.

The “new SW” produced in this project is a number “SW groups” (consisting of SW modules in Python or Java) and a number of config-files grouped according the product tree.

Plan

The first phase (3 months) allows presentation of work at the CDR (Critical Design Review) that includes the following: consolidated users’ needs functional and system requirements, analysis of existing data, design the implementation and implement a set of Proof of Concept functionalities.

The second phase (9 months) is the system implementation phase that is be validated during the TRR (Test Readiness Review), where the first prototypes for use cases are implemented and verified and the first data results are obtained and interpreted.

The last phase (2 months) focuses on the test (internal validation), verification (external validation) of the system and the follow-up project specification validated at the FR (Final Review).

Current status

The activity is completed. Outcome and conclusions are:

The infrastructure handles data from 12000 terminals for a 2 years period with a time resolution of 10 minutes, achieved by clustering and parallelization.
Storage system runs in a private cloud provisioned by user security features.
Software architecture allows flexibility in the processing steps and modules. Process of data ingestion, analyses, storage and multi-dimensional access to data works successfully via programming interface.
Graphical big data processing has been achieved through 2-D maps visualizations, waterfall diagrams, histograms, line plots and scatter plots for signal over time, including animated time-series.
A high understanding of the VSAT system has been achieved, being able to distinguish various influences on the signal and correct them. Nevertheless, a reliable methodology for the calculation of the footprint move and carrier level fluctuation needs to be further developed.
The accuracy of the footprint move data is to be further validated in a follow on project.
The conversion from weather data to attenuation data was successfully prototyped and verified in the test stations in Milan. Nevertheless for usage by a Satellite Operator improved quality results are required for the prediction of radio signals attenuation.
The correlation of weather data and attenuation data was successful implemented, but without sufficiently high correlation results as various systemic influences were not sufficiently filtered out.
VSAT installation quality metric map was successfully implemented, but in-field validation was not made possible during the project.
A prototype Graphical User Interface was implemented. Additional features are to be further developed according to suggestions from Satellite Operator.