ENART - Exploration of New Approaches to Real-Time Archiving of Satellite Data
Scientific and Earth observation missions produce more and more data during their lifetime, supported by constantly increasing downlink capabilities. While this trend is highly desired from the mission perspective, it permanently creates a demand for even larger systems with even better performance in the ground segment.
The ground segment systems of space agencies, satellite manufacturers and satellite operators often cannot keep up with this steadily increasing amount of data and the required real- time performance for storage and retrieval (which again is influenced by a number of factors such as the choice of the hardware and database, the reading and writing mechanisms, the real-time capability of the operating system, back-up strategies and system distribution).
The experience from various space projects shows that the needs of satellite operations frequently exceed the capabilities of currently used tools and technologies. The main problems are: insufficient write performance (restricting the real-time capabilities in archiving of telemetry parameters), unsatisfying retrieval performance (restricting usability and availability of data), and huge disk space consumption (restricting operability and maintainability).
The goal of the ENART activity is to investigate and show in a prototype how fulfilment of the high-level requirements on data archiving, such as:
- (Near) Real-time write performance, without degradation of retrieval performance even if the archive is huge (several TByte or more)
- Powerful retrieval functions for analysis and evaluation of historical data
- Efficient support of low-level data types as for example raw octets or bit strings
- Application of data reduction / compression to optimise disk space
- Support for parallel writing and retrieving
- Non-functional requirements such as reliability, availability, safety and security as imposed by space data systems (such as mission control systems, carrier monitoring systems, check-out systems), can be improved by alternate approaches to structuring, indexing, storage and retrieval algorithms (e.g. optimisation of data structures and algorithms due to specific properties of the data, use of structural approaches such as Patricia Tries / Radix Trees).
Benchmarking against existing (COTS) solutions and an analysis of integratability into existing systems and interfaces complete the work in the ENART project.
Siemens Communication, Media and Technology, Space - Gerfried Krames
Siemens Communication, Media and Technology, Space
Tel.: +43 (517) 07 - 42628