10th EAI International Conference on Testbeds and Research Infrastructures for the Development of Networks & Communities

Research Article

FITeagle: A Semantic Testbed Management Framework

  • @INPROCEEDINGS{10.4108/icst.tridentcom.2015.259748,
        author={Alexander Willner and Daniel Nehls and Thomas Magedanz},
        title={FITeagle: A Semantic Testbed Management Framework},
        proceedings={10th EAI International Conference on Testbeds and Research Infrastructures for the Development of Networks \& Communities},
        publisher={ACM},
        proceedings_a={TRIDENTCOM},
        year={2015},
        month={8},
        keywords={semantic testbed infrastructure federation management},
        doi={10.4108/icst.tridentcom.2015.259748}
    }
    
  • Alexander Willner
    Daniel Nehls
    Thomas Magedanz
    Year: 2015
    FITeagle: A Semantic Testbed Management Framework
    TRIDENTCOM
    ICST
    DOI: 10.4108/icst.tridentcom.2015.259748
Alexander Willner1,*, Daniel Nehls1, Thomas Magedanz2
  • 1: Technische Universität Berlin
  • 2: Fraunhofer FOKUS
*Contact email: alexander.willner@tu-berlin.de

Abstract

Approaches to share resources across administrative domains have been developed for many years. A specific field of application is the experimental evaluation of Future Internet related research within federated facilities. Testbeds that want to offer support for this experiment life cycle have to expose multiple interfaces along with mechanisms to allow handovers between them. Existing work rest upon self-contained software components that are bound to specific protocols and schematic data models that aggravate such handovers. As a result, for a testbed the process of joining one or multiple federations is an expensive process. We propose an approach that reutilizes insights of the Semantic Web research to address parts of this issue. In this paper, we describe a framework which abstracts from specific protocols and functionalities by managing heterogeneous resources based on a semantic information model. As a result, different API calls relate to the semantically same resources which allows theses handovers and further makes this approach applicable to related fields such a federated cloud computing. We have validated our work within several research projects and further show that the implementation is standard compliant.