Ad Hoc Networks. 7th International Conference, AdHocHets 2015, San Remo, Italy, September 1–2, 2015, Proceedings

Research Article

Cloud-Based Network Virtualization: An IoT Use Case

Download
431 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-25067-0_16,
        author={Giovanni Merlino and Dario Bruneo and Francesco Longo and Salvatore Distefano and Antonio Puliafito},
        title={Cloud-Based Network Virtualization: An IoT Use Case},
        proceedings={Ad Hoc Networks. 7th International Conference, AdHocHets 2015, San Remo, Italy, September 1--2, 2015, Proceedings},
        proceedings_a={ADHOCNETS},
        year={2015},
        month={9},
        keywords={IoT Cloud OpenStack network virtualization WebSocket},
        doi={10.1007/978-3-319-25067-0_16}
    }
    
  • Giovanni Merlino
    Dario Bruneo
    Francesco Longo
    Salvatore Distefano
    Antonio Puliafito
    Year: 2015
    Cloud-Based Network Virtualization: An IoT Use Case
    ADHOCNETS
    Springer
    DOI: 10.1007/978-3-319-25067-0_16
Giovanni Merlino,*, Dario Bruneo1,*, Francesco Longo1,*, Salvatore Distefano,*, Antonio Puliafito1,*
  • 1: Università di Messina
*Contact email: gmerlino@unime.it, dbruneo@unime.it, flongo@unime.it, salvatore.distefano@polimi.it, apuliafito@unime.it

Abstract

In light of an overarching scheme about extending the capabilities of Internet of things (IoT) with Cloud-enabled mechanisms, network virtualization is a key enabler of infrastructure-oriented IoT solutions. In particular, without network virtualization infrastructure cannot really be considered flexible enough to meet emerging requirements, and even administrative duties, such as management, maintenance and large-scale automation, would turn out to be brittle and addressed by special casing, leading to loss of generality and a variety of corner cases. We propose a Cloud-based network virtualization approach for IoT, based on the OpenStack IaaS framework, where its networking subsystem, Neutron, gets extended to accomodate virtual networks and arbitrary topologies among virtual machines and globally dispersed smart objects, whichever the setup and constraints of the underlying physical networks. This work outlines a motivating use case for our approach, and the ensuing discussion is provided to frame the benefits of the underlying design.