Loading...
Research Project
Evolving FIRE into a 5G-Oriented Experimental Playground for Vertical industries
Funder
Authors
Publications
MIGRATE: mobile device virtualisation through state transfer
Publication . Santa, José; Ortiz, Jordi; Fernandez, Pedro J.; Luís, Miguel; Gomes, Christian; Oliveira, Jorge; Gomes, Diogo; Sanchez-Iborra, Ramon; Sargento, Susana; Skarmeta, Antonio F.
Delegation of processing tasks to the network has moved from cloud-based schemes to edge computing solutions where nearby servers process requests in a timely manner. Virtualisation technologies have recently given data cloud and network providers the required flexibility to offer such on-demand resources. However, the maintenance of close computing resources presents a challenge when the served devices are on the move. In this case, if processing continuity is desired, a transference of processing resources and task state should be committed to maintain the service to end devices. The solution here presented, MIGRATE, proposes the concept of virtual mobile devices (vMDs) implemented as Virtual Functions (VxF) and acting as virtual representatives of physical processing devices. vMDs are instantiated at the edge of the access network, following a Multi-Access Edge Computing (MEC) approach, and move across different virtualisation domains. MIGRATE provides seamless and efficient transference of these software entities to follow the real location of mobile devices and continue supporting their physical counterparts. Software Defined Networks and Management and Operation functions are exploited to "migrate" vMDs to new virtualisation domains by forwarding data flows to the former domain until the new one is prepared, while a distributed data base avoids the transference of data. The solution has been deployed in a reference vehicular scenario at the Institute of Telecommunications Aveiro premises within the 5GINFIRE European project. In particular, the system has been evaluated under different virtualisation domains to study the operation of the migration approach in a vehicular monitoring scenario. The results validate the system from the application viewpoint with a Web monitoring tool, and the migration of the digital twin provided as VxF is analysed attending to the modification of data flows, indicating a seamless transition between virtualisation domains in a timely manner.
Using aerial and vehicular NFV infrastructures to agilely create vertical services
Publication . Nogales, Borja; Silva, Miguel; Vidal, Ivan; Luís, Miguel; VALERA, FRANCISCO; Sargento, Susana; Azcorra, Arturo
5G communications have become an enabler for the creation of new and more complex networking scenarios, bringing together different vertical ecosystems. Such behavior has been fostered by the network function virtualization (NFV) concept, where the orchestration and virtualization capabilities allow the possibility of dynamically supplying network resources according to its needs. Nevertheless, the integration and performance of heterogeneous network environments, each one supported by a different provider, and with specific characteristics and requirements, in a single NFV framework is not straightforward. In this work we propose an NFV-based framework capable of supporting the flexible, cost-effective deployment of vertical services, through the integration of two distinguished mobile environments and their networks: small sized unmanned aerial vehicles (SUAVs), supporting a flying ad hoc network (FANET) and vehicles, promoting a vehicular ad hoc network (VANET). In this context, a use case involving the public safety vertical will be used as an illustrative example to showcase the potential of this framework. This work also includes the technical implementation details of the framework proposed, allowing to analyse and discuss the delays on the network services deployment process. The results show that the deployment times can be significantly reduced through a distributed VNF configuration function based on the publish-subscribe model.
Organizational Units
Description
Keywords
Contributors
Funders
Funding agency
European Commission
Funding programme
H2020
Funding Award Number
732497