COPELABS DSpace Repository

UMOBILE D4.5: Report on Data Collection and Inference Models

Show simple item record

dc.contributor.author Sofia, Rute C.
dc.contributor.author Santos, Igor
dc.contributor.author Soares, José
dc.contributor.author Diamantopoulos, Sotiris
dc.contributor.author Sarros, Christos-Alexandro
dc.contributor.author Vardalis, Dimitris
dc.contributor.author Tsaoussidis, Vassilis
dc.contributor.author d'Angelo, Angela
dc.date.accessioned 2017-10-06T10:48:42Z
dc.date.available 2017-10-06T10:48:42Z
dc.date.issued 2017-09-30
dc.identifier.uri http://hdl.handle.net/20.500.11933/733
dc.description.abstract Background: This report is written in the framework of the UMOBILE project WP4 “Services enablement”, task 4.2 “Data Collection and Contextual Inference”. The task dealt with data mining in the networking context. Specifically, the task has worked upon i) devising mechanisms and tools that assist in the collection of data from sensors and other sources; ii) working on contextual awareness aspects relevant to networking in general, and to UMOBILE in particular; iii) the dissemination of the processed data to other UMOBILE services and modules, to assist in a future optimization of the network operation, based on contextualization. A relevant aspect worked in this task is the notion of usage contextualization and service personalization. A second relevant aspect concerns the capability to infer roaming behaviour in a way that keeps anonymity and privacy of the user as well as of the device. A third relevant aspect of the task was to understand how to assist UMOBILE, namely, which modules could benefit of such contextualization operation, and what/how to integrate this operation in the UMOBILE architecture. A fourth aspect worked upon in this task concerned deriving relevant guidelines for the community, based on data collection and experimentation. For the purpose of contextualization, the task started with an analysis of UMOBILE requirements and how to best fit such requirements when considering data capture on the network. Operationally, the task started with a model derived from the Senception’s product PerSense TM. Such model has been incorporated into the PerSense Mobile Light (PML) middleware, a first tool to assist in data capture and understanding sensing limitations. Based on such product, and derived from an analysis of the UMOBILE use-cases, as well as derived from an analysis of project requirements, the task then proceeded to propose the UMOBILE Contextual Manager module, which is specified in this deliverable and which shall be available until the end of the project as open-source software, technology readiness level 6, in the context of the UMOBILE proof- of-concept being defined in WP5 (month 36). This deliverable has the following goals: i) to introduce related literature concerning contextualization aspects in networking; ii) to describe the relevancy of such contextualization in networking in general as well as in UMOBILE; iii) to provide the specification of the contextual manager as our proposal for a flexible networking framework for contextual awareness in NDN/ICN; iv) to describe experiments carried out, detailing results obtained and the relevancy in the context of networking. en_US
dc.description.sponsorship European Commission H2020 en_US
dc.publisher UMOBILE Consortium en_US
dc.subject network mining, contextual manager, wireless en_US
dc.title UMOBILE D4.5: Report on Data Collection and Inference Models en_US
dc.type Technical Report en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account