On the cloud deployment of a session abstraction for service/data aggregation

• Author(s): Domingos, João Nuno Silva Tabar
• Date: 2013
• Persistent ID: http://hdl.handle.net/10362/9923
• Origin: Repositório Institucional da UNL
• Subject(s): Big data; Cloud computing; Sessions; Dynamic reconfigurations; Mobile platforms
• Document Type: Master thesis
• Language: English
• Advisor(s): Gomes, Maria Cecília; Paulino, Hervé
• Contributor(s): RUN

Description

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

The global cyber-infrastructure comprehends a growing number of resources, spanning over several abstraction layers. These resources, which can include wireless sensor devices or mobile networks, share common requirements such as richer inter-connection capabilities and increasing data consumption demands. Additionally, the service model is now widely spread, supporting the development and execution of distributed applications. In this context, new challenges are emerging around the “big data” topic. These challenges include service access optimizations, such as data-access context sharing, more efficient data filtering/ aggregation mechanisms, and adaptable service access models that can respond to context changes. The service access characteristics can be aggregated to capture specific interaction models. Moreover, ubiquitous service access is a growing requirement, particularly regarding mobile clients such as tablets and smartphones. The Session concept aggregates the service access characteristics, creating specific interaction models, which can then be re-used in similar contexts. Existing Session abstraction implementations also allow dynamic reconfigurations of these interaction models, so that the model can adapt to context changes, based on service, client or underlying communication medium variables. Cloud computing on the other hand, provides ubiquitous access, along with large data persistence and processing services. This thesis proposes a Session abstraction implementation, deployed on a Cloud platform, in the form of a middleware. This middleware captures rich/dynamic interaction models between users with similar interests, and provides a generic mechanism for interacting with datasources based on multiple protocols. Such an abstraction contextualizes service/users interactions, can be reused by other users in similar contexts. This Session implementation also permits data persistence by saving all data in transit in a Cloud-based repository, The aforementioned middleware delivers richer datasource-access interaction models, dynamic reconfigurations, and allows the integration of heterogenous datasources. The solution also provides ubiquitous access, allowing client connections from standard Web browsers or Android based mobile devices.