Distributed Data Storage in Support for Context-Aware Applications
Résumé
Context-aware computing is a new paradigm that relies on large amounts of data collected from a variety of sources, ranging from smartphones to sensors, to automatically take smart decisions. This usually leads to large volumes of data, that need to be further processed to derive higher-level context information. Clouds have recently emerged as interesting candidates to support the storage and aggregation of such data for large-scale context-aware applications. However, specific extensions to support context-aware data need to be designed in order to be able to fully exploit the clouds' potential. In this paper we introduce such a cloud-based system, designed to support real-time processing and persistent storage of context data. Context Aware Framework is designed as an extension of the BlobSeer storage system, building a context-aware layer on top of it to enable scalable high-throughput under high-concurrency for big context data. Our experimental evaluation validates the transparency, mobility and real-time guarantees provided by our approach to context-aware applications.