A context change detection approach for ubiquitous Environments

摘要

The increase of mobile devices leads to the growth of demands for pervasive applications. Due to the heterogeneity of devices in terms of resources, capabilities, the applications have to adapt themselves to the context changes. This requires the development of context aware applications able to perform four phases. First, collecting context; second, analysing context to detect changes; third, planning adaptation actions; finally executing the adaptation actions. We focus on the second phase. Its aim is to analyse context to detect changes. In this paper, we propose a context analysis approach that relies on different thresholds to detect changes and raise notifications when changes occur. The analysis approach requires three steps. Storing context; then classifying context; finally, calculating and updating thresholds to detect changes. The context classification depends on the user needs. For the threshold calculation, different mathematical models are applied in order to update thresholds.