A Threshold based Context Change Detection in Pervasive Environments: Application to a Smart Campus

Nesrine Khabou (1) , Ismael Bouassida Rodriguez (1, 2) , Ghada Gharbi (2) , Mohamed Jmaiel (1)

1 REDCAD - Unité de Recherche en développement et contrôle d'applications distribuées
2 LAAS-SARA - Équipe Services et Architectures pour Réseaux Avancés

Context awareness is an essential feature of pervasive applications and runtime detection of context changes is necessary for enabling context awareness. Context aware applications are applications that are able to adapt their behavior (self adaptivity property) according to the available resources, environmental conditions, user needs, etc. The self adaptivity property is enforced by four phases. Monitoring/Collecting context parameters, analyzing the collected context parameters to detect changes, planning appropriate adaptation actions and finally executing the planned actions to respond to the changes. In this paper, we focus and discuss the second phase (Analysis phase). Contrary to the most research activities in which the analysis phase is performed by comparing context parameters against fixed thresholds, we propose an analysis approach that uses a threshold comparison technique. Fixed as well as adaptive thresholds are applied to detect changes and trigger notifications that are used to deal with the changes.

Mots clés: context awareness; Ubiquitous computing; context analysis techniques; adaptation

Domaines: Génie logiciel [cs.SE]

Nesrine Khabou, Ismael Bouassida Rodriguez, Ghada Gharbi, Mohamed Jmaiel. A Threshold based Context Change Detection in Pervasive Environments: Application to a Smart Campus. 5th International Conference on Ambient Systems, Networks and Technologies (ANT-2014), Jun 2014, Hasselt, Belgium. pp.461 - 468, ⟨10.1016/j.procs.2014.05.448⟩. ⟨hal-01088780⟩