A Multi-Dimensional Architectural Approach to Behavior-Intensive Adaptive Pervasive Applications

Abstract

Pervasive applications are featured by their transient interactions, explicit surrounding contexts and by the flow of activities. Towards transparently tackling these features yet rigorously modelling and validating such applications, this contribution capitalizes on SE advances: (1) explicit separation of concerns (e.g. interaction-centric functionalities and context-awareness); (2) flexible description of intrinsic rules governing any behavior-intensive task; and (3) concern-based transient architectural connectors to cope with adaptability. More precisely, we propose a multi-concern architectural approach that explicitly separate interaction concerns from context-aware ones. Each concern is first independently conceived through tailored ECA architectural connectors. These concerns are then accordingly integrated at the fine-grained activity level.