A hybrid approach for resource-based comparison of adaptable Java applications

During the last decade, context-awareness and adaptation have been receiving significant attention in many research areas. For application developers, the heterogeneity of resource-constrained mobile terminals creates serious problems for the development of mobile applications able to run properly on a large number of different devices. Thus, resource awareness plays a crucial role when developing such applications. It identifies the capability of being aware of the resources offered by an execution environment, in order to decide whether that environment is suited to receive and execute the application. Within this line of research, we propose Chameleon, a framework that provides both an integrated development environment and a proper context-aware support to adaptable Java applications for limited devices. In this paper we present the novel hybrid (from static to dynamic) analysis approach that Chameleon uses for inspecting (adaptable) Java programs with respect to their resource consumption in a given execution environment. This analysis permits to quantitatively compare alternative versions of the same program. The analysis is based on a resource model for specifying resource provisions and consumptions, and a parametric transition system that performs the actual analysis.