Um framework para aquisição adaptativa e fracamente acoplada de informação contextual para dispositivos móveis

Abstract

Mobile devices, such as smartphones and tablets, with a number of sensors have become commonplace in our daily lives. This scenario promotes applications from these platforms to increasingly access contextual information of the environment, the user and the system, which adapt accordingly or offer relevant services. This behavior is known as context-awareness. Several infrastructures have been created to help in the development of context-aware applications. These infrastructures facilitate the acquisition and management of contextual information. However, many of them are not appropriated to the heterogeneous and particular environment of mobile devices. This work presents an infrastructure for context acquisition called CAM (Context Acquisition Manager). CAM is a framework designed for use in sensor rich mobile devices. Among its main features are the loosely coupling with the applications that use it and the possibility of adapting in deployment time or execution time. The deployment adaptation allows a developer to customize what features will be included in the installation of the framework. The adaptation in execution time allow the framework to enable or disable its features according to applications requirements. To evaluate this work we developed a context-sensitive application as a proof of concept that uses the framework CAM. Through the development of this application, it was possible to notice a clear separation between the context acquisition code, wrapped by the framework, and the application code. A prototype of another application in which tests were performed on the dynamic adaptation mechanism of the framework was also developed. In this experiment, the impact of the adaptation on the resources of the device was investigated, which showed an increased economy in memory and CPU.