Adaptive Deployment Infrastructure for Android Applications

Abstract

Due to the popularization of mobile devices, the number of mobile applications has increased in recent years. Mobile devices are constantly evolving on computing capabilities, requiring the ability of mobile applications to deal with variations in services availability and computing resources. However, most of these applications are deployed using static configurations, making difficult to tolerate failures during their execution. In this context, we propose an adaptive deployment infrastructure to deal with this problem. Our solution proposes an autonomic loop that manages the current configuration model of a service application using dynamic feature model associated with the deployed architectural model. During runtime, according to the dynamic feature selection, the deployed architectural model can be modified in order to provide a new deployment solution. A mobile application was implemented using our solution, and after changes in services availability, it proved to be able to tolerate failures by dynamically changing its deployed configuration.