Constraint-driven dynamic adaptation of mobile applications for quality of service

Modern mobile applications are executed in a variety of execution environments by users with different preferences for energy savings, performance efficiency, reliability, and privacy. Offloading a mobile application's functionality to execute at a remote server has become an important energy and performance optimization technique. Mobile applications, however, executed over networks with divergent latency/bandwidth characteristics, access cloud-based servers that offer different levels of performance, availability, and privacy. An effective offloading mechanism must consider all these factors when determining which functionality should be offloaded to which server. In this paper, we present a novel approach to configurable, adaptive offloading for mobile applications that is driven by constraint solving. The programmer annotates energy intensive functionality at the method boundary. The end user, via a configuration menu, specifies how to prioritize energy savings, performance efficiency, server availability, and privacy. The specified priorities are then automatically translated into constraints used at runtime to drive an adaptive offloading runtime system. Applying our approach to third-party applications enhanced them with adaptive offloading capabilities, thereby optimizing their respective energy and performance efficiencies. These results indicate that our approach presents a promising direction in improving the quality of service of mobile applications.