Abstract
Mobile applications are becoming more popular as they provide the convenience of accessing services and information anywhere and at anytime. However, due to size and weight restrictions, mobile computing devices are limited in terms of battery energy and processing power. Running complex applications on resource-limited mobile computing devices is a real challenge. This paper proposes a cooperative paradigm for ad hoc computing in which a set of heterogeneous computing devices form a cooperative system on the fly, and whenever a resource-limited computing device in such a system has a resource-consuming application to be run, it uses resources of other devices to surmount the problem of resource limitation. To study the potential capability of this paradigm, we have designed energy-aware allocation and scheduling algorithms to make the most of all available resources in such a cooperative environment. These algorithms are static in the sense that they are based on discrete snapshots of the system topology. Although we consider these algorithms as preliminary work towards our objective of enabling cooperative computing, they have shown exciting results that encourage us to pursue our study. The main contributions of this paper are the novel allocation and scheduling algorithms which form a pioneering work towards enabling energy-aware cooperative processing in mobile computing environments.