Collaborative mobile-to-mobile computation offloading

Abstract

It is common practice for mobile devices to offload computationally heavy tasks off to a cloud, which has greater computational resources. In this paper, we consider an environment in which computational offloading is made among collaborative mobile devices.We call such an environment a mobile device cloud (MDC). We highlight the gain in computation time and energy consumption that can be achieved by offloading tasks with given characteristics to nearby devices inside a mobile device cloud. We adopt an experimental approach to measure power consumption in mobile to mobile opportunistic offloading using MDCs. Then, we adopt a data driven approach to evaluate and assess various offloading algorithms in MDCs. We believe that MDCs are not replacing the Cloud, however they present an offloading opportunity for a set of tasks with given characteristics or simply a solution when the cloud is unacceptable or costly. The promise of this approach shown by evaluating these algorithms using real datasets that include contact traces and social information of mobile devices in a conference setting.