Application Partitioning for Offloading in Mobile Pervasive Environments

Abstract In this work, we proposed an algorithm to partition an application for offloading into exactly two clusters. One for local execution on the resource constraint mobile device and other one for remote execution on another device such as smartphone, laptop, desktop and so on. The aim is to reduce energy usage on the source mobile device and improve response time. The evaluation results show an overall application performance improvement, in terms of response time and energy consumption, of the proposed algorithm compared to a no offloading and full offloading strategy. We note that high bandwidth and remote devices with higher computational speed benefits offloading while in low connectivity, a no offloading approach is better.