Improving offloading in mobile cloud computing

Abstract

Increasing growth of applications and operating systems of mobile devices has created limitations such as processing, storage capacity, and energy on these devices which may interrupt future progresses in mobile devices. Mobile cloud computing (MCC) can resolve these problems by running mobile device applications on sources outside the mobile device. Offloading has been proposed as one of the techniques to deal with the limitations of mobile devices in the field of MCC. In this technique, the processes move to a remote execution environment with higher computing capacity to increase efficiency and reduce energy consumption in mobile devices.In this paper, an approach has been presented to improve offloading in MCC. Our proposed approach is an improved MpOS (Multiplatform Offloading System) framework that uses the LRU cache to store methods offloaded to the remote execution environment. According to the evaluation and analysis, our proposed approach has become better than the MpOS framework in all remote execution environments but this difference can be more and better observed in Desktop cloudlet, Laptop cloudlet, EC2 large (4G) and EC2 medium (4G) environments.