Context-Awareness in Opportunistic Mobile Cloud Computing

Abstract

Smartphones have shaped the mobile computing community. Unfortunately, their power consumption overreaches the limits of current battery technology. Most solutions for energy efficiency turn towards offloading code from the mobile device into the cloud. Although mobile cloud computing inherits all the Cloud Computing advantages, it does not treat user mobility, the lack of connectivity, or the high cost of mobile network traffic. In this chapter, the authors introduce mobile-to-mobile contextual offloading, a novel collaboration concept for handheld devices that takes advantage of an adaptive contextual search algorithm for scheduling mobile code execution over smartphone communities, based on predicting the availability and mobility of nearby devices. They present the HYCCUPS framework, which implements the contextual offloading model in an on-the-fly opportunistic hybrid computing cloud. The authors emulate HYCCUPS based on real user traces and prove that it maximizes power saving, minimizes overall execution time, and preserves user experience.