CRII: NeTS: Exploiting System and Network Dynamics in Mobile Clouds

University Of Pittsburgh, Pittsburgh PA

Investigators

Abstract

This project focuses on exploiting various aspects of mobile system and wireless network dynamics to improve the efficiency of mobile cloud computing. Exploitation of these dynamics has substantial impacts on practical integration of mobile devices into the cloud, by alleviating the local computation burden of mobile devices and allowing adaptation to the heterogeneous system and environmental contexts. The research results are likely to inspire novel theoretical and systematic studies that open up new areas of research in highly interactive mobile computing systems. The project plans to engage female and under-represented minority students in the research activities and disseminate the research results broadly to the community. Mobile cloud computing bridges the gap between the limited capabilities of mobile devices and the increasing complexity of mobile applications, by offloading the computational workloads from local devices to the remote cloud. However, the effectiveness of mobile cloud computing could be impaired by the dynamic nature of system and network contexts, which lead to heterogeneous mobile application behaviors and seriously reduce the appropriateness of workload offloading decisions. The goal of this research is to exploit these critical dynamics in mobile clouds that are indispensable to efficient, prompt, and reliable workload offloading. More specifically, the project addresses three closely intertwined research issues in mobile cloud computing. The first part investigates how to analytically formulate the stochastic characteristics of run-time application executions, based on which the workload offloading decisions are probabilistically made and systematic techniques are developed to practically enforce such decisions. The second part incorporates the contexts and performance requirements of mobile cloud applications into the design of wireless networks, so as to adaptively balance between the wireless energy cost and application performance in mobile clouds through fundamental redesign of wireless transmission scheduling algorithms. The third part focuses on testbed development to automatically investigate the run-time system and network dynamics of mobile cloud applications in practice. This testbed consists of off-the-shelf smartphones and wearable devices, and enables in-field experiments for evaluating the performance of the proposed techniques and system designs.