Dynamic resource orchestration for multi-task application in heterogeneous mobile cloud computing

• Published in: 2016 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS) pp. 221 - 226
• Main Authors: Qi Qi, Jianxin Liao, Jingyu Wang, Qi Li, Yufei Cao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-04-2016
• Subjects: Cloud computing; Energy consumption; heterogeneous resources; Measurement; Mobile cloud computing; Mobile communication; Mobile handsets; multi-objective optimization; multi-task application; Optimization; resource orchestration
• DOI: 10.1109/INFCOMW.2016.7562076

The mobile cloud computing (MCC) that takes wireless access network as transmission medium and uses mobile devices as client becomes the newest evolution trends of cloud computing. When offloading the complicated multi-task application to the MCC environment, each task executes individually in terms of its own computation, storage and bandwidth requirement. Due to user's mobility, the provided resources contain different performance metrics that may affect the destination choice. Nevertheless, these heterogeneous MCC resources lack integrated management and can hardly cooperate with each other. Thus, how to choose the appropriate offload destination and orchestrate the resources for multi-task is a challenging problem. This paper decouples resource control of mobile cloud from user plane, where a centralized controller is responsible for resource orchestration, offload and migration. The resource orchestration is formulated as multi-objective optimal problem that contains the metrics of energy consumption, cost and availability. Finally, a particle swarm algorithm is used to obtain the approximate optimal solutions. Simulation results show that the solutions can hit Pareto optimum of resource orchestration in acceptable time.