Traffic-aware task allocation for cooperative execution in mobile cloud computing

Cooperative application execution among mobile devices has recently received significant attention in mobile cloud computing. However, it may incur extra traffic among devices, e.g., exchanging the results of the tasks, especially for data-dependency tasks. In this paper, we focus on the task execution in a chain structure, where each task must be executed based on the output result of its precedent task in the chain. We design task allocation schemes to the mobile devices such that the total traffic incurred during task execution is minimum, while the constraint of the resources on mobile device is not violated. Based on that, we also aim to minimize the total number of mobile devices participated in task execution. A dynamic programming and a bin packing algorithm are proposed to complete the above two objectives respectively. Finally, simulation results demonstrate the effectiveness of the proposed schemes in reducing the traffic and the number of required mobile devices.