Energy-efficient Task Scheduling Strategy Based On Cloud Assistant In Mobile Devices

With the development of Wireless Communication and Information Technology, mobile devices with plenty applications bring great convenience to people’s daily life. However, these complex applications consume a lot of energy and further decrease the battery life. At the same time, the breakthrough of battery technology can hardly obtain in a short period. So the energy-efficient of mobile devices becomes to an urgent and key problem. The cloud computing bring a new idea to save energy and expand abilities of mobile devices.Aimed at solving the energy-efficient problem on mobile device, we propose an energy-efficient strategy of task scheduling with the assistant of mobile cloud computing in this paper. In most cases, one application in a mobile device can be decomposed into a set of fine-grained tasks which consists of sequential tasks and parallel tasks. With the cloud assistant, some tasks could be offloaded to the cloud side for speeding up the executions and saving the energy of mobile device. However, the task offloading results in some additional cost during the communication between cloud and mobile devices. However, due to the transmission process of these tasks from wireless channel, the task scheduling will bring in additional transmission energy consumption while decrease the execution energy. Before offloading to cloud, we should make sure if it can save energy for the contradiction between execution energy and transmission energy. So an energy-efficient task scheduling is necessary to decide which tasks should be offloaded to the cloud side for execution.This paper firstly build the system model for this task scheduling problem from three aspects, which is task model, execution model and transmission model respectively. The optimal scheduling aims to minimize the energy consumption of mobile device for one application within the total completion time. This task scheduling problem can be reconstructed into a constrained shortest path problem and the LARAC method is applied to get the approximate optimal solution.Compared with the pure execution strategy, our strategy with cloud assistant decreases82.47%of energy and25.70%of time at most. Moreover, the availability and effectiveness of the proposed strategy are verified in different applications, where the time constraint, the amount of data and computation workload change.