Research On Task Offloading Scheduling Strategy In Mobile Cloud Computing

With the rapid development of mobile cloud computing,mobile devices are becoming more and more popular.However,due to the limitations of computing resources and storage resources in mobile devices,many applications can not be done on mobile devices alone,so that complex applications are migrated to remote servers as needed to take full advantage of their rich computing resources It is necessary to get high performance results.So the task unloading technology came into being.However,task unloading is not beneficial in any case,and task dumping is meaningful only if the cost of the remote execution of the task is less than the cost of local execution.Task in the end do not want to uninstall,to uninstall where the task has become a task to uninstall the key issues,so the study of a task to uninstall scheduling strategy is particularly important.Based on the limitations of the existing offload scheduling strategy,this paper puts forward a joint decision strategy based on how to make the uninstall decision more flexible to cope with the fluctuation of cloud server node load and the change of network condition.Joint decision strategy by the mobile side and the server side to participate in the task of unloading scheduling,greatly reducing the burden on the mobile side.On the server side,this paper proposes a task scheduling algorithm based on memory optimization,and predicts the remote execution time according to the scheduling algorithm of the server and the demand information from the mobile side.As predictions are based on server real-time load changes,predictions are more accurate.The decision engine of the mobile terminal receives the execution time estimate from the server,combines the output of the local time /energy consumption valuation model,calculates the utility value according to the multi-attribute objective function,and selects the execution position of the minimized utility value for the task.In the process of uninstallation,this paper increases the deadline-based task switching mechanism when the task is executed abnormally,and avoids the dynamic timeout caused by the network fluctuation.Finally,this paper uses the data provided by the container-based task unloading system to verify the joint decision strategy.Compared with the RA strategy and CTM strategy,the decision of this paper has higher task completion rate,smaller execution time and energy consumption cost,and proves that the joint decision strategy can dynamically calculate the energy consumption of the balance based on the battery balance Weights.