移动云计算的任务迁移方法

移动设备在移动过程中受到网络波动的影响很大,由于工作流中任务会传输到云端执行,会带来传输与接收的能耗,再加上所处云域中微云受到不同请求数、剩余资源等因素的影响,造成了微云响应延迟、任务等待执行时间延迟和网络活跃能量损耗,再加上异构网络对数据传输造成的切换延迟影响,使任务迁移延迟更加严重,所以在上传与接收阶段采用延时传输的策略,在保证总完成时间的基础上减少传输时间;在任务迁移阶段提出了微云跳跃选择算法和网络切换算法相结合的MJSA-NHA算法,通过跳跃选择找到合适的微云以减少不必要的响应与等待延迟,并且通过切换网络空闲—活跃状态减少网络能量损耗;并在延时传输基础上加入移动速度调节算法(MSAA),在发生网络变化过程中变换移动设备的速度,这样可以在减少网络切换次数的同时减少网络切换延迟。实验结果证明,延时传输策略比Random算法在总完成时间和移动端总能耗问题上优化了很多;MJSA-NHA比MuSIC算法、task delegation和code-offloading在迁移时间上分别优化了66%~86%、36%~56%和5%~8%,在能耗方面分别优化了56%~78%、25%~46%和6%~8%;MSAA比MuSIC算法、task delegation和code-offloading在网络切换延迟上优化了50%。

Task migration method for mobile cloud computing

In the mobile process,mobile devices are greatly affected by network fluctuations.Because tasks in the workflow will be transferred to the cloud for execution,it will bring energy consumption for transmission and reception.In addition,the micro cloud in the cloud domain is affected by different requests,remaining resources and other factors,resulting in the micro cloud response delay,task waiting time delay and network active energy loss.Due to the impact of switching delay caused by heterogeneous network on data transmission,the task migration delay is more serious.This paper adopted the strategy of delay transmission in the stage of upload and receive to reduce the transmission time on the basis of ensuring the total completion time.In the stage of task migration,this paper proposed the MJSA-NHA algorithm,which combined the micro cloud jump selection algorithm and the network switching algorithm.Through the jump selection,it could find the appropriate micro cloud to reduce unnecessary response and waiting delay,and reduced the network energy loss by switching the idle active state of the network.This paper added the mobile speed adjustment algorithm(MSAA)on the basis of delay transmission,changed the speed of mobile devices in the process of network change,which could reduce the number of network switching and reduce the network switching delay.Experimental results show that the delay transmission strategy is much better than Random algorithm in terms of total completion time and total energy consumption of mobile terminal;MJSA-NHA algorithm is 66%~86%,36%~56%and 5%~8%better than MuSIC algorithm,task delegation and code offloading respectively in terms of migration time,56%~78%,25%~46%and 6%~8%in terms of energy consumption;MSAA is 50%better than MuSIC algorithm,task delegation and code offloading on the handover delay.