Research On Task Offloading And Migration Algorithm In Mobile Edge Computing

With the rapid development of mobile Internet and Internet of Things technologies,human beings have entered the era of Internet of Everything.In order to solve the high latency of traditional cloud computing and the limited processing power of users,the mobile Edge Computing(MEC)technology is proposed.The MEC migrates computing and storage capabilities from the remote data center to the edge of the network close to the user,which is aiming to provide computation service quickly and directly.Mobile user can offload task to the MEC for processing,and the task can also be migrated between MECs.Therefore,this thesis researches task offloading and migration issues based on MEC.First of all,users are moving at any time in a MEC network scenario,and the mobility of the user will have a great influence on the offloading of the task workload.To solve this problem,this thesis investigates the impact of randomness caused by movement of user on decision making for offloading and migration,where the connection between user and the served MEC is uncertain.Consequentially,if the assigned tasks cannot match with the real connection time well,task migration would be caused.First of all,in order to address the impact of this uncertainty,the multi-stage stochastic programming is used to resolve the effects of connection time uncertainty.Meanwhile,in MEC networks,considering multiple candidate MEC servers could be serviced simultaneously due to overlapping,this thesis also introduces multi-arm bandit for MEC selection.The extensive simulations validate the improvement and effectiveness of the proposed Task Offloading and Migration Algorithm based on Stochastic Programming and Multi-Arm Bandit(TMSM)for offloading in terms of service latency,energy consumption,cost,and reward.In addition,compared with the traditional cloud computing,the computing resources of the MEC is small.When the user task is very large,the computing and communication delay generated by the MEC will also increase,which will become the constraint MEC.Therefore,from the perspective of reducing the delay,this thesis proposes a Joint Optimization of Communication and Computational Delay of Task Offloading and Migration Algorithm(JOCCM).The algorithm can determine whether the task is migrated and where the task is migrated in the MEC.Simulation experiments show that the algorithm can effectively reduce the system delay in the MEC network environment.