Research On Balanced Offloading Algorithm For Multi-task In Mobile Edge Computing

The continuous development of mobile networks poses new challenges for end devices with limited computing power.Mobile edge computing(MEC)is the answer to this problem.Mobile edge computing technology disperses computing and storage resources near the mobile terminal,and the mobile terminal fills its own resource shortage by transferring the computing and storage functions to the MEC server.However,in reality,MEC servers have limited computing resources and the network environment is becoming increasingly complex.Therefore,how to allocate the limited resources to the tasks to be offloaded in an efficient and reasonable way and how to solve the unbalanced load caused by the uneven distribution of tasks are key factors to improve network performance and user quality of service(Qo S).On the other hand,different types of services(online video,interactive games,files,etc.)have different requirements on the processing power of MEC servers.Therefore,it is meaningful to consider the problem of different offloading requirements for multiple services with limited server resources,and thus to explore the study of balanced offloading methods for multiple tasks in mobile edge computing.In view of the above,this paper studies and solves the problems of MEC server load imbalance caused by uneven service distribution,which leads to long queuing time of some MEC servers(overload),idle computing resources of other MEC servers(underload)and low user coverage rate.In this paper,a balanced unloading algorithm for multi-type tasks(BOMT)is proposed.In this algorithm,priority of unloading tasks is firstly divided according to the type,size and maximum allowable delay of unloading tasks.Then,according to the priority size and the current load of MEC server,different algorithms are executed for tasks with different priorities.Finally,the optimal unloading strategy can be found through iterative method.Simulation results show that the proposed BOMT algorithm can effectively reduce system delay,increase user coverage and complete rate of unloading task.