Research On Optimization Of Task Offloading And Scheduling For Mobile Edge Computing

With the rapid development of mobile services in recent years,the mobile computing tasks require stronger computing ability and smaller processing delay.However,the computing ability and battery capacity of the mobile device is insufficient,computing tasks directly on a mobile device may not meet the latency requirements of the tasks,and it can also reduce the standby time of the mobile device.Mobile Cloud Computing(MCC)deploys the servers in the remote data center,which causes long data transmission delay between the mobile devices and the servers.Therefore,MCC cannot meet the latency requirement of mobile computing tasks.Mobile Edge Computing(MEC)deploys servers at the edge of mobile network(such as base stations),which reduces the data transmission delay between servers and mobile devices.Therefore,MEC solves the problem of mobile computing tasks' demand for computing ability and processing delay,and also reduces the energy consumption of mobile devices.However,the MEC servers have limited computing and storage resources and mobile networks have limited bandwidth resources,it is impossible to offload all mobile computing tasks to the MEC servers for processing,so the MEC servers need to offload and schedule mobile computing tasks reasonably to achieve efficient utilization of servers' resources.In practical scenarios,the calculation amount(the number of CPU execution cycles)required by the mobile computing tasks is uncertain,that is,the uncertain demand tasks.However,in the researches on MEC tasks offloading and scheduling optimization,most researchers assume that the calculation amount required by the tasks is certain,so this assumption is too idealistic,which reduces the practicality of the designed optimization algorithm.In addition,the existing studies on MEC tasks offloading and scheduling for uncertain demand tasks do not consider the delay constraint of mobile computing tasks,and model the problem as a Geometric Programming model,which is difficult to solve and cannot meet the delay requirement of mobile computing tasks in practical scenarios.Aiming at the shortcomings in the existing researches,this paper considers the mobile computing tasks are completed with a high probability within the maximum tolerable time as a constraint,and considers the average value of the time and energy consumption of the mobile computing tasks as the optimization goal of the model,and designs an algorithm to solve the model.Then simulation results show that the algorithm can obtain a good result in short time and with a small number of iterations.In addition,the existing studies on MEC tasks offloading and scheduling optimization consider all mobile computing tasks obtain the required resources at the same time and then release the occupied resources at the same time after all tasks have been calculated.This will result in that after some tasks with less computing time are completed,the resources occupied by them cannot be reused by other computing tasks,resulting in idle and waste of servers' resources.Aiming at the shortcomings in the existing researches,this paper considers dividing mobile computing tasks into multiple batches for offloading to the servers for calculating,and designs an algorithm and other three comparison algorithms for solving the model.Simulation results show that,compared with the other three comparison algorithms,the algorithm can obtain a result close to the optimal solution in a short time;meanwhile,the offload scheduling optimization model can improve the utilization efficiency of servers' resources to a certain extent.