Research On Computing Offloading Method Under Mobile Edge Computing

With the rapid development of mobile Internet technology and the popularization of various smart terminals,many computing-intensive applications appear in people’s daily lives,such as face recognition,virtual reality and autonomous driving.However,due to the demanding requirements of computing power and delay,the application scenarios of those application programs are greatly limited.In this context,the mobile edge computing(MEC),as an excellent paradigm to solve this problem,has attracted great attention from the academic community in recent years.As a key technology in MEC,the computing offloading technology is also a research hotspot in the current academic circle.How to adopt an appropriate computing offload strategy is the key to improving the user experience of the MEC system in different application environments.This thesis mainly studies the solution of computing offloading strategy in MEC system with multi-user and single server.The main work and innovation points are as follows:First,in view of the high cost of solving the computational offloading strategy in most of full offloading research,this thesis proposes a low-cost computational offloading solution based on the improved beetle algorithm.A multi-channel exploration strategy is introduced into the beetle algorithm to strengthen the global optimization ability of the algorithm,and at the same time,a simulated annealing strategy is adopted in the selection of candidate solutions to reduce the probability of falling into a local optimum.Simulation experiments show that compared with other algorithms,this algorithm can obtain better system benefits with lower time cost.Second,this thesis studies the problem of partial offloading of independent tasks in NOMA-MEC system.According to the task delay requirements and the energy consumption of mobile devices,this thesis constructs a system revenue maximization model.By introducing the chaotic optimization strategy and the distance hierarchical search selection strategy into the standard RSA algorithm,this thesis proposes a new computational offloading solution based on a new the improved reptile search algorithm.Experiments show that the solution has good performance in algorithm optimization and high-dimensional problems.Finally,in order to solve the mutual dependence of each subtask in the current computing task,this thesis proposes a task classification strategy that combining serial and parallel tasks.On the basis of the task classification strategy,the concept of weight coefficient is introduced for tasks of the same level.Maximize the system benefits of MEC system while ensuring the orderly completion of each sub-task.In addition,in order to obtain the optimal solution,this thesis proposes a computational offloading solution based on an improved arithmetic optimization algorithm.In view of the insufficient diversity of the algorithm population,this thesis introduces the strategy of reverse learning and elite particle competition into the traditional arithmetic optimization algorithm to improve the optimization performance of the algorithm.The simulation experiments demonstrate that the algorithm has good performance in this model.