Research On Computational Offloading Technology In Mobile Edge Computing

The development and implementation of 5G networks and the popularization of mobile devices(MD)lead to an increase in the demand for massive device connections and computing.The emergence of new computationally intensive applications,such as video recognition and telemedicine,has not only increased the requirements on the computing power of the equipment,but also put forward higher requirements on the time delay.Mobile Edge Computing(MEC)is a new computing paradigm.By deploying computing and storage resources at the edge of the network or near the terminal,it is a device with high-intensive computing requirements,low latency requirements,or lower energy consumption requirements so that computing offloading services can realize the processing of equipment tasks better.However,the computing resources in the MEC server are limited.As a result,the research on the problem of resource allocation and offloading strategy maximization plays a vital role in improving the efficiency of the MEC server.This thesis has designed two different communication scenarios under the 5G network,and carried out research based on different communication scenarios,and explored the optimization problem of computing offload in MEC.The results are as follows:When it comes to the problem of computing offloading of multiple users in a singlelayer communication network scenario,the following jobs are done under the condition of limited edge server resources.First and foremost,both the calculation model and communication model are constructed based on the task communication data volume and calculation data volume.Besides,a mathematical model of the server computing resource allocation problem is built,and the objective function is proved to be a convex function,which proves the solvability of the problem.What's more,considering the time delay and energy consumption of the user's equipment,the time delay and energy consumption are normalized together.The weighted sum of the two is used as the system utility function to represent the utility value obtained by offloading calculation.Finally,a joint computing resource allocation and offloading decision method are proposed and realized to maximize the total utility,which combines the process of computing resource allocation with the technology of simulated annealing algorithm.The MATLAB software is used to implement the algorithm simulation.The results show that the offloading strategy can not only adjust the weight value of the time delay and energy consumption according to the user's equipment dynamically,but also achieve the purpose of optimizing the offloading benefits.When it comes to the problem of multi-user computing offloading in the scenario of a double-layer communication network,the optimization model and optimization function are constructed.To begin with,the interval model of task assignment is established to realize the pre-assignment of offloading strategy based on the two task attributes of data volume and computation volume.What's more,a communication resource allocation method based on the potential game is designed to maximize the total throughput in the network,because there are many devices in a two-layer communication network,which will increase the competition for communication resources.Besides,the task priority model is established based on the data volume and delay constraints by using(0,1)standardization method so that the computing resource allocation scheme of the MEC server is realized.In the end,the preallocated offloading strategy is further optimized to maximize the gain value on the basis of the iterative comparison of the gain value of the optimization function.The experimental simulation of MATLAB software shows that the offloading strategy can achieve better throughput performance,and the gain value can be increased and stable when the number of users continues to increase.