Research On Task Offloading And Resource Allocation Algorithm In Mobile Edge Computing

The rapid development of mobile Internet and Internet of Things promotes a series of new applications with computing-intensive and delay-sensitive features such as virtual reality,augmented reality and face recognition.However,due to the limited computing capability and battery capacity of mobile device,it is difficult to meet the needs of users for these applications.As a new computing paradigm to solve this issue,Mobile edge computing(MEC)can deploy high performance MEC servers at the edge of mobile network and allow users to offload their tasks to the MEC servers for processing,thereby effectively reducing execution delay of tasks and energy consumption of device.Considering the limited resources of the MEC system and the dynamic nature of task requests,how to rationally and effectively design task offloading strategies and resource allocation schemes has become a critical issue in MEC research.This thesis focuses on the task offloading and resource allocation in MEC.The main research contents and innovations of the thesis are summarized as follows:1.To solve the problem that the current MEC system is difficult to meet the lowlatency and high-reliability requirements of users due to the unreasonable design of MEC system based on average metric,a joint task offloading and resource allocation algorithm based on Lyapunov is proposed.Firstly,under the multi-MEC servers and multi-users mobile edge system architecture,the algorithm designs a dynamic scheduling model of task queues for users and MEC servers,and a resource optimization model is established to maximize the average profit of the MEC system with the queue overflow probability and total resources as constraints.Then,the probabilistic mixed optimization problem is transformed into a non-probabilistic optimization problem using Markov's inequality,and a resource allocation algorithm based on a single time slot is designed by using Lyapunov optimization.Finally,the original optimization problem is transformed into three sub-problems,namely user local computing resource allocation,power and bandwidth allocation and MEC servers computing resource allocation problem,and which are solved separately.The simulation results show that the algorithm can improve the time average profit of the MEC system while meeting the quality of service(Qo S)requirements of users.2.To solve the problem of low utilization of computing resources at edge nodes due to uneven spatial distribution of service traffic in mobile edge networks,a joint task offloading and resource allocation algorithm with edge node cooperation is proposed.Firstly,considering an MEC system architecture based on SDN,the algorithm jointly optimizes offloading decisions and computing resources by taking computing resources of MEC servers,and establishes a fair-aware delay optimization model of users.Secondly,since the optimization problem is a mixed integer nonlinear programming(MINLP)problem.In order to reduce the solution difficulty,this thesis decouples the original optimization problem into task offloading and computing resource allocation subproblem.Further,an iterative solution schemes based on bisection method and simulated annealing algorithm to obtain an approximation optimal solution to the optimization problem.Finally,the simulation results verify the convergence of the algorithm and show that the mechanism of edge node cooperation effectively reduces the delay of tasks and improves the utilization of resources.