Research On Computation Offloading And Resource Allocation Of Mobile Edge Computing

With the continuous development of wireless communication technology and the increasing number and types of terminals,an increasing amount of data is generated at the edge of the network.At the same time,numerous,complex and diverse application scenarios have emerged,such as smart c ities,augmented reality,ultra-high definition live video and online games,etc.These applications generally have characteristics such as a large amount of calculation and delay sensitivity.However,mobile devices have shortcomings as well such as insufficient computing power,storage capacity and limited battery life,which makes it difficult to meet th e complex needs of applications.Meanwhile,the traditional single cloud computing model is no longer applicable,primarily because its computing power cannot match explosively growing edge data and the task transmission overhead is too large to meet the real-time needs of applications.In response to the problems described above,people are considering the possibility of the task being executed on the computing n odes at the edge of the network.Consequently,they have built a mobile edge computing model.The emergence of mobile edge computing can improve the capabilities of local data processing,reduce the time delay of data upload,as well as the energy consumption of mobile devices to a certain extent.This provides effective solutions for these applications.How to provide efficient computation offloading strategies and resource allocation schemes for complex and diverse new applications on the edge of the network is a key issue that needs to be urgently resolved in mobile edge computing.Computation offloading and resource allocation are the prerequisites for localized data processing and resource allocation.Notably,its execution efficiency and execution cost will directly affect the overall performance of the mobil e edge computing system.Therefore,in view of the characteristics of application delay sensitivity and high computing density,this paper focuses on the topics of computation offloading and resource allocation in mobile edge computing as a means to improve system performance and service quality.The main research work is as follows:(1)Aiming at the multi-cell network environment in mobile edge com puting,a hierarchical computation offloading and resource allocation optimization method framework is proposed,namely HIQCO.This paper initially modeled the mobile edge computing system in a multi-user,multi-cell network environment,and analyzed the problems of computation offloading and resource allocation under the model.Furthermore,we formulated a problem model that minimizes system overhead.Through analyzing the characteristics of the optimization problem,the problem is then transformed into two sub-problems.The first sub-problem is the computation offloading and channel allocation issue.The second sub-problem is the transmission power and computing resource allocation issue.Then,a hierarchical optimization method framework HIQCO based on immune algorithm,quasi-convex optimization technology and convex optimization technology is proposed to solve the optimization problem.Finally,the advantages of HIQCO in reducing system overhead are analyzed through the MATLAB simulation platform.(2)Aiming at the scenario of a multi-user and multi-server in mobile edge computing,a scheme for jointly optimizing computation offloading decision and MEC server resources is proposed,namely DCORA.Here,this paper first established an optimization function with the goal of minimizing system overhead s and transformed the original optimization problem into a Ma rkov decision problem.Moreover,we proposed a computation offloading and resource allocation scheme DCORA based on Deep Q Network(DQN).Finally,the experimental simulation results illustrate that the DCORA scheme can reduce system overhead while fully utilizi ng the limited resources of the MEC server.In summary,this paper mainly studies the offloading and resource allocation schemes in mobile edge computing,and proposes two solutions to minimize system overhead and improve system performance.In th e process of computation offloading,the computing resources of mobile devices and MEC servers are effectively used,which provides new ideas for the research on computation offloading and resource allocation in mobile edge computing.