Research On Multi-User Task Offloading Strategy In Mobile Edge Computing

With the rapid development of Io T technology and 5G mobile communication technology,various new mobile applications in mobile devices are putting higher demands on data-intensive and computeintensive tasks in terms of latency and energy consumption.However,the limited computing power and battery capacity of mobile devices cannot guarantee the quality of service of such applications.The emergence of mobile edge computing provides a new solution to this problem by offloading intensive tasks to nearby edge servers,enabling users to achieve low latency and low energy consumption.However,the edge server computing resources and mobile communication resources are limited,and it becomes a critical issue to make optimal decisions on how to rationally allocate various resources.In this thesis,we start the research from the perspective of multi-user task offloading,discuss the centralized and distributed computation offloading problems respectively,and propose corresponding offloading strategies.The main research contents are as follows:(1)In this thesis,the centralized computation offloading problem is studied and analyzed.To address the complexity of networks and the difficulty of managing heterogeneous resources in ultra-dense networks,this thesis combines software defined networking with mobile edge computing and proposes a centralized computation offloading strategy based on the sparrow search algorithm.Firstly,a mathematical model is established by considering task latency,energy consumption and communication resources.An optimization problem is proposed to minimize the system energy consumption under the delay constraint.Then,the sparrow search algorithm,which is relatively novel and efficient in current research,is used as a framework,and the algorithm is improved by introducing the Cauchy mutation operator and the discrete operator.The simulation results prove the effectiveness and feasibility of the proposed algorithm.(2)In this thesis,the distributed computation offloading problem is studied and analyzed.For the centralized computation offloading system has many single point problems,this thesis proposes a distributed computation offloading strategy based on the mixed strategy game.Firstly,the communication and computing resources are modeled to quantify the energy consumption overhead and delay overhead in the computation offloading process,and a problem is proposed to minimize the weighted sum of energy consumption and delay for each mobile device.Secondly,this problem is transformed into the mixed strategy game and fictitious play is used to solve the mixed strategy Nash Equilibrium.Finally,experimental results show that the strategy achieves a lower overhead for all devices compared to other offloading strategies.