Research On Joint Task Offloading And Data Cache Strategy In Mobile Edge Network

With the rapid growth of the number of mobile devices and the emergence of many emerging applications,mobile network traffic has grown exponentially,leading to an unprecedented increase in demand for data content and computing applications.The traditional centralized network architecture cannot meet the needs of mobile users due to the heavy load and long delay of the backhaul link.Therefore,a new architecture is proposed to expand network capabilities from the core network to the edge network,named Mobile Edge Computing(MEC).Mobile edge computing can provide low-capacity cloud computing and storage functions at the edge of mobile cellular networks to meet the needs of mobile devices capable of running latency-sensitive applications.However,the computing / storage resources on the edge server are limited and only the important application data can be cached on the edge server.Therefore,it is necessary to find optimal caching decisions to minimize the execution latency and energy consumption of mobile devices.How to combine task offloading and data caching to effectively and reasonably utilize the computing capability and storage resources of the edge server to meet the high experience requirements of mobile users is still an urgent issue to be addressed.The main research goal of this paper is to realize the joint computing task offloading and data caching of multi-user collaboration in mobile edge computing networks.The key issue in MEC data caching is how to balance the limited storage capacity of MEC servers with massive databases.This thesis considers a system where multiple mobile devices migrate duplicate computing tasks to an edge server and share the data content required for the computing tasks.In addition,this thesis proposes an effective Lyapunov online algorithm that can perform the joint scheduling task offloading and dynamic data caching strategies for computing tasks or data content.The main contributions of this thesis are described as follows:1.This thesis considers the task offloading strategy based on Lyapunov optimization in the D2D-assisted mobile edge networks.Firstly,this paper proposes a MEC-based mobile device queuing task sharing method.By offloading computation tasks to an MEC sever or an adjacent mobile device for execution,the task offloading optimization is modeled as the minimum average total energy consumption problem.Secondly,the task queue is stable under the condition,the Lyapunov optimization algorithm is used to solve the optimization problem.The proposed method dynamically adjusts the offloading decision of all tasks based on the weight parameters of the current task,and determines whether it is executed in adjacent mobile devices or the MEC server.Simulation results show that the algorithm can effectively reduce the average total energy consumption of task execution and greatly improve the offloading efficiency compared with the shortest queue waiting time algorithm and the all offloading algorithm of the MEC.2.This thesis considers the joint task offloading and data caching strategies based on user collaboration in mobile edge networks.Firstly,this thesis obtains the local computing latency of mobile computing tasks and the computing latency at edge servers,and the transmission delay of computing tasks based on the queuing theory.Secondly,under the conditions of meeting mobile device time and energy constraints,the task offloading and data caching policy problems are modeled as minimizing the average total execution delay problem of mobile devices.Finally,a real-time collaborative online algorithm based on Lyapunov optimization is proposed,which can efficiently perform the task offloading and data caching decisions online without the need for the future information.In particular,this thesis proposes a genetic algorithm as a key subroutine of the online algorithm.This algorithm effectively solves the optimal offloading and caching strategies and achieves collaboration between mobile devices.Simulation results show that our proposed scheme is superior to traditional strategies in joint task offloading and data caching,and can effectively reduce the average total execution delay of mobile devices while meeting long-term energy constraints.