Research On Task Offloading And Cache Management In Mobile Edge Computing

With the rapid development of 5G technology in recent years,modern wireless applications have proliferated,such as cognitive aids,mobile augmented reality,etc.However,the computing power,memory and battery life of mobile devices are limited,which cannot meet the requirements of these computationally intensive and delay-sensitive tasks.Mobile edge computing has been introduced as a new computing paradigm to solve these problems.Applications running on mobile devices can offload computation-intensive tasks to nearby edge servers at a lower cost,which will effectively save mobile device resources and improve task processing efficiency.At the same time,in order to ensure that delay-sensitive and computation-intensive tasks are carried out efficiently in the edge network,the edge network also provides computational content caching capabilities.Compared with centralized cloud computing,edge servers are scattered and have limited computing and storage resources.Unreasonable task offloading and computing content cache placement will cause problems such as low resource utilization and long service delay.We take mobile augmented reality tasks,which are computationally intensive and time-delay sensitive,as an example to study the task offloading and related parameters optimization of mobile augmented reality tasks in moving edge computing.On this basis,the joint optimization problem of task offloading and cache management is studied.The contributions of this thesis are as follows:(1)The thesis solves the offloading and parameters optimization problem of mobile augmented reality tasks in mobile edge computing.Considering the effects of wireless bandwidth resources,central processing unit frequency of clients and resolution size on energy consumption,service delay and detection accuracy of clients during task offloading and parameters optimization,a function is designed to evaluate the energy efficiency of mobile augmented reality clients.Under the limitation of mobile augmented reality task completion time and wireless bandwidth resources,the problem of task offloading and parameters optimization is formulated to minimize the energy efficiency function.We propose a server selection and parameters optimization algorithm based on block coordinate descent to solve this problem.The algorithm first generates the priority queue for the task.Based on the order of the priority queue,the tasks are offloaded to the appropriate mobile edge server according to analytic hierarchy process.Then,related parameters are optimized according to the block coordinate descent,and tasks are reassigned according to the completion time until the energy efficiency function converges.Simulation results show that our algorithm performs better than the compared algorithms.(2)The thesis solves the problem of task offloading and cache placement of mobile augmented reality tasks in multi-edge server cooperative system.In this thesis,the mobile augmented reality task is divided into five sub-tasks according to the mobile augmented reality application components,and the dependencies between the five sub-tasks are given.Hit ratio and service latency are designed to evaluate task offloading and cache placement on edge servers.Under the constraints of the computing resources and cache space of the MEC server,the problem of task offloading and cache placement is formulated to maximize hit ratio and minimize service latency.Therefore,a task offloading and cache placement algorithm based on multi-objective artificial bee colony is proposed.When initializing the cache placement,this thesis considers the influence of popularity,time and space on the cache placement strategy.In the optimization process,in order to avoid falling into the local optimum,this thesis improves the solution update formula and introduces the pareto optimal relation to find the optimal solution.The experimental results show that the algorithm has better performance than the contrast algorithm.