Research On Edge Computing Cache Configuration And Management

With the development of computing and data intensive applications such as virtual reality and augmented reality,higher requirements are put forward for the service ability of the network.Traditional cloud computing offloads computing tasks to the remote data center,which makes the processing delay of the whole task difficult to be guaranteed,so mobile edge computing came into being.Mobile edge computing places computing and storage on the edge of the network.Edge cache can reduce user access delay and reduce the traffic pressure of the core network by storing content.However,mobile edge cache network is a complex cache system,and the theoretical modeling and analysis are very complex,which makes the implementation of efficient mobile edge network cache system configuration and management more challenging.Therefore,based on the characteristics of the existing mobile edge network architecture,this thesis focuses on the cooperative cache configuration and management in mobile edge network.Based on the typical two-tier architecture of the existing mobile edge network,the third chapter first designs a Cooperative Cache Model for the mobile edge cache network system.In order to reduce cache redundancy,the model requires a content object to be cached on at most one node,and all cache nodes in the system cooperate with each other to provide content caching services.Then,based on the cooperative cache model,this thesis studies the optimization of Cooperative Cache management.Specifically,this the-sis describes the cooperative cache management optimization problem as a combinatorial optimization problem,and proves that it is NP hard.For this NP hard problem,an effi-cient approximation algorithm is proposed in this thesis.When the content request process follows Poisson process and Pareto process respectively,the approximation ratio of the approximation algorithm is 2 and^2?_?-1respectively.The experimental results show that compared with the existing solutions,the cache management scheme under the cooper-ative cache model proposed in this thesis can achieve better performance in the total hit rate of content and access delay.Mobile Edge cache network needs to meet the different performance objectives of users,content providers and network operators.The existing work mainly considers the total hit rate and delay of the request content from the perspective of users.However,network operators pay more attention to the total service revenue obtained by caching content.The existing research lacks to consider different optimization indicators from the perspective of operators.Therefore,the fourth chapter of this thesis designs a cooperative caching model for mobile edge caching network with the goal of maximizing service rev-enue.The model can not only meet the maximization of service revenue of operators,but also ensure the user service quality(cache hit rate)to a certain extent.This thesis describes the cache configuration and management problem under the Cooperative Cache Model as a convex optimization problem.In order to solve this problem effectively,this thesis uses the primal dual decomposition framework to decompose the convex optimization problem into multiple sub-problems for iterative solution.In order to further improve the efficiency of the algorithm,a greedy algorithm which can get the optimal solution is designed for the solution of the sub problem.The simulation results show that the cache configuration and management strategy proposed in this thesis can achieve better performance in terms of user satisfaction and service revenue compared with the existing solutions.