| With the development of 5G and the Internet of Things,the number of mobile users has increased dramatically,and users’ demand for various Internet services has exploded,which causing problems such as network congestion,large transmission delays,and poor user experience.Content caching has been unable to meet the ever-increasing content demands from users.By sinking the content to the edge side,the edge caching greatly relieves the transmission pressure of the core network and reduces the network delay caused by user content requests,which has become one of the key technologies in mobile edge computing.However,most traditional edge nodes work independently or provide caching services in a point-topoint collaborative manner,ignoring the high similarity of user content requests,resulting storing a large amount of duplicate content in adjacent edge nodes,which cause redundant caching.The storage capacity limitation of a single edge node can no longer meet the intensive content requests of mobile users,users directly obtaining content from the content center server will introduce high transmission delay.In addition,the mobility of users may lead to frequent switching of communication links in content delivery process,which makes it difficult to ensure stable transmission of content.It is of great significance to study the multi-mode caching and delivery technology of mobile terminal content based on edge collaboration.In view of the redundant caching problem caused by insufficient collaboration of edge nodes,this paper proposes an edge collaboration domain construction and content caching method based on multi-user regional content preference.Adjacent regions are merged into collaborative caching domains,and a domain management server is designated for each collaborative domain to realize resource sharing on the edge side;then,the content pre-caching algorithm is designed from the perspective of the global collaboration domain and each region to avoid repeated content caching.Finally,to fit the time-varying characteristics of user content requirements,an update algorithm of the collaboration domain is designed to realize a balance between the service capability and stability of the collaboration domain.Simulation results show that the proposed edge caching strategy based on collaborative domain can improve the resource hit rate on the edge side,reduce the load pressure on the content cloud center,and improve the diversity of network resources on the edge side.In order to solve the problem of high latency caused by unclear edge node collaboration mode and lack of unified management in dense content request scenarios,this paper proposes a multi-mode delivery strategy for mobile terminal content based on deep reinforcement learning.Based on this,four content delivery modes are further designed:edge direct delivery,edge collaborative indirect delivery,cross-domain collaborative delivery and cloud delivery,which define the collaboration process of edge side nodes.Furthermore,a terminal trajectory prediction model is designed to obtain terminal locations in advance to deploy edge collaboration and bandwidth resources in advance.Allocation scheme;finally,with the goal of maximizing average user satisfaction,based on deep reinforcement learning to solve the optimal delivery mode and bandwidth resource allocation scheme,it solves the problem that when a single edge node cannot meet the content demand,users obtain content from the content center high transmission delay problem.The simulation results show that the proposed algorithm is superior to other algorithms in terms of convergence and delay,which verifies the effectiveness of the proposed scheme.With the characteristics of edge cache network,this paper studies content caching and content delivery strategy from the perspective of multi-mode edge collaboration,which improves the cache and content delivery efficiency on the edge side,which is of great significance to the caching optimization of edge network. |
PDF Full Text Request