The Research On Distributed Cooperative Cache Mechanism For Mobile Communication Network

With the rapid development of mobile communication technology and new Internet applications,people’s demand for communication has gradually shifted from people to people to interactions and collaborations between people and things,things and things.The large-scale growth of Io T applications such as virtual reality,8K live broadcast,and telemedicine have gradually shifted data services from connectioncentric to content-centric.The global mobile data traffic is increasing exponentially,and the traditional cloud-end approach cannot cope with the high load and network congestion caused by the rapid development of the Internet of Things.Multi-access edge computing technology can provide computing,storage,and differentiated network service capabilities at the network edge,promote the deep integration of business platforms and mobile networks,and establish a carrier-grade service platform with low latency,high reliability,and high speed to improve the quality of users’ network experience.This thesis focuses on the challenges faced by the deployment strategy of distributed collaborative caching in mobile communication networks.From the perspective of joint edge caching and terminal caching technology,aiming to minimize the average delay of user content acquisition and system energy consumption.We study how to achieve collaborative configuration and optimization among cellular network,edge caching and terminal caching resources in depth to reduce the load on the backbone network,avoid network congestion,and improve the quality of users’ network experience.The main research content and work are summarized as follows:1.Aiming at the problem of limited edge cache server space and low content diversity in a single server,a distributed collaborative cache solution based on dynamic programming is proposed.First,we consider the advantages of D2 D communication in the edge caching networks and build the problem of minimizing the delay of user content acquisition by segmenting the more popular content to different network nodes.To solve this problem,the collaborative deployment of content is reduced to a complete knapsack problem,and a distributed collaborative deployment strategy based on dynamic programming is proposed.The simulation results show that the proposed scheme reduces the delay of user content acquisition while increasing the diversity of system content.2.Aiming at the issue of high energy consumption for cache deployment in mobile scenarios,a content deployment strategy based on deep reinforcement learning to minimize the energy consumption of the cache system is proposed.We establish a dual time-scale cache content deployment/update model based on the Markov decision process,use the Google S2-based traffic area division model to divide the users ’ mobile area,and pre-place the content with high popularity in the edge nodes according to the location of the mobile users,by analyzing content popularity and users’ willingness,a content collaborative deployment algorithm based on Deep Deterministic Policy Gradient(DDPG)is proposed to obtain dynamic content deployment decisions.Simulation results show that the proposed content deployment algorithm can quickly converge and effectively reduce cache energy consumption.