Joint Optimization Design Of Edge Content Caching And Recommendation In Wireless Networks

Edge caching can realize the local response of some user requests by caching part of content objects at the edge nodes of wireless networks,and reduce the transmission latency of requested content objects,improve the user experience and balance the network loadings.Moreover,content recommendation can guide users to make requests for specific content objects,and improves users' dependence and satisfaction with related businesses by designing personalized content recommendation lists according to users' interests and preferences in the content objects of different subjects.Therefore,the joint design of edge caching and content recommendation in wireless networks can guide users to request cached content objects through content recommendation,so as to further improve the utility of edge caching and user experience.However,there are still many technical challenges in the aspects of modeling the content popularity and user preference,evaluating the utility of content caching and recommendation,and compromising the performance and cost-benefit of joint optimization algorithm.Therefore,this paper studied the joint optimization method of content caching and recommendation in wireless networks,aiming to balance the utility and cost of content caching and recommendation,and realize the real-time control of edge caching,content recommendation and content transmission.The main work and innovation points of this paper are summarized as follows:First,a joint optimization algorithm of content caching and recommendation based on deep reinforcement learning is proposed in this paper to realize the real-time control of edge caching,content recommendation and content transmission.Firstly,a joint optimization model of content transmission and recommendation with joint edge caching in wireless networks is established,and a mixed integer nonlinear programming problem is presented.In wireless networks,the joint optimization problem of content caching and recommendation is established by taking into account factors such as different user preferences for different content objects,transmission latency from the central server to the base station,user waiting latency,and the communication status,etc.,with the optimization goal of minimizing the transmission latency and backhaul loadings.Secondly,a deep reinforcement learning based joint optimization algorithm of content caching and recommendation is proposed.In order to decouple the coupling variables in the proposed optimization problem,the optimization problem is divided into several sub-problems.Moreover,in order to realize dynamic and real-time content caching and recommendation decisions optimization at the edge,deep reinforcement learning algorithm is applied to solve the joint decision problem.By designing reasonable state,action and reward function,the dynamic,real-time and efficient optimization decisions of content caching and recommendation can be realized at the edge.Finally,the simulation results are given to evaluate the performance of the proposed joint optimization algorithm,which show that the proposed joint optimization algorithm can achieve certain performance gains in content transmission latency and backlink loadings compared with the traditional algorithms.Second,a content request perception based joint optimization algorithm of content caching and recommendation is proposed.By mining the characteristics of user behavior on the history data,artificial intelligence technology is applied to better understand user behavior and characteristics of the network and user context information,and better predict the request distribution of the content objects.Moreover,a content popularity perception based joint optimization strategy of content caching and recommendation is designed,further improving the network performance and user experience.Simulation results show that the proposed algorithm can achieve a certain performance gain and effectively improve the caching efficiency.