Storage And Recommendation Based Optimization Design Of Mobile Social Network

The combination of D2D（Device-to-Device）and social network is one of the hot spots in the future development of the communications industry,the application of D2D technology has characteristics of wide coverage,low energy consumption,high system capacity and high spectral efficiency.In cellular network,it's one of the key technologies in the future communication networks.However,most of the existing D2D communication cache optimization designs are based on the traditional physical network environment,but do not take into account the influence of social network characteristics built by D2D communication on cache optimization.While the social network built by mobile Internet users is bringing the explosive growth of mobile data traffic,it also has an impact on the success rate of D2D communication.Meanwhile,the cache optimization strategy of user terminal devices is facing new challenges.This paper first studies the optimization design of D2D communication system based on user's storage resources and user's social attributes.In terms of D2D communication access protocol,in addition to the constraints of connection time and communication radius,we also consider the restriction of social relations on D2D communication link,that is,only two users with social relationship can make D2D communication.On the other hand,taking into account that users'terminal devices has a certain storage space,the files the user cached can be passed to his"friends"through D2D communication.And these"friends"can transfer the file to their"friends"of friends",with infinite diffusion.To this end,this paper studies the problem of user-side storage design to maximize the data offloading through D2D communication in social networks.We derive that the optimization problem is a 0-1 knapsack problem,which can be solved by the traditional greedy algorithm（Greedy）.But when the number of users reaches a certain magnitude,the computational complexity of the algorithm will increase rapidly,resulting in very large delay.Therefore,this paper proposes a community based greedy caching strategy to reduce the complexity of the algorithm.This new weighted Fast Unfolding algorithm divides the user into community based on the users'social relationships,locations,and communication radii.Then we prove that the optimal problem after reconstructing is monotone submodular maximization subject to matroid constraint,and can use the community based greedy algorithm to solve it.Finally,the simulation results show that the proposed cache configuration scheme has a significant performance improvement over the existing schemes.Secondly,in order to get larger data offloading ratio,based on users'social attributes,this paper studies the impact of users'interests and behaviors on users'storage,and proposes an optimization storage strategy with recommendation system.This paper first gives the D2D network model based on social networks and recommender system,divides the research area into several parts,then selects an important user（Important User,IU）^[53]for each part.Then by using three algorithms:pre-filter,collaborative filtering algorithm and latent factor algorithm,designs a recommendation system based on mobile social networks and user'download history.Finally,the content produced by the recommendation system can be cached at the terminal of the important user,and the other users can get the requested file through the D2D communication.