Efficient Content Distribution Technology For Mobile Edge Computing

With recent advances in mobile communication technologies,as well as the efficient fast-rate 4G/LTE infrastructure,the demands for high-speed data applications,such as mobile multimedia services,have been growing explosively in recent years.The rapid growth of mobile data traffic will pose an enormous challenge to the next generation wireless systems(5G).To address this challenge,caching of popular contents during off-peak traffic periods at mobile network edges has emerged as an effective approach to cope with the tremendously increasing traffic over mobile networks.Caching technologies at the edge of wireless network can effectively reduce the duplicated transmissions of the content downloads demanded by mobile users and greatly improve user quality of experience.Recently,the emergence of mobile edge computing(MEC),which is based on the 5G evolution architecture,has accelerated the development of intelligent distributed caching at mobile network edges.The main goal of MEC is to reduce latency and network traffic congestion by bringing the computation and storage capacity from the core network to the edge network.MEC provides a highly distributed computing environment within the Radio Access Network(RAN)that can be used to deploy applications and services as well as to store and process content in close proximity to mobile subscribers.Furthermore,MEC can perform specific tasks that could not be achieved with traditional network infrastructure,such as application-aware performance optimization,big data analytics,and distributed content caching.Thus intuitively we can design new intelligent content caching and distribution mechanism by leveraging the data storage and compute capability of MEC.First of all,we propose a proactive caching mechanism named Learning based Cooperative Caching(LECC)strategy with aim of minimizing the transmission cost and improving content distribution performance in MEC architecture.We consider the scenario where several distributed MEC servers with local caches cooperate and share contents with each other.We formulate a proactive caching optimization model that needs no prior knowledge of content popularity matrix.We further show that the caching optimization problem is NP hard and use a two-step framework for designing the LECC strategy.We first propose a TL-based approach to learn content popularity.In our TL-based approach,to obtain the training data,we employ K-means clustering algorithm to classify all the content items based on access feature correlation.We formulate an optimal prediction model based on the input training data to learn the content popularity.We then formulate content caching for minimum transmission cost problem as a 0-1 integer-programming problem and design a greedy algorithm for a sub-optimal solution.From the numerical results,we find that the transfer learning approach is effective in popularity prediction and our proposed TL-based cooperative caching mechanism can effectively increase cache hit rate,reduce transmission cost and content delivery latency compared with existing known caching strategies.Next,in order to further reduce the high cost of base station transmission and improve the quality of network service,this paper proposes a MEC and D2 D co-caching mechanism based on user mobility.Unlike most studies of existing wireless cache networks,this paper does not adopt the fixed network topology,but fully considers the impact of user mobility on the D2 D link and caching strategy.In this paper,we develop a cooperative caching mechanism with aim of minimizing the downloading delay of users,where MECs and UEs are involved in local content caching.The hybrid genetic simulated annealing algorithm is used to solve the content caching scheme.The caching scheme is compared with traditional caching strategies,and the experimental results show that our caching scheme can effectively improve the cache hit rate and reduce the average downloading delay.