Research On Efficient Content Placement And Delivery Mechanism In Small Cell Networks

With the popularity of various smart handsets,and the emerging of new generation applications such as ultra-high-definition real-time video,augmented reality/virtual reality,and large-scale interactive online games,the number of mobile users and data traffic density in the popular area has explosively grown.These facts cause network signaling soaring,user data blasting,backhaul short-age,and energy consumption increasing.As a result,the performance of tradi-tional wireless communication networks deployed by operators has been dras-tically reduced,and the service experience of users has been further decreased.The content placement and delivery technology deploys the cache and process-ing servers on the access network,and thus can reduce the delay and improve the service experience of users.Meanwhile,the content placement and delivery technology can also reduce redundant signaling traffic and content data traffic in the access network,which in turn can save backhaul capacity and improve overall network performance.Therefore,the content placement and delivery technology receives extensive attention from the academic community and the industry,and becomes the focus in the wireless communication research field.The small cell network(SCN)uses small-cell base stations(SBSs)for networking,so that it has high spectrum efficiency and energy efficiency and makes itself a flexible and easy-expanding structure.Based on these advan-tages,the SCN can efficiently improve the quality of service and overall net-work performance.As content placement and delivery technologies are grad-ually applied to SCNs,these networks are also facing some new problems,such as ineffectiveness on management of tremendous equipments and back-haul overloaded etc.Therefore,it is necessary to study these issues and pro-pose more feasible mechanisms for efficient content placement and delivery in SCNs.Based on the characteristics of the SCN,this paper studies the con-tent placement,the content delivery,and the content update in the small cell network.The main work and innovation are summarized as follows:1.An iteration-based hierarchical content placement scheme is proposed to optimize the SCN content placement process.Firstly,we study the hierarchical content placement problem and divide it into two layers:the first layer is called Cache Selection layer and focuses on the process from content providers(CP)to service providers(SP)to reduce the costs and increase the utility of CP;the second layer is called Content Delivery layer and focuses on the process from SP to users to make the utilities of SP and users reach a trade-off.Furthermore,we combine the two processes by setting a content profit parameter used in both processes.Then,the utility functions of the CP,SP and users are all defined,and the hierarchical content placement problem is modeled by integer programming and auction,aiming at maximiz-ing each utility function respectively.After that,we design an iteration-based hierarchical content placement scheme including simple branch-bound algo-rithm and multi-commodity auction.Simulation results show that the scheme can achieve high content placement efficiency with low complexity.2.We propose an efficient distributed content delivery mechanism to im-prove the quality of service.Firstly,we study the content delivery problem in the small cell network,which aims at maximizing the utility functions of each small-cell base station and each user with the constraints of the backhaul capacity and the maximum transmitting power limitation.Secondly,we model this problem as a multiple commodity auction problem.Considering the limitation given by traditional auction model,a set of distributed auction protocols is used to convert the orig-inal problem into a distributed auction problem.Next,we propose a parallel distributed auction algorithm to address the post-converted problem,and fur-thermore prove that the proposed algorithm prevent the collusion of SBSs if the number of SBSs is less than a threshold.Finally,simulation results show that the proposed mechanism can guarantee the truthfulness of all participants in the content delivery process and improve the efficiency of content delivery.3.We propose a content updating strategy considering users' mobility.Firstly,we use the mankind mobility model and the hidden Markov model to acquire the parameter sets of user requirement transition probability and caching state transition probability.Subsequently,based on the parameter sets,the content update problem is formulated by the flow model in graph theory,while the objective is considered as maximizing the average hit rate of local SBSs' cache,with multiple constraints on and between SBSs taken into ac-count.Next,we prove that the original problem can be resolved into severalfinite simple max-flow problems.Then a content update scheme is designed by including both the content demand prediction and source finding algorithm,and the advanced preflow-push algorithm.Finally,simulation results verify the effectiveness of the proposed scheme in improving the performance of the local hit rate.