Research On Multimedia Content Distribution Algorithm With QoS Guarantee

In recent years,with the rapid development of Internet and multimedia technology,it has become a common phenomenon that users watching various video resources online.How to satisfy people's growing demand for video resources and how to enable users to get better video viewing experience.They have become the current research hotspot and have attracted widespread attention from many researchers.Cloud video delivery networks(CVDNs)apply cloud computing to video delivery networks,and new network architecture has been formed.The method makes full use of the advantages of VDNs and cloud computing,with lower cost to provide users with high-quality video content delivery services.CVDNs duplicating or caching videos that are accessed by users frequently to cloud data center,it effectively reduces the load of the source server and alleviates the congestion phenomenon of the network,which has been widely recognized in the industry.However,there are many problems in the existing CVDNs.First,the existing CVDNs do not guarantee the quality of service(QoS).In real life,most of the objects distributed by CVDNs are online video.Because of the diverse requests of users,the requirement of video delay is high,and higher QoS is needed.Traditional CVDNs simply provides video content delivery service,without considering the diversity of user requests and users' requirements for QoS.Secondly,the existing CVDNs delivery mechanism is less concerned with video content classification and user collaboration,and the high cost cannot attract users.They have become a bottleneck restricting the development of CVDNs technology.In view of the above questions,this article firstly proposes a new CVDNs network topology which support link cutting and priority oriented.This method can improve the QoS and the revenue of CVDNs provider.Secondly,this paper presents a CVDNs based on content perception and group buying strategies.This method can effectively reduce the cost of users.The main work and innovation of this article are as follows:1.Aiming at the problems that the CVDNs cannot meet users' diversity needs and QoS,a CVDNs network topology supporting link cutting and priority oriented is proposed.First,a method of supporting link cutting is added to the CVDNs to allow video streams to be distributed to the user through different links.This method can increase the network topology acceptance rate of CVDNs,and increase the QoS and the number of serviceable users.Secondly,a priority-oriented approach is used in CVDNs to divide users into two categories: ordinary users and paid users.This method makes the paid user with a higher priority and has a priority in the competitive use of the link.Finally,a complete objective function and constraint conditions are proposed,and a priority-oriented CVDNs problem is formulated as a mix integer programming(MIP)problem.2.To solve the problem of high cost for users to use CVDNs,a content-aware and group-buying(CG)method is proposed.Firstly,we use a content aware method to classify the content of video,and allow the users to purchase them by forming coalitions.In the end,wedefine cost formula,user purchase quantity constraint and link constraint to formulate the CG problem as a linear programming problem which can be solved by GLPK tools.Experimental results show that our methods can improve the acceptance rate of the CVDNs network topology,improve the QoS,and reduce the cost of users'.