The Research Of Video Streaming Content Distribution Strategy Based On CDN Network

With the rapid development of the Internet,the continuous upgrading of network infrastructure,streaming media technology is advancing day by day,Today,streaming media has become the main application in the Internet.A large number of online video,making the Internet bear more and more pressure.Content Delivery Network(CDN)has been adopted by many Internet companies.Compared with the traditional network architecture,CDN can improve the utilization of network bandwidth resources,especially the network bandwidth resources of the Backbone Network,improve the quality of network transmission,and can effectively reduce the load of the source server and improve the user's Quality of Experience(QoE).During the first two decades,the studies of CDN have been mainly focused on content distribution strategy,while the research on content pre-distribution has been ignored.At present,the basis of pre-distribution of content is only popularity or the user's interest,and the calculation method of user's interest degree is generally more rough and simple,there is a large deviation when compared with the user's actual situation,which will lead to the declination of pre-distributed content hit rate,it cannot better serve the gradual enhancement of the streaming media applications business.In order to solve the above problems,this paper proposes a new CDN streaming media content distribution strategy through researching the user interest degree in CDN.This strategy avoids the influence of non-interest factors such as system recommendation,user error operation and web navigation on the calculation of interest degree when constructing user interest model,which makes the calculated user interest more accurate and reasonable.Simulation experiments show that the strategy of streaming media pre-distribution based on this improved user interest model can effectively improve the hit rate of user requests.In practical application,the user's interest is not static,there is interest drift phenomenon,the user's interest in a thing will change with the scene or time to change.Time window method and forgetting function method is the solution to interest drift presently.Based on the forgetting function method,this paper proposes a dynamic user interest degree model.In this user interest model,as time goes on,using the natural attenuation function to modify the user interest,making the calculated user interest with the user's real interest as possible fitting.Simulation results show that the model can effectively improve the accuracy of content pre-distribution under the guidance of user interest model.In this paper,CDNsim simulation software is used to simulate the CDN network environment.The experimental data are derived from the real user access log of a network television station in Hong Kong.