Research On Coupled Video Distribution Method For Cost Optimization

In recent years,with the rapid development of bandwidth wireless access technology and intelligent terminal technology,Internet service content has become increasingly diversified.Videos have become a convenient way to deliver information.Faced with massive and abundant video content,a large number of concurrent service requests may cause congestion and bottlenecks in the Internet.In order to alleviate the conflict between the high demand of bandwidth and the response time of network services,researchers have proposed Video Delivery Networks(VDNs)technology to reduce response delay and bandwidth consumption effectively.VDNs deploys edge servers throughout the network and redirects user requests to the nearest edge servers with a comprehensive consideration of network load,response time and internet traffic,while improving the users' quality of experience.However,VDNs also have some limitations.First,because of the limited capacity of surrogate servers and the constraint of backbone network bandwidth,the origin server distribute videos to surrogate servers in real time will lead to overload of servers and long response time.What's more,the existing algorithms of video delivery are optimized for performance indicators such as delay,bandwidth and content popularity independently,resulting increased cost of delivery.Finally,the existed scheme of push and pull does not consider the mobility of users and it dissatisfied personalized requirement of mobile users for online video in dynamic network scenario.To overcome these problems,we propose a cost optimization method for coupled video delivery,which not only considers bandwidth consumption,delay performance and personalization demand comprehensively,but also meets the requirement of mobile users for online video in dynamic networks.The proposed method can minimize the video delivery cost as well.The main work and innovation of this article are as follows:1.Aiming at the conflict between bandwidth consumption and response time in traditional video delivery methods,we propose a coupled video delivery model which combines coupled proactive delivery and coupled reactive delivery method to minimize delivery cost in terms of bandwidth consumption,delay performance,personalized demand and construct a multicast delivery tree for the video delivery.The problem of minimizing coupled delivery cost under constraint conditions is proposed.Since solving this problem is NP-hard,we develop a Heuristic Multicast Delivery Tree(HMDT)algorithm to provide more optimized approximate cost.2.Since the existed push-pull delivery system have not considered the dynamic requests of mobile users,we truncate the continuous process for video delivery into a sequence of static graphs.Then,the dynamic demand model for users is constructed according to the sequence.The interest of users for video will change as time flying.Hence we analyze the interest attenuation of user over time and formulate the function of interest attenuation to track the change of user interest so as to minimize the cost for video delivery.3.A simulation system for the cost optimized coupled video delivery is constructed in this paper.This system includes the generation of simulation network topology,the implementation of the heuristic multicast delivery tree and the visual representation of simulation delivery path.Meanwhile,the DOT language is used to describe the network topology and the directed network topology graph is drawn to make the simulation results clear at a glance in this paper.The extensive simulations show that our proposed model can yield strategies with smaller expected cost,and the delivery cost is inversely proportional to the attenuation of user interest.