Research On Intelligent Video Transmission Technology On The Network Based On Edge Computing

With the diversification of video services and the upgrading of network infrastructure,video users have put forward higher requirements for viewing experience.However,based on existing network architectures,the rapid growth of video traffic has brought great challenges to the delivery of network content.As an important optimization method,dynamic adaptive video streaming maximizes bandwidth utilization through chunk-based transmission,multiple bitrate versions and other mechanisms to ensure user experience.However,it also has some problems such as the high cost of storage and the difficulty of bitrate selection.The emergence of edge computing alleviates the traffic pressure of video origin servers and also provides more chances for optimizing the adaptive video streaming.Effectively exploiting the storage,computing,bandwidth and other resources of edge servers can better reduce redundant transmission,fight against network jitters and accelerate the response to users.Therefore,it is of great significance to comprehensively consider the characteristics of dynamic adaptive video streaming and edge computing while utilizing the network resources and improving the user experience.Based on the paradigm of edge computing services,this paper is to optimize the user experience of network video from multiple perspectives,e.g.,architecture design of the video transmission,requirement modeling of the video users and resource scheduling of the edge computing.The main work and contributions are concluded as follows.·Through analyzing the transmission characteristics of dynamic adaptive video and the features of available resources from edge computing,we design the network video transmission architecture based on the edge computing.The users send video requests according to network information and client status.Then the origin server sends back corresponding content according to the requests.With the computing,storage,forwarding and other capabilities of the edge server,it can achieve the goals such as improvement of video quality,optimization of content cache and acceleration of request response.·Through analyzing the video service requirements under the mobile network,we design an edge selection scheme for optimizing user experience.For multi-access edge computing in the mobile network,content servers can be amounted in base stations.According to the handover of base stations,the cache status of edge servers and other information,the edge selection strategy is designed to find a proper edge server for the user and finally achieve the goal of optimizing the user experience.·Through analyzing the characteristics of chunk-based video requests,we design a single-access edge optimization strategy of responding to user requests.When watching adaptive video,an edge server will provide acceleration services for the covered users.The edge can cache popular content to reduce redundant transmission while it can also prefetch some content to fight against network jitters.According to the probability of bitrate switching and the chunk classification based on structural similarity,the gain model of user experience is built.The intelligent caching and prefetching decisions for the edge are implemented accordingly.·Through analyzing user differences and network dynamics,we design an optimization scheme of user experience based on multiple edge servers.Content providers serve the users in different regions and need to deploy multiple edge servers.According to the available storage and bandwidth capabilities of the neighbour edge servers,these edge nodes can cooperate with each other to maximize the utilization of their limited resources and finally improve the user experience under dynamic network conditions.