Research On Dynamic Adaptive Streaming Media Transmission Optimization Strategy Base On QoE

With the widespread popularity of mobile smart devices(tablet computers,smart phones,etc.)and the emergence of a large number of video applications(video-ondemand/live broadcast,online conferences,etc.),the proportion of video traffic in the total Internet traffic has gradually increased,and video users have increased sharply.How to guarantee and improve the quality of user experience(QoE)of video viewers is a major challenge for streaming media transmission systems.In this thesis,we study the rate adaptive algorithm deployed in the video client to improve the quality of video playback and improve the QoE of users based on the international standard of dynamic adaptive streaming media transmission and a video cache replacement and transmission optimization strategy deployed in the network transport layer.We analyzed the main characteristics of dynamic adaptive streaming media technology based on HTTP,introduced the existing enterprise solutions of adaptive streaming media technology and the international standard MPEG-DASH of dynamic adaptive streaming media technology and its related technologies.We studied and analyzed the advantages and disadvantages of existing bitrate adaptive algorithms,and proposed a video bitrate adaptive algorithm based on cahe compensation.The proposed algorithm analyzed the bandwidth fluctuation and obtained the estimated bandwidth according to the recent downloaded fragments,according to the estimated bandwidth and current bitrate level,two dynamic buffer threshold were set to control rate switching to higher level and buffer time was consumed,or to control rate switching down gradually and buffer time was accumulated,a consumption accumulation buffer state loop was formed in video buffer.The experimental results show that the proposed algorithm has high bandwidth utilization,switching smoothness and switching stability in dynamic network environment,which can effectively improve the quality of user experience.Secondly,we proposed a video bitrate adaptive algorithm based on content selection.The proposed algorithm is based on the assumption that different video content can bring different viewing experience to users.The purpose of the algorithm is to enable clients to selectively consume redundant cache to improve playback bitrate when requesting video segments containing high-quality video content.Firstly,the estimated available bandwidth at the next moment was estimated.Secondly,The process of subsequent video requests was simulated based on the calculated estimated bandwidth and video cache.In each simulation process,the strategy of consuming cache to improve the bitrate was adopted on the basis of preserving part of video cache.According to the set provided by the video server that identifies the content level of all video segments,video segments were selected as higher bitrate version in content level order from high to low.The optimal bitrate selection distribution of the remaining video segments requested by the estimated bandwidth was simulated.According to the simulated optimal bitrate selection distribution,the video bitrate level that should be selected at the current time can be determined.The experimental results show that the proposed algorithm can guarantee the highest video average bitrate for the video segments with the highest content quality under the assumption that the video content has experience differences,it can effectively improve the user experience quality.Finally,we proposed a user-oriented video cache replacement and transmission optimization strategy.The proposed strategy comprehensively analyzed the characteristics of video users requesting video in streaming media transmission system,divided the video slice set of a single video content into multiple video blocks,the corresponding video block content is cached in the edge server according to the order in which the user requested video fragmentation.When the video user was about to request the current video block,the active caching strategy was adopted to obtain the adjacent video blocks from the adjacent edge server or video cloud server to meet the subsequent requests of video user.When the storage capacity of the edge server reached the upper limit,the popularity change was estimated according to the historical requests of the video,and the idle video block with low popularity was selected for replacement.The experimental results show that the proposed strategy makes more full use of the storage resources of the edge server,improves the user request hit rate,reduces the user request response delay,and can effectively improve the user QoE.