Research And Design Of Steaming Media Server Optimization Strategy Over HTTP2.0

With the development of network video services,MPEG-DASH,which combines progressive streaming and real-time streaming features,has become the mainstream technology.MPEG-DASH based on HTTP1.1 has a large startup delay in the video buffering stage,and the video average bit rate will be greatly shaken as the network environment changes,which affects the user experience to some extent.The emergence of HTTP2.0 provides new features such as binary framing,server push,and multiplexing.However,for these features HTTP 2.0 does not give a specific optimization method in the specific scenario of streaming media transmission.To this end,this paper proposes four optimization strategies for MPEG-DASH streaming media server based on these new features of HTTP2.0,including server segmentation adaptive push,client buffer threshold threshold adaptive adjustment,client prediction network bandwidth,and calculation video frame weights are used for flow control.These optimization strategies have improved the startup delay and average bit rate fluctuations caused by improper push or insufficient buffering from different angles.Based on these four optimization strategies,this paper designs an adaptive transmission scheme based on MPEG-DASH,which includes the adaptive algorithm of the number of push segments and buffer start threshold,bandwidth prediction algorithm and video frame weight algorithm.Based on the scheme and supporting algorithms,the MPEG-DASH on-demand system is implemented.The comparison experiments verify the improvement of the average bit rate stability and startup delay.