Research On Buffer-based Rate Adaptation Algorithms For Dynamic Adaptive Streaming Over HTTP

Streaming media services often transmit huge amounts of data.In order to use the limited and constantly fluctuating bandwidth to transmit high-quality audio and video data,the MPEG organization and 3GPP jointly formulated the MPEG-DASH standard.The core part of the MPEG-DASH streaming media transmission system is the client-side adaptive algorithm,and the quality of the algorithm will affect the user's experience quality.How to design an MPEG-DASH client adaptive algorithm to improve user experience quality has attracted the attention of many domestic and foreign scholars.Aiming at the scenario where a single client plays video under a wireless network in a DASH streaming media system,in order to utilize limited bandwidth to transmit highquality audio and video data to improve user experience quality,a client-side adaptive algorithm based on buffer status is proposed.Firstly,the MPEG-DASH system framework and technical specifications are studied in depth,and then the existing classic MPEG-DASH rate adaptive algorithm is analyzed in detail,and the shortcomings of the traditional adaptive algorithm are pointed out.Then it analyzes the MPEG-DASH experience quality evaluation indicators,and specifically introduces four important indicators: video rebuffering time and times,average video bitrate,bitrate switching times,and startup delay.Through analysis,it is concluded that there is a restrictive relationship between these indicators,and the design goals of the client-side adaptive algorithm are proposed.Then analyze the dynamic characteristics of the buffer.On this basis,a buffer-based controller is proposed to realize DASH client bitrate adaptation,make bitrate decisions through buffer occupancy rate and buffer change rate,and introduce bitrate control factor to reduce unnecessary bitrate switching,so as to achieve balance among multiple DASH experience quality evaluation indicators.Simulation experiments are carried out to analyze the performance of the proposed algorithm in this paper.The experimental results show that the proposed algorithm not only reduces the occurrence of rebuffering events,but also selects a higher bitrate of video segments,thus improving the quality of user experience.It has certain practical value to improve the user experience quality of MPEG-DASH streaming media service.