论文部分内容阅读
流媒体用户通常有“预览”行为,对腾讯视频系统的测量结果显示即使在网络传输良好的情况下,约50%用户在离观看结束不到一半时就离开了.增加用户下载带宽在提高用户体验质量的同时也会带来服务器带宽浪费.本文主要贡献是:首先,建立了流媒体带宽分配的数学模型以刻画下载速率对带宽浪费率、单位时间服务用户数和QoE的影响;其次,发现增大下载速率,用户体验质量提升较大而带宽浪费较小(对腾讯视频系统的计算显示当下载速率是播放速率的两倍时带宽浪费率仅为25%),为适当增大下载速率提供了合理性依据;最后,给出了求解总用户体验质量最大化问题的最优下载速率的基本步骤.
Streaming media users often have “preview” behavior and measurements on Tencent’s video system show that about 50% of users left even less than half the time they watched, even with good network traffic. Improve the quality of user experience will also bring server bandwidth waste.The main contributions of this paper are as follows: Firstly, a mathematical model of streaming media bandwidth allocation is established to describe the effect of download rate on bandwidth waste rate, number of service users per unit time and QoE; , And found that increasing download speed, improving user experience quality and small bandwidth waste (calculation of Tencent’s video system shows that the bandwidth wasted rate is only 25% when the download rate is twice the play rate) The rationality is provided by the rate. Finally, the basic steps of solving the optimal download rate of the total quality of user experience are given.