Computer Engineering and Applications ›› 2007, Vol. 43 ›› Issue (10): 159-161.

• 网络、通信与安全 • Previous Articles     Next Articles

Study the Congestion Control of Large Quantity

Wei Sun   

  • Received:2006-05-10 Revised:1900-01-01 Online:2007-04-01 Published:2007-04-01
  • Contact: Wei Sun

大数据量实时视频流传输拥塞控制研究

孙为 王伟   

  1. 兰州理工大学 现代网络教育学院 哈尔滨工业大学(威海)计算机学院
  • 通讯作者: 孙为

Abstract: There is broad application perspective of large quantity video stream transmitting system and many problems in transmitting the large quantity and high bandwidth video stream.Netowrk overload equilibrium and the quality of video playback all affect the security of network and the application of video stream transmission system.On analysing status of the network overlaod to forecast the ratio of dropped packet by Kalman Filter and smoothing the sending rate by setup the buffer ,a TCP-Friendly and based RTCP feedback congestion control mechanism fitted large quantitiy video stream transmission is provided based on studying the TCP-Friendly congestion control .In the end, the character of TCP-Friendly and the quanlity of playback of this congestion control mechanism is tested and the result is analysed by network simulation.

Key words: large quantity, real-time, congestion control, tcp-friendly

摘要: 大数据量实时视频流传输系统有广泛的应用前景。大数据量、高带宽要求的视频流传输中存在许多问题。网络的负载均衡、视频的回放质量都将严重影响整个网络的安全以及传输系统的应用。通过对丢包率进行卡尔曼过滤(Kalman Filter)分析预测网络负载状况、结合接收缓冲区大小设置来平滑发送速率,在分析TCP友好拥塞控制的基础上,提出了一种基于RTCP反馈的TCP友好(TCP-Friendly)对该机制的TCP友好性、视频的回放质量进行了试验和结果分析。

关键词: 大数据量, 实时, 拥塞控制, TCP友好