中国邮电高校学报(英文) ›› 2012, Vol. 19 ›› Issue (3): 60-67.doi: 10.1016/S1005-8885(11)60266-5

• Networks • 上一篇    下一篇

Evolved data scheduling scheme based on layered P2P streaming in mobile network

夏海轮 WANG Ning ZENG Zhi-min   

  1. Beijing University of Posts and Telecommunications
  • 收稿日期:2011-11-30 修回日期:2012-03-23 出版日期:2012-06-30 发布日期:2012-06-08
  • 通讯作者: 夏海轮 E-mail:xiahailun@bupt.edu.cn
  • 基金资助:

    中央高校基本科研业务费专项资金

Evolved data scheduling scheme based on layered P2P streaming in mobile network

XIA Hai-Lun WANG Ning ZENG Zhi-min   

  1. Beijing University of Posts and Telecommunications
  • Received:2011-11-30 Revised:2012-03-23 Online:2012-06-30 Published:2012-06-08
  • Contact: XIA Hai-Lun E-mail:xiahailun@bupt.edu.cn

摘要:

为了缓解分层移动P2P流媒体服务中由复杂的无线网络环境和不稳定的用户行为带来的负面效果,本文提出了一种演进的分层流媒体数据调度算法E-LP2P。为保证基本的流媒体流畅播放,根据不同数据在播放过程中的重要性来调度基层流媒体数据;根据数据所处的位置及服务节点特征来调度增强层流媒体数据;为应对由无线带宽的波动所引起的层抖动,使用抖动防止机制在数据调度过程中根据网络环境而动态调整订阅层数。仿真结果表明本算法在吞吐量、层到达率、服务器负载及无用包率等方面都提供了较好的业务QoS保障。

关键词:

layered streaming, P2P, mobile network, data scheduling

Abstract:

To relieve the negative effect brought by the intricate wireless network environment and unstable user behavior in layered mobile P2P streaming service, an Evolved Layered P2P (E-LP2P) data scheduling scheme in the process of service delivery is introduced in this paper. The data in base layer is scheduled according to its importance in streaming play to guarantee the basic play of streaming. The data in enhancement layer is scheduled according to the characters of streaming data, including its position and amount in server peer set in a multiple tied way towards the data in enhancement layer. To cope with the layer jitter caused by the fluctuation of bandwidth, jitter prevent mechanism is used to adjust the highest layer dynamically during the process of data scheduling. Simulation results show that the E-LP2P can provide good QoS performance in terms of throughput, layer delivery ratio, server load and useless packet ratio.

Key words:

layered streaming, P2P, mobile network, data scheduling