›› 2013, Vol. 26 ›› Issue (11): 26-.

• 论文 • 上一篇    下一篇

基于Bloom Filter的云内容路由优化

俞冶,金逸超,尹丽英   

  1. (1.嘉兴市广播电视集团 总工程师办公室,浙江 嘉兴 314001;2.南洋理工大学,新加坡 639798;3.西安邮电大学 经济管理学院,陕西 西安 710121;4.西安电子科技大学 经济管理学院,陕西 西安 710071)
  • 出版日期:2013-11-15 发布日期:2013-11-19
  • 作者简介:俞冶(1959—),女,高级工程师。研究方向:云计算与新媒体。E-mail:njyuye@hotmail.com

Optimization of Cloud Content Routing Based on the Bloom Filter

YU Ye,JIN Yichao,YIN Liying   

  1. (1.Chief Engineer's Office,Jiaxing Broadcasting Station,Jiaxing 314000,China;2.Nanyang Technological University,Singapore 639798,Singapore;3.School of Economics and Management,Xi'an University of Posts and Telecommunications,Xi'an 710121,China;4.School of Economics and Management,Xidian University,Xi'an 710071,China)
  • Online:2013-11-15 Published:2013-11-19

摘要:

随着国家三网融合战略的进一步推进,传统的“广播式”媒体内容分发模式已无法满足用户日益增长的双向“互动式”业务需求。利用云计算技术构建媒体云内容分发网络,实现媒体内容的弹性部署、高效分发来解决该问题,已成为下一代广电网络(NGB)的核心内容。在构建媒体云网络过程中,如何进一步改善用户的响应时延是必须考虑的因素。文中在此背景下,提出利用全局Bloom Filter优化媒体云网络中的内容路由。通过两种优化路由设计,使用户的平均响应时延得到有效下降。并采用排队网络对传统以及优化后的路由策略进行理论建模,使用OMNeT++网络仿真器对提出的路由策略进行仿真。其结论与仿真结果一致性良好。结果显示,优化后的路由策略,在不同的场景下,最多可节省65.2%的平均响应时延。

关键词: 媒体云, 内容路由, 响应时延, Bloom Filter, 排队网络

Abstract:

With the rapid advances on Tri-networks integration in China,the traditional broadcasting content service mode can no longer meet the increasing requirements on interactive content service mode.The key to this problem is an elastic and efficient cloud-based media network by leveraging cloud computing technology.This solution has been considered as the key of the Next Generation Broadcasting Network (NGB).One of the critical design objectives is to further improve the mean response time in this cloud-based media network.This paper proposes the use of a global Bloom Filter to improve the content routing schemes.Specifically,we design two improved routing schemes (i.e.,parallel scheme and cut-through scheme) after introducing Bloom Filter,to reduce the mean response time.We formulate both the traditional and the proposed routing schemes as queuing models,and simulate them by using OMNeT++.The analytical results agree well with the numerical ones.The results show the proposed schemes can save up to 65.2% mean response time in different networking scenarios.

Key words: cloud based media network;content routing;response time;Bloom Filter;queuing model

中图分类号: 

  • TP393