北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2010, Vol. 33 ›› Issue (5): 22-27.doi: 10.13190/jbupt.201005.22.zhangx

• 论文 • 上一篇    下一篇

无线网络业务行为——特性分析与行为建模

张兴,颜志,肖静,蔡雯琦,王文博   

  1. 泛网无线通信教育部重点实验室(北京邮电大学)
  • 收稿日期:2009-11-01 修回日期:2009-12-02 出版日期:2010-08-28 发布日期:2010-06-15
  • 通讯作者: 张兴 E-mail:hszhang@bupt.edu.cn
  • 基金资助:

    国家级.国家重点基础研究发展计划项目;部级.教育部高等学校博士点基金项目

Traffic Behavior in Wireless Networks—Characteristics  Analysis and Behavior Modeling

  • Received:2009-11-01 Revised:2009-12-02 Online:2010-08-28 Published:2010-06-15

摘要:

为了研究移动无线网络中的业务行为特征,提出了一种连续时间半马尔科夫业务行为模型(CTSMM). 基于国内某地无线蜂窝网络中采集得到的业务数据,进行了深入的业务行为特性挖掘分析;从业务会话级层面分别针对单小区和多小区网络环境进行了会话到达、会话持续时间和会话量等重要特征的研究分析,并以此为基础提出CTSMM,该模型可以较好地反映无线网络中的业务行为特性. 同时基于实际采集的业务数据对所提业务行为模型进行了大量比较验证,结果表明,该行为模型基本符合无线网络中业务行为规律,可以基于该业务行为模型进行面向业务行为的无线网络设计、资源规划与优化、频谱动态利用等.

关键词: 业务行为, 行为模型; 自相似, 特性分析, 半马尔科夫模型

Abstract:

To study the characteristics of traffic behavior in wireless mobile networks, a continuoustime semiMarkov process model (CTSMM) is proposed. Based on the large amount of traffics gathered from real wireless networks in China, the characteristics of traffic behavior are analyzed deeply. Specifically, the session arrival, the session duration and the session traffic size for single cell and multiple cells environments are studied. CTSMM is proposed to characterize traffic behavior. Meanwhile, based on the collected traffic data, the proposed model is verified for reflecting the properties of the traffic behavior in wireless networks. Using this model, the wireless network design, the resource management optimization, and the dynamic wireless spectrum utilization, are well performed.

Key words: traffic behavior, behavior model, self similarity, characteristics analysis, semi Markov model