北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2013, Vol. 36 ›› Issue (6): 74-78.doi: 10.13190/j.jbupt.2013.06.016

• 论文 • 上一篇    下一篇

异质业务到达下共享服务器系统服务模式研究

迟学芬, 石佳琳, 张嘉盛, 石文孝   

  1. 吉林大学 通信工程学院, 长春 130012
  • 收稿日期:2012-12-25 出版日期:2013-12-31 发布日期:2013-10-08
  • 作者简介:迟学芬(1962—),女,教授,博士生导师,E-mail:chixf@jlu.edu.cn.
  • 基金资助:

    国家自然科学基金项目(61071074)

Service Mode of Server Sharing System with Heterogeneous Services Arrival

CHI Xue-fen, SHI Jia-lin, ZHANG Jia-sheng, SHI Wen-xiao   

  1. Department of Communication Engineering, Jilin University, Changchun 130012, China
  • Received:2012-12-25 Online:2013-12-31 Published:2013-10-08

摘要:

针对长期演进(LTE)的包交换和可变速率服务特点,将LTE网络对异质业务的服务过程抽象为可变速率服务器的包级排队系统,以研究异质业务共享服务器的方法. 把异质业务粗略划分为人对人(H2H)、机器对机器(M2M)2类,根据网络技术和业务特性,分别用MMPP/M/1/KM+ARQ、M/M/1/KH两种队列模型建模网络对M2M、H2H两种业务到达服务. 2类业务共享同一服务器,服务过程受H2H队长阈值控制,平均服务速率遵循指数或线性变化,总服务速率保持不变. 构建3维马尔可夫模型,推导服务器共享系统状态的生成矩阵,求解3维马尔可夫链. 仿真实验给出2种服务模式下的系统性能,同时展示了不同突发特性的M2M到达对系统性能的影响.

关键词: 长期演进, 可变服务速率, 服务器共享, 机器对机器

Abstract:

A packet-level variable service rate queuing system is established to model the process that the long-term evolution (LTE) network works for heterogeneous services traffic. The system reflects packets switching and variable service rate of LTE. The main research target is to explore the way that heterogeneous services share a common server. Heterogeneous services can be divided into two types, human to human (H2H) services and machine to machine (M2M) services. Based on network of LTE and its service characteristics, two queuing models MMPP/M/1/KM+ARQ and M/M/1/KH are established to depict different LTE network services under M2M and H2H traffic arrival respectively. Two kinds of services sharing the same server and the service process are controlled by the queue length threshold of H2H queue. The average service rate changes exponentially or in a linear fashion, while the total service rate remains constant. By deducing three-dimensional Markov chain and generator matrix of queuing system, the queuing model is solved. Algorithm simulation shows the system performance in several circumstances, including different service modes and different burst features of M2M arrival.

Key words: long-term evolution, variable service rate, processor sharing, machine to machine

中图分类号: