Computer Engineering and Applications ›› 2006, Vol. 42 ›› Issue (2期): 24-27.

• 博士论坛 • Previous Articles     Next Articles

Research on a Flow Control Algorithm based on Multi-followers Stackelberg Game Theory

Jun Tao,Qiang Wu,Qingliang Wu   

  1. 东南大学计算机系
  • Received:2005-07-27 Revised:1900-01-01 Online:2006-01-11 Published:2006-01-11
  • Contact: Jun Tao



  1. 东南大学计算机系
  • 通讯作者: 陶军 seutao seutao

Abstract: In the current network, the end users usually play different roles. Some users decide the flow control strategies for the whole network and others can only follow these strategies. So we cannot simply apply the non-cooperative game theory to study the greedy behaviors of non-cooperative users in Internet. After one-shot game model and the solutions for flow control game equilibrium are discussed, we study the behaviors of flow and congestion control based on multi-followers Stackelberg game model further. Subsequently, a flow control model based on Stackelberg game with multi-followers is presented. The existence of the Stackelberg Equilibrium is proved. Moreover, the solution of leader and multi-followers on Stackelberg equilibrium point is deduced. Finally, a flow control algorithm based on Stackelberg game with multi-followers, FCSM, is presented and the executing process of FCSM algorithm is discussed.

Key words: Stackelberg Equilibrium, Stackelberg game with multi-followers, utility, flow control

摘要: 由于现实网络中,端系统的角色往往并不平等,存在某些端系统制定整个网络的流速控制策略,其他端系统只能跟随这些策略,不能简单地使用非合作博弈理论研究Internet中非合作端系统的贪婪行为。本文在讨论了一次博弈模型和流速控制博弈均衡的解析解的基础上,对多跟随者Stackelberg博弈模型的流速与拥塞控制行为进行了深入研究。提出了多跟随者Stackelberg博弈流速控制模型,论证了Stackelberg均衡的存在性,并推导出在Stackelberg均衡点上领导者和多跟随者流速率的解析解。最后,本文设计了一种基于多跟随者Stackelberg博弈的流速控制算法(FCSM),并讨论了FCSM算法执行的过程。

关键词: Stackelberg均衡, 多跟随者Stackelberg博弈, 效用, 流速控制