无线传感器网络路由中合作性重复博弈模型的研究  

Repeated-Game Theory of Cooperative Model in Wireless Sensor Network Routing

作　　者：刘群[1] 先兴平[1] 郭松涛[2] 吴涛[1] 

机构地区：[1]重庆邮电大学计算机学院,重庆400065 [2]重庆大学计算机学院,重庆400010

出　　处：《传感技术学报》2010年第9期1322-1327,共6页Chinese Journal of Sensors and Actuators

基　　金：国家自然科学基金资助(60903213);重庆市自然科学基金资助(CSTC2007BB2386);重庆市教委基础研究项目资助(KJ080510)

摘　　要：无线传感器网络中,节点能耗、路径可靠度以及节点的死亡时间是传感器网络路由需要考虑的关键因素。为了提高能量利用率和传感器网络收益,在节点理性且自私的条件下,运用博弈论方法提出了一种基于节点合作的数据包发送/转发的重复博弈模型,设计了一个与路径连通度和节点能量消耗有关的收益函数,采用惩罚机制使重复博弈模型存在子博弈精炼纳什均衡,降低了自私节点背叛的可能性。实验结果表明:采用惩罚机制的重复博弈能够提高网络的收益,同时也提高了网络吞吐量,任何自私节点的不合作行为都导致节点的能量浪费和节点的整体收益下降。Energy consumption, path reliability and node’s life time are the key issues in wireless sensor network routing. We present a repeated game theoretic model based on cooperative packet forwarding under the conditions of selfish and rational nodes for improving energy efficiency and sensor networks payoff, we also design a payoff function on path reliability and energy consumption. Using the punishment mechanism, the repeated game model can propel a Nash Equilibrium and decrease the defection possibility of selfish nodes. The simulation results show that the repeated game model enhances the network payoff and increases the network throughput, and also show that the defection at any selfish node will degrade the network payoff and cause energy waste.

关 键 词：无线传感器网络 路由 重复博弈 纳什均衡 自私性 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]