传感器网络中基于帕累托最优效用的包转发研究  

Pareto Optimal Utility Based Packet Forwarding in Sensor Networks

作　　者：阎毓杰[1] 王殊[1] 

机构地区：[1]华中科技大学电子与信息工程系,武汉430074

出　　处：《计算机科学》2007年第5期32-35,80,共5页Computer Science

基　　金：国家自然科学基金(No.60472015);国家985二期工程"基于网格的高性能计算与复杂系统仿真平台建设"的资助

摘　　要：本文运用博弈论的观点和方法来解决传感器网络中的包转发问题。为传感器网络建立了包转发模型,分析了节点参与包转发会话所获得的帕累托最优效用,提出了基于帕累托最优效用的包转发算法POUPF,并证明了该算法能够建立纳什均衡以保证每个节点都获得帕累托最优效用。仿真结果表明:POUPF能够有效促进节点自发合作,确保了每个节点获得帕累托最优效用;任何偏离POUPF节点的包转发行为都会导致所有节点效用的下降。This paper focuses on the packet forwarding problem in sensor networks and provides a solution with the method of Game theory. It designs a system model for packet forwarding and derives the node＇s Pareto optimal utility. Further, it proposes a POUPF algorithm which propels the nodes to operate at the Pareto optimal utility and constitutes a Nash Equilibrium. The simulation results verify that POUPF assures the nodes to achieve the Pareto optimal utility by stimulating cooperation; deviation at any one node will decrease utilities of all related nodes.

关 键 词：传感器网络 包转发 博弈论 纳什均衡 帕累托最优效用 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]