摘要
该文提出了一种基于信誉的P2P网络信任管理模型。在P2P网络中,由于不存在中心节点,需要根据节点的行为来判断其是否可信。通过引入时间衰减算法,解决了对行为评估,本地信任值以及推荐的时间相关性问题。通过对近期表现和长期表现的对比,给出对该节点未来表现的主观预期,能够对节点异动做出反应。利用DHT资源发现算法,有效地降低了网络消耗,并使模型具有可扩展性。模型能够有效地提高P2P网络的交易成功率。
This paper presents a novel decentralized trust management model with reputation.In P2P（peer to peer） network,there is no trusted authority.Trust relations between peers should be established with peer＇s behaviors.There are three main contributions in this paper.Towards utilizing time self-decay function,the time-related problem is resolved.Also through comparing the nearly scores and general scores,the problem of servent＇s subjective expect can be resolved.After using the DHTs,bandwidth cost can be reduced and salability can be obtained.This model can promote the business succeed rate in P2P network efficiently.
出处
《电子与信息学报》
EI
CSCD
北大核心
2009年第11期2786-2790,共5页
Journal of Electronics & Information Technology
基金
国家973计划项目(2007CB310704)
国家自然科学基金(60821001
90718001)
北京市自然科学基金(4073037)资助课题
关键词
网络信任管理
P2P网络
时间衰减
主观预期
分布式哈希表
Network trust management
P2P network
Time self-decay
Subjective expect
Distributed hash table
