计算机应用 ›› 2011, Vol. 31 ›› Issue (01): 139-142.

• 信息安全 • 上一篇    下一篇

基于P2P网络的动态模糊综合信任模型

饶屾,王勇   

  1. 桂林电子科技大学
  • 收稿日期:2010-06-28 修回日期:2010-08-15 发布日期:2011-01-12 出版日期:2011-01-01
  • 通讯作者: 饶屾

Dynamic fuzzy comprehensive trust model based on P2P network

  • Received:2010-06-28 Revised:2010-08-15 Online:2011-01-12 Published:2011-01-01
  • Contact: Rao Shen

摘要: 针对现有的P2P网络信任模型在聚合节点信任值时对节点行为的差异性与动态性考虑不足,提出了一种动态的模糊综合信任模型——DFCTrust,即在静态模糊综合评价的基础上加入时间衰减因子及波动惩罚因子。首先,该模型通过静态模糊综合评价的方法来计算每次交易的满意度得分,由于在计算过程中加入了交易上下文的评价因素,所以可以很好地避免恶意节点在小交易上诚信在大交易上欺骗的恶意行为;其次,在综合计算节点信任度时,引入了时间衰减因子和波动惩罚因子,这样不仅可以避免与不活跃节点交易所造成的失败,而且还可以有效地抵御节点采用周期性振荡欺骗的恶意行为。仿真实验表明,DFCTrust具有更好的动态适应能力,能够有效地处理动态恶意节点策略性的行为改变,从而提高交易的成功率。

关键词: 对等网络, 动态综合评价, 信任模型, 模糊综合评价

Abstract: According to the deficiency of trust model employed by the existing P2P netwok, which inadquately consider the difference and the dynamic in peers’ behavior when aggregating the peer trust, a new dynamic fuzzy comprhensive trust model (DFCTrust) was proposed in this paper, that is the time attemiation factor and the fluctuation punisive measure were added on the basis of static fuzzy comprehensive evaluation. In the model, the satisfactory degree of every transction was firstly computed by static fuzzy comprehensive evaluation. Attribute to taking the transction context factor into consideration, the model is good at avoiding the malicious act which is trust in small transction while distrust in large one. Secondly, because the time attemiation factor and the fluctuation punisive measure were added when aggregating peertrust, the model do well not only in reducing the failure due to trading with inactive peers but also in resisting the periodic oscillation cheating from malicious peers. Simulation experiment shows that DFCTrust has good dynamic adaptive ability to cope whith the behavior changes with strategy and then can improve transction success rate.

Key words: Peer to Peer (P2P) network, dynamic comprehensive evaluation, trust model, Fuzzy evaluation method