摘要： 为提高P2P 信任模型对恶意节点的抑制能力,提出一种改进的分组P2P 信任模型。利用模糊推理规则结 合信任值和贡献值,将网络中节点划分为若干不同等级的小组,通过小组等级限制节点的资源访问权限。在直接 信任度的计算中引入时间衰减函数反映节点的实时情况,并设置惩罚因子对节点的恶意行为进行惩罚。在推荐信 任度的计算中结合小组等级计算推荐节点可信度,以降低算法的复杂度。数据分析结果表明,该模型能有效抑制 恶意节点的攻击,随着共谋节点、自私节点及震荡节点的增加,其文件下载成功率高于PeerTrust 模型和EigenTrust模型。
关键词:
信任模型,
双重属性值,
分组,
时间衰减函数,
惩罚因子
Abstract: To improve the P2P trust model’ s ability of inhibiting malicious nodes. Grouping nodes in the network
according to the level by using the basic fuzzy inference rule and combining trust value and contribution value,and limiting the resource access through the level of the node. Time attenuation function reflecting the actual situation is introduced and penalty factor punishing malicious behavior of the node in the calculation of the comprehensive trust is designed. In the calculation of recommend trust,it effectively reduces the complexity of the algorithm by using the recommend node’s credibility which is calculated by the level of node. According to the analysis of data,the proposed model can effectively inhibit the malicious nodes attack,with the increasing proportion of collusion nodes,free-rider nodes and concussion nodes in the network,file download successful rate of this model is higher than PeerTrust model and EigenTrust model.
Key words:
trust model,
dual attribute values,
grouping,
time decaying function,
penalty factor
中图分类号: