基于历史聚集簇的P2P网络信任模型

P2P网络具有开放、匿名、节点间松耦合等特性，这些特性导致许多恶意节点提供虚假资源、恶意评价、团伙作案等问题.为解决这一问题，提出了一种新的P2P网络信任模型RCTrust，该模型在将节点根据兴趣等历史信息以簇的形式聚集的基础上，根据节点通信的目标所在簇分为簇内信任和簇外信任两种类型，并根据节点的通信成功率、通信历史相关度、评价诚实度以及评价时效性等行为对节点的信誉进行评价和反馈.仿真分析表明，该信任模型具有抵抗恶意攻击的能力，并提供了较高的通信成功率.

Related Cluster Based Trust Model for P2P Networks

The characteristics of P2P networks are open, anonymous, and loosely coupled inter-node, which lead to false resources, malicious evaluation, and syndicates, etc. To solve this problem, a related cluster based trust model for P2P networks named RCTrust was presented. Nodes were gathered in clusters according to their interest similarity and communication history. Trust relation was categorized into two kinds according to the cluster which communication target was belonging to and subsequently each solution for these kinds was also put forward. In RCTrust, measure of node communication success rate, communications history similarity, honesty and timeliness of evaluation were used as reference parameters to evaluate the credit of peers. Simulation results showed that robustness against malicious attacks was obtained and a high rate of successful communication was provided by the proposed model.