摘要： 目前多数信任模型未考虑节点的异构性，不能有效识别具有复杂策略的恶意节点。针对该问题，提出一种面向对等网络的动态多层分组信任模型。从可信度和服务能力2个方面对节点进行评估，引入惩罚机制和时间权重，基于多层分组结构构建信任模型。仿真实验结果表明，该模型能有效提高节点的交互成功率。

关键词: 信任, 对等网络, 多层分组, 可信度, 服务能力

Abstract: Many existed trust models haven’t take into account the heterogeneity of peers, and can’t identify the malicious peers with complicated strategy. Aiming at this problem, this paper presents a dynamic multilayer and grouping trust model for Peer-to-Peer(P2P) networks to resolve these problems. The model evaluates the peers on the trustworthy and service capability, as well as puts the punishing mechanism and time factor into operation. This model establishes the trust model on the multilayer and grouping structure. Simulation experimental results proves that the model can improve the successful rate of peers’ activities effectively.

Key words: trust, Peer-to-Peer(P2P) networks, multilayer grouping, trustworthy, service capability

中图分类号: 

• TP309