摘要
针对P2P软件开发中的安全性问题及信任机制使用等问题,提出了一种P2P安全重叠网络数学模型。作为模型的核心部分,重叠网络被划分为4层:数据确认层、数据采集层、信任评估层和路由层。数据确认层执行对不可信数据的过滤,数据采集层对逻辑节点之间的交互数据进行向量化并进行相关的统计和分析,信任评估层通过判断某个逻辑节点是否可信来实现模型的信任机制,路由层在信任评估结果的基础上识别出恶意节点并完成逻辑节点到对应的物理节点的映射功能。在模型实现阶段,为提高P2P软件的安全性以及信任机制使用的便利性,采用简单向量距离分类法和Kademlia算法对全局信任模型进行了改进。仿真结果和实际使用情况表明,基于该模型开发的P2P软件具有较高的安全性。
To improve the security of software developing and the use of trust mechanism in P2P, a mathematical model of P2P security overlay network was proposed. As the core of the model, overlay network was divided into four parts, ineluding data confirmation layer, data capture layer, trust mechanism layer and routing layer. In the matter of functions, data confirmation layer filtered untrustworthy data; data capture layer run the vectorization of the exchanging data between logic nodes and processed relevant statistics and analysis; trust mechanism layer realized trust mechanism via identifying whether a logic node was trustworthy or not; routing layer build up mapping relationship on logic nodes and physical nodes. To enhance the security of P2P software and the convenience in the use of trust mechanism, simple vector distance classifications and kademlia algorithm were adopted on the basis of the global trust model. Simulating results show that P2P software developing can be more safe and easier under this model.
出处
《通信学报》
EI
CSCD
北大核心
2009年第5期99-104,共6页
Journal on Communications
基金
国家自然科学基金资助项目(60873231
60572131)
科技型中小企业创新基金资助项目(08C26213200495)
江苏省科技攻关基金资助项目(BE2007058)
江苏省高校自然科学基础研究基金资助项目(08KJB520005)~~
