对抗共谋及节点行为动态性的P2P信任模型

针对现有P2P信任模型存在的网络开销大、不能很好地处理节点行为的动态性改变及共谋攻击等问题，提出了一种新的信任模型。通过直接交互节点的局部评价加权其评价可信度计算节点的全局信誉值；采用基于局部评价标准差、局部评价集中度的方法识别和抑制共谋攻击；根据节点行为的改变动态更新其信誉值和评价可信度。仿真实验表明，提出的模型较现有模型在网络开销、抑制共谋攻击和节点行为的动态性等方面有较大改进。

P2P trust model for resisting collusion and dynamic behavior of node

To solve the problems of great network costs, the deficiency in dealing with dynamic characteristic and collusion attack in current P2P trust model, a new trust model was proposed. It computed global reputation value through that the direct interaction peers local evaluation weighted its evaluation credibility to avoid iterative computing process. The standard deviation and concentration ratio of local evaluation was used to identify and control collusion attack. Moreover, according to the node changes, its reputation value and evaluation credibility were updated dynamically. The simulation and experimental results indicate that the new model has advantages in reducing network costs, controlling dynamic characteristic of node and collusion attack over the existing trust models.