基于统计推断的P2P信誉评估机制

利用简单的数理统计学知识建立计算模型,为P2P网络中每一个节点提供一个信誉值依概率出现的区间,对节点未来的行为预测,从而使得节点可以规避风险,减少恶意行为的发生.并且针对网络中的恶意攻击,提出了相应的解决方法.通过模拟实验证明了该模型具有良好的性能.

P2P Reputation Estimation Based on Statistical Inference

In this paper, we will build a computational model based on simple mathematical statistics, and then assign each node an interval in which reputation value will occur with some probability, so that nodes can avoid risk and reduce occurrence of malicious behavior. We also introduce some methods against intentional attack and prove that these methods have high performance by simulation.