Peer-to-peer based intrusion detection modeling and analysis

This paper proposes a model of intrusion detection based on peer-to-peer (p2p) mechanism. The crux of our research is that quantitatively analyze the response rate of the global intrusion detection system (IDS). The principal approaches are modeling, deduction and numerical simulation. Theoretical proof and experiments demonstrate that our model is reliable and robust. Based on this model, we draw the following conclusions. First, this model is insensitive to network scale so it is adaptable to large-scale network. Second, response rate depends mostly on peer's degree. Third, our model does not care about how many peers are aware of intrusion in the beginning. In a word, with proper parameters our model can achieve an extremely rapid response rate.