Granular Computing based Intrusion Detection Model upon Network Monitor Data Streams

The network intrusion detection system is a reasonable supplement on firewall, it can monitor data stream in network in real-time. Considering the problem of the urgent requirement of intrusion detection techniques upon large unbounded network monitor data streams, we presented a novel granular computing based intrusion detection model, and successfully applied it into the project MAUDS (Multi-agent based Intelligent Intrusion Detection System). lots of experiments indicated that this model can work well for intrusion detection. Based on this model, many algorithms and techniques could be developed. On the other hand, it is also useful to understand the nature of decision rules and association rules mining in network monitor data streams.