Artificial immune system based on interval type-2 fuzzy set paradigm

This paper discusses the design and engineering of a biologically-inspired intrusion detection system, based on interval type-2 fuzzy set paradigm, for protecting computer networks. To this end, we have proposed a performance-based Artificial Immune System (AIS) that mimics the workings of an adaptive immune system and consists of a number of running artificial white blood cells, which search, recognize, store and deny anomalous behaviors on individual hosts. The proposed AIS monitors the system through analyzing the set of parameters to provide general information on its state. For the analysis, we have suggested a dynamic technique based on interval type-2 fuzzy set paradigm that enable identifying the system status - i.e. Non-Attack, Suspicious-Non-Attack, Non-Decidable, Suspicious-Attack, Attack. In conclusion, for proving the effectiveness of the suggested model, an exhaustive testing is conducted and results are reported.