AN ADAPTIVE FUZZY LOGIC MODEL FOR INTRUSION DETECTION IN COMPUTER NETWORKS BASED ON ARTIFICIAL IMMUNE SYSTEM
An adaptive model and the structure of the Intrusion Detection System (IDS), which is constructed based on immunological principles was proposed. Fuzzy rules classify objects belonging to several classes simultaneously with varying degrees of affiliation. Recognition of network traffic state is the shortage of a priori information about the properties of the source intrusion and the stochastic nature of recognizable events. To increase the level of confidence in the intrusion detection system was made adaptive tuning of decision rules for the classification of network traffic states. The system is designed for the detection and classification of network attacks classes: DoS, R2L, U2R, Probe . Setting up and testing of the model are based on the detection of anomalies in the data sets obtained from the real IPtraffic
of computer networks and contained in a certain database KDD'99.