Intrusion Detection in computer networks based Fuzzy systems and Tabu search

Due to the rapid development of computer networks¡ Intrusions and attacks into these networks have grown¡ and occur in various ways. To resist against hackers and intrusive behaviors¡ several algorithms have been introduced in literature known as intrusion detection methods. The aim of intrusion detection is to identify unauthorized use¡ misuse¡ and vulnerability made by internal users or external attackers. The proposed method¡ based on misuse detection¡ extracts required knowledge from fuzzy system which is a set of fuzzy if-then rules¡ and performs the intrusion detection process. . In fact¡ the mentioned knowledge is considered as a fuzzy rule base which is optimized during the data mining process by an optimization algorithm according to some criteria such as accuracy and comprehensibility. Tabu search algorithm is employed to optimize the obtained set of fuzzy rules. Finally¡ the proposed method is implemented and applied to the NSL-KDD dataset which contains some information about normal and intrusive behaviors in computer networks. The results are compared with those of well-known methods¡ and show the competitive accuracy and efficiency.