Network Intrusion Detection System Using Fuzzy Logic

Abstract

Computer and network security plays an important role in modern communication. An efficient network is free from malicious activities. Today’s network administrators have to pay most of their efforts to make the network secured. This paper presents a network intrusion detection system using fuzzy logic approach. Throughout this paper, we have developed an intrusion detection system that uses fuzzy rules to differentiate normal and abnormal activities. For building fuzzy rules, we use membership function depending on normal activities in network environment. A rule is matched with some degree of membership value. All data are combined and defuzzyfied. We have tested detection system with several datasets with 92% accuracy.