An Authorized Access Attack Detection Method for Realtime Intrusion Detection System

Abstract

Recently, a malicious user breaks into the network and destroys the entire network. In particular, it destroys the whole network through unauthorized access attacks with a contaminated system. In this paper, we propose a fast unauthorized access attack detection method for the real-time intrusion detection system. Conventionally, unauthorized access detection was performed by supervised learning to analyze all collected traffic characteristics. In the proposed method, the normal traffic is classified through unsupervised learning prior to a supervised learning, and the intrusion detection of supervised learning for unauthorized access to the normal traffic is not performed. Here, our scheme makes up arbitrary test traffic to pass through a gateway in order to classify the normal traffic. A supervised learning is performed as to classify unauthorized access attack types on abnormal traffic. Therefore, we verified that the proposed method can classify normal traffic and detect unauthorized access attacks against abnormal traffic and shorten the time than the conventional method by simulation.