An ANFIS-IDS against deauthentication DOS attacks for a WLAN

Abstract

In this paper, an intrusion detection system (IDS) based on adaptive neuro-fuzzy inference system (ANFIS) rule is realized to minimize the detection delay for the deauthentication denial-of-service (DOS) attacks on the medium access control (MAC) layer of a wireless local area network (WLAN). Both the average sequence number gap (SNG) between the successive packets and the average statistical value of the de-authentication packets received by an Access Point (AP) are used to detect the deauthentication DoS attack. The proposed ANFIS-IDS experimental platform is implemented and tested against real deauthentication DoS attack to empirically evaluate its average detection delay (ADD) and average false alert rate (FAR). The performance of the IDS using the proposed ANFIS method is compared with non-parametric sequential change point detection (NPSCPD) algorithm in a practical WLAN environment.