Fuzzy C-Means Based Feature Selection Mechanism for Wireless Intrusion Detection

Abstract

The wireless network devices grow rapidly and the security of these devices is quite crucial. Attackers employ new techniques and methods to deceive the system and to take the most important information. An intrusion detection model is required to monitor wireless security breaches when the prevention methods in the preparation phase are passed. It is critical to have an automatic detection policy in place to respond to network threats as quickly as feasible. In this study, we propose a Fuzzy C-Means (FCM) based feature selection mechanism for wireless intrusion detection. The proposed mechanism utilizes the distance of FCM center point and data point, calculates the difference of normal and attack center distances, and adopts the distances to select the features. We evaluate the algorithm with the benchmark Aegean Wi-Fi Intrusion Dataset (AWID), and the results show an impressive accuracy for the binary detection of flooding, impersonation and injection attacks.