Volume 17, Number 4

Using Combination of Fuzzy Set and Gravitational Algorithm for Improving Intrusion Detection

  Authors

Amin Dastanpour 1 and Raja Azlina Raja Mahmood 2, 1 Kerman Institute of Higher Education, Iran, 2 University Putra Malaysia, Malaysia

  Abstract

An intrusion detection system (IDS) is a tool used by administrators to protect networks from unknown activities. In signature-based systems, the detection of attacks relies on predefined patterns or behaviours associated with known threats, triggering an alert upon identification of a match. Conversely, anomaly detection systems initiate their process by establishing a baseline profile that reflects the normal operational behaviour of the system or network. These systems possess the capability to identify previously unrecognized attacks, rendering them more effective than their signature-based counterparts. Nevertheless, anomaly-based IDS must consider numerous characteristics when pinpointing attacks Despite these difficulties, machine learning techniques have demonstrated a strong ability to achieve highly accurate anomaly detection and have been employed to identify attacks over the past few decades. Intrusion detection systems are widely used methods to maintain network security. In this paper, the proposed IDS employs machine learning approaches, namely FUZZY are initially applied, followed by optimization algorithms such as Gravitational Search Algorithm (GSA) to determine the optimal subset of detection features. Comparison study on the performance of the FUZZY and FUZZY-GSA models using KDD dataset with selected optimal 27 total features, shows that the proposed model achieves the highest detection rate with the lowest false alarm rate. The highest detection rate for FUZZY-GSA on the KDD dataset is 98.94% in comparison to other recognition algorithm. In summary, the proposed FUZZY-GSA model attains the highest attack recognition percentage with the lowest false positive rate in KDD dataset.

  Keywords

Fuzzy, Gravitational Search Algorithm (GSA), Intrusion Detection System (IDS), Security, Networks