Journal ID : TRKU-11-08-2020-10987
[This article belongs to Volume - 62, Issue - 08]
Total View : 1206

Title : A GWO-based Attack Detection System Using K-means Clustering Algorithm

Abstract :

So far, various methods have been proposed to deal with cyber-attacks, but many of them are not capable of running in the real environment or do not have enough accuracy to detect different types of attacks. In this paper, using the features of k-means clustering algorithm, ineffective data in detection process are removed from the dataset. Then, the accuracy of attack detection increases by using the Gray Wolf Optimization (GWO) algorithm and replacing stronger wolves based on their degree of suitability. In each iteration of the algorithm, the fitness is computed and if it improves the algorithm is repeated again, otherwise the algorithm terminates. The main purpose of the proposed method is to increase the accuracy of detection as well as reduce the likelihood of getting stuck in local optimal points. The simulation results on 4 different types of attacks in the NSL-KDD and synthetic dataset show that about 3.2% better detection accuracy is obtained rather than other researches by adjusting the parameters of the gray wolf algorithm, as a conclusion, the proposed method has the necessary efficiency for detecting attacks on computer networks

Full article

//