A Novel Approach for Detection of DoS / DDoS Attack in Network Environment using Hellinger Distance Technique

Abstract

As the popularity of the Internet is increasing, the number of threats is also increasing at a rapid pace. DoS attack is one such threat. DoS attacks deplete the network bandwidth, thereby, preventing the genuine users from accessing the resources. These types of attacks exhaust the computing resources of its victim without advanced warning. The goal of our research is to detect DoS/DDoS attacks effectively using the Hellinger distance technique. We have tested the performance of our approach using real network traces. Our proposed model achieved approx 91.41% Accuracy in DoS detection.We have also analyzed and compared our own dataset with standard NSLKDD dataset and UNSW_NB15 Dataset and it is observed that our proposed model has better accuracy than that of NSLKDD dataset and UNSW_NB15 Dataset which shows an accuracy of 85.29% and 81.42% respectively.