Seybold Report ISSN: 1533-9211
Meenakshi1, Ramachandra AC2
1Research Scholar, Computer Science & Engineering Department,
Nitte Meenakshi Institute of Technology, Bangalore, India-560064. meenakshi.rao.kateel@gmail.com
2Professor & Head, Electronics & Communication Engineering Department,
Nitte Meenakshi Institute of Technology, Bangalore, India- 560064. ramachandra.ac@nmit.ac.in
Vol 17, No 10 ( 2022 ) | Licensing: CC 4.0 | Pg no:2270-2281 | Published on: 31-10-2022
Abstract
DDoS attack is launched remotely against public servers and effects legitimate network users. It is one of the greatest cyber threats to the availability of networks, applications, and services. Gaming, attack capability shows, fun, extortion, creating huge loss to business etc. are some of the top motivations for the criminals behind these attacks. Exploring DDoS attack detection and classifying attack types were main objective of this paper. To achieve this Deep Neural Network model is designed. Model is trained and tested using dataset CICDDoS-2019 which is appropriate to gain information about most recent DDoS attack types. Feature Engineering is done on dataset samples to make it fit for modelling and evaluation. Proposed model classifies packets as benign or attacked. Accuracy, precision, F1-measure and recall are some of the metrics considered here for performance measurement. Results show that proposed classification model performs well, as compare to the traditional machine learning classifiers. This study result contributes to a good understanding of IDS capacity using open-source DDoS dataset.
Keywords:
DDoS attacks, Deep Neural Networks, Feature selection, Pre-processing