Year: 2023
Type: article
Abstract: Due to the increasing use of networks at present, Internet systems have raised many security problems, and statistics indicate that the rate of attacks or intrusions has increased excessively annually, and in the event of any malicious attack on network vulnerabilities or information systems, it may lead to serious disasters, violating policies on network security, i.e., “confidentiality, integrity, and availability” (CIA). Therefore, many detection systems, such as the intrusion detection system, appeared. In this paper, we built a system that detects network attacks using the latest machine learning algorithms and a convolutional neural network based on a dataset of the CSE-CIC-IDS2018. It is a recent dataset that contains a set of common and recent attacks. The detection rate is 99.7%, distinguishing between aggressive attacks and natural assertiveness.
Authors: Teba Ali Jasim Ali, Muna M. Taher Jawhar
Institution: University of Mosul
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4318615325