209. OVERVIEW OF DEEP LEARNING TECHNIQUES FOR NETWORK INTRUSION DETECTION SYSTEMS

Goce Stevanoski; Aleksandar Risteski; Marko Porjazoski

Goce Stevanoski Military Academy General Mihailo Apostolski, Goce Delčev University in Štip, Vasko Karangelevski St. 1000 Skopje, Republic of Macedonia
Aleksandar Risteski Faculty of Electrical Engineering and Information Technologies, “Ss. Cyril and Methodius” University in Skopje, Rugjer Bošković bb, P.O.Box 574, 1001 Skopje, Republic of Macedonia
Marko Porjazoski Faculty of Electrical Engineering and Information Technologies, “Ss. Cyril and Methodius” University in Skopje, Rugjer Bošković bb, P.O.Box 574, 1001 Skopje, Republic of Macedonia

Keywords: Intrusion detection systems, Machine Learning Algorithms, Deep Learning

Abstract

The rapid advances in the new digital world are producing vast amounts of data. This gives more opportunities in business management, but it can also help in implementing new security techniques. Intrusion detection systems (IDS) are enforcing processes for analyzing network data. This study is reviewing the main Deep Learning approaches for intrusion detection in IT network traffic. In the beginning, the study gives an overview of the various IDS types and their usability in the IT network. Then it presents some of the most used Deep learning techniques proposed by the research community in recent years. By analyzing various papers on the subject, current achievements, and limitations in developing IDS are detected and presented. The study ends by providing the future reach of the newly proposed Deep Learning techniques in monitoring and detecting malicious activities in network traffic.