A Novel Approach to Prevention of Hello Flood Attack in IoT Using Machine Learning Algorithm Makine Öğrenmesi Algoritmasını Kullanarak IoT'de Hello Flood Saldırısının Önlenmesine Yönelik Yeni Bir Yaklaşım


Gönen S., Barişkan M. A., Karacayilmaz G., Alhan B., YILMAZ E. N., ARTUNER H., ...Daha Fazla

El-Cezeri Journal of Science and Engineering, cilt.9, sa.4, ss.1529-1541, 2022 (Scopus) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 9 Sayı: 4
  • Basım Tarihi: 2022
  • Doi Numarası: 10.31202/ecjse.1149925
  • Dergi Adı: El-Cezeri Journal of Science and Engineering
  • Derginin Tarandığı İndeksler: Scopus
  • Sayfa Sayıları: ss.1529-1541
  • Anahtar Kelimeler: Cyber Security, Flood Attacks, IoT, IoT Security, Machine learning
  • İstanbul Gelişim Üniversitesi Adresli: Evet

Özet

© 2022, TUBITAK. All rights reserved.With the developments in information technologies, every area of our lives, from shopping to education, from health to entertainment, has transitioned to the cyber environment, defined as the digital environment. In particular, the concept of the Internet of Things (IoT) has emerged in the process of spreading the internet and the idea of controlling and managing every device based on IP. The fact that IoT devices are interconnected with limited resources causes users to become vulnerable to internal and external attacks that threaten their security. In this study, a Flood attack, which is an important attack type against IoT networks, is discussed. Within the scope of the analysis of the study, first of all, the effect of the flood attack on the system has been examined. Subsequently, it has been focused on detecting the at-tack through the K-means algorithm, a machine learning algorithm. The analysis results have been shown that the attacking mote where the flood attack has been carried out has been successfully detected. In this way, similar flood attacks will be detected as soon as possible, and the system will be saved from the attack with the most damage and will be activated as soon as possible.