Machine learning-based detection of IoT botnet attacks


  • Hussein Maad Abdul Kadhim Department of Computer and Communications, Faculty of Engineering, Arts, Sciences and Technology University in Lebanon
  • Mona K. El Abbasi Department Electrical and Computer Engineering, American University of Beirut
  • Hayder Kareem Algabri Department of Artificial Intelligence, Al-Hilla University College


Machine Learning ,IoT Botnets, N-BaIoT Attacks.


A growing number of internet-connected devices are being used today as a result of technological developments. With the Internet of Things (IoT), life has become easier, because these devices can now communicate simultaneously with each other. With the Internet of Things, people can accomplish the tasks they need to - within a program - and make themselves comfortable. As long as the IoT devices work correctly and securely, all of their advantages are valid. As well as having their advantages and disadvantages, these devices can have disadvantages when they do not work properly or are misused. An excellent instance of this phenomenon was the IoT Botnet assaults in 2016. Machine learning techniques can effectively mitigate IoT-based assaults and premeditated attacks. This study endeavors to accurately identify regular network traffic and attack traffic by employing machine learning techniques. The N-BaIoT Provision 737E security camera has been analyzed in the literature using a data collection that includes both regular network traffic and malicious network activity. The provided dataset has been utilized for the purpose of conducting machine learning. Both supervised and unsupervised learning were conducted as part of the study. The EM (Expectation Maximization) method was employed, resulting in a success rate of 76.73% for the learning process. The decision tree (J48) algorithm in the Weka 3.8 software had a success rate of 99.95% in the supervised learning application.




How to Cite

Hussein Maad Abdul Kadhim, Mona K. El Abbasi, & Hayder Kareem Algabri. (2024). Machine learning-based detection of IoT botnet attacks. Chinese Journal of Computational Mechanics, (5), 604–614. Retrieved from