A botnet is a network of infected computers that are used by hackers to launch distributed denial of service (DDoS) attacks, phishing attacks, spambot attacks, chatbot attacks, etc., for the purpose of gaining access to steal confidential information or hook a system for ransom. Concentration of botnet attacks before was on traditional devices like personal computers (PCs), laptops, phones, and so on. The coming of the Internet of Things (IOT) changes the direction of attack and adds strong volume to the attack. When hackers discovered the flaws in IoT, especially in the area of configurations, they took the advantages and regularized the strong volume of attack. Many researchers have studied and contributed to botnet detection methods and techniques, using machine learning models, neural network models, deep learning models, blockchain, or intrusion detection system models. In this study we present an improved technique of detecting, classifying, and eliminating botnets in a network system. We hybridized and modified feature selection algorithms with machine learning to detect botnets in a network system. The accuracy achieved was 99%, precision was 99%, recall was 99.5%, and the F1 score was 99.5%. This means that the classification report shows nearly perfect performance on both normal and botnet traffic.
Machine Learning (ML), Deep Learning (DL), Internet of Things (IoT), Command and Control (C&C), Pair to Pair (P2P), Industrial Internet of Things (IIoT)
IRE Journals:
Abu Tasiu , Dr. Abdullahi Musa Yola , Abubakar Sadiq Nurudeen , Abdulrahman Buba
"Improved Algorithm Technique for Detecting and Eliminating Botnets in Network System" Iconic Research And Engineering Journals Volume 9 Issue 2 2025 Page 399-409
IEEE:
Abu Tasiu , Dr. Abdullahi Musa Yola , Abubakar Sadiq Nurudeen , Abdulrahman Buba
"Improved Algorithm Technique for Detecting and Eliminating Botnets in Network System" Iconic Research And Engineering Journals, 9(2)