Botnet Detection Using Hybrid Methods
Wassan Adnan Hashim, Seyed Ali, Ali T. Al-Khawaldeh, Faris K. AL-Shammri · 2024
This The proliferation of Internet of Things (IoT) devices has led to a corresponding rise in cyberattacks targeting these interconnected devices. Among the most concerning threats are botnet-based attacks, known for their complexity and destructive potential within the IoT ecosystem. Researchers have explored various machine learning (ML) and deep learning (DL) techniques to detect and classify botnet attacks in IoT environments. This study introduces an effective approach for detecting botnet attacks on IoT devices using the N-BaIoT dataset. Leveraging DL and hybrid models, we develop methods to identify three prevalent and hazardous IoT threats. Our findings highlight the superiority of the CNN-BiGRU-BiLSTM model, achieving a remarkable 98.87% accuracy in detecting botnet assaults, surpassing existing models in the field. Our research contributes to the ongoing efforts to devise efficient and precise methodologies for detecting botnet-based attacks on IoT devices. The proposed method holds significant promise for enhancing IoT security and mitigating the adverse impacts of botnet attacks within the IoT ecosystem