Review of Botnet Attack Detection in SDN-Enabled IoT Using Machine Learning.

Area of Science:

• Cybersecurity
• Network Engineering
• Artificial Intelligence

Background:

• Software-defined networking (SDN) and the Internet of Things (IoT) integration offers vast connectivity but introduces significant vulnerabilities.
• Botnet attacks, including DDoS and phishing, pose severe threats to SDN-enabled IoT networks, causing economic disruption.

Purpose of the Study:

• To review research on machine learning techniques for detecting and mitigating botnet attacks in SDN-enabled IoT environments.
• To analyze the performance of various machine learning models in identifying and countering these cyber threats.

Main Methods:

• Discussion of major botnet attack types in SDN-IoT networks.
• Analysis of commonly used machine learning (ML) and deep learning (DL) techniques for botnet detection.
• Evaluation of ML/DL model performance using standard metrics.

Main Results:

• Both classical ML and DL techniques demonstrate comparable performance in botnet detection.
• Classical ML methods necessitate extensive feature engineering and struggle with detecting novel, unforeseen attacks.
• Timely detection, real-time monitoring, and adaptability remain challenges for classical ML due to signature-based approaches.

Conclusions:

• Machine learning offers viable solutions for botnet detection in SDN-IoT networks.
• Deep learning models may offer advantages in adaptability and detecting new threats compared to classical ML.
• Further research is needed to address the limitations of classical ML in real-time, adaptive botnet defense.