Edge-Based Machine Learning for Immediate Botnet Detection and Response in IoT Networks
J. V. Anchitaalagammai, S. Kavitha, S. Murali, A Shifana, A Boomika · 2024
The proliferation of Internet of Things (IoT) devices has increased the risk of botnet attacks, posing a serious threat to network security. This paper presents a novel edge-based machine-learning framework aimed at detecting and mitigating botnet attacks in real-time. The proposed method uses a combination of existing machine learning techniques, such as Graph Neural Networks (GNNs), Long Short-Term Memory (LSTM) networks, and provides algorithms such as AdaBoost and XG-Boost, to better classify network traffic and detect botnet activity. To improve the model performance, Principal Component Analysis (PCA) is used for size reduction by eliminating redundant features, ensuring the model captures essential patterns without learning noise from the data. Experimental results show that different models can generate fewer false positives and achieve higher detection accuracy in IoT environments. The proposed method achieves a botnet detection rate of [insert detection rate] %, demonstrating significant improvements in real-time IoT network security.