Botnet Detection in IOT Using Unsupervised Learning Techniques DBSCAN and ANN

The rapid proliferation of Internet of Things (IoT) devices has significantly increased the attack surface for cyber threats, notably botnets. Traditional detection methods often fall short in identifying sophisticated and evolving botnet behaviors. This research explores the efficacy of unsupervised learning techniques, specifically Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Artificial Neural Networks (ANN), for botnet detection in IoT environments. DBSCAN is leveraged for its ability to identify anomalies in network traffic data without predefined labels, while ANN is utilized for its capacity to recognize complex patterns and adapt to new threats. The proposed framework combines the strengths of DBSCAN in clustering and noise handling with the adaptive learning capabilities of ANN. This hybrid approach is designed to detect both known and unknown botnet activities, providing a more dynamic and responsive security solution. Extensive experiments were conducted on diverse IoT datasets, encompassing various types of network traffic and botnet behaviors. The results indicate that the integrated DBSCAN-ANN model outperforms traditional detection methods in terms of accuracy, precision, and recall, while maintaining low false-positive rates. Furthermore, the study delves into the computational efficiency of the proposed model, demonstrating its scalability and suitability for real-time deployment in resource-constrained IoT environments. The results emphasize the potential of unsupervised learning techniques to enhance IoT security by offering a proactive and adaptive defense mechanism against the ever-evolving threat landscape posed by botnets. This research contributes to the ongoing efforts in securing IoT ecosystems and underscores the importance of innovative machine learning applications in cybersecurity.