Smart Shield: Enhancing IoT Security Against DDoS Attacks using AI techniques
Saikat Das, Raktim Ranjan Das · 2024
In the era of the internet, the rapid growth of the Internet of Things (IoT) has significantly enhanced connectivity and automation, but it has also exposed new vulnerabilities, especially through DDoS attacks. Addressing these security challenges, this paper presents “Smart Shield,” a robust framework designed to detect DDoS attacks in IoT networks using advanced Machine Learning (ML) and Deep Learning (DL) techniques. Leveraging the CICIoT2023 dataset, a comprehensive and up-to-date repository of IoT network traffic data, we meticulously trained and evaluated multiple ML models, such as, naive Bayes, logistic regression, neural networks, decision trees, and support vector machines and DL models, such as MLP, CNN, RNN, Bidrectional LSTM and GRU. Our findings demonstrate that the Smart Shield framework achieves outstanding performance, significantly surpassing existing research in terms of F1-score, precision, recall, and false positive rate. The results underline the efficiency of integrating ML and DL methodologies in fortifying IoT infrastructures against sophisticated DDoS threats, establishing a new benchmark in the field of IoT security.