Enhanced Hybrid Deep Learning Model for Distributed Denial of Service Attack Detection in IoT Environments
S N, H S Vimala, C N Pruthvi, J Shreyas · 2024
Security concerns have grown significantly with the proliferation of Internet of Things (IoT) devices, particularly in light of Distributed Denial of Service (DDoS) attacks. Network resources may be overloaded by these assaults, leading to significant disruptions, data leaks, and outages. Studies have demonstrated that a combined deep learning strategy utilizing Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks may successfully identify these DDoS assaults in Internet of Things environments. This study aims to make the hybrid model better by adding new techniques like feature selection, ensemble learning, and an improved attention mechanism. We thoroughly test the upgraded model against the original hybrid model and other traditional machine learning methods using different datasets. The results show that the improved model boosts detection accuracy, precision, recall, and F1-score, making it a stronger tool for spotting DDoS attacks in IoT networks. By using CIC-DDoS2019 dataset, our model yields results with an accuracy of 99%, precision, recall and F1 score of 98%.