Towards Spatio-Temporal Deep Learning with Feature Selection for Robust DDoS Attack Detection

Resumen

Distributed Denial of Service (DDoS) attacks represent a significant challenge to network security, making timely detection critical. In this paper, we present a hybrid deep learning approach that combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks with feature selection based on Principal Component Analysis (PCA). We apply this method to the CICDDoS2019 dataset, which includes various DDoS attack scenarios. Our approach aims to enhance detection performance by using CNN for spatial feature extraction and LSTM for temporal sequence learning. Our experiments show that deep learning can effectively identify DDoS attacks with high accuracy when the data is properly preprocessed. The results indicate that our model outperforms traditional methods, achieving high accuracy, precision, and recall.