DDoS-Net: Classifying DDoS Attacks in Wireless Sensor Networks with Hybrid Deep Learning

With the ubiquitous deployment of Wireless Sensor Networks (WSNs) in diverse applications, ensuring their security against cyber threats is imperative. One significant challenge is the susceptibility of WSNs to Distributed Denial of Service (DDoS) attacks, necessitating effective countermeasures. This research addresses the critical issue of DDoS attacks in WSNs, recognizing the diverse and imbalanced nature of WSN data. The imbalance poses challenges for intrusion detection systems, and existing solutions often fall short in providing comprehensive protection. The study introduces DDoS-Net, a hybrid Deep Learning (DL) model that integrates Long Short-Term Memory (LSTM) and Multi-Layer Perceptron (MLP) architectures. DDoS-Net is designed to handle data imbalance effectively and incorporates thorough feature analysis to enhance the model's detection capabilities. Evaluation on the WSN-BFSF dataset demonstrates the exceptional performance of DDoS-Net. The model achieves an accuracy of 98.12%, precision of 98.16%, recall of 98.12%, and an impressive f1-score of 0.98 in minimal epochs. These results surpass existing state-of-the-art models, showcasing the efficacy of the proposed DL approach in countering DDoS threats in WSNs.