"Detecting Distributed Denial of Service Attacks - Deep Learning-Based Distributed Denial-Of-Service Detection"
A new study in the International Journal of Networking and Virtual Organizations explores the use of deep learning to detect Distributed Denial-of-Service (DDoS) attacks, which could help service providers lessen the effects of these attacks. According to researchers Hanene Mennour and Sihem Mostefai at the University Abdelhamid Mehri in Constantine, Algeria, DDoS attacks are ongoing, and there is an increasingly urgent need to develop methods to detect and block these attacks, especially given the current state of world affairs. They have built and tested a deep Convolutional Neural Network (CNN), a Stacked Long Short-Term Memory (S-LSTM) neural network, and a third system, which is a hybrid of the CNN and the LSTM systems. The team tested their developments against CICIDS2017, CICDDoS2019, and BoT-IoT, three benchmarking tools. Testing revealed that the scaleable hybrid tool was more effective than both of the separate modules in detecting a DDoS attack. In addition, compared to other approaches, their novel tool has lower computational complexity and can also outperform earlier approaches in nearly all metrics. This article continues to discuss the concept of DDoS attacks and the deep learning-based DDoS detection tool developed by the researchers.