DDoS attack detection using deep learning

<span id="docs-internal-guid-58e12f40-7fff-ea30-01f6-fbbed132b03c"><span>Nowadays, IoT devices are widely used both in daily life and corporate industrial environments. The use of these has increased dramatically by 2030 it is estimated that their usage will rise to 125 billion causing enormous flow information. It likely also increase distributed denial-of-service (DDoS) attack surface. As have limited resources, impossible add additional security structures it. Therefore, the risk DDoS attacks malicious people who can take control devices, remain extremely high. In this paper, we CICDDoS2019 dataset as a improved bugs introducing new taxonomy for attacks, including classification based on flows network. We propose detection using deep neural network (DNN) long short-term memory (LSTM) algorithm. Our results show detect more than 99.90% all three types attacks. indicate learning another option detecting may cause disruptions future.</span></span>