DDoS Attack Detection using Enhanced Long-Short Term Memory with Hybrid Machine Learning Algorithms
M. Sinthuja, Suthendran Kannan · 2022 3rd International Conference on Smart Electronics and Communication (ICOSEC) · 2022
The Internet of Things (IoT) environment’s heterogeneous components make the distributed denial-of-service (DDoS) attack a security challenge. The dependency on the network topology is one of the various drawbacks of the DDoS detection techniques currently in use. In this research work, a new variant of long-short term memory (LSTM) is proposed to detect DDoS attacks in the IoT environment. The proposed method uses a new hybrid method for optimizing the parameters of LSTM based on bacterial foraging optimization (BFO) with a firefly algorithm (FA) called BFOFA-LSTM. The performance of comparison algorithms is examined using two types of datasets including BoT-IoT and CICDDoS2019 and four different performance indicators are considered such as sensitivity, specificity, accuracy, and F-Measure. The proposed BFOFALSTM detection method produced a high detection rate when compared with other literature algorithms.