Distributed denial of service attack detection in cloud computing using hybrid extreme learning machine

Authors: GOPAL SINGH KUSHWAH, VIRENDER RANGA

Abstract: One of the major security challenges in cloud computing is distributed denial of service (DDoS) attacks. In these attacks, multiple nodes are used to attack the cloud by sending huge traffic. This results in the unavailability of cloud services to legitimate users. In this research paper, a hybrid machine learning-based technique has been proposed to detect these attacks. The proposed technique is implemented by combining the extreme learning machine (ELM) model and the blackhole optimization algorithm. Various experiments have been performed with the help of four benchmark datasets namely, NSL KDD, ISCX IDS 2012, CICIDS2017, and CICDDoS2019, to evaluate the performance of our proposed technique. It achieves an accuracy of 99.23%, 92.19%, 99.50%, 99.80% with NSL KDD, ISCX IDS 2012, CICIDS2017, and CICDDoS2019, respectively. The performance comparison with other techniques based on ELM, artificial neural network (ANN) trained with blackhole optimization, backpropagation ANN, and other state-of-the-art techniques is also performed.

Keywords: Cloud computing, black hole optimization, DDoS attacks, artificial neural networks, extreme learning machine

Full Text: PDF