DDoS attack detection in the cloud environment using an optimised long short-term memory with an improved firefly algorithm

Distributed denial of service (DDoS) attacks are the most dangerous types of attacks on cloud computing. For cloud computing technology to be widely used, defences against these threats must be developed. Hence, this present research work proposes a new detection scheme based on a long short-term memory (LSTM) optimised by an improved firefly algorithm (IFA) called LSTM-IFA. The IFA uses opposition-based learning (OBL) to increase population diversity and the local search algorithm (LSA) for enhancing its exploitation is the second enhancement. The IFA is used to enhance the performance of LSTM by optimising hyperparameters that produce high detection accuracy with a fast convergence rate. Experimental findings were done over four distinct datasets to evaluate the proposed LSTM-IFA approach which is obtained 98.67% of average accuracy. The experiment's findings demonstrated that, compared to previous detection techniques, the suggested enhanced LSTM methodology achieved a greater detection rate and accuracy.