The Distributed Denial of Service (DDoS) attacks is a critical cyber security threat that blocks the services for users and leads to important damage for reputation and effective customers. The existing methods provides low classification accuracy due to irrelevant features in classification performance. The Improved Whale Optimization Algorithm (IWOA) is proposed for the selecting of relevant features which improved the classification performance. The selected features are detected and classified by Optimized Long Short-Term Memory (OLSTM) that provided high detection rate of DDoS attacks. The one-hot encoding and min-max normalization techniques are used in the data pre-processing stage to improve the performance of classification. The CIC-DDoS 2019 dataset are used for evaluating the proposed method. The IWOA-OLSTM method attained highest accuracy 97.12%, precision 96.74%, recall 96.27%. f1-score 96.54% and DR 93.26% on CIC-DDoS 2019 dataset which is effective than Convolutional Neural Network (CNN) - Bidirectional Long Short-Term Memory (BiLSTM) and CNN-LSTM.

Improved Whale Optimization Algorithm and Optimized Long Short-Term Memory for DDoS Cyber Security Threat