Circle Search Algorithm with Convolutional Neural Network for Distributed Denial-of-Service Attack Detection in Software-Defined Networks

The Distributed Denial-of-Service (DDoS) attack is a crucial cybersecurity threat for Software-Defined Network (SDN). However, DDoS in SDN severely affect network availability which reduces the classifier performance. Hence, Circle Search Algorithm with Convolutional Neural Network (CSA-CNN) is proposed for detecting DDoS anomaly attacks in SDN. The DDoS anomaly dataset such as CIC-DDoS2019 is preprocessed through label encoding which changes categorical data to arithmetical data types through allocating single numerical labels to every class. Then, CSA is applied for feature selection which helps to efficiently detect malicious traffic patterns by exploring local and global search spaces and improves the detection accuracy. The CNN has ability to detect known and unseen attacks due to handling multidimensional data which increase the security of SDN. The CSA-CNN achieved better performance in terms of 99.97%, 99.95%, 99.94% and 99.98% of precision, sensitivity, f1-score and accuracy for CIC-DDoS2019 dataset when compared to Gated Recurrent Unit (GRU).