Detection of DDoS Attacks using Artificial Gorilla Troops Optimizer based Deep Learning Model

Abstract

The importance of security has skyrocketed alongside the adoption of Internet of Things (IoT) services. Software-defined networking (S DN) provides a means of securely managing IoT devices, which were exposed in a current distributed denial-of-service (DDoS) attack. Many IoT devices unwittingly contributed to the DDoS attack. DDoS attacks, one of the most common types of cyberattack, are particularly pernicious since they can cripple a network’s ability to function and render many of its services inaccessible to users. This research used optimised deep learning-based models to detect DDoS in SDN. At first, the dataset’s normal and DDoS attack traffic characteristics were gathered from the SDN. The models are recommended to be simpler, easier to read, and to have a shorter training period when using an NSL-KDD dataset for feature selection approaches. Real-time DDoS attack detection in SDN is proposed in this research using an Long Short-Term Memory (LS TM) models. High accuracy in classification is achieved by utilising an artificial gorilla troop optimizer to pick the features of NSL-KDD. Using less time and material, the proposed IDS was able to achieve a detection accuracy of 97.59%. These findings provide encouraging evidence that DDoS attack detection in SDN could benefit from the use of deep learning and feature selection techniques, which could significantly reduce processing loads and runtimes.