An evolutionary-based technique to characterise an anomaly in internet of things networks

Abstract

Internet of things (IoT) can connect devices embedded in various systems to the internet. Distributed denial-of-service, commonly referred as DDoS is an attack to disrupt normal traffic of a victim in IoT system, facility, or network by crushing the target or its surrounding infrastructure with overflow of internet traffic. Unfortunately, modern DDoS attack detection strategies have been failed to rationalise the early detection of DDoS attacks. Therefore, in this study, teaching learning-based optimisation (TLBO) with the learning algorithm is integrated to mitigate denial of service attacks. Furthermore, TLBOIDS selects the most relevant features from the original IDS dataset which can help to distinguish typical low-rate DDoS attacks with the use of four classification algorithms. KDD Cup 99 dataset is used in the experimental study. From the simulation results, it is obvious that TLBOIDS with C4.5 achieves high detection and accuracy with a false positive rate.