Research on DDoS Attack Detection Based on ELM in IoT Environment
Zhihui Li, Lihong Wei, Wei Li, Lai Wei, Minshi Chen, Ming Lv, Xulong Zhi, Chenguang Wang, Nan Gao · 2019
There are many Distributed Denial of Service (DDoS) attack accidents in the world, which use the Internet of Things (IoT) devices to launch attacks and make network unavailable such as Mirai. IoT devices have reached more than 8 billion units by 2018 in the world. However, there exist some problems in the IoT environment, such as simple protocols, fixed bandwidth, a lot of devices, and easily controlled. Therefore, IoT devices are vulnerable to intrusion and become a puppet for DDoS attacks. The shortcomings of traditional methods include slow learning speed and low efficiency, and existing DDoS protection methods are applied to Ethernet, not to the IoT environment, so it is vitally important to research how to detect and respond to DDoS attacks in the IoT environment quickly. Based on the characteristics of IoT devices, joint entropy features that consist with DDoS attack in IoT innovatively are extracted and a method based on Extreme Learning Machine (ELM) to detect DDoS attacks is proposed in this manuscript. The experimental results show that the proposed method performs better than other methods in terms of detection accuracy, false positive rate and training duration.