Article Preview
TopIntroduction
A spree of huge distributed denial-of-service (DDoS) attacks in late 2016 basically changed the threat landscape in Internet security. Notably, the initial attack on Krebs exceeded volumes of 600 Gbps, halt major Internet service suppliers such as OVH and Dyn (Krebs, 2016; Millman, 2016; Moss, 2016). This destructive attack came from hundreds of thousands of Internet of Things (IoT) devices controlled by a brand new botnet named Mirai. The Mirai botnet is a wake-up call to the industry to better secure Internet of Things (IoT) devices or risk exposing the Internet infrastructure to increasingly disruptive distributed denial-of-service (DDoS) attacks (AsSadhan et al., 2017).
Today, the explosive growth of the Internet of Things market has connected billions of small devices to the Internet. In order to survive in such a rapidly growing market, developers usually ignore security issues and bring vulnerable products to the market. These vulnerabilities can be used for launching large-scale DDoS attacks by Mirai, threatening users' personal information and the company's data assets (Garg & Sharma, 2017; Silva, 2017). With the explosive growth of IoT devices, it is no longer realistic to respond with workforce tactics, and an innovative method is needed to greatly improve defense capabilities.
In this paper, we propose a machine learning-based white-hat worm launcher for Botnet Defense System (BDS). The main contributions of our research are as follows:
- 1.
A machine-learning based white-hat worm launcher is designed to predict white-hat worms’
- 2.
appropriate positions for BDS.
- 3.
In order to apply the launcher to large-scale IoT network, the divide-and-conquer algorithm is proposed to develop the launcher’s scalability.
- 4.
To tackle the correlation problem in divide-and-conquer algorithm, the boundary overlapping method is proposed to develop the launcher’s adaptivity.
- 5.
The effect of the proposed launcher is evaluated through the simulation of PN2 model.
Remaining part of this paper is organized as follows: Section 2 surveys the related work and BDS. Section 3 presents the launcher’s methodology. Section 4 presents the simulation results and discussion. Section 5 summarizes our key points and gives future work.