The rapid development of internet of things (IoT) is to be the next generation of the IoT devices are a simple target for attackers due to the lack of security. Attackers can easily hack the IoT devices that can be used to form botnets, which can be used to launch distributed denial of service (DDoS) attack against networks. Botnets are the most dangerous threat to the security systems. Software‐defined networking (SDN) is one of the developing filed, which introduce the capacity of dynamic program to the network. Use the flexibility and multidimensional characteristics of SDN used to prevent DDoS attacks. The DDoS attack is the major attack to the network, which makes the entire network down, so that normal users might not avail the services from the server. In this article, we proposed the DDoS attack detection model based on SDN environment by combining support vector machine classification algorithm is used to collect flow table values in sampling time periods. From the flow table values, the five‐tuple characteristic values extracted and based on it the DDoS attack can be detected. Based on the experimental results, we found the average accuracy rate is 96.23% with a normal amount of traffic flow. Proposed research offers a better DDoS detection rate on SDN.

中文翻译：

---

物联网的安全挑战：使用基于支持向量机的专家系统进行分布式拒绝服务攻击检测

物联网（IoT）的快速发展将成为下一代物联网设备，由于缺乏安全性，成为攻击者的简单目标。攻击者可以轻松破解可用于形成僵尸网络的物联网设备，僵尸网络可用于对网络发起分布式拒绝服务 (DDoS) 攻击。僵尸网络是对安全系统最危险的威胁。软件定义网络（SDN）是其中一个发展领域，它将动态程序的能力引入网络。利用SDN的灵活性和多维特性来防止DDoS攻击。DDoS攻击是对网络的主要攻击，它使整个网络瘫痪，使普通用户可能无法使用服务器的服务。在本文中，我们结合支持向量机分类算法提出了基于SDN环境的DDoS攻击检测模型，用于收集采样时间段的流表值。从流表值中提取出五元组特征值，并以此为基础检测 DDoS 攻击。根据实验结果，我们发现在正常流量下平均准确率为96.23%。拟议的研究在 SDN 上提供了更好的 DDoS 检测率。