Internet of Things (IoT) devices are highly developed and can be found in everyday life such as watches, smart lights and so on. For now, there are 24 billion IoT devices connected to the internet and the number will continue to grow. The number of IoT devices connected to the internet means there are many security holes that can be exploited by irresponsible people to carry out attacks that have a wide impact on the network. One of the attacks that can be done is Low Rate Attack. To solve these problems, many researchers have created a new paradigm in networking, which is to take advantage of the advantages of Software Defined Network (SDN) to be applied to IoT networks. This study proposes a classification method for detecting low rate attacks using machine learning using the K-Nearest Neighbors (KNN) algorithm. This study also proposes a new feature scheme for the dataset by utilizing the port statistics feature in the SDN environment. The results showed that the KNN classification model applied got good results, namely 92% when evaluating the model applied to the SD-IoT environment. On the other hand, the lowest packet loss is 1.6% and the highest packet loss is 99%, this can be greatly influenced by the hardware resources used because the detection system requires high hardware resources.
Copyrights © 2023