Botnet detection In IoT devices using machine learning

Abstract

Internet of Things (IoT) devices are a group of interconnected devices or machines that have the ability to transfer data over a network without the influence of any external factor. The technology makes use of sensor nodes embedded into everyday computing objects, which communicate in a wireless multi-hop fashion to exchange data over a local network or the internet. With the rapid technological advancements taking place around the globe, the use of IoT devices has also increased proportionately. Although the prevalence of IoT devices in human lives has influenced the IoT manufacturers to make it cheap an accessible, but on the other hand, the system provides minimal control with no substantial security measures due to its prodigious application, which in turn makes it susceptible to botnet attacks. Botnet is a network of interconnected malware contaminated IoT devices, individually referred to as a bot. These bots are used as instruments of malicious attack on a network of IoT devices which allows the group of hackers (referred to as Botmaster) to perform distributed denial-of-service attack (DDoS), data theft and spam by flooding the network with unnecessary information. As a result, botnet detection has risen as an essential ingredient of network security. In this paper, our motive is to use various Machine Learning algorithms to detect botnet attacks and filter out the algorithm which will be most suitable and accurate to detect such attacks by comparing the derived outputs.