IoT Device Classification Techniques and Traffic Analysis - A Review

Abstract

Classification of IOT devices is a trending topic nowadays. Gartner predicted that there will be 25 billion Internet of Things (IOT) devices in use. There is an increase in the risk of security breaches for devices connected to the internet, because of the proliferation of smart devices. To identify the malfunction of the devices, it is very important to identify every device connected to the network. There are some basic techniques that were traditionally used to classify the device connected to the network; they are port-based and payload-based. Today machine learning techniques are widely used as it gives more accuracy for identifying devices. The classification of IOT devices is still an immature topic, as many researchers have noted was the classification of network traffic. This article, look at emerging trends to classify IOT devices using various techniques. Also, the performance and accuracy of the various techniques used until now is discussed. This paper highlights the security issues for devices connected to the network. Finally, a discussion on an emerging subject is done, i.e. using Blockchain with IOT to secure communication between different IOT devices.