Automatic discovery and classifications of IoT devices

Abstract

In the last few years a new concept has gained ground in the technology world. With an increasingly dominant role in our days, Internet of Things (IoT) is a technological revolution that is changing our lives. The imagination is the limit regarding the new devices that may arise in the market. This phenomenon is derived from technology evolution and the growing acceptance of this type of products in our everyday life. Faced with such diversity, growing and with the existing uncertainty about the new devices emerging in the market, there is a need to integrate each new device in our lives. However, nowadays, there is no existing structure prepared for the expectable emerging diversity in the IoT world. This work aims to provide some help in this specific point by contributing to the automation of part of the device integration process. Applications that integrate devices from many different vendors are now available, but still rely on manual configuration by the application developers for every new device integrated, which is tedious and requires application updates to be rolled out frequently. An application that can discover the new device properties, and decide to which class of devices it belongs to can automatically generate an interface and the necessary integration drivers for the new device with no/less human intervention. The demonstration that it is possible to classify devices based on their communication properties is this works' main contribution. To achieve this goal we need to identify the device that is communicating inside a network using the information shared by IoT devices. The final results show that IoT communication data can be used to identify the device, mainly if in possession of a considerable sized device information database.