A MACHINE LEARNING FRAMEWORK BASED ON VARIOUS NETWORK TRAFFIC CHARACTERISTICS TO IDENTIFY AND CLASSIFY THE DEFAULT BEHAVIOR OF IOT DEVICES ON A NETWORK
Pagalla Bhavani Shankar, Yogi Reddy Maramreddy, Padala S Venkata Durga Gayatri · International Journal of Engineering Applied Sciences and Technology · 2021
The Internet of Things (IoT) is being well acquire to the next era of revolutionary generations amongst the new technologies. IoT technology being hailed so hard we had to stop in our society, smart homes, enterprises, and smart cities. Dynamics of smart one’s are increasingly being equipped with a profusion of IoT devices. Due to the tremendous upgradation of knowledge in various aspects impresarios of such smart environments may not even be fully aware of their working nature or principles of IoT devices, assets and functioning properly safe from cyberattacks. In this paper, we addressing this challenge by developing a robust framework for IoT device classification using traffic characteristics obtained at the level of network level. As a part of robust framework, firstly, we have a tendency to instrument a smart environment with 28 completely different IoT devices, spanning cameras, lights, plugs, motion sensors, appliances and health-monitors. We have a tendency to collect and synthesize traffic traces from this framework infrastructure for a period of 6 months, a type of subset of which we release as open data for the community to use. Second, we have to present or gifts the insights into the underlying network traffic characteristics using statistical and applied mathematical attributes such as activity cycles, port numbers, signaling patterns and cipher suites. Third, we have a tendency to develop a multi-stage machine learning based classification algorithm and demonstrate its ability to identify specific IoT devices with over 99% accuracy based on their network flow of activity. Finally, we have a tendency to discuss the trade-offs between cost, speed, and performance involved in deploying the classification network framework in real-time. Our study paves the way for impresarios of smart environments to monitor their IoT devices and assets for presence, functionality, and cyber-security without requiring any specialized devices or protocols.