Requirements framework for IoT device authentication using behavioral fingerprinting

Abstract

The Internet of Things (IoT) has more and more been integrated into our work and life. However, besides the benefits brought by the recent advancements, there is an increasing challenge for securing IoT devices and networks. A common security mechanism is authentication. However, IoT devices are often resource-constrained which make the use of state-of-the-art encryption technologies infeasible. Therefore, researchers are trying to develop lightweight authentication methods. A promising example of this is the use of behavioral device fingerprinting. Still, the remaining problem with this technology is that it is unclear which feature sets are most feasible to implement device fingerprinting schemes in practical systems. In short, the current research body lacks clearly defined requirements. To overcome this issue, this research aims to design a requirements framework for IoT authentication schemes using behavioral device fingerprinting. To do so, Design Science Research is used, incorporating a systematic literature review. In the end, a requirements framework for behavioral device fingerprinting authentication is presented. The proposed framework features 20 requirements in the four categories High-level IoT, Fingerprint sophistication, Machine learning sophistication, and Attack resistance. We have demonstrated the application of the requirements framework in this article. It is believed that the proposed framework will help researchers and practitioners to develop better IoT authentication solutions.