Physical Unclonable Function and Machine Learning Based Group Authentication and Data Masking for In-Hospital Segments

Abstract

The involvement of the Internet of things (IoT) in the development of technology makes systems automated and peoples’ lives easier. The IoT is taking part in many applications, from smart homes to smart industries, in order to make a city smart. One of the major applications of the IoT is the Internet of medical things (IoMT) which deals with patients’ sensitive information. This confidential information needs to be properly transferred and securely authenticated. For successful data protection and preserving privacy, this paper proposes multidevice authentication for the in-hospital segment using a physical unclonable function (PUF) and machine learning (ML). The proposed method authenticates multiple devices using a single message. Most of the protocols require PUF keys to be stored at the server, which is not required in the proposed framework. Moreover, authentication, as well as data, is sent to the server in the same message, which results in faster processing. Furthermore, a single ML model authenticates a group of devices at the same time. The proposed method shows 99.54% accuracy in identifying the group of devices. Moreover, the proposed method takes 2.6 ms and 104 bytes to complete the authentication of a device and takes less time with the increment of devices in the group. The proposed algorithm is analyzed using a formal analysis to show its resistance against various vulnerabilities.