Towards Trusted IoT Sensing Systems: Implementing PUF as Secure Key Generator for Root of Trust and Message Authentication Code

Abstract

Trustworthy sensor data is important for IoT sensing systems. As such, these systems need to guarantee that the sensor data is acquired by the correct device and has not been tampered with. However, IoT sensing systems can be quite complex and are often composed of multiple components, i.e., a main device and subordinate sensors. The main device is responsible for gathering and processing the data from the subordinate sensor and reports the result to a server. In order to guarantee data correctness, we introduce two types of physically unclonable function (PUF): one for the main device and one for the subordinate sensor. The main device has a trusted execution environment (TEE) for critical processing, and the correctness of the TEE is guaranteed by remote attestation based on a PUF. The subordinate sensor sends the sensor data to the main device with a message authentication code (MAC) based on a PUF. We implemented a trusted IoT sensing system using a RISC-V Keystone with a PRINCE Glitch PUF for the main device and a Raspberry Pi that simulates a CMOS image sensor PUF for the subordinate sensor.