An Experimental Framework for Demonstrating the Reliability and Efficacy of SRAM as Device Identifier in Secure PUF-Based IoT Authentication

The expansion of the Internet of Things (IoT) has significantly improved efficiency and connectivity across various sectors, but it has also increased the vulnerability of these systems to security threats such as unauthorized access and data tampering. Traditional security solutions often fall short due to the unique challenges of IoT environments, including heterogeneous device types, limited computational power, and the need for low-power operations. This paper explores the application of Physically Unclonable Functions (PUFs), utilizing inherent physical variations in hardware to generate unique, device-specific identifiers for robust device authentication. Specifically, this study employs Static Random-Access Memory (SRAM) as a reliable identifier by leveraging its unique power-up states.An experimental framework was established, involving Arduino Uno R3 boards equipped with ATmega328p microcontrollers, where 4,000 SRAM readouts were collected under various conditions to test the stability and uniqueness of SRAMbased PUF fingerprints. Preliminary results demonstrate high levels of stability and uniqueness, confirming the potential of SRAM-based PUFs for secure IoT applications. The findings not only validate the effectiveness of SRAM-based PUFs in enhancing IoT security but also pave the way for future research to optimize and expand the use of this technology in securing IoT devices against emerging threats.