Unified PUF and TRNG Design Using SOT-MTJ Crossbar Array
Kunal Kranti Das, Arshid Nisar, Brajesh Kumar Kaushik, Anupam Chattopadhyay, Furqan Zahoor · IEEE Transactions on Magnetics · 2025
In the era of the Internet of Things (IoT), Physically Unclonable Functions (PUFs) and True Random Number Generators (TRNGs) are essential for ensuring hardware-level security and device authentication. This paper presents a unified, simulation-based design of a dual-function PUF/TRNG using a Spin-Orbit Torque Magnetic Tunnel Junction (SOT-MTJ) crossbar array. The proposed architecture exploits the inherent stochastic switching behavior and process-induced resistance variations of SOT-MTJ devices to generate unique and random responses. Depending upon the application, the same hardware can seamlessly switch between PUF and TRNG functionalities, enhancing reconfigurability and area efficiency. The proposed design achieves low energy consumption of 0.2 pJ/bit for the PUF and 0.3 pJ/bit for the TRNG, with respective throughputs of 8 Gbps and 0.333 Gbps. The PUF demonstrates excellent performance with uniqueness of 50.15%, uniformity of 51.6%, bit-aliasing of 51.25%, and perfect reliability (100%). The TRNG output passes all NIST statistical randomness tests without requiring post-processing. Compared to existing STT-MRAM and RRAM-based solutions, the proposed design offers superior performance in terms of energy efficiency, randomness quality while providing reconfigurability between TRNG and PUF applications.