Abstract
Physically Unclonable Functions (PUFs) have become a fundamental security primitive for hardware authentication and cryptographic key generation. This study evaluates the statistical properties of static random-access memory (SRAM)-based PUFs, analyzing multiple instances under varying design-time decisions using the NIST 800-22 test suite. The results demonstrate that all evaluated PUFs meet the NIST randomness criteria, with p-values above 0.01 for the Monobit and Runs tests, min-entropy (H min ) values exceeding 0.91, and an average between-class Hamming distance (BCHD) near 50%, ensuring strong uniqueness. Additionally, design-time parameters such as column multiplexing ratios slightly influence entropy, while memory size and aspect ratio exhibit minimal impact on response stability.