A High Throughput STR-based TRNG by Jitter Precise Quantization Superposing

Abstract

With the rapid development of integrated circuits and the continuous progress of computing capability, higher demands have been placed on the security and speed of data encryption in security systems. As a basic hardware security primitive, the true random number generator (TRNG) plays an important role in the encryption system, which requires higher throughput and randomness with lower hardware overhead. However, the throughput of TRNG is related to the entropy source’s quality and the randomness extraction methodology. To quantify the randomness of the entropy source with higher efficiency and quality, we utilize the independent jitter of the self-timed ring (STR) to generate original entropy and propose a high throughput jitter-based TRNG which can extract random information at the pulse of oscillation signal by jitter precise quantization superposing and random oscillation sampling. The proposed TRNG has been implemented on Artix-7 and Virtex-6 FPGAs. The generated true random number successfully passes the NIST SP800-22 and NIST SP800-90B tests while also exhibiting a minimum entropy greater than 0.9947. The most prominent superiority of our proposed TRNG is that it achieves a high throughput of 330 Mbps with an ultra-low hardware overhead of only 35 LUTs and 12 DFFs.