A Lightweight and High‐Throughput True Random Number Generator Based on Hybrid Entropy Source

ABSTRACT The true random number generator (TRNG), as the fundamental component of an information security system, is a cryptographic primitive essential for generating keys, random numbers, initialization vectors, and more. With the advancement of lightweight applications and high‐speed communication, solving the problem of high resource overhead and low throughput of TRNG becomes increasingly important. To overcome the limitations in terms of resource overhead and throughput posed by traditional single‐entropy source TRNG, this paper proposes a hybrid entropy source TRNG based on FPGA. This architecture combines the jitter from a ring oscillator (RO) with metastability triggered by the XOR gate to enhance the entropy generation capability of TRNG. Experimental results demonstrate that this proposed structure can be implemented and tested on multiple FPGAs without manual placement and complex postprocessing circuits. It achieves a minimum throughput of 300 Mbps with the minimum entropy greater than 0.996 while only utilizing 19 lookup tables (LUTs) and 4 D‐flip‐flops (DFFs) as resource overhead.