Lightweight Hybrid Entropy Source True Random Number Generator Based on Jitter and Metastability
Shihao Yang, Huaguo Liang, Rong Hu, Liang Yao, Zhengfeng Huang, Maoxiang Yi, Yingchun Lu
Abstract
True random number generators (TRNGs) perform an irreplaceable role in securing information systems. In this brief, a lightweight hybrid entropy source TRNG implementation scheme based on jitter and metastability is proposed, which uses the underlying field programmable gate array (FPGA) primitive language to construct a dual cross-coupled XOR gate (DCCX) unit to generate a random output sequence and is deployed by automatic layout and routing with the Xilinx compiler. The experimental results show that the generated random sequence passes the NIST SP800-22 test, the TESTU01 test and the AIS-31 test. The structure has been extensively tested under voltage and temperature variations and shows excellent robustness. It is worth mentioning that the proposed TRNG consumes only 12 LUTs and 10 DFFs, and achieves 150Mbps and 200Mbps throughput on Xilinx Artix-7 and Kintex-7 FPGAs, respectively, which shows a highly efficient resource utilization compared to the previously published TRNGs.