High-Throughput FPGA-Compatible TRNG Architecture Exploiting Multistimuli Metastable Cells
Riccardo Della Sala, Davide Bellizia, Giuseppe Scotti
Abstract
This paper presents a True Random Number Generator (TRNG) exploiting latched-XOR (LX) gates and its implementation on a Xilinx Spartan 6 FPGA device. The proposed LX-TRNG aims at improving the Throughput (TP) of conventional ring oscillators (ROs) based TRNGs by combining the effect of latches metastability and ROs jitter. Measurements results have demonstrated that the generated bitstreams show very good randomness exhibiting a byte (bit) entropy of 7.9979 (0.9997), according to T8-test of AIS-31. The proposed TRNG has also been extensively tested under voltage and temperature variations showing very good robustness. In particular both NIST’s and AIS-31 tests are passed for all the considered supply voltage and temperature ranges. The FPGA implementation occupies only 9 Slices and, despite its compactness, it exhibits a throughput as high as 12.5 Mbit/s with a 50 MHz operating frequency. The computation of the figure of merit <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$FOM_{E}$ </tex-math></inline-formula> has shown the capability of the proposed TRNG to optimize the trade-off between hardware resources, bitstreams entropy and throughput, outperforming previous works.