An FPGA-Compatible TRNG with Ultra-High Throughput and Energy Efficiency

Zhaojun Lu,Houjia Qidiao,Qidong Chen,Zhenglin Liu,Jiliang Zhang
DOI: https://doi.org/10.1109/dac56929.2023.10247746
2023-01-01
Abstract:In this paper, we design an energy-efficient true random number generator with ultra-high throughput for FPGA. Only four ring oscillators constructed using eight LUTs are sampled by multiple sampling points to fully exploit the randomness of the entropy source, which provides high-quality and over 275 Mbps random sequences while consuming 13 slices. An end-to-end implementation and testing framework is tailored for easy deployment and portability on Xilinx 7 serials FPGAs. The proposed architecture passes the NIST SP 800-22 and 800-90B tests without post-processing and outperforms the state-of-the-art in terms of minimum entropy and energy efficiency.
What problem does this paper attempt to address?