Chaotic Analysis and Entropy Estimation of the Entropy Source Based on Semiconductor Superlattice Chaos

Liwei Xu,Han Wu,Jianguo Xie,Jing Liu,Qiong Yuan,Shengwei Xu,Xiaoming Chen,Shoushan Luo,Huamin Feng
DOI: https://doi.org/10.1016/j.mejo.2022.105556
2022-01-01
Abstract:A true random number generator (TRNG) is a fundamental security primitive. Recently, the semiconductor superlattice (SSL) has been investigated as an ideal physical entropy source with high speed. It shows multi-degree freedom nonlinearity based on the effects of quantum mechanics. However, there is still a lack of systematic evaluation schemes for SSL. In this study, a chaotic assessment model for SSL is proposed. Chaotic characteristic assessment and nonlinear dynamics estimation are carried out using the model. Furthermore, a set of entropy estimation procedures are introduced to evaluate the uncertainty of the SSL. The estimation results demonstrate that SSL is chaotic, highly complex, and nonlinear. SSL shows stable performance and has an entropy rate greater than 50% with 8 sample bits. Moreover, the throughput of the assessment prototype reaches 300 Mbit/s. In brief, SSL is an excellent physical entropy source for high-quality and high-speed TRNG.
What problem does this paper attempt to address?