Demonstration of 30 Gbit/s Generation of Superconductive True Random Number Generator

Demonstration of 30 Gbit/s Generation of Superconductive True Random Number Generator
复制标题

DOI:
10.1109/tasc.2010.2092401
复制
发表时间:
2011-06
影响因子:
1.8
通讯作者:
Tatsuro Sugiura;Y. Yamanashi;N. Yoshikawa
Tatsuro Sugiura;Y. Yamanashi;N. Yoshikawa
中科院分区:
物理与天体物理3区
文献类型:
--
作者:
Tatsuro Sugiura;Y. Yamanashi;N. Yoshikawa

文献摘要

被引文献

相似文献

真随机数发生器是从热噪声、电噪声等物理现象中提取熵来产生真随机数的一种方法,广泛应用于密码通信系统中。我们一直在开发一个真正的随机数发生器,它可以产生真正的随机数序列,不可能被预测,通过利用高速操作和高灵敏度的集成电路。在这项研究中,我们计算了输出随机比特的相关性的产生率的依赖关系。统计测试已经进行了NIST的统计测试套件的基础上,以评估的质量的随机数生成器在高生成率的伪随机数生成器生成的序列的随机性。采用ISTEC-SRL 2.5 kA/cm ~ 2 Nb标准工艺制造的可编程随机数发生器,以30 Gbit/s的产生速率产生了3.2 M比特的随机数序列。生成的随机数序列通过了NIST统计测试套中的13种统计测试,尽管由于生成的随机数不足而没有进行3种测试。该结果充分证明了一个伪真随机数发生器能够以高达30 Gbit/s的生成速率生成可用于实际密码应用的高质量随机数。
True random number generators, which output truly random numbers by extracting entropy from physical phenomena such as thermal and electronic noises, are widely used in the field of the cryptographic communication systems. We have been developing a superconductive true random number generator that can generate truly random number sequences, impossible to be predicted, by utilizing the high-speed operation and high-sensitivity of superconductive integrated circuits. In this study, we have calculated the dependences of correlation of output random bits on the generation rate. Statistical tests have been performed on the basis of the NIST statistical test suite in order to evaluate the quality of the randomness of sequences generated by the superconductive true random number generator at high generation rate. We have generated a random number sequence consisting of 3.2 M-bit at the generation rate of 30 Gbit/s using the superconductive true random number generator, fabricated by the ISTEC-SRL 2.5 kA/cm2 Nb standard process. The generated random number sequences passed 13 kinds of the statistical tests in the NIST statistical test suit, although the 3 tests were not performed because of the shortage of the generated random numbers. The result sufficiently proves that a superconductive true random number generator can generate a high quality of random numbers that can be used for practical cryptographic applications, at a generation rate of up to 30 Gbit/s.