Truly Random Number Generator (TRNG) accumulated jitter estimating method based on annular oscillator

Abstract

The invention discloses a truly Random Number Generator (TRNG) accumulated jitter estimating method and circuit based on an annular oscillator; the method comprises the following steps: 1, using sampling interval inner oscillation signal edge counting as the statistics sample; 2, calculating sample variance; 3, determining whether the sample variance is bigger than d (belongs to 1 / 12, 0.2098) or not, if the sample variance is bigger than d, the accumulated jitter estimated value can be resolved as count square root of the formula (sample variance-1 / 12), otherwise the estimated value equals to the sample standard deviation. The method uses sample frequency distribution to calculate the sample variance, thus eliminating mass multiplication operations needed by using direct formula to calculate sample variance; the designed estimating circuit uses a sampling signal and an oscillation signal as input, uses the accumulated jitter square estimate value as output, and comprises an edge counting module, a mean value estimate module, a sample space forming module, a sample frequency statistics module, and an accumulated jitter estimate module; the method and circuit can accurately and efficiently estimate truly Random Number Generator (TRNG) accumulated jitter based on the annular oscillator, and the entropy assessment theory can be combined so as to estimate the true randomness.

Description

technical field [0001] The invention relates to a method and circuit for estimating the accumulated jitter of a true random number generator (RO-based TRNG) based on a ring oscillator, which can be applied to the field of true randomness evaluation of the true random number generator. Background technique [0002] RO-based TRNG provides true random numbers for cryptosystems, and is widely used in various security hardware devices because of its simple structure and suitable for digital circuit implementation. RO-based TRNG samples the fast oscillation signal with jitter generated by the ring oscillator by a fixed frequency sampling signal, and the sampling result is output as random data, such as figure 1 and figure 2 shown. The true randomness that RO-based TRNG can provide determines the security of its output data. According to the latest RO-based TRNG entropy evaluation theory, the accumulated jitter in the sampling interval is essentially the standard deviation of t...

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the problem of inaccurate estimation caused by the introduction of quantization error when edge counting is used as RO-based TRNG cumulative jitter estimation samples, and at the same time reduce the high computational load of the cumulative jitter estimation method on chip

Images