CN1818853A - Device and method for producing true random codes - Google Patents

Abstract

A device for generating true random codes, including several multi-threshold random code generators and data processors, the noise generators are connected to each multi-threshold random code generator to form several random code generator units, each multi-threshold random code generator unit The threshold random code generator is connected with the data processor. The method of using this device to generate a true random code: the random level signal generated by the noise generator is screened by a multi-threshold random code generator to obtain a binary true random code sequence, and then several random codes are generated in each cycle The binary true random code sequence generated by the device unit takes one bit each, and is combined into a binary true random code sequence output by the data processor. The device obtains the output of a true random code through a multi-threshold random code generator, so as to improve the output code rate and randomness of the random code. The device can generate true random codes that do not need to be adjusted, work stably, output code rates can be expanded, randomness can be improved, and meet the needs of quantum secret communication.

Description

A device for generating true random codes and a method for generating true random codes

technical field

The invention relates to the technical field of optical quantum secure communication, in particular to a device for generating a true random code and a method for generating a true random code.

Background technique

The function of the true random code generator is to generate a code book that will never be repeated, ensuring that the one-time pad is absolutely unbreakable. The quantum channel of quantum secure communication is not used to transmit ciphertext or plaintext, but to establish and transmit codebook. To complete the entire quantum communication process, two communication channels are required, one is a quantum channel for transmitting quantum keys; the other is a classical channel for key verification and ciphertext transmission. Due to its physical principles, quantum channels cannot be tapped. However, the two processes carried out through the classical channel are not absolutely untappable. When verifying the key, the announced result is public. In order to prevent eavesdroppers from inferring the unpublished part based on the announced part, all quantum keys must be irrelevant. In an actual quantum key distribution system, Alice and Bob need to use random codes to randomly prepare and randomly detect the quantum states of single photons. The quantum cipher sequence is extracted from the random code sequence, so each bit of the random code is required to be statistically independent. At the same time, it has been proved mathematically that the one-time pad system is a completely safe and unbreakable password. The one-time pad system requires a key quantity equal to the corresponding plaintext quantity, and it is only used once. The absolute security of the quantum secure communication system is jointly determined by the absolute security of the quantum channel and the classical channel, and its theoretical basis is the quantum non-cloning principle and the one-time pad principle respectively. To ensure the absolute security of the classical channel, it is necessary to use a true random code generator to generate statistically independent true random codes, and provide a codebook that will never be repeated for the one-time pad encryption system, thereby ensuring the absolute security of the classical channel of quantum secure communication. Safety.

1) There are two main methods of generating random codes: mathematical methods and physical methods

(1) Mathematical method

The commonly used random number generation methods in mathematics include: RAND table to generate pseudo-random numbers, linear congruential generator to generate pseudo-random number sequences, feedback shift register to generate pseudo-random number sequences, pseudo-random number generator ANSIX9.17 and so on. These pseudo-random number sequences generated by mathematical methods are only determined by algorithms and seeds. Once the algorithm and seeds are determined, the value of each bit in the sequence is determined. Therefore, the information entropy is zero and cannot be used in quantum security systems. .

(2) Physical method

The physical method of generating random codes is to pass random signals through the transformation system to obtain binary random codes. At present, the research work of true random code generator is mainly concentrated on the research work of random source. For the transformation system, the scheme of comparator screening is generally used.

2) Common random signal sources

(1) Select the natural noise of the real world

There are abundant random phenomena in nature, and the randomness of various noise signals can be used to obtain true random codes.

(2) Quantum random events

The resonant fluorescence radiation generated by trapped ions has a random distribution of photon interstitial times, randomness of photons passing through optical beam splitters, and so on.

(3) Surrounding computer-generated random events

A copy of every keystroke, mouse command, sector count, time-of-day seek latency for every disk operation, actual mouse position, monitor scan line count, actual displayed image content, and more.

3) The progress of hardware generation of true random numbers at home and abroad

G.B.Agnew proposed to put two metal-insulated-semiconductor capacitors (MISC) very close together, and the random bit is a function of the difference in charge between them; bit random number generator; Manfield Richter has developed a random number generator based on the thermal noise of semiconductor diode; Nanjing University has also successfully developed a random number generator based on the principle of chaos, which is summarized by the deterministic equation (referring to The uncertainty of the solution of the equation caused by the nonlinear effect of the formal determination), the typical examples are Lorenz differential equation and the image Logist, they can use the small deviation of the initial value to cause the large deviation of the solution to produce the chaotic effect , the period of the random sequence is theoretically infinitely long, this method can be used for synchronous communication, and the working process is chaos→chaos synchronization→chaos; the Institute of Physics, Chinese Academy of Sciences has developed an optical quantum random source based on the randomness of a single photon passing through a beam splitter; East China Normal University Developed a random source based on polarized light quanta; Lepley et al. used broadband white noise generated by erbium-doped fiber amplifiers and pin detectors as the input of high-speed D flip-flops, and produced 2.5Gb/s and 5Gb/s non-return-to-zero true random codes .

4) Comparison of various methods for generating true random codes

True random code generators are divided into two categories: traditional and quantum according to the physical random source they use. Traditional random code generators use noise signals as random sources, and output random codes through a single-threshold discriminator. Although it is easy to implement and fast, it is difficult to control and maintain stability due to various statistical characteristics of noise. The randomness of the code is poor. The quantum random source is a completely random source that never repeats. Although it has neither rules nor repeats, it is difficult to collect data and the equipment is complicated. At the same time, the stability and generation rate of the quantum random source are far from quantum secrecy. Requirements, and affected by factors such as interference noise, the randomness of the actual random code is far less than the ideal value of the theoretical value. For example, quantum security system experiments usually use single photons as a random source. Although the randomness is good, this method is limited by stability and speed, and it is still in the laboratory stage. The correlation coefficient is not ideal, even greater than 0.05. That is, if the upper limit of irrelevance is exceeded, the output random codes can be considered to be correlated.

In order to achieve the practical stage of quantum secure communication, random code generators with good randomness, stable operation and low energy consumption are needed. However, it is difficult for the random code generators mentioned above to meet these requirements at the same time.

Contents of the invention

The purpose of the present invention is to provide a device for generating true random codes, which obtains the output of true random codes through multi-threshold random code generator screening, so as to improve the output code rate and randomness of random codes.

Another object of the present invention is to provide a method for generating true random codes with the above-mentioned device, which can generate codes that do not require adjustment, work stably, output code rates can be expanded, randomness can be improved, and meet the needs of quantum secret communication. True random code.

The device for producing true random codes of the present invention includes a noise generator, several multi-threshold random code generators, and a data processor, and the noise generator is connected to each multi-threshold random code generator respectively to form several random code generator units , each multi-threshold random code generator is connected to the data processor.

The multi-threshold random code generator is an analog-to-digital converter, and the analog-to-digital converter outputs with the lowest bit; or comprises a comparator array and a decoder, a noise generator, a comparator array, a decoder, and a data processor in sequence Connection; the comparator array is a number of comparators connected in parallel, the noise generator is connected to each comparator separately, the comparator array outputs the corresponding TTL level, and after being decoded by the decoder, a binary true random code sequence is obtained.

The method of generating a true random code with the above-mentioned device:

The random level signal generated by the noise generator is screened by the multi-threshold random code generator to obtain a binary true random code sequence, and then in each cycle, the binary true random code sequences generated by several random code generator units are respectively One bit is taken and combined into a binary sequence by a data processor to output a true random code sequence.

The random level signal generated by the noise generator is screened by the comparator array, and the comparator array outputs the corresponding TTL level, and after being decoded by the decoder, a binary true random code sequence is obtained.

The random level signal generated by the noise generator can also be screened by the analog-to-digital converter, and the lowest bit of the analog-to-digital converter can be used as an output to obtain a binary true random code sequence.

The data processor is a parallel-to-serial converter or a data acquisition card.

The present invention is characterized in the conversion system and output characteristic extension of the multi-threshold random code generator, the conversion system is an analog-to-digital converter, or a comparator array and a decoder are connected together.

The present invention digitizes the noise signal and then divides it into two groups according to its parity, the odd group corresponds to the random code "1" (or "0"), and the even group corresponds to the random code "0" (or "1"). The principle is as follows:

It is assumed that the input noise voltage V falls within the interval [V _min , V _max ], and its probability distribution density is p(V). Divide this interval into 2 ⁿ small intervals with multiple threshold voltages, and number each interval from bottom to top (from 1 to 2 ⁿ ). Then the width of each interval is

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}{{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}}$

The i-th interval is [V _min +(i-1)D _v , V _min +iD _v ], where 0<i≤2 ⁿ . We stipulate: when the noise falls in the i interval, if i is an odd number, the output of the random code generator is "1" (or "0"), and when i is an even number, the output is "0" (or "1 "). In this way, the input random noise is digitized into a random code.

Multi-threshold discrimination refers to setting the threshold voltages of the comparators in the comparator array to the threshold voltages mentioned above, so that there is a one-to-one correspondence between the output of the comparator array and the input noise voltage. When the noise falls in the i interval, there are i comparators whose outputs are high level, and 2 ⁿ -i comparators whose outputs are low level.

The decoder refers to the ability to decode the output of the comparator so that when the noise falls in the i interval, if i is an odd number, the output of the random code generator is "1" (or "0"), when i is When the number is even, output "0" (or "1"), that is, a true random code output.

The analog-to-digital converter digitizes the noise signal, and then divides it into two groups according to its parity, the odd group corresponds to the random code "1" (or "0"), and the even group corresponds to the random code "0" (or "1") . The lowest bit of the analog-to-digital converter represents whether the noise signal is in an odd or even interval, so the analog-to-digital converter is used to realize the functions of the comparator array and the decoder.

Through multi-threshold screening, the information entropy of the output random code can be automatically stabilized. According to the calculation formula of information entropy:

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

In the formula, p( _xi ) is the probability of "0" and "1" in the random code. When "0" and "1" are equally likely to appear, the entropy of the random code reaches the maximum, and there is the maximum uncertainty. The multi-threshold screening method can keep the information entropy at a large value, and its analysis is as follows:

The probability that the input noise voltage falls in the i-th interval is

$\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; \&Integral;</span>}_{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; i</span>}{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}}^{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; dV</span>}$

When n is large enough, that is, D _v is small enough,

p[V _min +(i-1)D _v ]≈p(V _min +iD _v )

but

P(i)≈D _v p(V _min +iD _v )

so,

P(i)≈P(i+1)

From the above, suppose the probability that the input noise voltage falls in the interval where i is odd is P _O , and the probability that it falls in the even interval is P _C , then

P _O =P _C and when the elements in the sample space of the random variable appear equally, the information entropy is the largest. In this way, when n is large enough, a random code containing the maximum amount of information can be obtained. For 16-bit interval division, without any compensation or adjustment circuit, the entropy value of a random number can reach more than 0.99.

The normalized autocorrelation coefficient of the output random code can be reduced by multi-threshold screening. The correlation coefficient of the random code is jointly determined by the randomness of the physical random source and the system transfer function of the transformation system.

The traditional single-threshold discrimination circuit uses threshold crossing events to generate random code output, and the average number of positive threshold crossings per unit time

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; \&Integral;</span>}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&infin;</span>}\stackrel{<span\; class="notranslate">\; \&CenterDot;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; ,</span>\stackrel{<span\; class="notranslate">\; \&Center\; Dot;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; d</span>}\stackrel{<span\; class="notranslate">\; \&Center\; Dot;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}$

For a stationary Gaussian random process

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\stackrel{<span\; class="notranslate">\; \&Center\; Dot;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;\&pi;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; ]</span>$

For a stationary Gaussian random process, the average number of positive threshold crossings per unit time is equal to the average number of negative threshold crossings, so the average number of total threshold crossings per unit time is 2 N.

For a stationary random process within any time interval τ=|t _t -t ₁ |, the correlation coefficient has nothing to do with the time, but is only a function of the time interval.

The correlation function of the output random code is

${\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;\&tau;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}$

$<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; \&tau;</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\frac{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; !</span>}<span\; class="notranslate">\; +</span>\frac{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 4</span>}}}{\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; !</span>}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ]</span>$

$<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}\mathrm{<span\; class="notranslate">\; \&tau;</span>}}$

Its normalized autocorrelation function is

r _x (τ)＝e ^{-4 Nτ}

In the formula N is the number of positive threshold crossings per unit time, for a stationary Gaussian process, when the threshold

When V _a = m _x , N = f ₀

In the formula, f ₀ is the average frequency of Gaussian white noise. F

r _x (ττ＝exp[-4f ₀ τ]

It can be seen from the above formula that the normalized autocorrelation function of the output random code is a function of the period τ of the random code and the average frequency f ₀ of Gaussian white noise.

In engineering applications, when τ is large to a certain extent, if r _x (τ) is already small, it can be approximately considered that there is no correlation between X(t) and X(t+τ), and they are statistically independent. The correlation time τ ₀ is introduced here. When τ>τ ₀ , it can be considered that there is no correlation between X(t) and X(t+τ). τ ₀ is defined as

|r _x (τ ₀ )|≤0.05

From the above analysis, we can see that when the sampling time of the single-threshold random code generator is greater than τ ₀ , that is

When f ₀ ·τ≥0.749, it can be considered that the bits of the output random code sequence are irrelevant, that is, the output is a true random code. However, when τ ₀ is too large, the generation rate of random codes is limited, and it needs to be solved by increasing the bandwidth of the random noise signal source, which puts forward higher requirements for the random noise signal source.

The multi-threshold discrimination circuit has multiple threshold voltages, the number of Gaussian white noise falling in the i-th area and the positive-threshold crossing at the i+1th threshold voltage per unit time is

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; \&Integral;</span>}_{{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}}\mathrm{<span\; class="notranslate">\; dV</span>}{<span\; class="notranslate">\; \&Integral;</span>}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{\frac{{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}}{\mathrm{<span\; class="notranslate">\; \&Delta;t</span>}}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; ,</span>\stackrel{<span\; class="notranslate">\; \&Center\; Dot;</span>}{\mathrm{<span\; class="notranslate">\; V</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; d</span>}\stackrel{<span\; class="notranslate">\; \&CenterDot;</span>}{\mathrm{<span\; class="notranslate">\; V</span>}}$

When the breakover voltage D _v (i.e. resolution) is not very small, approximately

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; v</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; V</span>}}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

The average number of total positive threshold crossings per unit time is

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}$

Its normalized autocorrelation function is

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;\&tau;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; ]</span>$

because

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; ></span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; N</span>}}$

Therefore, when a multi-threshold comparator is used for discrimination, the normalized autocorrelation function of the output random code decays with a period τ much faster than when a single threshold is used. Therefore, at the same sampling frequency, the correlation coefficient of the random code generated by the multi-threshold discrimination circuit is small. Correlation time τ ₀ is correspondingly greatly reduced.

较单门限时大很多，所以相关时间τ₀减小，随即码的采样点可以变得密集。The code rate of the output random code is improved through multi-threshold screening. The maximum output rate of the random code depends on the correlation time τ ₀ of the output random code. For the multi-threshold voltage screening method, when the average frequency f ₀ of Gaussian white noise remains unchanged, because

Much larger than that of a single threshold, so the correlation time τ ₀ decreases, and then the sampling points of the code can become dense.

By combining the outputs of multiple random code generators, the code rate of the output random code is increased. Because n random code generators generate n random code sequences, one bit is sequentially taken from each of the n random code sequences each time to the final output random code sequence, and so on, so that the n random code sequences are combined into one Random code sequence. In this way, the output code rate is n times the output code rate of a single random code generator.

By combining the outputs of multiple random code generators, the randomness of the output random code is improved. Because n random code generators generate n random code sequences, one bit is sequentially taken from each of the n random code sequences each time to the final output random code sequence, and so on, so that the n random code sequences are combined into one Random code sequence. In this way, under the condition of the same output code rate, the code rate of a single random code generator is reduced to one-nth of the original, the sampling time interval τ is increased to n times of the original, and the normalized correlation coefficient is greatly reduced. Sex is improved.

Compared with the prior art, the present invention has the following advantages and beneficial effects:

(1) the random code that the present invention outputs is true random code, and the information entropy of the random code of output can be automatically stabilized, and the bandwidth requirement to noise source is lower;

(2) the normalized correlation coefficient of the random code that the present invention outputs can reduce greatly by increasing the quantity of threshold voltage and increasing the number of the random code generator unit of combining output, close to zero;

(3) the present invention is by increasing the number of the random code generator unit of merging output, and the random code of output can reach great code rate;

(4) The present invention works stably, is cheap, has good expansibility, and is convenient for integration.

Description of drawings

Fig. 1 is the structural representation of the device that produces true random code of the present invention;

Fig. 2 is the structural representation of the random code generator unit of device of the present invention;

Fig. 3 is another structural schematic diagram of the random code generator unit of the device of the present invention.

Detailed ways

In order to better understand the present invention, the present invention will be further described below in conjunction with the accompanying drawings.

As shown in Figure 1, the device that produces true random code of the present invention comprises noise generator, several multi-threshold random code generators, data processor, and noise generator is connected with each multi-threshold random code generator respectively, forms several A random code generator unit, each multi-threshold random code generator is connected to the data processor.

As shown in Figure 2, the multi-threshold random code generator is an analog-to-digital converter, and the analog-digital converter outputs with the lowest bit; as shown in Figure 3, the multi-threshold random code generator includes a comparator array and a decoder, and the noise The generator, comparator array, decoder, and data processor are connected in sequence; the comparator array is used in parallel with several comparators, the noise generator is connected to each comparator separately, and the comparator array outputs the corresponding TTL level. After being decoded by the decoder, a binary true random code sequence is obtained.

The data processor in Figure 1 can be a parallel-to-serial converter or a data acquisition card.

Example 1

The random noise source is a broadband white noise source, the model is HP346, the conversion system uses a comparator array, the comparator array is composed of eight comparators, the model is max961, the decoder uses 741s148, and the output is a true random code sequence, and its code rate is 1Mbit /s.

Example 2

The random noise source is a broadband white noise source, the model is HP346; the conversion system uses an analog-to-digital converter, the model is max150; the parallel-to-serial converter is a 16:1 converter, the model is max3891; the white noise signal output by HP346 is passed through max150 Multi-threshold screening, output a true random code sequence, the code rate is 1Mbit/s. 16 random code sequences generated by HP346 and max150 are respectively added to 16 parallel input ports of max3891, and at the serial output port of max3891, a random code output with a code rate of 16Mbit/s is obtained.

Example 3

The random noise source is a broadband white noise source, the model is HP346; the conversion system uses an analog-to-digital converter, the model is max150, and the random code sequence is input through the data acquisition card, and the model of the data acquisition card is AD7812L. The white noise signal output by HP346 passes through the multi-threshold screening of max150, and outputs a true random code sequence with a code rate of 1Mbit/s. The random code sequences generated by 8 HP346 and max150 are respectively added to the 8 digital input ports of AD7812, and one bit is sequentially taken from each of the 8 random code sequences to the final output random code sequence, and so on. Eight random code sequences are combined into one random code sequence. At the digital output port of AD7812L, a random code output with a code rate of 8Mbit/s is obtained.

Claims (10)

1. device that produces real random code, comprise noise generator, it is characterized in that also comprising several multi-threshold random code generators, data processor, noise generator links to each other respectively with each multi-threshold random code generator, form several random code generator unit, each multi-threshold random code generator links to each other with data processor.

2. the device of generation real random code according to claim 1 is characterized in that described multi-threshold random code generator is an analog-digital converter, with the lowest order of analog-digital converter as output.

3. the device of generation real random code according to claim 1 is characterized in that described multi-threshold random code generator comprises comparator array and demoder, and noise generator, comparator array, demoder, data processor connect successively.

4. the device of generation real random code according to claim 1 is characterized in that described data processor is parallel-to-serial converter or data collecting card.

5. the device of generation real random code according to claim 3 is characterized in that described comparator array is that several comparers are connected in parallel, and noise generator links to each other respectively with each comparer.

6. produce the method for real random code with the described device of claim 1, it is characterized in that process is as follows:

After screening by the multi-threshold random code generator, the level signal at random that noise generator is produced obtains binary real random code sequence, then in each cycle, binary real random code sequence that several random code generator unit produce is respectively got one, merge into a binary real random code sequence output by data processor.

7. the method for generation real random code according to claim 6, it is characterized in that the examination of the process of the level signal at random comparator array that described noise generator produces, comparator array is exported corresponding Transistor-Transistor Logic level, through obtaining binary real random code sequence after the decoder decode.

8. the method for generation real random code according to claim 6, it is characterized in that the examination of the process of the level signal at random analog-digital converter that described noise generator produces, as output, obtain binary real random code sequence with the lowest order of analog-digital converter.

9. the method for generation real random code according to claim 6 is characterized in that the noise signal digitizing, is divided into two groups according to its parity then, the corresponding random code " 1 " of odd number group, the corresponding random code " 0 " of even number set; The perhaps corresponding random code " 0 " of odd number group, the corresponding random code " 1 " of even number set.

10. the method for generation real random code according to claim 6 is characterized in that described data processor is parallel-to-serial converter or data collecting card.

Images