Random number generator
A multi-entropy source system with integrated random number generation and easy component replacement addresses stability and power efficiency issues in physical random number generators, ensuring reliable operation and versatility.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Y D KKK
- Filing Date
- 2026-04-14
- Publication Date
- 2026-06-25
AI Technical Summary
Physical random number generators face challenges in ensuring stable random number generation over a long period due to potential malfunctions or failures, and they are limited by high power consumption and low mass-production capabilities, making them less widely adopted compared to pseudo-random number generators.
A system comprising multiple sub-substrates with entropy sources and a main substrate that integrates and diversifies entropy sources, incorporating a mix-entropy generation method to enhance safety and speed up random number generation, while allowing for easy replacement of malfunctioning components.
The system ensures stable random number generation over a long duration despite failures, reduces power consumption, and supports easy replacement of malfunctioning components, thereby improving reliability and versatility.
Smart Images

Figure 2026104952000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a technique of a random number generation device that generates random numbers using a physical random number source.
Background Art
[0002] In the field of information security, techniques for protecting digital data by utilizing random numbers are frequently used. Currently, the most widespread is a "pseudo-random number generation device" that uses an algorithm for generating random numbers. However, the pseudo-random number generation device uses an algorithm artificially created for generating random numbers. Therefore, there is a risk (vulnerability) that the seed (the seed input into the algorithm) may be guessed or the algorithm itself may be elucidated if enough time is spent.
[0003] On the other hand, in contrast to the pseudo-random number generation device that uses an algorithm for random number generation, there is a "physical random number generation device 20" that uses a physical phenomenon as shown in the block diagram of FIG. 1. It includes an entropy source 22 that obtains electrical energy supplied from a power supply 21 and generates physical random numbers. As this entropy source, there are mainly types that include an entropy source using thermal noise and types that include an entropy source using quantum random numbers. In an environment where security needs to be ensured over a long period, a pseudo-random number generation device that leaves regularity is unsuitable, and it is desirable to adopt a physical random number generation device.
[0004] Physical phenomenon random number sequences such as thermal noise and quantum noise generated by the entropy source 22 are tested by a first random number testing device 23. If it is "OK" by the first random number testing device 23, the physical phenomenon random number sequence is output to an FPGA 24 which is a programmable logic device. When it becomes "NG", the output to the FPGA 24 is stopped and a notification is made that an abnormality has occurred.
[0005] The FPGA 24 is equipped with a distillation processing device 25. The "distillation process" performed by this distillation apparatus 25 is a process that improves the quality of random numbers (physical phenomenon random numbers) generated by the entropy source 22, and is carried out by a circuit that performs matrix operations called an extractor. The signal output by the entropy source 22 may be imperfect, so the process is performed to compensate for the degradation caused by that imperfection and to improve the quality.
[0006] The signal (output random number) output from the programmable logic device is passed through the second random number verification device 26. If the second random number verification device 26 determines it to be "OK", the converted random number is output to the external device 40 via the output interface 27. If the result is "NG", output to output interface 27 is stopped and a notification is sent indicating that an error has occurred.
[0007] The aforementioned physical random number generator can produce random number sequences that are close to ideal random numbers (true random numbers). Therefore, random numbers generated by a physical random number generator have higher security performance than those generated by a pseudorandom number generator. However, it has drawbacks such as being expensive due to its low mass-production capability and having limitations in reducing the overall physical size of the device. For this reason, pseudo-random number generators are currently more widely used.
[0008] Now, if the random seed is insufficient, the security of the encryption system may be lost, or the sampling results may become inaccurate. Patent Document 1 discloses a random number generator that ensures the accuracy of the sampling results without compromising security.
[0009] Physical random number generators equipped with an entropy source are known to consume more power than pseudorandom number generators. Patent document 2 discloses a random number generator capable of efficiently generating Bernoulli sequences while suppressing power consumption compared to conventional devices. [Prior art documents] [Patent Documents]
[0010] [Patent Document 1] Patent No. 7006887 [Patent Document 2] Special Publication No. 2016-513313 [Disclosure of the Invention] [Problems that the invention aims to solve]
[0011] Figure 2(a) shows an overview of the conventional technology and its challenges. In other words, a physical random number generator has a complex process of securing the necessary power supply for the entropy source, generating physical phenomena in the entropy source, and then extracting those physical phenomena to generate random numbers. As a result, the possibility of malfunction or failure is higher compared to a pseudo-random number generator. On the other hand, environments requiring physical random number generators often necessitate long-term continuous use. Therefore, it is desirable to ensure stable random number generation even in the event of malfunctions or failures.
[0012] Figure 2(b) illustrates the different types of entropy sources. However, domestic products utilizing physical phenomena as entropy sources have limited versatility. Furthermore, coupled with declining demand, there are concerns about the continued supply. On the other hand, with overseas products, it is necessary to consider supply-side risks and reliability (randomness) risks from the perspective of their strategic product aspect in terms of information security.
[0013] The problem that this invention aims to solve is to provide a physical random number generator and control method that can stably ensure random number generation over a long period of time, even when encountering malfunctions or failures. [Means for solving the problem]
[0014] To solve the aforementioned problems, the present invention comprises a plurality of sub-substrates (20A, 20B, 20C) equipped with an entropy source for generating a sequence of random numbers representing physical phenomena, and a main substrate (10B) that merges the sequences of random numbers representing physical phenomena generated by the plurality of sub-substrates (20A, 20B, 20C) and integrates the merged random numbers (see Figure 3).
[0015] The system can accommodate multiple entropy sources (multi-entropy generation) and multiple types of entropy sources simultaneously (mix-entropy generation). This diversifies the procurement of entropy sources, improving the safety of random numbers and speeding up random number generation.
[0016] (First invention) The first invention comprises multiple random number output boards (20A, 20B) that generate a sequence of random numbers based on physical phenomena, The present invention relates to a random number generator (10A) equipped with a main board (10B) that enables the random number sequences of physical phenomena output by the multiple random number output boards (20A, 20B) to be used as physical random numbers in an external device (40). The aforementioned random number output boards (20A, 20B) are, An entropy source (12A, 12B) that generates a random sequence of numbers based on physical phenomena, The entropy source (12A, 12B) has a power regulator (11A, 11B) to adjust the electrical energy it receives from the power supply (11) so that it can output a random sequence of physical phenomena, The I / F conversion unit (13A, 13B) outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source (12A) as a converted signal so that it can be used on the main board (10B), It is equipped with. The aforementioned main board (10B) is The first random number verification device (15A, 15B) verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from the aforementioned I / F conversion unit (13A, 13B), can be used as random numbers. When the first random number testing device (15A, 15B) tests and it is a conversion signal that can be used as a random number, a data processing unit (16) that mixes or maxes the physical phenomenon random number sequence as the conversion signal as random number data; A distillation processing device (17) that performs a distillation process with an extractor on the physical phenomenon random number mixed or maxed by the data processing unit (16) to obtain an output random number in order to improve the safety as a random number; A second random number testing device (18) that tests whether the output random number can be used as a random number by the external device (40); An external interface (19) that converts the tested output random number tested by the second random number testing device (18) into a converted output random number that can be used by the external device (40); It is provided (refer to FIGS. 4, FIGS. 5, FIGS. 6).
[0017] (Term Explanation) Regarding the "random number output board (20A, 20B)", "plural" means two (refer to FIG. 4) or more (refer to FIG. 13). The "I / F conversion unit" in the random number output board (20A, 20B, 20C) as a sub-board is a part that converts electrical characteristics, timing, data format, etc. so that they can be used on the main board (10B). The "I / F conversion unit" on the main board (10B) receives a signal for identifying which of the plurality of random number output boards (20A, 20B, 20C) it is connected to. Also, through the "I / F conversion unit" in the random number output board (20A, 20B, 20C), initialization and output control for the entropy source in the random number output board (20A, 20B) are executed (refer to FIG. 4).
[0018] The "extractor" is a device (circuit) that performs post-processing to improve the quality of the random number sequence generated by the entropy source. The "distillation process" is to improve the randomness (random number property) of the output random number by reducing the number of bits of the output random number itself. A specific example of the "distillation process" is to extract some bits from a physical random number and multiply them by an extractor matrix in which 0s or 1s are arranged in a matrix to generate an output random number. Using the Hash-Based DRBG algorithm on the signals output by the random number output boards (20A, 20B) to the main board (10B) ensures that each output signal has a 50% probability.
[0019] "Mixing" refers to a method of enhancing security by using the exclusive OR of multiple conversion signals output from multiple random number output boards (20A, 20B) to the main board (10B). This provides a level of security similar to a one-time pad, which is an encryption method that uses a random number sequence only once. For example, even if one of the random number output boards fails for some reason, the randomness of the other board will not be compromised as long as it maintains its randomness.
[0020] "Maximizing" refers to maximizing the generation speed of multiple conversion signals output by multiple random number output boards (20A, 20B) to the main board (10B). For example, the generation speed can be maximized by using time-division multiplexing.
[0021] The term "mixing or maximizing" means that either mixing or maximizing is selectively adopted. Mixing and maximizing are never performed simultaneously. The selection between mixing and maximizing is generally handled by an automatic switching process (see Figure 12). For example, if all of the multiple conversion signals output by multiple random number output boards (20A, 20B) to the main board (10B) are normal, maximizing the generation speed is more beneficial, so maximizing is selected.
[0022] The ability to "mix or maximize" the random number generation process yields two benefits. Firstly, it increases the speed at which random numbers are generated. Secondly, even if one physical random number chip malfunctions (for example, due to poor electrical contact), the other chip can still generate random numbers. This contributes to increasing the reliability of the physical random number generator according to the present invention.
[0023] (action) The power regulators (11A, 11B) adjust the electrical energy obtained from the power supply (11). The entropy sources (12A, 12B) that receive the electrical energy adjusted by the power regulators (11A, 11B) generate a random sequence of physical phenomena based on physical phenomena. The I / F conversion unit (13A, 13B) outputs the random sequence of physical phenomena output by the entropy source (12A) as a converted signal so that it can be used on the main board (10B). The first random number verification device (15A, 15B) verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from the I / F conversion unit (13A, 13B), can be used as random numbers. If the first random number verification device (15A, 15B) verifies that the converted signal can be used as a random number, the data processing unit (16) mixes or maxes out the sequence of random numbers representing physical phenomena as a converted signal for use as random number data. The data processing unit (16) mixes or maximizes the random numbers representing physical phenomena, and the distillation processing unit (17) performs a distillation process on these random numbers via an extractor. This distillation process results in output random numbers with improved safety from the mixed or maximized random numbers representing physical phenomena. The second random number verification device (18) verifies whether the output random numbers can be used as random numbers by the external device (40). The external interface (19) converts the verified output random numbers verified by the second random number verification device (18) into physical random numbers (converted output random numbers) that can be used by the external device (40).
[0024] (Variation 1 of the first invention) The main board (10B) is equipped with sub-board I / F conversion units (14A, 14B) that convert the conversion signals output by the I / F conversion units (13A, 13B) in the random number output boards (20A, 20B) so that they can be used on the main board (10B). The I / F conversion units (13A, 13B) in the aforementioned random number output boards (20A, 20B) are configured to standardize the signals output to the aforementioned sub-board I / F conversion units (14A, 14B).
[0025] (action) The signals output to the sub-board I / F conversion unit (14A, 14B) are standardized by the I / F conversion unit (13A, 13B) on the random number output board (20A, 20B). Therefore, even if the types of entropy sources (12A, 12B) used in the random number output board (20A, 20B) are different, they can still be used on the main board (10B). As a result, a highly versatile physical random number generator can be provided.
[0026] (Variation 2 of the first invention) The first invention may be formed as follows: In other words, the outer periphery of the side of the main board (10B) that is equipped with the first random number verification device (15A, 15B) is provided with an outer periphery frame (30) erected in a direction that covers the first random number verification device (15A, 15B). The outer frame (30) has a height dimension that covers the random number output boards (20A, 20B) connected to the first random number verification device (15A, 15B), The structure is designed to seal the random number output boards (20A, 20B) by providing an outer cover material (33) that covers the surface facing the main board (10B) with the outer frame (30) in between (see Figures 19 and 20).
[0027] (action) The side of the main board (10B) that houses the first random number verification device (15A, 15B) is sealed by the outer frame (30) and the outer cover material (33). As a result, the random number output boards (20A, 20B) are not exposed and are surrounded by the back surface of the main board (10B), the outer surface of the outer frame (30), and the top surface of the outer cover material, thus protecting the entropy sources (12A, 12B).
[0028] (Second invention) The second invention relates to a random number generator (10A) that replaces the random number output boards (20A, 20B) in the random number generator (10A) of the first invention with a plurality of slots (20M) on which random number output boards (20A, 20B) can be mounted. In other words, it is equipped with multiple slots (20M) that can accommodate random number output boards (20A, 20B) that generate random number sequences based on physical phenomena, The present invention relates to a random number generator (10A) equipped with a main board (10B) that enables the conversion signals for random number generation output by the random number output boards (20A, 20B) to be used as random numbers by an external device (40) when the random number output boards (20A, 20B) are mounted in the aforementioned slot (20M). The aforementioned random number output boards (20A, 20B) include an entropy source (12A, 12B) that generates a sequence of random numbers based on physical phenomena, The entropy source (12A, 12B) has a power regulator (11A, 11B) to adjust the electrical energy it receives from the power supply (11) so that it can output a random sequence of physical phenomena, The I / F conversion unit (13A, 13B) outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source (12A) as a converted signal so that it can be used on the main board (10B), It is equipped with. The main board (10B) includes a first random number verification device (15A, 15B) that verifies whether the physical phenomenon random numbers output as conversion signals from the I / F conversion unit (13A, 13B) can be used as random numbers, The first random number verification device (15A, 15B) verifies that the converted signal can be used as a random number, and the data processing unit (16) mixes or maximizes the physical phenomenon random number as random number data. The data processing unit (16) performs a distillation process via an extractor to convert the mixed or maxed physical phenomenon random numbers into output random numbers, thereby improving their safety as random numbers, and the distillation processing unit (17) A second random number verification device (18) that verifies whether the aforementioned output random numbers can be used as random numbers by an external device (40), The second random number verification device (18) has an external interface (19) that converts the output random numbers verified by the external device (40) into converted output random numbers that can be used by the external device (40), It is equipped with (see Figure 9).
[0029] (action) By installing the random number output board (20A, 20B) in slot (20M), it will function similarly to the random number generator (10A) according to the first invention. If either of the random number output boards (20A, 20B) stops working, you can remove the non-functioning board (20A, 20B) and replace it to restore functionality.
[0030] (Variation 1 of the second invention) The main board (10B) is equipped with sub-board I / F conversion units (14A, 14B) that convert the conversion signals output by the I / F conversion units (13A, 13B) in the random number output boards (20A, 20B) so that they can be used on the main board (10B). The I / F conversion unit (13A, 13B) in the random number output board (20A, 20B) mounted in the aforementioned slot (20M) is configured to standardize the output to the sub-board I / F conversion unit (14A, 14B).
[0031] (action) The random number output boards (20A, 20B) installed in multiple slots (20M) can be mixed and matched.
[0032] (Variation 2 of the second invention) The second invention may also be formed in the same way as the first invention, as follows: In other words, the outer periphery of the side of the main board (10B) that is equipped with the first random number verification device (15A, 15B) is provided with an outer periphery frame (30) erected in a direction that covers the first random number verification device (15A, 15B). The outer frame (30) has a height dimension that covers the random number output boards (20A, 20B) connected to the random number verification devices (15A, 15B), The structure is designed to seal the random number output boards (20A, 20B) by providing an outer cover material (33) that covers the surface facing the main board (10B) with the outer frame (30) in between (see Figures 16 and 17).
[0033] (Third invention) The third invention comprises multiple random number output boards (20A, 20B) that generate a sequence of random numbers based on physical phenomena, The present invention relates to a method for controlling a random number generator (10A) equipped with a main board (10B) that enables the random number sequences of physical phenomena output by multiple random number output boards (20A, 20B) to be used as random numbers in an external device (40). The aforementioned random number output boards (20A, 20B) are, An entropy source (12A, 12B) that generates a random sequence of numbers based on physical phenomena, The entropy source (12A, 12B) has a power regulator (11A, 11B) to adjust the electrical energy it receives from the power supply (11) so that it can output a random sequence of physical phenomena, The system includes an I / F conversion unit (13A, 13B) that outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source (12A) as a converted signal so that it can be used on the main board (10B). The aforementioned main board (10B) is The first random number verification device (15A, 15B) verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from the aforementioned I / F conversion unit (13A, 13B), can be used as random numbers. The first random number verification device (15A, 15B) verifies that the converted signal can be used as a random number, and the data processing unit (16) mixes or maximizes the sequence of physical phenomena random numbers as the converted signal into random number data. The data processing unit (16) performs a distillation process via an extractor to convert the mixed or maxed physical phenomenon random numbers into output random numbers, thereby improving their safety as random numbers, and the distillation processing unit (17) A second random number verification device (18) that verifies whether the output random numbers can be used as random numbers by the external device (40), The system includes an external interface (19) that converts the verified output random numbers verified by the second random number verification device (18) into converted output random numbers that can be used by the external device (40). The control method for the random number generator (10A) is as follows: If the conversion signal from any of the aforementioned random number output boards (20A, 20B) to the I / F conversion unit (13A, 13B) is verified by the first random number verification device (15A, 15B) as a normal output, the data processing unit (16) selects the maximization process. If the first random number testing device (15A, 15B) detects an abnormality in either output, the data processing unit (16) will select the mixing process.
[0034] (Variation of the third invention) In the third invention, the following may also be done. In other words, if the output of the conversion signal to the I / F conversion unit by any of the random number output boards in the random number verification device is delayed by a predetermined time, the data processing unit may select the maximization process. [Effects of the Invention]
[0035] According to the first invention, we were able to provide a physical random number generator that can stably ensure the generation of random numbers over a long period of time, even if it encounters malfunctions or failures. According to the second invention, it was possible to provide a physical random number generator that allows for easy replacement of the random number output board if it malfunctions or fails. The third invention provides a control method for a physical random number generator that can stably ensure the generation of random numbers over a long period of time, even when it encounters malfunctions or failures. [Brief explanation of the drawing]
[0036] [Figure 1] This is a block diagram showing the basic structure of a conventional physical random number generator. [Figure 2] (a) is an overview and challenges of the conventional technology, and (b) is a comparison table showing the type of entropy source, manufacturer, etc., and performance. [Figure 3] A branch diagram outlining the present invention. [Figure 4] This is a block diagram showing the physical random number generator (basic form) related to the present application. [Figure 5] This is a block diagram detailing the internal workings of a physical random number generator, specifically an entropy source. [Figure 6] This is a block diagram of the physical random number generator (Variation 1) according to the present application. [Figure 7] This is a block diagram of the physical random number generator (variation 2) according to the present application. [Figure 8] The following are examples of output from the data processing unit, where (a) shows automatic switching and (b) shows an increase in data volume. [Figure 9] The following are examples of mixed output from the data processing unit: (a) shows the case where one output board is not functioning, and (b) shows the case where one output is delayed. [Figure 10] These are examples of output from the data processing unit; (a) shows mixing when one output board is not functioning, and (b) shows maximization when one output board is delayed. [Figure 11] This is a flowchart showing the process from random number generation to output. [Figure 12] This is a block diagram showing a physical random number generator, indicating that it has a slot for replacing the random number output board. [Figure 13] This is a block diagram showing an embodiment of the physical random number generator according to the present application, comprising three physical random number output boards. [Figure 14] This is a conceptual diagram illustrating that a physical random number generator equipped with three physical random number output boards can function even if the vendors of the physical random number output boards are different or mixed. (a) shows an example of multi-vendor configuration of the entropy source, and (b) shows an example of mixed-vendor configuration of the entropy source. [Figure 15] This is a block diagram showing a random number generator according to the present invention, in which only the entropy source in the random number output board can be replaced. [Figure 16] This is an assembled perspective view showing the main components of the physical random number generator according to the present invention. [Figure 17] This is a perspective view showing the assembled physical random number generator according to the present invention. [Modes for carrying out the invention]
[0037] The present invention will be described below based on embodiments. The drawings used herein are Figures 3 to 17. Figures 1 and 2 may be referred to as necessary.
[0038] (Figure 3) Figure 3 shows the following structure and function. In other words, random numbers are generated from entropy sources A, B, and C on the three sub-boards, converted into signals via I / F conversion units A, B, and C (omitted in this diagram), and sent to the main board. These signals are then sent to the data processing unit via the I / F conversion unit (omitted in this diagram) of the main board as random number data, where they are mixed or maxed out, and then distilled to become output random numbers. These converted output random numbers are then output to external devices via an external interface (shown in Figure 4, etc.) which is not shown in the diagram.
[0039] The random number output boards (sub-boards 20A, 20B, 20C) connected to and mounted on the main board (labeled "10B" in Figure 4) are made multiple and multi- and mixed to function with multiple types of entropy sources.
[0040] Multi- and mixed approaches make it possible to overcome problems such as the difficulty in obtaining entropy sources and absorb the need for ELCs, etc. Furthermore, by incorporating multiple entropy sources, safety can be improved through exclusive OR operations, and speed can be increased through redundancy. These points will be discussed later.
[0041] (Figure 4) Figure 4 shows a physical random number generator according to an embodiment of the present invention. This diagram shows a random number generator 10A that generates random numbers based on an entropy source, a power supply 11 that provides electrical energy to the random number generator 10A, and an external device 40 that uses the random numbers generated by the random number generator 10A.
[0042] The random number generator 10A consists of a main board 10B and two random number output boards 20A and 20B connected to the main board 10B. The random number output boards 20A and 20B are detachably attached to the random number generator 10A (detailed in Figure 15).
[0043] The random number output board 20A includes an entropy source 12A (for example, random number generation based on thermal noise), a power supply regulator 11A that adjusts the electrical energy supplied from the power supply 11 so that it can be used by the entropy source 12A in order to provide electrical energy to the entropy source 12A, and a main board I / F conversion unit 13A that converts it into a fixed signal required by the main board 10B. Furthermore, the process by which the random number sequence of physical phenomena generated by the entropy source 12A is processed will be explained in detail using Figure 5.
[0044] The random number output board 20B also has the same configuration as the random number output board 20A, and is equipped with an entropy source 12B, a power regulator 11B, and a main board I / F conversion unit 13B. The power regulator 11B is supplied with electrical energy directly from the power supply 11, and the random number output boards 20A and 20B are independent of each other.
[0045] The random number output boards 20A, 20B, and the main board 10B are designed on the premise that they will function even in a different manner than entropy source 12A, for example, if entropy source 12B performs random number generation based on quantum noise.
[0046] The main board 10B includes sub-board I / F units 14A and 14B that receive formatted signals from the main board I / F conversion units 13A and 13B of the random number output boards 20A and 20B, first random number verification devices 15A and 15B that verify the formatted signals received by the sub-board I / F units 14A and 14B, a data processing unit 16 that processes the formatted signals after verification, an extractor 17 that improves the safety of the data processed by the data processing unit 16, a second random number verification device 18 that verifies the data processed by the extractor 17, and an external interface 19 that adjusts the formatted data so that it can be used by an external device 40.
[0047] The data processing unit 16 is a device (circuit) that, when a converted signal is a converted signal that can be verified and used as a random number, mixes or maxes out the sequence of physical phenomenon random numbers as random number data. The mixing and maxing out processes will be described in detail using Figures 5 to 7. In this embodiment, the data processing unit 16 automatically selects whether to mix or max out the data.
[0048] The distillation processing apparatus (extractor) 17 is a device (circuit) that performs distillation processing. Distillation processing is the process of performing post-processing to improve the quality of the random number sequence generated by the entropy source. The signals output by the random number output boards 20A and 20B to the main board 10B may be incomplete, and processing is performed to compensate for the degradation caused by this incompleteness and to improve the quality.
[0049] (Figure 5) Figure 5 is a block diagram showing the sequence of processes, along with the equipment, in which the random number sequence of physical phenomena generated by the entropy source is processed before being used by external equipment, and details the entropy source 12B.
[0050] The entropy source 12B is generated by a laser light generator (abbreviated as "laser" in the figure) that irradiates a beam splitter with laser light, which then passes through an optically balanced receiver to extract quantum noise as a sequence of physical random numbers. The extracted quantum noise is converted into a usable conversion signal on the main board 10B via the I / F conversion unit 13B. This conversion signal is then sent to the data processing unit 16 as a verified conversion signal via the I / F conversion unit 14B and the first random number verification device 15B on the main board 10B.
[0051] Thermal noise is extracted from the entropy source 12A and converted into a usable conversion signal on the main board 10B via the I / F conversion unit 13A. This conversion signal is then sent to the data processing unit 16 as a verified conversion signal via the I / F conversion unit 14A and the first random number verification device 15A on the main board 10B.
[0052] Upon receiving the two verified conversion signals, the data processing unit 16 performs a mix-up or maximization process, as detailed in Figures 8-10, to create random number data. The created random number data passes through the distillation processing unit 17 to become output random numbers, which are then verified by the second random number verification unit 18. These verified output random numbers are then output to the external device 40 as converted output random numbers via the external interface 19.
[0053] (Figure 6) Figure 6 shows variation 1 of the physical random number generator shown in Figure 4. The only difference from the physical random number generator shown in Figure 6 is the data processing unit 16.
[0054] If both entropy sources 12A and 12B are functioning correctly, safe random numbers can be provided to the external device 40 regardless of whether mixing or maximization is selected. Maximization allows the data processing unit 16 to generate random numbers faster than mixing. Therefore, the data processing unit 16 of the physical random number generator shown in Figure 6 indicates that it selects the maximization process when it confirms that both entropy sources 12A and 12B are functioning correctly.
[0055] (Figure 7) Figure 7 shows variation 2 of the physical random number generator shown in Figure 4. The difference from the physical random number generator shown in Figure 4 lies in the random number verification device. In other words, in Figure 4, the random number testing device is divided into a first random number testing device 15A, 15B and a second random number testing device 18. On the other hand, in Figure 7, only the output random numbers output from the distillation processing device 17 are passed through the second random number testing device 18. This has the advantage of saving implementation capacity.
[0056] As shown in Figure 4, the method of verifying the output from sub-board I / F sections 14A and 14B, and also verifying the compressed random numbers output from them, is called the "parallel type." As shown in Figure 7, the method of verifying only the output random numbers output from the distillation processing apparatus 17 (using the same verification circuit repeatedly) is called the "serial type."
[0057] (Figure 8) Figure 8(a) shows that when the converted signal obtained via the I / F conversion unit 14A and the first random number verification device 15A is "0x1537", and the converted signal obtained via the I / F conversion unit 14B and the first random number verification device 15B is "0x4291", the data processing unit mixes them and outputs random number data "0x1537 XOR 0x4291", i.e., "0x57A6".
[0058] Mixing refers to the process of enhancing safety by using exclusive OR to connect multiple conversion signals output by multiple random number output boards 20A and 20B to the main board 10B.
[0059] Figure 8(b) also shows a specific example of maximization by the data processing unit when the converted signal obtained via the I / F conversion unit 14A and the random number verification device 15A is "0x1537", and the converted signal obtained via the I / F conversion unit 14B and the random number verification device 15B is "0x4291".
[0060] Maximization refers to maximizing the generation speed of multiple conversion signals output by multiple random number output boards 20A and 20B to the main board 10B. For example, the generation speed can be maximized by time-division multiplexing. This can result in signals like "0x15374291" or "0x42911537" simply by concatenating "0x1537" and "0x4291," as well as randomly generated signals like "0x14223971."
[0061] Furthermore, it can generate random numbers in units of 2 bits, 4 bits, 8 bits, or other units required by the external device 40.
[0062] (Figure 9) Figure 9 shows an example of the mixing process when one output board is not functioning or when the output is delayed.
[0063] Figure 9(a) shows the case where the I / F conversion unit 14A and the first random number verifier 15A did not function (outputting ALL "0" or ALL "1"). That is, if the conversion signal obtained via the I / F conversion unit 14B and the first random number verifier 15B outputs "0x4291", the data processing unit 16 outputs random number data of "0x4291" or "0xBS6E".
[0064] Figure 15(b) shows that the converted signal obtained via the I / F conversion unit 14A and the first random number verification device 15A is "0x1537". However, this is the case when the converted signal obtained via the I / F conversion unit 14B and the first random number verification device 15B outputs "0x4291" with a delay compared to "1537". When the delayed converted signal becomes "0xA429", mixing results in the output of random number data "0x1537 XOR 0xA429", or "0xB112".
[0065] (Figure 10) Figure 10 shows an example of maximization processing when one output board is not functioning or when the output is delayed.
[0066] Figure 10(a) shows the case where the I / F conversion unit 14A and the random number verification device 15A do not function (outputting ALL "0" or ALL "1"). In this case, even if the data processing unit 16 performs maximization, it will output "0x00004291" or "0xFFFF4291", which will not be suitable for random number generation. Therefore, an algorithm is implemented that does not perform maximization if one output board does not function.
[0067] Figure 10(b) shows the case where the converted signal obtained via the I / F conversion unit 14A and the first random number verification device 15A is "0x1537", but the converted signal obtained via the I / F conversion unit 14B and the first random number verification device 15B is "0x4291", which is output later than "0x1537". If the delayed converted signal becomes "0xA429", maximizing will output a random number data sequence of "0x1537A429" or "0xA4291537".
[0068] Figures 15 and 16 illustrate the cases in a random number generator 10A equipped with two random number output boards 20A and 20B, specifically when one board is not functioning and when the output of one board is delayed. With a random number generator 10C equipped with three random number output boards 20A, 20B, and 20C as shown in Figure 22 later, it is possible to perform both mixed and maxed-out random number generation even if one of the three boards does not function or has a delayed output.
[0069] (Figure 11) Figure 11 is a flowchart showing a simple processing procedure for performing the first and second random number tests, and the hardware configuration corresponds to Figure 4.
[0070] The power is turned on, and electrical energy is supplied from power regulators A and B to entropy sources A and B. That is, a random number generation command is issued (S1, S1'). The first random number verification device checks whether the random number sequences generated by entropy sources A and B are normal or not (S2, S2'). If an abnormality occurs, the output of random numbers will be stopped (S7).
[0071] (Figure 12) Figure 12 shows that the physical random number generator 10A is designed to have multiple random number output boards that are interchangeable. In other words, the physical random number generator 10A, before the random number output board 20B is incorporated, has a slot 20M into which the random number output board 20B can be incorporated.
[0072] When the random number output board 20B is installed in slot 20M, the power regulator 11B is electrically connected to the power supply 11. At the same time, the main board I / F section 13B is electrically connected to the sub-board I / F section 14B on the main board 10B.
[0073] Figure 12 shows that if the random number output board 20B is malfunctioning, it can be easily replaced. Furthermore, the replacement random number output board does not need to be of the same type as the original, as long as it can output a conversion signal that the sub-board I / F section 14B can receive (i.e., the main board I / F section 13B is provided).
[0074] (Figure 13) The random number generator 10C shown in Figure 13 differs from the embodiments shown in Figures 4 and 7 in that it has three random number output boards 20A, 20B, and 20C (triple entropy source). In this invention, while two and three random number output boards are given as examples, it is also possible to provide a random number generator equipped with four or more random number output boards.
[0075] (Figure 14) Figure 14(a) shows that "multi-vendorization" is possible for the three slots, where the same entropy source from the same manufacturer is used. In other words, if an entropy source is readily available from any of the manufacturers, it can be uniformly adopted.
[0076] Figure 14(b) shows that the three slots can utilize a "mix-vendor" approach, employing different entropy sources from multiple manufacturers. In other words, instead of unifying on an entropy source from any single manufacturer, it is possible to mix and use entropy sources from readily available manufacturers.
[0077] (Figure 15) Figure 15 shows an embodiment equipped with an entropy source slot 12M that allows only the malfunctioning entropy source to be replaced. Only an entropy source B of the same type as the original entropy source B can be installed in this entropy source slot 12M. This is because the power regulator 11B and the main board I / F conversion unit 13B will not be replaced and will continue to be used after the replacement.
[0078] (Figure 16) Figure 16 is an assembled perspective view of the components that make up the physical random number generator 1 according to this embodiment. The top surface of the main board 10B, as mentioned above, incorporates electronic circuits (not shown in the diagram), and numerous pin headers 20N are erected for connecting to these electronic circuits.
[0079] As mentioned above, the outer frame 30 is a rectangular frame that is erected and fixed inside the outer edge of the top surface of the main board 10B, and its height is greater than the pin header 20N mentioned above. The outer frame 30 and the main board 10B are fixed together all around by soldering so that they cannot be easily separated.
[0080] The break plate 31 has eight bends, with the outermost horizontal portion, which is placed on the upper end of the outer peripheral frame 30 in the short-side direction, at both ends. It is equipped with a header support that supports the aforementioned pin header 20N while passing through it, and when assembled, it will be located above the main board 10B.
[0081] The two spaces formed by the break plate 31 are where the random number output boards 20A and 20B are located. Both the random number output boards 20A and 20B have a rectangular shape in plan, and are equipped with through-holes along their longer sides through which pin headers 20N are electrically connected. By passing the pin headers 20N through these through-holes, the random number output boards 20A and 20B are electrically connected to the main board 10B, even though they are separated by the break plate 31.
[0082] The present invention, by providing multiple random number output boards 20A and 20B, offers the following advantages. Firstly, by incorporating two identical physical random number chips (entropy sources), it becomes possible to process random number generation in parallel, effectively doubling the generation speed. Furthermore, even if one entropy source fails, random numbers can still be generated as long as the other entropy source is functioning correctly.
[0083] Secondly, by incorporating different types of physical random number chips (entropy sources), the reliability of the generated random numbers is improved. Furthermore, even if, for example, one entropy source is tampered with by someone without legitimate authority, parallel processing with other entropy sources can generate random numbers that eliminate the intent of tampering, thus ensuring security.
[0084] In this embodiment, it is assumed that random number output boards 20A and 20B are incorporated, but it can function even if only one physical random number chip is incorporated. Even if physical random number chips become difficult to obtain, this will contribute to stabilizing the supply to meet the demand for physical random number generators.
[0085] (Figure 17) After assembling the random number output boards 20A and 20B onto the break plate 31, the inner cover material 32 is fixed onto the break plate 31. Then, after resin injection (not shown in the diagram), the outer cover material 33 is fixed. The completed assembly is shown in Figure 17.
[0086] According to the random number generator of the embodiment described above, even if malfunctions or failures occur, stable random number generation can be ensured for a long period of time. Furthermore, if it becomes difficult to obtain an entropy source, the random number generator can be manufactured using an entropy source from a readily available manufacturer, or it can be manufactured using a mixture of entropy sources from multiple manufacturers. [Industrial applicability]
[0087] The present invention has applicability in industries such as the manufacturing of information and communication equipment, the information and communication service industry including the installation of information and communication equipment, and the software industry that creates computer software for information and communication services. [Explanation of Symbols]
[0088] 10; Random number generator 10A; Random number generator (double entropy source) 10B: Main board 10C: Random number generator (triple entropy source) 11;Power supply 11A;Power regulator A 11B;Power regulator B 12; Entropy source 12A; Entropy source A 12B; Entropy source B 12M; Entropy source slot 13A; Main board I / F conversion unit A 13B; Main board I / F conversion unit B 14A; Sub-board I / F conversion unit A 14B; Sub-board I / F conversion unit B 15A; First random number verification device 1 15B; First random number verification device 1 16; Multi / Mix device 17; Distillation apparatus (extractor) 18; Second Random Number Verification Device 19; External Interface 20A; Random number output board 20B; Random number output board 20C; Random number output board 20M; Slot 20N; pin header 21; power supply 22; Entropy source 23; First Random Number Verification Device 24; FPGA 25; Distillation Apparatus 26; Second Random Number Verification Device 27; Output Interface 30; outer frame 31; Break plate 32; Inner cover material 33; Outer cover material 40; External equipment
Claims
1. It is equipped with multiple random number output boards that generate random number sequences based on physical phenomena, A random number generator equipped with a main board that enables the sequence of physical phenomenon random numbers output by multiple random number output boards to be used as physical random numbers in an external device, The aforementioned random number output board is An entropy source that generates a random sequence of physical phenomena based on physical phenomena, The entropy source is a power regulator that adjusts the electrical energy obtained from the power source in order to output a random sequence of physical phenomena, An I / F conversion unit that outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source as a converted signal so that it can be used on the main board, Each of the multiple units is equipped with The aforementioned main board is A first random number verification device that verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from each of the aforementioned I / F conversion units, can be used as random numbers, When the first random number verification device has verified that there is only one conversion signal that can be used as a random number, a data processing unit mixes that one conversion signal, which is a sequence of random numbers based on physical phenomena, as random number data. The data processing unit performs a distillation process via an extractor to convert the mixed random numbers of physical phenomena into output random numbers in order to improve their safety as random numbers. A second random number verification device that verifies whether the output random numbers can be used as random numbers by the external device, An external interface that converts the verified output random numbers verified by the second random number verification device into converted output random numbers that can be used by the aforementioned external device, A random number generator equipped with the following features.
2. It is equipped with multiple random number output boards that generate random number sequences based on physical phenomena, A random number generator equipped with a main board that enables the sequence of physical phenomenon random numbers output by multiple random number output boards to be used as physical random numbers in an external device, The aforementioned random number output board is An entropy source that generates a random sequence of physical phenomena based on physical phenomena, The entropy source is a power regulator that adjusts the electrical energy obtained from the power source in order to output a random sequence of physical phenomena, An I / F conversion unit that outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source as a converted signal so that it can be used on the main board, Each of the multiple units is equipped with The aforementioned main board is A first random number verification device that verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from each of the aforementioned I / F conversion units, can be used as random numbers, When the first random number verification device has verified that there are multiple conversion signals that can be used as random numbers, a data processing unit maximizes the sequence of physical phenomenon random numbers that serve as multiple conversion signals as random number data, The data processing unit performs a distillation process via an extractor to convert the maximized random numbers of physical phenomena into output random numbers in order to improve their safety as random numbers. A second random number verification device that verifies whether the output random numbers can be used as random numbers by the external device, An external interface that converts the verified output random numbers verified by the second random number verification device into converted output random numbers that can be used by the aforementioned external device, A random number generator equipped with the following features.
3. The main board is equipped with a sub-board I / F conversion unit that converts the conversion signal output by the I / F conversion unit of the random number output board so that it can be used on the main board. The random number generator according to claim 1 or claim 2, wherein the I / F conversion unit in the random number output board standardizes the signal to be output to the sub-board I / F conversion unit.
4. On the outer periphery of the side of the main board on which the first random number verification device is installed, an outer frame is provided that is erected in a direction that covers the first random number verification device. The outer frame has a height dimension that covers the random number output board connected to the first random number verification device, The random number output board is enclosed by an outer cover material that covers the surface opposite to the main board, sandwiching the outer frame. The random number generator according to claim 1.
5. On the outer periphery of the side of the main board on which the first random number verification device is installed, an outer frame is provided that is erected in a direction that covers the first random number verification device. The outer frame has a height dimension that covers the random number output board connected to the first random number verification device, The random number output board is enclosed by an outer cover material that covers the surface opposite to the main board, sandwiching the outer frame. The random number generator according to claim 2.
6. It is equipped with multiple slots that can accommodate multiple random number output boards that generate random number sequences based on physical phenomena, A random number generator comprising a main board that enables the conversion signal for random number generation output by a random number output board to be used as a physical random number in an external device when a random number output board is installed in the aforementioned slot, The multiple random number output boards mounted in the multiple slots described above are an entropy source that generates a sequence of random numbers based on physical phenomena, The entropy source is a power regulator that adjusts the electrical energy obtained from the power source in order to output a random sequence of physical phenomena, An I / F conversion unit that outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source as a converted signal so that it can be used on the main board, Each random number output board is equipped with this, The aforementioned main board is A first random number verification device that verifies whether the physical phenomenon random numbers output as conversion signals from the aforementioned multiple I / F conversion units can be used as random numbers, When the first random number verification device has verified that there is only one conversion signal that can be used as a random number, the data processing unit mixes that one conversion signal, which is a random number of physical phenomena, into random number data. The data processing unit performs a distillation process via an extractor to convert the mixed random numbers of physical phenomena into output random numbers in order to improve their safety as random numbers. A second random number verification device that verifies whether the aforementioned output random numbers can be used as random numbers by an external device, The second random number verification device converts the output random numbers verified by it into converted output random numbers that can be used by the aforementioned external device, and A random number generator equipped with the following features.
7. It is equipped with multiple slots that can accommodate multiple random number output boards that generate random number sequences based on physical phenomena, A random number generator comprising a main board that enables the conversion signal for random number generation output by a random number output board to be used as a physical random number in an external device when a random number output board is installed in the aforementioned slot, The multiple random number output boards mounted in the multiple slots described above are an entropy source that generates a sequence of random numbers based on physical phenomena, The entropy source is a power regulator that adjusts the electrical energy obtained from the power source in order to output a random sequence of physical phenomena, An I / F conversion unit that outputs a sequence of random numbers representing physical phenomena output by the aforementioned entropy source as a converted signal so that it can be used on the main board, Each random number output board is equipped with this, The aforementioned main board is The aforementioned main board is A first random number verification device that verifies whether the sequence of random numbers representing physical phenomena, output as a conversion signal from the aforementioned multiple I / F conversion units, can be used as random numbers, When the first random number verification device has verified that there are multiple conversion signals that can be used as random numbers, a data processing unit maximizes the sequence of physical phenomenon random numbers that serve as multiple conversion signals as random number data, The data processing unit performs a distillation process via an extractor to convert the maximized random numbers of physical phenomena into output random numbers in order to improve their safety as random numbers. A second random number verification device that verifies whether the output random numbers can be used as random numbers by the external device, An external interface that converts the verified output random numbers verified by the second random number verification device into converted output random numbers that can be used by the aforementioned external device, A random number generator equipped with the following features.
8. The main board is equipped with a sub-board I / F conversion unit that converts the conversion signal output by the I / F conversion unit of the random number output board so that it can be used on the main board. The I / F conversion unit in the random number output board mounted in the aforementioned slot is configured to standardize the output to the sub-board I / F conversion unit. A random number generator according to claim 6 or claim 7.