Quantum random number generation rate enhancement apparatus and method

Abstract

This invention provides a device and method for enhancing the generation rate of quantum random numbers, primarily relating to the field of quantum information technology. The device includes: a quantum entropy source module for generating a first quantum random analog signal; an analog-to-digital conversion module, including an equalization filtering unit; the equalization filtering unit is used to: flatten the power spectral density of the first quantum random digital signal based on the target amplitude-frequency response characteristics, obtaining a filtered first quantum random digital signal; wherein the equalization filtering unit is loaded with filtering coefficients such that the target amplitude-frequency response characteristics of the equalization filtering unit are opposite to the amplitude-frequency response characteristics of the analog link, and that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal; and a processing module for performing randomness extraction processing on the filtered first quantum random digital signal, outputting a quantum random number sequence.

Description

Technical Field

[0001] This invention relates to the field of quantum information technology, and more specifically, to a device and method for enhancing the rate of quantum random number generation. Background Technology

[0002] Quantum random number generators utilize the intrinsic randomness of quantum mechanics to produce truly random numbers, and have important applications in fields such as cryptography. The continuous-variable vacuum state scheme has become one of the mainstream technical approaches due to its simple structure and high speed. However, the non-ideal frequency response of the analog link can introduce time dependence into the original data, leading to a decrease in minimum entropy and limiting the final real-time rate.

[0003] The proposed solution uses a deterministic extraction algorithm based on matrix multiplication for post-processing. However, when the quality of the original data is low, it requires a large amount of hardware computing resources with a low extraction ratio and introduces significant processing latency, which becomes the main bottleneck for the real-time performance of the system. Summary of the Invention

[0004] In view of this, the present invention provides a device and method for enhancing the rate of quantum random number generation.

[0005] One aspect of the present invention provides a quantum random number generation rate enhancement device, comprising: a quantum entropy source module for generating a first quantum random analog signal; an analog-to-digital conversion module electrically connected to the quantum entropy source module via an analog link, the analog-to-digital conversion module including an equalization filtering unit; the analog-to-digital conversion module being configured to: perform analog-to-digital conversion on the first quantum random analog signal to obtain an initial first quantum random digital signal, and use the equalization filtering unit to flatten the power spectral density of the first quantum random digital signal based on a target amplitude-frequency response characteristic, so as to reduce the time correlation between the sampling points of the first quantum random digital signal, thereby obtaining a filtered first quantum random digital signal; wherein the equalization filtering unit is loaded with filtering coefficients such that the target amplitude-frequency response characteristic of the equalization filtering unit is opposite to the amplitude-frequency response characteristic of the analog link, and the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal; and a processing module electrically connected to the analog-to-digital conversion module for performing randomness extraction processing on the filtered first quantum random digital signal and outputting a quantum random number sequence.

[0006] According to an embodiment of the present invention, the quantum random number generation rate enhancement device further includes: a coefficient generation module, which is communicatively connected to the processing module, for generating the filtering coefficients and providing the filtering coefficients to the processing module so that the processing module can load the filtering coefficients into the equalization filtering unit.

[0007] According to an embodiment of the present invention, the coefficient generation module is configured to: obtain the amplitude-frequency response characteristics of the simulated link; determine the target amplitude-frequency response characteristics of the equalization filter unit based on the amplitude-frequency response characteristics of the simulated link, and generate initial filter coefficients based on the target amplitude-frequency response characteristics of the equalization filter unit, wherein the target amplitude-frequency response characteristics of the equalization filter unit are the reciprocal of the amplitude-frequency response characteristics of the simulated link; and scale the initial filter coefficients using a gain factor to obtain the filter coefficients.

[0008] According to an embodiment of the present invention, the quantum entropy source module is further configured to generate a second quantum random analog signal; the analog-to-digital conversion module is further configured to: perform analog-to-digital conversion on the second quantum random analog signal without configuring the filtering coefficients and without configuring the initial filtering coefficients, to obtain an initial second quantum random digital signal; the coefficient generation module is further configured to: obtain the power spectral density of the initial second quantum random digital signal and the total variance of the initial second quantum random digital signal; calculate the total variance of the filtered second quantum random digital signal based on the power spectral density of the initial second quantum random digital signal and the amplitude-frequency response characteristics corresponding to the initial filtering coefficients; wherein the amplitude-frequency response characteristics of the initial filtering coefficients are obtained by mathematical transformation of the initial filtering coefficients; and obtain the gain factor based on the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

[0009] According to an embodiment of the present invention, the quantum entropy source module is further configured to generate a second quantum random analog signal; the analog-to-digital conversion module is further configured to: perform analog-to-digital conversion on the second quantum random analog signal with only the initial filtering coefficients configured, to obtain an initial second quantum random digital signal, and flatten the power spectral density of the initial second quantum random digital signal to obtain a filtered second quantum random digital signal; wherein the equalization filtering unit is loaded with the initial filtering coefficients; the coefficient generation module is further configured to: obtain the total variance of the initial second quantum random digital signal and the total variance of the filtered second quantum random digital signal; and obtain the gain factor based on the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

[0010] According to an embodiment of the present invention, the coefficient generation module is further configured to: determine a plurality of equally spaced frequency points within the sampling rate range of the analog-to-digital conversion module, wherein the number of the plurality of equally spaced frequency points is the same as the number of taps of the equalization filter unit; determine the amplitude-frequency response characteristics at the plurality of equally spaced frequency points based on the target amplitude-frequency response characteristics of the equalization filter unit; and perform a discrete inverse Fourier transform on the amplitude-frequency response characteristics at the plurality of equally spaced frequency points to obtain the initial filter coefficients.

[0011] According to an embodiment of the present invention, the processing module includes a field-programmable gate array; the processing module is further configured to: generate a pseudo-random sequence using a pre-stored seed; and perform a bitwise XOR operation between the pseudo-random sequence and the filtered first quantum random digital signal to obtain the quantum random number sequence.

[0012] According to an embodiment of the present invention, the above-mentioned equalization filtering unit is a programmable finite impulse response filter.

[0013] According to an embodiment of the present invention, the above-mentioned quantum entropy source module is an integrated chip of a balanced zero-difference detector based on vacuum state fluctuations.

[0014] One aspect of the present invention provides a method for enhancing the rate of quantum random number generation, applicable to the quantum random number generation rate enhancement apparatus described in any of the preceding claims, comprising: responding to receiving a first quantum random analog signal, performing analog-to-digital conversion on the first quantum random analog signal to obtain an initial first quantum random digital signal; using the equalization filtering unit to flatten the power spectral density of the first quantum random digital signal based on a target amplitude-frequency response characteristic to reduce the time correlation between the sampling points of the first quantum random digital signal, thereby obtaining a filtered first quantum random digital signal; wherein the equalization filtering unit is loaded with filtering coefficients such that the target amplitude-frequency response characteristic of the equalization filtering unit is opposite to the amplitude-frequency response characteristic of the analog link, and such that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal; outputting the filtered first quantum random digital signal so that a processing module performs randomness extraction processing on the filtered first quantum random digital signal and outputs a quantum random number sequence.

[0015] Embodiments of the present invention perform real-time flattening processing on quantum random digital signals at the analog-to-digital conversion front-end by loading specific filter coefficients into an equalization filter unit integrated within the analog-to-digital conversion module. This processing reconstructs the power spectral density of the signal while maintaining the total signal energy, effectively reducing the temporal correlation between different sampling points. This improves the minimum entropy of the original data at the information theory level, enabling the back-end processing module to perform randomness extraction with a higher extraction ratio or a more efficient algorithm, ultimately significantly enhancing the real-time generation rate of quantum random numbers. Attached Figure Description

[0016] The above and other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings.

[0017] Figure 1 A schematic diagram of a quantum random number generation rate enhancement device according to an embodiment of the present invention is shown.

[0018] Figure 2 A schematic diagram of a quantum random number generation rate enhancement device according to another embodiment of the present invention is shown.

[0019] Figure 3 The diagram illustrates the power spectral density of the original quantum random signal and the uniformization filtering principle of the equalization filtering unit according to a specific embodiment of the present invention.

[0020] Figure 4 A schematic diagram of a quantum random number generation rate enhancement device according to a specific embodiment of the present invention is shown.

[0021] Figure 5 A schematic diagram of signal processing and data flow during the operation phase of a quantum random number generation rate enhancement device according to a specific embodiment of the present invention is shown.

[0022] Figure 6 A flowchart of a method for enhancing the rate of quantum random number generation according to an embodiment of the present invention is shown. Detailed Implementation

[0023] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the invention. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the invention for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0024] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0025] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0026] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0027] Quantum random number generators (QRNGs) utilize the intrinsic randomness of quantum mechanics to generate truly random numbers, finding important applications in quantum key distribution, cryptography, and scientific simulations. The continuous-variable vacuum-state QRNG scheme has become one of the mainstream technologies due to its relatively simple structure, ease of integration, and high speed. In this scheme, vacuum-state quantum noise is detected by a balanced zero-difference detector and then collected by an analog-to-digital converter to obtain the original random number sequence.

[0028] However, due to limitations in the bandwidth of practical detectors, the non-ideal gain flatness of amplifiers, and the non-uniformity of the frequency response of transmission links, temporal dependencies are introduced into the raw data. According to information theory, the minimum entropy of the data determines the final rate from which information-theoretically secure true random numbers can be extracted. The presence of temporal dependencies leads to a decrease in the minimum entropy of the raw data, directly limiting the final real-time output rate of the QRNG system.

[0029] To improve the real-time speed of QRNG, related technologies first optimize analog front-end devices to enhance signal bandwidth and flatness. However, this approach is limited by the inherent physical characteristics of analog devices, resulting in limited improvement potential. Therefore, a post-processing algorithm based on the Toplitz matrix hash function is further employed to extract truly random numbers from the original data. However, when the original data has a low minimum entropy due to the aforementioned reasons, an extremely low extraction ratio must be used to ensure security. This consumes massive amounts of logic and storage resources on the hardware platform and introduces significant processing latency, becoming a major bottleneck restricting the system's real-time performance.

[0030] In view of this, embodiments of the present invention provide a quantum random number generation rate enhancement device. By loading specific filter coefficients into an equalization filter unit integrated within the analog-to-digital conversion module, the quantum random digital signal is flattened in real time at the analog-to-digital conversion front-end. This filtering process can reconstruct the power spectral density of the signal while maintaining the total signal energy, effectively reducing the temporal correlation between signal sampling points. This improves the minimum entropy of the original data at the information theory level, enabling the back-end processing module to perform randomness extraction with a higher extraction ratio or a more efficient algorithm, ultimately significantly enhancing the real-time generation rate of quantum random numbers.

[0031] Specifically, embodiments of the present invention provide a quantum random number generation rate enhancement device, comprising: a quantum entropy source module for generating a first quantum random analog signal; an analog-to-digital conversion module electrically connected to the quantum entropy source module via an analog link, the analog-to-digital conversion module including an equalization filtering unit; the analog-to-digital conversion module is used to: perform analog-to-digital conversion on the first quantum random analog signal to obtain an initial first quantum random digital signal, and use the equalization filtering unit to flatten the power spectral density of the first quantum random digital signal based on the target amplitude-frequency response characteristics, so as to reduce the time correlation between the sampling points of the first quantum random digital signal, thereby obtaining a filtered first quantum random digital signal; wherein, the equalization filtering unit is loaded with filtering coefficients so that the target amplitude-frequency response characteristics of the equalization filtering unit are opposite to the amplitude-frequency response characteristics of the analog link, and so that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal; and a processing module electrically connected to the analog-to-digital conversion module for performing randomness extraction processing on the filtered first quantum random digital signal and outputting a quantum random number sequence.

[0032] It should be noted that the quantum random number generation rate enhancement device and method provided in this embodiment of the invention can be used in the field of quantum information technology, particularly in the field of quantum key distribution. The quantum random number generation rate enhancement device and method provided in this embodiment of the invention can also be used in any field other than quantum information technology, such as cryptography, scientific simulation, statistical sampling, and other fields requiring high-speed, high-quality true random numbers. The application fields of the quantum random number generation rate enhancement device and method provided in this embodiment of the invention are not limited.

[0033] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0034] Figure 1 A schematic diagram of a quantum random number generation rate enhancement device according to an embodiment of the present invention is shown.

[0035] like Figure 1As shown in the figure, the quantum random number generation rate enhancement device illustrated in this embodiment includes a quantum entropy source module 110, an analog-to-digital conversion module 120, and a processing module 130.

[0036] The quantum entropy source module 110 is a physical unit that generates truly random signals based on the intrinsic randomness of quantum mechanics. In this embodiment, the output terminal of the quantum entropy source module 110 is electrically connected to the input terminal of the analog-to-digital converter module 120, and is used to generate a first quantum random analog signal and transmit the first quantum random analog signal to the analog-to-digital converter module 120.

[0037] Quantum random analog signals refer to analog electrical signals output after quantum vacuum fluctuations are detected by equilibrium zero-difference detection. These signals carry intrinsic, unpredictable quantum randomness.

[0038] The first quantum random analog signal refers to the quantum random analog signal generated by the quantum entropy source module 110, transmitted via the analog link, and received and processed by the analog-to-digital conversion module 120 during the normal operation phase of the quantum random number generation rate enhancement device, which is used to ultimately generate a quantum random number sequence.

[0039] The input terminal of the analog-to-digital converter module 120 is electrically connected to the output terminal of the quantum entropy source module 110 via an analog link, and the output terminal of the analog-to-digital converter module 120 is electrically connected to the input terminal of the processing module 130.

[0040] An analog link refers to the complete analog signal transmission path from the output of the quantum entropy source module 110 to the input of the analog-to-digital converter module 120. This analog signal transmission path may include analog devices such as detector output stages, amplifiers, transmission lines, and impedance matching networks. In practical operation, these analog devices exhibit non-ideal frequency response characteristics; that is, signals of different frequency components are amplified or attenuated to varying degrees when passing through the analog link, resulting in a non-flat amplitude-frequency response characteristic within the target frequency band. The amplitude-frequency response characteristic refers to the relationship between amplitude gain and frequency.

[0041] In this embodiment, the analog-to-digital conversion module 120 can be used to perform analog-to-digital conversion on the first quantum random analog signal to obtain the initial first quantum random digital signal x[n].

[0042] Analog-to-digital conversion (ADC) processing includes sampling and quantization. During sampling, a sample-and-hold circuit can be used to perform quantization based on a preset sampling rate f. s The first quantum random analog signal is discretized in the time domain to obtain a series of sampled values. During quantization, each sampled value is converted into a digital codeword using a quantizer, thus obtaining the initial first quantum random digital signal. Here, the initial first quantum random digital signal refers to the original digital signal before filtering.

[0043] Due to the non-ideal frequency response of the analog link, the power spectral density S of the initial first quantum random digital signal x[n] is reduced. x (f) It exhibits uneven characteristics within the target frequency band, with signal energy concentrated in certain frequency bands, resulting in strong time correlation between sampling points.

[0044] In this embodiment, the analog-to-digital conversion module 120 includes an equalization filter unit 121. This equalization filter unit 121 can be a digital filter used to filter signals in the digital domain to alter the signal's spectral characteristics. The equalization filter unit 121 is loaded with filter coefficients, which can be used to define the weights of the convolution operation of the equalization filter unit 121. In one example, the filter coefficients can be a set of preset values, such as h[n] (n=0,1,…,N-1,N), where N is the number of taps.

[0045] In this embodiment, the analog-to-digital conversion module 120 can be used to flatten the power spectral density of the initial first quantum random digital signal x[n] based on the target amplitude-frequency response characteristics using the equalization filtering unit 121, and output the filtered first quantum random digital signal y[n].

[0046] The target amplitude-frequency response characteristic refers to the gain or attenuation characteristics that the equalization filter unit 121 expects to apply to different frequency signal components within the target frequency band, in order to compensate for the uneven distribution of the power spectral density of the initial first quantum random digital signal x[n], thereby making the power spectral density S of the output filtered first quantum random digital signal y[n] equal to... y (f) tends to be flat.

[0047] In this embodiment, the flattening of the power spectral density indicates that the energy distribution of the signal at different frequency components tends to be uniform. The low-frequency large-scale fluctuations of the signal caused by energy concentration are suppressed, and the linear correlation between each sampling point is significantly reduced. Therefore, through this flattening process, a filtered first quantum random digital signal y[n] with reduced time correlation between each sampling point is obtained.

[0048] In this embodiment, the filter coefficients loaded by the equalization filter unit 121 are not arbitrary general parameters, but customized coefficients obtained through specific design based on the frequency response characteristics of the simulated link between the output of the quantum entropy source module 110 and the input of the analog-to-digital conversion module 120.

[0049] In this embodiment, the filtering coefficients can make the target amplitude-frequency response characteristics of the equalization filtering unit 121 opposite to those of the analog link. That is, the amplitude gain of the equalization filtering unit in the target frequency band and the amplitude attenuation of the analog link compensate for each other, thereby achieving flattening of the signal power spectral density in the cascaded response.

[0050] In this embodiment, the filtering coefficient can also keep the total signal energy of the signal before and after filtering equal, that is, the filtering operation only reconstructs the frequency domain energy distribution of the signal without changing the overall energy level of the signal.

[0051] Under the combined constraints of the above two conditions, the filtering coefficients can eliminate the time correlation between signal sampling points while maintaining the original strength of quantum noise, thereby improving the minimum entropy of the initial first quantum random digital signal x[n] at the information theory level, and providing a high-quality data foundation for subsequent randomness extraction processing.

[0052] The input terminal of the processing module 130 is electrically connected to the output terminal of the analog-to-digital converter module 120.

[0053] In this embodiment, the processing module 130 is used to receive the filtered first quantum random digital signal y[n] output by the analog-to-digital conversion module 120 and perform randomness extraction processing on it.

[0054] Randomness extraction processing refers to using information-theory-safe extraction algorithms to remove residual classical noise and correlation effects from the original data, extracting a truly random, statistically uniform quantum random number sequence as the final output of a quantum random number generation rate enhancement device.

[0055] Based on this, the quantum random number generation rate enhancement device proposed in this embodiment of the invention loads specific filter coefficients into the equalization filter unit integrated within the analog-to-digital conversion module, performing real-time flattening processing on the quantum random digital signal at the analog-to-digital conversion front end. This processing reconstructs the power spectral density of the signal while maintaining the total signal energy constant, effectively reducing the temporal correlation between different sampling points of the signal. This improves the minimum entropy of the original data at the information theory level, enabling the back-end processing module to perform randomness extraction with a higher extraction ratio or a more efficient algorithm, ultimately significantly enhancing the real-time generation rate of quantum random numbers.

[0056] In a preferred embodiment, the equalization filtering unit may include a programmable finite impulse response (PFIR) filter.

[0057] The PFIR filter is a digital filter that can be integrated into a high-speed analog-to-digital converter (ADC) chip as a hardware filtering module. Its output depends only on a finite number of current and past input values, and its impulse response is of finite length. Its filtering operation is a convolution operation, which can be used to multiply several consecutive sampling points of the initial first quantum random digital signal x[n] with the filter coefficients one by one and then sum them to obtain the filtered first quantum random digital signal. This convolution operation is performed point-by-point in a pipelined manner synchronized with the sampling clock. Each new input sampling point outputs a corresponding filtering result, thus achieving real-time digital filtering processing. Specifically, the specific expression of the filtered first quantum random digital signal y[n] is as follows:

[0058] (1);

[0059] Where h[k] represents the weight value of the k-th tap of the filter coefficient. The first quantum random digital signal represents the initial quantum random digital signal. There are sampling points, where N represents the total number of taps in the PFIR filter, where , .

[0060] In this preferred embodiment, the PFIR filter internally includes a set of programmable coefficient registers, which can load custom coefficient values ​​through an external control interface to change the frequency response characteristics of the filter. Because it uses only a feedforward path, the PFIR filter has a simple structure, fixed time delay, and is easy to integrate into a high-speed ADC chip in hardware, making it suitable for the real-time signal flattening requirements of this invention.

[0061] In one specific implementation, the PFIR filter may include a real coefficient filter with 96 or 48 taps. The larger the number of taps, the higher the frequency resolution, and the more accurately it can approximate the complex inverse characteristic curve.

[0062] Based on this, embodiments of the present invention use a programmable finite impulse response filter as the equalization filtering unit, and utilize the existing built-in hardware of the high-speed ADC chip to realize the flattening filtering function. There is no need to design an additional dedicated filtering circuit, which not only reduces the system complexity and cost, but also allows the flattening process and analog-to-digital conversion to be completed in the same pipeline within the same chip, minimizing processing delay and ensuring real-time performance.

[0063] In a preferred embodiment, the quantum entropy source module is an integrated chip of a balanced zero-difference detector based on vacuum state fluctuations.

[0064] The vacuum state is the ground state of the quantum light field, and even at absolute zero, it exhibits inherent quantum fluctuations that are theoretically completely unpredictable. In this preferred embodiment, a balanced homodyne detector can efficiently extract the orthogonal amplitude or phase fluctuations of the vacuum state by interfering the vacuum state signal with a local oscillating light signal on a 50:50 beam splitter and using differential detection of two photodiodes to eliminate classical noise in the local oscillating light. This fluctuation signal is intrinsically random, constituting an information-theoretically secure true random number entropy source. The integrated chip implementation integrates the aforementioned optical components and detector onto a single chip, offering advantages such as small size, low power consumption, and high stability.

[0065] Based on this, embodiments of the present invention employ a vacuum-state balanced zero-difference detector as a quantum entropy source, providing the device with a rigorously provable source of randomness based on fundamental principles of quantum mechanics. Compared to entropy sources based on classical noise or pseudo-random algorithms, this physically guarantees the true randomness of the original data. Furthermore, the integrated chip implementation allows for the integration of optical components and detectors from the analog link onto a single chip, offering advantages such as small size, low power consumption, and high stability.

[0066] Figure 2 A schematic diagram of a quantum random number generation rate enhancement device according to another embodiment of the present invention is shown.

[0067] like Figure 2 As shown, in Figure 1 Based on the device shown, the quantum random number generation rate enhancement device also includes a coefficient generation module 140.

[0068] The coefficient generation module 140 is communicatively connected to the processing module 130. In some specific embodiments, this communication connection can be implemented through various means such as the Peripheral Component Interconnect Express (PCIe) bus, Ethernet, internal bus, or shared memory.

[0069] In this embodiment, the coefficient generation module 140 is used to generate filter coefficients and provide the generated filter coefficients to the processing module 130.

[0070] During the system initialization phase, i.e., before the device officially begins generating quantum random number sequences, the processing module 130, in response to receiving the filtering coefficients, loads the filtering coefficients into the coefficient register inside the equalization filtering unit 121 via the Serial Peripheral Interface (SPI). For example, for a high-speed ADC chip, the processing module 130 can sequentially write the filtering coefficients into the coefficient register inside the equalization filtering unit 121 (e.g., addresses 0x0E00~0x0E7F) via the SPI interface. After completing the loading of the filtering coefficients, the processing module 130 also configures the control register of the analog-to-digital conversion module 120 to enable the filtering function of the equalization filtering unit 121. At this point, the equalization filtering unit 121 is in a configuration-ready state.

[0071] Based on this, embodiments of the present invention separate the design and generation function of filter coefficients from the processing module by setting up an independent coefficient generation module, forming a modular device architecture. The processing module does not need to consume its own computing resources to calculate coefficients in real time; it is only responsible for loading the generated coefficients into the equalization filtering unit and performing subsequent random extraction processing, thereby improving the overall operating efficiency of the system. At the same time, the coefficient generation module can flexibly generate dedicated coefficients according to the actual simulation link characteristics of different systems, enhancing the adaptability and configurability of the device.

[0072] The specific deployment location of the coefficient generation module 140 can be flexibly selected according to the actual needs of the system. In one example, the coefficient generation module 140 can be deployed in a standalone host computer (such as a personal computer or server), communicating with the processing module 130 via a PCIe bus or Ethernet. In another example, the coefficient generation module 140 can be deployed in an embedded processor (e.g., a RISC-8 processor), interacting with the processing module 130 via an internal bus or shared memory. In yet another example, the coefficient generation module 140 can be deployed as a soft core within the field-programmable gate array (FPGA) of the processing module 130, operating as a functional module within the FPGA and communicating with other logic units of the processing module 130 through the FPGA's internal interconnect resources.

[0073] According to an embodiment of the present invention, the coefficient generation module is used to: obtain the amplitude-frequency response characteristics of the analog link; determine the target amplitude-frequency response characteristics of the equalization filter unit based on the amplitude-frequency response characteristics of the analog link, and generate initial filter coefficients based on the target amplitude-frequency response characteristics of the equalization filter unit; and scale the initial filter coefficients using a gain factor to obtain the filter coefficients.

[0074] The amplitude-frequency response characteristics of a simulated link can be used to describe the relationship between the amplitude gain of the simulated link and frequency within a target frequency band. In some specific implementations, these characteristics can be obtained by performing actual frequency sweep measurements on the simulated link using measurement equipment such as a spectrum analyzer, or they can be obtained through circuit simulation during the system design phase. The obtained amplitude-frequency response characteristics can be stored in the coefficient generation module in the form of amplitude values ​​at discrete frequency points.

[0075] The following combination Figure 3 The core principle of achieving power spectral density flattening using the above-mentioned filter coefficients is explained in detail.

[0076] Figure 3 The diagram illustrates the power spectral density of the original quantum random signal and the uniformization filtering principle of the equalization filtering unit according to a specific embodiment of the present invention.

[0077] like Figure 3 As shown, the horizontal axis represents frequency, and the vertical axis represents power spectral density. The dashed curve represents the ideal power spectral density distribution of the quantum random signal generated by the quantum entropy source module, which appears as a flat straight line within the target frequency band (e.g., the power spectral density is constant at approximately -70 dBm / Hz), consistent with the white noise characteristics of vacuum quantum noise in the classical limit.

[0078] The solid curve represents the power spectral density distribution of the quantum random signal after transmission via an analog link, without processing by an equalization filter unit; that is, the power spectral density of the initial first quantum random digital signal. This solid curve visually illustrates the distortion effect caused by the non-ideal frequency responses of devices such as detectors, amplifiers, and transmission lines in the analog link: the originally flat power spectral density becomes uneven, and the signal energy is not uniformly distributed across the entire frequency band, but is significantly concentrated in the low-frequency band, while the energy in the high-frequency band is relatively weak. This concentrated distribution of energy in the frequency domain manifests in the time domain as a strong linear correlation between adjacent sampling points of the signal, i.e., time correlation. This time correlation directly leads to a decrease in the information-theoretic minimum entropy of the original data, limiting the efficiency of subsequent randomness extraction and the real-time generation rate of the final random numbers.

[0079] The function of the equalization filter unit is to compensate for the non-ideal frequency response of the analog link, so that the power spectral density of the filtered signal is restored to a flat state. To this end, the filter coefficients loaded in the equalization filter unit are specifically designed based on the amplitude-frequency response characteristics of the analog link. Its target amplitude-frequency response characteristics are opposite to those of the analog link, in order to counteract the non-ideal influence of the analog link on the signal spectrum, redistribute the energy concentrated in the low-frequency band to the entire target frequency band, and restore the power spectral density of the output signal to a flat characteristic consistent with the dashed curve.

[0080] In this embodiment, the target amplitude-frequency response characteristic of the equalization filter unit is determined to be the reciprocal of the amplitude-frequency response characteristic of the analog link. Specifically, the expression for this mathematical relationship is as follows:

[0081] (2);

[0082] Among them, H target (f) represents the target amplitude-frequency response characteristics of the equalization filter unit, H front (f) represents the amplitude-frequency response characteristics of the analog link.

[0083] Based on this mathematical relationship, it can be seen that in frequency bands where the analog link attenuation is large, the equalization filter unit should provide a corresponding multiple of gain. Conversely, in frequency bands where the analog link gain is high, the equalization filter unit should provide a corresponding multiple of attenuation. Based on the target amplitude-frequency response characteristics as described in formula (2), a set of initial filter coefficients h can be generated through the filter design algorithm. init [n], whose frequency response can approximate the target amplitude-frequency response characteristic H of the equalization filter unit within the target frequency band. target (f) is used to compensate for the non-ideal effects of the analog link on the signal spectrum. The filter design algorithms include, but are not limited to, frequency sampling and least squares methods.

[0084] The gain factor is a scalar constant used to scale all initial filter coefficients as a whole. The scaled filter coefficients are the final filter coefficients applied to the equalization filter unit. Specifically, the expression for the filter coefficients is as follows:

[0085] (3);

[0086] Where h[n] represents the filter coefficient, G represents the gain factor, and h init [n] represents the initial filter coefficients.

[0087] In this embodiment, while flattening the power spectral density using only the initial filter coefficients can flatten the signal, it may result in the total energy of the output signal being unequal to the total energy of the input signal. Scaling the initial filter coefficients using a gain factor is precisely to compensate for this energy difference, ensuring that when filtering with the scaled filter coefficients, the total energy of the filtered first quantum random digital signal is equal to the total energy of the initial first quantum random digital signal.

[0088] Figure 4 A schematic diagram of a quantum random number generation rate enhancement device according to a specific embodiment of the present invention is shown.

[0089] like Figure 4As shown, in this specific embodiment, the coefficient generation module 140 may include a feature measurement unit 141, an equalization filter design unit 142, and a coefficient formatting unit 143 to collaboratively complete the above-mentioned filter coefficient generation process.

[0090] Specifically, the feature measurement unit 141 is used to perform the above-mentioned operation of acquiring the amplitude-frequency response characteristics of the simulated link, that is, to acquire the actual amplitude-frequency response characteristics of the simulated link from the output of the quantum entropy source module to the input of the analog-to-digital conversion module in the target frequency band by measurement or simulation, i.e., H. front (f).

[0091] The equalization filter design unit 142 is used to perform the above-mentioned operation of determining the target amplitude-frequency response characteristics and generating initial filter coefficients based on the amplitude-frequency response characteristics of the analog link. Based on the amplitude-frequency response characteristics provided by the feature measurement unit 141, this unit determines the target amplitude-frequency response characteristic H of the equalization filter unit according to the reciprocal relationship described in formula (2). target (f). The initial filter coefficients h are generated using a filter design algorithm. init [n], making the cascaded response of the equalization filter and the front-end analog link flat within the target frequency band. Furthermore, the equalization filter design unit 142 is also used to determine the gain factor G and to adjust the initial filter coefficients h using the gain factor G. init [n] is scaled to obtain the final filter coefficients h[n].

[0092] The coefficient formatting unit 143 is used to perform the operation of converting the filter coefficients into a format that can be loaded by the analog-to-digital converter module. Since the equalization filter unit built into the analog-to-digital converter module has specific format requirements for the loaded coefficients, the coefficient formatting unit 143 is responsible for converting the filter coefficients h[n] generated by the equalization filter design unit 142 into a suitable fixed-point format, such as the Q1.15 format. The Q1.15 format is a fixed-point number representation format. "Q" represents a fixed-point number (Quantized number), "1" represents the number of bits in the integer part, and "15" represents the number of bits in the fractional part. In the Q1.15 format, each filter coefficient is represented as a 16-bit signed fixed-point number, with the highest bit being the sign bit and the remaining 15 bits representing the fractional part, so that the processing module 130 can load it into the coefficient register of the analog-to-digital converter module 120 through the SPI interface.

[0093] Based on this, embodiments of the present invention refine the generation process of filter coefficients into a collaborative process involving multiple clearly defined sub-units, such as link measurement, inverse filter design, energy compensation scaling, and format conversion. This ensures that the final generated filter coefficients can accurately compensate for the non-ideal characteristics of a specific analog link, and achieve a flattened power spectral density within the target frequency band while maintaining the total signal energy. The time correlation between sampling points is effectively reduced. Since the filtering process does not change the total signal energy but only reconstructs its spectral distribution, the filtered signal significantly reduces its time correlation while maintaining the original quantum noise intensity, thereby improving the minimum entropy of the original data.

[0094] According to an embodiment of the present invention, the quantum entropy source module is further configured to generate a second quantum random analog signal; the analog-to-digital conversion module is further configured to: perform analog-to-digital conversion on the second quantum random analog signal without configuring filter coefficients and without configuring initial filter coefficients, to obtain an initial second quantum random digital signal; the coefficient generation module is further configured to: obtain the power spectral density of the initial second quantum random digital signal and the total variance of the initial second quantum random digital signal; calculate the total variance of the filtered second quantum random digital signal based on the power spectral density of the initial second quantum random digital signal and the amplitude-frequency response characteristics corresponding to the initial filter coefficients; and obtain the gain factor based on the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

[0095] In this embodiment, the gain factor can be determined based on theoretical calculations.

[0096] Specifically, during the system initialization phase, the quantum entropy source module is also used to generate a second quantum random simulation signal.

[0097] The second quantum random simulation signal refers to the quantum random simulation signal generated by the quantum entropy source module, transmitted via the analog link, and received and processed by the analog-to-digital conversion module during the initialization phase of the quantum random number generation rate enhancement device, used to determine the gain factor. It should be noted that the second quantum random simulation signal and the first quantum random simulation signal are both generated by the same quantum entropy source module, and they have the same physical characteristics, differing only in their usage phase and purpose.

[0098] In this embodiment, the analog-to-digital conversion module is used to perform analog-to-digital conversion on the second quantum random analog signal without configuring filter coefficients or initial filter coefficients, to obtain the initial second quantum random digital signal.

[0099] During this initialization phase, the equalization filter unit 121 is either in a bypass state or has a unit impulse response coefficient, and no filtering is performed on the signal. Therefore, the initial second quantum random digital signal obtained through analog-to-digital conversion completely retains all the non-ideal frequency response characteristics introduced by the analog link.

[0100] In this embodiment, the coefficient generation module is used to obtain the power spectral density S of the initial second quantum random digital signal. init (f) and the total variance of the initial second quantum random digital signal In some specific implementations, the power spectral density S of the initial second quantum random digital signal init (f) can be obtained by estimating the power spectrum of the initial second quantum random digital signal. The total variance of the initial second quantum random digital signal... It can be obtained by performing variance statistics on the sample values ​​of the initial second quantum random digital signal.

[0101] In this embodiment, the coefficient generation module is also used to perform the following calculations to determine the gain factor G.

[0102] In this embodiment, the initial filter coefficients h can be adjusted first. init [n] undergoes mathematical transformations to obtain the result with respect to the initial filter coefficients h. init The amplitude-frequency response characteristic H corresponding to [n] init (f) This mathematical transformation is performed only in the digital domain on a known sequence of coefficients to obtain its corresponding amplitude-frequency response characteristics, without relying on the actual hardware operation of the equalization filter unit. In one example, the initial filter coefficients h can be... init [n] undergoes a discrete-time Fourier transform, and the specific expression is as follows:

[0103] (4);

[0104] Among them, H init (f) represents the amplitude-frequency response characteristic corresponding to the initial filter coefficients, h init [n] represents the initial filter coefficients, f s This indicates the sampling rate of the analog-to-digital conversion module.

[0105] In this embodiment, the power spectral density S of the initial second quantum random digital signal can be used as a basis. init (f), and the amplitude-frequency response characteristics H corresponding to the initial filter coefficients obtained by the above mathematical transformation. init (f) Calculate the total variance of the filtered second quantum random digital signal. .

[0106] In one example, according to Parseval's theorem, the power spectral density S of the initial second quantum random digital signal can be expressed as... init (f) and the amplitude-frequency response characteristic H corresponding to the initial filter coefficients init Multiply the squares of the modulus of (f) and integrate in the frequency domain. The result of the integration is the total variance of the filtered second quantum random digital signal. The specific expression is as follows:

[0107] (5);

[0108] in, H represents the total variance of the filtered second quantum random digital signal. init (f) represents the amplitude-frequency response characteristic corresponding to the initial filter coefficients, S init (f) represents the power spectral density of the initial second quantum random digital signal.

[0109] In this embodiment, the total variance of the initial second quantum random digital signal is calculated. The total variance of the filtered second quantum random digital signal The square root of the ratio is the gain factor G. This gain factor G is used to adjust the initial filter coefficients h. init Scaling [n] yields the final filter coefficients. The specific expression for the gain factor G is as follows:

[0110] (6).

[0111] Based on this, embodiments of the present invention provide a gain factor determination method based on theoretical calculation. This method samples a quantum random simulation signal during the initialization phase, statistically analyzes the power spectral density of the sampled second quantum random simulation signal, and integrates this power density with the mathematical transformation results of the initial filter coefficients in the frequency domain. This allows the gain factor to be obtained without actually running the equalization filter unit. This method does not rely on real-time hardware feedback; the calculation process is completed entirely within the computation unit of the coefficient generation module, making it suitable for scenarios requiring rapid initialization or with limited hardware resources.

[0112] According to an embodiment of the present invention, the quantum entropy source module is further configured to generate a second quantum random analog signal; the analog-to-digital conversion module is further configured to: perform analog-to-digital conversion on the second quantum random analog signal with only initial filtering coefficients configured to obtain an initial second quantum random digital signal, and flatten the power spectral density of the initial second quantum random digital signal to obtain a filtered second quantum random digital signal; the coefficient generation module is further configured to: obtain the total variance of the initial second quantum random digital signal and the total variance of the filtered second quantum random digital signal; and obtain the gain factor based on the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

[0113] In this embodiment, the gain factor can also be determined based on online measurement.

[0114] Specifically, during the system initialization phase, the quantum entropy source module generates a second quantum random simulation signal. It should be noted that the source and physical characteristics of the second quantum random simulation signal are consistent with those described in the aforementioned theoretical calculations, and will not be repeated here.

[0115] In this embodiment, the analog-to-digital conversion module only configures the initial filter coefficients h. init In the case of [n], the second quantum random analog signal is converted from analog to digital to obtain the initial second quantum random digital signal. The equalization filter unit is loaded with the initial filter coefficients h that have not been scaled by the gain factor G. init [n], rather than the final complete filter coefficients.

[0116] The equalization filter unit uses the pre-loaded initial filter coefficients to adjust the power spectral density S of the initial second quantum random digital signal. init (f) Flattening is performed to obtain the filtered second quantum random digital signal. Although the power spectrum of the output signal has become flat, the initial filter coefficient h... init [n] Without energy compensation, its total energy may differ from the input signal.

[0117] In this embodiment, the coefficient generation module is used to obtain the total variance of the initial second quantum random digital signal. and the total variance of the filtered second quantum random digital signal. Both are obtained by performing variance statistics on the sample values ​​of their respective signals, without the need for frequency domain integration calculations. Subsequently, the total variance of the initial second quantum random digital signal is calculated. The total variance of the filtered second quantum random digital signal The square root of the ratio is used as the gain factor G, and the specific calculation method is shown in formula (6).

[0118] Based on this, embodiments of the present invention provide a gain factor determination method based on online measurement. This method obtains the gain factor by directly calculating the variance of the input and output signals through an actual run of a flattening process. Compared with theoretical calculations, this method is intuitive and reliable, does not rely on theoretical calculations such as power spectrum estimation and frequency domain integration, and can adaptively compensate for various non-ideal factors in actual hardware operation, making it particularly suitable for systems supporting online calibration.

[0119] According to an embodiment of the present invention, the coefficient generation module is further configured to: determine multiple equally spaced frequency points within the sampling rate range of the analog-to-digital conversion module, wherein the number of multiple equally spaced frequency points is the same as the number of taps of the equalization filter unit; determine the amplitude-frequency response characteristics at the multiple equally spaced frequency points based on the target amplitude-frequency response characteristics of the equalization filter unit; and perform a discrete inverse Fourier transform on the amplitude-frequency response characteristics at the multiple equally spaced frequency points to obtain the initial filter coefficients.

[0120] In this embodiment, the specific algorithm for generating the initial filter coefficients can employ a frequency sampling method. This method involves sampling the desired frequency response at equal intervals in the frequency domain to obtain a set of discrete frequency response sample values. These sample values ​​are then subjected to an inverse discrete Fourier transform to obtain the filter coefficients in the time domain. Since the frequency domain sampling point positions directly correspond to the time domain characteristics of the filter coefficients, the calculation process is simple and efficient.

[0121] Specifically, the coefficient generation module can be used within the sampling rate range of the analog-to-digital conversion module (i.e., from DC to f). s / 2 frequency range, where f s (Representing the sampling rate), multiple equally spaced frequency points are selected. These frequency points are a set of discrete frequency positions uniformly selected on the frequency axis, used to evaluate the target amplitude-frequency response characteristics H of the equalization filter unit. target (f) Perform frequency domain sampling. The number of frequency points is the same as the number of taps in the equalization filter unit, which is N. For example, when the equalization filter unit is a finite impulse response filter with 96 taps, the number of frequency points is also 96.

[0122] In one example, the location of the frequency point can be represented as:

[0123] (7);

[0124] Among them, f k This represents the frequency value of the k-th frequency point.

[0125] For each frequency point f k The coefficient generation module is based on the target amplitude-frequency response characteristics H. target(f) Determine the amplitude-frequency response characteristic H[k] at this frequency point. This frequency response value is a complex number, and its amplitude is determined by the target amplitude-frequency response characteristic H[k] at this frequency point. target (f k The phase can be set to a linear phase. This ensures the causal realizability of the designed filter. Specifically, the expression for the amplitude-frequency response characteristic at each frequency point is as follows:

[0126] (8);

[0127] Where H[k] represents the amplitude-frequency response characteristic at the k-th frequency point, This represents the target amplitude-frequency response characteristic at the k-th frequency point, where, That is, the target amplitude-frequency response characteristic at the k-th frequency point is the reciprocal of the amplitude-frequency response characteristic of the analog link at the k-th frequency point.

[0128] In this specific embodiment, the coefficient generation module can also be used to perform an inverse discrete Fourier transform (IDFT) on the amplitude-frequency response characteristics H[k] at the above N equally spaced frequency points. The time-domain sequence obtained by the transform is the initial filter coefficient h. init [n] maps discrete sampled values ​​in the frequency domain to a discrete impulse response sequence in the time domain. The IDFT transforms the resulting sequence, and the length of the sequence is equal to the number of frequency points N, which is also equal to the number of taps in the equalization filter unit. Specifically, the initial filter coefficients h init The expression for [n] is as follows:

[0129] (9).

[0130] Based on this, embodiments of the present invention construct complex amplitude-frequency response characteristics containing linear phase directly at discrete sampling points of the target frequency response, and obtain time-domain coefficients using inverse discrete Fourier transform. The principle is clear and the computation is efficient. The frequency sampling method can directly generate filter coefficients that meet the requirements based on the known target frequency response, making it suitable for embedded systems with limited hardware resources. This provides an accurate and reliable foundation for subsequent calculation of the gain factor and the final generation of the filter coefficients.

[0131] After the generation, formatting, and loading of the filter coefficients are completed in the system initialization phase, and the filtering function of the equalization filter unit is enabled, the device enters the real-time operation phase.

[0132] Figure 5 A schematic diagram of signal processing during the operation phase of a quantum random number generation rate enhancement device according to a specific embodiment of the present invention is shown.

[0133] like Figure 5 As shown, the analog-to-digital conversion module 120 includes an equalization filtering unit 121, an analog-to-digital conversion unit 122, and a uniformized raw random number unit 123.

[0134] The analog-to-digital converter 122 receives the first quantum random analog signal transmitted via an analog link, performs analog-to-digital conversion on it, and obtains the initial first quantum random digital signal. The analog-to-digital conversion includes two sub-steps: sampling and quantization. The sample-and-hold circuit discretizes the analog signal in the time domain at a preset sampling rate, and the quantizer converts each sampled value into a digital codeword.

[0135] The equalization filter unit 121 is loaded with filter coefficients to flatten the power spectral density of the initial first quantum random digital signal and output the filtered first quantum random digital signal. The target amplitude-frequency response characteristics of the equalization filter unit 121 are opposite to those of the analog link, and the filter coefficients are scaled to make the total energy of the signal before and after filtering equal.

[0136] The uniformized raw random number unit 123 is used to receive the filtered first quantum random digital signal, and to format it to generate a standardized data stream that conforms to subsequent transmission and processing. In one embodiment, the uniformized raw random number unit 123 can package the filtered data, add frame identifiers, or perform bit width conversion so that it can be transmitted to the processing module through a high-speed serial interface.

[0137] like Figure 5 As shown, during the operation phase, the first quantum random analog signal output by the quantum entropy source module 110 enters the analog-to-digital conversion module 120. The analog-to-digital conversion unit 122 inside the analog-to-digital conversion module 120 samples and converts the first quantum random analog signal at a preset sampling rate (e.g., 2.4 GSa / s) to obtain the initial first quantum random digital signal x[n].

[0138] The digital signal after analog-to-digital conversion is fed into the equalization filtering unit 121 built into the analog-to-digital conversion module 120. The equalization filtering unit 121 performs a real-time convolution operation on the initial first quantum random digital signal x[n] synchronized with the sampling clock, based on the loaded filtering coefficients h[n], and performs flattening processing. This filtering process, under the constraint of keeping the total signal energy constant, reconstructs the power spectrum distribution of the signal, making the power spectral density of the filtered first quantum random digital signal y[n] tend to be flat, effectively reducing the time correlation between signal samples.

[0139] The filtered first quantum random digital signal y[n] is then sent to the uniformized original random number unit 123 for data format processing to generate a standardized data stream that conforms to subsequent transmission and processing.

[0140] Processing module 130 receives the filtered first quantum random digital signal y[n] from analog-to-digital converter module 120 via a high-speed serial interface. Specifically, analog-to-digital converter module 120 internally sends the data processed by the normalized original random number unit 123 to a high-speed serial transmitter, which transmits it to processing module 130 in the form of a high-speed serial differential signal via a high-speed serial interface. The high-speed serial receiver in processing module 130 is responsible for restoring the clock, deframes, and reconstructing the filtered first quantum random digital signal y[n] for subsequent randomness extraction processing.

[0141] In one specific embodiment, the high-speed serial transmitter, high-speed serial receiver, and high-speed serial interface described above all conform to the JESD204B standard. The analog-to-digital conversion module 120 integrates a JESD204B transmitter, and the processing module 130 integrates a JESD204B receiver; the two transmit data through the JESD204B interface.

[0142] In this embodiment, the processing module may include a Field Programmable Gate Array (FPGA). An FPGA is a programmable logic device with abundant programmable logic units, registers, and high-speed serial transceivers, making it suitable for performing real-time pipelined processing of high-speed data streams.

[0143] In this embodiment, the specific method of randomness extraction performed by the processing module is as follows. First, a pseudo-random sequence p[n] is generated using a pre-stored seed. The seed is a preset initial value used to initialize the internal state of the pseudo-random number generator. The pseudo-random sequence is generated by a deterministic algorithm and has statistically approximate random characteristics. Second, the pseudo-random sequence p[n] is XORed with the filtered first quantum random digital signal y[n] to obtain the output quantum random number sequence z[n]. Specifically, the expression for the quantum random number sequence z[n] is as follows:

[0144] (10);

[0145] in, This indicates a bitwise XOR logical operation. For example, if two bits are the same, the output is 0; if they are different, the output is 1.

[0146] In this embodiment, the filtered first quantum random digital signal y[n] carries quantum true randomness. XORing it with a pseudo-random sequence can purify the randomness and achieve statistical homogenization of the original signal. Since the filtered first quantum random digital signal y[n] has been flattened by the equalization filtering unit, the time correlation is greatly reduced and the minimum entropy is improved. The FPGA can use this simple and efficient streaming XOR algorithm to achieve real-time processing to match the high-speed data rate of the front end.

[0147] Based on this, the embodiments of the present invention make full use of the parallel pipeline processing capability of FPGA, simplifying complex hash operations into bitwise XOR operations, which greatly reduces the hardware resource consumption and processing latency of post-processing, enabling the post-processing throughput to match the high sampling rate of the front-end analog-to-digital conversion module, thereby ensuring the overall real-time performance of quantum random number generation.

[0148] The embodiments of the present invention provide a method for enhancing the rate of quantum random number generation, applicable to the apparatus of any of the foregoing embodiments.

[0149] Figure 6 A flowchart of a method for enhancing the rate of quantum random number generation according to an embodiment of the present invention is shown.

[0150] like Figure 6 As shown, the method for enhancing the quantum random number generation rate includes operations S610~S630.

[0151] In operation S610, in response to receiving the first quantum random analog signal, the first quantum random analog signal is converted from analog to digital to obtain the initial first quantum random digital signal.

[0152] In operation S620, the power spectral density of the initial first quantum random digital signal is flattened based on the target amplitude-frequency response characteristics using the equalization filter unit, so as to reduce the time correlation between the sampling points of the first quantum random digital signal and obtain the filtered first quantum random digital signal.

[0153] In this embodiment, the equalization filter unit is loaded with filter coefficients so that the target amplitude-frequency response characteristics of the equalization filter unit are opposite to the amplitude-frequency response characteristics of the analog link, and so that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal.

[0154] During operation S630, a filtered first quantum random digital signal is output so that the processing module can perform randomness extraction processing on the filtered first quantum random digital signal and output a quantum random number sequence.

[0155] It should be noted that the quantum random number generation rate enhancement method part in the embodiments of the present invention corresponds to the quantum random number generation rate enhancement device part in the embodiments of the present invention. For a detailed description of the quantum random number generation rate enhancement method part, please refer to the quantum random number generation rate enhancement device part, which will not be repeated here.

[0156] Based on this, embodiments of the present invention embed power spectral density flattening under energy constraints into the analog-to-digital conversion process, thereby reducing time correlation and increasing minimum entropy at the very beginning of signal sampling. The method does not rely on complex equalization calculations by external processing modules, freeing up subsequent processing resources and enabling the overall system to generate high-quality quantum random number sequences in real time with higher throughput.

[0157] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of the present invention can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or pairings fall within the scope of this invention.

[0158] The embodiments of the present invention have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of the invention. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the invention, and all such substitutions and modifications should fall within the scope of the invention.

Claims

1. A device for enhancing the rate of quantum random number generation, characterized in that, The quantum random number generation rate enhancement device includes: The quantum entropy source module is used to generate the first quantum random analog signal; An analog-to-digital conversion module is electrically connected to the quantum entropy source module via an analog link. The analog-to-digital conversion module includes an equalization filtering unit. The analog-to-digital conversion module is used to: perform analog-to-digital conversion on the first quantum random analog signal to obtain an initial first quantum random digital signal, and use the equalization filtering unit to flatten the power spectral density of the first quantum random digital signal based on the target amplitude-frequency response characteristics to reduce the time correlation between the sampling points of the first quantum random digital signal, thereby obtaining a filtered first quantum random digital signal; wherein, the equalization filtering unit is loaded with filtering coefficients so that the target amplitude-frequency response characteristics of the equalization filtering unit are opposite to the amplitude-frequency response characteristics of the analog link, and so that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal; The processing module, electrically connected to the analog-to-digital conversion module, is used to perform randomness extraction processing on the filtered first quantum random digital signal and output a quantum random number sequence.

2. The quantum random number generation rate enhancement device according to claim 1, characterized in that, The quantum random number generation rate enhancement device also includes: The coefficient generation module is communicatively connected to the processing module and is used to generate the filter coefficients and provide the filter coefficients to the processing module so that the processing module can load the filter coefficients into the equalization filter unit.

3. The quantum random number generation rate enhancement device according to claim 2, characterized in that, The coefficient generation module is used for: Obtain the amplitude-frequency response characteristics of the simulated link; Based on the amplitude-frequency response characteristics of the simulated link, the target amplitude-frequency response characteristics of the equalization filter unit are determined, and based on the target amplitude-frequency response characteristics of the equalization filter unit, initial filter coefficients are generated, wherein the target amplitude-frequency response characteristics of the equalization filter unit are the reciprocal of the amplitude-frequency response characteristics of the simulated link. The initial filter coefficients are scaled using a gain factor to obtain the filter coefficients.

4. The quantum random number generation rate enhancement device according to claim 3, characterized in that, The quantum entropy source module is also used to generate a second quantum random simulation signal; The analog-to-digital conversion module is also used for: Without configuring the filter coefficients and without configuring the initial filter coefficients, the second quantum random analog signal is converted from analog to digital to obtain the initial second quantum random digital signal; The coefficient generation module is also used for: Obtain the power spectral density of the initial second quantum random digital signal and the total variance of the initial second quantum random digital signal; Based on the power spectral density of the initial second quantum random digital signal and the amplitude-frequency response characteristics corresponding to the initial filter coefficients, the total variance of the filtered second quantum random digital signal is calculated; wherein, the amplitude-frequency response characteristics of the initial filter coefficients are obtained by mathematical transformation of the initial filter coefficients; The gain factor is obtained by taking the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

5. The quantum random number generation rate enhancement device according to claim 3, characterized in that, The quantum entropy source module is also used to generate a second quantum random simulation signal; The analog-to-digital conversion module is also used for: With only the initial filter coefficients configured, the second quantum random analog signal is subjected to analog-to-digital conversion to obtain an initial second quantum random digital signal, and the power spectral density of the initial second quantum random digital signal is flattened to obtain a filtered second quantum random digital signal. The equalization filtering unit is loaded with the initial filtering coefficients; the coefficient generation module is further used for: Obtain the total variance of the initial second quantum random digital signal and the total variance of the filtered second quantum random digital signal; The gain factor is obtained by taking the square root of the ratio of the total variance of the initial second quantum random digital signal to the total variance of the filtered second quantum random digital signal.

6. The quantum random number generation rate enhancement device according to claim 3, characterized in that, The coefficient generation module is also used for: Within the sampling rate range of the analog-to-digital conversion module, multiple equally spaced frequency points are selected, wherein the number of the multiple equally spaced frequency points is the same as the number of taps of the equalization filter unit; Based on the target amplitude-frequency response characteristics of the equalization filtering unit, the amplitude-frequency response characteristics at the plurality of equally spaced frequency points are determined; The initial filter coefficients are obtained by performing a discrete inverse Fourier transform on the amplitude-frequency response characteristics at the multiple equally spaced frequency points.

7. The quantum random number generation rate enhancement device according to claim 1, characterized in that, The processing module includes a field-programmable gate array; The processing module is also used for: Generate pseudo-random sequences using pre-stored seeds; The pseudo-random sequence is XORed with the filtered first quantum random digital signal to obtain the quantum random number sequence.

8. The quantum random number generation rate enhancement device according to claim 1, characterized in that, The equalization filtering unit is a programmable finite impulse response filter.

9. The quantum random number generation rate enhancement device according to claim 1, characterized in that, The quantum entropy source module is an integrated chip of a balanced zero-difference detector based on vacuum state fluctuations.

10. A method for enhancing the rate of quantum random number generation, applicable to the quantum random number generation rate enhancement device according to any one of claims 1 to 9, characterized in that, The method for enhancing the quantum random number generation rate includes: In response to receiving a first quantum random analog signal, the first quantum random analog signal is converted from analog to digital to obtain an initial first quantum random digital signal; The power spectral density of the first quantum random digital signal is flattened based on the target amplitude-frequency response characteristics using the equalization filtering unit to reduce the time correlation between the sampling points of the first quantum random digital signal, thereby obtaining a filtered first quantum random digital signal. The equalization filtering unit is loaded with filtering coefficients such that the target amplitude-frequency response characteristics of the equalization filtering unit are opposite to the amplitude-frequency response characteristics of the analog link, and that the total signal energy of the filtered first quantum random digital signal is equal to the total signal energy of the initial first quantum random digital signal. A filtered first quantum random digital signal is output so that the randomness of the filtered first quantum random digital signal can be extracted by the processing module, and a quantum random number sequence can be output.

Images