An integrated photonic chip-based source-device-independent quantum random number generator

Abstract

This invention relates to quantum random number generators, specifically to a source-device-independent quantum random number generator based on an integrated photonic chip. The quantum source, trustless, provides quantum state input to an integrated heterodyne receiver. The integrated heterodyne receiver, trusted and fully characterized, consists of an integrated photonic chip (PIC) and its associated high-speed, low-noise electronic readout chain, responsible for heterodyne measurement of the input quantum state, extracting the Q and P orthogonal components of quantum fluctuations, and converting them into corresponding digital signals. A digital processing and random number extraction unit receives the Q and P digital signals, performs system calibration and signal preprocessing, and executes secure random number extraction. The technical solution provided by this invention effectively overcomes the shortcomings of existing technologies, such as difficulty in balancing high security and high random number generation rate, as well as the complexity, poor stability, and high power consumption of the equipment.

Description

Technical Field

[0001] This invention relates to quantum random number generators, and more specifically to a source-device-independent quantum random number generator based on an integrated photonic chip. Background Technology

[0002] Quantum random number generators (QRNGs) are the only physical means of generating truly random numbers and play an irreplaceable role in cryptography, secure communication, and scientific computing. However, all existing types of QRNGs have significant drawbacks:

[0003] 1) Fully Trusted QRNG: Based on continuous variable schemes such as vacuum fluctuations, it can achieve a high random number generation rate of several Gbps to tens of Gbps. However, its security is entirely based on absolute trust and accurate characterization of all hardware (especially light sources). Any unmodeled hardware defects, performance drifts and potential backdoors can be exploited by attackers to predict random numbers, resulting in a limited level of security.

[0004] 2) Device-independent QRNG: Based on Bell's theorem, it provides the highest level of security without trusting any device. However, its implementation requires a flawless Bell test, making the system complex and costly. Furthermore, the random number generation rate is limited to below Mbps, which cannot meet the needs of modern high-speed communication.

[0005] 3) Semi-device-independent QRNG: As a compromise, it reduces some trust assumptions (such as source device independence), but its implementation is mostly based on discrete optical components (such as beam splitters, phase modulators, etc.), resulting in a large system size, poor stability and high power consumption. In order to maintain the phase stability of optical interference, complex active feedback control loops (such as phase-locked loops) are usually required, which are more sensitive to environmental vibration and temperature changes, and are difficult to deploy stably in harsh environments outside the laboratory.

[0006] Therefore, there is an urgent need in this field for a technical solution that can simultaneously overcome the above-mentioned bottlenecks: while ensuring high security independent of source devices, achieving a high random number generation rate (over 10Gbps) comparable to fully trusted QRNGs, and achieving high reliability, passive stability, small size and low power consumption through chip integration. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a source device-independent quantum random number generator based on an integrated photonic chip, which can effectively overcome the shortcomings of the existing technology, such as difficulty in achieving high security and high random number generation rate, as well as the complexity, poor stability and high power consumption of the device.

[0009] (II) Technical Solution

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A source-device-independent quantum random number generator based on an integrated photonic chip includes the following functional modules:

[0012] A quantum source, trustless, provides quantum state input to an integrated heterodyne receiver;

[0013] The integrated heterodyne receiver, which is trusted and fully characterizable, consists of an integrated photonic chip (PIC) and its accompanying high-speed, low-noise electronic readout chain. It is responsible for heterodyne measurement of the input quantum state, extracting the Q and P orthogonal components of quantum fluctuations, and converting them into corresponding digital signals.

[0014] The digital processing and random number extraction unit receives Q and P digital signals, performs system calibration and signal preprocessing, and executes secure random number extraction.

[0015] Preferably, the integrated photonic chip PIC is manufactured using SOI technology. Its core is a 4×4 multimode interferometer (MMI) as a 90° optical mixer. The four output channels of the MMI are routed to four grating couplers (GC) through symmetrical waveguides to ensure consistent optical path length and loss.

[0016] Preferably, two pairs of high-speed photodiodes (PDs) are integrated on the four grating couplers (GCs) via vertical coupling to form two balanced detectors (BPDs). The balanced detectors (BPDs) suppress common-mode noise and local oscillator intensity noise through differential detection and output a weak current signal.

[0017] Preferably, the electronic read chain includes a low-noise transimpedance amplifier (TIA), a three-stage amplifier chain, and a radio frequency interface;

[0018] The transimpedance amplifier (TIA) performs current-to-voltage conversion and pre-amplification on the current signal output from the balanced detector (BPD).

[0019] A three-stage amplifier chain further amplifies the pre-amplified voltage signal and outputs it to subsequent circuits via an RF interface.

[0020] Preferably, the security of the quantum random number generator does not depend on the quantum source, but on the physical characteristics of heterodyne measurement, where the upper limit of the probability of an attacker guessing the result of each measurement is [value missing]. ;

[0021] in, The upper bound of the probability guess for the attacker represents the maximum probability of successfully guessing a single measurement result X, given that the attacker possesses side information E. , These represent the resolutions of the integrated heterodyne receiver on the Q and P quadrature components, respectively.

[0022] Preferably, the digital processing and random number extraction unit determines the resolution of the integrated heterodyne receiver on the Q and P quadrature components through system calibration. , ,include:

[0023] Injecting quantum states into an integrated heterodyne receiver and scanning the optical power P of the local oscillator. LO Simultaneously, the variances of the Q and P digital signals were measured and compared with the optical power P. LO Linear fitting is performed to obtain the receiver calibration curve;

[0024] The scaling factor required to convert the ADC readings to phase space vacuum units is calculated based on the slope of the receiver calibration curve, thus obtaining the resolution of the integrated heterodyne receiver on the Q and P quadrature components. , .

[0025] Preferably, the quantum conditional minimum entropy of the quantum random number generator is: ;

[0026] in, Let be the quantum conditional minimum entropy, representing the minimum remaining uncertainty of a single measurement result X when the attacker possesses side information E. At the maximum local oscillator optical power of 21.15 mW, the quantum conditional minimum entropy is... ;

[0027] After preprocessing and resampling to 2 GSps, the final safe random number generation rate is: .

[0028] Preferably, in order to maximize randomness and avoid classical noise, the digital processing and random number extraction unit selects a frequency window of 400MHz to 1400MHz for random number extraction, with a signal-to-noise ratio higher than 8dB within this frequency window, and finally uses the Toeplitz hash function to extract the final secure random number.

[0029] (III) Beneficial Effects

[0030] Compared with existing technologies, the source-device-independent quantum random number generator based on an integrated photonic chip provided by this invention has the following advantages:

[0031] 1) Leap in security: It realizes a source device-independent security framework. Even if the quantum source is completely maliciously controlled, it can ensure the unpredictability and privacy of the output random numbers based on the physical characteristics of the integrated heterodyne receiver. The security is far higher than that of a fully trusted QRNG.

[0032] 2) Leading performance: While ensuring the above-mentioned high security, the secure random number generation rate exceeds 20Gbps, which is the fastest semi-device-independent / source-device-independent QRNG at present, and can directly meet the real-time requirements of the next generation of high-speed QKD systems.

[0033] 3) A revolution in integration and reliability: By custom-integrated photonic chip (PIC), the complex heterodyne measurement system that traditionally occupies the entire optical platform is integrated onto a microchip, realizing a chip-level solution. The size and weight of the device are reduced by several orders of magnitude. The core optical component (4×4 multimode interferometer MMI) is a completely passive structure, requiring no active feedback. It has stable performance in a wide temperature range of 15°C to 80°C, is inherently vibration resistant, and is particularly suitable for harsh environments such as satellites and drones.

[0034] 4) Low power consumption and high sensitivity: Passive optical design and high-efficiency electronic interface reduce overall power consumption. The system can still maintain a high random number generation rate under low local oscillator optical power, which is beneficial to energy saving and system design. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0036] Figure 1 This is a schematic diagram of the system of the present invention;

[0037] Figure 2 This is a schematic diagram of the integrated photonic chip (PIC) structure in this invention;

[0038] Figure 3 This is a calibration curve of the receiver in this invention;

[0039] Figure 4 This is a graph showing the power spectral density amplitude and signal-to-noise ratio as a function of frequency in this invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0041] The core of this invention lies in achieving source device independence at the hardware level through a customized integrated heterodyne receiver based on a 4×4 multimode interferometer (MMI). This ensures a high level of security while achieving industry-leading random number generation rate, chip-level integration, and passive operational stability.

[0042] The following describes the specific functional modules of the source-device-independent quantum random number generator based on an integrated photonic chip provided by this invention, using concrete examples (such as...). Figure 1 (as shown) and technical effects.

[0043] The functional modules of the quantum random number generator include:

[0044] A quantum source, trustless, provides quantum state input to an integrated heterodyne receiver (in practical applications, a vacuum state is typically injected to take advantage of its inherent vacuum fluctuations).

[0045] The integrated heterodyne receiver, which is trusted and fully characterizable, consists of an integrated photonic chip (PIC) and its accompanying high-speed, low-noise electronic readout chain. It is responsible for heterodyne measurement of the input quantum state, extracting the Q and P orthogonal components of quantum fluctuations, and converting them into corresponding digital signals.

[0046] The digital processing and random number extraction unit receives Q and P digital signals, performs system calibration and signal preprocessing, and executes secure random number extraction.

[0047] I. Integrated Photonic Chip (PIC)

[0048] like Figure 2 As shown, the integrated photonic chip PIC is manufactured using SOI technology (with a top silicon thickness of 220nm and a buried oxide layer thickness of 2μm, suitable for the 1550nm communication band). The core is a 4×4 multimode interferometer (MMI) as a 90° optical mixer. The four output channels of the MMI are routed to four grating couplers (GC) via symmetrical waveguides (designed for the 1550nm TE mode, with a typical coupling loss of about 3dB / coupling point, which can be further optimized through subsequent end-face coating) to ensure consistent optical path length and loss.

[0049] Two pairs of high-speed photodiodes (PDs) are integrated on four grating couplers (GCs) via vertical coupling to form two balanced detectors (BPDs). The balanced detectors (BPDs) suppress common-mode noise and local oscillator intensity noise through differential detection and output a weak current signal.

[0050] The advantages of integrating a photonic chip (PIC) in the technical solution of this application are as follows:

[0051] Passive stabilization: Unlike interferometers that require active temperature control and phase feedback, the multimode interferometer (MMI) is a completely passive component. Experiments have shown that its phase deviation is less than 6% over a wide temperature range of 15°C to 80°C, giving it an inherent ability to resist environmental interference.

[0052] High precision: The test structure shows that the average phase spacing of the multimode interferometer (MMI) is approximately 89.99°, which is very close to the ideal 90°, ensuring the accuracy of heterodyne measurements;

[0053] Compactness and scalability: The entire optical system is integrated on a microchip, making the device much smaller than discrete optical component solutions, laying the technological foundation for miniaturized QRNG modules.

[0054] II. Electronic Read Chain

[0055] The electronic read chain includes a low-noise transimpedance amplifier (TIA), a three-stage amplifier chain, and an RF interface;

[0056] The transimpedance amplifier (TIA) (bandwidth 2.5GHz) performs current-to-voltage conversion and pre-amplification on the current signal output from the balanced detector (BPD).

[0057] A three-stage amplifier chain further amplifies the pre-amplified voltage signal and outputs it to subsequent circuits via an RF interface.

[0058] In the technical solution of this application, the electronic read chain design achieves an optimal balance between bandwidth, gain, and noise, realizing approximately 6pA / The low input reference noise ensures that the system can distinguish weak quantum shot noise.

[0059] III. Security Basics

[0060] The security of a quantum random number generator does not depend on the quantum source, but rather on the physical properties of heterodyne measurements. An attacker's probability of guessing the outcome of each measurement is capped at [value missing]. ;

[0061] in, The upper bound of the probability guess for the attacker represents the maximum probability of successfully guessing a single measurement result X, given that the attacker possesses side information E. , These represent the resolutions of the integrated heterodyne receiver on the Q and P quadrature components, respectively.

[0062] IV. System Calibration

[0063] The digital processing and random number extraction unit determines the resolution of the integrated heterodyne receiver on the Q and P quadrature components through system calibration. , ,include:

[0064] Injecting quantum states into an integrated heterodyne receiver and scanning the optical power P of the local oscillator. LO Simultaneously, the variances of the Q and P digital signals were measured and compared with the optical power P. LO Perform linear fitting to obtain the receiver calibration curve (e.g.) Figure 3 (as shown)

[0065] The scaling factor required to convert the ADC readings to phase space vacuum units is calculated based on the slope of the receiver calibration curve, thus obtaining the resolution of the integrated heterodyne receiver on the Q and P quadrature components. , .

[0066] With a maximum local oscillator optical power of 21.15mW, the Q-channel resolution of the integrated heterodyne receiver, after system calibration, is [value missing]. Vacuum unit (per ADC codeword), P-channel resolution Vacuum unit (per ADC codeword).

[0067] V. Entropy Calculation and Random Number Generation Rate

[0068] The quantum conditional minimum entropy of a quantum random number generator is ;

[0069] in, Let be the quantum conditional minimum entropy, representing the minimum remaining uncertainty of a single measurement result X when the attacker possesses side information E. At the maximum local oscillator optical power of 21.15 mW, the quantum conditional minimum entropy is... ;

[0070] After preprocessing and resampling to 2 GSps, the final safe random number generation rate is: .

[0071] VI. Spectrum Selection and Random Extraction

[0072] like Figure 4 As shown, in order to maximize randomness and avoid classical noise, the digital processing and random number extraction unit selects a spectrum window of 400MHz~1400MHz for random number extraction. The signal-to-noise ratio within this spectrum window is higher than 8dB. Finally, the Toeplitz hash function (e.g., matrix size 17600×11008) is used to extract the final secure random number.

[0073] To better illustrate the technical solution of this application, a specific example will be used for detailed explanation below.

[0074] In this example, a 1550nm distributed feedback laser is used as an external local oscillator. Its output power is set to a maximum of 21.15mW via a variable optical attenuator. Polarization is optimized by a polarization controller before being coupled into an integrated photonic chip (PIC) via an optical fiber array. A vacuum state is injected into the PIC. Two pairs of differential RF signals output from the integrated heterodyne receiver are used; one path is terminated, and the other is passed through a 48MHz high-pass filter before being digitized by an oscilloscope at a sampling rate of 25GSps and 8-bit resolution. The acquired raw data is transmitted to a computer for offline processing, which includes:

[0075] 1) Preprocessing: Resample the data to 2GSps and extract the data within the 400MHz~1400MHz spectrum window;

[0076] 2) Random number extraction: A Toeplitz matrix of size 17600×11008 is used to perform a general hash operation on the preprocessed data. This parameter setting ensures the security of the entire extraction process. ε'~10 -10 ;

[0077] 3) Specific process and result verification: The acquisition time was 10 seconds. The oscilloscope digitized the two orthogonal components at a sampling rate of 25 GSps, acquiring approximately 50 GB of raw data. After preprocessing and resampling to 2 GSps, and extracting data within the 400 MHz to 1400 MHz spectrum window, approximately 2 × 10⁻⁶ effective sampling points were obtained. 10 The data were extracted using a Toeplitz matrix of size 17600×11008. The extraction process took about 1 second, and about 2.5 GB of final random numbers were extracted from 50 GB of original data. The calculated secure random number generation rate was about 20.015 Gbps. These data passed all 15 tests of the NIST SP 800-22 test suite (as shown in Table 1 below), demonstrating their good statistical randomness.

[0078] Table 1. Results of all 15 tests in the NIST SP 800-22 test suite.

[0079]

[0080] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A source-device-independent quantum random number generator based on an integrated photonic chip, characterized in that: Includes the following functional modules: A quantum source, trustless, provides quantum state input to an integrated heterodyne receiver; The integrated heterodyne receiver, which is trusted and fully characterizable, consists of an integrated photonic chip (PIC) and its accompanying high-speed, low-noise electronic readout chain. It is responsible for heterodyne measurement of the input quantum state, extracting the Q and P orthogonal components of quantum fluctuations, and converting them into corresponding digital signals. The digital processing and random number extraction unit receives Q and P digital signals, performs system calibration and signal preprocessing, and executes secure random number extraction.

2. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 1, characterized in that: The integrated photonic chip PIC is manufactured using SOI technology. Its core is a 4×4 multimode interferometer (MMI) as a 90° optical mixer. The four output channels of the MMI are routed to four grating couplers (GC) through symmetrical waveguides to ensure consistent optical path length and loss.

3. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 2, characterized in that: Two pairs of high-speed photodiodes (PDs) are integrated on the four grating couplers (GCs) via vertical coupling to form two balanced detectors (BPDs). The balanced detectors (BPDs) suppress common-mode noise and local oscillator intensity noise through differential detection and output a weak current signal.

4. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 3, characterized in that: The electronic read chain includes a low-noise transimpedance amplifier (TIA), a three-stage amplifier chain, and an RF interface. The transimpedance amplifier (TIA) performs current-to-voltage conversion and pre-amplification on the current signal output from the balanced detector (BPD). A three-stage amplifier chain further amplifies the pre-amplified voltage signal and outputs it to subsequent circuits via an RF interface.

5. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 4, characterized in that: The security of the quantum random number generator does not depend on the quantum source, but on the physical characteristics of heterodyne measurements. An attacker's guess of the upper limit of the probability of each measurement result is... ; in, The upper bound of the probability guess for the attacker represents the maximum probability of successfully guessing a single measurement result X, given that the attacker possesses side information E. , These represent the resolutions of the integrated heterodyne receiver on the Q and P quadrature components, respectively.

6. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 5, characterized in that: The digital processing and random number extraction unit determines the resolution of the integrated heterodyne receiver on the Q and P quadrature components through system calibration. , ,include: Injecting quantum states into an integrated heterodyne receiver, scanning the optical power P of the local oscillator LO At the same time, the variances of Q, P digital signals are measured and compared with the optical power P LO Linear fitting is performed to obtain the receiver calibration curve; The scaling factor required to convert the ADC readings to phase space vacuum units is calculated based on the slope of the receiver calibration curve, thus obtaining the resolution of the integrated heterodyne receiver on the Q and P quadrature components. , .

7. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 6, characterized in that: The quantum conditional minimum entropy of the quantum random number generator is ; in, Let be the quantum conditional minimum entropy, representing the minimum remaining uncertainty of a single measurement result X when the attacker possesses side information E. At the maximum local oscillator optical power of 21.15 mW, the quantum conditional minimum entropy is... ; After preprocessing and resampling to 2 GSps, the final safe random number generation rate is: .

8. The source-device-independent quantum random number generator based on an integrated photonic chip according to claim 7, characterized in that: To maximize randomness and avoid classical noise, the digital processing and random number extraction unit selects a frequency window of 400MHz to 1400MHz for random number extraction. The signal-to-noise ratio within this frequency window is higher than 8dB. Finally, the Toeplitz hash function is used to extract the final secure random number.

Images