CN113810186B - High-precision real-time prediction method and system for self-adaptive quantum efficiency - Google Patents
High-precision real-time prediction method and system for self-adaptive quantum efficiency Download PDFInfo
- Publication number
- CN113810186B CN113810186B CN202111100710.8A CN202111100710A CN113810186B CN 113810186 B CN113810186 B CN 113810186B CN 202111100710 A CN202111100710 A CN 202111100710A CN 113810186 B CN113810186 B CN 113810186B
- Authority
- CN
- China
- Prior art keywords
- neural network
- detector
- quantum efficiency
- deep neural
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000013528 artificial neural network Methods 0.000 claims abstract description 62
- 238000012549 training Methods 0.000 claims abstract description 48
- 230000010355 oscillation Effects 0.000 claims abstract description 40
- 238000010521 absorption reaction Methods 0.000 claims description 22
- 230000003287 optical effect Effects 0.000 claims description 15
- 239000000969 carrier Substances 0.000 claims description 13
- 229910052732 germanium Inorganic materials 0.000 claims description 12
- GNPVGFCGXDBREM-UHFFFAOYSA-N germanium atom Chemical compound [Ge] GNPVGFCGXDBREM-UHFFFAOYSA-N 0.000 claims description 12
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 claims description 11
- 229910052710 silicon Inorganic materials 0.000 claims description 11
- 239000010703 silicon Substances 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000002800 charge carrier Substances 0.000 claims description 6
- 239000000463 material Substances 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 210000005036 nerve Anatomy 0.000 claims description 4
- 230000005684 electric field Effects 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims 5
- 230000006978 adaptation Effects 0.000 claims 1
- 238000012544 monitoring process Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000001427 coherent effect Effects 0.000 description 3
- 230000010354 integration Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000001208 nuclear magnetic resonance pulse sequence Methods 0.000 description 2
- 230000010287 polarization Effects 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/08—Key distribution or management, e.g. generation, sharing or updating, of cryptographic keys or passwords
- H04L9/0816—Key establishment, i.e. cryptographic processes or cryptographic protocols whereby a shared secret becomes available to two or more parties, for subsequent use
- H04L9/0852—Quantum cryptography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/70—Photonic quantum communication
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Electromagnetism (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Biophysics (AREA)
- Computer Security & Cryptography (AREA)
- Optics & Photonics (AREA)
- Computational Mathematics (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Optical Modulation, Optical Deflection, Nonlinear Optics, Optical Demodulation, Optical Logic Elements (AREA)
- Optical Communication System (AREA)
Abstract
The application provides a method and a system for high-precision real-time prediction of self-adaptive quantum efficiency, which mainly comprise the following steps: the method comprises the steps of constructing a local dictionary set by utilizing the light intensity of an input detector, the external current of an output detector under the light intensity and the fluctuation percentage of local oscillation light photons of the input detector, carrying out initial training and joint training on a deep neural network by utilizing the local dictionary set, carrying out high-precision real-time prediction on the quantum efficiency of a chip CVQKD system by utilizing the trained deep neural network, and accurately evaluating the actual security key rate of the system according to the accurate real-time quantum efficiency prediction result. The application considers the actual safety existing in the application of the chip CVQKD system, and can greatly extract the self-adaptive characteristic performance of the system by utilizing the deep neural network, thereby providing a method for accurately estimating the quantum efficiency along with the fluctuation of local oscillation light in the chip CVQKD system in real time. The scheme is simple in implementation mode, convenient to popularize on a large scale and suitable for large-scale business.
Description
Technical Field
The application relates to a security hole defense method, in particular to a self-adaptive quantum efficiency high-precision real-time prediction method and a self-adaptive quantum efficiency high-precision real-time prediction system, and especially relates to a self-adaptive quantum efficiency high-precision real-time prediction method based on light intensity monitoring and a deep neural network.
Background
In the field of quantum cryptography, quantum key distribution (QKD, quantum key distribution) technology has achieved rapid development and tremendous effort in recent years due to its unconditional security based on quantum mechanical guarantees. Quantum key distribution technology is now mature, and it enables authenticated parties Alice and Bob to share a key through an insecure quantum channel. In particular, this quantum channel can be freely controlled and handled by a potential eavesdropper. Currently, quantum key distribution systems are largely divided into two major categories, namely, a discrete-variable quantum key distribution (DVQKD) system and a continuous-variable quantum key distribution (CVQKD) system. CVQKD systems that utilize a weak coherent state and a balanced homodyne detector are well compatible with classical optical communication systems, as compared to DVQKD systems. It is therefore an urgent task to continue to explore CVQKD systems and to promote early commercialization thereof. Continuous variable quantum key component protocols based on gaussian modulated coherent states (GMCS-CVQKD) have been demonstrated to have unconditional security under single, collective and coherent attacks. Meanwhile, CVQKD has also made good progress in long-distance transmission experiments at the level of 100-200 km, and has made rapid progress in recent years. In recent years, photonic integration technology has provided an important technical approach to solve the miniaturization, cost-effectiveness and compatibility problems of conventional fiber-based CVQKD systems in existing optical communication systems. In addition, silicon-based photoelectric integration is a mature branch of integrated photonics technology, and rapid progress is also made in quantum source, detection and other aspects. In particular, chip-based silicon-based CVQKD systems have recently been first validated in 2m optical fibers, which means that CVQKD systems take an important step in the direction of integration.
However, since CVQKD does not take into account the actual drawbacks of the system in detail in theoretical security certification, almost all CVQKD systems can face potential practical security risks. Fortunately, research into the defects that third party attacker Eve may use to hide attacks has been conducted more thoroughly. But the latest breakthrough of CVQKD systems, chip-based CVQKD systems, as a new type of CVQKD system, may also face serious potential practical security issues. In recent years, although scholars have raised practical security issues for chip-based CVQKD systems. Unfortunately, however, the actual security problem study of chip-based CVQKD is almost blank as compared to the mature study of the actual security of fiber-based CVQKD. In fact, as the size of a CVQKD system is reduced to an on-chip level, it will be very different from systems built using discrete components. This is because many previously ignored effects will be highlighted, which may lead to system security vulnerabilities.
Current research into chip-based CVQKD systems is focused mainly on how to physically implement it, but many practical security issues need to be carefully considered simultaneously. For example, non-uniform or rough waveguides or heavy doping in an integrated detector can result in non-negligible free carrier absorption and scattering losses, etc. The scholars find that the jitter of the local oscillation light in the CVQKD system can cause practical safety problems, but the current academy only considers the monitoring of the lens noise calibration deviation, and does not consider the influence of the jitter on the quantum efficiency of the detector. However, a change in carrier mobility caused by a minute local oscillation optical jitter eventually leads to a change in quantum efficiency. These imperfections, which were not previously considered in chip-based CVQKD systems, can lead to deviations in the evaluation of excessive noise and other parameters between legitimate communicating parties. Chip-based CVQKD systems may face severe practical security vulnerabilities.
Fortunately, although the quantum efficiency in the silicon-based integrated detector in the chip-based CVQKD system is changed due to the fluctuation of local oscillator light intensity, the application provides a self-adaptive quantum efficiency high-precision real-time prediction method based on light intensity monitoring and a deep neural network, which can predict the quantum efficiency in the integrated detector in the chip CVQKD system in real time with high precision, thereby thoroughly defending the potential security vulnerability of the system caused by the fluctuation of local oscillator light to the variable sub-efficiency and further accurately evaluating the actual security key rate of the system with strict actual security.
In the chinese patent document with publication number CN110635895a, a CVQKD transmission device and method based on self-stabilizing intensity modulation, a CVQKD system, comprising a pulse laser, a first beam splitter, an adjustable optical attenuator, a self-stabilizing intensity modulation device, a first phase modulator, a delay device and a first polarizing beam splitter; the pulse laser is used for generating a periodic pulse sequence; the first beam splitter is used for splitting the pulse sequence; the adjustable optical attenuator is used for attenuating the signal light and inputting the signal light into the self-stabilizing intensity modulation device; the self-stabilizing intensity modulation device comprises a second beam splitter and a second phase modulator; the first phase modulator generates Gaussian modulated signal light from the signal light; the delay device delays the Gaussian modulated signal light and inputs the delayed signal light into the first polarization beam splitter; and the first polarization beam splitter outputs the delayed Gaussian modulated signal light and the local oscillation beam after combining.
Disclosure of Invention
Aiming at the defects in the prior art, the application aims to provide a self-adaptive quantum efficiency high-precision real-time prediction method and a self-adaptive quantum efficiency high-precision real-time prediction system.
The application provides a self-adaptive quantum efficiency high-precision real-time prediction method, which comprises the following steps:
step S1: collecting data required for deep neural network training, including the light intensity P input to the detector 0 And outputting the external current I of the detector under the light intensity ph ；
Step S2: calculating the fluctuation percentage of local oscillation light photons of an input detectorUsing the intensity P of the input detector 0 External current I of output detector under light intensity ph The local oscillation photon number fluctuation percentage of the input detectorConstructing a local dictionary set D: />
Step S3: construction of deep nerve modelPerforming initial training on the deep neural network by using the local dictionary set obtained in the step S2, entering the step S4 when the deep neural network meets the judging condition capable of accurately predicting the local oscillation photon fluctuation percentage according to the light intensity of the input detector, otherwise, continuing to expand the local dictionary set until the condition is met;
step S4: performing joint training on the deep neural network, completing the joint training stage when the deep neural network meets the judgment condition that the quantum efficiency can be accurately predicted according to the input local oscillator light intensity, otherwise, continuing training;
step S5: and carrying out high-precision real-time prediction on the quantum efficiency of the chip CVQKD system by using the trained deep neural network, and accurately evaluating the actual security key rate of the system according to the accurate real-time quantum efficiency prediction result.
Preferably, the local oscillation photon number fluctuation percentage of the input detector in the step S2The calculation is as follows:
wherein:
percentage of power fluctuation of local oscillation light input to detector, wherein P 0 Representing the intensity of the light input to the detector, I ph Representing the external current generated by the detector due to the input light intensity;
ρ 0 : bulk density of free carriers;
V eff : effective volume measurement of the region in which the free carriers are located;
m * : the effective mass of the charge carrier;
f rep : the repetition frequency of the chip CVQKD system;
N LO : the number of photons contained in each local oscillation optical pulse;
hv: h is the Planck constant, v is the frequency of the photon, and the product of the two represents the energy of the photon;
represents the average initial velocity of the free carrier ensemble before photon absorption;
after element replacement, the method comprises the following steps:wherein n represents the refractive index, ε, of the germanium material of a silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge quantity of electrons beta IB Representing the absorption coefficient between the bands->Represents the initial quantum efficiency constant, beta s Representing the scattering absorption coefficient.
Preferably, the decision condition in the step S3 is: satisfy epsilon for a given precision 1 And delta 1 For any meetingIs set to the input light intensity P of 0 If the mapping result of the deep neural network under the input light intensity meets +.>And when the training is finished, the deep neural network completes initial training.
Preferably, the decision condition in the step S4 is: epsilon for a given precision 2 If the following conditions are satisfied: wherein->Represents->And the second component in the function value is considered to be the end of the deep neural network joint training.
Preferably, in the step S5, the input light intensity is monitored according to beam splittingThe trained deep neural network is utilized to accurately predict the quantum efficiency in real time: />Wherein->
The application provides a self-adaptive quantum efficiency high-precision real-time prediction system, which comprises the following modules:
module M1: collecting data required for deep neural network training, including the light intensity P input to the detector 0 And outputting the external current I of the detector under the light intensity ph ；
Module M2: calculating the fluctuation percentage of local oscillation light photons of an input detectorUsing the intensity P of the input detector 0 External current I of output detector under light intensity ph The local oscillation photon number fluctuation percentage of the input detectorConstructing a local dictionary set D: />
Module M3: construction of deep nerve modelPerforming initial training on the deep neural network by using the local dictionary set in the module M2, executing the module M4 when the deep neural network meets the judging condition capable of accurately predicting the local oscillation photon fluctuation percentage according to the light intensity of the input detector, and otherwise, continuing to expand the local dictionary set until the condition is met;
module M4: performing joint training on the deep neural network, completing the joint training stage when the deep neural network meets the judgment condition that the quantum efficiency can be accurately predicted according to the input local oscillator light intensity, otherwise, continuing training;
module M5: and carrying out high-precision real-time prediction on the quantum efficiency of the chip CVQKD system by using the trained deep neural network, and accurately evaluating the actual security key rate of the system according to the accurate real-time quantum efficiency prediction result.
Preferably, the module M2 inputs the local oscillation photon number fluctuation percentage of the detectorThe calculation is as follows:
wherein:
percentage of power fluctuation of local oscillation light input to detector, wherein P 0 Representing the intensity of the light input to the detector, I ph Representing the external current generated by the detector due to the input light intensity;
ρ 0 : bulk density of free carriers;
V eff : effective volume measurement of the region in which the free carriers are located;
m * : the effective mass of the charge carrier;
f rep : the repetition frequency of the chip CVQKD system;
N LO : the number of photons contained in each local oscillation optical pulse;
hv: h is the Planck constant, v is the frequency of the photon, and the product of the two represents the energy of the photon;
represents the average initial velocity of the free carrier ensemble before photon absorption;
after element replacement, the method comprises the following steps:wherein n represents the refractive index, ε, of the germanium material of a silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge quantity of electrons beta IB Representing the absorption coefficient between the bands->Represents the initial quantum efficiency constant, beta s Representing the scattering absorption coefficient.
Preferably, the decision condition in the module M3 is: satisfy epsilon for a given precision 1 And delta 1 For any meetingIs set to the input light intensity P of 0 If the mapping result of the deep neural network under the input light intensity meets +.>And when the training is finished, the deep neural network completes initial training.
Preferably, the decision condition in the module M4 is: epsilon for a given precision 2 If the following conditions are satisfied: wherein->Representative ofAnd the second component in the function value is considered to be the end of the deep neural network joint training.
Preferably, in the module M5, the input light intensity is monitored according to beam splittingThe trained deep neural network is utilized to accurately predict the quantum efficiency in real time: />Wherein->
Compared with the prior art, the application has the following beneficial effects:
1. the current chip-based CVQKD system is still in the experimental verification stage and has not entered into large-scale business. Therefore, the research on the actual safety problem in the chip CVQKD is almost blank, the application considers the actual safety existing in the application of the chip CVQKD system, and can greatly extract the performance of the self-adaptive characteristics of the system by utilizing the deep neural network, and provides a self-adaptive quantum efficiency high-precision real-time prediction method based on light intensity monitoring and the deep neural network, thereby providing a method for real-time accurate estimation of the quantum efficiency along with local oscillation fluctuation in the chip CVQKD system and thoroughly avoiding the vulnerability from the source.
2. Although some mathematical methods can be adopted to perform preliminary rough estimation on quantum efficiency which varies with local oscillator light intensity fluctuation in the chip CVQKD system, the method is not accurate and thorough all the time. Such mathematical based approaches often require conservative estimates, which means that security in achieving the key rate needs to be guaranteed at the expense of system part security key rate performance, which is not tolerable in many situations. The method can estimate the real quantum efficiency of the system in real time with high precision, so that the quantum efficiency of the system can be estimated in real time with high precision without compromising the performance of the system fundamentally, and the method has great significance in constructing and guaranteeing the actual safety of a high-performance chip CVQKD.
3. The mathematical essence of the method is based on mathematical modeling of the whole physical mechanism process of the change of the quantum efficiency of the integrated detector in the chip CVQKD system along with the fluctuation of local oscillation light, so that the method is essentially suitable for thoroughly solving the actual safety problem generally faced by the integrated detector in most chip-based continuous variable quantum key distribution systems. Has quite universality.
4. The scheme is simple in implementation mode, and the prediction neural network in the self-adaptive quantum efficiency high-precision real-time prediction method based on the light intensity monitoring and the deep neural network can be popularized on a large scale once being trained, and the marginal cost is almost zero, so that the method is suitable for large-scale business.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the detailed description of non-limiting embodiments, given with reference to the accompanying drawings in which:
FIG. 1 is a flow chart of a method according to an embodiment of the application.
Detailed Description
The present application will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the present application, but are not intended to limit the application in any way. It should be noted that variations and modifications could be made by those skilled in the art without departing from the inventive concept. These are all within the scope of the present application.
The specific implementation mode of the quantum efficiency attack step of the local oscillator light intensity fluctuation variable detector is as follows: eve uses Alice and Bob to ignore the influence of the change of the actual security problem caused by the fluctuation of the local oscillator light intensity input into the silicon-based integrated detector on the quantum efficiency in the integrated detector in the receiving end of the silicon-based integrated CVQKD chip system, so that interception retransmission attack or other attack modes can be adopted to mask part of excessive noise caused by the self-interception system, and the interception of part of security keys is realized without discovery.
The high-precision real-time prediction method of the self-adaptive quantum efficiency is based on light intensity monitoring and a deep neural network, and referring to FIG. 1, comprises the following steps:
step S1: data required for training the deep neural network is collected, and the beam splitter is utilized to input the light intensity P of the detector 0 Beam splitting monitoring is performed while measuring the light intensity P 0 External current I of lower output detector ph Collecting a large amount of dataWhere i represents the sequence number of the data set, for a total of N sets of data.
Step S2: calculating the fluctuation percentage of local oscillation light photons of an input detectorAnd constructing a local dictionary set to prepare for subsequent deep neural network training. The calculation method comprises the following steps:
wherein:
the power fluctuation percentage of the local oscillation light input into the detector is equal to the power fluctuation percentage of the local oscillation light input into the detector due to the input light intensity P 0 The generated external current I ph Correlation;
ρ 0 : bulk density of free carriers;
V eff : effective volume measurement of the region in which the free carriers are located;
m * : the effective mass of the charge carrier;
f rep : the repetition frequency of the chip CVQKD system;
N LO : the number of photons contained in each local oscillation optical pulse;
hv: represents the energy of a photon, where h is the Planck constant and v is the frequency of the photon;
represents the average initial velocity of the free carrier ensemble before photon absorption;
after element replacement, the method comprises the following steps:
wherein n represents the refractive index, ε, of the germanium material of a silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge quantity of electrons beta IB Represents the absorption coefficient between the sheets,represents the initial quantum efficiency constant, beta s Representing the scattering absorption coefficient.
When the input light intensity and the external current collected according to the preparation step are finished, the local oscillation light photon fluctuation percentage of the input detector is calculatedBy means of "input local oscillator light intensity P 0 Output under the intensity of the input local oscillationExternal current I ph The local oscillation photon number fluctuation percentage of the input local oscillation light intensity is +.>"three-dimensional data elements form a set, namely, a local dictionary set D is constructed: /> Preparation is made for initial training and joint training of the subsequent deep neural network. Wherein->Representing a local dictionary set, the ith element is: when the input light intensity is +.>The corresponding output external current is +.>The input local oscillator light intensity +.>The fluctuation percentage of the photon number of the lower local oscillation light is +.>The superscript dic means the abbreviation "dictionary" is commonly used.
Step S3: and (3) performing initial training on the deep neural network by using the local dictionary set obtained in the step S2. For the constructed deep neural network modelInitial training is carried out, and when the deep neural network meets the judgment condition capable of accurately predicting the fluctuation percentage of the local oscillation optical photons according to the light intensity input into the integrated detector, epsilon for given precision is met 1 And delta 1 For arbitrary satisfaction->Is set to the input light intensity P of 0 In this condition->Representing local dictionary concentration and input light intensity P 0 The difference in size between them is smaller than any small positive number delta 1 If the mapping result of the deep neural network under the input light intensity meets the following conditions: />|| || 2 Representing euclidean norms in vector space, where +.>Representing an input light intensity of P 0 Under the condition of (1), the two-dimensional real vector sequence formed by the output current value predicted by the neural network model and the value of the local oscillation photon fluctuation percentage is obtained, and then the vector is matched with P 0 Nearest neighbor +.>Corresponding vector +.>The euclidean norm therebetween being less than a given arbitrarily small positive number epsilon 1 (representing the prediction accuracy), the deep neural network completes the initial training phase and enters the joint training phase. Otherwise, continuing to expand the local dictionary set until the above condition is met.
Step S4: and carrying out joint training on the deep neural network, completing the joint training stage when the deep neural network meets the judgment condition that the quantum efficiency can be accurately predicted according to the input local oscillator light intensity, and otherwise, continuing training.
When the deep neural network meets the judgment condition capable of accurately predicting quantum efficiency according to the input local oscillator light intensity, the combination is completedTraining stage. I.e. epsilon for a given precision 2 If the following conditions are satisfied:wherein-> Represents->The second component in the function value, < > in the calculation of the above formula>Alice and Bob, which represent quantum efficiencies, have fixed constant values of quantum efficiencies, β IB Representing the absorption coefficient between the bands, beta s Representing the scattering absorption coefficient, beta fc Represents the free carrier absorption coefficient ρ 0 Representing the bulk density of free carriers, V eff An effective volume measure representing the region in which the free carriers are located, m * Representing the effective mass of the charge carrier. f (f) rep Representing the repetition frequency, N, of a chip CVQKD system LO Represents the number of photons contained in each local oscillation optical pulse, h is Planck constant, v is the frequency of photons, and the product of the two represents the energy of photons, < >>Represents the average initial velocity of the free carrier ensemble before photon absorption, n represents the refractive index, ε, of the germanium material of the silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge of the electrons, lambda representing the wavelength of the optical signal and E representing the intensity of the electric field in which the free carriers are located. And at this time, the deep neural network joint training is considered to be finished, otherwise, training is continued.
Step S5: high precision quantum efficiency of chip CVQKD system using trained deep neural networkAnd (5) predicting in real time. After the combined training is finished, according to the input light intensity monitored by beam splittingThe trained deep neural network is utilized to accurately predict the quantum efficiency in real time: />Wherein the method comprises the steps ofAnd according to the accurate real-time quantum efficiency prediction result, accurately evaluating the actual security key rate of the system.
Those skilled in the art will appreciate that the application provides a system and its individual devices, modules, units, etc. that can be implemented entirely by logic programming of method steps, in addition to being implemented as pure computer readable program code, in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Therefore, the system and various devices, modules and units thereof provided by the application can be regarded as a hardware component, and the devices, modules and units for realizing various functions included in the system can also be regarded as structures in the hardware component; means, modules, and units for implementing the various functions may also be considered as either software modules for implementing the methods or structures within hardware components.
The foregoing describes specific embodiments of the present application. It is to be understood that the application is not limited to the particular embodiments described above, and that various changes or modifications may be made by those skilled in the art within the scope of the appended claims without affecting the spirit of the application. The embodiments of the application and the features of the embodiments may be combined with each other arbitrarily without conflict.
Claims (8)
1. The high-precision real-time prediction method for the self-adaptive quantum efficiency is characterized by comprising the following steps of:
step S1: collecting data required for deep neural network training, including the light intensity P input to the detector 0 And outputting the external current I of the detector under the light intensity ph ；
Step S2: calculating the fluctuation percentage of local oscillation light photons of an input detectorUsing the intensity P of the input detector 0 External current I of output detector under light intensity ph The local oscillation photon number fluctuation percentage of the input detectorConstructing a local dictionary set D: />Wherein i represents the sequence number of the data set, and there are N sets of data in total;
local oscillator photon number fluctuation percentage of input detectorThe calculation is as follows:
wherein:
percentage of power fluctuation of local oscillation light input to detector, wherein P 0 Representing the intensity of the light input to the detector, I ph Representing the external current generated by the detector due to the input light intensity;
ρ 0 : bulk density of free carriers;
V eff : effective volume measurement of the region in which the free carriers are located;
m * : the effective mass of the charge carrier;
f rrep : the repetition frequency of the chip CVQKD system;
N LO : the number of photons contained in each local oscillation optical pulse;
hv: h is the Planck constant, v is the frequency of the photon, and the product of the two represents the energy of the photon;
represents the average initial velocity of the free carrier ensemble before photon absorption;
after element replacement, the method comprises the following steps:wherein n represents the refractive index, ε, of the germanium material of a silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge quantity of electrons beta IB Representing the absorption coefficient between the bands->Represents the initial quantum efficiency constant, beta s Representing the scattering absorption coefficient; step S3: construction of deep nerve modelPerforming initial training on the deep neural network by using the local dictionary set obtained in the step S2, entering the step S4 when the deep neural network meets the judging condition capable of accurately predicting the local oscillation photon fluctuation percentage according to the light intensity of the input detector, otherwise, continuing to expand the local dictionary set until the condition is met;
step S4: performing joint training on the deep neural network, completing the joint training stage when the deep neural network meets the judgment condition that the quantum efficiency can be accurately predicted according to the input local oscillator light intensity, otherwise, continuing training;
step S5: and carrying out high-precision real-time prediction on the quantum efficiency of the chip CVQKD system by using the trained deep neural network, and accurately evaluating the actual security key rate of the system according to the accurate real-time quantum efficiency prediction result.
2. The adaptive quantum efficiency high precision real time prediction method according to claim 1, characterized in that: the judging conditions in the step S3 are as follows: satisfy epsilon for a given precision 1 And delta 1 For any meeting Is set to the input light intensity P of 0 If the mapping result of the deep neural network under the input light intensity meets the requirementAnd when the training is finished, the deep neural network completes initial training.
3. The adaptive quantum efficiency high precision real time prediction method according to claim 1, characterized in that: the judging conditions in the step S4 are as follows: epsilon for a given precision 2 If the following conditions are satisfied:wherein the method comprises the steps ofRepresentative ofThe second component in the function values is considered to be the end of the deep neural network joint training, ++>Alice and Bob, which represent quantum efficiencies, have fixed constant values of quantum efficiencies, β fc Represents the free carrier absorption coefficient, lambda represents the wavelength of the optical signal, and E represents the strength of the electric field in which the free carrier is located.
4. The adaptation of claim 1The high-precision real-time prediction method of the quantum efficiency is characterized by comprising the following steps of: in the step S5, the input light intensity is monitored according to beam splittingThe trained deep neural network is utilized to accurately predict the quantum efficiency in real time: />Wherein->
5. The self-adaptive quantum efficiency high-precision real-time prediction system is characterized by comprising the following modules:
module M1: collecting data required for deep neural network training, including the light intensity P input to the detector 0 And outputting the external current I of the detector under the light intensity ph ；
Module M2: calculating the fluctuation percentage of local oscillation light photons of an input detectorUsing the intensity P of the input detector 0 External current I of output detector under light intensity ph The local oscillation photon number fluctuation percentage of the input detectorConstructing a local dictionary set D: />Wherein i represents the sequence number of the data set, and there are N sets of data in total;
local oscillator photon number fluctuation percentage of input detectorThe calculation is as follows:
wherein:
percentage of power fluctuation of local oscillation light input to detector, wherein P 0 Representing the intensity of the light input to the detector, I ph Representing the external current generated by the detector due to the input light intensity;
ρ 0 : bulk density of free carriers;
V eff : effective volume measurement of the region in which the free carriers are located;
m * : the effective mass of the charge carrier;
f rep : the repetition frequency of the chip CVQKD system;
N LO : the number of photons contained in each local oscillation optical pulse;
hv: h is the Planck constant, v is the frequency of the photon, and the product of the two represents the energy of the photon;
represents the average initial velocity of the free carrier ensemble before photon absorption;
after element replacement, the method comprises the following steps:wherein n represents the refractive index, ε, of the germanium material of a silicon-based germanium detector 0 Represents the dielectric constant in vacuum, c represents the speed of light in vacuum, q e Representing the charge quantity of electrons beta IB Representing the absorption coefficient between the bands->Represents the initial quantum efficiency constant, beta s Representing scattering absorption systemA number;
module M3: construction of deep nerve modelPerforming initial training on the deep neural network by using the local dictionary set in the module M2, executing the module M4 when the deep neural network meets the judging condition capable of accurately predicting the local oscillation photon fluctuation percentage according to the light intensity of the input detector, and otherwise, continuing to expand the local dictionary set until the condition is met;
module M4: performing joint training on the deep neural network, completing the joint training stage when the deep neural network meets the judgment condition that the quantum efficiency can be accurately predicted according to the input local oscillator light intensity, otherwise, continuing training;
module M5: and carrying out high-precision real-time prediction on the quantum efficiency of the chip CVQKD system by using the trained deep neural network, and accurately evaluating the actual security key rate of the system according to the accurate real-time quantum efficiency prediction result.
6. The adaptive quantum efficiency high precision real time prediction system according to claim 5, wherein: the decision conditions in the module M3 are: satisfy epsilon for a given precision 1 And delta 1 For any meeting Is set to the input light intensity P of 0 If the mapping result of the deep neural network under the input light intensity meets the requirementAnd when the training is finished, the deep neural network completes initial training.
7. The adaptive quantum efficiency high precision real time prediction system according to claim 5, wherein: the decision conditions in the module M4 are: epsilon for a given precision 2 If the following conditions are satisfied:wherein the method comprises the steps ofRepresentative ofThe second component in the function values is considered to be the end of the deep neural network joint training, ++>Alice and Bob, which represent quantum efficiencies, have fixed constant values of quantum efficiencies, β fc Represents the free carrier absorption coefficient, lambda represents the wavelength of the optical signal, and E represents the strength of the electric field in which the free carrier is located.
8. The adaptive quantum efficiency high precision real time prediction method according to claim 6, wherein: in the module M5, the input light intensity is monitored according to beam splittingThe trained deep neural network is utilized to accurately predict the quantum efficiency in real time: />Wherein->
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111100710.8A CN113810186B (en) | 2021-09-18 | 2021-09-18 | High-precision real-time prediction method and system for self-adaptive quantum efficiency |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111100710.8A CN113810186B (en) | 2021-09-18 | 2021-09-18 | High-precision real-time prediction method and system for self-adaptive quantum efficiency |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113810186A CN113810186A (en) | 2021-12-17 |
| CN113810186B true CN113810186B (en) | 2023-11-07 |
Family
ID=78939764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111100710.8A Active CN113810186B (en) | 2021-09-18 | 2021-09-18 | High-precision real-time prediction method and system for self-adaptive quantum efficiency |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113810186B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114912608B (en) * | 2022-05-16 | 2023-04-07 | 南京邮电大学 | Global phase tracking prediction method suitable for double-field quantum key distribution system |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1825633A1 (en) * | 2004-12-15 | 2007-08-29 | Thales | Continuously-variable system for encryption key quantum distribution |
| EP2460069A1 (en) * | 2009-08-27 | 2012-06-06 | Universite Libre De Bruxelles | Network distributed quantum random number generation |
| CN108365953A (en) * | 2018-02-06 | 2018-08-03 | 中南大学 | Adaptive differential phase shift quantum key dissemination system based on deep neural network and its implementation |
| CN108428023A (en) * | 2018-05-24 | 2018-08-21 | 四川大学 | Trend forecasting method based on quantum Weighted Threshold repetitive unit neural network |
| CN108491185A (en) * | 2018-06-08 | 2018-09-04 | 中国科学技术大学 | The real-time quantum random number generator of high speed based on photoelectricity hybrid integrated |
| CN109033632A (en) * | 2018-07-26 | 2018-12-18 | 北京航空航天大学 | A kind of trend forecasting method based on depth quantum nerve network |
| CN110365473A (en) * | 2019-05-31 | 2019-10-22 | 南京邮电大学 | A kind of active feedback control method of the quantum communication system based on machine learning |
| CN110391903A (en) * | 2019-07-16 | 2019-10-29 | 上海循态信息科技有限公司 | Method, system and the medium of laser sowing attack are resisted in CVQKD system |
| WO2020140851A1 (en) * | 2018-12-30 | 2020-07-09 | 华南师范大学 | Quantum communication and quantum time-frequency transmission fusion network system and method |
| CN112929160A (en) * | 2021-01-22 | 2021-06-08 | 西安电子科技大学 | Plug-and-play reference system and measuring equipment independent quantum key distribution system and method |
| CN113055167A (en) * | 2021-03-22 | 2021-06-29 | 上海循态信息科技有限公司 | Defense method and system based on security vulnerability in chip CVQKD actual system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101833956B1 (en) * | 2017-05-19 | 2018-03-02 | 한국과학기술원 | System for phase compensation in continuous variable quantum key distribution |
| CA3088135A1 (en) * | 2018-01-31 | 2019-08-08 | Google Llc | Quantum computation through reinforcement learning |
| CN109361515B (en) * | 2018-11-23 | 2022-05-27 | 山西大学 | Pulsed light high-speed polarization locking method for continuous variable quantum key distribution system |
| US11776666B2 (en) * | 2019-03-07 | 2023-10-03 | Volkswagen Aktiengesellschaft | Simulating electronic structure with quantum annealing devices and artificial neural networks |
-
2021
- 2021-09-18 CN CN202111100710.8A patent/CN113810186B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1825633A1 (en) * | 2004-12-15 | 2007-08-29 | Thales | Continuously-variable system for encryption key quantum distribution |
| EP2460069A1 (en) * | 2009-08-27 | 2012-06-06 | Universite Libre De Bruxelles | Network distributed quantum random number generation |
| CN108365953A (en) * | 2018-02-06 | 2018-08-03 | 中南大学 | Adaptive differential phase shift quantum key dissemination system based on deep neural network and its implementation |
| CN108428023A (en) * | 2018-05-24 | 2018-08-21 | 四川大学 | Trend forecasting method based on quantum Weighted Threshold repetitive unit neural network |
| CN108491185A (en) * | 2018-06-08 | 2018-09-04 | 中国科学技术大学 | The real-time quantum random number generator of high speed based on photoelectricity hybrid integrated |
| CN109033632A (en) * | 2018-07-26 | 2018-12-18 | 北京航空航天大学 | A kind of trend forecasting method based on depth quantum nerve network |
| WO2020140851A1 (en) * | 2018-12-30 | 2020-07-09 | 华南师范大学 | Quantum communication and quantum time-frequency transmission fusion network system and method |
| CN110365473A (en) * | 2019-05-31 | 2019-10-22 | 南京邮电大学 | A kind of active feedback control method of the quantum communication system based on machine learning |
| CN110391903A (en) * | 2019-07-16 | 2019-10-29 | 上海循态信息科技有限公司 | Method, system and the medium of laser sowing attack are resisted in CVQKD system |
| CN112929160A (en) * | 2021-01-22 | 2021-06-08 | 西安电子科技大学 | Plug-and-play reference system and measuring equipment independent quantum key distribution system and method |
| CN113055167A (en) * | 2021-03-22 | 2021-06-29 | 上海循态信息科技有限公司 | Defense method and system based on security vulnerability in chip CVQKD actual system |
Non-Patent Citations (1)
| Title |
|---|
| 连续变量量子密钥分发实际安全性研究进展;黄鹏;曾贵华;;信息网络安全(11);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113810186A (en) | 2021-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Flamini et al. | Photonic quantum information processing: a review | |
| Jin et al. | Simple method of generating and distributing frequency-entangled qudits | |
| CN109361515B (en) | Pulsed light high-speed polarization locking method for continuous variable quantum key distribution system | |
| Kalita et al. | Solitons in magnetized plasma with electron inertia under weakly relativistic effect | |
| CN113810186B (en) | High-precision real-time prediction method and system for self-adaptive quantum efficiency | |
| Peng et al. | Satellite-to-submarine quantum communication based on measurement-device-independent continuous-variable quantum key distribution | |
| Karimi et al. | Near-inertial parametric subharmonic instability of internal wave beams | |
| Pan et al. | Secret key distillation over satellite-to-satellite free-space optics channel with a limited-sized aperture eavesdropper in the same plane of the legitimate receiver | |
| Zhao et al. | Monte Carlo-based security analysis for multi-mode continuous-variable quantum key distribution over underwater channel | |
| Zhong et al. | Predictive learning of multi-channel isochronal chaotic synchronization by utilizing parallel optical reservoir computers based on three laterally coupled semiconductor lasers with delay-time feedback | |
| Feng et al. | Modeling of a multi-parameter chaotic optoelectronic oscillator based on the Fourier neural operator | |
| Zhou et al. | Neural network-based prediction of the secret-key rate of quantum key distribution | |
| CN113055167B (en) | Defense method and system based on security vulnerability in chip CVQKD actual system | |
| Vyvlecka et al. | Robust excitation of C-band quantum dots for quantum communication | |
| Li et al. | On prediction of chaotic dynamics in semiconductor lasers by reservoir computing | |
| Bao et al. | Finite-key analysis of a practical decoy-state high-dimensional quantum key distribution | |
| Zhu et al. | Real-time selection for free-space measurement device independent quantum key distribution | |
| Wei et al. | High-precision data acquisition for free-space continuous-variable quantum key distribution | |
| CN113836524B (en) | Method and system for defending security vulnerabilities in chip CVQKD (continuously variable network QKD) actual system | |
| Gao et al. | Experimental collision-free dominant boson sampling | |
| Li et al. | Photonic reservoir computing enabled by silicon micro-rings | |
| Ma et al. | Chaotic dynamical enhanced optical physical layer encryption in OFDM-PON system based on echo state network | |
| Wang et al. | Existence and multiplicity of solutions for (p, q)\left (p, q\right)‐Laplacian Kirchhoff‐type fractional differential equations with impulses | |
| CN109901289B (en) | Design method of phase hologram for generating multi-mode superimposed vortex beam | |
| Wang et al. | Silicon photonic secure communication using artificial neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |