Continuous variable quantum key distribution parameter optimization method and system based on GA-BP neural network, medium and electronic equipment

Abstract

The invention provides a continuous variable quantum key distribution parameter optimization method and system based on a GA-BP neural network, a medium and electronic equipment. The method comprises the following steps: acquiring original data from a CV-QKD system, normalizing the original data, acquiring normalized data, and dividing the normalized data into a training set and a test set; establishing a BP neural network by using the training set; a genetic algorithm is adopted to optimize the weight and threshold of the BP neural network, and a GA-BP neural network with the optimal fitting effect is obtained; using the test set to test the GA-BP neural network, and obtaining the tested GA-BP neural network; according to the tested GA-BP neural network, obtaining a mapping relation between the target modulation variance and a corresponding target key rate, and finding out an optimal modulation variance and a corresponding maximum key rate by using a dichotomy based on the mapping relation; according to the invention, the complexity of calculating the key rate is reduced, and the key rate and the transmission distance in the CV-QKD system are improved.

Description

Technical Field

[0001] The present invention relates to the field of quantum communication, to artificial intelligence, and in particular to neural networks, and specifically to a continuous variable quantum key distribution parameter optimization method, system, medium and electronic equipment based on a GA-BP neural network. Background Art

[0002] Quantum key distribution (QKD), as a one-time key distribution method, has become one of the best solutions to ensure absolute security. QKD enables two remotely separated parties named Alice and Bob to extract a symmetric secret key string using a quantum channel. At present, according to the implementation method of quantum key distribution, it can be divided into discrete variable quantum key distribution (DVQKD) and continuous variable quantum key distribution (CVQKD). The CV-QKD system has the advantages of high key rate and convenient signal carrier preparation. Therefore, the CV-QKD technology is relatively mature and low-cost. It can be transmitted in ordinary optical fibers with classical optical communications, and has great practical advantages.

[0003] Key rate and transmission distance are important parameters to measure the quality of CV-QKD systems. In recent decades, the calculation of key rate is usually completed by numerical analysis methods. For example, for quantum light passing through the atmospheric channel, researchers mostly use numerical methods to analyze the impact of transmittance and excess noise on the key rate. Although numerical methods are widely used in practice to calculate the key rate of many quantum key distribution protocols, they consume a lot of computing resources and are too time-consuming. Therefore, it is particularly important to develop tools that are more efficient than numerical methods in calculating key rates. Summary of the invention

[0004] The purpose of the present invention is to provide a continuous variable quantum key distribution parameter optimization method, system, medium and electronic device based on GA-BP neural network, which are used to solve the problems pointed out in the above background technology.

[0005] In a first aspect, the present invention provides a method for optimizing parameters of continuous variable quantum key distribution based on a GA-BP neural network, the method comprising: obtaining original data from a CV-QKD system, normalizing the original data, obtaining normalized data, and dividing the normalized data into a training set and a test set; wherein the original data at least comprises: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least comprises: at least two normalized modulation variances and at least two normalized key rates; both the training set and the test set comprise at least one normalized modulation variance and the corresponding at least one normalized key rate; the normalized modulation variance in the training set is used as an input variable, and the training set is The normalized key rate in the test set is used as the output variable to establish a BP neural network; the weights and thresholds of the BP neural network are optimized by a genetic algorithm to obtain a GA-BP neural network with the best fitting effect; the normalized modulation variance in the test set is used as the input variable of the GA-BP neural network, and the normalized key rate in the test set is used as the output variable of the GA-BP neural network to test the GA-BP neural network and obtain a tested GA-BP neural network; a mapping relationship between a target modulation variance and a corresponding target key rate is obtained according to the tested GA-BP neural network, so as to find the optimal modulation variance and the corresponding maximum key rate based on the mapping relationship by using a dichotomy method to achieve continuous variable quantum key distribution parameter optimization.

[0006] In the present invention, a continuous variable quantum key distribution parameter optimization method based on GA-BP neural network is proposed to analyze the key rate of the CV-QKD system under the condition of finite code length, and the back propagation neural network (GA-BP) optimized based on genetic algorithm is applied to the CV-QKD system, and then the optimal modulation variance corresponding to the maximum key rate is found by using the dichotomy method; by utilizing machine learning technology to optimize the CV-QKD system parameters, the complexity of calculating the key rate in the prior art is reduced, and the key rate and transmission distance in the CV-QKD system are improved to a certain extent.

[0007] In an implementation manner of the first aspect, the modulation variance is a positive number.

[0008] In an implementation of the first aspect, the BP neural network is a three-layer BP neural network, and the BP neural network includes: an input layer, a hidden layer, and an output layer, wherein the calculation formula for the number of neurons in the hidden layer is:

[0009]

[0010] Among them, n 1represents the number of neurons in the hidden layer; n represents the number of neurons in the input layer; m represents the number of neurons in the output layer; a is a constant ranging from 1 to 10.

[0011] In an implementation of the first aspect, the use of a genetic algorithm to optimize the weights and thresholds of the BP neural network to obtain a GA-BP neural network with the best fitting effect includes: generating an initial population; the initial population includes a plurality of individuals; each of the individuals represents a parameter set of the BP neural network; each of the parameter sets includes a weight and a threshold; determining a termination condition of the genetic algorithm, initializing the genetic algorithm and calculating a fitness value of each of the individuals; judging whether the genetic algorithm has reached the termination condition based on the fitness value; and when judging that the genetic algorithm has not reached the termination condition, initializing the genetic algorithm; and The algorithm sequentially performs selection, crossover, and mutation operations to generate a new generation of chromosomes until it is determined that the genetic algorithm reaches the termination condition; when it is determined that the genetic algorithm reaches the termination condition, the calculation of the fitness value is terminated, the weights and thresholds optimized by the genetic algorithm are obtained, and the BP neural network is used to adjust the weights and thresholds optimized by the genetic algorithm to improve the accuracy of the weights and thresholds optimized by the genetic algorithm, until the accuracy meets the preset accuracy requirement, and the adjusted weights and thresholds are obtained to obtain the GA-BP neural network based on the adjusted weights and thresholds.

[0012] In an implementation of the first aspect, the GA-BP neural network uses a hyperbolic tangent function as an activation function.

[0013] In this implementation, the hyperbolic tangent function tanh is used as the activation function. Since tanh is a concave function in the positive number domain, the key rate obtained by the modulation variance passing through the GA-BP neural network is also a concave function of the modulation variance.

[0014] In an implementation of the first aspect, the mapping relationship is: the target key rate is a concave function about the target modulation variance; the method of using the dichotomy method based on the mapping relationship to find the optimal modulation variance and the corresponding maximum key rate includes: representing the mapping relationship with a relationship graph; the relationship graph is a schematic diagram of the concave function; wherein the target key rate is the vertical coordinate and the target modulation variance is the horizontal coordinate; on the relationship graph, starting from a preset initial range of the target modulation variance, the midpoint of the preset initial range is calculated, and the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint are calculated; the range of the target modulation variance is adjusted according to the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint, and repeated until the target modulation variance corresponding to the maximum target key rate is found; the maximum target key rate is the maximum key rate; the target modulation variance corresponding to the maximum target key rate is the optimal modulation variance.

[0015] In a second aspect, the present invention provides a continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network, the system comprising: a normalization module, used to obtain original data from a CV-QKD system, and normalize the original data to obtain normalized data, and divide the normalized data into a training set and a test set; wherein the original data at least comprises: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least comprises: at least two normalized modulation variances and at least two normalized key rates; the training set and the test set both comprise at least one normalized modulation variance and at least one corresponding normalized key rate; a network establishment module, used to take the normalized modulation variance in the training set as an input variable, and the normalized modulation variance in the training set as an input variable. The invention relates to a method for optimizing quantum key distribution parameters of a quantum computer. The method comprises the following steps: first, using a normalized modulation variance in the test set as an input variable of the GA-BP neural network, and second, using a normalized key rate in the test set as an output variable to establish a BP neural network; a network optimization module, which is used to optimize the weights and thresholds of the BP neural network by using a genetic algorithm to obtain a GA-BP neural network with the best fitting effect; a network testing module, which is used to use the normalized modulation variance in the test set as an input variable of the GA-BP neural network, and the normalized key rate in the test set as an output variable of the GA-BP neural network, so as to test the GA-BP neural network and obtain a tested GA-BP neural network; and a parameter optimization module, which is used to obtain a mapping relationship between a target modulation variance and a corresponding target key rate according to the tested GA-BP neural network, so as to find out an optimal modulation variance and a corresponding maximum key rate by using a dichotomy method based on the mapping relationship, so as to optimize the parameters of continuous variable quantum key distribution.

[0016] In a third aspect, the present invention provides an electronic device, comprising: a processor and a memory; the memory is used to store a computer program; the processor is used to execute the computer program stored in the memory, so that the electronic device executes the above-mentioned continuous variable quantum key distribution parameter optimization method based on the GA-BP neural network.

[0017] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by an electronic device, implements the above-mentioned continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network.

[0018] As described above, the continuous variable quantum key distribution parameter optimization method, system, medium and electronic device based on GA-BP neural network described in the present invention have the following beneficial effects:

[0019] (1) Compared with the prior art, the present invention optimizes the parameters of the CV-QKD system through a GA-BP-based method, wherein given the modulation variance, the key rate can be effectively obtained from the trained GA-BP neural network; secondly, since GA-BP uses the hyperbolic tangent function as the activation function and the key rate is derived from the GA-BP neural network, it is concave in the modulation variance. This theoretical result makes it possible to use the bisection method to effectively find the maximum key rate and the optimal modulation variance for a given GA-BP neural network.

[0020] (2) The continuous variable quantum key distribution parameter optimization scheme based on GA-BP neural network proposed in this invention can improve the key rate to a certain extent. The research on this scheme has far-reaching significance for the practical application of CV-QKD system, and further promotes the development of CV-QKD system in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Shown is a flow chart of a continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network according to an embodiment of the present invention.

[0022] Figure 2 It shows a flowchart of using a genetic algorithm to optimize the weights and thresholds of the BP neural network to obtain a GA-BP neural network with the best fitting effect as described in an embodiment of the present invention.

[0023] Figure 3 Shown is a schematic diagram of the structure of a continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0025] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention. The illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0026] For the CV-QKD system, the key rate of finite code length is calculated as follows:

[0027]

[0028] Where N is the total number of signals exchanged by Alice and Bob during the protocol; n is the number of segments of N signals used to establish the key; Δ(n) is related to the security of privacy amplification; FER is the frame error rate; the negotiation efficiency is measured by β = R / C(s), where R is the bit rate and C(s) is the channel capacity at the signal-to-noise ratio (SNR) value s; I AB represents the mutual information between Alice and Bob; the maximum value of the Holevo information leaked to Eve is denoted by χ BE Represents. The mutual information I AB and Holevo bound BE They are all related to the modulation variance, so by focusing on effectively optimizing the modulation variance and maximizing the key rate, we can ultimately achieve improved performance of CV-QKD.

[0029] With the simultaneous rise of artificial intelligence and quantum information science, the two fields are converging in a synergistic way. In this growing trend, several works attempt to design new theoretical models based on quantum algorithms to improve classical machine learning to achieve the desired quantum speedup. At the same time, with the increasing complexity of quantum systems, advanced quantum information technology also requires powerful data processing and data analysis tools. Therefore, we urgently need to use existing classical machine learning techniques to solve practical but difficult problems in quantum information science, such as tomography, quantum state classification, quantum metrology, quantum control, and quantum cryptography.

[0030] See also Figures 1 to 3. The following embodiments of the present invention provide a continuous variable quantum key distribution parameter optimization method, system, medium and electronic device based on a GA-BP neural network, wherein the parameters of the CV-QKD system are optimized by a GA-BP-based method, wherein given the modulation variance, the key rate can be effectively obtained from the trained GA-BP neural network; secondly, since GA-BP uses a hyperbolic tangent function as an activation function, and the key rate is derived from the GA-BP neural network, it is concave in the modulation variance, and this theoretical result enables the use of a dichotomy method to effectively find the maximum key rate and the optimal modulation variance of a given GA-BP neural network; the continuous variable quantum key distribution parameter optimization scheme based on a GA-BP neural network proposed in the present invention can improve the key rate to a certain extent, and the research on this scheme has far-reaching significance for the practical application of the CV-QKD system, and further promotes the development of the CV-QKD system in practical applications.

[0031] The technical solutions in the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0032] like Figure 1 As shown, in one embodiment, the present invention provides a continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network, the method comprising:

[0033] Step S1, obtaining original data from the CV-QKD system, and normalizing the original data to obtain normalized data, and dividing the normalized data into a training set and a test set.

[0034] Among them, the original data at least includes: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least includes: at least two normalized modulation variances and at least two normalized key rates; the training set and the test set both include at least one normalized modulation variance and the corresponding at least one normalized key rate.

[0035] In one embodiment, the modulation variance is a positive number.

[0036] Step S2: Using the normalized modulation variance in the training set as an input variable and the normalized key rate in the training set as an output variable to establish a BP neural network.

[0037] In one embodiment, the BP neural network is a three-layer BP neural network, and the BP neural network includes: an input layer, a hidden layer and an output layer, wherein the calculation formula for the number of neurons in the hidden layer is:

[0038]

[0039] Among them, n 1 represents the number of neurons in the hidden layer; n represents the number of neurons in the input layer; m represents the number of neurons in the output layer; a is a constant ranging from 1 to 10.

[0040] It should be noted that the specific value of a is not a condition to limit the present invention, as long as it is guaranteed to be between 1 and 10. In practical applications, it is limited by the complexity of the task and the risk of overfitting.

[0041] In this embodiment, a method for calculating neuron data of a hidden layer based on an input layer and an input layer scale is provided.

[0042] It should be noted that the BP (Back Propagation) neural network is a concept proposed by scientists led by Rumelhart and McClelland in 1986. It is a multi-layer feedforward neural network trained according to the error back propagation algorithm.

[0043] Step S3: using a genetic algorithm to optimize the weights and thresholds of the BP neural network to obtain a GA-BP neural network with the best fitting effect.

[0044] It should be noted that the Genetic Algorithm (GA) was first proposed by John Holland of the United States in the 1970s. The algorithm was designed and proposed based on the laws of evolution of organisms in nature. It is a computational model of biological evolution that simulates the natural selection and genetic mechanism of Darwin's theory of biological evolution. It is a method of searching for the optimal solution by simulating the natural evolution process. The algorithm uses mathematical methods and computer simulation operations to convert the problem-solving process into processes such as crossover and mutation of chromosome genes in biological evolution. When solving more complex combinatorial optimization problems, it can usually obtain better optimization results faster than some conventional optimization algorithms. Genetic algorithms have been widely used in combinatorial optimization, machine learning, signal processing, adaptive control, artificial life and other fields.

[0045] like Figure 2 As shown, in one embodiment, the weights and thresholds of the BP neural network are optimized by using a genetic algorithm to obtain a GA-BP neural network with the best fitting effect, including:

[0046] Step S301: Generate an initial population.

[0047] In this embodiment, the initial population includes a plurality of individuals; each of the individuals represents a parameter set of the BP neural network; each of the parameter sets includes a weight and a threshold.

[0048] Step S302: determine the termination condition of the genetic algorithm, initialize the genetic algorithm and calculate the fitness value of each individual.

[0049] It should be noted that the fitness value indicates how well an individual performs in a problem, and its calculation will directly affect the selection operation of the genetic algorithm, thereby determining which individuals will be retained and generate the next generation. In this embodiment, the fitness value is calculated using conventional technical means in the field, so its working principle will not be described in detail here.

[0050] Step S303: judging whether the genetic algorithm has reached the termination condition based on the fitness value.

[0051] When it is determined that the genetic algorithm has not reached the termination condition, step S304 is executed.

[0052] Step S304: performing selection, crossover, and mutation operations on the genetic algorithm in sequence to generate a new generation of chromosomes, until it is determined that the genetic algorithm reaches the termination condition.

[0053] When it is determined that the genetic algorithm reaches the termination condition, step S305 is executed.

[0054] Step S305, end the calculation of the fitness value, obtain the weights and thresholds optimized by the genetic algorithm, and use the BP neural network to adjust the weights and thresholds optimized by the genetic algorithm to improve the accuracy of the weights and thresholds optimized by the genetic algorithm, until the accuracy meets the preset accuracy requirement, obtain the adjusted weights and thresholds, and obtain the GA-BP neural network based on the adjusted weights and thresholds.

[0055] It should be noted that the specific setting of the preset accuracy requirement is not a condition to limit the present invention, and in practical applications, it can be set according to specific application scenarios.

[0056] Step S4: using the normalized modulation variance in the test set as the input variable of the GA-BP neural network, and using the normalized key rate in the test set as the output variable of the GA-BP neural network, so as to test the GA-BP neural network and obtain a tested GA-BP neural network.

[0057] It should be noted that in step S4, the test of the GA-BP neural network adopts conventional technical means in the field of neural networks, so its working principle will not be described in detail here; through the test, the performance of the GA-BP neural network can be evaluated and the generalization ability of the GA-BP neural network on unseen data can be verified.

[0058] Specifically, by taking the normalized modulation variance in the training set as the input variable and the normalized key rate in the training set as the output variable, a BP neural network is established, and then the weights and thresholds of the BP neural network are optimized by a genetic algorithm to obtain a GA-BP neural network; finally, by taking the normalized modulation variance in the test set as the input variable of the GA-BP neural network and the normalized key rate in the test set as the output variable of the GA-BP neural network, the GA-BP neural network is tested to train the GA-BP neural network to learn the mapping relationship between the target modulation variance and the target key rate.

[0059] In one embodiment, the GA-BP neural network uses a hyperbolic tangent function as an activation function tanh.

[0060] It should be noted that, since tanh is a concave function in the positive number domain, the key rate obtained by the modulation variance passing through the GA-BP neural network is also a concave function of the modulation variance.

[0061] In one embodiment, the mapping relationship is: the target key rate is a concave function of the target modulation variance.

[0062] The present invention optimizes the parameters of the CV-QKD system by a GA-BP-based method, where the key rate can be effectively obtained from the trained GA-BP neural network given the modulation variance. Secondly, since GA-BP uses the hyperbolic tangent function as the activation function and the key rate is derived from the GA-BP neural network, it is concave in the modulation variance. This theoretical result enables us to use the bisection method to effectively find the maximum key rate and the optimal modulation variance for a given GA-BP neural network.

[0063] Step S5: obtaining a mapping relationship between a target modulation variance and a corresponding target key rate according to the tested GA-BP neural network, and finding an optimal modulation variance and a corresponding maximum key rate based on the mapping relationship using a dichotomy method to achieve continuous variable quantum key distribution parameter optimization.

[0064] In one embodiment, the use of the binary search method based on the mapping relationship to find the optimal modulation variance and the corresponding maximum key rate includes: representing the mapping relationship with a relationship graph; the relationship graph is a schematic diagram of the concave function; wherein the target key rate is the ordinate and the target modulation variance is the abscissa; on the relationship graph, starting from a preset initial range of the target modulation variance, calculating the midpoint of the preset initial range, and calculating the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint; adjusting the range of the target modulation variance according to the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint, and repeating until the target modulation variance corresponding to the maximum target key rate is found; the maximum target key rate is the maximum key rate; the target modulation variance corresponding to the maximum target key rate is the optimal modulation variance.

[0065] It should be noted that the preset initial range is a fixed range that is set in advance. The specific setting is not a condition to limit the present invention. In practical applications, it can be set according to specific application scenarios. For example, in one embodiment, the preset initial range is set to 0 to 20.

[0066] In this embodiment, for the calculation of the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint, reference may be made to the above-mentioned key rate calculation formula, so it will not be described in detail here.

[0067] It should be noted that the present invention proposes a continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network to analyze the key rate of a CV-QKD system under finite code length conditions. When optimizing the parameters, the error correction code and code rate are given, the lower bound of the frame error rate is used as the FER value in the key rate, and the back propagation neural network (GA-BP) optimized based on a genetic algorithm is applied to the CV-QKD system, and then the dichotomy method is used to find the optimal modulation variance corresponding to the maximum key rate. By utilizing machine learning technology to optimize the parameters of the CV-QKD system, the complexity of calculating the key rate in the prior art is reduced, and the key rate and transmission distance in the CV-QKD system are improved to a certain extent.

[0068] The protection scope of the continuous variable quantum key distribution parameter optimization method based on GA-BP neural network described in the embodiment of the present invention is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, reducing or replacing steps in the prior art based on the principles of the present invention are included in the protection scope of the present invention.

[0069] An embodiment of the present invention also provides an electronic device, comprising: a processor and a memory; the memory is used to store a computer program; the processor is used to execute the computer program stored in the memory, so that the electronic device executes the above-mentioned continuous variable quantum key distribution parameter optimization method based on the GA-BP neural network.

[0070] An embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by an electronic device, the above-mentioned continuous variable quantum key distribution parameter optimization method based on GA-BP neural network is implemented.

[0071] A person of ordinary skill in the art can understand that all or part of the steps in the method for implementing the above-mentioned embodiment can be completed by instructing a processor through a program, and the program can be stored in a computer-readable storage medium, and the storage medium is a non-transitory medium, such as a random access memory, a read-only memory, a flash memory, a hard disk, a solid-state hard disk, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The above-mentioned storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a digital video disc (DVD)), or a semiconductor medium (for example, a solid-state disk (SSD)), etc.

[0072] The embodiment of the present invention further provides a continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network. The continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network can implement the continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network described in the present invention. However, the implementation device of the continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network described in the present invention includes but is not limited to the structure of the continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network listed in this embodiment. All structural deformations and replacements of the prior art made according to the principles of the present invention are included in the protection scope of the present invention.

[0073] like Figure 3 As shown, in one embodiment, the present invention provides a continuous variable quantum key distribution parameter optimization system based on a GA-BP neural network, the system comprising:

[0074] The normalization module 31 is used to obtain original data from the CV-QKD system, normalize the original data, obtain normalized data, and divide the normalized data into a training set and a test set; wherein the original data at least includes: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least includes: at least two normalized modulation variances and at least two normalized key rates; both the training set and the test set include at least one normalized modulation variance and the corresponding at least one normalized key rate.

[0075] The network establishment module 32 is used to establish a BP neural network by taking the normalized modulation variance in the training set as an input variable and the normalized key rate in the training set as an output variable.

[0076] The network optimization module 33 is used to optimize the weights and thresholds of the BP neural network using a genetic algorithm to obtain a GA-BP neural network with the best fitting effect.

[0077] The network testing module 34 is used to use the normalized modulation variance in the test set as the input variable of the GA-BP neural network, and the normalized key rate in the test set as the output variable of the GA-BP neural network, so as to test the GA-BP neural network and obtain the tested GA-BP neural network.

[0078] The parameter optimization module 35 is used to obtain the mapping relationship between the target modulation variance and the corresponding target key rate according to the tested GA-BP neural network, so as to find the optimal modulation variance and the corresponding maximum key rate based on the mapping relationship using the dichotomy method to achieve continuous variable quantum key distribution parameter optimization.

[0079] It should be noted that the structures and principles of the normalization module 31, the network establishment module 32, the network optimization module 33, the network test module 34 and the parameter optimization module 35 correspond one to one with the steps (steps S1 to S5) in the above-mentioned continuous variable quantum key distribution parameter optimization method based on GA-BP neural network. The specific working principle can also be referred to the introduction of the continuous variable quantum key distribution parameter optimization method based on GA-BP neural network in the aforementioned embodiment, so it will not be repeated here.

[0080] In the several embodiments provided by the present invention, it should be understood that the disclosed system, device or method can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules / units is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules or units, which can be electrical, mechanical or other forms.

[0081] The modules / units described as separate components may or may not be physically separated, and the components displayed as modules / units may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules / units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention. For example, the functional modules / units in the various embodiments of the present invention may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0082] Those of ordinary skill in the art should further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0083] The descriptions of the processes or structures corresponding to the above-mentioned figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0084] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A continuous variable quantum key distribution parameter optimization method based on GA-BP neural network, characterized in that: The method comprises: Obtaining raw data from a CV-QKD system, normalizing the raw data to obtain normalized data, and dividing the normalized data into a training set and a test set; wherein the raw data at least includes: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least includes: at least two normalized modulation variances and at least two normalized key rates; both the training set and the test set include at least one normalized modulation variance and the corresponding at least one normalized key rate; Using the normalized modulation variance in the training set as an input variable and the normalized key rate in the training set as an output variable, a BP neural network is established; Using genetic algorithm to optimize the weights and thresholds of the BP neural network to obtain a GA-BP neural network with the best fitting effect; Using the normalized modulation variance in the test set as an input variable of the GA-BP neural network, and the normalized key rate in the test set as an output variable of the GA-BP neural network, so as to test the GA-BP neural network and obtain a tested GA-BP neural network; The mapping relationship between the target modulation variance and the corresponding target key rate is obtained according to the tested GA-BP neural network, so as to find the optimal modulation variance and the corresponding maximum key rate based on the mapping relationship using the dichotomy method to achieve continuous variable quantum key distribution parameter optimization.

2. The continuous variable quantum key distribution parameter optimization method based on GA-BP neural network according to claim 1 is characterized in that: The modulation variance is a positive number.

3. The continuous variable quantum key distribution parameter optimization method based on GA-BP neural network according to claim 1 is characterized in that: The BP neural network is a three-layer BP neural network, comprising: an input layer, a hidden layer and an output layer, wherein the calculation formula for the number of neurons in the hidden layer is: Among them, n1 represents the number of neurons in the hidden layer; n represents the number of neurons in the input layer; m represents the number of neurons in the output layer; and a is a constant ranging from 1 to 10.

4. The continuous variable quantum key distribution parameter optimization method based on GA-BP neural network according to claim 1 is characterized in that: The method of using a genetic algorithm to optimize the weights and thresholds of the BP neural network to obtain a GA-BP neural network with the best fitting effect includes: Generate an initial population; the initial population includes a plurality of individuals; each of the individuals represents a parameter set of the BP neural network; each of the parameter sets includes a weight and a threshold; Determining a termination condition of the genetic algorithm, initializing the genetic algorithm and calculating a fitness value of each individual; Determining whether the genetic algorithm reaches the termination condition based on the fitness value; When it is determined that the genetic algorithm has not reached the termination condition, sequentially performing selection, crossover, and mutation operations on the genetic algorithm to generate a new generation of chromosomes, until it is determined that the genetic algorithm has reached the termination condition; When it is determined that the genetic algorithm reaches the termination condition, the calculation of the fitness value is terminated, the weights and thresholds optimized by the genetic algorithm are obtained, and the weights and thresholds optimized by the genetic algorithm are adjusted using the BP neural network to improve the accuracy of the weights and thresholds optimized by the genetic algorithm until the accuracy meets the preset accuracy requirements, and the adjusted weights and thresholds are obtained to obtain the GA-BP neural network based on the adjusted weights and thresholds.

5. The continuous variable quantum key distribution parameter optimization method based on GA-BP neural network according to claim 1 is characterized in that: The GA-BP neural network adopts the hyperbolic tangent function as the activation function.

6. The continuous variable quantum key distribution parameter optimization method based on GA-BP neural network according to claim 1 is characterized in that: The mapping relationship is: the target key rate is a concave function of the target modulation variance; the method of finding the optimal modulation variance and the corresponding maximum key rate based on the mapping relationship by using the dichotomy method includes: The mapping relationship is represented by a relationship diagram; the relationship diagram is a schematic diagram of the concave function; wherein the target key rate is the ordinate and the target modulation variance is the abscissa; On the relationship graph, starting from a preset initial range of the target modulation variance, calculating the midpoint of the preset initial range, and calculating the target key rate corresponding to the midpoint and the target key rates of adjacent points on both sides of the midpoint; The range of the target modulation variance is adjusted according to the target key rate corresponding to the midpoint and the target key rates of the adjacent points on both sides of the midpoint, and repeated until the target modulation variance corresponding to the maximum target key rate is found; the maximum target key rate is the maximum key rate; the target modulation variance corresponding to the maximum target key rate is the optimal modulation variance.

7. A continuous variable quantum key distribution parameter optimization system based on GA-BP neural network, characterized in that: The system comprises: A normalization module, used for obtaining raw data from a CV-QKD system, normalizing the raw data, obtaining normalized data, and dividing the normalized data into a training set and a test set; wherein the raw data at least includes: at least two modulation variances and at least two key rates; one modulation variance corresponds to one key rate; the normalized data at least includes: at least two normalized modulation variances and at least two normalized key rates; both the training set and the test set include at least one normalized modulation variance and the corresponding at least one normalized key rate; A network establishment module, used to establish a BP neural network by taking the normalized modulation variance in the training set as an input variable and the normalized key rate in the training set as an output variable; A network optimization module, used to optimize the weights and thresholds of the BP neural network using a genetic algorithm to obtain a GA-BP neural network with the best fitting effect; A network testing module, used to use the normalized modulation variance in the test set as an input variable of the GA-BP neural network, and the normalized key rate in the test set as an output variable of the GA-BP neural network, so as to test the GA-BP neural network and obtain a tested GA-BP neural network; A parameter optimization module is used to obtain a mapping relationship between a target modulation variance and a corresponding target key rate according to the tested GA-BP neural network, so as to find an optimal modulation variance and a corresponding maximum key rate based on the mapping relationship using a dichotomy method, thereby realizing continuous variable quantum key distribution parameter optimization.

8. An electronic device, characterized in that: The electronic device comprises: a processor and a memory; The memory is used to store computer programs; The processor is used to execute the computer program stored in the memory so that the electronic device executes the continuous variable quantum key distribution parameter optimization method based on the GA-BP neural network according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by an electronic device, the continuous variable quantum key distribution parameter optimization method based on a GA-BP neural network as described in any one of claims 1 to 6 is implemented.

Images