Quantum and quantum key fusion secure transmission method driven by quantum computing model

Abstract

The invention discloses a quantum computing model-driven quantum and quantum key fusion secure transmission method, which is applied to the technical field of data processing, and takes a quantum computing model as a core to construct a quantum and quantum key fusion secure transmission technical system. Firstly, quantum signal multi-dimensional features and key core data are collected, and noise reduction and efficiency improvement are performed by means of a quantum deep learning neural network; and a four-dimensional correlation model is established by using a quantum convolutional neural network and the like, and early warning and key optimization requirements are generated in combination with migration and meta-learning. Then, calculating an evolutionary trajectory by variable component sub-calculation and the like, solving a safe optimal solution, and obtaining a self-adaptive key distribution scheme; based on a quantum reinforcement learning structure decision agent, generating a control scheme giving consideration to safety and efficiency; according to the method, federated learning and secret sharing technologies are fused, global collaborative strategy adaptation features are obtained, a quantum probability graph model and other design evaluation systems are adopted, and security level classification and multi-dimensional comprehensive evaluation are achieved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a quantum computing model-driven secure transmission method for quantum and quantum key fusion. Background Technology

[0002] With the rapid development of the digital economy, the demand for secure and efficient data transmission is becoming increasingly urgent, especially for highly sensitive data in fields such as finance, government affairs, and healthcare, which places stringent requirements on the level of security protection. During quantum signal transmission and key generation, interference from factors such as channel noise and quantum state decoherence can lead to a large amount of high-dimensional redundant data. Current technologies lack efficient quantum data processing mechanisms and fail to deeply integrate quantum deep learning neural networks with feature extraction, relying solely on traditional filtering or simple quantum transformations. This makes it difficult to simultaneously optimize noise reduction, redundancy removal, and key information enhancement. Consequently, quantum data exhibits low signal-to-noise ratios and poor core feature recognition, increasing the complexity of subsequent model construction and affecting the stability of key generation and the reliability of transmission security due to insufficient data quality.

[0003] Existing models rely heavily on post-event log analysis for monitoring anomalies such as quantum channel interference and key leakage, lacking real-time and proactive early warning mechanisms. Furthermore, maintenance strategies and key distribution schemes are largely based on fixed rules, failing to utilize quantum computing models to dynamically simulate the evolution of transmission security, nor incorporating techniques like quantum reinforcement learning for real-time parameter adjustments. This results in insufficient targeted and timely strategy optimization when facing dynamically changing transmission environments (such as sudden changes in channel noise and fluctuations in node resources), making it difficult to quickly adapt to changing scenarios and potentially leading to security risks.

[0004] In distributed transmission scenarios, factors such as heterogeneity of node resources and stability of cross-node communication exacerbate the difficulty of global management. Existing technologies have failed to effectively integrate the advantages of quantum federated learning and quantum secret sharing technologies, making it impossible to achieve efficient collaborative scheduling of distributed nodes or to achieve global policy uniformity while ensuring data privacy. This results in low resource utilization and high communication latency during cross-node transmission, as well as security risks such as key share leakage and data tampering, failing to meet the requirements for large-scale application of distributed quantum communication networks.

[0005] Current evaluations of quantum and quantum key distribution transmission often focus on single metrics, such as key distribution rate and bit error rate, lacking a multi-dimensional quantitative evaluation system. They fail to cover core security metrics like security confidence and attack resistance, as well as key practical application metrics such as resource utilization efficiency, model adaptability, and global collaborative management fit. This results in an inability to comprehensively reflect the overall performance of the transmission system, making it difficult to accurately pinpoint technical shortcomings and thus limiting the iterative optimization of the entire transmission system. Summary of the Invention

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A quantum computing model-driven secure transmission method for quantum and quantum key fusion includes: acquiring multi-dimensional feature parameters of quantum signals and core data for quantum key generation; introducing a quantum deep learning neural network feature extraction module; and using quantum activation functions and quantum pooling layers to achieve noise reduction, redundancy removal, and key information enhancement of high-dimensional quantum data; processing the optimized data based on a quantum convolutional neural network; fusing quantum entanglement distillation technology and a deep quantum error correction coding algorithm driven by quantum attention mechanism to construct a four-dimensional correlation model of quantum signal-key security-computational resources-transmission delay; transferring parameters of the cross-scenario secure transmission model through quantum transfer learning; combining the rapid adaptation capability of the quantum meta-learning optimization model to generate dynamic security warning information and key hierarchical optimization requirements; and using a variable quantum computing model to simulate the evolution trajectory of transmission security in multiple scenarios, combined with a dynamic security model throughout the entire lifecycle of the quantum key. This paper proposes a quantum key distribution scheme. It employs a quantum approximation optimization algorithm coupled with a quantum simulated annealing algorithm to solve for the optimal security solution under qubit resource constraints and transmission delay thresholds. A quantum reinforcement learning algorithm is used to construct an intelligent decision-making agent, embedding a quantum reward function to adjust the quantum error correction code rate, key update frequency, and quantum signal modulation strategy in real time, generating a deep fusion control scheme that balances security and efficiency. A global collaborative management strategy is constructed by integrating a quantum federated learning mechanism and quantum secret sharing technology, generating quantum global collaborative management strategy adaptation and fusion features. Based on a quantum probabilistic graphical model coupled with a quantum Bayesian inference algorithm, a quantitative evaluation system is designed, incorporating quantum global collaborative management strategy adaptation and fusion features. Combined with a quantum support vector machine, transmission security level classification is implemented, generating comprehensive evaluation information including security confidence, anti-attack strength, resource utilization efficiency, model fit, and global collaborative control fit.

[0007] Its beneficial effects are as follows: This invention provides a quantum computing model-driven secure transmission method for quantum and quantum key fusion, using a quantum computing model as the core driver to construct a full-process secure transmission system for quantum and quantum key fusion. First, quantum signals and key core data are collected, and noise reduction and key feature enhancement are achieved through a quantum deep learning neural network. Then, a four-dimensional correlation model is constructed using a quantum convolutional neural network that integrates multiple technologies, combining transfer and meta-learning to generate early warning and key optimization requirements. Evolutionary trajectories are simulated through variable quantum computing, and coupled optimization algorithms are used to solve for the optimal security solution, forming an adaptive key distribution scheme. A decision agent is constructed using quantum reinforcement learning to dynamically adjust core parameters and generate a control scheme that balances security and efficiency. Quantum federated learning and secret sharing technologies are integrated to construct a global collaborative management strategy. Finally, a multi-dimensional quantitative evaluation system is designed based on a quantum probabilistic graphical model and Bayesian inference to achieve classification and comprehensive evaluation of transmission security levels.

[0008] Data processing efficiency has been significantly improved, with the quantum data signal-to-noise ratio increasing from 15dB to 32dB and the proportion of redundant information decreasing to 8%, providing high-quality data support for subsequent modeling. Security protection capabilities have been greatly enhanced, with attack resistance strength increased to level 9, and the key leakage risk warning response time ≤10ms, enabling real-time and proactive monitoring of abnormal states. Attached Figure Description

[0009] Figure 1 A flowchart of a quantum computing model-driven secure transmission method for quantum and quantum key fusion is provided in this embodiment of the invention. Figure 2 This is a schematic diagram of a quantum computing model-driven quantum and quantum key fusion secure transmission device provided in an embodiment of the present invention. Detailed Implementation

[0010] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Figure 1 This application describes a quantum computing model-driven secure transmission method for quantum and quantum key fusion, based on an exemplary embodiment of the present application.

[0011] In this application embodiment, a quantum computing model-driven secure transmission method for quantum and quantum key fusion is described, such as... Figure 1 As shown: S101 collects multidimensional feature parameters of quantum signals and core data for quantum key generation. It introduces a quantum deep learning neural network feature extraction module and uses quantum activation functions and quantum pooling layers to achieve noise reduction, redundancy removal, and key information enhancement of high-dimensional quantum data.

[0012] In one implementation, the entire process of quantum signal transmission and quantum key generation is used as the acquisition boundary. Multidimensional characteristic parameters of quantum signals and core data of quantum key generation are systematically collected to ensure coverage of key stages such as transmission links, key generation, and storage interaction. Core acquisition indicators and examples are explained below: Acquiring a single quantum bit | 0 State, |1 Preparing precision data for states and multi-qubit entangled states (such as Bell states), for example, measured by quantum state tomography, single-qubit |0 The fidelity of the prepared state was 99.2%, and the fidelity of the prepared two-qubit Bell state was 98.7%. Phase fluctuation data of fiber optic quantum channels and free-space quantum channels at different transmission distances were collected. For example, the phase noise figure was 0.02 rad / ns at a transmission distance of 10 km and 0.08 rad / ns at a transmission distance of 50 km.

[0013] Error data from the key negotiation phase of mainstream quantum key distribution protocols such as BB84 and E91 is collected. For example, the average error rate of the BB84 protocol is 1.8% in 1000 key negotiations, with a set error rate threshold of 3%. Entanglement degree data of entangled states (such as two-particle entangled states and multi-particle GHZ states) used for encoding during quantum key generation is also collected. For example, the von Neumann entanglement of two-particle entangled states... The von Neumann entanglement entropy is 1.0 (ideal maximum entanglement state), while the entanglement entropy measured in actual transmission is 0.92. Data on the coherent retention time of single photons and entangled photon pairs used for quantum signal transmission were collected. For example, the coherence time of a single photon at 1550 nm wavelength is 20 ns at room temperature, and extends to 50 ns at low temperature (-20℃). Data on the writing and reading accuracy of quantum states in quantum random access memory and solid-state quantum memory were collected. For example, the writing accuracy of a single quantum bit in a solid-state quantum memory is... After writing the entangled state of two qubits, the readout fidelity is 97.5%; after writing the entangled state of two qubits, the readout fidelity is 96.3%.

[0014] A quantum deep learning neural network feature extraction module is introduced to perform specialized optimization processing on the collected high-dimensional, high-noise quantum data. Through the synergistic effect of quantum activation functions and quantum pooling layers, data noise reduction, redundancy removal, and key information enhancement are achieved. A specific example is illustrated below: Quantum step activation function and quantum sigmoid activation function are used to perform nonlinear transformations on the original quantum data to enhance the discriminative power of effective features. For example, quantum sigmoid activation is applied to the original quantum channel phase noise coefficient data (including random noise interference) to compress abnormal fluctuations caused by noise (such as sudden bursts of 0.5 rad / ns of instantaneous noise) to a reasonable range while preserving the true phase change trend. A 2×2 quantum pooling window is designed to perform feature aggregation and dimensionality compression on the activated quantum data, eliminating redundant information. For example, for the 6-dimensional original feature vector composed of 6 core data types such as quantum bit state preparation fidelity and quantum entanglement entropy, quantum maximum pooling operation is used to select key feature points for each data type, compressing the 6-dimensional vector into a 3-dimensional core feature vector while preserving the core change patterns of each indicator.

[0015] After processing by the quantum deep learning neural network feature extraction module, the signal-to-noise ratio of quantum data increased from the original 15dB to 32dB, the proportion of redundant information decreased from 28% to 8%, and the quantum signal feature recognition and the stability of the core data for key generation were significantly improved, providing high-quality data support for subsequent model construction and analysis.

[0016] S102, based on the quantum convolutional neural network to process the above-mentioned optimized data, integrates quantum entanglement distillation technology and quantum attention mechanism-driven deep quantum error correction coding algorithm, constructs a four-dimensional correlation model of quantum signal-key security-computation resources-transmission delay, transfers cross-scenario quantum secure transmission model parameters through quantum transfer learning, and combines the rapid adaptation capability of quantum meta-learning optimization model to generate dynamic security early warning information and key hierarchical optimization requirements.

[0017] In one implementation, a correlation analysis is performed on the core technical parameters related to quantum-optimized data processing and model adaptation parameters to generate dynamic correlation analysis results between quantum data processing and model adaptation. The core technical parameters related to quantum-optimized data processing include quantum convolutional neural network processing parameters, quantum entanglement distillation technology parameters, quantum attention mechanism parameters, and deep quantum error correction coding algorithm parameters. The model adaptation parameters include quantum transfer learning cross-scenario parameters and quantum meta-learning optimization adaptation parameters. First, the specific scope of the two types of parameters is clarified, and then a correlation analysis is conducted using the Pearson correlation coefficient method and grey relational analysis to generate dynamic correlation analysis results.

[0018] Quantum convolutional neural network processing parameters include kernel size (e.g., 4×4 qubit matrix), number of quantum convolutional layers (e.g., 3 layers), and quantum activation function type (e.g., quantum ReLU activation function); quantum entanglement distillation technique parameters include the number of distillation iterations (e.g., 5 times) and entanglement state selection threshold (e.g., entanglement entropy ≥ 0.9); quantum attention mechanism parameters include attention weight allocation coefficients (e.g., quantum signal feature weight 0.6, key security feature weight 0.4) and attention window size (e.g., covering 6 consecutive data frames); deep quantum error correction coding algorithm parameters include error correction code length (e.g., 64 bits), coding rate (e.g., 1 / 2), and fault tolerance threshold (e.g., bit error rate ≤ 2%).

[0019] Quantum transfer learning cross-scene parameters include the feature mapping coefficients between the source and target scenes (e.g., the mapping coefficient from fiber channel to free space channel is 0.85) and the pre-trained model parameter transfer ratio (e.g., 80% core parameter transfer); quantum meta-learning optimization adaptation parameters include the meta-learning iteration steps (e.g., 100 steps) and the scene adaptation learning rate (e.g., 0.01).

[0020] Taking "quantum convolutional neural network kernel size" and "quantum transfer learning feature mapping coefficient" as examples, the correlation coefficient calculated by the Pearson correlation coefficient method is 0.78, indicating a strong positive correlation between the two; the grey correlation coefficient between "quantum entanglement distillation iteration count" and "quantum meta-learning iteration steps" is 0.65, which is a moderate positive correlation; the correlation coefficient between "deep quantum error correction coding rate" and "quantum transfer learning parameter transfer ratio" is 0.32, which is a weak correlation. Finally, a dynamic correlation analysis result table containing the pairwise correlation coefficients of all parameters is formed.

[0021] The correlation analysis results are filtered to generate a target simulation variable list. This list includes core technical parameters for quantum data processing, quantum model adaptation and optimization parameters, four-dimensional correlation model construction parameters, dynamic security early warning generation parameters, and key hierarchical optimization requirement parameters. A correlation filtering threshold is set (e.g., correlation ≥ 0.5), eliminating weakly correlated or uncorrelated parameter combinations and invalid variables, and retaining variables that have a significant impact on subsequent model construction, early warning generation, and key optimization. An example is provided below: Variables corresponding to parameter combinations with correlation ≥ 0.5 are retained, and core variables required for four-dimensional correlation model construction, early warning generation, and key optimization are added. Redundant variables with correlation < 0.5 are eliminated (e.g., "quantum attention mechanism window size" has a correlation of 0.35 with other core parameters and is therefore eliminated).

[0022] The core technical parameters for quantum data processing include quantum convolution kernel size, entanglement distillation iterations, attention weight allocation coefficients, and error correction coding rate. The parameters for quantum model adaptation and optimization include feature mapping coefficients and meta-learning iteration steps. The parameters for constructing the four-dimensional correlation model include the quantum signal-key security correlation coefficient, computational resource occupancy threshold (e.g., quantum bit utilization ≤ 80%), and transmission delay limit (e.g., ≤ 50ms). The parameters for generating dynamic security warnings include the warning trigger threshold (e.g., key leakage risk value ≥ 0.7) and warning response time (e.g., ≤ 10ms). The parameters for key hierarchical optimization requirements include the high-security-level key length threshold (e.g., 256 bits) and the ordinary-security-level key update cycle (e.g., 1 hour / time).

[0023] The correlation analysis results and the target simulation variable list are processed to generate parameter combination constraints, including synergistic adaptation constraints for quantum technology integration, compatibility constraints for cross-scenario parameter transfer, timeliness matching constraints for rapid model adaptation, and logical consistency constraints for four-dimensional correlation model construction. Combining the characteristics of quantum technology, cross-scenario adaptation requirements, and model construction logic, four types of constraints are formulated based on the correlation analysis results and the target simulation variable list. Examples are illustrated below: The kernel size of the quantum convolutional neural network must be adapted to the code length of the deep quantum error correction code, such as a 4×4 quantum convolution kernel corresponding to a 64-bit error correction code length, to ensure the synergy between data processing and error correction coding. The entanglement state screening threshold of quantum entanglement distillation must be higher than the fault tolerance threshold of deep quantum error correction coding, such as an entanglement entropy ≥ 0.9 corresponding to a bit error rate ≤ 2%, to avoid the high risk of bit error in the distilled entangled state.

[0024] In quantum transfer learning, the feature mapping coefficient between the source and target scenarios must be between 0.7 and 0.95. For example, the mapping coefficient from fiber channel to free space channel must meet this range to ensure that the transfer parameters are effectively adapted in the target scenario. The transfer ratio of pre-trained model parameters must not be less than 70%, and the transfer parameters must be compatible with the quantum hardware resources of the target scenario (e.g., the number of quantum gates corresponding to the transfer parameters does not exceed the maximum support of the target hardware).

[0025] The number of iterations in quantum meta-learning needs to be controlled between 50 and 200, with each iteration taking ≤1ms, to ensure that the model can complete cross-scenario adaptation within 100-200ms; the total time for parameter transfer and adaptation in quantum transfer learning is ≤500ms, meeting the requirements for rapid model adjustment in real-time transmission scenarios.

[0026] The quantum signal-key security correlation coefficient and the computational resource-transmission delay correlation coefficient must maintain logical consistency. For example, when the quantum signal feature weight increases, the key security feature weight should be adjusted synchronously, and the computational resource utilization rate should not exceed the threshold as the transmission delay decreases (e.g., when the transmission delay decreases from 50ms to 30ms, the quantum bit utilization rate should still be ≤80%). The quantization range of the four-dimensional parameters must match. For example, the quantum signal feature value (0-1), key security level (1-5), computational resource utilization rate (0%-100%), and transmission delay (0-100ms) must be logically coordinated through normalization processing.

[0027] S103 uses a variable quantum computing model to simulate the evolution trajectory of transmission security in multiple scenarios. Combined with a dynamic security model for the entire life cycle of quantum keys, it uses a quantum approximation optimization algorithm coupled with a quantum simulated annealing algorithm to solve for the optimal security solution under the constraints of quantum bit resources and transmission delay threshold, thereby generating an adaptive key distribution scheme.

[0028] In one implementation, parameters of the variable quantum computing model, quantum key lifecycle security parameters, quantum optimization algorithm coupling parameters, and resource and delay constraint parameters are extracted and classified to generate transmission security evolution simulation dimension information, key dynamic security monitoring item information, optimization algorithm adaptation evaluation dimension information, and optimal solution constraint basis information. First, the specific scope of the four core parameters is clarified, and then they are classified and organized according to functional attributes to generate corresponding dimension information. An example is illustrated below: Parameters of the variational quantum computation model include the number of quantum variational circuit layers (e.g., 4 layers), the number of variational parameters (e.g., 32 trainable qubit parameters), the choice of quantum measurement basis (e.g., combined Z-basis and X-basis measurements), and the iterative convergence threshold (e.g., loss function value ≤ 0.001). Quantum key lifecycle security parameters include the entropy source randomness index during key generation (e.g., minimum entropy ≥ 256 bits), the upper limit of the bit error rate during key distribution (e.g., ≤ 1.5%), the encryption strength during key storage (e.g., AES-256 encryption protection), and the thoroughness verification threshold during key destruction (e.g., data overwriting ≥ 3 times).

[0029] Coupling parameters for quantum optimization algorithms include the number of iterations for the quantum approximation optimization algorithm (e.g., 200 steps), the cooling rate for the quantum simulated annealing algorithm (e.g., a temperature decrease of 0.1 per 10 iterations), the weight allocation ratio between the two algorithms (e.g., 0.6 for the quantum approximation optimization algorithm and 0.4 for the quantum simulated annealing algorithm), and the algorithm termination condition (e.g., no change in the optimal solution after 5 consecutive iterations). Resource and delay constraint parameters include the upper limit of qubit resources (e.g., ≤64 qubits per transmission), the constraint on the number of quantum gate operations (e.g., ≤100 quantum gates per line), the transmission delay threshold (e.g., end-to-end delay ≤30ms), and the upper limit of computational resource utilization (e.g., quantum processor utilization ≤75%).

[0030] The classification and processing of generated dimensional information includes transmission security evolution simulation dimensional information covering simulation scenario type (e.g., fiber channel, free space channel, hybrid channel), evolution time step (e.g., one time step every 10ms), and security status assessment dimensions (e.g., key integrity, transmission confidentiality, and anti-attack capability). Key dynamic security monitoring information includes monitoring items such as key generation entropy monitoring, real-time distribution error rate monitoring, storage encryption status monitoring, and destruction thoroughness verification monitoring, along with their corresponding monitoring frequencies (e.g., monitoring once every 500ms).

[0031] The optimization algorithm adaptation evaluation dimensions include algorithm convergence speed, optimal solution accuracy, resource consumption ratio, and scenario adaptability score. The optimal solution solution constraint information clarifies the specific quantitative standards and priorities of constraints such as qubit resources, number of quantum gate operations, transmission delay, and computational resource utilization (e.g., transmission delay constraints have higher priority than computational resource utilization constraints).

[0032] Based on quantum secure transmission technology specifications, the information on transmission security evolution simulation dimensions, key dynamic security monitoring items, optimization algorithm adaptation evaluation dimensions, and optimal solution constraint basis is processed to generate evolution simulation accuracy matching information and key security monitoring real-time evaluation information. According to quantum secure transmission technology specifications (such as the ISO / IEC 19790 quantum secure communication standard), the four types of dimensional information generated above are subjected to compliance verification and adaptation processing to generate evolution simulation accuracy matching information and key security monitoring real-time evaluation information. An example is illustrated below: In accordance with the technical specification requirement of "transmission security evolution simulation error ≤ 5%", the simulation accuracy under different scenarios was verified. For example, in the fiber optic channel scenario, the simulated key error rate had an error of 3.2% compared with the actual test value, which met the accuracy requirement. In the free space channel scenario, the initial simulation error was 7.8%. By adjusting the iterative convergence threshold of the variable quantum computing model (from 0.001 to 0.0005), the error was reduced to 4.5%. Finally, an evolution simulation accuracy matching result table for each scenario was generated, clarifying the matching status (qualified / unqualified) and adjustment measures.

[0033] In accordance with the technical specification requirement of "monitoring response latency ≤ 10ms", the real-time performance of each monitoring item was evaluated. For example, the response latency for monitoring key generation entropy was 3ms, which meets the requirement; the initial response latency for monitoring key storage encryption status was 12ms. By optimizing the sampling frequency of the monitoring algorithm (from once every 500ms to once every 300ms), the latency was reduced to 8ms. A real-time performance evaluation report for each monitoring item was generated, including response latency data, compliance with the specification, and optimization scheme.

[0034] Based on evolutionary simulation accuracy matching information and key security monitoring real-time evaluation information, target data in the parameters of the variable quantum computing model and the security parameters of the entire quantum key lifecycle are marked and filtered to generate a secure transmission parameter correlation screening result. Using evolutionary simulation accuracy matching information (accuracy ≥ 95% is acceptable) and key security monitoring real-time evaluation information (response latency ≤ 10ms is acceptable) as the screening criteria, target data in the parameters of the variable quantum computing model and the security parameters of the entire quantum key lifecycle are marked and filtered. An example is given below: parameters that meet the accuracy and real-time requirements are marked as "valid parameters," those that do not are marked as "parameters to be optimized," and core parameters that are strongly related to accuracy and real-time performance are marked as "critical parameters."

[0035] In the variational quantum computing model parameters, the simulation accuracy corresponding to "4 layers of quantum variational circuit" and "32 variational parameters" is 96.8%, marked as "effective parameters + key parameters"; the "iteration convergence threshold of 0.001" leads to substandard simulation accuracy in the free-space channel scenario, marked as "parameter to be optimized". In the quantum key lifecycle security parameters, the monitoring response latency corresponding to "minimum entropy for key generation ≥ 256 bits" and "distribution error rate ≤ 1.5%" is 3ms, marked as "effective parameters + key parameters"; the "AES-256 encryption for key storage" results in a 12ms response latency due to monitoring algorithm redundancy, marked as "parameter to be optimized". Finally, 28 effective parameters were selected, including 15 key parameters and 6 parameters to be optimized, forming a correlation screening result for secure transmission parameters, clarifying the status of each parameter, the associated accuracy / real-time indicators, and the screening criteria.

[0036] The results of the correlation screening of secure transmission parameters are integrated and quantified to generate design results for a quantum key adaptive distribution scheme. The integration follows the logic of "quantum computing model parameters - key security parameters - optimization algorithm parameters - constraint parameters," clarifying the relationships between parameters. For example, "4 layers of quantum variational circuit" is associated with "key distribution error rate ≤ 1.5%" and "200 iteration steps of quantum approximation optimization algorithm," forming a parameter combination chain of "model-security-algorithm."

[0037] A weighted summation method was used to quantify and score the security performance, resource consumption, and transmission efficiency of parameter combinations (with weights of 0.5, 0.3, and 0.2, respectively). For example, parameter combination A (4-layer quantum variational circuit + 256-bit key minimum entropy + 200-step optimization algorithm iteration + 64 qubits) scored 92 points for security performance, 78 points for resource consumption, and 85 points for transmission efficiency, for a total score of 86.4 points. Parameter combination B (3-layer quantum variational circuit + 256-bit key minimum entropy + 150-step optimization algorithm iteration + 48 qubits) scored 81.2 points. Ultimately, parameter combination A, with the highest total score, was selected as the core configuration.

[0038] The adaptive key distribution scheme explicitly includes key generation configuration (minimum entropy of 256 bits, generation frequency once every 1 second), distribution strategy (adaptively adjusting the bit error rate threshold based on channel type, ≤1.2% for fiber optic channels and ≤1.5% for free space channels), optimization algorithm configuration (200 iterations of quantum approximation optimization algorithm + 0.1 cooling rate of quantum simulated annealing algorithm), and resource adaptation rules (dynamically adjusting the distribution rate according to the remaining amount of quantum bit resources, distributing at full speed when ≥64 quantum bits remain, and reducing the speed by 50% when 32-64 quantum bits remain). It also includes quantitative results such as the scheme's security performance indicators (attack resistance strength ≥9 levels), resource consumption indicators (quantum bit utilization ≤70%), and transmission efficiency indicators (distribution rate ≥100kbps).

[0039] S104 constructs an intelligent decision-making agent through a quantum reinforcement learning algorithm, embeds a quantum reward function to adjust the quantum error correction code rate, key update frequency and quantum signal modulation strategy in real time, and generates a deep fusion control scheme that balances security and efficiency.

[0040] In one implementation, the data related to the construction of the decision agent is adapted based on a quantum reinforcement learning model. A dynamic curve of the security level is used to present the trend of transmission security fluctuations, a line graph of efficiency improvement shows the effect of strategy optimization, and a heat map of parameter adjustments marks key control nodes. Feature identifiers are added to the core decision data to generate a visual representation of the decision agent construction data. The original collected quantum error correction code rate data (range 0.3-0.9) is normalized to the [0,1] interval, and outliers caused by sudden interference in the quantum channel (such as invalid data with an instantaneous code rate of 0.1) are removed, retaining 3000 sets of valid data samples. The key update frequency data (unit: times / minute) is smoothed to eliminate short-term fluctuation interference.

[0041] The visualization results are as follows: For the dynamic curve of the security level, time is plotted on the horizontal axis (each node is 5 seconds) and security level (levels 1-10) on the vertical axis, showing the trend of transmission security fluctuations within one hour. The curve shows that the security level is stable at level 8-9 for the first 30 minutes, drops to level 6 due to channel interference at the 35th minute, and recovers to level 8 within 10 minutes after adjusting the signal modulation strategy. For the efficiency improvement line graph, the number of strategy iterations (0-50 times) is plotted on the horizontal axis and transmission efficiency (unit: Mbps) is plotted on the vertical axis, showing the optimization effect. The initial efficiency is 50Mbps, increasing to 85Mbps after 30 iterations, and stabilizing at 90Mbps after 50 iterations. The slope of the line gradually slows down, indicating that the optimization tends to converge. For the parameter adjustment heatmap, the quantum error correction code rate (0.5-0.9) is used as the horizontal axis and the key update frequency (5-20 times / minute) is used as the vertical axis. The color depth represents the overall transmission performance (dark color is better). The heatmap marks the key control nodes: when the error correction code rate is 0.7 and the update frequency is 15 times / minute, the color is the darkest and the overall performance is the best.

[0042] For core data such as parameter combinations with a security level below 7 and iterative steps that improve efficiency by more than 10Mbps, mark them with characteristic labels such as "high risk" and "high efficiency optimization" to form a decision agent to build a data visualization report.

[0043] Based on the quantum secure transmission technology specifications, the compliance of the strategy adjustment was verified. The standard for setting the quantum error correction code rate, the key update frequency threshold, and the adaptability of the signal modulation strategy were checked one by one to determine whether they met the requirements of quantum secure transmission technology. Non-compliant items were marked with optimization directions, and the compliance verification results of the strategy adjustment were generated. The technical specifications require the quantum error correction code rate to be in the range of 0.5-0.9, and the error tolerance code rate ≤2%. The verification found that the initial error correction code rate of 0.45 was lower than the lower limit of the specification, and the optimization direction was marked as "increase the code rate to the range of 0.5-0.9, and it is recommended to set the initial value to 0.7"; another setting of 0.85, corresponding to an error tolerance code rate of 1.8%, met the specification requirements and was marked as "compliant".

[0044] The standard stipulates that the key update frequency should be ≥10 times / minute in high-security scenarios and ≥5 times / minute in ordinary scenarios. The review found that in financial data transmission scenarios, the key update frequency was set to 8 times / minute, which is lower than the requirement for high-security scenarios; the optimization direction is marked as "increase to 10-15 times / minute". In office data transmission scenarios, the frequency was set to 6 times / minute, which meets the requirements for ordinary scenarios and is marked as "compliant".

[0045] The specification requires that the signal modulation strategy be compatible with the quantum channel type (e.g., QPSK modulation for fiber optic channels and BPSK modulation for free-space channels). The review found that QPSK modulation was used in the free-space channel, indicating a compatibility issue; the optimization direction was marked as "switching to BPSK modulation to improve signal transmission stability." QPSK modulation was used in the fiber optic channel, showing good compatibility and was marked as "compliant." A compliance verification report for the strategy adjustment was generated, clearly identifying 3 non-compliant items, their corresponding reasons for violation and optimization directions, and 12 compliant items and their verification basis.

[0046] A significance verification method based on the correlation between security performance and transmission efficiency was adopted. Evaluation weights were defined for security and efficiency indicators in the decision data, and a weighted scoring method was used to calculate the overall suitability, enhancing the objectivity of the evaluation results. Considering the needs of practical application scenarios, a weight of 0.6 was set for security performance and 0.4 for transmission efficiency. Security performance includes three secondary indicators: attack resistance strength, key confidentiality, and data integrity, with weights of 0.3, 0.2, and 0.1, respectively. Transmission efficiency includes two secondary indicators: transmission rate and latency control, with weights of 0.25 and 0.15, respectively.

[0047] For a given parameter combination (error correction rate 0.7, update frequency 12 times / minute, BPSK modulation), the following scores are given: attack resistance 9 points (out of 10), key confidentiality 8.5 points, and data integrity 9.2 points; transmission rate 88Mbps (corresponding to a score of 8.8), latency 25ms (corresponding to a score of 9). The overall security performance score is calculated as follows: 9×0.3 + 8.5×0.2 + 9.2×0.1 = 2.7 + 1.7 + 0.92 = 5.32 points; the overall transmission efficiency score is calculated as follows: 8.8×0.25 + 9×0.15 = 2.2 + 1.35 = 3.55 points; and the overall adaptability score is calculated as follows: 5.32×0.6 + 3.55×0.4 = 3.192 + 1.42 = 4.612 points (out of 6). By adjusting the parameter combinations multiple times for scoring, the correlation coefficient between security performance and transmission efficiency was calculated to be 0.72, indicating a strong positive correlation between the two. The verification results are significant, which strengthens the objectivity of the evaluation.

[0048] By integrating and analyzing the data visualization results from the decision-making agent, the compliance verification results of strategy adjustments, and the verification results of the significance of the security-efficiency correlation, an evaluation result of the deep integration of quantum security and efficiency control scheme is generated, which includes security protection enhancement values, transmission efficiency improvement parameters, and quantum security compliance levels. Based on the compliance verification results, compliant parameter combinations are prioritized; combined with the key control nodes in the visualization results, the parameter range with the best overall performance is selected; and the final parameter configuration is determined with reference to the security-efficiency weighted score.

[0049] The evaluation results are as follows: After optimization, the anti-attack strength improved from level 7.5 to level 9, the key confidentiality score improved from 7 to 8.5, and the overall security protection enhancement value increased by 18%. The transmission rate increased from the initial 50Mbps to 90Mbps, an improvement of 80%; the transmission latency decreased from 40ms to 25ms, a reduction of 37.5%. All parameters after optimization meet the technical specifications, and the compliance level improved from "Level B" to "Level A". A final evaluation report on the quantum security and efficiency deep integration control scheme was generated, clarifying the core advantages, quantitative indicators, and applicable scenarios of the scheme.

[0050] S105 integrates quantum federated learning mechanism and quantum secret sharing technology to construct a global collaborative management strategy and generate quantum global collaborative management strategy adaptation and fusion features.

[0051] In one implementation, information quantification analysis is performed on the parameters of the quantum federated learning mechanism and the parameters of the quantum secret sharing technology to generate collaborative management quantification factors, privacy protection quantification factors, and resource adaptation quantification factors. The collaborative management quantification factor characterizes the degree of cooperation between the two technologies in global management; the privacy protection quantification factor characterizes the key security protection advantage of the quantum secret sharing technology; and the resource adaptation quantification factor characterizes the scheduling correlation of the quantum federated learning mechanism with distributed resources. First, the specific parameters of the two technologies are defined, and then information quantification is performed using the analytic hierarchy process (AHP) and entropy weight method to generate three types of quantification factors: collaborative management, privacy protection, and resource adaptation. An example is given below: the parameters of the quantum federated learning mechanism include the participation rate of distributed nodes (e.g., 6 out of 8 nodes participate stably, participation rate 75%), the aggregation frequency of node model parameters (e.g., aggregation once every 10 minutes), the number of federated learning iteration rounds (e.g., 50 rounds), and the encryption strength of cross-node data interaction (e.g., AES-256 encryption based on quantum keys). The parameters of quantum secret sharing technology include the number of key splits (e.g., splitting a 256-bit key into 6 parts), the key recombination threshold (e.g., at least 4 parts are required for recombination), the number of security checks for share transmission (e.g., 3 checks per transmission), and the encryption level of key share storage (e.g., encrypted storage in a quantum memory).

[0052] For the collaborative management quantification factor, the analytic hierarchy process (AHP) is used to calculate the degree of collaboration between the two technologies, with a quantification range of 0-1 (the closer to 1, the better the collaboration). For example, the node aggregation frequency of quantum federated learning matches the key recombination timeliness of quantum secret sharing by 89%, resulting in a collaborative management quantification factor of 0.89. For the privacy protection quantification factor, the entropy weight method is used to evaluate the key protection capability of quantum secret sharing technology, with a quantification range of 0-10 (the higher the score, the stronger the protection). For example, with a key splitting share of 6 and a recombination threshold of 4, the ability to resist collusion attacks scores 9.2, resulting in a privacy protection quantification factor of 9.2. For the resource adaptation quantification factor, the resource scheduling adaptation of quantum federated learning is quantified by combining the distributed node resource occupancy, with a quantification range of 0-1. For example, during the iteration process of quantum federated learning, the node computing resource utilization rate is stable at 60%-70%, and the storage resource occupancy rate is ≤50%, resulting in a resource adaptation quantification factor of 0.78.

[0053] Based on collaborative management quantization factors, privacy protection quantization factors, and resource adaptation quantization factors, the distributed node collaboration parameters in the quantum federated learning mechanism and the key splitting and recombination parameters in the quantum secret sharing technology are fused to generate information quantization and technical parameter correlation fusion features. Using the three types of quantization factors as weights, a weighted fusion process is performed on the distributed node collaboration parameters of quantum federated learning and the key splitting and recombination parameters of quantum secret sharing. An example is given below: the collaborative management quantization factor weight is set to 0.4, the privacy protection quantization factor weight to 0.3, and the resource adaptation quantization factor weight to 0.3.

[0054] For the distributed node collaboration parameters, with a node participation rate of 75% and a model aggregation frequency of 10 minutes / time, the weighted average is 75%×0.4+(1 / 10)×60×0.3 (converting the aggregation frequency to a standardized value)=0.3+0.18=0.48. For the key splitting and reassembly parameters, with a key splitting share of 6 parts and a reassembly threshold of 4 parts, the weighted average is (6 / 10)×0.3+(4 / 6)×0.4=0.18+0.27=0.45 (standardizing the share number and threshold).

[0055] The weighted results of the two types of parameters are correlated and integrated to form a feature vector [0.48, 0.45, 0.89, 9.2, 0.78] that integrates information quantification and technical parameters. This vector covers the parameter fusion results and three types of core quantification factors, and fully represents the correlation between technical parameters and quantification information.

[0056] Based on cross-node communication factors and resource heterogeneity factors in quantum-secure transmission scenarios, this paper fuses collaborative management quantification factors, privacy protection quantification factors, and resource adaptation quantification factors to generate scenario-resource factor and information quantification correlation fusion features. Cross-node communication factors include communication latency and channel stability, while resource heterogeneity factors include node computing power and storage resource differences. First, the cross-node communication factors and resource heterogeneity factors in quantum-secure transmission scenarios are defined, and then coupled and fused with the three types of quantification factors. An example is given below: For the cross-node communication factor, with a communication latency of 25ms (normalized value 0.7, lower latency means higher value) and channel stability of 92% (normalized value 0.92), the weighted fusion (weights 0.5 each) yields: 25ms × 0.5 + 92% × 0.5 = 0.7 × 0.5 + 0.92 × 0.5 = 0.81. For the resource heterogeneity factor, the node computing power difference coefficient is 0.2 (the smaller the difference, the lower the value, with a standardized value of 0.8), and the storage resource difference coefficient is 0.3 (standardized value of 0.7). After weighted fusion (with each weight being 0.5), the result is 0.8×0.5+0.7×0.5=0.75.

[0057] A multiplicative fusion method is employed to multiply the fusion result of scenario factors with quantitative factors for collaborative management and control, privacy protection, and resource adaptation, generating a fusion feature that correlates scenario and resource factors with quantitative information. For example, 0.81 (cross-node communication factor) × 0.75 (resource heterogeneity factor) × 0.89 (collaborative management and control factor) × 9.2 (privacy protection factor) × 0.78 (resource adaptation factor) = 3.68 (0.86 after standardization), ultimately forming a feature value of 0.86 and the corresponding feature description "high collaboration-high privacy protection fusion feature under high communication stability and low resource heterogeneity scenarios".

[0058] The fusion features of information quantification and technical parameters, and the fusion features of scenario and resource factors and information quantification are analyzed and processed to generate quantum global collaborative management strategy adaptation fusion features. Principal component analysis (PCA) is used to reduce the dimensionality and extract core information of these features, generating adaptation fusion features. An example is shown below: The two types of fusion features (feature vector [0.48, 0.45, 0.89, 9.2, 0.78] and eigenvalue 0.86) are normalized to a unified quantization range of [0, 1]. PCA identifies the top three core components with the highest variance contribution rates: "Technology Collaboration - Parameter Adaptation Principal Component," "Scenario Adaptation - Resource Scheduling Principal Component," and "Privacy Protection - Security Enhancement Principal Component," with variance contribution rates of 42%, 35%, and 18%, respectively, for a cumulative contribution rate of 95%.

[0059] Based on the core components and contribution rates, quantum global collaborative management strategy adaptation and fusion features are generated, specifically: collaborative control adaptation degree 0.87 (technical collaboration - parameter adaptation core component score), scenario resource adaptation degree 0.86 (scenario adaptation - resource scheduling core component score), and privacy and security adaptation degree 0.93 (privacy protection - security enhancement core component score). The feature description is formed as "global collaborative management adaptation features that support efficient collaboration of distributed nodes, adapt to high-stability and low-heterogeneity scenarios, and have strong privacy protection capabilities", providing a core basis for the subsequent formulation of global collaborative strategies.

[0060] S106 is based on a quantum probabilistic graphical model coupled with a quantum Bayesian inference algorithm. It incorporates a quantum global collaborative management strategy adaptation and fusion feature design quantitative evaluation system, and combines a quantum support vector machine to realize transmission security level classification, generating comprehensive evaluation information including security confidence, anti-attack strength, resource utilization efficiency, model fit, and global collaborative management fit.

[0061] In one implementation, the adaptation and fusion characteristics of the quantum global collaborative management strategy and the core technical parameters of quantum secure transmission are quantitatively analyzed to generate basic evaluation quantitative factors. These basic evaluation quantitative factors characterize the degree of fit between the collaborative fusion characteristics and the transmission technical parameters. First, the specific content of the two types of core information is clarified. Then, the adaptation fit is quantified using the cosine similarity method and linear regression analysis to generate the basic evaluation quantitative factors. An example is illustrated below: The quantum global collaborative management strategy adaptation and fusion features include a collaborative control adaptation degree of 0.87, a scene resource adaptation degree of 0.86, and a privacy and security adaptation degree of 0.93 (derived from the adaptation and fusion features generated by S105). Core technical parameters for quantum secure transmission include quantum convolutional neural network error correction accuracy (e.g., 98.5%), variable quantum computing model optimization efficiency (e.g., convergence after 200 iterations), quantum reinforcement learning decision response speed (e.g., ≤8ms), and quantum key distribution rate (e.g., 100kbps).

[0062] The cosine similarity method is used to calculate the fit between two types of information, with a quantization range of 0-1 (the closer to 1, the better the fit). For example, the similarity between a collaborative management fit of 0.87 and a quantum key distribution rate of 100kbps is 0.89; the similarity between a scene resource fit of 0.86 and a quantum convolutional neural network error correction accuracy of 98.5% is 0.91; and the similarity between a privacy and security fit of 0.93 and a quantum reinforcement learning decision response speed ≤8ms is 0.94. Through linear regression analysis and weighted fusion (each weight 1 / 3), the final evaluation basis quantization factor is (0.89+0.91+0.94) / 3=0.91.

[0063] Real-time security response data and resource scheduling efficiency data for quantum-secure transmission are collected and compared with benchmark data for the same scenario in the historical transmission database to generate a quantitative factor for transmission performance trends. The process involves collecting real-time quantum-secure transmission data and comparing it with historical benchmark data for the same scenario to quantify transmission performance trends. An example is as follows: Real-time collection of security response data (e.g., anti-attack response latency: 22ms, 23ms, 21ms…) and resource scheduling efficiency data (e.g., quantum processor utilization: 68%, 70%, 69%…) within one hour; and retrieving benchmark data for the same scenario (e.g., fiber optic channel financial data transmission) from the historical database: a benchmark value of 25ms for security response latency and 65% for resource scheduling efficiency, and performing spatiotemporal matching and alignment along the time dimension (one node every 10 minutes).

[0064] The performance trend is quantified by calculating the deviation rate and slope of real-time data from the baseline data. The quantification range is 0-1 (0.5 is the baseline, greater than 0.5 indicates performance improvement, less than 0.5 indicates performance degradation). For example, the average real-time security protection response latency is 22ms, a 12% decrease from the baseline value of 25ms, with a slope of -0.005 (continuous optimization); the average real-time resource scheduling efficiency is 69%, a 6.15% increase from the baseline value of 65%, with a slope of 0.003 (continuous improvement). After weighted fusion (security protection weight 0.6, resource scheduling weight 0.4), the transmission performance trend quantification factor is ((25-22) / 25×0.6+(69-65) / 65×0.4)+0.5=(0.072+0.0246)+0.5=0.5966, indicating that the transmission performance is showing an improving trend.

[0065] Based on the evaluation foundation quantization factor and the transmission performance trend quantization factor, and combined with the coupled architecture of quantum probabilistic graphical model-quantum Bayesian inference and the classification hierarchy design of quantum support vector machine, this method fuses the accuracy of security level determination and the dynamic adjustment effect of evaluation indicators for different transmission scenarios to generate collaborative-performance-classification correlated quantization features. Using the evaluation foundation quantization factor and the transmission performance trend quantization factor as input, and combining the coupled architecture of quantum probabilistic graphical model-quantum Bayesian inference and the classification hierarchy design of quantum support vector machine, the method fuses the accuracy of security level determination and the adjustment effect of evaluation indicators to generate correlated quantization features. An example is illustrated below: A quantum probabilistic graphical model-quantum Bayesian inference coupling architecture is used to quantify the uncertainty in security level determination, such as posterior probability estimation of transmission security level. A quantum support vector machine classification hierarchy is divided into 5 security levels (level 1 is the lowest, level 5 is the highest) for accurate classification of transmission security levels.

[0066] The evaluation base quantization factor of 0.91 and the transmission performance trend quantization factor of 0.5966 are input into the coupling architecture. The posterior probabilities of different security levels are calculated through quantum Bayesian inference. For example, the posterior probability of security level 5 is 0.85 and that of security level 4 is 0.13. Combined with the classification hierarchy of quantum support vector machine, the dynamic adjustment effect of evaluation indicators (such as security confidence and resource utilization efficiency) is quantified (the indicator improvement rate after adjustment is 15%). Finally, a collaboration-performance-classification correlation quantified feature vector [0.91, 0.5966, 0.85, 15%] is generated, which covers the dual quantization factors, the posterior probability of the highest security level, and the indicator adjustment improvement rate.

[0067] Based on a quantitative evaluation system, the correlation between collaboration, performance, and classification is analyzed and processed to generate multi-dimensional comprehensive evaluation information for quantum secure transmission. Using the correlation between collaboration, performance, and classification as the core quantitative feature, and combining it with the multi-dimensional index weights of the quantitative evaluation system, comprehensive analysis generates evaluation information containing core quantitative results. An example is provided below: Security confidence weight is set at 0.3, attack resistance weight at 0.25, resource utilization efficiency weight at 0.2, model fit weight at 0.15, and global collaborative management fit weight at 0.1. Scoring is performed based on the correlation quantitative feature vector [0.91, 0.5966, 0.85, 15%]. Security confidence: posterior probability of the highest security level 0.85 × 100 = 85 points; Anti-attack strength: combined with a 12% decrease in security protection response latency, score 88 points; Resource utilization efficiency: combined with a 6.15% increase in processor utilization, score 75 points; Model fit: evaluation of basic quantitative factors 0.91 × 100 = 91 points; Global collaborative management fit: indicator adjustment improvement rate 15% × 100 = 85 points.

[0068] The weighted comprehensive score is calculated as follows: 85×0.3+88×0.25+75×0.2+91×0.15+85×0.1=25.5+22+15+13.65+8.5=84.65 points. The final multi-dimensional comprehensive evaluation information is generated: security confidence 85%, attack resistance level 8.8 (out of 10), resource utilization efficiency 75%, model adaptability 91%, and global collaborative management fit 85%. The comprehensive evaluation level is "A". An evaluation note is also attached: "In the current transmission scenario, the security protection, model adaptability, and global collaborative performance are excellent, but resource utilization efficiency needs further optimization."

[0069] like Figure 2 As shown, a quantum computing model-driven quantum and quantum key fusion secure transmission device includes: The acquisition module 201 is used to collect multidimensional feature parameters of quantum signals and core data for quantum key generation. It introduces a quantum deep learning neural network feature extraction module and uses quantum activation functions and quantum pooling layers to achieve noise reduction, redundancy removal and key information enhancement of high-dimensional quantum data. Processing module 202 is used to process the optimized data based on quantum convolutional neural networks. It integrates quantum entanglement distillation technology with a deep quantum error correction coding algorithm driven by quantum attention mechanism to construct a four-dimensional correlation model of quantum signal-key security-computational resources-transmission delay. Through quantum transfer learning, it transfers parameters of the cross-scenario quantum secure transmission model. Combined with the rapid adaptation capability of the quantum meta-learning optimization model, it generates dynamic security warning information and key hierarchical optimization requirements. A variable quantum computing model is used to simulate the evolution trajectory of transmission security in multiple scenarios. Combined with a dynamic security model of the entire quantum key lifecycle, a quantum approximation optimization algorithm coupled with a quantum simulated annealing algorithm is used to solve for the optimal security solution under qubit resource constraints and transmission delay thresholds, generating an adaptive... The key distribution scheme utilizes quantum reinforcement learning algorithms to construct an intelligent decision-making agent, embedding a quantum reward function to adjust the quantum error correction code rate, key update frequency, and quantum signal modulation strategy in real time, generating a deep fusion control scheme that balances security and efficiency. It integrates quantum federated learning mechanisms and quantum secret sharing technology to construct a global collaborative management strategy, generating quantum global collaborative management strategy adaptation and fusion features. Based on a quantum probabilistic graphical model coupled with a quantum Bayesian inference algorithm, it designs a quantitative evaluation system by integrating quantum global collaborative management strategy adaptation and fusion features, and combines quantum support vector machines to achieve transmission security level classification, generating comprehensive evaluation information including security confidence, anti-attack strength, resource utilization efficiency, model fit, and global collaborative control fit.

[0070] A computing device includes a memory for storing computer program instructions and a processor for executing the computer program instructions, wherein when the computer program instructions are executed by the processor, the device is triggered to execute a quantum computing model-driven quantum and quantum key fusion secure transmission method.

[0071] The methods and / or embodiments in this application can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowchart. When the computer program is executed by a processing unit, it performs the functions defined in the methods of this application.

[0072] It should be noted that the computer-readable medium described in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0073] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0074] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within this application.

Claims

1. A quantum computing model-driven secure transmission method for quantum and quantum key fusion, characterized in that, include: We collect multidimensional feature parameters of quantum signals and core data for quantum key generation, introduce a quantum deep learning neural network feature extraction module, and use quantum activation functions and quantum pooling layers to achieve noise reduction, redundancy removal and key information enhancement of high-dimensional quantum data. Based on the quantum convolutional neural network processing of the above optimized data, the quantum entanglement distillation technology and the quantum attention mechanism-driven deep quantum error correction coding algorithm are integrated to construct a four-dimensional correlation model of quantum signal-key security-computation resources-transmission delay. The parameters of the cross-scenario quantum secure transmission model are transferred through quantum transfer learning, and combined with the rapid adaptation capability of the quantum meta-learning optimization model, dynamic security early warning information and key hierarchical optimization requirements are generated. A variable quantum computing model is used to simulate the evolution trajectory of transmission security in multiple scenarios. Combined with a dynamic security model of the entire life cycle of quantum keys, a quantum approximation optimization algorithm coupled with a quantum simulated annealing algorithm is used to solve the optimal security solution under the constraints of quantum bit resources and transmission delay threshold, and an adaptive key distribution scheme is generated. Intelligent decision-making agents are constructed by using quantum reinforcement learning algorithms, and quantum reward functions are embedded to adjust the quantum error correction code rate, key update frequency and quantum signal modulation strategy in real time, thereby generating a deep fusion control scheme that balances security and efficiency. A global collaborative management strategy is constructed by integrating quantum federated learning mechanism and quantum secret sharing technology, and adaptive and fusion features of quantum global collaborative management strategy are generated. Based on a quantum probabilistic graphical model coupled with a quantum Bayesian inference algorithm, a quantitative evaluation system is designed by integrating quantum global collaborative management strategy adaptation and fusion features. Combined with quantum support vector machines, transmission security level classification is realized, generating comprehensive evaluation information including security confidence, anti-attack strength, resource utilization efficiency, model fit, and global collaborative management fit.

2. The quantum computing model-driven secure transmission method for quantum and quantum key fusion according to claim 1, characterized in that, Based on the optimized data processed by quantum convolutional neural networks, a deep quantum error correction coding algorithm driven by quantum entanglement distillation technology and quantum attention mechanism is integrated to construct a four-dimensional correlation model of quantum signal-key security-computational resources-transmission delay. Quantum transfer learning is used to transfer parameters of the cross-scenario quantum secure transmission model. Combined with the rapid adaptation capability of the quantum meta-learning optimization model, dynamic security early warning information and key hierarchical optimization requirements are generated, including: A correlation analysis was conducted on the core technical parameters related to quantum-optimized data processing and model adaptation parameters to generate dynamic correlation analysis results between quantum data processing and model adaptation. The core technical parameters related to quantum-optimized data processing include quantum convolutional neural network processing parameters, quantum entanglement distillation technology parameters, quantum attention mechanism parameters, and deep quantum error correction coding algorithm parameters. The model adaptation parameters include quantum transfer learning cross-scene parameters and quantum meta-learning optimization adaptation parameters. The correlation analysis results are filtered and processed to generate a target simulation variable list, which includes core technical parameters for quantum data processing, quantum model adaptation and optimization parameters, four-dimensional correlation model construction parameters, dynamic security early warning generation parameters, and key layering optimization requirement parameters. The correlation analysis results and the target simulation variable list are processed to generate parameter combination constraints, including the collaborative adaptation constraints of quantum technology integration, the compatibility constraints of cross-scenario parameter transfer, the timeliness matching constraints of rapid model adaptation, and the logical consistency constraints of four-dimensional correlation model construction.

3. The quantum computing model-driven quantum and quantum key fusion secure transmission method according to claim 1, characterized in that, A variable quantum computing model is used to simulate the evolution trajectory of transmission security in multiple scenarios. Combined with a dynamic security model of the entire quantum key lifecycle, a quantum approximation optimization algorithm coupled with a quantum simulated annealing algorithm is used to solve for the optimal security solution under qubit resource constraints and transmission delay thresholds, generating an adaptive key distribution scheme, including: The parameters of the variable quantum computing model, the quantum key lifecycle security parameters, the quantum optimization algorithm coupling parameters, and the resource and delay constraint parameters are extracted and classified to generate transmission security evolution simulation dimension information, key dynamic security monitoring item information, optimization algorithm adaptation evaluation dimension information, and optimal solution constraint basis information. Based on the quantum secure transmission technology specifications, the transmission security evolution simulation dimension information, key dynamic security monitoring item information, optimization algorithm adaptation evaluation dimension information, and optimal solution solution constraint basis information are processed to generate evolution simulation accuracy matching information and key security monitoring real-time evaluation information. Based on the accuracy matching information of evolution simulation and the real-time evaluation information of key security monitoring, target data in the parameters of the variable quantum computing model and the security parameters of the quantum key throughout its life cycle are marked and filtered to generate the correlation filtering results of secure transmission parameters; The results of the correlation screening of secure transmission parameters are integrated and quantified to generate the design results of the quantum key adaptive distribution scheme.

4. The quantum computing model-driven secure transmission method for quantum and quantum key fusion according to claim 1, characterized in that, Intelligent decision-making agents are constructed using quantum reinforcement learning algorithms. A quantum reward function is embedded to adjust the quantum error correction code rate, key update frequency, and quantum signal modulation strategy in real time, generating a deeply integrated control scheme that balances security and efficiency. This includes: Based on the quantum reinforcement learning model, the relevant data for decision agent construction are adapted and processed. The dynamic curve of security level presents the trend of transmission security fluctuation, the line graph of efficiency improvement shows the effect of strategy optimization, the heat map of parameter adjustment marks the key control nodes, the core decision data is labeled with feature identifiers, and the data visualization results of decision agent construction are generated. Based on the quantum secure transmission technology specifications, the compliance of the strategy adjustment basis is verified. The quantum error correction code rate setting standard, key update frequency threshold, signal modulation strategy adaptability and other dimensions are checked one by one to determine whether they meet the requirements of quantum secure transmission technology. Non-compliant items are marked with optimization directions and the strategy adjustment compliance verification results are generated. Significance verification of the correlation between security performance and transmission efficiency is adopted. Evaluation weights are defined for security and efficiency indicators in decision data. The comprehensive fit is calculated by weighted scoring method to enhance the objectivity of evaluation results. By integrating and analyzing the data visualization results of the decision agent construction, the compliance verification results of the strategy adjustment, and the significant verification results of the security-efficiency correlation, an evaluation result of the quantum security and efficiency deep integration control scheme is generated, which includes security protection enhancement value, transmission efficiency improvement parameter, and quantum security compliance level.

5. The quantum computing model-driven secure transmission method for quantum and quantum key fusion according to claim 1, characterized in that, A global collaborative management strategy is constructed by integrating quantum federated learning mechanisms and quantum secret sharing technology. This quantum global collaborative management strategy is adapted to and incorporates the following features: Information quantification analysis is performed on the parameters of the quantum federated learning mechanism and the parameters of the quantum secret sharing technology to generate collaborative management quantification factor, privacy protection quantification factor, and resource adaptation quantification factor. Among them, the collaborative management quantification factor represents the degree of cooperation between the two technologies in global management, the privacy protection quantification factor represents the key security protection advantage of the quantum secret sharing technology, and the resource adaptation quantification factor represents the degree of scheduling correlation of the quantum federated learning mechanism for distributed resources. Based on collaborative management quantization factor, privacy protection quantization factor, and resource adaptation quantization factor, the distributed node collaboration parameters in the quantum federated learning mechanism and the key splitting and recombination parameters in the quantum secret sharing technology are fused to generate information quantization and technical parameter correlation fusion features. Based on cross-node communication factors and resource heterogeneity factors in quantum secure transmission scenarios, collaborative management quantification factors, privacy protection quantification factors, and resource adaptation quantification factors are fused to generate scenario-resource factor and information quantification correlation fusion features. Among them, cross-node communication factors include communication delay and channel stability, and resource heterogeneity factors include node computing power and storage resource differences. The characteristics of information quantification and technical parameter correlation and fusion, and the characteristics of scenario and resource factors correlation and fusion with information quantification are analyzed and processed to generate quantum global collaborative management strategy adaptation and fusion characteristics.

6. The quantum computing model-driven secure transmission method for quantum and quantum key fusion according to claim 5, characterized in that, Based on a quantum probabilistic graphical model coupled with a quantum Bayesian inference algorithm, a quantitative evaluation system is designed by integrating quantum global collaborative management strategy adaptation and fusion features. Combined with quantum support vector machines, transmission security level classification is implemented, generating comprehensive evaluation information including security confidence, anti-attack strength, resource utilization efficiency, model fit, and global collaborative management suitability. The quantum global collaborative management strategy adaptation and integration characteristics and the core technical parameters of quantum secure transmission are quantitatively analyzed and processed to generate basic evaluation quantitative factors. Among them, the basic evaluation quantitative factors characterize the degree of adaptation and fit between the collaborative integration characteristics and the transmission technical parameters. The security protection response data and resource scheduling efficiency data collected in real time for quantum secure transmission are compared with the benchmark data of the same scenario in the historical transmission database to generate a transmission performance trend quantification factor. Based on the evaluation of basic quantitative factors and transmission performance trend quantitative factors, combined with the coupling architecture of quantum probabilistic graphical model-quantum Bayesian inference and the classification hierarchy design of quantum support vector machine, the accuracy of security level determination and the dynamic adjustment effect of evaluation indicators in different transmission scenarios are fused to generate collaborative-performance-classification related quantitative features. Based on the quantitative evaluation system, the quantitative characteristics of the correlation between collaboration, performance and classification are analyzed and processed to generate multi-dimensional comprehensive evaluation information for quantum secure transmission.

Images