Quantum attack resistant data protection method

Abstract

The invention relates to the field of quantum computing, in particular to an anti-quantum-attack data protection method. According to the method, a configurable double-layer Hash protocol is provided for data protection, and the configurable double-layer Hash protocol is combined with a post-quantum digital signature algorithm to serve as second-layer Hash on the basis of standard SHA-3 Hash so as to enhance the defense capacity for quantum computing attacks. And dynamic selection can be carried out between a standard SHA-3 Hash layer and a post-quantum digital signature algorithm, so that the method not only can provide rapid Hash calculation in a scene with a low security demand, but also can provide additional anti-quantum attack protection in a scene with a high security demand. According to the method, the quantum attack resistance is improved, the adaptability of the Hash algorithm is enhanced, and a more efficient and safer encryption scheme is provided for practical application.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of quantum computing, and in particular to a data protection method against quantum attacks. BACKGROUND

[0002] With the development of quantum computing, traditional encryption algorithms are facing unprecedented security threats. Quantum computers have the ability to significantly accelerate certain computing tasks, the most famous of which are Shor's algorithm and Grover's algorithm. Shor's algorithm can break encryption algorithms based on large number factorization and discrete logarithm (such as RSA and ECC) in polynomial time, while Grover's algorithm can significantly speed up hash collision search. Therefore, under the threat of quantum computers, the security of traditional encryption methods becomes unreliable, and it is urgent to develop quantum-resistant encryption algorithms and protocols.

[0003] Post-Quantum Cryptography (PQC) refers to encryption algorithms that can maintain security in a quantum computer environment. Currently, encryption schemes based on lattice theory, especially post-quantum digital signature algorithms (PQC) (such as Dilithium, Falcon) and homomorphic encryption, are considered one of the most promising quantum-resistant encryption technologies. These algorithms use the mathematical difficulty of lattices, and their security is not affected by quantum computers. In terms of hash algorithms, quantum computers can use Grover's algorithm to accelerate the search for hash collisions with a time complexity of square root. Therefore, in order to improve the quantum computing resistance of hash algorithms, researchers have proposed post-quantum hash algorithms to compensate for the weaknesses of traditional hash algorithms in a quantum computing environment.

[0004] In the prior art, hash and signature mechanisms are usually combined to form multi-layer encryption. The most common post-quantum digital signature algorithms (PQC) include Dilithium and Falcon, both of which are based on lattice encryption technology and still maintain their security under the threat of quantum computers. Some existing quantum-resistant hash algorithms have proposed schemes that combine traditional hash algorithms (such as SHA-3) with post-quantum digital signature algorithms (PQC), but these schemes mostly focus on the encryption process and lack flexible configuration of encryption levels. That is, the configuration ability of the double-layer hash protocol is weak, and the combination of hash and signature is usually fixed, lacking the ability to dynamically adjust and configure. There is no ready-made protocol that can flexibly choose single or double-layer hash mode to adapt to different security needs. SUMMARY

[0005] In view of the defects of the prior art, the present application provides an anti-quantum attack data protection method, and proposes a configurable double-layer hash protocol for data protection. The configurable double-layer hash protocol can dynamically select between a standard SHA-3 hash layer and a post-quantum digital signature algorithm (PQC) (such as Dilithium or Falcon), so that the present method can provide fast hash calculation in a low-security requirement scenario and additional protection against quantum attacks in a high-security requirement scenario. Through such flexible configuration, the present application not only improves the anti-quantum attack capability, but also enhances the adaptability of the hash algorithm, providing a more efficient and secure encryption scheme for practical applications.

[0006] To solve the technical problem, the technical solution adopted by the present application is: an anti-quantum attack data protection method, comprising: A. First layer hash processing, using a hash algorithm to perform hash operation on input data to generate an intermediate hash value H1; B. Second layer hash processing, performing hash compression on the intermediate hash value H1 of step A through a lattice-based post-quantum digital signature algorithm to generate a final hash value; H 1 C. Dynamically selecting a hash mode as a single-layer hash mode or a double-layer hash mode based on a security evaluation engine, a. Single-layer mode, only performing first layer hash processing, outputting H_final = H1; b. Double-layer mode, sequentially performing first layer hash processing and second layer hash processing, outputting H_final = PQC_Sign(H1), PQC_Sign(.) being a lattice-based post-quantum digital signature algorithm.

[0007] Further, the basis for the security evaluation engine to select the hash mode is one or more of the following parameters: Data tag, real-time security risk score, client request credibility, system security posture.

[0008] Further, the security evaluation engine uses a rule-based security evaluation method to evaluate the security requirement level, and selects the hash mode based on the security requirement level: The rule-based security evaluation method uses a set of rules for scoring, and the set of rules includes multiple factors and the weight of each factor. When a certain factor reaches a set threshold, the score of the factor in the security requirement level is automatically increased. When the security requirement level is lower than the set threshold, the single-layer hash mode is used. When the security requirement level is higher than the set threshold, the double-layer hash mode is used.

[0009] ​Further, the security evaluation engine evaluates the security requirement level using a machine learning model, and selects a hash mode based on the security requirement level; the process of evaluating the security requirement level using the machine learning model is as follows: a classifier is trained using historical attack samples, the trained classifier first extracts access data features from the input data, the access data features include IP reputation, request frequency and data type, the IP reputation is the reputation of the access device IP address, the request frequency is the number of requests per unit time, the data type is classified according to the sensitivity of the data, and then the security requirement level of the data is predicted according to the access data features, when the security requirement level is lower than a set threshold, a single-layer hash mode is used, and when the security requirement level is higher than the set threshold, a double-layer hash mode is used.

[0010] Further, the security evaluation engine evaluates the security requirement level using a deep learning model, and selects a hash mode based on the security requirement level; the process of evaluating the security requirement level using the deep learning model is as follows: first, historical data is collected, the historical data includes data of multiple attack types and data of multiple sensitivities, and the deep learning model is trained and evaluated based on the historical data, the trained deep learning model analyzes the features of the newly entered request, calculates the security requirement level, when the security requirement level is lower than a set threshold, a single-layer hash mode is used, and when the security requirement level is higher than the set threshold, a double-layer hash mode is used.

[0011] Further, the security requirement level uses a weighted sum model, and the expression formula is as follows: ; wherein represents the security requirement level, , and are features extracted from data labels, real-time attack scores and client reputation, , , are , and weights, and .

[0012] Further, in the second layer hash processing, the lattice-based post-quantum digital signature algorithm realizes hash compression through the following steps: B1, performing lattice basis vector mapping on the intermediate hash value H 1 to generate a compressed multi-dimensional lattice vector; B2, performing signature operation on the multi-dimensional lattice vector according to a pre-defined private key matrix to generate a final hash value H_final.

[0013] Further, in the first layer hash processing, the output length of the hash algorithm is fixed at 256 bits, and the splicing method of the intermediate hash value H1 is to divide the input data into fixed blocks, and then generate the hash value H1 through the sponge structure absorption and extrusion mechanism.

[0014] Further, the method includes a key management module for dynamically generating and distributing the public-private key pair required by the post-quantum digital signature algorithm, and the public key is verified for integrity through a hash tree structure.

[0015] Further, the method reserves multiple levels of hash slots.

[0016] The beneficial effects of the present application are: the present method combines the traditional hash algorithm with the post-quantum digital signature algorithm (PQC), improving the security, flexibility and efficiency of data encryption, and the specific technical effects are as follows.

[0017] 1. Enhanced quantum resistance, introducing the post-quantum digital signature algorithm (PQC) (such as Dilithium or Falcon), effectively improving the defense capability of the protocol against quantum computing attacks. Even if a quantum computer can break SHA-3 hash, the post-quantum digital signature algorithm (PQC) can still provide additional protection.

[0018] 2. Flexible configuration, automatically selecting single-layer hash mode or double-layer hash mode through a security evaluation engine, selecting single-layer hash (SHA-3) or double-layer hash (SHA-3 + post-quantum digital signature algorithm (PQC)) mode, improving computing efficiency in low security requirements, and enhancing quantum attack resistance in high security requirements.

[0019] 3. Improve data integrity and tamper resistance, double-layer hash mode ensures that data not only prevents quantum computing attacks, but also effectively prevents data tampering and enhances data protection.

[0020] 4. Adapt to different application scenarios, suitable for financial, cloud computing, big data storage and other fields, and can flexibly select hash mode to ensure security requirements in different scenarios.

[0021] 5. Compatible with existing encryption architecture, based on standard SHA-3 hash and post-quantum digital signature algorithm (PQC), compatible with existing encryption protocols, easy to integrate and promote.

[0022] 6. Prevent quantum computing from being broken, the post-quantum digital signature algorithm (PQC) resists the risk of being broken by a quantum computer, providing long-term security for future encryption needs. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 Flowchart for the method described in Example 1. DETAILED DESCRIPTION

[0024] The application will be further described below in connection with specific embodiments.

[0025] Embodiment 1 The embodiment discloses a data protection method against quantum attacks, which provides a configurable double-layer hash protocol for data protection in a quantum computer environment (with quantum attacks). The configurable double-layer hash protocol combines a post-quantum digital signature algorithm (PQC) (such as Dilithium or Falcon) as a second layer hash algorithm on the basis of a standard SHA-3 hash to enhance the defense capability against quantum computing attacks. The protocol not only effectively improves the security of data, but also has the characteristics of flexible configuration, and users can select appropriate hash modes according to different security requirements.

[0026] As shown in Figure 1 The implementation steps of the method include: A, first layer hash processing, using a hash algorithm (SHA-3) to perform hash operation on input data to generate an intermediate hash value H1. In this step, the SHA-3 hash algorithm divides the input data according to a fixed block, and generates a hash value H1 through a sponge structure absorption-extrusion mechanism. The output length of the SHA-3 hash algorithm is fixed at 256 bits, which ensures the consistency and unpredictability of the data.

[0027] B, second layer hash processing, performing hash compression on the intermediate hash value H1 of step A through a lattice-based post-quantum digital signature algorithm to generate a final hash value. In this step, the lattice-based post-quantum digital signature algorithm realizes hash compression through the following steps: B1, performing lattice basis vector mapping on the intermediate hash value H1 to generate a compressed multi-dimensional lattice vector; B2, performing signature operation on the multi-dimensional lattice vector according to a pre-defined private key matrix to generate a final hash value H_final.

[0028] The post-quantum digital signature algorithm (PQC) is based on lattice-based encryption technology, which has high resistance to quantum computing attacks. The algorithm effectively prevents quantum computers from being cracked by Shor algorithm through the introduction of complex mathematical structures (such as lattices, rings, etc.).

[0029] C, in order to analyze the sensitivity of input data, external attack risk, access request identity and other factors in real time, a security evaluation engine is set, and the hash mode is dynamically selected as a single-layer hash mode or a double-layer hash mode based on the security evaluation engine, a), single-layer mode, only first layer hash processing is performed, and H_final = H1 is outputted; b) In the two-layer mode, the first-layer hash processing and the second-layer hash processing are performed in sequence, and the output H_final = PQC_Sign(H1). PQC_Sign(.) is a lattice-based post-quantum digital signature algorithm.

[0030] D. Set up a key management module. The key management module is used to dynamically generate and distribute the public key-private key pairs required for the post-quantum digital signature algorithm. The public key is verified for integrity through a hash tree structure.

[0031] E. Reserve multi-level hash slots to support the integration of future new post-quantum algorithms (such as SPHINCS+ and Rainbow), with good scalability and foresight.

[0032] In this embodiment, the security assessment engine evaluates the security requirements of the data by analyzing multiple real-time indicators to determine the hash pattern. The real-time indicators analyzed include: Data label: A label that comes with the data. For example, financial data has label A, and medical data has label B. Label A and label B represent different security levels. The data label is used to determine whether the data is sensitive (such as financial data, medical data, etc.), thereby deciding whether security protection needs to be strengthened.

[0033] Real-time security risk scoring: Through intrusion scoring based on log analysis, the system dynamically assesses whether the current system is facing external attacks, thereby obtaining a real-time security risk score.

[0034] Credibility of client requests: The credibility of the access is determined based on information such as whether the source IP address of the access request is trusted and the user's reputation level, thereby obtaining the credibility of the client request.

[0035] System security status: Analyze the current security status of the system, such as whether it is in a high-risk operating state, to obtain the system security status.

[0036] Based on real-time security indicators, the system calculates a security requirement level (RiskLevel). The security requirement level has a set threshold (0.7 in this example). When RiskLevel ≥ 0.7, the dual-layer hash mode (SHA-3 + post-quantum digital signature algorithm (PQC)) is enabled to provide higher security; when RiskLevel < 0.7, the single-layer hash mode (using only SHA-3) is enabled to improve computing efficiency.

[0037] The embodiment adopts an automatic identification mechanism to evaluate the security requirements of data, reduces manual intervention through the automatic mechanism, and improves the efficiency and accuracy of the deployment of the hash protocol. The hash mode can be flexibly switched according to different security requirements, especially in dynamic environments such as cloud computing platforms and edge nodes, and has better adaptability. Different security threats and risks can be responded to in real time, and the protocol can continue to operate stably under changing security situations. This intelligent configuration method makes the double-layer hash protocol of the application have strong adaptability in various complex and variable environments, further enhancing the security and flexibility of the system.

[0038] In the embodiment, the security evaluation engine evaluates the real-time security requirement level in the following way: 1. Rule-based security evaluation, using a set of rules for scoring, each set of rules including multiple factors (such as data type, user identity, request source, etc.) and the weight of each factor. When a certain factor reaches a set threshold, its score in the security requirement level is automatically increased. When the security requirement level is less than the set threshold, a single-layer hash mode is used, and when the security requirement is greater than the set threshold, a double-layer hash mode is used. For example, due to different data types, the security requirement level of financial transaction data may be higher than that of ordinary text data.

[0039] 2. Machine learning model, using a trained classifier to evaluate the risk level, the classifier can use support vector machine (SVM), random forest (RF) or neural network (NN). First, train the machine learning model with historical data, and predict the security requirement level of the data according to the access behavior, data attributes, etc. When the security requirement level is less than the set threshold, a single-layer hash mode is used, and when the security requirement is greater than the set threshold, a double-layer hash mode is used. The machine learning model predicts the risk level of the data according to the following features: IP reputation: Evaluate the reputation of the IP address based on historical data.

[0040] Request frequency: Number of requests per unit time.

[0041] Data type: classified according to the sensitivity of the data (such as financial data, medical data, etc.).

[0042] Using machine learning models (such as random forests, support vector machines, etc.) to classify requests and predict their security risk levels.

[0043] For example, use historical attack samples to train a model that can evaluate the security risk of requests in real time and give a prediction score. The prediction score can be part of the RiskLevel.

[0044] 3. Deep learning model, using deep neural network (DNN) or convolutional neural network (CNN) to process complex input data such as network traffic data or file content. Deep learning model can automatically extract high-dimensional features from data to more efficiently and accurately determine the security requirement level. When the security requirement level is less than the set threshold, single-layer hash mode is used, and when the security requirement is greater than the set threshold, double-layer hash mode is used.

[0045] The process of predicting the security requirement level using the deep learning model is as follows: Training data: In order to train the evaluation engine, a set of historical data needs to be collected, including various types of attacks (such as intrusion detection logs, malicious IPs, low-reputation users, etc.) and the sensitivity of data (such as financial data, medical data, etc.).

[0046] Algorithm selection: Random forest or support vector machine (SVM) can be selected as the initial model, as they usually perform well in handling classification problems, especially when the data features are diverse.

[0047] Real-time evaluation: In actual application, when a new request enters, the evaluation engine will analyze the characteristics of the request in real time and calculate the RiskLevel, and then decide to enable single-layer hash or double-layer hash mode according to the RiskLevel value (such as 0.7 as the threshold).

[0048] This embodiment uses a weighted sum model to calculate the security requirement level, and the calculation formula is: ; Where represents the security requirement level, , and are features extracted from data labels, real-time attack scores, and client reputation, , , are the weights of , and , and .

[0049] In this embodiment, According to the sensitivity of data type and the experience evaluation of security field, usually, The value range of can be 0.4 to 0.6. Through historical attack data and attack impact evaluation, the weight of β can be set. Usually, the attack risk weight The value can be set at 0.3 to 0.5. The weight of the first layer can be set to 0.1 to 0.3.

[0050] For each input data item, its security requirement level is calculated, and a dynamic decision is made whether to enable the double-layer hash mode after real-time evaluation.

[0051] The configurable double-layer protocol described in this embodiment allows dynamic selection of single-layer hash (such as SHA-3) or double-layer hash (such as SHA-3 + post-quantum digital signature algorithm (PQC)) mode according to the security requirement level (RiskLevel) calculated by the security evaluation engine. In the low security requirement scenario, the single-layer hash mode is dynamically selected, and only the hash algorithm (such as SHA-3) is used for data processing. In the high security requirement scenario, the double-layer hash mode is dynamically enabled, and the post-quantum digital signature algorithm (PQC) provides additional protection.

[0052] The double-layer hash mode provides additional protection against quantum attacks. By introducing the post-quantum digital signature algorithm (PQC) after the SHA-3 hash layer, it can effectively deal with potential threats from quantum computers, especially Shor algorithm and Grover algorithm attacks. The post-quantum digital signature algorithm (PQC) (such as Dilithium or Falcon) uses lattice-based encryption technology, which has been widely recognized in current quantum computing theory research for its ability to resist quantum attacks, and can effectively resist attacks from quantum computers.

[0053] Embodiment 2 This embodiment discloses an anti-quantum attack data protection method, comprising the following steps: 1. Input data, assuming that the input data is M, which can be any data that needs to be processed for secure hash, such as file content, transaction data or message, etc.

[0054] 2. First layer SHA-3 hash processing, first, use SHA-3-256 hash algorithm to process input data M, generate intermediate hash value H1. SHA-3-256 uses sponge structure, output length is 256 bits. In other embodiments, SHA-3-224, SHA-3-512, SHA-3-384 algorithms can also be used, and the output length of the first layer hash processing is 224, 512 or 384 bits.

[0055] Formula: H1 = SHA3-256 (M); This process ensures the irreversibility of the input data D, preventing data leakage or tampering.

[0056] 3. Second layer post-quantum hash (post-quantum digital signature algorithm (PQC)), The hash value H1 generated by the first layer is compressed using a post-quantum digital signature algorithm (PQC) such as Dilithium or Falcon to generate the final hash value H final .

[0057] Formula: H final = PQC_Sign(H1); Post-quantum digital signature algorithm (PQC) such as Dilithium or Falcon is a lattice-based encryption technology that is naturally resistant to quantum computing attacks, providing additional protection to ensure that the hash result remains secure even under quantum computing threats.

[0058] 4. Configurable mode, during implementation, the system deploys a security assessment engine to analyze the sensitivity of current input data, external attack risk, access request identity, etc. in real time, and dynamically decides to use single-layer or double-layer hash mode. The assessment engine can use rule engine, strategy model or machine learning method for security level assessment. When RiskLevel < 0.7, the system selects only SHA-3-256 hash algorithm for single-layer hash processing: H final = H1 = SHA3-256(M); When RiskLevel ≥ 0.7, the system automatically selects to enable double-layer hash mode, first performs SHA-3-256 hash processing, and then combines post-quantum digital signature algorithm (PQC) for encryption to obtain the final hash value H final 。

[0059] H final = PQC_Sign(SHA3-256(M)).

[0060] Example 3 This embodiment discloses the application of the above method in specific fields, such as in the financial field. In financial transactions, when customers submit transaction data, the system uses double-layer hash protocol for data hashing to ensure the integrity of transaction data. In high-security demand transactions, double-layer hash mode is enabled, combined with post-quantum digital signature algorithm (PQC) to enhance protection against quantum computing attacks.

[0061] Cloud storage field, in cloud storage, when users upload files, the system performs hash processing on the file content. The system selects single-layer or double-layer hash mode according to file type and importance. For example, for storing sensitive data (such as medical records, financial data, etc.), double-layer hash mode is enabled to ensure data security.

[0062] In the field of blockchain, the hash calculation of block data and transaction data is the key to guarantee the security and tamper resistance of the system in blockchain applications. According to the security requirements of the block, the system can choose to use single-layer hash in ordinary transactions, and enable double-layer hash mode in high-value transactions or important data.

[0063] In order to verify the performance of the method, the inventors conducted experimental verification, and the experimental results are as follows: From the above experimental data, although the double-layer hash mode has a slight increase in time and memory consumption, the performance loss is acceptable (the delay increases by about 25 milliseconds), and in the scene with high security requirements, this sacrifice is worthwhile.

[0064] Anti-quantum attack experiment, in the experiment, the attack of the simulated quantum computer (using Shor algorithm to crack single-layer hash) is successfully carried out in the traditional single-layer hash protocol. In contrast, in the double-layer hash mode, the existence of the post-quantum digital signature algorithm (PQC) effectively resists the attack.

[0065] The above description is only the basic principle and preferred embodiment of the present application, and the improvements and substitutions made by those skilled in the art according to the present application belong to the protection scope of the present application.

Claims

1. A data protection method against quantum attacks, characterized by: include: A. The first layer of hash processing uses a hash algorithm to perform a hash operation on the input data and generate an intermediate hash value H1; B. The second layer of hash processing uses a lattice-based post-quantum digital signature algorithm to hash and compress the intermediate hash value H1 in step A to generate the final hash value; C. Dynamically select the hash mode as single-layer hash mode or double-layer hash mode based on the security assessment engine. a) Single-layer hash mode: only the first layer of hash processing is performed, and the output is H_final = H1; b) Double-layer hash mode, performing the first-layer hash processing and the second-layer hash processing in sequence, and the output is H_final = PQC_Sign(H1), where PQC_Sign(.) is a lattice-based post-quantum digital signature algorithm.

2. The data protection method against quantum attacks according to claim 1, characterized in that: The posture engine selects a hash mode based on one or more of the following parameters: Data labels, real-time security risk scores, reputation of client requests, and system security posture.

3. The data protection method against quantum attacks according to claim 1, characterized in that: The security assessment engine uses a rule-based security assessment method to evaluate the security requirement level and selects the hash mode based on the security requirement level: The rule-based security assessment method uses a set of rules for scoring. A set of rules includes multiple factors and the corresponding weights of each factor. When a factor reaches the set threshold, its score in the security requirement level is automatically increased. When the security requirement level is lower than the set threshold, a single-layer hash mode is used. When the security requirement level is higher than the set threshold, a double-layer hash mode is used.

4. The data protection method against quantum attacks according to claim 1, characterized in that: The security assessment engine uses a machine learning model to evaluate the security requirement level and selects a hash mode based on the security requirement level. The process of using a machine learning model to evaluate the security requirement level is as follows: a classifier is trained with historical attack samples. The trained classifier first extracts access data features from the input data. Access data features include IP reputation, request frequency, and data type. IP reputation is the reputation of the IP address of the access device. Request frequency is the number of requests per unit time. Data types are classified according to the sensitivity of the data. The security requirement level of the data is then predicted based on the access data features. When the security requirement level is lower than the set threshold, a single-layer hash mode is used. When the security requirement level is higher than the set threshold, a double-layer hash mode is used.

5. The data protection method against quantum attacks according to claim 1, characterized in that: The security assessment engine uses a deep learning model to evaluate the security requirement level and selects a hash mode based on the security requirement level. The process of using a deep learning model to evaluate the security requirement level is as follows: first, historical data is collected, which includes data on various attack types and data on various sensitivities. The deep learning model is trained and evaluated based on the historical data. The trained deep learning model performs feature analysis on newly incoming requests and calculates the security requirement level. When the security requirement level is lower than the set threshold, a single-layer hash mode is used. When the security requirement level is higher than the set threshold, a double-layer hash mode is used.

6. The data protection method against quantum attacks according to any one of claims 3 to 5, characterized in that: The safety requirement level uses a weighted sum model, which is expressed as follows: ； in Indicates the level of security requirements, 、 and It is the feature extracted from data labels, real-time attack scores, and client reputation. 、 、 for 、 and The weight of .

7. The data protection method against quantum attacks according to claim 1, characterized in that: In the second layer of hashing, the lattice-based post-quantum digital signature algorithm achieves hash compression through the following steps: B1, perform lattice basis vector mapping on the intermediate hash value H1 to generate a compressed multi-dimensional lattice vector; B2. Perform signature operation on the multidimensional lattice vector according to the predefined private key matrix to generate the final hash value.

8. The data protection method against quantum attacks according to claim 1, characterized in that: In the first layer of hash processing, the intermediate hash value H1 is spliced ​​by dividing the input data into fixed blocks, and then generating the hash value H1 through the sponge structure absorption and squeezing mechanism. The hash value H1 is a fixed length.

9. The data protection method against quantum attacks according to claim 1, characterized in that: This method includes a key management module, which is used to dynamically generate and distribute the public key-private key pair required by the post-quantum digital signature algorithm, and the public key is verified for integrity through a hash tree structure.

10. The data protection method against quantum attacks according to claim 1, characterized in that: This method reserves multiple levels of hash slots.

Images