Market supervision data security protection method and system based on dynamic encryption strategy

Abstract

The invention discloses a market supervision data security protection method and system based on a dynamic encryption strategy, and relates to the technical field of information security. The method comprises the following steps: generating an initial seed input improved four-dimensional hyperchaotic system, and generating a dynamic encryption parameter set through chaotic mapping iteration; the method comprises the following steps: carrying out adaptive fragmentation on market supervision data to be encrypted to generate a data sub-block set, carrying out dynamic S-box-based symmetric encryption on odd index sub-blocks, carrying out a chaotic stream encryption mode associated with a current dynamic sub-key on even index sub-blocks, and generating a unique quantum chaotic fingerprint for each encrypted sub-block; a searchable encryption index is jointly constructed by using a post-quantum encryption algorithm and a lattice-based encryption algorithm and is used for carrying out query operation under the condition that data is not decrypted; the encryption security, privacy protection capability and quantum computing attack resistance of the market supervision data are remarkably improved, and meanwhile, an efficient, flexible and extensible data security protection scheme is provided.

Description

Technical Field

[0001] The present invention relates to the field of information security technology, and particularly to a method and system for market supervision data security protection based on a dynamic encryption policy. Background Art

[0002] The market supervision data exchange platform bears important business data of market supervision departments at all levels and related institutions. These data include enterprise registration information, supervision and law enforcement records, complaint and reporting content, and credit evaluation data, etc., which have high sensitivity and confidentiality requirements. However, most of the existing platforms adopt traditional static encryption and access control means, and still have the following deficiencies in the face of increasingly complex network security threats:

[0003] There is a risk of static encryption being broken: Currently, a large number of systems use fixed keys and fixed algorithms (such as symmetric encryption with static S-boxes) to protect data. If an attacker monitors or obtains ciphertext for a long time, the unchanged encryption mechanism may be analyzed and cracked. Some studies have shown that introducing dynamically changing encryption elements can significantly improve security.

[0004] Key management issues: The key management of encryption algorithms is still a challenge, especially in large-scale systems, where the generation, distribution, and storage of keys are vulnerable to attacks. Traditional methods lack flexible key update and automated management mechanisms, resulting in the encryption protection of data not being able to adapt to potential security threats in real time.

[0005] Privacy protection issues: With the continuous development of quantum computing and advanced computing capabilities, traditional encryption methods can no longer meet the needs of future information protection. Especially in the face of possible quantum computing attacks, existing encryption algorithms show weak anti-quantum attack capabilities and are difficult to effectively protect data privacy.

[0006] Conflict between query and data protection: After traditional encryption technology encrypts data, it often cannot directly perform efficient queries on the encrypted data, or some parts of the data need to be exposed during querying. Although existing searchable encryption technologies provide the ability to query encrypted data, their security and efficiency are still restricted, and their anti-quantum attack capabilities have not been effectively verified. Summary of the Invention

[0007] In view of the above problems, the present invention proposes a method and system for market supervision data security protection based on a dynamic encryption policy. By introducing technical means such as a dynamic encryption parameter generation mechanism, a quantum random number generator, chaotic mapping iteration, and a dynamic key update mechanism, it aims to provide a higher level of security protection, while taking into account the usability and performance of the system, and meeting the dual requirements of market supervision business for data security and data utilization.

[0008] The present invention realizes the above object through the following technical solutions:

[0009] A method for securing market supervision data based on a dynamic encryption strategy, the method comprising:

[0010] Generating an initial seed through a quantum random number generator and inputting it into an improved four-dimensional hyperchaotic system. Through chaotic mapping iteration, a set of dynamic encryption parameters is generated, including dynamic sub-keys and a dynamic S-box. At the same time, the entropy value change of the set of dynamic encryption parameters is monitored in real time. When the entropy value drops below a set threshold, the initial seed is automatically refreshed and a new set of dynamic encryption parameters is iteratively generated;

[0011] Adapting to fragment the market supervision data to be encrypted based on the set of dynamic encryption parameters to generate a set of data sub-blocks. Symmetric encryption based on the dynamic S-box is used for the odd-index sub-blocks in the set of data sub-blocks, and a chaotic stream encryption method associated with the current dynamic sub-key is used for the even-index sub-blocks. A unique quantum chaotic fingerprint is generated for each encrypted sub-block to verify the integrity and uniqueness of the encrypted data. A redundancy backup strategy is set according to the importance level of each encrypted sub-block;

[0012] For the query keywords submitted by the user, a searchable encryption index is jointly constructed using a post-quantum encryption algorithm and a lattice-based encryption algorithm for query operations without decrypting the data. After the client submits a zero-knowledge proof for the server to verify the query legality, the server performs index retrieval based on chaotic topological rules and returns a set of encrypted result identifiers after removing duplicates.

[0013] Preferably, the method for generating a set of dynamic encryption parameters through chaotic mapping iteration comprises:

[0014] Generating an initial seed Q_Seed using a multi-source quantum random number generator that combines a photon source and a charge source;

[0015] Generating multiple initial vectors of the chaotic system by hashing and expanding the initial seed Q_Seed, performing invertible expansion based on a Gaussian random matrix mapping, and performing sensitive perturbation in combination with the platform master key;

[0016] During the perturbation process, a non-linear permutation masking mechanism is introduced to perform secondary confusion on the expanded initial vectors;

[0017] Evolving the initial vectors according to the following four-dimensional hyperchaotic mapping iteration relationship:

[0018]

[0019] Wherein, X(n), Y(n), Z(n), and W(n) are the four variables of the improved four-dimensional hyperchaotic system, and X(n+1), Y(n+1), Z(n+1), and W(n+1) are the updated variables; a, b, c, and d are adjustment parameters; Φ is a quantum gate control perturbation function; represents a combination operation; represents an exclusive OR operation; H(·) is a lightweight hash compression function;

[0020] After a preset number of iteration steps, sample each chaotic orbit to generate a dynamic sub-key group and a dynamic S-box mapping table, constituting a dynamic encryption parameter set P(t).

[0021] Preferably, when the entropy value drops below a set threshold, automatically refresh the initial seed and re-iterate to generate a new dynamic encryption parameter set. The method includes:

[0022] Perform a joint evaluation of the multi-dimensional entropy values of Shannon entropy H Shannon , Renyi entropy H Renyi , and sample entropy H Sample for the dynamically generated encryption parameter set P(t) in each round, and construct a comprehensive entropy value H total . The formula is:

[0023] H total = w1·H Shannon + w2·H Renyi + w3·Norm(H Sample );

[0024] Wherein, w1, w2, and w3 are the entropy weights of each item, and w1 + w2 + w3 = 1; Norm(H Sample ) represents sample entropy normalization;

[0025] Set a multi-level entropy value response mechanism, including:

[0026] Safety interval: The comprehensive entropy value H total is higher than a preset first threshold, and the current parameters remain unchanged;

[0027] Warning interval: The comprehensive entropy value H total is between a preset first threshold and a preset second threshold, and dynamically adjust the adjustment parameters a, b, c, and d of the chaotic mapping;

[0028] Danger interval: The comprehensive entropy value H total is lower than a preset second threshold, triggering an initial seed Q_Seed refresh action;

[0029] Before the system executes the refresh action, dynamically adjust the refresh frequency and the preset number of iteration steps according to the load information of the current platform's concurrent access quantity and data traffic;

[0030] When a refresh is triggered, the quantum random number generator is called again to generate a new initial seed, and a random segment is selected from the dynamic encryption parameter set P(t) whose entropy value was in the safe range in the previous round as the perturbation factor, and combined with the new initial seed to generate the initial vector;

[0031] The selection of the refresh strategy is optimized based on a lightweight reinforcement learning model, and the lightweight reinforcement learning model decides whether to adopt the re-hashing, gene migration or full restart mode according to the historical entropy value trend and the refresh effect feedback;

[0032] Perform chaotic iterative evolution on the new initial vector, and perform entropy value verification on the generated results for multiple consecutive rounds. If the entropy values of two consecutive rounds are both lower than the safe range, execute the enhanced restart mechanism.

[0033] Preferably, the enhanced restart mechanism specifically includes:

[0034] Discard all current states;

[0035] Call the quantum random number generator again to generate a new initial seed;

[0036] Reset the initial vector of the chaotic mapping;

[0037] Increase the initial number of iteration steps;

[0038] Switch to the backup chaotic mapping system;

[0039] Until the comprehensive entropy value of the dynamic encryption parameter set P(t) returns to the safe range.

[0040] Preferably, the method for adaptively fragmenting the market supervision data to be encrypted based on the dynamic encryption parameter set includes:

[0041] According to the comprehensive entropy value H extracted from the dynamic encryption parameter set P(t) total , determine the number of fragments N of the market supervision data D, and the formula is:

[0042]

[0043] In the formula, DataSize(D) represents the size of the market supervision data D to be encrypted;

[0044] Generate a pseudo-random mask sequence Mask(t) based on the dynamic encryption parameter set P(t) for defining the splitting positions of the market supervision data D;

[0045] Using the pseudo-random mask sequence Mask(t), split the market supervision data D to be encrypted into a set of data sub-blocks {D i} according to the mask change points, and the boundary of each data block is determined by the mask value;

[0046] After fragmentation, according to the chaotic trajectory π(t), the logical order of the data sub-block set {D i} is perturbed; specifically, the chaotic perturbation permutation function π is used to disrupt the physical order of the data sub-blocks and generate a new logical order {D π(i)}.

[0047] Preferably, after encrypting the data sub-blocks, combined with the previous encryption historical data, based on the multi-hash mechanism, a quantum chaotic fingerprint based on the historical encryption chain is generated for each sub-block;

[0048] And calculate the entropy density for each data sub-block, evaluate the importance of each data sub-block according to the entropy density, and divide it into three levels: high, medium, and low;

[0049] According to the importance level of the data sub-blocks, set the corresponding redundancy backup strategy, specifically:

[0050] High level: For critical and sensitive data, adopt a triple-copy storage strategy, and at the same time superimpose an erasure code mechanism based on exclusive-or operation. Each copy is stored on different physical nodes respectively, and redundant check blocks are generated through exclusive-or encoding to ensure fast recovery through the remaining copies and check data in case of a single-node failure;

[0051] Medium level: Adopt a dual-copy storage mechanism and attach a timestamp chain structure. A hash association with increasing timestamps is established between the two copies to form an immutable time sequence record. Any modification of the copy will trigger a chain verification to prevent data tampering and historical version forgery;

[0052] Low level: Execute single-copy storage, and only use a lightweight hash algorithm to generate data fingerprints, and implement basic integrity verification through hash values with low computational overhead.

[0053] Preferably, the post-quantum encryption algorithm and the lattice-based encryption algorithm are jointly used to construct a searchable encryption index. The method includes:

[0054] Use the post-quantum encryption algorithm to encrypt the query keyword Q to obtain the encrypted query keyword E PQE (Q);

[0055] Use the lattice-based encryption algorithm to encrypt the query keyword Q to obtain the encrypted query keyword E LWE (Q);

[0056] Merge the encrypted query keywords to generate a searchable encryption query index Index(Q), and the expression is:

[0057]

[0058] Wherein, P(t) is a set of dynamic encryption parameters;

[0059] The client submits a query request to the server through the searchable encryption query index Index(Q).

[0060] Preferably, after the client submits a zero-knowledge proof for the server to verify the query legality, the server performs index retrieval based on the chaotic topology rule and returns an encrypted result identifier set after removing duplicates. The method includes:

[0061] After receiving the searchable encryption query index Index(Q) and the zero-knowledge proof ZKP(Q) submitted by the client, the server verifies the query legality;

[0062] The server generates a chaotic topology rule T chaos (t) based on the current set of dynamic encryption parameters P(t), and performs encrypted index retrieval according to the chaotic topology rule T chaos (t) to query the encrypted data set;

[0063] When the server returns the query result, it applies an intelligent deduplication mechanism, removes duplicate encrypted result identifiers through similarity evaluation, and dynamically adjusts the returned content through fuzzy query technology to ensure a high match between the query result and the user's needs, while avoiding revealing the data scale;

[0064] The client decrypts the deduplicated query result and verifies the integrity of the query result through quantum chaotic fingerprints to ensure that the query result is consistent with the historical version of the encrypted data.

[0065] Preferably, the method further includes privacy-preserving computing and dynamic access control and authentication;

[0066] Among them, the privacy-preserving computing is specifically:

[0067] Apply hierarchical functional encryption processing to the encrypted data sub-blocks, and introduce chaotic differential privacy noise controlled by the dynamic encryption parameter set P(t) during the federated learning training process to update the global model weights. The formula is as follows:

[0068]

[0069] In the formula, W t+1 is the weight of the global model at t + 1, and W t is the weight of the global model at t; η represents the learning rate; is the gradient of the loss function; λ represents the control factor of the differential privacy noise; ChaosNoise(P(t)) represents the chaotic differential privacy noise generated based on the dynamic encryption parameter set P(t);

[0070] The dynamic access control and authentication are specifically:

[0071] Dynamically adjust the access policy by combining the user access behavior pattern, query frequency, and user role, and perform authentication using a post-quantum secure authentication protocol;

[0072] Combining multi-factor authentication with a post-quantum authentication protocol to provide a multi-level authentication mechanism, ensuring a more secure authentication process. During the communication process, use the post-quantum authentication protocol to verify the user's identity, ensuring the reliability of the user's identity and the ability to resist quantum attacks;

[0073] The main key management and rotation are executed by a hardware security module HSM based on a chaos-enhanced mechanism, which hierarchically stores and periodically replaces the chaotic main key and functional encryption keys to ensure the security of the keys and the randomness of updates;

[0074] Generate log records for all operations involving key generation, data encryption and decryption, query, and analysis;

[0075] Regularly check the logs to detect abnormal behaviors, including abnormally frequent decryption requests and continuous large numbers of query hits, and give timely warnings.

[0076] A market supervision data security protection system based on a dynamic encryption policy, the system includes:

[0077] A dynamic encryption parameter generation module, used to generate an initial seed and input it into an improved four-dimensional hyperchaotic system, and generate a set of dynamic encryption parameters through chaotic mapping iteration;

[0078] An entropy value monitoring module, used to monitor the entropy value change of the set of dynamic encryption parameters in real time. When the entropy value drops below the set threshold, automatically refresh the initial seed and re-iterate to generate a new set of dynamic encryption parameters;

[0079] A dynamic encryption module, used to adaptively fragment the market supervision data to be encrypted based on the set of dynamic encryption parameters, generate a set of data sub-blocks, and use symmetric encryption based on a dynamic S-box for the odd-index sub-blocks in the set of data sub-blocks, and use a chaotic stream encryption method associated with the current dynamic sub-key for the even-index sub-blocks. At the same time, generate a quantum chaotic fingerprint for each encrypted sub-block to verify the integrity and uniqueness of the encrypted data;

[0080] An encrypted query index construction module, used to jointly construct a searchable encrypted index using a post-quantum encryption algorithm and a lattice-based encryption algorithm, and after submitting a zero-knowledge proof by the client to the server to verify the query legality, perform index retrieval based on chaotic topological rules and return a set of encrypted result identifiers after removing duplicates;

[0081] The privacy protection computing module is used to apply hierarchical functional encryption processing to encrypted data sub - blocks, introduce chaotic differential privacy noise controlled by a dynamic encryption parameter set during the federated learning training process, update the global model weights, and adjust the noise intensity according to the differential privacy noise control factor;

[0082] The dynamic access control and authentication module is used to dynamically adjust the access policy by combining the user access behavior pattern, query frequency, and user role, perform identity authentication using a post - quantum secure authentication protocol, and provide a multi - factor authentication - based multi - level identity authentication mechanism to ensure a more secure identity authentication process;

[0083] The key management module is used to perform master key management and rotation based on the hardware security module HSM with a chaotic enhancement mechanism, hierarchically store and periodically replace the chaotic master key and functional encryption keys to ensure the security of the keys and the randomness of updates;

[0084] The log recording and security auditing module is used to record logs for all operations related to key generation, data encryption and decryption, query, and analysis, and regularly check the logs to detect abnormal behaviors, including abnormally frequent decryption requests and continuous large - number query hits, and give alarms in a timely manner;

[0085] The data storage and backup module is used to set redundancy backup strategies according to the importance level of data sub - blocks, including triple - copy storage, double - copy storage, and single - copy storage mechanisms.

[0086] The beneficial effects of the present invention are as follows: By dynamically generating an encryption parameter set, it avoids the security vulnerabilities brought by the fixed encryption mode in traditional encryption algorithms, improves the unpredictability and anti - attack ability of the encryption process. Especially in the face of quantum computing threats, it can effectively prevent the encryption system from being cracked; introducing chaotic differential privacy noise controlled by a chaotic map during the encryption process effectively enhances the privacy protection of data during processing and analysis; using a searchable encryption index jointly constructed by a post - quantum encryption algorithm and a lattice - based encryption algorithm to achieve efficient query operations without decrypting the data. Compared with traditional encrypted query methods, the innovative chaotic topology rules and deduplication query mechanism greatly improve the efficiency and privacy protection ability of the query process. When querying encrypted data, it can effectively prevent information leakage, and at the same time improve the accuracy and relevance of query results; hierarchically store and periodically rotate keys through the hardware security module HSM with a chaotic enhancement mechanism. Combining the dynamic characteristics of the chaotic system, the key update process is more flexible and random, improving the anti - attack ability of the keys and avoiding the risk that the keys in traditional key management schemes are too fixed and easy to be cracked. In addition, the present invention further improves the overall security and monitorability of the system through log recording and security auditing of key generation, data encryption and decryption, and query operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0088] Among them:

[0089] Figure 1 is the method flowchart in the embodiments of the present invention;

[0090] Figure 2 is the system structure block diagram in the embodiments of the present invention. Specific embodiments

[0091] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions of the embodiments of the present invention in conjunction with the drawings of the embodiments of the present invention. Obviously, the described embodiments are some, rather than all, of the embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present invention.

[0092] As Figure 1 shown, it is an embodiment of the present invention. This embodiment provides a method for protecting the security of market supervision data based on a dynamic encryption strategy, including the following:

[0093] S1: Quantum-enhanced chaotic parameter generation

[0094] Generate an initial seed through a quantum random number generator and input it into an improved four-dimensional hyperchaotic system. After chaotic mapping iteration, a dynamic encryption parameter set is generated, including a dynamic sub-key and a dynamic S-box. At the same time, the entropy value change of the dynamic encryption parameter set is monitored in real time. When the entropy drops below a set threshold (such as 7.8 bit / byte), the initial seed is automatically refreshed and a new dynamic encryption parameter set is generated by re-iteration.

[0095] In one of the embodiments, after chaotic mapping iteration to generate a dynamic encryption parameter set, the method includes:

[0096] SA1: Generate an initial seed Q_Seed using a multi-source quantum random number generator that combines a photon source and a charge source. The photon source is responsible for providing high-speed and unpredictable physical random events, and the charge source provides an enhanced entropy basis based on thermal noise fluctuations. The combination of the two can enhance the anti-predictability and true randomness of seed generation;

[0097] SA2: Generate multiple initial vectors of the chaotic system from the initial seed Q_Seed through hash expansion (such as Keccak (SHA-3)), and perform invertible expansion based on the Gaussian random matrix mapping to distribute them in the high-dimensional state space, improving the distribution uniformity and untraceability of the initial state. The expanded initial value is further combined with the platform master key K0 for perturbation (such as XOR or cross mapping) to bind the key sensitivity;

[0098] SA3: During the perturbation process, introduce a non-linear permutation mask mechanism, including operations such as non-linear permutation, bit reversal, and non-uniform bit perturbation of the vector, simulate multiple non-linear path perturbations, and perform secondary confusion on the expanded initial vector to improve the complexity of the initial state of the system;

[0099] SA4: Evolve the initial vector according to the following four-dimensional hyperchaotic mapping iteration relationship:

[0100]

[0101] In the formula, X(n), Y(n), Z(n), W(n) are the four variables of the improved four-dimensional hyperchaotic system, and +1X(n+1), Y(n+1), Z(n+1), W(n+1) are the updated variables; a, b, c, d are adjustment parameters; Φ is the quantum gate control perturbation function; Denotes the combination operation; Denotes the XOR operation; H(·) is a lightweight hash compression function;

[0102] SA5: After completing the preset number of iteration steps (such as 10,000 rounds), sample each chaotic orbit to generate a dynamic sub-key group and a dynamic S-box mapping table (sample the output chaotic value sequence at a set interval from the chaotic orbit, sort the sequence according to the value size and re-number it, use it to construct an 8-bit input → 8-bit output dynamic S-box, and at the same time select specific bits in the sequence to construct the dynamic sub-keys required for multiple rounds of encryption), forming a dynamic encryption parameter set P(t) for subsequent heterogeneous encryption operations on data sub-blocks.

[0103] Furthermore, according to the real-time entropy value change of the dynamic encryption parameter set, dynamically adjust the adjustment parameters a, b, c, d of the chaotic system to make the system evolve along a higher chaos degree evolution path when the entropy decreases; during the dynamic S-box generation process, a local perturbation optimization strategy can also be introduced to improve the non-linearity and differential resistance characteristics of the final substitution box by perturbing the order and permutation method of the sampled values.

[0104] In this embodiment, through a three-layer hybrid mechanism of multi-source quantum entropy input + Gaussian mapping expansion + non-linear perturbation, the sensitivity and irreversibility of the chaotic system to the initial value are greatly enhanced; the four-dimensional coupled hyperchaotic mapping structure has a higher Lyapunov exponent, which improves the uncertainty of the sequence; the real-time adjustable parameters and state entropy monitoring mechanism enable the generation process to have an adaptive "self-healing" ability, solving the risk of weak entropy degradation in static systems; the generated S-box and sub-key structure are based on time-series dynamic output, effectively resisting linear and differential attacks, and at the same time supporting the replacement of high-strength symmetric encryption algorithms (such as SM4, AES, etc.).

[0105] The core purpose of comprehensively judging multiple entropy values is: to judge both the randomness quality (Shannon), identify the frequency skew (Renyi), and avoid being deceived by pseudo-random structures (Sample Entropy). Therefore, instead of simply averaging, a decision-making model or a fusion rule judgment model based on a weighted confidence mechanism should be used. The example method is as follows:

[0106] Define the minimum qualified value (such as bit / byte) for each entropy:

[0107] H Shannon ≥7.85, H Renyi ≥7.8, H Sample ≥0.7 (unit is the output after standardization)

[0108] Define the judgment rules:

[0109] All satisfied → high entropy judgment, enter the safe interval;

[0110] Any one falling into the sub-optimal interval (such as H Renyi ∈[7.5, 7.8]) → enter the warning interval;

[0111] Any one seriously lower than the threshold → low entropy judgment, trigger refresh.

[0112] This embodiment preferably adopts a weighted entropy fusion scoring model. In one embodiment, when the entropy value drops below the set threshold, the initial seed is automatically refreshed and a new set of dynamic encryption parameters is generated by re-iteration. The method includes:

[0113] SB1: Conduct a joint evaluation of the multi-dimensional entropy values of Shannon entropy H Shannon , Renyi entropy H Renyi and sample entropy H Sample for the set of dynamic encryption parameters P(t) generated in each round, and construct a comprehensive entropy value H total . The formula is:

[0114] H total = w1·H Shannon + w2·H Renyi+ w3·Norm(H Sample );

[0115] wherein, w1, w2, and w3 are the entropy weights of each, and w1 + w2 + w3 = 1 (the default suggestion is w1 = 0.4, w2 = 0.4, w3 = 0.2); Norm(H Sample ) represents the normalization of sample entropy (such as mapping to the interval [0, 8]);

[0116] SB2: Set a multi-level entropy value response mechanism, including:

[0117] Safety interval: The comprehensive entropy value H total is higher than the preset first threshold (such as H total > 7.8), and the current parameters remain unchanged;

[0118] Warning interval: The comprehensive entropy value H total is between the preset first threshold and the preset second threshold (such as H total ∈ [7.3, 7.8]), and dynamically adjust the adjustment parameters a, b, c, and d of the chaotic mapping;

[0119] Danger interval: The comprehensive entropy value H total is lower than the preset second threshold (that is, H total < 7.3), and trigger the initial seed Q_Seed refresh action;

[0120] SB3: Before the system executes the refresh action, dynamically adjust the refresh frequency and the preset number of iteration steps according to the load information such as the current platform concurrent access quantity and data traffic (optimization adjustment before refreshing, controlling the refresh overhead and time window) to ensure that the refresh behavior achieves a balance between security and performance;

[0121] SB4: When the refresh is triggered, re - call the quantum random number generator to generate a new initial seed, and select a random segment from the dynamic encryption parameter set P(t) whose entropy value was in the safe interval in the previous round as the perturbation factor, and combine it with the new initial seed to generate an initial vector (generate new initial conditions when refreshing, preparing to enter a new round of evolution);

[0122] SB5: The selection of the refresh strategy is optimized according to a lightweight reinforcement learning model. The lightweight reinforcement learning model decides whether to adopt the re - hashing, gene migration, or full restart mode based on the historical entropy value trend and the refresh effect feedback (after the new seed is generated, the reinforcement learning model decides which strategy to use to refine the initial state adjustment) to reduce mis - triggering and improve efficiency;

[0123] SB6: Perform chaotic iterative evolution on the new initial vector, and continuously verify the entropy value of the generated result for multiple rounds. If the entropy values of two consecutive rounds are both lower than the safe interval, execute the enhanced restart mechanism (if the effect is still not good after refreshing, start a complete enhanced restart). This mechanism ensures that the generated dynamic encryption parameter set P(t) always maintains a high level of statistical randomness and unpredictability.

[0124] Furthermore, the enhanced restart mechanism specifically includes:

[0125] Discard all current states;

[0126] Re - call the quantum random number generator to generate a new initial seed;

[0127] Reset the initial vector of the chaotic map;

[0128] Increase the initial iteration steps;

[0129] Switch to the alternative chaotic map system;

[0130] Until the comprehensive entropy value of the dynamic encryption parameter set P(t) returns to the safe interval.

[0131] S2: Dynamic fractal encryption storage

[0132] Based on the dynamic encryption parameter set, adaptively slice the market supervision data to be encrypted to generate a set of data sub - blocks. Use symmetric encryption based on the dynamic S - box for the odd - indexed sub - blocks in the set of data sub - blocks, and use the chaotic stream encryption method associated with the current dynamic sub - key for the even - indexed sub - blocks. And generate a unique quantum chaotic fingerprint for each encrypted sub - block to verify the integrity and uniqueness of the encrypted data; Set the redundancy backup strategy according to the importance level of each encrypted sub - block.

[0133] Step S2 specifically includes:

[0134] S21: Determine the number of slices N of the market supervision data D according to the comprehensive entropy value H extracted from the dynamic encryption parameter set P(t). The formula is: total , where DataSize(D) represents the size of the market supervision data D to be encrypted;

[0135]

[0136] In the formula, DataSize(D) represents the size of the market supervision data D to be encrypted;

[0137] S22: Generate a pseudo - random mask sequence Mask(t) based on the dynamic encryption parameter set P(t) to define the splitting positions of the market supervision data D;

[0138] S23: Using the pseudo-random mask sequence Mask(t), the market supervision data D to be encrypted is segmented into a set of data sub-blocks {D i} according to the mask change points, and the boundary of each data block is determined by the mask value;

[0139] S24: After segmentation, according to the chaotic trajectory π(t), the logical order of the set of data sub-blocks {D i} is perturbed; specifically, the chaotic perturbation permutation function π is used to disrupt the physical order of the data sub-blocks, generating a new logical order {D π(i)}, ensuring that the physical storage order of the data sub-blocks is not consistent with their logical order, thereby increasing the anti-analysis ability of the encrypted data;

[0140] S25: For the set of data sub-blocks after adding perturbation, the odd-index sub-blocks are encrypted symmetrically based on the dynamic S-box (such as SM4, AES), and the even-index sub-blocks are encrypted by the chaotic stream encryption method associated with the current dynamic sub-key (such as XOR encryption);

[0141] S26: Combining the previous encrypted historical data, based on the multiple hashing mechanism, a quantum chaotic fingerprint based on the historical encryption chain is generated for each sub-block to prevent data tampering and forgery;

[0142] S27: Calculate the entropy density for each data sub-block, and evaluate the importance of each data sub-block according to the entropy density, and divide it into three levels: high, medium, and low;

[0143] S28: Set the corresponding redundancy backup strategy according to the importance level of the data sub-blocks, specifically:

[0144] High level: For critical and sensitive data (such as enterprise finance, market analysis data), a triple-copy storage strategy is adopted, and at the same time, an erasure code mechanism based on XOR operation is superimposed. Each copy is stored on different physical nodes, and redundant check blocks are generated through XOR encoding to ensure quick recovery through the remaining copies and check data in case of a single-node failure;

[0145] Medium level: For moderately sensitive data (such as commodity quality data, consumer feedback), a dual-copy storage mechanism is adopted, and a timestamp chain structure is attached. A hash association with increasing timestamps is established between the two copies to form an immutable time series record. Any modification of the copy will trigger chain verification to prevent data tampering and historical version forgery;

[0146] Low level: For low-sensitive data (such as statistical data, public reports), single-copy storage is performed, and only a lightweight hashing algorithm is used to generate data fingerprints, and basic integrity verification is achieved through hash values with low computational overhead (such as xxHash or BLAKE3), reducing storage and computational resource consumption while ensuring basic protection capabilities.

[0147] S3: Quantum-resistant Searchable Encryption Query

[0148] For the query keywords submitted by the user, a searchable encryption index is jointly constructed using a post-quantum encryption algorithm and a lattice-based encryption algorithm for query operations without decrypting the data. After the client submits a zero-knowledge proof for the server to verify the query legality, the server performs index retrieval based on the chaotic topology rules and returns an encrypted result identifier set after removing duplicates.

[0149] Specifically, the implementation method of step S3 includes:

[0150] S31: The client user inputs the query keyword Q through the application interface. For example: "Enterprise registration date" or "Market product quality inspection report". These query keywords need to maintain privacy during transmission and should not be exposed to the server or other third parties. Therefore, in this embodiment, the post-quantum encryption algorithm is used to encrypt the query keyword Q to obtain the encrypted query keyword E PQE (Q), and the lattice-based encryption algorithm is used to encrypt the query keyword Q to obtain the encrypted query keyword E LWE (Q). Then, the encrypted query keywords are merged to generate a searchable encryption query index Index(Q), and the expression is:

[0151] Index(Q) = [E PQE (Q) ⊕ E LWE (Q)] ⊕ P(t);

[0152] In the formula, P(t) is a set of dynamic encryption parameters to enhance the security of the query index and ensure consistency with the dynamic characteristics of the encryption system;

[0153] The client submits a query request to the server through the searchable encryption query index Index(Q);

[0154] S32: To verify the query legality, the client uses the zero-knowledge proof (ZKP) protocol to generate a zero-knowledge proof ZKP(Q) for the corresponding query keyword Q. After the server receives the searchable encryption query index Index(Q) and the zero-knowledge proof ZKP(Q) submitted by the client, it verifies the query legality. The verification process ensures that the query content conforms to the system rules without exposing the actual query content;

[0155] S33: The server generates a chaotic topology rule T chaos (t) based on the current set of dynamic encryption parameters P(t). This rule consists of a sequence generated by a chaotic system and is used to guide the encrypted index retrieval operation of the query. According to the chaotic topology rule T chaos(t) Perform encrypted index retrieval to query the encrypted data set; during the query process, the server does not decrypt any data and only matches the encrypted query index with the encrypted database data;

[0156] S34: When the server returns the query results, it applies an intelligent deduplication mechanism to remove duplicate encrypted result identifiers through similarity evaluation (such as hash value matching, cosine similarity, etc.), ensuring that the returned query results are unique and relevant, and dynamically adjusts the returned content through fuzzy query technology, so that the returned results are not only exact match items but also can include approximate match items with a relatively high degree of relevance, increasing the flexibility of the query, ensuring a high degree of match between the query results and the user's needs, and at the same time avoiding revealing the data scale;

[0157] S35: The server generates an encrypted identifier set from the deduplicated query result identifiers. These identifiers do not contain any actual data but are encrypted identifiers to ensure data privacy. The encrypted result identifier set is returned to the client through a secure encrypted channel. The client decrypts the deduplicated query results to restore the data that meets the query conditions, and verifies the integrity of the query results through quantum chaotic fingerprints to ensure that the query results are consistent with the historical versions of the encrypted data. If the data is tampered with or inconsistent, the client will receive a warning message and abort the operation.

[0158] S4: Privacy-preserving computing

[0159] The system divides the original market supervision data into blocks, and each data sub-block set {D i} undergoes hierarchical functional encryption processing, and different encryption layers adopt different encryption strategies to ensure data privacy and operability. For example, symmetric encryption algorithms can be used to encrypt some sub-blocks, while asymmetric encryption and other strategies can be adopted for other sub-blocks to ensure multi-level data protection.

[0160] For certain specific data types (such as numerical data), SM4 or AES encryption can be selected; while for sensitive information (such as user identity information), encryption methods based on public key infrastructure are adopted to ensure privacy protection at different levels.

[0161] During the federated learning training process, chaotic differential privacy noise controlled by a dynamic encryption parameter set P(t) is introduced to update the global model weights. The formula is as follows:

[0162]

[0163] In the formula, W t+1 is the weight of the global model at t + 1 (or model parameters, that is, the updated model weights), W tis the weight of the global model at time \(t\) (i.e., the current model weight); \(\eta\) represents the learning rate, which is a constant used to control the step size in each update and usually ranges between 0 and 1; is the gradient of the loss function, which is used to measure the prediction error of the model under the current weight and guide the model on how to adjust the parameters to reduce the error; \(\lambda\) represents the control factor of the differential privacy noise, which is used to adjust the intensity of the privacy noise introduced by the chaotic differential privacy noise;

[0164] ChaosNoise(P(t)) represents the chaotic differential privacy noise generated based on the dynamic encryption parameter set P(t). This noise has unpredictability and time-variability, aiming to enhance the privacy protection ability of the model. By introducing chaotic noise, data leakage is prevented;

[0165] In each round of training, by using the learning rate \(\eta\) and the gradient of the loss function the weight of the model is updated, and at the same time, the noise \(\lambda\cdot ChaosNoise(P(t))\) based on chaotic differential privacy is added to ensure privacy protection. The scheduling of the noise intensity is responsible for the chaotic scheduler. According to different privacy requirements and changes in the attack model, the noise intensity is dynamically adjusted to optimize the privacy protection effect and improve the efficiency of the training process.

[0166] S5: Dynamic Access Control and Authentication

[0167] Combining the user access behavior pattern, query frequency, and user role, the access policy is dynamically adjusted, and a post-quantum secure authentication protocol is used for authentication. This is the premise and entry point of the access control mechanism. The system needs to perceive information such as user behavior and role differences to make subsequent control policy judgments;

[0168] The combination of multi-factor authentication and the post-quantum authentication protocol provides a multi-level authentication mechanism to ensure a more secure authentication process. During the communication process, the post-quantum authentication protocol is used to authenticate the user's identity to ensure the reliability of the user's identity and the anti-quantum attack ability;

[0169] The main key management and rotation are executed by the hardware security module HSM based on the chaotic enhancement mechanism. The chaotic main key and functional encryption keys are stored hierarchically and replaced regularly to ensure the security of the keys and the randomness of the update (after authentication, to ensure data security, a secure key management system is required);

[0170] Log records are generated for all operations involving key generation, data encryption and decryption, query, and analysis (during key operations such as encryption operations and data calls, the system should leave traces for auditing);

[0171] Regularly check the logs to detect abnormal behaviors, including abnormally frequent decryption requests and consecutive large numbers of query hits, and issue alarms in a timely manner; the log information is used for post-event auditing, to discover potential security risks and respond in a timely manner, which is a key link at the end of the security closed-loop.

[0172] As Figure 2 shown, this is another embodiment of the present invention. This embodiment provides a security protection system for the market supervision data security protection method based on a dynamic encryption policy as described above, including:

[0173] A dynamic encryption parameter generation module, which is used to generate an initial seed and input it into an improved four-dimensional hyperchaotic system, and generate a set of dynamic encryption parameters through chaotic mapping iteration;

[0174] An entropy value monitoring module, which is used to monitor the change of the entropy value of the dynamic encryption parameter set in real time. When the entropy value drops below the set threshold, it automatically refreshes the initial seed and re-iterates to generate a new set of dynamic encryption parameters;

[0175] A dynamic encryption module, which is used to adaptively slice the market supervision data to be encrypted based on the dynamic encryption parameter set, generate a set of data sub-blocks, and use symmetric encryption based on a dynamic S-box for the odd-index sub-blocks in the set of data sub-blocks, and use a chaotic stream encryption method associated with the current dynamic sub-key for the even-index sub-blocks. At the same time, a quantum chaotic fingerprint is generated for each encrypted sub-block to verify the integrity and uniqueness of the encrypted data;

[0176] An encrypted query index construction module, which is used to jointly construct a searchable encrypted index using a post-quantum encryption algorithm and a lattice-based encryption algorithm, and after submitting a zero-knowledge proof by the client for the server to verify the query legality, perform index retrieval based on chaotic topological rules, and return a set of encrypted result identifiers after removing duplicates;

[0177] A privacy protection calculation module, which is used to apply hierarchical functional encryption processing to the encrypted data sub-blocks, introduce chaotic differential privacy noise controlled by the dynamic encryption parameter set during the federated learning training process, update the global model weights, and adjust the noise intensity according to the differential privacy noise control factor;

[0178] A dynamic access control and authentication module, which is used to dynamically adjust the access policy by combining the user access behavior pattern, query frequency, and user role, perform identity authentication using a post-quantum security authentication protocol, and provide a multi-level identity authentication mechanism in combination with multi-factor authentication to ensure a more secure identity authentication process;

[0179] A key management module, which is used to perform main key management and rotation based on a hardware security module HSM with a chaotic enhancement mechanism, hierarchically store and periodically replace the chaotic main key and functional encryption keys to ensure the security of the keys and the randomness of updates;

[0180] The logging and security auditing module is used to record logs for all operations involving key generation, data encryption / decryption, querying, and analysis, and regularly check the logs to detect abnormal behaviors, including abnormally frequent decryption requests and consecutive large numbers of query hits, and issue alarms in a timely manner;

[0181] The data storage and backup module is used to set redundancy backup strategies according to the importance levels of data sub-blocks, including triple-copy storage, dual-copy storage, and single-copy storage mechanisms.

[0182] In summary, by introducing innovative technologies such as dynamic encryption policies, post-quantum encryption algorithms, differential privacy protection, and multi-factor authentication, the present invention not only significantly improves the encryption security, privacy protection ability, and resistance to quantum computing attacks of market supervision data, but also provides an efficient, flexible, and scalable data security protection solution.

[0183] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of various changes or substitutions, and these should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A market supervision data security protection method based on a dynamic encryption strategy, characterized in that: The method comprises: Generate an initial seed through a quantum random number generator, input it into an improved four-dimensional hyperchaotic system, and generate a dynamic encryption parameter set, including a dynamic subkey and a dynamic S-box, through chaotic mapping iteration; monitor the entropy value change of the dynamic encryption parameter set in real time, and when the entropy value drops below a set threshold, automatically refresh the initial seed and iterate to generate a new dynamic encryption parameter set; adaptively segment the encrypted market regulatory data based on the dynamic encryption parameter set to generate a data sub-block set, adopt symmetric encryption based on dynamic S-box for odd-indexed sub-blocks in the data sub-block set, adopt chaotic stream encryption associated with the current dynamic sub-key for even-indexed sub-blocks, and generate a unique quantum chaotic fingerprint for each encrypted sub-block to verify the integrity and uniqueness of the encrypted data; set a redundant backup strategy according to the importance level of each encrypted sub-block; For the query keywords submitted by the user, a post-quantum encryption algorithm and a lattice-based encryption algorithm are used to jointly construct a searchable encrypted index, which is used for query operations without decrypting the data; after the client submits the zero-knowledge proof to the server to verify the legitimacy of the query, the server performs index retrieval based on chaotic topology rules, and returns a set of encrypted result identifiers after removing duplicates.

2. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 1 is characterized in that: The method of iteratively generating a dynamic encryption parameter set through chaotic mapping includes: A multi-source quantum random number generator that integrates a photon source and a charge source is used to generate the initial seed Q_Seed; The initial seed Q_Seed is used to generate multiple initial vectors of the chaotic system through hash expansion, and reversibly expanded based on Gaussian random matrix mapping, and sensitively disturbed in combination with the platform master key; During the perturbation process, a nonlinear permutation mask mechanism is introduced to perform secondary confusion on the expanded initial vector; The initial vector is evolved according to the following four-dimensional hyperchaotic mapping iterative relationship: In the formula, X(n), Y(n), Z(n), and W(n) are the four variables of the improved four-dimensional hyperchaotic system, +1X(n+1), Y(n+1), Z(n+1), and W(n+1) are the updated variables; a, b, c, and d are adjustment parameters; Φ is the quantum gate perturbation function; Represents a combinatorial operation; represents the XOR operation; H(·) is a lightweight hash compression function; After a preset number of iterations, each chaotic orbit is sampled to generate a dynamic subkey group and a dynamic S-box mapping table to form a dynamic encryption parameter set P(t).

3. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 2 is characterized in that: When the entropy value drops below a set threshold, the initial seed is automatically refreshed and a new dynamic encryption parameter set is iteratively generated, the method comprising: Perform Shannon entropy H on the dynamic encryption parameter set P(t) generated in each round Shannon , Renyi entropy H Renyi And the sample entropy H Sample The multi-dimensional entropy value is jointly evaluated to construct the comprehensive entropy value H total , the formula is: H total =w1·H Shannon +w2·H Renyi +w3·Norm(H Sample ); Where w1, w2, w3 are the entropy weights, and w1+w2+w3=1; Norm(H Sample ) represents sample entropy normalization; Set up a multi-level entropy response mechanism, including: Safety interval: the comprehensive entropy value H total If the value is higher than the preset first threshold, the current parameters are maintained unchanged; Warning interval: the comprehensive entropy value H total Dynamically adjust adjustment parameters a, b, c, d of the chaotic mapping between a preset first threshold and a preset second threshold; Danger zone: the comprehensive entropy value H total When the value is lower than the preset second threshold, the initial seed Q_Seed refresh action is triggered; Before executing the refresh action, the system dynamically adjusts the refresh frequency and the preset number of iterations according to the current platform concurrent access number and data traffic load information; When a refresh is triggered, the quantum random number generator is called again to generate a new initial seed, and a random fragment is selected from the dynamic encryption parameter set P(t) whose entropy value is in the security interval in the previous round as a perturbation factor, which is combined with the new initial seed to generate an initial vector; The selection of refresh strategy is optimized based on a lightweight reinforcement learning model, which decides whether to adopt rehashing, gene migration or full restart mode based on historical entropy trends and refresh effect feedback; The chaotic iterative evolution is performed on the new initial vector, and the entropy value of the generated result is verified for multiple consecutive rounds. If the entropy value of two consecutive rounds is lower than the safety interval, the enhanced restart mechanism is executed.

4. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 3 is characterized in that: The enhanced restart mechanism specifically includes: Discard all current states; Re-call the quantum random number generator to generate a new initial seed; Reset the initial vector of the chaos map; Increase the number of initial iteration steps; Switch to the backup chaos mapping system; Until the comprehensive entropy value of the dynamic encryption parameter set P(t) is restored to the safe range.

5. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 3 is characterized in that: The method of adaptively sharding the market supervision data to be encrypted based on the dynamic encryption parameter set includes: According to the comprehensive entropy value H extracted from the dynamic encryption parameter set P(t), total , determine the number of shards N of market supervision data D, the formula is: Where DataSize(D) represents the size of the market supervision data D to be encrypted; Generate a pseudo-random mask sequence Mask(t) based on the dynamic encryption parameter set P(t) to define the segmentation position of the market supervision data D; Using the pseudo-random mask sequence Mask(t), the market supervision data D to be encrypted is divided into a data sub-block set {D i }, the boundary of each data word block is determined by the mask value; After sharding, according to the chaotic trajectory π(t), the data sub-block set {D i Specifically, the chaotic perturbation permutation function π is used to disrupt the physical order of the data sub-blocks and generate a new logical order {D π(i) }.

6. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 5 is characterized in that: After encrypting the data sub-blocks, combined with the previous encryption history data, based on the multiple hashing mechanism, a quantum chaos fingerprint based on the historical encryption chain is generated for each sub-block; The entropy density of each data sub-block is calculated, and the importance of each data sub-block is evaluated according to the entropy density, and divided into three levels: high, medium and low; Set the corresponding redundant backup strategy according to the importance level of the data sub-block, specifically: High level: For key sensitive data, a triple copy storage strategy is adopted, and an erasure coding mechanism based on XOR operation is superimposed. Each copy is stored in a different physical node, and a redundant check block is generated through XOR coding to ensure that rapid recovery can be achieved through the remaining copies and check data when a single node fails. Medium level: A dual-copy storage mechanism is used with an additional timestamp chain structure. A hash association with increasing timestamps is established between the two copies to form an unalterable time series record. Any modification of the copy will trigger chain verification to prevent data tampering and historical version forgery. Low level: Single copy storage is performed, only lightweight hash algorithm is used to generate data fingerprint, and basic integrity verification is implemented through hash value with low computational overhead.

7. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 1 is characterized in that: The method of using a post-quantum encryption algorithm and a lattice-based encryption algorithm to jointly construct a searchable encrypted index includes: using a post-quantum encryption algorithm to encrypt a query keyword Q to obtain an encrypted query keyword E PQE (Q); Encrypt the query keyword Q using a lattice-based encryption algorithm to obtain the encrypted query keyword E LWE (Q); Combine the encrypted query keywords to generate a searchable encrypted query index Index(Q), the expression is: Index(Q)=[E PQE (Q)⊕E LWE (Q)]⊕P(t); Where P(t) is the dynamic encryption parameter set; The client submits a query request to the server through the searchable encrypted query index Index(Q).

8. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 7 is characterized in that: After the client submits the zero-knowledge proof to the server for query legitimacy verification, the server performs index retrieval based on the chaotic topology rule and returns the encrypted result identification set after removing duplicate items. The method includes: After receiving the searchable encrypted query index Index(Q) and zero-knowledge proof ZKP(Q) submitted by the client, the server verifies the legitimacy of the query; The server generates the chaotic topology rule T based on the current dynamic encryption parameter set P(t) chaos (t), and according to the chaotic topological rule T chaos (t) perform encrypted index retrieval to query the encrypted data set; The server applies an intelligent deduplication mechanism when returning query results. It removes duplicate encrypted result identifiers through similarity evaluation and dynamically adjusts the returned content through fuzzy query technology to ensure that the query results are highly matched with user needs while avoiding data leakage. The client decrypts the deduplicated query results and verifies the integrity of the query results through quantum chaos fingerprints to ensure that the query results are consistent with the historical version of the encrypted data.

9. The method for protecting market supervision data security based on dynamic encryption strategy according to claim 1 is characterized in that: The method also includes privacy-preserving computation and dynamic access control and authentication; The privacy protection calculation is specifically as follows: Hierarchical functional encryption is applied to the encrypted data sub-blocks, and chaotic differential privacy noise controlled by the dynamic encryption parameter set P(t) is introduced in the federated learning training process to update the global model weights. The formula is as follows: Where W t+1 is the weight of the global model at t+1, W t is the weight of the global model at t; η represents the learning rate; is the gradient of the loss function; λ represents the control factor of differential privacy noise; ChaosNoise(P(t)) represents the chaotic differential privacy noise generated based on the dynamic encryption parameter set P(t); The dynamic access control and authentication are specifically as follows: Dynamically adjust access policies based on user access behavior patterns, query frequency, and user roles, and use post-quantum security authentication protocols for identity authentication; Multi-factor authentication is combined with post-quantum authentication protocol to provide a multi-level identity authentication mechanism to ensure a more secure identity authentication process. During the communication process, the post-quantum authentication protocol is used to verify the user's identity to ensure the reliability of the user's identity and the ability to resist quantum attacks; The master key management and rotation are performed by the hardware security module HSM based on the chaos enhancement mechanism. The chaotic master key and functional encryption key are stored hierarchically and replaced regularly to ensure the security of the key and the randomness of the update. Generate log records for all operations involving key generation, data encryption and decryption, query and analysis; Check logs regularly to detect abnormal behavior, including unusually frequent decryption requests and continuous large number of query hits, and issue timely alerts.

10. The security protection system of the market supervision data security protection method based on dynamic encryption strategy according to any one of claims 1 to 9, characterized in that: The system comprises: A dynamic encryption parameter generation module is used to generate an initial seed and input it into the improved four-dimensional hyperchaotic system, and to iteratively generate a dynamic encryption parameter set through chaotic mapping; An entropy value monitoring module is used to monitor the entropy value change of the dynamic encryption parameter set in real time. When the entropy value drops below a set threshold, the initial seed is automatically refreshed and a new dynamic encryption parameter set is re-iteratively generated; A dynamic encryption module, used to adaptively slice the market supervision data to be encrypted based on the dynamic encryption parameter set, generate a data sub-block set, and use symmetric encryption based on dynamic S-box for odd-indexed sub-blocks in the data sub-block set, and use chaotic stream encryption associated with the current dynamic sub-key for even-indexed sub-blocks, and generate a quantum chaotic fingerprint for each encrypted sub-block to verify the integrity and uniqueness of the encrypted data; The encrypted query index construction module is used to jointly construct a searchable encrypted index using a post-quantum encryption algorithm and a lattice-based encryption algorithm. After submitting a zero-knowledge proof on the client side and verifying the legitimacy of the query on the server side, the index retrieval is performed based on the chaotic topology rules, and the encrypted result identification set is returned after removing duplicate items. The privacy-preserving computing module is used to apply hierarchical functional encryption processing to the encrypted data sub-blocks, introduce chaotic differential privacy noise controlled by a dynamic encryption parameter set during the federated learning training process, update the global model weights, and adjust the noise intensity according to the differential privacy noise control factor; Dynamic access control and authentication module, which is used to dynamically adjust access policies based on user access behavior patterns, query frequency, and user roles, adopt post-quantum security authentication protocols for identity authentication, and provide a multi-level authentication mechanism in combination with multi-factor authentication to ensure a more secure authentication process; The key management module is used to perform master key management and rotation based on the hardware security module HSM of the chaos enhancement mechanism, hierarchically store and regularly replace the chaotic master key and functional encryption key to ensure the security of the key and the randomness of the update; The logging and security auditing module is used to log all operations involving key generation, data encryption and decryption, query and analysis, and regularly check the logs to detect abnormal behaviors, including abnormally frequent decryption requests and continuous large number of query hits, and issue timely alarms; The data storage and backup module is used to set redundant backup strategies according to the importance level of data sub-blocks, including triple copy storage, double copy storage and single copy storage mechanisms.

Images