A data security method and system based on privacy computing and multi-functional encryption

Abstract

The application discloses a kind of data security method and system based on privacy calculation and multi-functional encryption, it is related to privacy security computing field, including;The operation authority of encrypted data block is verified to computing node by zero-knowledge proof protocol, and strategy compliance proof is generated;Using differential privacy budget real-time verification mechanism, the privacy protection strength of strategy compliance proof is verified, and the temporary permission token of authorized passage is obtained;Based on temporary permission token, the strategy compliance proof allowed cryptomancy calculation is carried out to encrypted data block by homomorphic operation, and the encrypted calculation result and calculation correctness proof are generated;Encrypted calculation result, strategy compliance proof and calculation correctness proof are jointly cryptography encapsulation, and the output security data is passed through verifiable computing protocol;The application realizes the dual guarantee of privacy protection and data security by differential privacy budget real-time verification mechanism and homomorphic cryptomancy calculation based on temporary permission token.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of privacy security computing, in particular to a data security method and system based on privacy computing and multi-functional encryption. BACKGROUND

[0002] Currently, the privacy computing field widely adopts joint modeling technology based on homomorphic encryption (HE) and secure multi-party computing (MPC), such as the Cryptonets framework proposed by Microsoft Research, which realizes the computability of data in an encrypted state by converting the neural network inference process into homomorphic encryption operations. In addition, the industry gradually introduces policy-driven access control mechanisms, such as attribute-based encryption (ABE) and zero-knowledge proof (ZKP), to enhance the compliance verification capability in the data use process.

[0003] The limitations of the prior art mainly lie in that natural language policies need to be manually converted into cryptographic parameters, and the computing process needs to be interrupted and re-encrypted when the policy changes; for example, when a new "statistical access only" policy is added in a medical data sharing scenario, the traditional method needs to decrypt the data and re-execute the encryption process, which not only destroys the continuity of the calculation, but also increases the risk of privacy leakage. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a data security method based on privacy computing and multi-functional encryption to solve the problem that data security and multi-functionality are difficult to balance due to insufficient coupling of policy and encryption.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a data security method based on privacy computing and multi-functional encryption, which includes: based on a natural language policy, a policy compiler is used to perform structured programmable rule conversion to obtain a policy code, and a key derivation function is used to generate an encryption key; based on the encryption key, a binding encryption algorithm is used to perform policy-based encapsulation on the data to be encrypted to generate an encryption data block that fuses policy features; a zero-knowledge proof protocol is used to verify the operation permission of the computing node on the encryption data block to generate a policy compliance proof; a differential privacy budget real-time verification mechanism is used to verify the privacy protection strength of the policy compliance proof to obtain a temporary permission token that is authorized to pass; based on the temporary permission token, a homomorphic operation is performed on the encryption data block to allow a policy-compliant proof to perform a ciphertext calculation, and an encryption calculation result and a calculation correctness proof are generated; the encryption calculation result, the policy compliance proof and the calculation correctness proof are jointly encrypted, and a secure data is output through a verifiable calculation protocol.

[0008] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, wherein: the natural language strategy is converted into a structured programmable rule through a policy compiler to obtain a policy code, and a key derivation function is used to generate an encryption key. The specific steps are as follows:

[0009] Based on natural language strategies, a strategy compiler is used to perform semantic analysis and logical transformation to obtain structured programmable rules.

[0010] The structured programmable rules are serialized by a policy encoder to obtain a policy code;

[0011] The policy encoding is parameterized using a key derivation function to obtain the encryption key.

[0012] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, wherein: based on the encryption key, the binding encryption algorithm is used to strategically encapsulate the encrypted data to generate an encrypted data block that integrates the policy features. The specific steps are as follows:

[0013] Based on the encryption key, the encryption parameters of the encrypted data are initialized through the binding encryption algorithm to generate the encryptor initialization parameters;

[0014] Combine the policy code with the encryptor initialization parameters, perform policy binding encryption on the encrypted data through the binding encryption algorithm, and generate an intermediate encrypted data block that carries the complete policy constraints.

[0015] The policy features are extracted from the intermediate encrypted data block, and are fused and integrity encapsulated through the policy coding embedding and integrity signature mechanism to obtain the encrypted data block with the fused policy features.

[0016] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, wherein: the operation authority of the computing node on the encrypted data block is verified by the zero-knowledge proof protocol to generate a policy compliance proof, the specific steps are as follows:

[0017] Through the zero-knowledge proof protocol, the computing node permission declaration is matched with the policy constraints in the encrypted data block to generate a permission verification request;

[0018] The policy verification engine performs policy rule matching and logic verification on permission verification requests to generate preliminary policy compliance proof;

[0019] The preliminary policy compliance proof is cryptographically bound to the encrypted data block, and integrity binding and tamper-proof encapsulation are performed through digital signatures to generate a verifiable policy compliance proof.

[0020] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, the differential privacy budget real-time verification mechanism is used to verify the privacy protection strength of the policy compliance proof and obtain a temporary authorization token. The specific steps are as follows:

[0021] Through policy encoding parsing and parameter standardization conversion, the privacy protection strength parameters declared in the policy compliance proof are extracted to generate structured privacy parameters;

[0022] Through the real-time verification mechanism of differential privacy budget, the structured privacy parameters are verified using the privacy budget threshold specified by the policy compliance proof, and the privacy budget verification results are output;

[0023] The privacy budget verification result is verified by the token generation engine, and the security token is cryptographically encapsulated to obtain a temporary permission token that is authorized.

[0024] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, wherein: based on the temporary permission token, the encrypted data block is subjected to a homomorphic operation to perform a secret state calculation allowed by the policy compliance proof, and the encrypted calculation result and the calculation correctness proof are generated. The specific steps are as follows:

[0025] The token parsing engine verifies the authorization scope and privacy budget of the temporary permission token in real time, and generates a confidential computing authorization certificate containing valid operation instructions.

[0026] Based on the secret computing authorization certificate, the encrypted data block is subjected to secret computing permitted by the policy compliance proof through homomorphic encryption to generate the encrypted computing result;

[0027] The encrypted calculation results are encrypted through a verifiable computing protocol, and the calculation trace is cryptographically encapsulated and zero-knowledge proof is generated to generate proof of calculation correctness.

[0028] As a preferred solution of the data security method based on privacy computing and multifunctional encryption described in the present invention, the encryption calculation results, policy compliance proof and calculation correctness proof are jointly cryptographically encapsulated, and secure data is output through a verifiable computing protocol. The specific steps are as follows:

[0029] Based on the encrypted calculation results, policy compliance proof and calculation correctness proof, the joint packaging engine is used to assemble structured data to generate a data packet to be verified containing the calculation evidence chain;

[0030] Through the verifiable computing protocol, the digital signature of the data packet to be verified is bound to the zero-knowledge proof to generate a tamper-proof secure data package;

[0031] The secure data encapsulation is verified on and off the chain through a lightweight verification circuit, and secure data with dual constraints of privacy protection and computational correctness is output.

[0032] In the second aspect, the present invention provides a data security system based on privacy computing and multifunctional encryption, including a compilation module, an encryption module, a verification module, a privacy control module, a secret computing module and an output module; the compilation module is used to perform structured programmable rule conversion based on natural language policies through a policy compiler to obtain policy encoding, and use a key derivation function to generate an encryption key; the encryption module is used to perform policy-based encapsulation of encrypted data using a binding encryption algorithm based on the encryption key, and generate an encrypted data block that integrates policy features; the verification module is used to verify the operation authority of the computing node on the encrypted data block through a zero-knowledge proof protocol, and generate a policy compliance proof; the privacy control module is used to use a differential privacy budget real-time verification mechanism to verify the privacy protection strength of the policy compliance proof and obtain an authorized temporary license token; the secret computing module is used to perform secret computing allowed by the policy compliance proof on the encrypted data block through homomorphic operations based on the temporary license token, and generate an encrypted computing result and a calculation correctness proof; the output module is used to jointly cryptographically encapsulate the encrypted computing result, the policy compliance proof and the calculation correctness proof, and output secure data through a verifiable computing protocol.

[0033] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the data security method based on privacy computing and multifunctional encryption as described in the first aspect of the present invention.

[0034] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the data security method based on privacy computing and multifunctional encryption as described in the first aspect of the present invention.

[0035] The beneficial effects of the present invention are: through the differential privacy budget real-time verification mechanism and homomorphic secret computing based on temporary permission tokens, dual protection of privacy protection and data security is achieved: the differential privacy budget mechanism ensures that the data usage process is strictly controlled to prevent the leakage of private information; homomorphic secret computing completes data processing in an encrypted state to ensure that the original data is invisible throughout the process, which not only meets the privacy compliance requirements of highly sensitive data, but also supports safe and efficient cross-institutional data collaboration, fully releasing data value while protecting data privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0037] Figure 1 Flowchart of a data security method based on privacy computing and multifunctional encryption.

[0038] Figure 2 Schematic diagram of a data security system based on privacy computing and multifunctional encryption.

[0039] Figure 3 Flowchart for binding encryption and data encapsulation to policies.

[0040] Figure 4 Flowchart for verification of secret state computation. DETAILED DESCRIPTION

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0042] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0043] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0044] Reference Figures 1 to 4 , is an embodiment of the present invention, which provides a data security method based on privacy computing and multifunctional encryption, including the following steps:

[0045] S1. Based on the natural language strategy, the strategy compiler is used to convert the structured programmable rules to obtain the strategy code, and the key derivation function is used to generate the encryption key.

[0046] S1.1. Based on the natural language strategy, the strategy compiler is used to perform semantic analysis and logical transformation to obtain structured programmable rules.

[0047] Specifically, after the natural language policy is input into the policy compiler, the policy compiler uses semantic parsing technology to extract the entities, attributes, and constraint relationships in the policy, parsing "allow department A to access type B data" into a triple structure of "subject = department A, object = type B, operation = access"; the parsing results are mapped into discrete rule forms through the logic conversion engine, such as "IF the request subject matches department A AND the request object matches type B THEN allow access to the operation", forming machine-executable conditional judgment logic; the policy syntax tree construction method is used to organize the discrete rules into a hierarchical tree structure, with the root node being the policy type, the child nodes containing specific constraints, and the leaf nodes associated with execution actions; the generated policy syntax tree is standardized and encoded, and the policy elements are encapsulated in the JSON-LD format to ensure that each policy component has a unique and identifiable semantic identifier, and the structured programmable rules are output;

[0048] It should be noted that the process of organizing discrete rules into a hierarchical tree structure using the policy syntax tree construction method is as follows: the root node is set as the policy type decision point, such as "data access control policy"; the first-level child nodes are branched according to the policy element classification, including subject branch, object branch and operation branch; the subordinate nodes of the subject branch expand specific subject constraints according to the actual policy content, such as the constraint condition of "role = manager" is mounted under the "department A" node; the subordinate nodes of the object branch are classified according to the data type, such as the attribute constraint of "sensitivity level = high" is associated with the "type B data" node; the operation branch contains "allow / deny" execution action nodes, and the depth-first traversal algorithm is used to check the inclusion relationship between discrete rules. When a discrete rule conflict is detected, a priority mark is added to the conflicting node; the generated policy syntax tree satisfies the hierarchical relationship in which the parent node represents the abstract policy category, the child node carries the specific constraint conditions, and the leaf node determines the execution action, forming a hierarchical tree structure with a clear policy execution path.

[0049] S1.2. Serialize the structured programmable rules through the policy encoder to obtain the policy code.

[0050] Specifically, a depth-first traversal algorithm is used to access each node of the policy syntax tree, starting from the root node to record the hierarchical relationship between nodes; non-leaf nodes in the policy syntax tree are converted into structured data containing policy type and child node location information, such as the storage location of the data access control policy node to record the subordinate subject constraints, object constraints and operation constraints; leaf nodes are converted into clear conditional expressions; predefined serialization format specifications are used to convert structured data into byte streams to ensure that special characters in conditional expressions are escaped; length identification information is added before the byte stream to form a policy data unit with clear boundary divisions; a hash algorithm is applied to each policy data unit to generate a content verification value; all policy data units are combined in the original topological order of the policy syntax tree, and a policy code that meets the binary format requirements is output;

[0051] It should be explained that the process of the predefined serialization format is: analyzing the three core data types of node type, conditional expression and reference relationship contained in the structured programmable rules; stipulating the order of fields in the byte stream as node type identifier, number of child nodes and conditional expression text; using the escape method of adding hexadecimal value for special characters in the conditional expression text; writing a 4-byte unsigned integer before the conditional expression text to represent the text length; calculating and appending a 32-byte SHA-256 hash value at the end of the complete serialized data; and finally forming a serialization format.

[0052] S1.3. Parameterize the policy encoding using a key derivation function to obtain an encryption key.

[0053] Specifically, the HMAC-SHA256 algorithm is used to take the policy code as the input message; a fixed-length random salt value, exemplarily 32 bytes, is selected and concatenated with the policy code to form the derived input; an iteration count parameter, exemplarily 10,000, is set to enhance the derivation strength; a key derivation function is executed, and the output is truncated to a specified length, exemplarily the first 32 bytes being used as the intermediate derivation result; a key stretching algorithm is applied to the intermediate derivation result, exemplarily using the PBKDF2 algorithm to perform secondary processing on all 32 bytes of the intermediate derivation result, enhancing the key strength through multiple iterative hashing; and finally, a fixed-length byte sequence, exemplarily 256 bits, is output as the encryption key. The encryption key maintains a strict correspondence with the original policy code; identical policy code inputs always produce identical encryption key outputs.

[0054] S2. Based on the encryption key, the binding encryption algorithm is used to strategically encapsulate the encrypted data to generate an encrypted data block that integrates the policy features.

[0055] S2.1. Based on the encryption key, the encryption parameters of the encrypted data are initialized through the binding encryption algorithm to generate the encryptor initialization parameters.

[0056] Specifically, the policy code and the encryption key are aligned byte by byte. If the lengths are inconsistent, zeros are added to the end of the shorter data until the two are of the same length. A logical exclusive OR operation is performed on the exemplary 8 bits of each corresponding byte. The Nth byte of the policy code is independently compared with each bit of the Nth byte of the encryption key. If they are different, a 1 is output, and if they are the same, a 0 is output, to generate an 8-bit result byte.

[0057] All 8-bit result bytes are concatenated in their original order to form a binding factor, and the length of the binding factor is the same as the input data to be encrypted. In the exemplary process, the 32-byte policy code and the 32-byte encryption key generate a 32-byte binding factor. The binding factor is processed by the SHA3-256 algorithm, and the SHA3-256 algorithm performs 24 rounds of Keccak-f

[1600] permutation operations to output a fixed 256-bit binding key. The first 128 bits of the binding key are used as the initialization vector component, and the last 128 bits are added to the original encryption key byte by byte and then modulo 256 to generate Encryption material; the policy code is input into the Keccak-1600 sponge function through the SHA3-256 algorithm, and the first 256 bits are truncated after 24 rounds of permutation to generate a 32-byte hash value. The first 16 bytes are used as the binding identifier; 16 bytes of space are reserved in the encryptor initialization parameter structure to store the CRC32 check value of the policy code; the 16-byte binding identifier, 16-byte check value reserved bit, 16-byte initialization vector component and 16-byte encryption material are written into the data structure according to the specified offset, and the 64-byte encryptor initialization parameters are output.

[0058] S2.2. Combine the policy code with the 64-byte encryptor initialization parameter, perform policy binding encryption on the encrypted data through the binding encryption algorithm, and generate an intermediate encrypted data block that carries the complete policy constraints.

[0059] Specifically, the 64-byte encryptor initialization parameter is parsed to obtain a 128-bit initialization vector and a 128-bit original encryption key; the AES-256 algorithm is configured to work in the GCM mode, the authentication tag length is set to 128 bits as an example, the initialization vector is used as the initial counter value of GCM, the data to be encrypted is divided into fixed-size intermediate data blocks, the size of each intermediate data block is set to 4096 bytes as an example, and the insufficient part is filled to a complete block by using PKCS#7; a policy encoding segment is inserted at the starting position of each intermediate data block, 32 bytes of policy encoding content are inserted as an example, and the policy encoding bytes are distributed in a cyclic manner; the original encryption key is used for encryption preprocessing of the data block, and the preprocessing process simultaneously calculates an HMAC-SHA256 check value by taking the policy encoding segment as additional authentication data; a 128-bit GCM pre-authentication tag is generated for each preprocessed intermediate data block, and the tag calculation includes the preprocessed intermediate data block and the corresponding policy encoding segment; a 256-bit SHA3-256 hash value of the complete policy encoding is appended to the tail of the final preprocessed data; and the intermediate encrypted data block composed of the preprocessed intermediate data block (containing the embedded policy encoding), the 128-bit pre-authentication tag and the tail policy integrity check value is output.

[0060] S2.3, extract the policy feature from the intermediate encrypted data block, and fuse and integrity-encapsulate the policy feature by embedding the policy encoding and the integrity signature mechanism to obtain an encrypted data block of the fused policy feature.

[0061] Specifically, the policy encoding segment in the intermediate encrypted data block is extracted; a CRC32 check value of the policy encoding segment is calculated, and a 4-byte check value is generated as an example; a 32-byte policy feature digest is generated by performing a hash operation on the complete policy encoding by using the SHA3-256 algorithm; a 64-byte signature is generated by using the ECDSA algorithm to digitally sign the 32-byte policy feature digest, and the secp256r1 elliptic curve is used as an example; the policy encoding segment, the CRC32 check value, the 32-byte policy feature digest and the ECDSA signature are written into the header of the intermediate encrypted data block according to a predetermined offset; a 32-byte check value is generated by recalculating the HMAC-SHA256 integrity check value of the intermediate encrypted data block as an example; the 32-byte check value is appended to the tail of the intermediate encrypted data block; and the encrypted data block of the fused policy feature containing the policy encoding segment, the 32-byte check value and the encrypted content is output;

[0062] The process of using the SHA3-256 algorithm to hash the complete policy code and generate a 32-byte policy feature summary should be explained: the complete policy code is arranged in byte order; the SHA3-256 state matrix is ​​initialized to a 1600-bit all-zero matrix; the complete policy code data is processed in 1152-bit blocks, exemplified by processing 144-byte encrypted data blocks; each encrypted data block is updated with the state matrix through an XOR operation; 24 rounds of Keccak-f

[1600] permutation are performed on the updated state matrix; finally, the first 256 bits from the updated state matrix are extracted as the policy feature summary; and the policy feature summary with a fixed length of 32 bytes is output;

[0063] Furthermore, the predetermined process of the predetermined offset is as follows: the four components of the policy coding segment, CRC32 check value, policy feature summary and ECDSA signature have fixed length characteristics; the policy coding segment is exemplarily 32 bytes in length, the CRC32 check value is fixed at 4 bytes, the policy feature summary is fixed at 32 bytes, and the ECDSA signature is fixed at 64 bytes; the arrangement order of each component is determined according to the functional relevance, the policy coding segment is placed at the starting position, the CRC32 check value follows immediately, and the policy feature summary and ECDSA signature are arranged in the verification logic order; a 1-byte separation gap is reserved between components to prevent data adhesion; the offset value record is statically configured, and the policy coding segment offset is 0 bytes, the CRC32 check value offset is 32 bytes, the policy feature summary offset is 36 bytes, and the ECDSA signature offset is 68 bytes.

[0064] S3. Verify the computing node's operation authority over the encrypted data block through the zero-knowledge proof protocol and generate a policy compliance certificate.

[0065] S3.1. Through the zero-knowledge proof protocol, the computing node permission declaration is matched with the policy constraints in the encrypted data block to generate a permission verification request.

[0066] Specifically, the policy encoding fragment embedded in the header of the encrypted data block is extracted; the policy encoding fragment is parsed to obtain the policy constraints, and the plain text rule of "Department A can access type B data" is parsed as an example; the computing node generates a permission statement containing its own identity credentials, and the example statement is "the current node belongs to Department A"; the zk-SNARK algorithm is used to construct an arithmetic circuit, and the arithmetic circuit input includes policy constraints and permission statements; the policy feature matching logic is implemented in the arithmetic circuit, and the identity attributes in the permission statement are non-interactively compared with the policy constraints; the arithmetic circuit outputs a 1-bit verification result, outputting 1 for successful matching and 0 for failure; the verification result, the arithmetic circuit proof, and the hash value of the policy encoding fragment are combined into structured data; and a permission verification request containing a complete verification evidence chain is output.

[0067] Furthermore, the process of constructing an arithmetic circuit using the zk-SNARK algorithm is as follows: the policy constraints parsed from the policy encoding fragment are mapped to logic gates, and the attribute values ​​in the permission declaration are used as private inputs of the circuit; the R1CS constraint expresses the logical judgment of the policy rule, and each R1CS constraint is in the form of a×b=c; the department affiliation verification is exemplarily expressed as the product relationship between the department identifier and the access permission bit; the QAP polynomial converts the R1CS constraint into a polynomial relationship; the verification result is set as the output public variable; and finally, an arithmetic circuit that encodes the policy matching calculation logic is formed.

[0068] S3.2. The policy verification engine performs policy rule matching and logic verification on the permission verification request to generate a preliminary policy compliance certificate.

[0069] Specifically, parse the zk-SNARK proof and policy feature hash contained in the permission verification request; extract the original policy code from the encrypted data block that integrates the policy features, re-obtain the original policy code SHA3-256 hash value and compare it with the policy feature hash in the request; verify the integrity of the zk-SNARK proof and confirm the validity of the output result of the arithmetic circuit; match the permission attributes declared by the computing node with the access control rules in the policy code item by item, and execute the logical judgment defined by the policy rules to match the permission attributes declared by the computing node (such as "Department A") with the access control rules in the policy code (such as "Allow Department A to access"). If the judgment result is that the conditions are met, output the access authorization conclusion (such as "Allow access"), and if they are not met, output the rejection conclusion (such as "Reject access"). The judgment result directly determines the final access control decision; record the key judgment nodes and judgment results in the matching process; output a preliminary policy compliance certificate containing policy feature identification, permission matching conclusion and verification time information;

[0070] Furthermore, the process of setting access control rules: convert policy statements such as "allow department A to access type B data" into machine-processable logical expressions; define three basic elements: subject attributes (such as department, role), object attributes (such as data type, sensitivity level) and operation type (such as read, write); use policy syntax tree to organize rules, with the root node as policy type, branch nodes as conditional judgments, and leaf nodes as execution actions; set matching operators (such as equals, contains) for each condition, and exemplarily set department matching to string exact matching; combine multiple conditions to form a compound logical expression, and exemplarily set "department A AND role manager" as the access prerequisite; finally output access control rules containing complete subject-object-operation triples and conditional logic.

[0071] S3.3. Cryptographically bind the preliminary policy compliance proof to the encrypted data block, perform integrity binding and tamper-proof encapsulation through digital signatures, and generate a verifiable policy compliance proof.

[0072] Specifically, three core data are extracted from the preliminary policy compliance proof: an exemplary 32-byte SHA3-256 hash value policy feature identifier, an exemplary 1-byte Boolean value permission determination conclusion (0x01 for allow / 0x00 for deny), and an exemplary 8-byte Unix timestamp format verification time information; an exemplary 32-byte length integrity feature value (stored policy encoding hash) is obtained from the encrypted data block header; these four data are arranged in strict offset order: an exemplary 0-31 byte offset stores the policy feature identifier, an exemplary 32-byte offset stores the permission determination conclusion, and an exemplary 33-40 byte offset stores the verification time information. , the integrity characteristic value is stored at an exemplary 41-72 byte offset to form verification data of an exemplary 72-byte length; the verification data is digitally signed using the secp256r1 elliptic curve, and the signature private key is an exemplary 256-bit policy administrator key to generate an exemplary 64-byte ECDSA signature; the digital signature is directly concatenated with the verification data, with the exemplary first 64 bytes storing the signature and the last 72 bytes storing the verification data, forming a complete policy compliance proof of an exemplary 136-byte length; the mathematical correspondence between the signature and the verification data is verified through the elliptic curve public key to confirm the validity of the signature; and finally a verifiable policy compliance proof is generated.

[0073] S4. Use the differential privacy budget real-time verification mechanism to verify the privacy protection strength of the policy compliance proof and obtain a temporary permission token for authorization.

[0074] S4.1. Through policy encoding parsing and parameter standardization conversion, extract the privacy protection strength parameters declared in the policy compliance proof and generate structured privacy parameters.

[0075] Specifically, the proof content of the verifiable policy compliance proof is extracted to obtain the original policy code; the privacy protection statement part clearly defined in the policy code is located; the privacy parameter type mark and corresponding value in the privacy protection statement are identified, and the text description of differential privacy parameter = 1.0 is identified as an example; the privacy parameter value is extracted and converted into a standard format, and the differential privacy parameter value is converted into a floating point representation as an example; the privacy parameter value is verified to be within the predefined valid range, and the differential privacy parameter value is confirmed to be between the minimum allowed value of 0.1 and the maximum allowed value of 10.0; the verified values ​​are stored according to the parameter type, and the differential privacy parameters, k-anonymity parameters, etc. are stored as examples; and the structured privacy parameters containing the complete parameter type and the corresponding value are output;

[0076] It should be noted that the process of determining the effective range is as follows: examining the minimum noise addition and maximum information leakage of the privacy parameter numerical processing operation; determining the parameter boundaries based on the privacy parameter numerical classification, with an exemplary differential privacy parameter range of 1.0-10.0 for ordinary data, 0.5-5.0 for sensitive data, and 0.1-2.0 for highly sensitive differential privacy; testing the privacy protection effect under different parameter values, and exemplarily verifying that the re-identification risk is less than 0.1% when the differential privacy parameter = 1.0; determining the minimum available value and maximum allowable value of the parameter, and exemplarily setting the allowable range of the global differential privacy parameter to 0.1-10.0.

[0077] S4.2. Through the real-time verification mechanism of differential privacy budget, the structured privacy parameters are verified using the privacy budget threshold specified by the policy compliance proof, and the privacy budget verification results are output.

[0078] Specifically, the privacy parameter type and value in the structured privacy parameter are parsed, and a differential privacy parameter = 1.0 is obtained as an example; a predefined privacy budget threshold is extracted from the policy compliance proof, and a privacy budget threshold of 5.0 is obtained as an example; the current parameter value is compared with the corresponding privacy budget threshold, and the comparison is 1.0≤5.0 as an example; the comparison result is recorded and the over-limit status is marked, and the differential privacy parameter value is recorded as not exceeding the limit; a combined verification is performed on multiple associated parameters, and the combined compliance of the k-anonymity parameter = 3 and the differential privacy parameter = 1.0 is verified as an example; a verification record containing the verification results of each parameter is generated, and the record "Differential privacy parameter verification passed" is recorded as an example; and a formatted privacy budget verification result is output, and the output is "Differential privacy budget remaining 80%". The privacy budget verification result reflects the consumption of the privacy budget by the current operation;

[0079] It should be noted that the process of pre-defining the privacy budget threshold is as follows: classify the sensitivity level of privacy parameters by analyzing data processing scenarios, and exemplarily divide them into three levels: normal, sensitive, and highly sensitive; perform quantitative privacy risk assessment on privacy parameter operations at different levels, and exemplarily measure the probability of re-identification when the differential privacy parameter = 1.0; perform parameter balance test in combination with the privacy parameter utility requirements, and exemplarily verify the availability of analysis results when the differential privacy parameter = 0.5; perform hierarchical settings based on the test results, and exemplarily set the single budget of differential privacy parameters for normal data as 1.0, sensitive privacy parameters as 0.5, and highly sensitive privacy parameters as 0.2; perform statistical analysis on historical operation privacy parameters, and exemplarily calculate the average differential privacy parameter consumption of 0.3 per query; determine the final privacy budget threshold through actual scenario verification, and exemplarily confirm that the total daily differential privacy parameter budget of 10.0 can meet business needs; and obtain the privacy budget threshold.

[0080] S4.3. The privacy budget verification result is cryptographically encapsulated into a security token through the token generation engine to obtain a temporary authorization token.

[0081] Specifically, read the privacy budget verification result field position, the exemplary status field is located at the 1st-2nd bytes (0x01 means pass / 0x00 means reject), and the remaining budget percentage is located at the 3rd-4th bytes (0x50 means 80%); convert the binary value of the status field to Boolean type, and convert 0x01 to "pass"; convert the remaining budget percentage byte value to a decimal value; verify the validity of the field value, and confirm that the percentage value is in the range of 0-100; output the structured privacy parameter verification status and remaining budget percentage value; read the timestamp field in the policy compliance certificate to extract the authorization start and end time; convert the verification status, remaining budget, and remaining budget percentage into a decimal value; The calculation ratio and time information are parsed and spliced. The exemplary format is "parameter type: differential privacy parameter | status: passed | remaining: 80% | start: 202303011200 | end: 202303012000"; obtain the SHA3-256 hash value of the spliced ​​privacy budget verification result as the token unique identifier; use the preset 256-bit elliptic curve private key to digitally sign the hash value; combine the digital signature result with the privacy budget verification result, exemplarily placing the signature in the first 64 bytes and the privacy budget verification result in the subsequent bytes; append a 4-byte CRC checksum to verify data integrity; output a temporary license token containing the signature, the data to be encrypted, and the checksum.

[0082] S5. Based on the temporary permission token, perform secret calculations allowed by the policy compliance proof on the encrypted data block through homomorphic operations to generate encrypted calculation results and proof of calculation correctness.

[0083] S5.1. Perform real-time verification of the authorization scope and privacy budget of the temporary permission token through the token parsing engine, and generate a confidential computing authorization credential containing valid operation instructions.

[0084] Specifically, verify the validity of the digital signature of the temporary license token, and use the preset public key to check the matching of the signature and the temporary license token; parse the token data to obtain the authorization parameters, and exemplarily read the data in the format of "differential privacy parameters|pass|80%|202303011200|202303012000"; verify the validity of the time, and compare the current time with the start time and end time in the temporary license token; confirm that the remaining budget is sufficient, and exemplarily verify that 80% of the remaining budget meets the current operation requirements; extract the operation permission list in the policy compliance certificate, and exemplarily obtain "allow aggregate computing|prohibit access to encrypted data"; combine the verified authorization parameters and operation permissions, and exemplarily generate the "allow execution of aggregate computing with differential privacy parameters ≤ 1.0" instruction; use the AES-256 algorithm to encrypt the authorization instruction, and exemplarily use a 256-bit random number temporarily generated for the session as the key; output the secret computing authorization certificate containing the encrypted authorization instruction and the remaining budget status;

[0085] It should be noted that the provisioning process of the preset public key is: using a random source to generate key material; the key pair generation device automatically performs the key derivation function and transfers the public key part to the verification node storage area; the public key storage area is deployed in a secure environment with physical protection and access control; the key lifecycle management process ensures that the public key remains available during the validity period; and finally, the preset public key for token verification is formed.

[0086] S5.2. Based on the secret computing authorization certificate, perform secret computing permitted by the policy compliance proof on the encrypted data block through homomorphic encryption to generate an encrypted computing result.

[0087] It should be explained that the expression for generating the encrypted calculation result by performing the secret calculation allowed by the policy compliance proof on the encrypted data block through homomorphic encryption is:

[0088]

[0089] Where M is the result of encryption calculation, C1 is the ciphertext of original plaintext 1m1 after homomorphic encryption, C2 is the ciphertext of original plaintext 2m2 after homomorphic encryption, E is the encryption operation on plaintext m1 and m2, m1 is the original plaintext 1 to be encrypted, and m2 is the original plaintext 2 to be encrypted;

[0090] Specifically, parse the secret computing authorization certificate to obtain operation instructions and permission constraints; verify the matching of the policy compliance certificate carried by the encrypted data block and the operation instructions; initialize the homomorphic encryption operation environment, and exemplarily configure the Paillier algorithm parameters to a 2048-bit modulus; perform homomorphic multiplication operations on the encrypted data block according to the instruction requirements, and exemplarily process two encrypted data blocks to realize ciphertext multiplication; monitor and record the privacy budget consumption status in real time, and exemplarily increase the cumulative differential privacy parameter value by 0.5; output the operation result in an encrypted state; attach metadata including the operation type and actual budget consumption; and generate an encrypted calculation result that meets the policy constraints.

[0091] S5.3. Use a verifiable computing protocol to cryptographically encapsulate the calculation trace of the encrypted calculation results and generate a zero-knowledge proof to generate a proof of calculation correctness.

[0092] Specifically, the steps of the homomorphic encryption calculation process are recorded, and the execution trajectory of "multiplication operation-input block 1-input block 2-output block" is recorded as an example; the operation trajectory is converted into an arithmetic circuit representation, and the multiplication operation is converted into an R1CS constraint relationship as an example; a zero-knowledge proof is generated using the zk-SNARK protocol, and the zero-knowledge proof key length is set to 256 bits as an example; an integrity check value is added to the encryption calculation result, and a SHA3-256 hash is calculated as an example; the zero-knowledge proof and the integrity check value are combined in a predetermined format, and the first 128 bytes are stored as a proof of calculation correctness, and the last 32 bytes are stored as a hash; the logical consistency of the proof of calculation correctness and the encryption calculation result is verified; and a proof of calculation correctness containing complete proof data and verification information is output;

[0093] Furthermore, the specific process of verifying the logical consistency of the proof of computational correctness and the result of the encryption calculation is as follows: extract the zero-knowledge proof part and the result hash value in the proof of computational correctness; recalculate the SHA3-256 hash value from the result of the encryption calculation; compare the consistency of the two hash values; use the preset verification key to verify the zero-knowledge proof to confirm the validity of the zero-knowledge proof; check the correspondence between the arithmetic circuit output in the zero-knowledge proof and the result of the encryption calculation; verify the matching of the privacy budget consumption record and the computational complexity; confirm that the logical consistency is established after all verification steps have passed; if any verification step fails (for example, the hash does not match or the zero-knowledge proof verification fails), the process is terminated immediately and the "computational correctness verification failed" error status code is output, marking the encrypted data block as untrusted.

[0094] S6. Cryptographically encapsulate the encrypted calculation results, policy compliance proof, and calculation correctness proof, and output secure data through a verifiable computing protocol.

[0095] S6.1. Based on the encrypted calculation results, policy compliance proof, and calculation correctness proof, structured data is assembled through the joint packaging engine to generate a data packet to be verified containing the calculation evidence chain.

[0096] Specifically, parse the calculation correctness proof to obtain the zero-knowledge proof part and verification information, extract the arithmetic circuit parameters and verification key in the zero-knowledge proof; extract the 32-byte policy feature identifier and 8-byte authorization tag from the policy compliance proof; divide the encrypted calculation result into blocks of predetermined size, exemplarily every 4096 bytes as an encrypted data block, and add a 2-byte sequence number and a 1-byte type tag to each encrypted data block; arrange the data components in the order of policy proof (exemplarily 128 bytes), calculation result block (variable length), and correctness proof (exemplarily 160 bytes); add a 4-byte length prefix and a 1-byte type identifier to each component; splice the sequentially arranged data components in sequence, and append a 32-byte integrity check value to the end of the data packet to be verified; output the data packet to be verified containing the complete evidence chain.

[0097] S6.2. Digitally sign the data packet to be verified and bind it with zero-knowledge proof through a verifiable computing protocol to generate a tamper-proof secure data package.

[0098] Specifically, 128 bytes of policy compliance proof are extracted from the header of the data packet to be verified, including the policy feature identifier and authorization tag; the encrypted calculation result block in the middle part is read, and multiple 4096-byte encrypted data blocks are obtained, each with a serial number tag; the calculation correctness proof of the last 160 bytes is parsed to obtain the zero-knowledge proof and verification parameters; the SHA3-512 algorithm is used to process the complete data packet to generate a 64-byte hash value; the secp256r1 elliptic curve is used to digitally sign the hash value to generate 64-byte signature data; the zero-knowledge proof and the signature data are combined in a fixed format, and the exemplary zero-knowledge proof occupies the first 192 bytes and the signature occupies the subsequent 64 bytes; it is verified whether the combined data conforms to the predefined format specification; and a 256-byte security data package consisting of the data to be encrypted, the zero-knowledge proof and the digital signature is output.

[0099] S6.3. Use lightweight verification circuits to verify the secure data encapsulation on and off the chain, and output secure data with dual constraints of privacy protection and computational correctness.

[0100] Specifically, extract the 64-byte digital signature and 192-byte zero-knowledge proof data from the header of the security data encapsulation; use the secp256r1 elliptic curve public key to verify the digital signature and confirm the correspondence between the digital signature and the content of the security data encapsulation; parse the arithmetic circuit parameters and verification key in the zero-knowledge proof to verify that the calculation process meets the requirements of the policy compliance proof; check the privacy budget consumption record, and verify that the cumulative value of the differential privacy parameter 0.8 does not exceed the privacy budget threshold 5.0; compare the consistency of the 32-byte policy hash value stored on the chain with the policy feature identifier in the encapsulation; after the verification is passed, append an 8-byte verification result mark to the security data encapsulation, and the exemplary mark content is 0x01 (privacy passed) and 0x01 (calculation passed); output the final security data with a double verification mark.

[0101] This embodiment also provides a data security system based on privacy computing and multifunctional encryption, including:

[0102] Compilation module, encryption module, verification module, privacy control module, secret state calculation module and output module;

[0103] A compilation module is used to convert structured programmable rules based on natural language policies through a policy compiler to obtain policy encoding, and to generate encryption keys using a key derivation function;

[0104] The encryption module is used to strategically encapsulate the encrypted data using a binding encryption algorithm based on the encryption key to generate an encrypted data block that integrates the policy features;

[0105] The verification module is used to verify the computing node's operation authority on the encrypted data block through the zero-knowledge proof protocol and generate a policy compliance certificate;

[0106] The privacy control module is used to verify the privacy protection strength of the policy compliance proof using a real-time verification mechanism of differential privacy budget and obtain a temporary authorization token.

[0107] A secret state computing module is used to perform secret state computing permitted by policy compliance proof on encrypted data blocks through homomorphic operations based on temporary permission tokens, and generate encrypted computing results and computing correctness proof;

[0108] The output module is used to cryptographically encapsulate the encrypted calculation results, policy compliance proof, and calculation correctness proof, and output secure data through a verifiable computing protocol.

[0109] This embodiment also provides a computer device, which is suitable for the data security method based on privacy computing and multifunctional encryption, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the data security method based on privacy computing and multifunctional encryption proposed in the above embodiment.

[0110] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0111] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the data security method based on privacy computing and multi-functional encryption proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.

[0112] In summary, the present invention achieves dual protection of privacy protection and data security through: a real-time verification mechanism for differential privacy budget and homomorphic secret computing based on temporary permission tokens: the differential privacy budget mechanism ensures that the data usage process is strictly controlled to prevent the leakage of private information; homomorphic secret computing completes data processing in an encrypted state to ensure that the original data is invisible throughout the process, which not only meets the privacy compliance requirements of highly sensitive data, but also supports safe and efficient cross-institutional data collaboration, fully releasing the value of data while protecting data privacy.

[0113] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A data security method based on privacy computing and multifunctional encryption, characterized by: include, Based on the natural language strategy, the strategy compiler is used to convert the structured programmable rules to obtain the strategy code, and the key derivation function is used to generate the encryption key; Based on the encryption key, the binding encryption algorithm is used to strategically encapsulate the encrypted data to generate an encrypted data block that integrates the policy features; Verify the computing node's operation rights on the encrypted data block through the zero-knowledge proof protocol and generate policy compliance proof; A real-time verification mechanism for differential privacy budget is used to verify the privacy protection strength of the policy compliance proof and obtain a temporary authorization token. Based on the temporary permission token, the encrypted data block is subjected to the secret state calculation allowed by the policy compliance proof through homomorphic operation, and the encrypted calculation result and the calculation correctness proof are generated; The encrypted calculation results, policy compliance proof and calculation correctness proof are jointly cryptographically encapsulated, and secure data is output through a verifiable computing protocol.

2. The data security method based on privacy computing and multifunctional encryption according to claim 1, wherein: The natural language strategy is converted into a structured programmable rule through a strategy compiler to obtain a strategy code, and a key derivation function is used to generate an encryption key. The specific steps are as follows: Based on natural language strategies, a strategy compiler is used to perform semantic analysis and logical transformation to obtain structured programmable rules. The structured programmable rules are serialized by a policy encoder to obtain a policy code; The policy encoding is parameterized using a key derivation function to obtain the encryption key.

3. The data security method based on privacy computing and multifunctional encryption according to claim 2, wherein: Based on the encryption key, the binding encryption algorithm is used to strategically encapsulate the encrypted data to generate an encrypted data block that integrates the policy features. The specific steps are as follows: Based on the encryption key, the encryption parameters of the encrypted data are initialized through the binding encryption algorithm to generate the encryptor initialization parameters; Combine the policy code with the encryptor initialization parameters, perform policy binding encryption on the encrypted data through the binding encryption algorithm, and generate an intermediate encrypted data block that carries the complete policy constraints. The policy features are extracted from the intermediate encrypted data block, and are fused and integrity encapsulated through the policy coding embedding and integrity signature mechanism to obtain the encrypted data block with the fused policy features.

4. The data security method based on privacy computing and multifunctional encryption according to claim 3, wherein: The zero-knowledge proof protocol is used to verify the computing node's operation authority on the encrypted data block and generate a policy compliance certificate. The specific steps are as follows: Through the zero-knowledge proof protocol, the computing node permission declaration is matched with the policy constraints in the encrypted data block to generate a permission verification request; The policy verification engine performs policy rule matching and logic verification on permission verification requests to generate preliminary policy compliance proof; The preliminary policy compliance proof is cryptographically bound to the encrypted data block, and integrity binding and tamper-proof encapsulation are performed through digital signatures to generate a verifiable policy compliance proof.

5. The data security method based on privacy computing and multifunctional encryption according to claim 4, characterized in that: The differential privacy budget real-time verification mechanism is used to verify the privacy protection strength of the policy compliance proof and obtain a temporary authorization token. The specific steps are as follows: Through policy encoding parsing and parameter standardization conversion, the privacy protection strength parameters declared in the policy compliance proof are extracted to generate structured privacy parameters; Through the real-time verification mechanism of differential privacy budget, the structured privacy parameters are verified using the privacy budget threshold specified by the policy compliance proof, and the privacy budget verification results are output; The privacy budget verification result is verified by the token generation engine, and the security token is cryptographically encapsulated to obtain a temporary permission token that is authorized.

6. The data security method based on privacy computing and multifunctional encryption according to claim 5, characterized in that: Based on the temporary permission token, the encrypted data block is subjected to a homomorphic operation to perform a secret calculation allowed by the policy compliance proof, and an encrypted calculation result and a proof of calculation correctness are generated. The specific steps are as follows: The token parsing engine verifies the authorization scope and privacy budget of the temporary permission token in real time, and generates a confidential computing authorization certificate containing valid operation instructions. Based on the secret computing authorization certificate, the encrypted data block is subjected to secret computing permitted by the policy compliance proof through homomorphic encryption to generate the encrypted computing result; The encrypted calculation results are encrypted through a verifiable computing protocol, and the calculation trace is cryptographically encapsulated and zero-knowledge proof is generated to generate proof of calculation correctness.

7. The data security method based on privacy computing and multifunctional encryption according to claim 6, wherein: The encryption calculation results, policy compliance proof and calculation correctness proof are jointly cryptographically encapsulated and secure data is output through a verifiable computing protocol. The specific steps are as follows: Based on the encrypted calculation results, policy compliance proof and calculation correctness proof, the joint packaging engine is used to assemble structured data to generate a data packet to be verified containing the calculation evidence chain; Through the verifiable computing protocol, the digital signature of the data packet to be verified is bound to the zero-knowledge proof to generate a tamper-proof secure data package; The secure data encapsulation is verified on and off the chain through a lightweight verification circuit, and secure data with dual constraints of privacy protection and computational correctness is output.

8. A data security system based on privacy-preserving computing and multifunctional encryption, based on the data security method based on privacy-preserving computing and multifunctional encryption according to any one of claims 1 to 7, characterized in that: Including compilation module, encryption module, verification module, privacy control module, secret state calculation module and output module; A compilation module is used to convert structured programmable rules based on natural language policies through a policy compiler to obtain policy encoding, and to generate encryption keys using a key derivation function; The encryption module is used to strategically encapsulate the encrypted data using a binding encryption algorithm based on the encryption key to generate an encrypted data block that integrates the policy features; The verification module is used to verify the computing node's operation authority on the encrypted data block through the zero-knowledge proof protocol and generate a policy compliance certificate; The privacy control module is used to verify the privacy protection strength of the policy compliance proof using a real-time verification mechanism of differential privacy budget and obtain a temporary authorization token. A secret state computing module is used to perform secret state computing permitted by policy compliance proof on encrypted data blocks through homomorphic operations based on temporary permission tokens, and generate encrypted computing results and computing correctness proof; The output module is used to cryptographically encapsulate the encrypted calculation results, policy compliance proof, and calculation correctness proof, and output secure data through a verifiable computing protocol.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the data security method based on privacy computing and multifunctional encryption are implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the data security method based on privacy computing and multifunctional encryption according to any one of claims 1 to 7 are implemented.

Images