Blockchain-based semi-honest privacy protection method and system for secure multi-party computation

Abstract

This invention discloses a semi-honest privacy protection method and system based on blockchain secure multi-party computation. The method includes: constructing a general security architecture and configuring a blockchain identity generation module to obtain user attribute credentials; deploying a reconfigurable computing engine through an off-chain computing layer, dynamically decomposing the computation logic into sub-task units according to task type, and executing local computation through a semi-honest security model; adapting to different computing scenarios through dynamic task decomposition based on the reconfigurable engine, defining the scope of credential disclosure through a programmable policy language to achieve selective disclosure; implementing segmented encryption of sub-task inputs or outputs using a process privacy protection protocol and binding it to the blockchain ledger layer; the semi-honest model allows participants to follow the process privacy protection protocol, ensuring that even if a single node is compromised, the original data remains unrecoverable through segmented encryption; and finally, recombining the sub-task results through a smart contract layer and verifying the consistency of the computation results to achieve end-to-end consistent response.

Description

Technical Field

[0001] This invention belongs to the field of blockchain digital identity, specifically relating to a semi-honest privacy protection method and system based on blockchain secure multi-party computation. Background Technology

[0002] With the surge in demand for multi-institutional data collaboration in fields such as finance and healthcare, secure multi-party computation (MPC) technology has become a core means of privacy protection. While traditional MPC solutions (such as obfuscated circuits and secret sharing) can achieve "data usable but not visible," their static computation models struggle to adapt to heterogeneous computing environments (such as collaboration between IoT devices and cloud servers), leading to rigid task decomposition and response latency reaching the second level. Furthermore, existing privacy protection technologies often focus on a single objective—either protecting only the computation process (such as homomorphic encryption) or controlling only access to the result (such as attribute encryption)—lacking a design that is compatible with both process and result-based privacy.

[0003] Current mainstream technologies have three main shortcomings:

[0004] Architectural limitations: Most systems adopt centralized or simple distributed architectures, lacking versatility. For example, while MicroAlgorithm Technology's privacy computing blockchain integrates TEE hardware to improve security, its computation process relies on fixed task allocation and cannot dynamically reorganize subtasks according to data dimensions, making it difficult to support high-concurrency scenarios.

[0005] Privacy-Performance Imbalance: To enhance security, existing solutions often sacrifice efficiency. For example, verification mechanisms based on zero-knowledge proofs (such as zk-SNARKs) consume a large amount of computing resources, resulting in verification latency exceeding 500ms; while on-chain data storage (such as medical information sharing systems) exacerbates network load due to redundant writes, making it difficult to meet real-time requirements.

[0006] The semi-honest model suffers from weak protection: It lacks sufficient protection against semi-honest participants (those who follow the protocol but attempt to steal secrets). Existing models often rely on a single encryption mechanism (such as SHA-256 hashing) and lack multi-level verification of local computational behavior. Once subjected to a collusion attack (such as ≥20% of nodes conspiring), the system's confidence level plummets to below 80%. Real-world testing in the financial sector shows that such flaws can lead to data tampering risks.

[0007] In summary, three major contradictions urgently need to be addressed: the contradiction between general-purpose computing architecture and customized scenarios, the contradiction between end-to-end privacy protection and millisecond-level response, and the contradiction between semi-honest threats and lightweight verification. While there have been attempts in recent years to integrate blockchain and MPC, these efforts have not resolved key technological gaps such as reconfigurable computing engines and dual-verification mechanisms. Summary of the Invention

[0008] This invention addresses the shortcomings of existing technologies by providing a semi-honest privacy protection method and system based on secure multi-party computation using blockchain.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A semi-honest privacy protection method based on blockchain-secure multi-party computation includes the following steps: A general security architecture is constructed and a blockchain identity generation module is configured. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. The off-chain computing layer deploys a reconfigurable computing engine that dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computation on the sub-task unit through a semi-honest security model to obtain the sub-task result. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments. These encrypted fragments are then bound to the blockchain ledger layer to obtain the binding result. The calculation result is obtained by recombining the results of subtasks at the smart contract layer, and the consistency of multiple calculation results is verified based on the binding result.

[0010] As one possible implementation, the blockchain ledger layer adopts a sharded storage structure, including the following steps: The identity information of the participants is stored in the public shard, and the metadata of the computation task is stored in the private shard. The public shard and the private shard are associated with each other through a Merkle root hash. Privately partitioned data uses a threshold decryption mechanism, while publicly partitioned data is based on zero-knowledge authentication, ensuring that the computation process is auditable and identity privacy is controlled.

[0011] As one possible implementation method, the reconfigurable computing engine is achieved through the following steps: Based on the dimensions of the input data and the computational complexity, the system automatically selects either a parallel or serial reassembly strategy. The subtask unit is encapsulated as a lightweight container, and the image hash of the lightweight container is stored on the blockchain. Based on historical performance data, the computation time is predicted, and the allocation of CPU / GPU resources is dynamically adjusted.

[0012] As one possible implementation, the process privacy protection protocol includes the following steps: Distributed homomorphic encryption is used to encrypt the locally encrypted data of each participant and upload the ciphertext to the off-chain computing layer; Homomorphic operations are performed on ciphertext using a secure multi-party computation protocol, including addition or multiplication. The result of homomorphic computation is subjected to differential privacy perturbation to obtain the final result, and Laplace noise is added to the final result. The parameters of the Laplace noise are obtained through smart contract consensus.

[0013] As one possible implementation, the semi-honest security model allows participants to follow a process privacy protection protocol, ensuring that even if a single node is compromised, the original data remains unrecoverable through segmented encryption. This is achieved through a dual authentication mechanism, including the following steps: The participants generate zero-knowledge proofs, and use the results of the zero-knowledge proofs to prove that the participants have performed the computation process in accordance with the protocol, thus obtaining the output commitment values ​​of each participant. The smart contract compares the output commitments of each participant and detects malicious tampering by comparing the results. A semi-honest security model supports local proofs to prevent nodes from forging computation processes, and global verification is performed to identify collusion attacks based on the results of global verification.

[0014] As one possible implementation, the verification of the consistency of multiple calculation results is achieved through hardware acceleration, including the following steps: Run critical computing in a trusted execution environment, isolating potentially malicious operating systems; Improve throughput by using GPU homomorphic acceleration to parallelize homomorphic encryption operations; High-frequency interaction data is temporarily stored in a pre-authorized RAM cache to reduce on-chain read and write latency.

[0015] As one possible implementation, a licensing mechanism for protecting compatible results is also included, comprising the following steps: The semi-honest security model performs calculations on the sub-task units formed by the decomposition of the reconfigurable computing engine to obtain the calculation results, thus obtaining the sub-task results. The results of the subtasks are encrypted using attribute base encryption to generate ciphertext. Data users must possess an attribute key that satisfies the access policy in order to decrypt the data. The rules of the access policy are enforced by the smart contract.

[0016] A semi-honest privacy protection system based on blockchain secure multi-party computation includes: Blockchain Identity Generation Module: Construct a general security architecture and configure a blockchain identity generation module. The blockchain identity generation module generates decentralized identifiers and associates them with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. Off-chain computing module: The off-chain computing layer deploys a reconfigurable computing engine, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing on the sub-task unit through a semi-honest security model to obtain the sub-task result. Reconfigurable Engine Module: The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. Encryption binding module: The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments, and the encrypted fragments are bound to the blockchain ledger layer to obtain the binding result; Result verification module: Obtain the calculation result by recombining the sub-task results through the smart contract layer, and verify the consistency of multiple calculation results based on the binding result.

[0017] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the following method: A general security architecture is constructed and a blockchain identity generation module is configured. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. The off-chain computing layer deploys a reconfigurable computing engine that dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computation on the sub-task unit through a semi-honest security model to obtain the sub-task result. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments. These encrypted fragments are then bound to the blockchain ledger layer to obtain the binding result. The calculation result is obtained by recombining the results of subtasks at the smart contract layer, and the consistency of multiple calculation results is verified based on the binding result.

[0018] A semi-honest privacy protection device based on blockchain secure multi-party computation includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the following method: A general security architecture is constructed and a blockchain identity generation module is configured. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. The off-chain computing layer deploys a reconfigurable computing engine that dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computation on the sub-task unit through a semi-honest security model to obtain the sub-task result. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments. These encrypted fragments are then bound to the blockchain ledger layer to obtain the binding result. The calculation result is obtained by recombining the results of subtasks at the smart contract layer, and the consistency of multiple calculation results is verified based on the binding result.

[0019] Compared with existing technologies, the general architecture of this invention supports heterogeneous computing environments (such as IoT devices and cloud servers working together), blockchain ledger layer sharding storage reduces on-chain load, threshold decryption (such as BLS signature aggregation) prevents single-point data leakage, and zero-knowledge verification (such as zk-SNARKs) allows verifiers to confirm identity legitimacy without obtaining plaintext identity, meeting compliance requirements; the reconfigurable engine adapts to different computing scenarios through dynamic task decomposition (such as splitting matrix operations into vector dot products), while containerization ensures the consistency of the computing environment, hash-based evidence storage prevents tampering, and resource elastic allocation (based on LSTM prediction models) compresses the response time of complex computing tasks; the semi-honest model allows participants to follow process privacy protection protocols, and through segmented encryption (such as Paillier homomorphic encryption combined with Shamir secret sharing), it ensures that even if a single node is compromised, the original data is still irrecoverable. Among them, the process privacy protection protocol achieves a privacy-utility balance: homomorphic encryption protects the computing process, and differential noise, while protecting individual privacy (satisfying ε-differential privacy), ensures a reduction in the error rate of aggregation results. Attached Figure Description

[0020] Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the modules of the system of the present invention. Detailed Implementation

[0021] To clearly illustrate the present invention and make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings, so that those skilled in the art can implement the invention based on the description. The technology of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Example 1: A semi-honest privacy protection method based on secure multi-party computation in blockchain, such as Figure 1 As shown, it includes the following steps: S100. Construct a general security architecture and configure a blockchain identity generation module. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. S200. The off-chain computing layer deploys a reconfigurable computing engine, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing on the sub-task units through a semi-honest security model to obtain the sub-task results. S300. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. S400. The input or output of the subtask unit is segmented and encrypted using a process privacy protection protocol to obtain encrypted segments. The encrypted segments are then bound to the blockchain ledger layer to obtain a binding result. S500 obtains the calculation result by recombining the sub-task results through the smart contract layer, and verifies the consistency of multiple calculation results based on the binding result.

[0023] S100. Construct a general security architecture and configure a blockchain identity generation module. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. By configuring the blockchain identity generation module through the security architecture, enterprises or individuals can apply for attribute credentials as needed. The blockchain ledger layer adopts a sharded storage structure, and the implementation method includes: The identity information of the participants is stored in the public shard, and the metadata of the computation task is stored in the private shard. The two are linked by a Merkle root hash. Private sharded data employs a threshold decryption mechanism (requiring ≥k participants to jointly decrypt), sharded storage reduces on-chain load, and threshold decryption (such as BLS signature aggregation) prevents single points of data leakage; public sharded data is based on zero-knowledge authentication, ensuring auditable computation processes and controlled identity privacy. Zero-knowledge verification (such as zk-SNARKs) allows verifiers to confirm identity legitimacy without obtaining plaintext identity, meeting compliance requirements.

[0024] S200. The off-chain computing layer deploys a reconfigurable computing engine, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing on the sub-task units through a semi-honest security model to obtain the sub-task results. A reconfigurable computing engine is deployed in the off-chain computing layer, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing through a semi-honest security model. The reconfigurable computing engine implementation includes: Configure a dynamic task scheduler: automatically select a parallel / serial reorganization strategy based on the dimension of the input data and the computational complexity (e.g., decompose a 10^6 level matrix multiplication into 10^3 level block operations). Containerization of computing units: Subtasks are encapsulated into lightweight Docker containers, and the container images are stored on the blockchain using hashes; containerization ensures the consistency of the computing environment, and hash storage prevents tampering. The resource elastic allocation module, based on an LSTM prediction model, dynamically adjusts CPU / GPU resource allocation based on historical performance data to predict computation time. This resource elastic allocation (based on an LSTM prediction model) compresses the response time of complex computational tasks.

[0025] S300. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. The semi-honest model allows participants to follow process privacy protection protocols and ensures that the original data is still unrecoverable even if a single node is compromised through segmented encryption; The semi-honest security model is implemented through a dual-verification mechanism, including: Local computation proof: Participants generate zero-knowledge proofs (zk-STARKs) to prove that they performed computation according to the protocol; Global result verification: The smart contract compares the committed values ​​of each participant to detect malicious tampering. Enhanced security: Local proofs prevent nodes from forging computation processes, and global verification identifies collusion attacks; The dual mechanism enables the system to maintain a safe level of confidence even in a 20% semi-honest scenario.

[0026] S400. The input or output of the subtask unit is segmented and encrypted using a process privacy protection protocol to obtain encrypted segments. The encrypted segments are then bound to the blockchain ledger layer to obtain a binding result. The process privacy protection protocol includes: Input phase: Distributed homomorphic encryption is used, and each participant locally encrypts the data {E(pk_i, x_i)}, and the ciphertext is uploaded to the off-chain computing layer; Computation phase: Execute secure multi-party computation protocol to complete homomorphic addition / multiplication operations in ciphertext state; homomorphic encryption (such as the BFV scheme) protects the computation process; Output phase: Differential privacy perturbation is implemented, adding Laplace noise (ε≤0.1) to the final result. The noise parameters are generated by smart contract consensus. Differential noise reduces the error rate of the aggregation result while protecting individual privacy (satisfying ε-differential privacy).

[0027] Achieving a balance between privacy and utility.

[0028] S500 obtains the calculation result by recombining the sub-task results through the smart contract layer, and verifies the consistency of multiple calculation results based on the binding result.

[0029] Specifically, the verification of the consistency of the calculation results is achieved through hardware acceleration, including: Trusted Execution Environment (TEE): Runs critical computing within an Intel SGX or ARM TrustZone, isolating potentially malicious operating systems; TEE ensures key security.

[0030] GPU homomorphic acceleration: Parallelizes homomorphic encryption operations by calling the CUDA library, improving throughput; GPU homomorphic acceleration can improve the efficiency of homomorphic multiplication operations; Off-chain cache channel: High-frequency interactive data is temporarily stored in a pre-authorized RAM cache area, reducing on-chain read / write latency. The cache channel can also reduce network latency.

[0031] As one embodiment of the present invention, the present invention also supports a licensing mechanism that protects against compatibility issues: The calculation results are generated into ciphertext through attribute-based encryption (ABE). Data users need to possess an attribute key that meets the access policy (such as "department=finance&level>5") to decrypt the data. The policy rules are enforced by the smart contract.

[0032] By combining the ABE mechanism of blockchain, fine-grained result control can be achieved, supporting dynamic policy updates (such as adding time constraints), which can reduce the response time for permission changes.

[0033] Example 2: A semi-honest privacy protection system based on secure multi-party computation using blockchain, such as Figure 2 As shown, it includes: Blockchain identity generation module 100: Constructs a general security architecture and configures the blockchain identity generation module. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. Off-chain computing module 200: The off-chain computing layer deploys a reconfigurable computing engine, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing on the sub-task unit through a semi-honest security model to obtain the sub-task result. Reconfigurable engine module 300: The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. Encryption binding module 400: Employs a process privacy protection protocol to perform segmented encryption on the input or output of the subtask unit to obtain encrypted fragments, and binds the encrypted fragments to the blockchain ledger layer to obtain a binding result; Result verification module 500: Obtains the calculation result by recombining the sub-task results through the smart contract layer, and verifies the consistency of multiple calculation results based on the binding result.

[0034] Example 3: A semi-honest privacy protection device based on blockchain-secure multi-party computation is disclosed. This device can be a server or a mobile terminal. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and the database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database contains all data of the computer device. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a semi-honest privacy protection method based on blockchain-secure multi-party computation. It stores verifiable and secure machine learning models for node behavior pattern analysis; The media is partitioned into a public area and a secure area. Access permissions for the secure area are controlled by a physical write-protection switch.

[0035] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0036] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. It will be apparent to those skilled in the art that various modifications can be made to the above embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A semi-honest privacy protection method based on blockchain secure multi-party computation, characterized in that, Includes the following steps: A general security architecture is constructed and a blockchain identity generation module is configured. The blockchain identity generation module generates a decentralized identifier and associates it with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. The off-chain computing layer deploys a reconfigurable computing engine that dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computation on the sub-task unit through a semi-honest security model to obtain the sub-task result. The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments. These encrypted fragments are then bound to the blockchain ledger layer to obtain the binding result. The calculation result is obtained by recombining the results of subtasks at the smart contract layer, and the consistency of multiple calculation results is verified based on the binding result.

2. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, The blockchain ledger layer adopts a sharded storage structure, including the following steps: The identity information of the participants is stored in the public shard, and the metadata of the computation task is stored in the private shard. The public shard and the private shard are associated with each other through a Merkle root hash. Privately partitioned data uses a threshold decryption mechanism, while publicly partitioned data is based on zero-knowledge authentication, ensuring that the computation process is auditable and identity privacy is controlled.

3. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, The reconfigurable computing engine is implemented through the following steps: Based on the dimensions of the input data and the computational complexity, the system automatically selects either a parallel or serial reassembly strategy. The subtask unit is encapsulated as a lightweight container, and the image hash of the lightweight container is stored on the blockchain. Based on historical performance data, the computation time is predicted, and the allocation of CPU / GPU resources is dynamically adjusted.

4. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, The process privacy protection agreement includes the following steps: Distributed homomorphic encryption is used to encrypt the locally encrypted data of each participant and upload the ciphertext to the off-chain computing layer; Homomorphic operations are performed on ciphertext using a secure multi-party computation protocol, including addition or multiplication. The result of homomorphic computation is subjected to differential privacy perturbation to obtain the final result, and Laplace noise is added to the final result. The parameters of the Laplace noise are obtained through smart contract consensus.

5. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, The semi-honest security model allows participants to follow a process privacy protection protocol, ensuring that even if a single node is compromised, the original data remains unrecoverable through segmented encryption. This is achieved through a dual-verification mechanism, including the following steps: The participants generate zero-knowledge proofs, and use the results of the zero-knowledge proofs to prove that the participants have performed the computation process in accordance with the protocol, thus obtaining the output commitment values ​​of each participant. The smart contract compares the output commitments of each participant and detects malicious tampering by comparing the results. A semi-honest security model supports local proofs to prevent nodes from forging computation processes, and global verification is performed to identify collusion attacks based on the results of global verification.

6. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, The verification of the consistency of multiple calculation results is achieved through hardware acceleration, including the following steps: Run critical computing in a trusted execution environment, isolating potentially malicious operating systems; Improve throughput by using GPU homomorphic acceleration to parallelize homomorphic encryption operations; High-frequency interaction data is temporarily stored in a pre-authorized RAM cache to reduce on-chain read and write latency.

7. The semi-honest privacy protection method based on blockchain secure multi-party computation according to claim 1, characterized in that, It also includes a licensing mechanism for compatibility result protection, comprising the following steps: The semi-honest security model performs calculations on the sub-task units formed by the decomposition of the reconfigurable computing engine to obtain the calculation results, thus obtaining the sub-task results. The results of the subtasks are encrypted using attribute base encryption to generate ciphertext. Data users must possess an attribute key that satisfies the access policy in order to decrypt the data. The rules of the access policy are enforced by the smart contract.

8. A semi-honest privacy protection system based on blockchain secure multi-party computation, characterized in that, include: Blockchain Identity Generation Module: Construct a general security architecture and configure a blockchain identity generation module. The blockchain identity generation module generates decentralized identifiers and associates them with a verifiable credential template library to obtain the user's attribute credentials. The general security architecture includes a blockchain ledger layer, a smart contract layer, and an off-chain computing layer. Off-chain computing module: The off-chain computing layer deploys a reconfigurable computing engine, which dynamically decomposes the computing logic into sub-task units according to the task type. Each participant performs local computing on the sub-task unit through a semi-honest security model to obtain the sub-task result. Reconfigurable Engine Module: The reconfigurable engine is based on attribute credentials, adapts to different computing scenarios through dynamic task decomposition, defines the scope of credential disclosure through a programmable policy language, and achieves selective disclosure through zero-knowledge proofs and disclosure scope. Encryption binding module: The input or output of the subtask unit is encrypted in segments using a process privacy protection protocol to obtain encrypted fragments, and the encrypted fragments are bound to the blockchain ledger layer to obtain the binding result; Result verification module: Obtain the calculation result by recombining the sub-task results through the smart contract layer, and verify the consistency of multiple calculation results based on the binding result.

9. A semi-honest privacy protection device based on blockchain secure multi-party computation, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that the processor, when executing the computer program, implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a program, wherein the computer-readable storage medium stores a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

Images