Blockchain-based secure multi-party computation financial data privacy protection system

Abstract

The application discloses a financial data privacy protection system based on a blockchain secure multi-party computation, and belongs to the technical field of data protection, and specifically comprises the following steps: a data preprocessing module is used for segmenting original financial data and adding a unique watermark identifier; a distributed ledger is used for storing an encrypted logical data segment hash digest and a watermark mapping relationship, and a digital signature and a timestamp are fused to ensure reliability; a computing node cluster receives data through a special interface, dynamically forms a computing group according to business rules, and outputs an encrypted calculation result; a zero-knowledge proof verifier and the computing node cluster are bidirectionally authenticated, and data possession is verified; an encryption proxy gateway processes the output result of the computing node cluster, and completes format conversion and ciphertext superposition; a cross-chain routing module constructs a virtual overlay layer, performs protocol conversion and three-way handshake verification, realizes cross-chain transmission of the encrypted result, and a heterogeneous chain protocol converter and a verification component guarantee transmission safety; and the application comprehensively guarantees the privacy and safety of financial data, and improves data processing efficiency and reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data protection, and in particular to a financial data privacy protection system based on blockchain secure multi-party computation. BACKGROUND

[0002] In the wave of digital transformation, financial data as the core asset of enterprise operation, its privacy protection and safe sharing has become the focus of global enterprises. With the development of blockchain, secure multi-party computation (MPC) and other cryptography technologies, a new technical path is provided to solve the privacy protection problem in cross-organizational data collaboration. The traditional centralized data storage and processing mode faces challenges such as single point of failure, data leakage risk, etc., and emerging distributed computing technology is reshaping the paradigm of inter-enterprise data collaboration.

[0003] At present, the financial data sharing system based on blockchain mainly adopts two technical routes: one is to realize distributed storage and authorized access of data through alliance chain, and each participant stores encrypted data on the chain and defines access rules through smart contract; the second is to combine secure multi-party computation technology to realize multi-party collaborative computation without revealing the original data. For example, some systems use homomorphic encryption technology to allow computation on ciphertext, and the decrypted result is equivalent to the result of computation on plaintext; some systems use secret sharing scheme to divide data into multiple fragments and store them in different computing nodes.

[0004] However, the existing technical solutions still have significant deficiencies. On the one hand, in the data preprocessing stage, most systems only use a single encryption strategy, which is difficult to cope with complex and variable security threats, and lack effective data lifecycle management mechanisms, resulting in security risks of residual expired data. On the other hand, the collaboration efficiency of computing nodes in the secure multi-party computation process is low, and the trusted verification mechanism of the calculation result is imperfect, which is easy to cause disputes over the calculation result. These technical bottlenecks need to be broken through to build a more secure, efficient and trusted financial data privacy protection system. SUMMARY

[0005] The purpose of the present application is to provide a financial data privacy protection system based on blockchain secure multi-party computation, which solves the following technical problems:

[0006] The existing technical solutions in the data preprocessing stage, most systems only use a single encryption strategy, which is difficult to cope with complex and variable security threats, and lack effective data lifecycle management mechanisms, resulting in security risks of residual expired data. And the collaboration efficiency of computing nodes in the secure multi-party computation process is low, and the trusted verification mechanism of the calculation result is imperfect, which is easy to cause disputes over the calculation result.

[0007] The purpose of the present application can be achieved by the following technical solutions:

[0008] The financial data privacy protection system based on blockchain secure multi-party computation comprises a data preprocessing module, a distributed ledger, a computing node cluster, a zero-knowledge proof verifier, an encryption proxy gateway and a cross-chain routing module, and specifically:

[0009] The data preprocessing module is used for splitting original financial data into at least two logical data segments, adding a unique watermark identifier to each logical data segment, and performing asymmetric encryption on the logical data segment with the unique watermark identifier by using an encryption key independently generated by a participant to obtain an encrypted logical data segment of the participant.

[0010] The distributed ledger is used for storing a hash digest of the encrypted logical data segment and a mapping relationship between the logical data segment and the watermark identifier, wherein the hash digest is generated by fusing a digital signature of the participant and a timestamp in the generation process.

[0011] The computing node cluster is used for receiving encrypted logical data segments of different participants through a special physical interface, dynamically forming a multi-party computing group based on a preset smart contract business computing rule, performing computation on the encrypted logical data segments and outputting an encrypted computation result.

[0012] The zero-knowledge proof verifier is used for establishing a two-way authentication connection with the computing node cluster, and verifying the legal possession of the logical data segment by the participant before executing the computing task.

[0013] The encryption proxy gateway is used for performing format conversion and ciphertext superposition on the encrypted computation result output by the computing node cluster.

[0014] The cross-chain routing module is used for performing transmission of the encrypted computation result between different chains, and comprises a heterogeneous chain protocol converter and a data integrity checking component.

[0015] As a further scheme of the application, the operation process of the data preprocessing module specifically comprises:

[0016] In the data segmentation stage, the amount field and the business entity identifier in the original financial data are separated and processed by a financial data type identification rule to generate a first logical data segment containing pure numerical values and a second logical data segment containing business-related information.

[0017] The unique watermark identifier is spliced from a blockchain identity certificate fingerprint obtained by the participant during registration and a hash value of a data generation time window, and each watermark identifier is associated with a public key certificate copy of the participant and stored in the distributed ledger.

[0018] The layered encryption strategy is executed in the encryption process, the logical data segment is signed by using the private key of the participant first, and then is encrypted by using the public key of the receiver, so that an encrypted logical data segment with a nested ciphertext structure is formed; the encrypted logical data segment is verified by the local compliance checking engine before being sent to the computing node cluster, whether it conforms to the preset data desensitization rule is verified, and an audit mark is generated and written into the blockchain.

[0019] As a further scheme of the application, the data preprocessing module is used for attaching life cycle metadata to each encrypted logical data segment, and the metadata includes a maximum retention period, a list of allowed use scenarios and an associated business contract identifier.

[0020] The data preservation monitor deployed in the blockchain periodically scans the account book record, and sends an update reminder to the relevant participants in advance for the logical data segment approaching the expiration period.

[0021] When the smart contract detects a preset business completion condition, a data destruction instruction is triggered, and all copies of the data to be destroyed are synchronously erased by the cooperation mechanism of the computing node cluster; the destruction process adopts a cryptography forgetting technology, deletes the data content, and clears the physical residual information in the storage medium by an overwrite writing mode, and an independent audit node issues a verifiable destruction certificate.

[0022] As a further scheme of the application, the working mode of the computing node cluster is:

[0023] When receiving the encrypted logical data segment, each computing node verifies the validity period and revocation status of the sender's identity certificate through a secure isolated hardware environment, and constructs a temporary computing sandbox in the memory;

[0024] In the process of forming the multi-party computing group, the computing nodes are dynamically elected according to the business computing rules defined in the smart contract, the computing nodes collect the computing parameters of each participant and generate a task allocation topology graph; when executing the computing task, each computing node exchanges the necessary relay parameters in the temporary computing sandbox, transmits the intermediate computing results by secret sharing, and generates a Merkle tree evidence in real time and uploads it to the blockchain;

[0025] After the completion of the computing task, the computing nodes send the ciphertext digest of the encrypted computing result to all participants for joint signature confirmation, and after the joint signature confirmation is completed, the subsequent processing of the encryption proxy gateway is triggered.

[0026] As a further scheme of the application, a special log collector is deployed in the computing node cluster to capture key events in the computing process in real time, including data receiving time, computing step execution state and abnormal interruption record.

[0027] The collected log data is encrypted and desensitized to generate compressed data packets containing time sequence fingerprints, and the compressed data packets are uploaded to multiple blockchain nodes for storage through a random polling mechanism;

[0028] The log analyzer running in the blockchain network periodically performs correlation analysis on the dispersed log segments, detects potential security threat patterns, and feeds the analysis results to the risk control centers of the participants in an encrypted form; all access operations of the audit logs need to be verified through zero-knowledge proof, and the query results are returned after being processed by differential privacy technology.

[0029] As a further scheme of the application, the execution logic of the smart contract is:

[0030] The contract instance deployed in the blockchain network includes a dynamic access strategy engine, which automatically adjusts the access control rules according to the real-time acquired participant reputation score, data sensitivity level and environmental risk indicators;

[0031] The pre-defined calculation task trigger condition in the smart contract is bound to the real business event, and when a specific business state change reported by the off-chain oracle is detected, a multi-party calculation request is automatically initiated;

[0032] All state change records generated during the execution of the smart contract are stored in the form of encrypted logs, and the log entries are connected through a forward secure chain structure; the smart contract also includes an automated dispute resolution mechanism, which detects abnormal calculation results, calls backup redundant calculation nodes to re-execute the task, and determines the final effective result by comparing multiple calculation copies.

[0033] As a further scheme of the application, the running mechanism of the zero-knowledge proof verifier specifically includes:

[0034] The dynamic challenge parameter is generated in the verification phase, which includes the watermark identifier in the logical data segment required by the participant to prove that the mapping relationship is consistent with the blockchain storage, and the generation process of the encryption key conforms to the key management strategy submitted during pre-registration, and the logical data segment is verified not to exceed the preset valid use period;

[0035] An interactive proof protocol is executed in the verification process, the zero-knowledge proof verifier sends a challenge instruction containing a random salt value to the participant, and the participant returns the proof data after a specific transformation within a specified time;

[0036] The verification result is divided into three levels of credibility according to the verification strength, different credibility levels correspond to different calculation task permissions, and the calculation task permissions are written into the execution condition judgment logic of the smart contract in an encrypted form.

[0037] As a further scheme of the application, the processing process of the encryption proxy gateway is:

[0038] After receiving the encrypted computing result output by the cluster of computing nodes, a ciphertext format standardization process is performed to convert the structural differences generated by different participant encryption algorithms into a unified standard format;

[0039] In the ciphertext superposition stage, a random noise factor is injected into the encrypted computing result after standardization processing through blinding processing technology, and the final result after superposition contains a linear combination of the encrypted computing result and the noise factor;

[0040] The original encryption hierarchy is preserved during the format conversion process, and the participant signature information is converted into a verifiable nested certificate chain form;

[0041] The processed encrypted computing result is split into multiple data packets, which are distributed to each relevant blockchain computing node through a polling mechanism, and each data packet is attached with an integrity check code for the receiver to verify the transmission integrity.

[0042] As a further scheme of the present application, the process of the cross-chain routing module for data transmission is:

[0043] A virtual overlay layer covering multiple blockchain networks is established, and the virtual overlay layer contains a relay chain structure for storing metadata indexes of cross-chain interactions;

[0044] When the encrypted computing result needs to be transmitted across the chain, the cluster of verification computing nodes on the source chain performs fragmentation processing on the encrypted computing result to generate a smart contract transaction package containing cross-chain routing instructions;

[0045] The heterogeneous chain protocol converter parses the control instructions in the smart contract transaction package and converts the control instructions into a transaction format recognizable by the target chain while preserving the original encryption hierarchy unchanged;

[0046] The data integrity verification component implements three-way handshake verification during transmission, including source chain outbound verification, relay chain transfer verification, and target chain inbound verification, and each verification needs to obtain the signature confirmation of the majority of nodes on the relevant chain.

[0047] The beneficial effects of the present application are:

[0048] The data preprocessing module of the application adopts logical segmentation, hierarchical encryption and life cycle management technology, solves the security hidden danger of single encryption strategy, separates the amount field and the business entity identifier through the financial data type identification rule, combines the participant private key signature and the receiver public key encryption to form a nested ciphertext structure, and realizes the whole cycle safety control through the data preservation monitor and the cryptography forgetting technology; the computing node cluster constructs a secure isolated hardware environment and a temporary computing sandbox, dynamically elects and coordinates the computing nodes to generate a task distribution topology graph, adopts secret sharing to transfer intermediate results and generates a Merkle tree evidence, and improves the multi-party computing efficiency and result credibility; the zero-knowledge proof verifier realizes three-level trusted level verification through dynamic challenge parameters and interactive proof protocol, ensures the legal possession right of the participants and the timeliness of the data; the encryption proxy gateway processes the ciphertext format standardization and blind superposition technology, retains the original encryption hierarchy and enhances the data confusion degree; the cross-link routing module constructs a virtual overlay layer and a relay chain structure, combines a protocol converter and a three-way handshake verification mechanism, solves the heterogeneous chain compatibility and data integrity problem, and realizes the safe and reliable cross-chain transmission. BRIEF DESCRIPTION OF DRAWINGS

[0049] The application will be further described below in conjunction with the drawings.

[0050] Figure 1 It is a module schematic diagram of the application. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the application will be clearly and completely described below in conjunction with the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0052] Please refer to Figure 1 The application is a financial data privacy protection system based on blockchain secure multi-party computation, which comprises a data preprocessing module, a distributed ledger, a computing node cluster, a zero-knowledge proof verifier, an encryption proxy gateway and a cross-link routing module, specifically:

[0053] The data preprocessing module is the first processing link for the original financial data entering the system, responsible for the secure segmentation and encryption of data. According to the data characteristics and security requirements, it divides the original financial data into at least two logical data segments, such as separating the amount field and the business entity identifier, generating a first logical segment containing pure numerical values and a second logical segment containing business-related information, and implementing data classification protection. Each logical data segment will add a unique watermark identifier, which is composed of the participant's blockchain identity certificate fingerprint and the hash value of the data generation time window, used for data tracing and rights confirmation. Then, use the encryption key generated independently by the participant for asymmetric encryption. When encrypting, first sign the logical data segment with the participant's private key, then encrypt it with the receiver's public key, forming a nested ciphertext structure, ensuring data confidentiality and verifiable source. After encryption, the logical data segment is checked by the local compliance checking engine to confirm that it meets the preset data desensitization rules, and an audit mark is generated and written to the blockchain, providing data support for auditing and supervision.

[0054] The distributed ledger is the core of system data storage and management. It stores the hash digest of encrypted logical data segments and the watermark identifier mapping relationship. When generating the hash digest, the participant's digital signature and timestamp are integrated. The participant's digital signature ensures the reliability of the data source, and the timestamp gives the data a unique time dimension, making the hash digest tamper-proof and traceable, achieving distributed secure storage of data, and laying the foundation for data verification, tracing, and other operations.

[0055] The computing node cluster is the execution unit of the system to realize multi-party computation. It receives encrypted logical data segments from different participants through a special physical interface with security protection mechanism, preventing external illegal attacks and data theft. After receiving the data, the computing node cluster dynamically forms a multi-party computation group according to the preset business computing rules, selects the computing nodes through the election algorithm according to the business rules in the smart contract. After the computing nodes collect the computing parameters of each participant, a multi-dimensional evaluation model is first constructed to quantify the computing capacity, network conditions, security level and historical performance of each node. Then, the graph theory algorithm is used to decompose the computing task into data preprocessing, ciphertext computation, result aggregation and other sub-task nodes. Based on the evaluation results, a weighted directed graph is generated, where the nodes represent computing resources, the edges represent data flow and cooperation relationship, and the weights reflect task priority and resource demand. Through the simulated annealing algorithm to iteratively optimize the topology structure, the data transmission delay and redundant computation are minimized, and finally a task allocation topology graph is formed, which includes task scheduling path, data encryption transmission channel and exception handling link, to ensure the orderly execution of the computing task. During the computing process, each computing node exchanges relay parameters in an encrypted state, transmits intermediate computing results using secret sharing method, and generates Merkle tree evidence in real time for interaction record and uploads to the blockchain for verification of the computing process. After the completion of the computing task, the coordinating computing nodes send the ciphertext digest of the encrypted computing result to all participants for joint signature confirmation to ensure the credibility of the computing result.

[0056] The zero-knowledge proof verifier is used to ensure the legality of the participant's data use. It establishes a two-way authentication connection with the computing node cluster and verifies the participants before the execution of the computing task. During the verification, dynamic challenge parameters are generated, requiring the participants to prove that the watermark identifier in their logical data segment is consistent with the mapping relationship stored in the blockchain, the encryption key generation process conforms to the pre-registered key management strategy, and the logical data segment is within the valid use period. The verification adopts an interactive proof protocol, and the verifier sends challenge instructions containing random salt values to the participants, who need to return proof data transformed by a specific transformation within a specified time. According to the verification strength, three levels of trust level are divided, and different levels correspond to different computing task permissions. The permission information is written into the smart contract execution condition judgment logic in an encrypted form to realize fine-grained management of participant permissions.

[0057] The encryption proxy gateway optimizes the processing of the encrypted calculation results output by the cluster of computing nodes. After receiving the primary encrypted results, it first performs standardization processing on the ciphertext format, unifying the structures generated by different participants due to differences in encryption algorithms. In the ciphertext superposition stage, random noise factors are injected through blinding processing technology, so that the final result is a linear combination of the original calculation result and noise data, enhancing data confidentiality. The original encryption hierarchy is preserved during format conversion, and the participant signature information is converted into a verifiable nested certificate chain form. The processed encrypted results are split into multiple data packets, which are distributed to related blockchain computing nodes through a polling mechanism. Each data packet is accompanied by an integrity check code to ensure data transmission integrity.

[0058] The cross-link routing module enables secure transmission of encrypted calculation results between different blockchain networks. It builds a virtual overlay layer covering multiple blockchain networks, which contains a relay chain structure for storing cross-chain interaction metadata indexes, providing guidance for cross-chain transmission. When cross-chain transmission is required, the cluster of verification computing nodes on the source chain processes the data shards to generate a smart contract transaction package containing cross-chain routing instructions. The protocol converter parses the control instructions of the source chain transaction package and converts them into a transaction format recognizable by the target chain, while preserving the original encryption hierarchy, completing protocol conversion between heterogeneous chains. The data integrity verification component performs three-way handshake verification in the transmission, including source chain outbound verification, relay chain transfer verification, and target chain inbound verification. Each verification requires the signature confirmation of a majority of computing nodes on the relevant chain to ensure the security of cross-chain transmission.

[0059] In a preferred embodiment of the present application, the operation flow of the data preprocessing module specifically includes:

[0060] In the data segmentation stage, the financial data type recognition rules built-in the data preprocessing module use a combination of regular expressions and semantic analysis to deeply analyze the original financial data. Through the pre-set amount field characteristic pattern, the amount value in the data is accurately identified and separated from the business entity identifier. For example, for financial data containing customer order amount and customer number, the order amount value is extracted to generate a first logical data segment containing pure numerical values, and the customer number and other business-related information form a second logical data segment. This separation processing enables the classification and protection of data with different sensitivity levels, reducing the potential risks of data leakage.

[0061] In the watermark identification generation link, the construction of unique watermark identification is based on cryptographic hash algorithm and digital certificate technology. Participants will obtain a blockchain identity certificate when they register with the system, and the certificate fingerprint is unique. The data preprocessing module splices the fingerprint with the hash value of the data generation time window to form a unique watermark identification. For example, the SHA-256 algorithm is used to hash the data generation time window, and then combined with the identity certificate fingerprint in a specific format. Each watermark identification is stored in association with a copy of the participant's public key certificate in the distributed ledger, which can be used for data tracing, ownership verification, and operation auditing in the future, ensuring the traceability of the data flow process.

[0062] The encryption process strictly implements a layered encryption strategy. First, the participant uses their own private key to digitally sign the logical data segment. The signing process follows the ECDSA (Elliptic Curve Digital Signature Algorithm) standard, which ensures the authenticity and integrity of the data source by signing the hash value of the logical data segment. After signing, the public key of the recipient is used for asymmetric encryption using mature encryption algorithms such as RSA or ECC, forming a nested ciphertext structure. This structure ensures the confidentiality of the data during transmission and storage, and also verifies the ownership of the data. After encryption, the logical data segment is sent to the local compliance check engine before being sent to the computing node cluster. The engine has a pre-set data desensitization rule library, which detects and processes sensitive information in the logical data segment through pattern matching and compliance verification algorithms, such as anonymizing names, ID numbers, etc. After passing the check, an audit mark is generated, containing information such as data processing time and operator identification, which is written into the blockchain for future supervision and auditing.

[0063] In a preferred case of the present embodiment, the data preprocessing module further enhances the data lifecycle management capability. Life cycle metadata is attached to each encrypted segment, which is stored in a structured format such as JSON or XML, containing key information such as maximum retention period, allowed use scenarios, and associated business contract identifier. For example, a certain financial logical data segment has a maximum retention period of 3 years, and the allowed use scenarios are limited to internal financial analysis and tax reporting, and the associated business contract identifier points to a specific cooperation project contract.

[0064] The data preservation monitor deployed in the blockchain is implemented based on smart contracts and runs automatically. It periodically scans the distributed ledger records at preset intervals (such as daily, weekly), compares the current time of the logical data segment with the maximum retention period, and automatically generates an update reminder for the logical data segment approaching the expiration period (such as 30 days before expiration), and sends it to the relevant participants through a secure communication channel. The reminder contains information such as logical data segment identifier and remaining validity period, so that participants can handle it in a timely manner.

[0065] When the smart contract monitors that the preset business completion condition is met, a data destruction instruction is triggered automatically. The instruction is transmitted and executed through the consensus mechanism and the cooperation mechanism of the computing node cluster, so as to ensure that the data copies stored on the computing nodes are synchronized and erased. The destruction process adopts a cryptographic forgetting technology, specifically, the original financial data storage area is covered by using randomly generated binary data for multiple times in a covering writing manner, so as to eliminate the physical residual information in the storage medium. After the operation is completed, an independent audit node verifies the data destruction process and result based on a zero-knowledge proof technology, and issues a verifiable destruction certificate containing verification time, verifier signature and the like, so as to ensure the effectiveness and compliance of the data destruction.

[0066] In another preferred embodiment of the present application, the working mode of the computing node cluster is:

[0067] When receiving the encrypted logical data segment, each computing node uses a secure isolated hardware environment, such as a trusted execution environment (TEE) technology, to provide a physically and logically isolated secure space for data reception and processing. In this environment, the computing node strictly verifies the validity period and revocation status of the sender's identity certificate. Through real-time interaction with the certificate authority (CA) or local cached certificate status information, the validity of the certificate is ensured. At the same time, a temporary computing sandbox is constructed in the memory, which is based on the memory management and virtualization technology of the operating system, and allocates independent memory space for each computing task to prevent memory conflict and data leakage between different tasks.

[0068] In the process of forming the multi-party computing group, the computing node cluster dynamically elects computing nodes according to the business rules predefined in the smart contract using a distributed election algorithm. For example, a variant of the Byzantine fault tolerance (BFT) algorithm can be used to ensure that reliable coordination computing nodes can be elected even in the presence of partial computing node failures or malicious behavior. The coordination computing node is responsible for collecting the computing parameters of each participant, including but not limited to the encryption key of the logical data segment, the configuration information of the computing task, etc. The coordination computing node generates a task allocation topology graph according to these parameters, which is based on graph theory and optimization algorithms, considering factors such as computing power and network bandwidth of the computing nodes, to reasonably plan the allocation of computing tasks among the computing nodes, so as to maximize the computing efficiency.

[0069] In the process of performing computing tasks, each computing node interacts with data in an encrypted state. Through secret sharing, intermediate computing results are securely transmitted between computing nodes. Secret sharing uses cryptographic algorithms such as the Shamir secret sharing scheme to divide a secret value into multiple shares, with each computing node holding a portion of the shares. Only when a sufficient number of computing nodes come together can the original secret value be restored. During data interaction, interaction records generate Merkle tree evidence in real time and upload to the blockchain. Merkle tree is based on hash algorithm, which combines the hash values of a series of data blocks layer by layer to form a binary tree structure, and the hash value of the root node represents the integrity of the entire data set. By uploading Merkle tree evidence to the blockchain, the traceability and tamper resistance of the computing process are ensured.

[0070] After the completion of the computing task, the coordinating computing node sends the ciphertext digest of the final result to all participants for joint signature confirmation. Participants use their own private keys to sign the ciphertext digest, and the signature process follows the Digital Signature Standard (DSS). After joint signature confirmation, the subsequent processing of the encrypted proxy gateway is triggered. Homomorphic encryption technology allows specific computing operations to be performed on ciphertext without decryption, thereby protecting data privacy.

[0071] In a preferred case of the present embodiment, the computing node cluster further enhances the monitoring and management capabilities of the computing process. By deploying a dedicated log collector, key events in the computing process are captured in real time. These events include data reception time, computing step execution status, and abnormal interruption records. The log collector uses a high-performance log recording library and event-driven architecture to ensure accurate recording of each key event without affecting computing performance.

[0072] The collected log data is first encrypted using symmetric encryption algorithms such as AES to ensure the confidentiality of the log content during transmission and storage. Then, desensitization is performed through data replacement, masking, and other techniques to remove sensitive information such as participant identities and specific financial data from the logs. The processed log data generates compressed data packets containing time sequence fingerprints, which are generated based on timestamps and data hash values to uniquely identify each log record. Through a random polling mechanism, compressed data packets are uploaded to multiple blockchain computing nodes for storage. The random polling mechanism ensures uniform distribution of log data in the blockchain network, improving storage reliability and attack resistance.

[0073] The log analyzer running in the blockchain network periodically performs correlation analysis on the dispersed log fragments. The log analyzer employs big data analysis techniques and machine learning algorithms such as clustering analysis, anomaly detection, etc. to detect potential security threat patterns. For example, by analyzing abnormal fluctuations in data reception time, frequent errors in step execution status, etc., it can be found that there may be network attacks or system failures. The analysis results are fed back to the risk control center of each participant in an encrypted form to ensure the security of information transmission. All access operations of audit logs need to be verified by zero-knowledge proof, which allows the prover to prove that a statement is correct without revealing any useful information to the verifier. The query result is returned after being processed by differential privacy technology, which protects sensitive information in the query result by adding noise, etc. while ensuring the usability of the query result.

[0074] In another preferred case of the embodiment, the execution logic of the smart contract is:

[0075] The execution logic of the smart contract provides strong support for the automation and intelligent operation of the entire system. The contract instance deployed in the blockchain network contains a dynamic access strategy engine that automatically adjusts access control rules based on real-time acquired participant reputation scores, data sensitivity levels, and environmental risk indicators. The reputation score is obtained by evaluating the historical behavior and transaction records of the participants; the data sensitivity level is determined according to the type and importance of the data; and the environmental risk indicator considers network security conditions, system stability, etc. The dynamic access strategy engine uses rule engine technology and decision tree algorithms to dynamically adjust access control rules based on these real-time indicators, ensuring that only authorized participants can access and operate data under appropriate conditions.

[0076] The pre-defined computing task trigger conditions in the contract are closely bound to the business events of the physical world. When a specific business state change reported by the off-chain oracle is detected, a multi-party computing request is automatically initiated. The off-chain oracle obtains business data in real time through an interface with external data sources and converts it into a format recognizable by the blockchain. For example, when the status of a sales order changes from "created" to "shipped", the oracle sends this status change information to the blockchain, triggering the relevant financial data computing task.

[0077] All state change records generated during the contract execution process are stored in the form of encrypted logs, and the log entries are connected through a forward secure chain structure. The forward secure chain structure is based on the cryptographic hash chain technology, and the hash value of each log entry depends on the hash value of the previous entry, ensuring the orderliness and tamper resistance of the log. The contract also contains an automated dispute resolution mechanism that calls backup redundant computing nodes to re-execute tasks when abnormal computing results are detected. Redundant computing nodes use hot backup or cold backup strategies to ensure that they can take over tasks in a timely manner when the main computing node fails or the computing result is abnormal. By comparing multiple computing copies to determine the final valid result, a majority vote or consensus algorithm is used to ensure the accuracy and reliability of the result.

[0078] In another preferred embodiment of the present application, the operation mechanism of the zero-knowledge proof verifier specifically includes:

[0079] In the verification phase, the zero-knowledge proof verifier dynamically generates challenge parameters based on a secure random number generation algorithm. The parameters verify the watermark identification, encryption key compliance, and usage period of the logical data segment: for the watermark identification, the participants are required to compare and analyze the hash and digital certificate to prove that the watermark in the logical data segment is consistent with the mapping relationship stored in the distributed ledger, ensuring traceability of the data source; for the encryption key, it verifies whether the generation process meets the key management strategy submitted to the blockchain during pre-registration, such as key length and generation algorithm meeting security standards; and according to the life cycle metadata attached to the logical data segment, it checks whether the data is within the preset valid usage period to prevent expired data from participating in illegal computing.

[0080] The verification process performs an interactive proof protocol based on the zero-knowledge proof principle. The verifier sends a challenge instruction containing a random salt value to the participant, which is generated by a cryptographically secure pseudo-random number generator to increase the randomness and unpredictability of the verification process. After receiving the instruction, the participant needs to generate proof data using the local encryption key and logical data segment within the specified time through specific cryptographic transformation algorithms, such as discrete logarithm, hash collision, and polynomial commitment. The discrete logarithm uses mathematical structures such as elliptic curves to enable the prover to prove possession of the private key corresponding to the public key without revealing the private key, such as signature compliance verification. Hash collision uses the one-way and collision resistance of hash functions to convert data or random salt values into unique fingerprints for generating unpredictable challenge parameters or compressed proof data, such as watermark identification hash comparison. Polynomial commitment allows the prover to "commit" a polynomial and verify complex logic without disclosing the polynomial itself by submitting specific point evaluations or coefficient relationships, such as data validity period range proof. All of the above are achieved through cryptographic one-way puzzles to realize the "verification without revealing details" feature. The verifier verifies the validity of the proof data without obtaining the original financial data to confirm the participant's legal possession of the logical data segment.

[0081] The verification result is divided into high, medium, and low levels of trust levels according to the verification strength. The high-level trust level requires the participant to pass the watermark identification, key compliance, and usage period verification completely, and the verification time of the proof data is less than the threshold. The medium-level trust level allows passing in the case of minor flaws in some secondary verification items that do not affect the overall security. The low-level trust level is suitable for scenarios that only pass the basic identity verification. Different trust levels correspond to different computing task permissions, such as high-level trust level participating in core financial data calculation and low-level trust level only performing auxiliary data processing. The permission information is encrypted by the AES-256 encryption algorithm and written into the execution condition judgment logic of the smart contract to ensure the security and dynamics of the permission management.

[0082] In another preferred embodiment of the present application, the processing process of the encryption proxy gateway is:

[0083] After receiving the primary encryption result output by the computing node cluster, the encryption proxy gateway starts the ciphertext format standardization process. To address the structural differences caused by the use of various encryption algorithms such as RSA and ECC by different participants, the format parsing and conversion engine is used to convert the ciphertext into a JSON-Web Encryption (JWE) format that conforms to the system specification. During the conversion process, the original encryption hierarchy is strictly preserved to ensure data confidentiality.

[0084] In the ciphertext superposition stage, the blinding processing technology is used to further confuse the data. Through the noise injection algorithm, random noise factors conforming to the Gaussian distribution are added to the original calculation results, so that the final result after superposition presents a linear combination of the original calculation result and the noise data. The noise parameter is dynamically adjusted according to the data sensitivity level, and higher intensity noise is injected for high sensitive data, which greatly improves the data confidentiality while ensuring the availability of the calculation result.

[0085] In the format conversion process, the encryption proxy gateway converts the participant signature information into a verifiable nested certificate chain form. Based on the X.509 certificate standard, the signature information is hierarchically nested with the participant's public key certificate, certificate authority signature and other information to form a traceable and verifiable certificate chain structure, which facilitates subsequent auditing of data sources and operation permissions.

[0086] The processed encrypted result is split into multiple data packets of appropriate size by the data packet splitting module according to the data size and network transmission characteristics. Through the polling mechanism, the data packets are distributed to each relevant blockchain computing node. Each data packet is attached with an integrity check code generated based on the Hash Message Authentication Code (HMAC), and the receiver verifies whether the data packet has been tampered with during transmission through the check code, ensuring the integrity of data transmission.

[0087] In another preferred embodiment of the present application, the cross-link routing module performs the process of data transmission as follows:

[0088] First, a virtual overlay layer covering multiple blockchain networks is established, which is based on a peer-to-peer (P2P) architecture, and the relay chain structure in it uses directed acyclic graph (DAG) technology to store metadata indexes of cross-chain interactions. The metadata includes source chain and target chain identification, logical data segment hash value, cross-chain transaction timestamp and other information, providing path guidance and state tracking for cross-chain data transmission.

[0089] When the encrypted calculation result needs to be transmitted across the chain, the verification computing node cluster on the source chain performs sharding processing on the data according to the data size and the storage characteristics of the target chain. Each logical data segment after sharding is packaged together with the cross-chain routing instruction (containing target chain address, transmission priority and other information) as a smart contract transaction package. The transaction package adopts the smart contract bytecode format to ensure standardized execution in the blockchain network.

[0090] The protocol converter serves as a bridge for heterogeneous chain communication, and based on syntax-semantic parsing technology, it deeply analyzes the control instructions in the source chain transaction package. Through a rule mapping table, it converts the transaction format of the source chain (such as the EVM format of Ethereum) into a transaction format recognizable by the target chain (such as the chaincode format of Fabric), while strictly preserving the original encryption hierarchy to avoid data leakage risks caused by format conversion.

[0091] The data integrity verification component implements a three-way handshake verification mechanism. In the source chain outbound verification phase, the source chain computing node cluster signs the logical data segment through a multi-signature algorithm (such as BLS signature), verifies the data integrity, and sends it to the relay chain; in the relay chain relay verification phase, the relay chain computing node compares the hash value of the logical data segment and verifies the validity of the source chain signature, and forwards it to the target chain after confirmation; in the target chain inbound verification phase, the target chain computing node cluster verifies the signature through the same hash verification and signature verification process as the source chain, and receives the logical data segment after obtaining the majority of computing nodes (such as 2 / 3 or more computing nodes) on the chain. The three verification processes ensure the security and integrity of the data throughout the cross-chain transmission.

[0092] The above describes one embodiment of the present application in detail, but the content described is only the preferred embodiment of the present application, and cannot be considered as limiting the scope of the present application. Any equivalent changes and improvements made within the scope of the present application should still be within the scope of the present patent.

Claims

1. A financial data privacy protection system based on blockchain secure multi-party computation, characterized in that, The system comprises a data preprocessing module, a distributed ledger, a computing node cluster, a zero-knowledge proof verifier, an encryption proxy gateway, and a cross-chain routing module. The data preprocessing module is configured to split original financial data into at least two logical data segments, add a unique watermark identifier to each logical data segment, and perform asymmetric encryption on the logical data segment with an encryption key independently generated by a participant to obtain an encrypted logical data segment of the participant. In the data segmentation stage, the amount field and the business entity identifier in the original financial data are separated and processed by a financial data type identification rule to generate a first logical data segment containing pure numerical values and a second logical data segment containing business-related information. In the encryption process, a hierarchical encryption strategy is executed, i.e., the logical data segment is first signed using the private key of the participant, and then encrypted using the public key of the recipient, forming an encrypted logical data segment with a nested ciphertext structure. Before being sent to the computing node cluster, the encrypted logical data segment is verified by a local compliance checking engine to determine whether it meets the preset data desensitization rules, and an audit mark is generated and written to the blockchain. Each encrypted logical data segment is attached with lifecycle metadata, which includes the maximum retention period, the list of allowed usage scenarios, and the associated business contract identifier. A data preservation monitor deployed in the blockchain periodically scans the ledger records and sends an update reminder to the relevant participant distributed ledger for the logical data segment approaching the expiration date. The blockchain is used to store the hash digest of the encrypted logical data segment and the mapping relationship between the logical data segment and the watermark identifier. The hash digest generation process incorporates the digital signature of the participant and the timestamp. The computing node cluster is configured to receive encrypted logical data segments from different participants through a dedicated physical interface, dynamically form a multi-party computing group based on the business computing rules of the preset smart contract, perform computation on the encrypted logical data segments, and output encrypted computation results. The zero-knowledge proof verifier is configured to establish a two-way authentication connection with the computing node cluster and verify the legal possession of the logical data segment by the participant before performing the computation task. The encryption proxy gateway is configured to perform format conversion and ciphertext superposition on the encrypted computation results output by the computing node cluster. The cross-chain routing module is configured to perform transmission of the encrypted computation results between different chains, including a heterogeneous chain protocol converter and a data integrity checking component. 2.The blockchain-based secure multi-party computation financial data privacy protection system of claim 1, wherein, The operation flow of the data preprocessing module specifically includes: The unique watermark identifier is formed by concatenating the blockchain identity certificate fingerprint obtained by the participant during registration and the hash value of the data generation time window. Each watermark identifier is associated with a copy of the public key certificate of the participant and stored in the distributed ledger.

3. The financial data privacy protection system based on blockchain secure multi-party computation according to claim 2, wherein When the smart contract detects the preset business completion condition, a corresponding data destruction instruction is triggered, and the data destruction instruction synchronously erases all copies of the to-be-destroyed data through a cooperative mechanism of a computing node cluster; the destruction process adopts a cryptography forgetting technology, deletes data content, and clears physical residual information in a storage medium through an overwrite writing manner, and an independent audit node issues a verifiable destruction certificate. 4.The blockchain-based secure multi-party computation financial data privacy protection system of claim 3, wherein, The working mode of the computing node cluster is: When receiving an encrypted logical data segment, each computing node verifies the validity period and revocation state of a sender identity certificate through a secure isolation hardware environment, and constructs a temporary computing sandbox in the memory; In the group formation process of the multi-party computing group, computing nodes are dynamically elected according to a business computing rule defined in the smart contract, the computing nodes collect computing parameters of each participant and generate a task allocation topology graph; when executing a computing task, each computing node exchanges necessary relay parameters in the temporary computing sandbox, transmits intermediate computing results through a secret sharing manner, and generates a Merkle tree evidence in real time and uploads the blockchain through interaction recording; After the computing task is completed, the computing nodes send a ciphertext digest of the encrypted computing result to all participants for joint signature confirmation, and after the joint signature confirmation is completed, subsequent processing of the encrypted proxy gateway is triggered. 5.The blockchain-based secure multi-party computation financial data privacy protection system of claim 4, wherein, A special log collector is deployed in the computing node cluster, and key events in the computing process are captured in real time, including data receiving time, computing step execution state and abnormal interruption record; After the collected log data is encrypted and desensitized, a compressed data packet containing a time sequence fingerprint is generated and uploaded to multiple blockchain nodes for storage through a random polling mechanism; A log analyzer running in the blockchain network regularly performs correlation analysis on the dispersedly stored log segments, detects potential security threat patterns, and feeds back the analysis results to the risk control center of each participant in an encrypted form; Access operations of all audit logs need to be verified through zero-knowledge proof, and the query results are returned after being processed by differential privacy technology. 6.The blockchain-based secure multi-party computation financial data privacy protection system of claim 4, wherein, The execution logic of the smart contract is: The contract instance deployed in the blockchain network includes a dynamic access strategy engine, which automatically adjusts the access control rules according to the real-time acquired participant credit score, data sensitivity level and environmental risk indicators; The computing task trigger condition predefined in the smart contract is bound with a real business event, and when a specific business state change reported by an off-chain oracle is detected, a multi-party computing request is automatically initiated; All state change records generated in the smart contract execution process are stored in the form of encrypted logs, and the log entries are connected through a forward secure chain structure; the smart contract also includes an automatic dispute resolution mechanism, which calls backup redundant computing nodes to re-execute the task when detecting an abnormal computing result, and determines the final effective result by comparing multiple computing copies. 7.The blockchain-based secure multi-party computation financial data privacy protection system of claim 3, wherein, The running mechanism of the zero-knowledge proof verifier specifically includes: The verification phase generates dynamic challenge parameters, which include requirements for the participant to prove that the watermark identification in the provided logical data segment is consistent with the mapping relationship stored in the blockchain, and the generation process of the encryption key conforms to the key management policy submitted during pre-registration, while verifying that the logical data segment does not exceed the preset valid use period; During the verification process, an interactive proof protocol is executed, and the zero-knowledge proof verifier sends a challenge instruction containing a random salt value to the participant, who returns the proof data after a specific transformation within a specified time; The verification result is divided into three levels of trust level according to the verification strength, and different trust levels correspond to different computing task permissions, which are written into the execution condition judgment logic of the smart contract in an encrypted form. 8.The blockchain-based secure multi-party computation financial data privacy protection system of claim 1, wherein, The processing process of the encryption proxy gateway is as follows: After receiving the encrypted calculation result output by the computing node cluster, the standardization processing of the ciphertext format is executed, and the structural differences generated by different participant encryption algorithms are converted into a unified standard format; In the ciphertext superposition stage, a random noise factor is injected into the standardized encrypted calculation result through the blinding processing technology, and the final result after superposition contains the linear combination of the encrypted calculation result and the noise factor; The original encryption hierarchy is preserved during the format conversion process, and the participant's signature information is converted into a verifiable nested certificate chain form; The processed encrypted calculation result is split into multiple data packets, which are distributed to each relevant blockchain computing node through a polling mechanism, and each data packet is attached with an integrity check code for the receiver to verify the transmission integrity. 9.The blockchain-based secure multi-party computation financial data privacy protection system of claim 1, wherein, The process of the cross-chain routing module for data transmission is as follows: A virtual overlay layer covering multiple blockchain networks is established, which contains a relay chain structure for storing metadata indexes of cross-chain interactions; When the encrypted calculation result needs to be transmitted across chains, the verification computing node cluster on the source chain performs fragmentation processing on the encrypted calculation result and generates a smart contract transaction package containing cross-chain routing instructions; The heterogeneous chain protocol converter parses the control instructions in the smart contract transaction package and converts them into a transaction format recognizable by the target chain, while preserving the original encryption hierarchy unchanged; The data integrity check component implements three-way handshake verification during transmission, including source chain outbound verification, relay chain transfer verification, and target chain inbound verification, and each verification requires signature confirmation from a majority of nodes on the relevant chain.

Images