Financial data privacy protection system based on block chain security multi-party computing

Abstract

The invention discloses a financial data privacy protection system based on block chain security multi-party computing, and belongs to the technical field of data protection, and the system specifically comprises the steps that a data preprocessing module segments original financial data, and adds a unique watermark identifier; the distributed account book stores the Hash abstract and watermark mapping relation of the encrypted logic data fragment, and the digital signature and the timestamp are fused to ensure reliability; the computing node cluster receives data through a special interface, dynamically builds a computing group according to a service rule and outputs an encrypted computing result; bidirectional authentication is carried out on the zero-knowledge proof verifier and the computing node cluster to verify the data holding right; the encryption proxy gateway processes an output result of the computing node cluster to complete format conversion and ciphertext superposition; the cross-link routing module constructs a virtual coverage layer, cross-link transmission of encryption results is achieved through protocol conversion and three-way handshake verification, and a heterogeneous chain protocol converter and a verification assembly guarantee transmission safety; the financial data privacy security is comprehensively guaranteed, and the data processing efficiency and reliability are improved.

Description

Technical Field

[0001] The present invention relates to the field of data protection technology, and in particular to a financial data privacy protection system based on blockchain secure multi-party computing. Background Art

[0002] Amidst the wave of digital transformation, financial data, as a core asset for business operations, faces increasing scrutiny for its privacy and secure sharing. The development of cryptographic technologies such as blockchain and secure multi-party computation (MPC) has provided new technical solutions for addressing privacy concerns in cross-organizational data collaboration. Traditional centralized data storage and processing models face challenges such as single points of failure and high risk of data leakage. Emerging distributed computing technologies are reshaping the paradigm for inter-enterprise data collaboration.

[0003] Currently, blockchain-based financial data sharing systems primarily utilize two technical approaches: one is to achieve distributed data storage and authorized access through consortium chains, where each participant stores encrypted data on the chain and defines access rules through smart contracts; the other is to integrate secure multi-party computing technology to enable multi-party collaborative computing without leaking the original data. For example, some systems use homomorphic encryption technology to allow calculations to be performed on ciphertext, with the decrypted result being equivalent to the result of the plaintext calculation; others use secret sharing schemes to split data into multiple fragments and store them in different computing nodes.

[0004] However, existing technical solutions still have significant shortcomings. For one thing, during the data preprocessing phase, most systems employ only a single encryption strategy, making it difficult to cope with complex and changing security threats. Furthermore, they lack effective data lifecycle management mechanisms, leading to residual security risks from expired data. Furthermore, the computational node collaboration efficiency during secure multi-party computation is low, and the trusted verification mechanism for computational results is imperfect, which can easily lead to disputes over computational results. These technical bottlenecks urgently need to be overcome to build a more secure, efficient, and reliable financial data privacy protection system. Summary of the Invention

[0005] The purpose of this invention is to provide a financial data privacy protection system based on blockchain secure multi-party computing to solve the following technical problems:

[0006] Existing technical solutions employ only a single encryption strategy during data preprocessing, making them incapable of addressing complex and ever-changing security threats. They also lack effective data lifecycle management mechanisms, leading to security risks associated with expired data. Furthermore, the computational node collaboration efficiency during secure multi-party computation is low, and the trusted verification mechanism for computational results is imperfect, easily leading to disputes over computational results.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

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

[0009] A data preprocessing module is configured to divide the original financial data into at least two logical data segments, add a unique watermark to each logical data segment, and then asymmetrically encrypt the logical data segments with the unique watermark using an encryption key independently generated by each participant to obtain the encrypted logical data segments of the participant;

[0010] A distributed ledger for storing hash summaries of encrypted logical data segments and mappings between logical data segments and watermark identifiers, wherein the hash summaries are generated by integrating digital signatures and timestamps of the participants;

[0011] A computing node cluster is used to receive encrypted logical data fragments 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, calculate the encrypted logical data fragments and output the encrypted calculation results;

[0012] A zero-knowledge proof verifier, used to establish a two-way authentication connection with the computing node cluster, verifying the legal ownership of the logical data fragments by the participants before executing the computing task;

[0013] The encryption proxy gateway is used to perform format conversion and ciphertext superposition on the encryption calculation results output by the computing node cluster;

[0014] The cross-chain routing module is used to perform the transmission of cryptographic calculation results between different chains, including heterogeneous chain protocol converters and data integrity verification components.

[0015] As a further solution of the present invention: the operation process of the data preprocessing module specifically includes:

[0016] In the data segmentation phase, the amount field and the business entity identifier in the original financial data are separated and processed according to the financial data type identification rules, generating 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 composed of the blockchain identity certificate fingerprint obtained by the participant when registering and the hash value of the data generation time window. Each watermark identifier is associated with a copy of the participant's public key certificate stored in the distributed ledger;

[0018] A layered encryption strategy is implemented during the encryption process. The logical data fragments are first signed using the private keys of the participants, and then encrypted using the public keys of the recipients to form encrypted logical data fragments with a nested ciphertext structure. Before being sent to the computing node cluster, the encrypted logical data fragments are verified by the local compliance check engine to see whether they comply with the preset data desensitization rules, and an audit mark is generated and written into the blockchain.

[0019] As a further solution of the present invention: the data pre-processing module adds lifecycle metadata to each encrypted logical data segment, the metadata including a maximum retention period, a list of allowed usage scenarios, and an associated business contract identifier;

[0020] The data preservation monitor deployed in the blockchain regularly scans the ledger records and sends update reminders to relevant participants in advance for logical data fragments that are approaching expiration dates;

[0021] When the smart contract detects the preset business completion conditions, it triggers the corresponding data destruction instructions. The data destruction instructions synchronously erase all copies of the data to be destroyed through the collaborative mechanism of the computing node cluster; the destruction process uses cryptographic forgetting technology to delete the data content and clear the physical residual information in the storage medium by overwriting. The independent audit node issues a verifiable destruction certificate.

[0022] As a further solution of the present invention: the working mode of the computing node cluster is:

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

[0024] During the formation of a multi-party computation group, computation nodes are dynamically elected based on the business computation rules defined in the smart contract. These computation nodes collect the computation parameters of each participant and generate a task allocation topology. When executing computation tasks, each computation node exchanges necessary relay parameters within a temporary computation sandbox, transmits intermediate computation results through secret sharing, and generates Merkle tree evidence in real time based on the interaction records and uploads it to the blockchain.

[0025] After the computing task is completed, the computing node sends the ciphertext summary of the encrypted computing result to all participants for joint signature confirmation. After the joint signature confirmation is completed, the subsequent processing of the encryption proxy gateway is triggered.

[0026] As a further solution of the present invention: a dedicated log collector is deployed in the computing node cluster to capture key events in the computing process in real time, including data reception time, computing step execution status and abnormal interruption records;

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

[0028] The log analyzer running in the blockchain network regularly performs correlation analysis on decentralized log fragments to detect potential security threat patterns, and the analysis results are fed back to the risk control center of each participant in encrypted form; all access operations to audit logs must be verified through zero-knowledge proof, and the query results are returned after processing using differential privacy technology.

[0029] As a further solution of the present invention: the execution logic of the smart contract is:

[0030] Contract instances deployed in the blockchain network include a dynamic access policy engine that automatically adjusts access control rules based on real-time participant reputation scores, data sensitivity levels, and environmental risk indicators.

[0031] The predefined computing task triggering conditions in the smart contract are bound to real-world business events. When a specific business status change reported by the off-chain oracle is detected, a multi-party computing 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. When an abnormal calculation result is detected, the backup redundant computing node is called to re-execute the task, and the final valid result is determined by comparing multiple calculation copies.

[0033] As a further solution of the present invention: the operating mechanism of the zero-knowledge proof verifier specifically includes:

[0034] During the verification phase, dynamic challenge parameters are generated. These dynamic challenge parameters require the participant to prove that the watermark identifier in the provided logical data segment is consistent with the mapping relationship stored in the blockchain, and that the encryption key generation process complies with the key management policy submitted during pre-registration. At the same time, it verifies that the logical data segment has not exceeded the preset validity period.

[0035] During the verification process, an interactive proof protocol is executed. 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 results are divided into three levels of trustworthiness according to the verification strength. Different trustworthiness levels correspond to different computing task permissions. The computing task permissions are written into the execution condition judgment logic of the smart contract in encrypted form.

[0037] As a further solution of the present invention: the processing process of the encryption proxy gateway is:

[0038] After receiving the encrypted calculation results output by the computing node cluster, it performs ciphertext format standardization processing to convert the structural differences generated by the encryption algorithms of different participants into a unified standard format;

[0039] During the ciphertext superposition phase, a random noise factor is injected into the standardized encryption calculation result through blinding processing technology. The final result after superposition contains a linear combination of the encryption calculation result and the noise factor.

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

[0041] The processed encryption calculation results are split into multiple data packets and distributed to the relevant blockchain computing nodes through a polling mechanism. Each data packet is accompanied by an integrity check code for the recipient to verify the transmission integrity.

[0042] As a further solution of the present invention: the process of performing data transmission by the cross-link routing module is:

[0043] Establishing a virtual overlay covering multiple blockchain networks, the virtual overlay includes a relay chain structure for storing metadata indexes of cross-chain interactions;

[0044] When cross-chain transmission of encrypted calculation results is required, the verification computing node cluster on the source chain shards the encrypted calculation results and generates 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 keeping the original encryption hierarchy unchanged;

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

[0047] Beneficial effects of the present invention:

[0048] The data preprocessing module of the present invention adopts logical segmentation, layered encryption and lifecycle management technology to solve the security risks of a single encryption strategy, separates the amount field from the business entity identifier through the financial data type identification rule, combines the private key signature of the participants and the public key encryption of the recipient to form a nested ciphertext structure, and realizes full-cycle security management and control through data preservation monitor and cryptographic forgetting technology; the computing node cluster builds a secure isolated hardware environment and a temporary computing sandbox, dynamically elects and coordinates computing nodes to generate a task allocation topology map, uses secret sharing to transmit intermediate results and generate Merkle tree evidence, thereby improving the efficiency of multi-party computing and the credibility of the results; the zero-knowledge proof verifier implements three-level trust level verification through dynamic challenge parameters and interactive proof protocol, ensuring the legal ownership of the participants and the timeliness of the data; the encryption proxy gateway retains the original encryption hierarchy and enhances the data obfuscation through ciphertext format standardization processing and blind superposition technology; the cross-chain routing module constructs a virtual overlay layer and relay chain structure, combines the protocol converter and the three-way handshake verification mechanism, solves the compatibility and data integrity problems of heterogeneous chains, and realizes the security and credibility of cross-chain transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be further described below with reference to the accompanying drawings.

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

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] See also Figure 1 As shown, the present invention is a financial data privacy protection system based on blockchain secure multi-party computing, including 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. Specifically:

[0053] The data preprocessing module is the first step in processing raw financial data entering the system, responsible for secure data segmentation and encryption. Based on data characteristics and security requirements, it segments the raw financial data into at least two logical data segments. For example, it separates the amount field from the business entity identifier, generating a first logical segment containing pure numeric values ​​and a second logical segment containing business-related information, thus achieving data classification protection. Each logical data segment is uniquely watermarked. This watermark is composed of the fingerprint of the participating party's blockchain identity certificate and the hash value of the data generation time window. This watermark is used for data traceability and ownership verification. Subsequently, asymmetric encryption is performed using encryption keys independently generated by the participating parties. The logical data segment is first signed with the participating party's private key and then encrypted with the recipient's public key, forming a nested ciphertext structure to ensure data confidentiality and verifiable origin. The encrypted logical data segments are then verified by the local compliance check engine to comply with pre-set data desensitization rules. Audit tags are then generated and written to the blockchain, providing data support for audits and regulatory oversight.

[0054] The distributed ledger is the core of the system's data storage and management. It stores hash summaries and watermark mappings of encrypted logical data fragments. The hash summaries are generated by integrating the digital signatures and timestamps of participating parties. The digital signatures of participating parties ensure the authenticity of the data source, while the timestamps uniquely identify the data's time dimension, making the hash summaries tamper-resistant and traceable. This enables distributed and secure data storage and lays the foundation for operations such as data verification and traceability.

[0055] The compute node cluster is the execution unit for multi-party computation (MPC). It receives encrypted logical data fragments from different participants through dedicated physical interfaces equipped with security mechanisms to prevent external attacks and data theft. Upon receiving the data, the compute node cluster dynamically forms a MPC group based on pre-set business computation rules. Computing nodes are selected through an election algorithm based on the business rules in the smart contract. After the coordinating compute nodes collect computational parameters from each participant, they first construct a multi-dimensional evaluation model to quantify each node's computing power, network conditions, security level, and historical performance. Graph theory algorithms are then used to decompose the computational task into subtask nodes, such as data preprocessing, encrypted computation, and result aggregation. Based on the evaluation results, a weighted directed graph is generated, in which nodes represent computing resources, edges represent data flows and collaborations, and weights reflect task priorities and resource requirements. A simulated annealing algorithm is used to iteratively optimize the topology, minimizing data transmission delays and redundant computations. Ultimately, a task allocation topology is formed, including task scheduling paths, data encryption transmission channels, and exception handling links, ensuring the orderly execution of computational tasks. During the computation process, each computing node exchanges relay parameters in an encrypted state and transmits intermediate computation results using secret sharing. The interaction records generate Merkle tree evidence in real time and upload it to the blockchain to verify the computation process. Upon completion of the computation task, the coordinating computing node sends a ciphertext summary of the encrypted computation result to all participating parties for joint signature confirmation, ensuring the reliability of the computation result.

[0056] The zero-knowledge proof verifier is used to ensure the legitimacy of data use by participants. It establishes a two-way authentication connection with the computing node cluster and verifies participants before the computing task is executed. During verification, a dynamic challenge parameter is generated, requiring the participant to prove that the watermark identifier in its logical data fragment is consistent with the mapping relationship stored on the blockchain, that the encryption key generation process complies with the pre-registered key management policy, and that the logical data fragment is within the valid usage period. Verification uses an interactive proof protocol. The verifier sends a challenge instruction containing a random salt value to the participant, and the participant must return a specifically transformed proof data within the specified time. Three levels of trust are divided according to the verification strength. Different levels correspond to different computing task permissions. The permission information is written into the smart contract execution condition judgment logic in encrypted form to achieve refined management of the participant's permissions.

[0057] The encryption proxy gateway optimizes the processing of the encrypted calculation results output by the compute node cluster. After receiving the primary encryption results, it first standardizes the ciphertext format to unify the structures generated by different encryption algorithms among different participants. During the ciphertext overlay phase, it injects random noise factors through blinding processing techniques, making the final result a linear combination of the original calculation results and the noise data, enhancing data confidentiality. The format conversion process preserves the original encryption hierarchy and converts the participant signature information into a verifiable nested certificate chain. The processed encryption results are split into multiple data packets and distributed to the relevant blockchain compute nodes via a polling mechanism. Each data packet is accompanied by an integrity check code to ensure data transmission integrity.

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

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

[0060] During the data segmentation phase, the data preprocessing module's built-in financial data type identification rules utilize a combination of regular expressions and semantic analysis to perform in-depth analysis of the raw financial data. Preset feature patterns for amount fields accurately identify monetary values ​​within the data and separate them from business entity identifiers. For example, for financial data containing customer order amounts and customer numbers, the order amount is extracted to generate a first logical data segment containing purely numerical values, while business-related information such as the customer number forms a second logical data segment. This separation allows data of varying sensitivity to be classified and protected, mitigating the potential risk of data breaches.

[0061] During the watermark generation phase, unique watermarks are constructed based on cryptographic hashing algorithms and digital certificate technology. Upon system registration, participants receive a blockchain identity certificate with a unique fingerprint. The data preprocessing module concatenates this fingerprint with the hash value of the data generation time window to form a unique watermark. For example, the data generation time window is hashed using the SHA-256 algorithm and then combined with the identity certificate fingerprint in a specific format. Each watermark is stored in the distributed ledger in association with a copy of the participant's public key certificate. This can be used for data provenance, ownership verification, and operational auditing, ensuring traceability throughout the data flow.

[0062] The encryption process strictly adheres to a layered encryption strategy. First, each participant digitally signs the logical data segment using their own private key. This signing process adheres to the ECDSA (Elliptic Curve Digital Signature Algorithm) standard, ensuring the authenticity and integrity of the data source by signing the hash value of the logical data segment. After signing, asymmetric encryption is performed using the recipient's public key, employing established encryption algorithms such as RSA or ECC, to form a nested ciphertext structure. This structure not only ensures data confidentiality during transmission and storage, but also verifies data ownership through signature verification. Before being sent to the compute node cluster, the encrypted logical data segment enters a local compliance check engine. This engine, with its built-in pre-configured data desensitization rule library, uses pattern matching and compliance verification algorithms to detect and process sensitive information within the logical data segment, such as anonymizing names and ID numbers. Upon successful verification, an audit tag is generated, containing information such as data processing time and operator identification, and is written to the blockchain for subsequent oversight and auditing.

[0063] In a preferred embodiment of this embodiment, the data preprocessing module further enhances data lifecycle management capabilities. Lifecycle metadata is appended to each encrypted segment. This metadata is structured and stored in JSON or XML format and contains key information such as the maximum retention period, a list of permitted usage scenarios, and associated business contract identifiers. For example, a financial logic data segment might have a maximum retention period of three years, with permitted usage limited to internal financial analysis and tax filing, and an associated business contract identifier pointing to a specific collaborative project agreement.

[0064] The data preservation monitor deployed on the blockchain operates automatically based on smart contracts. It scans distributed ledger records at preset intervals (e.g., daily or weekly). By comparing the current time of logical data segments with the maximum retention period, it automatically generates update reminders for logical data segments approaching expiration (e.g., 30 days before expiration) and sends them to relevant participants via a secure communication channel. The reminders contain the logical data segment identifier and the remaining validity period, allowing participants to respond promptly.

[0065] When the smart contract detects that the preset business completion conditions have been met, it automatically triggers a data destruction instruction. This instruction is transmitted and executed through the consensus and collaboration mechanisms of the computing node cluster, ensuring that the data copies stored on each computing node are synchronously erased. The destruction process utilizes cryptographic forgetting technology. Specifically, through overwriting, the original financial data storage area is repeatedly overwritten with randomly generated binary data to eliminate any physical residual information on the storage medium. After the operation is completed, an independent audit node verifies the data destruction process and results using zero-knowledge proof technology and issues a verifiable destruction certificate that includes the verification time and the verifier's signature to ensure the effectiveness and compliance of the data destruction.

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

[0067] When receiving encrypted logical data fragments, each compute node utilizes a secure, isolated hardware environment, such as Trusted Execution Environment (TEE) technology, to provide a physically and logically isolated secure space for data reception and processing. Within this environment, the compute node rigorously verifies the validity and revocation status of the sender's identity certificate. Certificate validity is ensured through real-time interaction with the Certificate Authority (CA) or locally cached certificate status information. Simultaneously, a temporary computational sandbox is constructed in memory. This sandbox, leveraging the operating system's memory management and virtualization technologies, allocates independent memory space to each computational task, preventing memory conflicts and data leaks between different tasks.

[0068] During the formation of a multi-party computation group, the computing node cluster dynamically selects computing nodes using a distributed election algorithm based on the business rules predefined in the smart contract. For example, a variant of the Byzantine Fault Tolerance (BFT) algorithm can be used to ensure that a reliable coordinating computing node can be elected even in the event of partial computing node failure or malicious behavior. The coordinating computing node is responsible for collecting computing parameters from each participant, including but not limited to encryption keys for logical data fragments and configuration information for computing tasks. Based on these parameters, the coordinating computing node generates a task allocation topology. This topology, based on graph theory and optimization algorithms, considers factors such as computing node computing power and network bandwidth to rationally plan the distribution of computing tasks among the computing nodes to maximize computing efficiency.

[0069] When executing a computational task, each computing node exchanges data in an encrypted state. Intermediate computation results are securely transmitted between computing nodes through secret sharing. Secret sharing utilizes cryptographic algorithms, such as the Shamir secret sharing scheme, to split a secret value into multiple shares. Each computing node holds a portion of the shares, and the original secret value can only be recovered when a sufficient number of computing nodes are brought together. During this data exchange process, Merkle tree evidence is generated in real time and uploaded to the blockchain. The Merkle tree, based on a hashing algorithm, combines the hash values ​​of a series of data blocks layer by layer to form a binary tree structure. The hash value of the root computing node represents the integrity of the entire data set. Uploading the Merkle tree evidence to the blockchain ensures the traceability and immutability of the computation process.

[0070] After the computation task is completed, the coordinating computing node sends the ciphertext summary of the final result to all participating parties for joint signature confirmation. Each participant signs the ciphertext summary using their own private key, following the Digital Signature Standard (DSS). Once the joint signature confirmation is complete, the encryption proxy gateway triggers subsequent processing. Homomorphic encryption technology allows specific computational operations to be performed on ciphertext without decryption, thereby protecting data privacy.

[0071] In a preferred embodiment of this embodiment, the computing node cluster further enhances the monitoring and management capabilities of the computing process. By deploying a dedicated log collector, critical events during the computing process are captured in real time. These events include data reception time, computational step execution status, and abnormal interruption records. The log collector utilizes a high-performance logging library and event-driven architecture to ensure accurate recording of every critical event without compromising computing performance.

[0072] Collected log data is first encrypted using a symmetric encryption algorithm, such as AES, to ensure confidentiality during transmission and storage. Desensitization is then performed, using techniques such as data replacement and masking to remove sensitive information from the logs, such as participant identities and specific financial data. The processed log data is then converted into a compressed data package containing a time series fingerprint. This fingerprint, generated based on the timestamp and data hash value, uniquely identifies each log record. This compressed data package is then uploaded to multiple blockchain compute nodes for storage via a randomized round-robin mechanism. This randomized round-robin mechanism ensures even distribution of log data across the blockchain network, improving storage reliability and attack resistance.

[0073] Log analyzers running within the blockchain network regularly perform correlation analysis on decentralized log fragments. They employ big data analysis techniques and machine learning algorithms, such as cluster analysis and anomaly detection, to detect potential security threat patterns. For example, by analyzing unusual fluctuations in data reception times or frequent errors in the execution status of computational steps, they can identify potential network attacks or system failures. The analysis results are fed back to the risk control centers of each participant in an encrypted format to ensure the security of information transmission. All access to audit logs is verified using zero-knowledge proofs. Zero-knowledge proofs allow a prover to prove a statement to a verifier without revealing any useful information. Query results are processed and returned using differential privacy techniques. Differential privacy protects sensitive information in the query results by, for example, adding noise, while ensuring the availability of the results.

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

[0075] The execution logic of smart contracts provides strong support for the automated and intelligent operation of the entire system. Contract instances deployed within the blockchain network incorporate a dynamic access policy engine, which automatically adjusts access control rules based on real-time participant reputation scores, data sensitivity levels, and environmental risk indicators. Reputation scores are derived by evaluating participants' historical behavior and transaction records; data sensitivity levels are determined based on factors such as data type and importance; and environmental risk indicators consider factors such as network security and system stability. The dynamic access policy engine utilizes 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 manipulate data under appropriate conditions.

[0076] The predefined triggering conditions for computational tasks in the contract are tightly coupled to physical-world business events. Upon detecting a specific business status change reported by an off-chain oracle, a multi-party computation request is automatically initiated. Off-chain oracles, through interfaces with external data sources, acquire business data in real time and convert it into a format recognizable by the blockchain. For example, when a sales order status changes from "Created" to "Shipped," the oracle transmits this status change information to the blockchain, triggering the associated financial data computation task.

[0077] All state changes generated during contract execution are stored as encrypted logs, with log entries linked by a forward-secure chain structure. This forward-secure chain structure is based on cryptographic hash chain technology, where the hash value of each log entry depends on the hash value of the previous entry, ensuring log order and immutability. The contract also includes an automated dispute resolution mechanism. When an anomaly is detected in a calculation result, a redundant backup compute node is called upon to re-execute the task. Redundant compute nodes utilize hot or cold backup strategies to ensure timely takeover in the event of a failure of the primary compute node or an abnormal calculation result. The final valid result is determined by comparing multiple computation replicas, using a majority vote or consensus algorithm to ensure accuracy and reliability.

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

[0079] During the verification phase, the zero-knowledge proof verifier dynamically generates challenge parameters based on a secure random number generation algorithm. These parameters are verified across three aspects: the watermark identification of the logical data segment, the compliance of the encryption key, and its expiration date. Regarding watermark identification, participants are required to verify the consistency of the mapping relationship between the watermark in the logical data segment and the distributed ledger storage through hash comparison and digital certificate analysis, ensuring data traceability. Regarding encryption keys, the generation process is verified to comply with the key management policy submitted to the blockchain during pre-registration, such as whether the key length and generation algorithm meet security standards. Furthermore, based on the lifecycle metadata attached to the logical data segment, the data is verified to be within the preset valid usage period to prevent expired data from being illegally used in calculations.

[0080] The verification process uses an interactive proof protocol based on zero-knowledge proof principles. The validator sends a challenge instruction to the participant, which contains a random salt value. The random salt value 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 must generate and return proof data within a specified timeframe using a local encryption key and a logical data fragment through a specific cryptographic transformation algorithm. Examples include discrete logarithm, hash collision, and polynomial commitment. Discrete logarithm, using mathematical structures such as elliptic curves, enables the prover to prove possession of the corresponding public key through a challenge-response process without revealing the private key, such as in signature compliance verification. Hash collision leverages the one-way and collision-resistant properties of hash functions to transform data or a random salt value into a unique fingerprint, used to generate unpredictable challenge parameters or compress proof data, such as in hash comparisons of watermarks. Polynomial commitment allows the prover to "commit" to a polynomial and, by submitting proofs of evaluation at specific points or coefficient relationships, verify complex logic without disclosing the polynomial itself, such as range proofs for data validity periods. All three algorithms utilize cryptographic one-way puzzles to achieve specific transformations that "verify without revealing details." By verifying the validity of the proof data, the verifier confirms the participant's legal ownership of the logical data fragment without accessing the original financial data.

[0081] Verification results are divided into three levels of trustworthiness based on verification strength: high, medium, and low. The high trustworthiness level requires participants to fully pass watermark identification, key compliance, and expiration verification, and to prove that the data verification time is below a threshold. The medium trustworthiness level allows for passing even if some minor verification items have minor flaws but do not affect overall security. The low trustworthiness level is suitable for scenarios where only basic identity verification is passed. Different trustworthiness levels correspond to different computing task permissions. For example, high trustworthiness levels allow participation in core financial data calculations, while low trustworthiness levels can only perform auxiliary data processing. Permission information is encrypted using the AES-256 encryption algorithm and written into the smart contract's execution condition judgment logic to ensure the security and dynamism of permission management.

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

[0083] After receiving the primary encryption results from the compute node cluster, the encryption proxy gateway initiates the ciphertext format standardization process. To address structural differences caused by various encryption algorithms, such as RSA and ECC, different participants utilize a format parsing and conversion engine to convert the ciphertext into the JSON-Web Encryption (JWE) format that complies with system specifications. This conversion process strictly preserves the original encryption hierarchy to ensure data confidentiality.

[0084] During the ciphertext overlay phase, blinding techniques are employed to further obfuscate the data. A noise injection algorithm adds a random noise factor conforming to a Gaussian distribution to the original calculation result, making the final result appear as a linear combination of the original calculation result and the noise data. The noise parameter is dynamically adjusted based on the data's sensitivity level, with higher noise levels injected into highly sensitive data. This significantly enhances data confidentiality while ensuring the usability of the calculation results.

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

[0086] The processed encrypted results are split into multiple packets of appropriate size by the packet splitting module, based on data size and network transmission characteristics. These packets are distributed to the relevant blockchain computing nodes via a polling mechanism. Each packet is accompanied by an integrity check code generated using a hash message authentication code (HMAC). The receiver uses this check code to verify whether the packet has been tampered with during transmission, ensuring data transmission integrity.

[0087] In another preferred embodiment of the present invention, the process of performing data transmission by the cross-link routing module is:

[0088] First, a virtual overlay covering multiple blockchain networks is established. This overlay is built on a peer-to-peer (P2P) architecture, with a relay chain structure that uses directed acyclic graph (DAG) technology to store metadata indexes for cross-chain interactions. This metadata includes information such as the source and target chain identifiers, hash values ​​of logical data fragments, and cross-chain transaction timestamps, providing path guidance and status tracking for cross-chain data transmission.

[0089] When encrypted computation results need to be transferred across blockchains, the validating computing node cluster on the source chain shards the data based on the data size and storage characteristics of the target chain. Each logical data fragment is packaged into a smart contract transaction package along with cross-chain routing instructions (including information such as the target chain address and transfer priority). This transaction package uses the smart contract bytecode format to ensure standardized execution across the blockchain network.

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

[0091] The data integrity verification component implements a three-way handshake verification mechanism. During the source chain outbound verification phase, the source chain computing node cluster signs the logical data fragment using a multi-signature algorithm (such as BLS signature), verifies the data integrity, and sends it to the relay chain. During the relay chain transfer verification phase, the relay chain computing node compares the hash value of the logical data fragment and verifies the validity of the source chain signature. After confirmation, it is forwarded to the target chain. During the target chain inbound verification phase, the target chain computing node cluster uses the same hash checksum and signature verification process as the source chain. After obtaining signature confirmation from the majority of computing nodes on the chain (such as more than 2 / 3 computing nodes), it receives the logical data fragment. The three-way verification process ensures the security and integrity of data throughout the cross-chain transmission process.

[0092] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A financial data privacy protection system based on blockchain secure multi-party computing, characterized by: It includes data preprocessing module, distributed ledger, computing node cluster, zero-knowledge proof verifier, encryption proxy gateway and cross-chain routing module. Specifically: A data preprocessing module is configured to divide the original financial data into at least two logical data segments, add a unique watermark to each logical data segment, and then asymmetrically encrypt the logical data segments with the unique watermark using an encryption key independently generated by each participant to obtain the encrypted logical data segments of the participant; A distributed ledger for storing hash summaries of encrypted logical data segments and mappings between logical data segments and watermark identifiers, wherein the hash summaries are generated by integrating digital signatures and timestamps of the participants; A computing node cluster is used to receive encrypted logical data fragments 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, calculate the encrypted logical data fragments and output the encrypted calculation results; A zero-knowledge proof verifier, used to establish a two-way authentication connection with the computing node cluster, verifying the legal ownership of the logical data fragments by the participants before executing the computing task; The encryption proxy gateway is used to perform format conversion and ciphertext superposition on the encryption calculation results output by the computing node cluster; The cross-chain routing module is used to perform the transmission of cryptographic calculation results between different chains, including heterogeneous chain protocol converters and data integrity verification components.

2. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 1 is characterized in that: The operation process of the data preprocessing module specifically includes: In the data segmentation phase, the amount field and the business entity identifier in the original financial data are separated and processed according to the financial data type identification rules, generating a first logical data segment containing pure numerical values ​​and a second logical data segment containing business-related information; The unique watermark identifier is composed of the blockchain identity certificate fingerprint obtained by the participant when registering and the hash value of the data generation time window. Each watermark identifier is associated with a copy of the participant's public key certificate stored in the distributed ledger; A layered encryption strategy is implemented during the encryption process. The logical data fragments are first signed using the private keys of the participants, and then encrypted using the public keys of the recipients to form encrypted logical data fragments with a nested ciphertext structure. Before being sent to the computing node cluster, the encrypted logical data fragments are verified by the local compliance check engine to see whether they comply with the preset data desensitization rules, and an audit mark is generated and written into the blockchain.

3. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 2 is characterized in that: The data pre-processing module adds lifecycle metadata to each encrypted logical data fragment, wherein the metadata includes a maximum retention period, a list of allowed usage scenarios, and an associated business contract identifier; The data preservation monitor deployed in the blockchain regularly scans the ledger records and sends update reminders to relevant participants in advance for logical data fragments that are approaching expiration dates; When the smart contract detects the preset business completion conditions, it triggers the corresponding data destruction instructions. The data destruction instructions synchronously erase all copies of the data to be destroyed through the collaborative mechanism of the computing node cluster; the destruction process uses cryptographic forgetting technology to delete the data content and clear the physical residual information in the storage medium by overwriting. The independent audit node issues a verifiable destruction certificate.

4. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 3 is characterized in that: The working mode of the computing node cluster is: When receiving encrypted logical data fragments, each computing node verifies the validity period and revocation status of the sender's identity certificate through a secure isolated hardware environment and builds a temporary computing sandbox in memory; During the formation of a multi-party computation group, computation nodes are dynamically elected based on the business computation rules defined in the smart contract. These computation nodes collect the computation parameters of each participant and generate a task allocation topology. When executing computation tasks, each computation node exchanges necessary relay parameters within a temporary computation sandbox, transmits intermediate computation results through secret sharing, and generates Merkle tree evidence in real time based on the interaction records and uploads it to the blockchain. After the computing task is completed, the computing node sends the ciphertext summary of the encrypted computing result to all participants for joint signature confirmation. After the joint signature confirmation is completed, the subsequent processing of the encryption proxy gateway is triggered.

5. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 4 is characterized in that: A dedicated log collector is deployed in the computing node cluster to capture key events in the computing process in real time, including data reception time, computing step execution status, and abnormal interruption records; The collected log data is encrypted and desensitized to generate compressed data packets containing time series fingerprints, which are uploaded to multiple blockchain nodes for storage through a random polling mechanism. The log analyzer running in the blockchain network regularly performs correlation analysis on decentralized log fragments to detect potential security threat patterns. The analysis results are fed back to the risk control center of each participant in an encrypted form. All access operations to audit logs must be verified through zero-knowledge proof, and the query results are returned after processing using differential privacy technology.

6. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 4 is characterized in that: The execution logic of the smart contract is: Contract instances deployed in the blockchain network include a dynamic access policy engine that automatically adjusts access control rules based on real-time participant reputation scores, data sensitivity levels, and environmental risk indicators. The predefined computing task triggering conditions in the smart contract are bound to real-world business events. When a specific business status change reported by the off-chain oracle is detected, a multi-party computing request is automatically initiated. 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. When an abnormal calculation result is detected, the backup redundant computing node is called to re-execute the task, and the final valid result is determined by comparing multiple calculation copies.

7. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 3 is characterized in that: The operating mechanism of the zero-knowledge proof verifier specifically includes: During the verification phase, dynamic challenge parameters are generated. These dynamic challenge parameters require the participant to prove that the watermark identifier in the provided logical data segment is consistent with the mapping relationship stored in the blockchain, and that the encryption key generation process complies with the key management policy submitted during pre-registration. At the same time, it verifies that the logical data segment has not exceeded the preset validity period. During the verification process, an interactive proof protocol is executed. 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. The verification results are divided into three levels of trustworthiness according to the verification strength. Different trustworthiness levels correspond to different computing task permissions. The computing task permissions are written into the execution condition judgment logic of the smart contract in encrypted form.

8. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 1 is characterized in that: The processing process of the encryption proxy gateway is as follows: After receiving the encrypted calculation results output by the computing node cluster, it performs ciphertext format standardization processing to convert the structural differences generated by the encryption algorithms of different participants into a unified standard format; During the ciphertext superposition phase, a random noise factor is injected into the standardized encryption calculation result through blinding processing technology. The final result after superposition contains a linear combination of the encryption calculation result and the noise factor. The original encryption hierarchy is retained during the format conversion process, and the signature information of the participants is converted into a verifiable nested certificate chain format; The processed encryption calculation results are split into multiple data packets and distributed to the relevant blockchain computing nodes through a polling mechanism. Each data packet is accompanied by an integrity check code for the recipient to verify the transmission integrity.

9. The financial data privacy protection system based on blockchain secure multi-party computing according to claim 1 is characterized in that: The process of performing data transmission by the cross-link routing module is as follows: Establishing a virtual overlay covering multiple blockchain networks, the virtual overlay includes a relay chain structure for storing metadata indexes of cross-chain interactions; When cross-chain transmission of encrypted calculation results is required, the verification computing node cluster on the source chain shards the encrypted calculation results 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 the control instructions into a transaction format recognizable by the target chain while keeping the original encryption hierarchy unchanged; The data integrity verification component implements three-way handshake verification during the transmission process, including source chain outbound verification, relay chain transfer verification and target chain inbound verification. Each verification must obtain signature confirmation from the majority of nodes on the relevant chain.

Images