Blockchain-based multi-party secure computation method and system

Abstract

The present application relates to the technical field of secure computing, and in particular to a multi-party secure computing method and system based on a block chain, which discloses a secure computing task determination module, a block node screening module, a splitting rule analysis module and a multi-party secure computing module. After the system receives a secure computing task, the basic computing performance of the nodes in the block chain is analyzed, and the comprehensive computing capacity of each node for different smart contracts is deeply analyzed, so that the nodes that can perform secure computing work are accurately and quickly screened out. The splitting rule analysis module and the multi-party secure computing module are set, the comprehensive matching degree of the nodes in the block chain and the encrypted fragments under different splitting rules is deeply analyzed, the allocation and utilization capacity of the node resources in the block chain are ensured, and the distributed computing potential of the block chain network is released.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of secure computing, more specifically, it relates to a multi-party secure computing method and system based on a block chain. BACKGROUND

[0002] With the rapid development of information technology, data as an important production factor, the value in various fields is increasingly prominent. In order to realize more efficient business cooperation, mine data potential value and improve the scientific nature of decision-making, different institutions often need to carry out cross-institutional data cooperation and joint calculation. However, in the multi-party data computing scenario, there are many technical challenges to be solved, which highlights the importance and innovation of the multi-party secure computing method and system based on the block chain of the present application.

[0003] In the traditional multi-party computing environment, when multiple participants want to jointly use their own data for joint analysis, modeling and other computing tasks, data privacy protection has always been one of the core problems. Since each participant needs to share data or transmit data to a unified computing platform for processing, there is inevitably a risk of data leakage, so that sensitive information may be obtained by other participants or external malicious attackers, seriously threatening the privacy rights and interests of enterprises, institutions and individuals. For example, in the financial field, if multiple banks want to jointly build a risk assessment model and share user credit records, transaction records and other data, once the data is leaked, not only the privacy of the user will be damaged, but also serious financial security problems may be caused; in the medical industry, when different medical institutions share patient medical record data to carry out joint medical research, the privacy of patients needs to be strictly protected, and it is difficult to ensure data availability while ensuring privacy in the traditional way.

[0004] To cope with the above privacy and security problems, multi-party secure computing technology has emerged, aiming to complete joint computing tasks under the condition of "available but invisible" data of each participant. However, the existing multi-party secure computing scheme still has many limitations, which is difficult to meet the increasingly complex practical application requirements.

[0005] On the one hand, in terms of node resource allocation, most schemes lack a fine screening mechanism for participating computing nodes. Usually, only based on part of the basic computing performance indicators of the node (such as computing power, storage capacity, etc.) to allocate computing tasks, without fully considering the comprehensive adaptability of the node when facing different types of smart contracts and diversified computing tasks. This is easy to cause resource mismatching problems, such as assigning complex deep learning related computing tasks to nodes that are good at such algorithms although they have high computing power, or assigning tasks with high requirements for data privacy to nodes with insufficient security protection capabilities, thereby affecting the computing efficiency, increasing the computing delay, and even possibly causing data security hazards due to improper handling of the node.

[0006] On the other hand, the processing and distribution of data ciphertext are not reasonable enough. In the process of splitting and distributing to different nodes for calculation after data encryption, the existing scheme often fails to deeply analyze the comprehensive matching degree between nodes and encrypted fragments under different splitting rules, and the splitting rule selection is relatively blind and lacks scientific basis. This makes the node resources unable to be fully and effectively utilized, and some nodes may be overloaded or have idle resources due to the distribution of unmatched encrypted fragments, which cannot fully release the distributed computing potential of the blockchain network, limits the performance improvement of the entire multi-party secure computing system, and is difficult to cope with large-scale and high-complexity data collaboration computing scenarios.

[0007] Based on the above, the present application provides a multi-party secure computing method and system based on a blockchain. SUMMARY

[0008] In view of the deficiencies in the prior art, the purpose of the present application is to provide a multi-party secure computing method and system based on a blockchain.

[0009] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0010] The multi-party secure computing system based on a blockchain comprises a secure computing task determination module, a block node screening module, a splitting rule analysis module and a multi-party secure computing module.

[0011] The secure computing task determination module determines the secure computing task of the initiator and the original data, encapsulates the secure computing task as a smart contract, determines the input data format, calculation accuracy requirement and result delivery condition of the smart contract.

[0012] The block node screening module determines the preferred secure computing node in the blockchain.

[0013] The splitting rule analysis module is used for encrypting the original data, generating data ciphertext after encryption, determining all applicable splitting rules of the data ciphertext, and obtaining the comprehensive secure computing matching index of the blockchain under each splitting rule.

[0014] The multi-party secure computing module marks the splitting rule with the largest comprehensive secure computing matching index value as the secure splitting rule, splits the data ciphertext into multiple encrypted fragments using the secure splitting rule, and arranges the preferred secure computing node to perform secure computing.

[0015] Further, the preferred security calculation node in the blockchain is determined, specifically as follows: the initial election index of each node in the blockchain is determined, the initial election boundary index is set, when the initial election index of the node is greater than or equal to the initial election boundary index, the corresponding node is marked as a preliminary election security calculation node, and the preliminary election security calculation node without a security calculation task at present is marked as a preferred security calculation node.

[0016] Further, the initial election index of a node is determined as follows: a node in the blockchain is selected, all security calculation records of the node within a time length of t before the current time of the system are obtained, all security calculation records are matched two by two, the two security calculation records matched are marked as a record comparison group, the difference calculation capability index of each record comparison group is obtained, all difference calculation capability indexes are summed and averaged, the average difference calculation capability index is calculated, the consistency of the pre-encryption hash value and the post-encryption hash value in the security calculation record is compared, when the pre-encryption hash value and the post-encryption hash value are consistent, the security calculation number is increased by one, and the product result of the average difference calculation capability index and the security calculation number is obtained to obtain the initial election index of the node.

[0017] Further, the difference calculation capability index of the record comparison group is obtained as follows: the security calculation efficiency of the two security calculation records in the record comparison group is summed to obtain the group security calculation efficiency, the contract feature set of the two security calculation records in the record comparison group is matched to obtain a contract feature difference group, the contract feature difference model is obtained, the contract feature difference group is taken as the input data of the contract feature difference model, and the contract feature difference index is output, and the difference calculation capability index of the record comparison group is calculated based on the group security calculation efficiency and the contract feature difference index.

[0018] Further, the security calculation record includes security calculation efficiency, pre-encryption hash value, post-encryption hash value, and contract feature set.

[0019] Further, the contract feature set of the security calculation record is obtained as follows: the encryption shard to which the security calculation record is directed is determined, and the smart contract corresponding to the encryption shard is determined, the input data format, calculation precision requirement, and result delivery condition of the smart contract are extracted to obtain the input data format feature, calculation precision requirement feature, and result delivery condition feature, and the input data format feature, calculation precision requirement feature, and result delivery condition feature are combined to obtain the contract feature set.

[0020] Further, the comprehensive security calculation matching index of the blockchain under a split rule is obtained in the following manner: a split rule applicable to the data ciphertext is selected, the data ciphertext is split according to the split rule, after the splitting, a plurality of encrypted fragments are generated, the total number of the encrypted fragments is obtained, when the total number of the preferred security calculation nodes is greater than or equal to the total number of the encrypted fragments, the total number of the encrypted fragments is calculated by ratio with the total number of the preferred security calculation nodes, a residual allocation matching index is calculated, a competitive calculation difference gradient value is calculated, and the comprehensive security calculation matching index of the blockchain under the split rule is determined based on the residual allocation matching index and the competitive calculation difference gradient value.

[0021] Further, the competitive calculation difference gradient value is calculated in the following manner: all the preferred security calculation nodes are sequentially sorted in descending order of the initial competitive index values, the initial competitive index of the next preferred security calculation node after sorting is calculated by ratio with the initial competitive index of the previous preferred security calculation node, a competitive gradient ratio is calculated, all the competitive gradient ratios are sequentially sorted in descending order of the values, the absolute difference value of the two adjacent competitive gradient ratios after sorting is calculated, a competitive gradient gap ratio is calculated, the sum average of all the competitive gradient gap ratios is calculated, an average competitive gradient gap ratio is calculated, the fragment volume of the encrypted fragment is obtained, all the encrypted fragments are sequentially sorted in descending order of the fragment volume, the fragment volume of the next encrypted fragment after sorting is calculated by ratio with the fragment volume of the previous encrypted fragment, a volume gradient ratio is calculated, all the volume gradient ratios are sequentially sorted in descending order of the values, the absolute difference value of the two adjacent volume gradient ratios after sorting is calculated, a volume gradient gap ratio is calculated, the sum average of all the volume gradient gap ratios is calculated, an average volume gradient gap ratio is calculated, the absolute difference value of the average competitive gradient gap ratio and the average volume gradient gap ratio is calculated, and the competitive calculation difference gradient value is calculated.

[0022] Further, the preferred security calculation nodes are arranged to perform security calculation in the following manner: all the encrypted fragments are sequentially sorted in descending order of the fragment volume, all the preferred security calculation nodes are sequentially sorted in descending order of the initial competitive index values, the first encrypted fragment after sorting is allocated to the first preferred security calculation node, and so on, and after the allocation, each preferred security calculation node performs security calculation on the encrypted fragments.

[0023] Further, the multi-party security calculation method based on the blockchain comprises the following steps:

[0024] Step one: determining the security calculation task of the initiator and the original data, and encapsulating the security calculation task as a smart contract;

[0025] Step two: determine the preferred security calculation node in the blockchain;

[0026] Step three: encrypt the original data, generate data ciphertext after encryption, determine all applicable split rules of the data ciphertext, and obtain the comprehensive security calculation matching index of the blockchain under each split rule;

[0027] Step four: mark the split rule with the largest comprehensive security calculation matching index value as the security split rule, split the data ciphertext into multiple encrypted fragments using the security split rule, and arrange the preferred security calculation node to perform security calculation.

[0028] Compared with the prior art, the present application has the following beneficial effects:

[0029] 1. The setting of the security calculation task determination module and the block node screening module can accurately and quickly screen out nodes that can perform security calculation work after the system receives a security calculation task, analyze the basic calculation performance of the nodes in the blockchain, and deeply analyze the comprehensive calculation ability of each node for different smart contracts. The split rule analysis module and the multi-party security calculation module deeply analyze the comprehensive matching degree of the data ciphertext under different split rules, the nodes in the blockchain and the encrypted fragments, ensure the allocation and utilization ability of the node resources in the blockchain, and release the distributed calculation potential of the blockchain network.

[0030] 2. In the process of deeply analyzing the matching degree to reasonably allocate node resources and encrypted fragments, since the data ciphertext is always processed according to the optimal split rule and allocated to the most suitable node, each node only needs to process the encrypted fragment allocated to itself, without touching other irrelevant data, further strengthening the data privacy protection. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 The system of the present application is a schematic diagram of the principle;

[0032] Figure 2 The principle flow chart of the multi-party security calculation module of the present application;

[0033] Figure 3 The method flow chart of the method of the present application. DETAILED DESCRIPTION

[0034] Example one: refer to Figures 1-2 The multi-party security calculation system based on the blockchain includes a security calculation task determination module, a block node screening module, a split rule analysis module, and a multi-party security calculation module.

[0035] A security calculation task determination module determines a security calculation task of an initiator and original data, encapsulates the security calculation task as a smart contract, and determines an input data format, a calculation precision requirement, and a result delivery condition of the smart contract.

[0036] A block node screening module determines an initial election index of each node in a blockchain, sets an initial election boundary index (the initial election boundary index is a preset index used for comparison with the initial election index), marks a corresponding node as a preliminary election security calculation node when the initial election index of the node is greater than or equal to the initial election boundary index (does not mark when the initial election index of the node is less than the initial election boundary index), and marks a preliminary election security calculation node that currently has no security calculation task as a preferred security calculation node.

[0037] The initial election index of a node is determined as follows: a node in the blockchain is selected, all security calculation records of the node within a time length of t before the current time of the system are obtained, all the security calculation records are matched two by two, the two matched security calculation records are marked as a record comparison group, the difference calculation capability index of each record comparison group is obtained, all the difference calculation capability indexes are summed and averaged, the average difference calculation capability index Dsry is calculated, the consistency of the pre-hash value and the post-hash value in the security calculation record is compared, the security calculation number is increased by one when the pre-hash value and the post-hash value are consistent, the security calculation number is marked as Nucms, and the initial election index of the node is calculated by Fshu=Dsry*Nucms.

[0038] The difference calculation capability index of the record comparison group is obtained as follows: the security calculation efficiency of the two security calculation records in the record comparison group is summed and calculated to obtain the in-group security calculation efficiency Bpset, the contract feature set of the two security calculation records in the record comparison group is matched into a contract feature comparison group, the contract feature difference model is obtained, the contract feature comparison group is used as the input data of the contract feature difference model, the contract feature difference index Lwzb is output, and the difference calculation capability index Sgtd of the record comparison group is calculated by Sgtd=(Bpset+ec1)*(Lwzb+ec2), wherein ec1 is a first coefficient, ec2 is a second coefficient, the value of ec1 is 1.12, and the value of ec2 is 1.25.

[0039] The construction process of the contract feature difference model: a deep learning model is constructed, k contract feature comparison groups are obtained, the deep learning model is trained on the contract feature comparison groups, and a contract feature difference index is assigned to each contract feature comparison group. The value range of the contract feature difference index is (1.1-6.0). The larger the contract feature difference index, the greater the feature difference between the two contract feature sets in the contract feature comparison group. The k contract feature comparison groups are divided into a training set, a validation set, and a test set in a ratio of 70:15:15. The training set, the validation set, and the validation set are trained, and the contract feature difference model is constructed after the training is completed.

[0040] The security calculation record includes security calculation efficiency, pre-encryption hash value (a fixed-length string obtained by performing hash operation on the encrypted fragment data received by the node before the start of the calculation task), post-encryption hash value (a value obtained by performing hash operation on the decrypted plaintext result or encrypted state calculation result after the completion of the calculation), and contract feature set.

[0041] The contract feature set of the security calculation record is obtained as follows: the encrypted fragment corresponding to the security calculation record is determined, and the smart contract corresponding to the encrypted fragment is determined. The input data format, calculation accuracy requirement, and result delivery condition of the smart contract are extracted to obtain the input data format feature (including data type, data structure, data format constraint, etc.), calculation accuracy requirement feature (including data accuracy, time accuracy, and calculation logic accuracy), and result delivery condition feature (including delivery form and delivery actuality). The input data format feature, calculation accuracy requirement feature, and result delivery condition feature are combined to form the contract feature set.

[0042] The splitting rule analysis module performs encryption processing (such as using a symmetric encryption algorithm) on the original data, generates a data ciphertext after encryption processing, determines all applicable splitting rules for the data ciphertext (such as splitting according to time, geographic location, data type, and other features, and the applicable splitting rules are different depending on the features), and obtains the comprehensive security calculation matching index of the blockchain under each splitting rule.

[0043] The multi-party secure calculation module marks the splitting rule with the largest comprehensive security calculation matching index value as the secure splitting rule, splits the data ciphertext into multiple encrypted fragments using the secure splitting rule, sorts all encrypted fragments in descending order of fragment size, sorts all preferred secure calculation nodes in descending order of initial bidding index value, assigns the first encrypted fragment to the first preferred secure calculation node, and so on. After the assignment is completed, each preferred secure calculation node performs security calculation (including fragment decryption, calculation processing, and result aggregation) on the encrypted fragments.

[0044] The process for obtaining the comprehensive security computing matching index of a blockchain under a splitting rule is as follows: Select a splitting rule applicable to the encrypted data (e.g., splitting the encrypted data according to time characteristics). Split the encrypted data according to this rule. After splitting, generate multiple encrypted shards. Obtain the total number of encrypted shards. When the total number of preferred secure computing nodes is greater than or equal to the total number of encrypted shards, calculate the ratio between the total number of encrypted shards and the total number of preferred secure computing nodes to obtain the residual allocation matching index Sgpe. Sort all preferred secure computing nodes in descending order of their initial election index values. Calculate the ratio between the initial election index of the next adjacent preferred secure computing node and the initial election index of the previous preferred secure computing node to obtain the election gradient ratio. Sort all election gradient ratios in descending order of their values. Calculate the absolute difference between two adjacent election gradient ratios after sorting to obtain the election gradient ratio. The gradient difference ratio is calculated by summing and averaging all the election gradient difference ratios to obtain the average election gradient difference ratio. The encrypted shard volume is obtained by sorting all encrypted shards in descending order of volume. The volume gradient ratio is calculated by comparing the volume of the next adjacent encrypted shard with the volume of the previous encrypted shard. All volume gradient ratios are then sorted in descending order of value. The absolute difference between any two adjacent volume gradient ratios is calculated to obtain the volume gradient difference ratio. Finally, the average volume gradient difference ratio is calculated by summing and averaging all volume gradient difference ratios. The absolute difference between the average election gradient difference ratio and the average volume gradient difference ratio is calculated to obtain the election computation difference gradient value Ficgv (a smaller election computation difference gradient value indicates a closer match between the gradient distribution of the encrypted shard volume and the initial election exponential gradient of the preferred secure computing node). The comprehensive security computing matching index Rwsz of the blockchain under this splitting rule is calculated, where ec3 is the third coefficient and the value of ec3 is 0.51.

[0045] The system includes a secure computing task determination module and a block node screening module. After receiving a secure computing task, the system analyzes the basic computing performance of nodes in the blockchain and deeply analyzes the comprehensive computing capabilities of each node for different smart contracts. This allows for the accurate and rapid screening of nodes that can perform secure computing tasks. The system also includes a splitting rule analysis module and a multi-party secure computing module. These modules deeply analyze the comprehensive matching degree between nodes and encrypted shards in the blockchain under different splitting rules, ensuring the allocation and utilization of node resources in the blockchain and releasing the distributed computing potential of the blockchain network.

[0046] Example 2: Refer toFigure 3 A multi-party secure calculation method based on a blockchain, comprising the following steps:

[0047] Step 1: determining a secure calculation task of an initiator and original data, and encapsulating the secure calculation task as a smart contract;

[0048] Step 2: determining preferred secure calculation nodes in the blockchain;

[0049] Step 3: performing encryption processing on the original data, generating data ciphertext after the encryption processing, determining all applicable splitting rules of the data ciphertext, and obtaining a comprehensive secure calculation matching index of the blockchain under each splitting rule;

[0050] Step 4: marking the splitting rule with the largest comprehensive secure calculation matching index value as a secure splitting rule, splitting the data ciphertext into multiple encrypted fragments using the secure splitting rule, and arranging the preferred secure calculation nodes to perform secure calculation.

[0051] Based on the above method, in the process of reasonably allocating node resources and encrypted fragments by deeply analyzing the matching degree, the data ciphertext is always processed according to the optimal splitting rule and allocated to the most suitable node, and each node only needs to process the encrypted fragment allocated to itself without touching other irrelevant data, further strengthening the data privacy protection.

[0052] The above formulas are all dimensionless to calculate the numerical values, and the preset parameters in the formulas are set by a person skilled in the art according to the actual situation.

[0053] The above-described embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, the above-described embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center through a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0054] It should be understood that the size of the sequence number of each process described above in various embodiments of the present application does not mean the order of execution, and the execution order of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0055] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be realized in electronic hardware, or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0056] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0057] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the division of the above-described device embodiments is only a logical function division, and there can be another division manner for actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the among different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0058] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and various other media that can store program codes.

[0059] The above describes only the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A blockchain-based multi-party secure computation system, characterized in that, It includes a secure computation task determination module, a block node filtering module, a splitting rule analysis module, and a multi-party secure computation module; The secure computing task determination module determines the secure computing task and raw data of the initiator, encapsulates the secure computing task into a smart contract, and determines the input data format, computing accuracy requirements and result delivery conditions of the smart contract. The block node filtering module determines the preferred secure computing nodes in the blockchain; The preferred secure computing nodes in the blockchain are determined as follows: the initial election index of each node in the blockchain is determined, and an initial election boundary index is set. When the initial election index of a node is greater than or equal to the initial election boundary index, the corresponding node is marked as a preliminary secure computing node, and the preliminary secure computing nodes that currently have no secure computing tasks are marked as preferred secure computing nodes. The initial election index of a node is determined as follows: Select a node in the blockchain, obtain all secure computation records of that node within a time period t before the current system time, match all secure computation records pairwise, mark the two matched secure computation records as a record comparison group, obtain the difference computing power index of each record comparison group, sum and average all difference computing power indices to calculate the average difference computing power index, compare the consistency of the pre-secret hash value and the post-secret hash value in the secure computation records, when the pre-secret hash value and the post-secret hash value are consistent, increase the number of secure computations by one, and obtain the initial election index of the node based on the product of the average difference computing power index and the number of secure computations. The splitting rule analysis module is used to encrypt the original data, generate ciphertext after encryption, determine all applicable splitting rules for the ciphertext, and obtain the comprehensive security calculation matching index of the blockchain under each splitting rule. The multi-party secure computation module marks the splitting rule with the largest comprehensive secure computation matching index value as the secure splitting rule, uses the secure splitting rule to split the encrypted data into multiple encrypted fragments, and arranges the preferred secure computation nodes to perform secure computation. The process for obtaining the comprehensive security computing matching index of a blockchain under a splitting rule is as follows: Select a splitting rule applicable to the encrypted data, split the encrypted data according to the splitting rule, generate multiple encrypted fragments after splitting, obtain the total number of encrypted fragments, and when the total number of preferred secure computing nodes is greater than or equal to the total number of encrypted fragments, calculate the ratio between the total number of encrypted fragments and the total number of preferred secure computing nodes to obtain the surplus allocation matching index. Then, calculate the candidate computing difference gradient value. Based on the surplus allocation matching index and the candidate computing difference gradient value, determine the comprehensive security computing matching index of the blockchain under the splitting rule. The election gradient value is calculated as follows: All preferred secure computing nodes are sorted in descending order of their initial election index values. The ratio of the initial election index of the next adjacent preferred secure computing node to the initial election index of the previous preferred secure computing node is calculated to obtain the election gradient ratio. All election gradient ratios are then sorted in descending order of their values. The absolute difference between any two adjacent election gradient ratios is calculated to obtain the election gradient difference ratio. Finally, all election gradient difference ratios are summed and averaged to obtain the average election gradient difference ratio. The encrypted shards are then obtained. For volume, all encrypted fragments are sorted in descending order of fragment volume. The volume of the next adjacent encrypted fragment is calculated as the ratio of the volume of the previous encrypted fragment to obtain the volume gradient ratio. All volume gradient ratios are then sorted in descending order of value. The absolute difference between any two adjacent volume gradient ratios is calculated to obtain the volume gradient difference ratio. All volume gradient difference ratios are summed and averaged to obtain the average volume gradient difference ratio. The absolute difference between the average volume gradient difference ratio and the average volume gradient difference ratio is calculated to obtain the competition calculation difference gradient value.

2. The blockchain-based multi-party secure computation system according to claim 1, characterized in that, The difference computing capability index of the record comparison group is obtained as follows: The security computing efficiency of the two security computing records in the record comparison group is summed to obtain the security computing efficiency within the group. The contract feature sets of the two security computing records in the record comparison group are matched into a contract feature comparison group to obtain the contract feature difference model. The contract feature comparison group is used as the input data of the contract feature difference model to obtain the contract feature difference index. The difference computing capability index of the record comparison group is calculated based on the security computing efficiency within the group and the contract feature difference index.

3. The blockchain-based multi-party secure computation system according to claim 1, characterized in that, Secure computation records include secure computation efficiency, pre-secret hash value, post-secret hash value, and contract feature set.

4. The blockchain-based multi-party secure computation system according to claim 2, characterized in that, The process of obtaining the contract feature set of the secure computation record is as follows: determine the encrypted shard to which the secure computation record is applied, and then determine the smart contract corresponding to the encrypted shard. Extract features from the input data format, computation precision requirements, and result delivery conditions of the smart contract. Extract the input data format features, computation precision requirement features, and result delivery condition features. Combine the input data format features, computation precision requirement features, and result delivery condition features into the contract feature set.

5. The blockchain-based multi-party secure computation system according to claim 1, characterized in that, Arrange preferred secure computing nodes to perform secure computations: Sort all encrypted fragments in descending order of fragment size, sort all preferred secure computing nodes in descending order of their initial election index values, assign the encrypted fragment with the highest value to the preferred secure computing node with the highest value, and so on. After the allocation is completed, each preferred secure computing node performs secure computations on the encrypted fragments.

6. A blockchain-based secure multi-party computation method, applied to the blockchain-based secure multi-party computation system according to any one of claims 1-5, characterized in that, Includes the following steps: Step 1: Determine the initiator's secure computing task and the original data, and encapsulate the secure computing task into a smart contract; Step 2: Determine the preferred secure computing nodes in the blockchain; Step 3: Encrypt the original data to generate ciphertext. Determine all applicable splitting rules for the ciphertext and obtain the comprehensive security calculation matching index of the blockchain under each splitting rule. Step 4: Mark the splitting rule with the largest comprehensive security calculation matching index value as the security splitting rule, use the security splitting rule to split the encrypted data into multiple encrypted fragments, and arrange the preferred security calculation node to perform the security calculation.

Images