Bid secrecy management and control method and device based on multi-party secure computation

Abstract

The application provides a bidding secret control method and device based on multi-party secure computation, which comprises the following steps: an initial key matrix is generated by a tendering party, a random disturbance value is input into the initial key matrix by each bidder, and a dynamically changing key pool is formed; the tendering party converts bidding data into a feature vector associated with an element in the key pool, and embeds the feature vector into a transmission stream and transmits the transmission stream to a security control node; after receiving the transmission stream, the security control node generates a matching mask based on the tendering rules; a matching result is obtained by performing a modulo operation on the feature vector and the matching mask, and the matching result is transmitted to the tendering party; after obtaining the matching result, the tendering party generates a winning feature code and sends the winning feature code to the security control node; the security control node only partially decrypts the feature vector matched with the winning feature code, generates result information containing the winning identifier but not leaking the bidding data, and pushes the result information to the corresponding tendering party and the tendering party. In the application, the defect that the bidding secret control method is prone to leaking bidding data is overcome.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a bidding secret control method and device based on multi-party secure computation. BACKGROUND

[0002] In the bidding activity, the secrecy of the bidding data is directly related to the fairness and impartiality of the bidding process. The traditional bidding secret control method mainly relies on the encryption transmission mode, but this mode has many limitations in practical application.

[0003] The traditional encryption transmission mode encrypts the bidding data by using a fixed key, and the generation and management of the key are concentrated in a single subject (such as the bidding party or a third-party platform). Once the key is leaked or cracked, the security of all bidding data will be threatened. At the same time, when the encrypted data participates in the bidding rule matching calculation, it often needs to be decrypted into plaintext first. There is a risk of data interception or tampering in this process, and it is difficult to ensure the secrecy of the bidding data in the whole process. SUMMARY

[0004] The main purpose of the present application is to provide a bidding secret control method and device based on multi-party secure computation, which aims to overcome the defect that the current bidding secret control method is easy to leak bidding data.

[0005] To achieve the above purpose, the present application provides a bidding secret control method based on multi-party secure computation, comprising the following steps:

[0006] The bidding party inputs a random disturbance value into the initial key matrix to form a dynamic changing key pool;

[0007] The bidding party converts the bidding data into a feature vector associated with the elements in the key pool and embeds it into a transmission stream for transmission to a security control node;

[0008] The security control node receives the transmission stream and generates a matching mask based on the bidding rules without decrypting the feature vector. The matching result is obtained by the modulo operation of the feature vector and the matching mask and is transmitted to the bidding party;

[0009] After the bidding party obtains the matching result, it generates a winning feature code and sends it to the security control node;

[0010] The security control node only partially decrypts the feature vector matching the winning feature code, generates a result information containing the winning identifier but not leaking the bidding data, and pushes it to the corresponding bidding party and bidding party.

[0011] Further, the key elements in the key pool are only activated temporarily based on real-time network features at the moment of bidding data submission, and automatically deactivated after single activation.

[0012] Further, the bidder converts the bidding data into a feature vector associated with the elements in the key pool, including:

[0013] The bidder splits the bidding data by pre-set dimensions to obtain multiple data sub-items;

[0014] A unique identifier is assigned to each data sub-item, which forms a mapping relationship with a selected element in the key pool, and the mapping rule is determined by a hash function generated by the bidder based on its private key;

[0015] Based on the mapping relationship, the data sub-items and the corresponding key pool elements are subjected to XOR operation to generate feature components;

[0016] All feature components are combined in the order of splitting of each data sub-item to form a feature vector associated with the elements in the key pool.

[0017] Further, the matching result obtained by the feature vector and the matching mask modulo operation is transmitted to the tenderer, including:

[0018] The security management node decomposes the feature vector into sub-vectors corresponding to each sub-item of the tendering rules, and each sub-vector corresponds to an evaluation dimension in the bidding data;

[0019] A weight matrix consistent with the dimension of the sub-vector is set for the matching mask, and the weight values are dynamically generated based on the importance of each evaluation dimension;

[0020] The modulo operation is performed on each sub-vector and the corresponding sub-mask in the matching mask to obtain a dimension matching value;

[0021] The dimension matching values are weighted and summed with the weight matrix to generate a comprehensive matching degree;

[0022] If the comprehensive matching degree is greater than a pre-set threshold, a matching result containing a matching success identifier and key dimension compliance information is generated.

[0023] Further, the key dimension compliance information is obtained through the modulo operation remainder range mapping, and does not contain the original bidding data.

[0024] Further, after the tenderer obtains the matching result, a winning feature code is generated, including:

[0025] The tenderer extracts the feature vectors corresponding to the bidding parties with a comprehensive matching degree greater than a pre-set threshold from the matching result to form a candidate set;

[0026] A ranking algorithm based on the priority of bidding rules is used to sort the candidate set by weight to generate a ranking sequence; wherein, the ranking parameters of the ranking algorithm are embedded in the ranking sequence after being encrypted with the private key of the bidding party.

[0027] Select the feature vector corresponding to the first bidder in the sorted sequence, extract its core mapping relationship with the key pool elements, and generate the basic feature code;

[0028] The basic feature code is salted, and the salted feature code is encrypted using the public key of the security control node to generate the winning feature code.

[0029] Furthermore, the digital signature of the bidding party is embedded in the public key encryption process, and a signature verification request is attached when the public key is sent to the security control node.

[0030] Furthermore, the security control node only partially decrypts the feature vector matching the winning bid feature code, generating result information containing the winning bid identifier but without disclosing the bidding data, and pushes it to the corresponding bidder and tendering party, including:

[0031] The security control node extracts the core mapping relationship from the winning bid feature code through decryption;

[0032] The core mapping relationship is compared with the key pool to locate the corresponding feature vector, and the components of the bidding core data involved in the feature vector are masked by a preset field masking rule.

[0033] Perform targeted decryption on the feature vector after masking the components, decrypting only the identifier fields related to the determination of the winning bid status;

[0034] The result information is generated, which includes the winning bid identifier, the bidder code, and an encrypted compliance statement; wherein the compliance statement is encrypted using the bidder's public key and can only be decrypted and viewed by the corresponding bidder.

[0035] The system pushes result information containing the complete winning bid identifier to the bidding party, and pushes encrypted information containing only its own winning bid status to the corresponding bidders.

[0036] Furthermore, the bidding party generates an initial key matrix, and each bidder inputs a random perturbation value into the initial key matrix to form a dynamically changing key pool, including:

[0037] The bidding party extracts the technical field feature vectors of each bidder based on the summary information of the pre-uploaded bid documents, and dynamically determines the dimension parameters of the initial key matrix according to the Euclidean distance clustering results of the technical field feature vectors; the matrix elements are generated using encryption parameters to obtain the initial key matrix.

[0038] The bidding party generates a random disturbance value, analyzes the distribution entropy of the numerical field in the bidding data, adjusts the Lyapunov exponent of the chaotic mapping based on the distribution entropy, and sets the length parameter of the random disturbance value as the prime factor decomposition result of the number of words in the bidding file.

[0039] The tendering party aggregates to generate a key pool, introduces a weighting factor based on the historical bidding deviation rate of the bidding party, calculates the final value of the key pool element as the weighted product of the matrix element and the random disturbance value, and dynamically generates the weight parameter through the sigmoid function of the historical bidding deviation rate of the bidding party.

[0040] Further, the tendering party generates an initial key matrix, and each bidding party inputs a random disturbance value to the initial key matrix to form a dynamically changing key pool, which comprises:

[0041] The tendering party analyzes the technical complexity level of the bidding project, and sets the size of the initial key matrix according to the complexity from low to high; the matrix element is generated based on the encryption parameter of lattice cryptography, and the encryption strength is adjusted in sections according to the project budget amount;

[0042] When the bidding party generates a random disturbance value, the bidding party first identifies the most frequently modified sensitive paragraph in the bidding file, adjusts the randomness of the chaotic algorithm used to generate the random disturbance value based on the modification frequency, and determines the length of the random disturbance value according to the total number of pages of the bidding file;

[0043] During the injection of the random disturbance value into the matrix, the number of current bidding parties is monitored in real time, and based on the number of bidding parties, the encryption threshold used in the operation of the disturbance value and the matrix element and the length of the hash value of the operation result are determined;

[0044] When the tendering party aggregates to generate a key pool, the compliance records of the historical bidding of each bidding party are combined to determine the weight proportion of the random disturbance value of each bidding party in the final result of the key pool; and the size of the sliding window used in the aggregation operation is adjusted according to the average length of time for the bidding party to respond to the tender;

[0045] After the generation of the key pool, the update interval gradually shortens as the tender deadline approaches, and the anti-interference parameter of the encryption algorithm is automatically enhanced each time the key pool is updated.

[0046] The application also provides a bidding secret control device based on multi-party secure computation, which comprises a tendering party, a plurality of bidding parties and a security control node.

[0047] The tendering party generates an initial key matrix, and each bidding party inputs a random disturbance value to the initial key matrix to form a dynamically changing key pool.

[0048] The bidding party converts the bidding data into a feature vector associated with an element in the key pool, and embeds the transmission stream into a secure management node;

[0049] The secure management node receives the transmission stream, and generates a matching mask based on the bidding rules without decrypting the feature vector; a matching result is obtained by a modulo operation of the feature vector and the matching mask, and is transmitted to the bidding party;

[0050] The bidding party obtains the matching result, generates a winning feature code, and sends it to the secure management node;

[0051] The secure management node only partially decrypts the feature vector matched with the winning feature code, generates result information containing the winning identifier but not leaking the bidding data, and pushes it to the corresponding bidding party and bidding party.

[0052] The bidding party converts the bidding data into a feature vector associated with an element in the key pool, and embeds the transmission stream into a secure management node; BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is a bidding secret management method based on multi-party secure computing in an embodiment of the present application;

[0054] Figure 2 is a bidding secret management device structure block diagram based on multi-party secure computing in an embodiment of the present application;

[0055] Figure 3 is a structural schematic block diagram of a computer device in an embodiment of the present application.

[0056] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0057] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0058] With reference to Figure 1 In an embodiment of the present application, a bid secret management method based on multi-party secure computation is provided, comprising the following steps:

[0059] Step S1, the tenderer generates an initial key matrix, and each bidder inputs a random disturbance value into the initial key matrix to form a dynamically changing key pool;

[0060] Step S2, the tenderer converts the bid data into a feature vector associated with the elements in the key pool, and embeds the transmission stream into the secure management node;

[0061] Step S3, after receiving the transmission stream, the secure management node generates a matching mask based on the tendering rules without decrypting the feature vector; the matching result is obtained by the modulo operation of the feature vector and the matching mask, and is transmitted to the tenderer;

[0062] Step S4, after obtaining the matching result, the tenderer generates a winning feature code and sends it to the secure management node;

[0063] Step S5, the secure management node only partially decrypts the feature vector matching the winning feature code, generates result information containing the winning identifier but not leaking the bid data, and pushes it to the corresponding tenderer and tenderer.

[0064] In the embodiment, as described in step S1, a key base dynamically changing and generated by multiple parties is constructed to provide security support for subsequent encryption of bidding data. Specifically, the tendering party first generates an initial key matrix according to the number of bidding parties, the security level of the project, and other factors. The matrix is composed of elements with encryption characteristics, and there is a specific mathematical correlation between the elements. After obtaining the initial key matrix, each bidding party independently generates a random disturbance value based on the characteristics of its own bidding data (such as data sensitivity, its own identity, etc.). The disturbance value is unique and does not contain any information that can be directly associated with the bidding content. Subsequently, each bidding party injects the random disturbance value it generates into the specified position of the initial key matrix through a secure channel. Distributed processing mechanisms can be used in the injection process to ensure that any participating party (including the tendering party) cannot obtain the complete set of disturbance values alone. Finally, the matrix after all bidding parties inject the disturbance values forms a dynamic key pool, and the element values of the matrix will be updated in real time as the disturbance values change. Any element that is accessed will trigger local dynamic adjustment, thereby avoiding static cracking of the key pool and providing dynamic security for subsequent data encryption.

[0065] As described in step S2, the transformation of bidding data from the original form to the secure transmission form is achieved. The core is to establish the association and binding of data and the key pool, while ensuring the concealment of the transmission process. In an embodiment, after completing the arrangement of the bidding data, the bidding party first structures and splits the original bidding data according to the preset dimensions (such as price, qualifications, technical parameters, etc.), obtaining multiple independent data sub-items. For each data sub-item, the bidding party establishes a unique association with a randomly selected element in the key pool through a preset mapping rule. The mapping rule is generated by the bidding party based on its own private key, ensuring that the association can only be recognized by itself. Subsequently, the bidding party standardizes each data sub-item and encrypts it in combination with the associated key pool element to generate a series of feature components. Then, the feature components are combined into a complete feature vector in the logical order of the data sub-items, so that the vector contains both the core information of the bidding data and the encryption protection through the binding with the key pool. In the transmission stage, the bidding party embeds the feature vector in the redundancy field of the regular data transmission stream instead of generating an independent bidding data packet, so that the transmission process of the feature vector is hidden in normal communication, reducing the risk of targeted interception, and finally securely transmitted to the security control node.

[0066] As described in the above step S3, the security calculation and preliminary screening of the bid data are implemented, and the core is to complete the matching verification of the bidding rules without decrypting the data. After the security control node receives the transmission stream containing the feature vector, it first extracts the feature vector of each bidder from the transmission stream through a special analysis algorithm, and no decryption operation is involved in the extraction process, ensuring that the feature vector is always in an encrypted state. Subsequently, the security control node generates a matching mask according to the bidding rules preset by the bidder, which is a structured data consistent with the dimension of the feature vector, and contains the eligible judgment standard of each evaluation dimension. In the calculation stage, the security control node performs a modulo operation on the feature vector and the matching mask: for each component of the feature vector, a modulo operation is performed on the judgment standard of the corresponding dimension in the matching mask, and the dimension matching result is obtained by judging whether the dimension meets the bidding rules according to the numerical range of the operation result; and then all dimension matching results are integrated and weighted to obtain the matching result reflecting the overall compliance of the bid data. In the entire calculation process, the feature vector remains in an encrypted state, and the security control node cannot analyze any original bid data, but only obtains the compliance judgment result, which is securely transmitted to the bidder.

[0067] As described in the above step S4, the preliminary determination of the winning bidder and the generation of the directional decryption credential are implemented, and the core is to ensure the security of the winning bid determination and the uniqueness of the winning feature code. After the bidder receives the matching result transmitted by the security control node, it first verifies the integrity and authenticity of the result to confirm that it has not been tampered with. Subsequently, the bidder selects all candidate bidders that meet the pre-set qualified threshold from the matching result to form a candidate set, and sorts the candidate set according to the priority of each evaluation dimension in the bidding rules to determine the optimal candidate object. To realize the directional decryption of the optimal candidate object, the bidder extracts the core association between the feature vector of the candidate object and the key pool element, generates a basic feature code combined with the current timestamp, the unique information of the bidding project identifier, and then encrypts the basic feature code through its own private key to generate the final winning feature code. The winning feature code only has a unique matching relationship with the feature vector of the optimal candidate object, and contains the digital signature of the bidder to ensure that it cannot be forged. Finally, the bidder sends the winning feature code to the security control node through an encrypted channel as the only credential for subsequent partial decryption.

[0068] As described in the above step S5, the security disclosure of the bid result is realized, and the core is to complete the targeted push of the result under the premise of protecting the privacy of the bid data. After the security control node receives the bid-winning feature code, it first verifies the digital signature of the tendering party contained therein to confirm its legality and integrity. After verification, the security control node locates the corresponding target feature vector among all feature vectors according to the association relationship contained in the bid-winning feature code. Subsequently, the security control node performs partial decryption on the target feature vector. For example, by using a preset field mask rule, the components of the feature vector related to the core data of the bid (such as specific bid, technical parameter details, etc.) are shielded, and only the necessary information related to the bid-winning state determination (such as the unique identifier of the bidder, the compliance conclusion, etc.) is decrypted. After decryption, the security control node generates result information, which contains the bid-winning identifier, the unique code of the bidder, and the encrypted compliance description (the compliance description can only be decrypted by the corresponding bidder through its own private key). Finally, a double-channel push mechanism is used to push the result information containing the complete bid-winning identifier to the tendering party, and to push the encrypted information containing only the bid-winning state of the corresponding bidder to the corresponding bidder. The dynamic session key is used to ensure the security of the transmission during the push process, so that the bid result disclosure is completed while strictly protecting the bid data privacy of the non-winning bidder and the core business information of the winning bidder.

[0069] In an embodiment, the key elements in the key pool are only temporarily activated based on real-time network characteristics at the moment of submitting the bid data, and automatically deactivated after a single activation.

[0070] In this embodiment, each key element in the key pool is not in a continuously available state, and its activation has strict time effectiveness and condition restrictions. Only at the moment when the bidder submits the bid data, the real-time network characteristics (such as network delay, data packet transmission sequence, etc.) at that time are captured as trigger conditions to temporarily activate the corresponding key element to participate in data encryption operations. Once the encryption of the bid data is completed, the key element is automatically deactivated and cannot be used for encryption or decryption of other data, thereby ensuring the security and uniqueness of key usage.

[0071] In an embodiment, the bidder converts the bid data into a feature vector associated with an element in the key pool, including:

[0072] The bidder splits the bid data according to a preset dimension to obtain a plurality of data sub-items;

[0073] A unique identifier is assigned to each data sub-item, and the identifier forms a mapping relationship with a selected element in the key pool, and the mapping rule is determined by a hash function generated by the bidder based on its private key;

[0074] Based on the mapping relationship, the data sub-item is XORed with the corresponding key pool element to generate a feature component;

[0075] All feature components are combined according to the splitting order of each data sub-item to form a feature vector associated with the elements in the key pool.

[0076] In this embodiment, the bidder first needs to perform structured processing on the original bid data, and split the complete bid data into multiple independent data sub-items according to the preset dimensions (such as bid information, qualification certificate, technical scheme parameter, service commitment, etc.). The splitting process needs to follow the internal logical association of the data to ensure that each data sub-item contains only information of a single category, for example, the bid information is split into sub-items such as total price, unit price, payment period, etc., and the qualification certificate is split into sub-items such as enterprise qualification level, relevant certification certificate, past performance, etc. Through dimension splitting, the originally complex bid data is converted into structured data units with clear classification.

[0077] After completing the data sub-item splitting, the bidder assigns a unique identifier to each data sub-item, which is in the form of a non-repetitive string or code, used to uniquely identify the corresponding data sub-item. Subsequently, the bidder randomly selects an element from the key pool, so that the identifier forms a one-to-one mapping relationship with the selected key pool element. It is worth noting that the establishment of the mapping relationship follows a specific rule generated by the bidder, for example, the bidder generates a special hash function based on the private key it holds, which takes the identifier as input and directly points to the corresponding element in the key pool through hash operation, ensuring the uniqueness and privacy of the mapping process. Since the hash function is generated by the private key, other participants (including the tenderer and other bidders) cannot crack the mapping rule, thereby ensuring the security of the association between the data sub-item and the key pool element.

[0078] After determining the mapping relationship between the data sub-item and the key pool element, the bidder performs encryption transformation on each data sub-item. According to the above mapping relationship, the key pool element corresponding to the data sub-item is called, and the original value of the data sub-item (after digital processing, text data needs to be converted into corresponding numerical encoding) is XORed with the key pool element. XOR operation as a basic encryption operation, its result (i.e. feature component) contains the information of data sub-item and key pool element, and cannot be individually deduced from the original data sub-item or key pool element through the feature component. Through the above operation, the original data sub-item is converted into a feature component with encryption characteristics, which not only retains the core information of the data sub-item, but also realizes the preliminary encryption of the data through the combination with the key pool element, providing support for the confidentiality in the subsequent transmission and calculation process.

[0079] After generating the feature components corresponding to all data sub-items, the bidder arranges and combines all feature components in the original splitting order of the data sub-items (such as the order of the bid sub-item, the qualification sub-item, and the technical parameter sub-item) to form a complete vector structure, that is, a feature vector. Each component in the feature vector corresponds to the result of the original data sub-item after encryption processing, and since each feature component is generated by an XOR operation with a key pool element, the entire feature vector is closely associated with multiple elements in the key pool. The above combination in the original order not only ensures the logical consistency of the feature vector and the original bid data, but also realizes the encryption protection of the entire data through the association with the key pool elements, so that the feature vector cannot be directly parsed as the original bid data even if it is intercepted during subsequent transmission and calculation.

[0080] In an embodiment, the matching result obtained by the modulo operation of the feature vector and the matching mask is transmitted to the tendering party, including:

[0081] The security management node decomposes the feature vector into sub-vectors corresponding to each sub-item of the tendering rules, and each sub-vector corresponds to an evaluation dimension in the bid data;

[0082] A weight matrix with the same dimension as the sub-vectors is set for the matching mask, and the weight values are dynamically generated based on the importance of each evaluation dimension;

[0083] The modulo operation is performed on each sub-vector and the corresponding dimension sub-mask in the matching mask to obtain a dimension matching value;

[0084] The dimension matching values are weighted and summed with the weight matrix to generate a comprehensive matching degree;

[0085] If the comprehensive matching degree is greater than a preset threshold, a matching result containing a matching success identifier and key dimension compliance information is generated.

[0086] In this embodiment, after receiving the feature vector, the security management node first structurally decomposes the feature vector according to each evaluation sub-item (such as price reasonableness, qualification compliance, and technical scheme feasibility) included in the tendering rules. The decomposition process needs to strictly follow the original arrangement logic of each feature component in the feature vector to ensure that each sub-vector obtained after decomposition forms a unique corresponding relationship with one evaluation sub-item in the tendering rules, and each sub-vector corresponds to an independent evaluation dimension in the bid data. For example, if the tendering rules include three sub-items of “bid range”, “enterprise qualification level”, and “technical parameter compliance rate”, the feature vector will be decomposed into three sub-vectors corresponding to the three evaluation dimensions of bid, qualification, and technology. Through this decomposition, the structure of the feature vector is precisely mapped to the evaluation system of the tendering rules.

[0087] When generating the matching mask, the security management node synchronously constructs a weight matrix that is completely consistent with the dimension of the sub-vector. Each element in the weight matrix represents the importance of the corresponding evaluation dimension in the overall bidding rules. For example, in a bidding that focuses on price as the core consideration factor, the weight value corresponding to the bid price dimension will be higher than that of other dimensions. The generation of the weight value is not fixed, but is dynamically adjusted according to the specific needs of the bidding project (such as project nature, fund size, technical difficulty, etc.): for projects that focus on technological innovation, the weight value of the technical scheme dimension will be correspondingly increased; for budget-sensitive projects, the weight value of the price dimension will be given higher priority. This dynamic generation mechanism ensures that the weight matrix can accurately reflect the core demand of the bidding rules, so that the subsequent comprehensive matching calculation result is more in line with the actual evaluation needs.

[0088] After completing the sub-vector decomposition and weight matrix setting, the security management node performs matching calculation for each evaluation dimension: for each sub-vector, the sub-mask of the corresponding dimension in the matching mask is called (the sub-mask contains the qualified judgment standard of the dimension, such as the price upper limit, the qualification level threshold, etc.), and the sub-vector and the sub-mask are executed. The result of the modulo operation (i.e., the dimension matching value) can quantitatively reflect the matching degree of the sub-item of the bid data corresponding to the sub-vector and the corresponding sub-item in the bidding rules. If the sub-item of the bid data completely meets the judgment standard, the dimension matching value is in the preset qualified interval; if there is a deviation, the dimension matching value will deviate from the interval, and the greater the deviation, the farther the distance between the dimension matching value and the qualified interval. Since the sub-vector and the sub-mask are in an encrypted state, the modulo operation process does not involve decryption of the original data, thereby ensuring the privacy and security of the bid data.

[0089] After obtaining the dimension matching values of all evaluation dimensions, the security management node performs weighted sum calculation on the dimension matching values according to the weight values of the dimensions in the weight matrix. Specifically, the dimension matching value of each dimension is multiplied by the corresponding weight value to obtain the weighted matching value of the dimension; then all weighted matching values of the dimensions are accumulated, and the final result is the comprehensive matching degree. The comprehensive matching degree is a quantitative reflection of the degree to which the bid data as a whole meets the bidding rules. The dimensions with higher weight values have a greater impact on the comprehensive matching degree. For example, if the price dimension has the highest weight value, the high or low of its dimension matching value will significantly determine the final result of the comprehensive matching degree. This weighted sum mechanism enables the comprehensive matching degree to comprehensively and selectively reflect the degree of fit between the bid data and the bidding rules.

[0090] The security management node compares the calculated comprehensive matching degree with a preset threshold (the threshold is set by the tendering party according to the minimum requirement of the project, representing the basic qualified standard that the bidding data needs to meet) : if the comprehensive matching degree is greater than the preset threshold, it means that the bidding data as a whole meets the requirements of the tendering rules, and the security management node will generate a matching result containing a "matching success" identifier, while the result is accompanied by the key dimension compliance information (such as which high-weight dimensions fully meet the requirements, which dimensions have deviations but do not affect the overall eligibility, etc.); if the comprehensive matching degree is less than or equal to the preset threshold, a "matching failure" identifier is generated, and only the key dimension information that does not meet the standard is recorded. The generated matching result can be transmitted to the tendering party through an encrypted channel, and the result does not contain any original bidding data, only reflects the matching status and key dimension compliance, which meets the evaluation needs of the tendering party and maximally protects the privacy of the bidding data.

[0091] In an embodiment, the key dimension compliance information is obtained through a modulus operation remainder range mapping, and does not contain the original bidding data.

[0092] In this embodiment, the generation of key dimension compliance information relies on the mapping relationship between the modulus operation remainder and the preset range, and the entire process does not involve the analysis of the original bidding data. Specifically, in the modulus operation process of the sub-vector and the corresponding sub-mask, different matching degrees will produce remainders in a specific numerical interval: when the bidding data sub-item fully meets the requirements of the tendering rule sub-item, the remainder will fall within the preset compliance range; when there is a slight deviation but still within the acceptable range, the remainder will fall within the basic compliance range; when the deviation exceeds the allowed limit, the remainder will fall within the non-compliance range. The security management node can determine the compliance status of the key dimension by identifying the range to which the remainder belongs. Based on the information obtained through the remainder range mapping, only the eligibility determination result of the dimension is reflected, and no content related to the original bidding data is included, thereby accurately feeding back the key dimension compliance while strictly protecting the privacy of the bidding data.

[0093] In an embodiment, after the tendering party obtains the matching result, a winning feature code is generated, including:

[0094] The tendering party extracts the feature vectors corresponding to the bidding parties whose comprehensive matching degrees are greater than the preset threshold from the matching result to form a candidate set;

[0095] A weight sorting algorithm based on the priority of the tendering rules is used to sort the candidate set to generate a sorting sequence; wherein the sorting parameters of the sorting algorithm are embedded in the sorting sequence after being encrypted by the private key of the tendering party;

[0096] The feature vector corresponding to the bidding party at the top of the sorting sequence is selected, and the core mapping relationship associated with the key pool elements is extracted to generate a basic feature code;

[0097] The base feature code is salted, and the salted feature code is encrypted by the public key of the security management node to generate the winning feature code.

[0098] In this embodiment, after receiving the matching result transmitted by the security management node, the tenderer first parses the result and screens out all the tender information with a comprehensive matching degree exceeding a preset threshold. The tender information corresponding to the tenderers all meet the basic requirements of the tender rules and have the qualifications to become the winning candidate. For each qualified tenderer, the tenderer extracts the feature vector generated in the transmission process. The feature vector is the carrier of the encrypted tender data and is closely associated with the key pool elements and has not been decrypted. The tenderer integrates these feature vectors to form a candidate set containing the encrypted information of all qualified tenderers. The formation process of the candidate set only involves the extraction and aggregation of feature vectors and does not involve the analysis of any original tender data, ensuring the confidentiality of the tender information.

[0099] According to the priority of each evaluation dimension in the tender rules (such as the importance ranking of the price, qualification, technical scheme and other dimensions), the tenderer adopts a preset sorting algorithm to sort the feature vectors in the candidate set by weight. In the sorting process, the algorithm calculates the comprehensive sorting score of each feature vector by combining the comprehensive matching degree corresponding to the feature vector of each tenderer and the weight value of each dimension, and then generates a sorting sequence in descending order of the score. To ensure the confidentiality and tamper resistance of the sorting rules, the core parameters involved in the sorting algorithm (such as the quantization values of the priority of each dimension and the score calculation coefficients) are first encrypted by the private key of the tenderer and then embedded in the sorting sequence in a hidden form. The above encryption embedding mechanism makes it impossible for other participants to crack the sorting logic, ensuring the fairness of the sorting process and preventing the risk of malicious tampering with the sorting parameters.

[0100] After obtaining the sorting sequence, the tenderer selects the feature vector ranked first in the sequence. The tenderer corresponding to the feature vector performs best in the comprehensive evaluation and is the preliminary winning candidate. Then, the tenderer analyzes the feature vector and extracts the core mapping relationship between the feature vector and the key pool elements. The core mapping relationship is established by the hash function generated by the private key of the tenderer when generating the feature vector, and is directly related to the correspondence between the components of the feature vector and the key pool elements. It is the key basis for subsequent decryption operations. Based on the extracted core mapping relationship, the tenderer generates a base feature code that can uniquely identify the feature vector. The base feature code contains key information for positioning the corresponding key pool elements and the structure of the feature vector, but does not involve any original tender data content.

[0101] To further enhance the security of the feature code, the tenderer performs salted processing on the basic feature code: introduces randomly generated interference information (i.e. salt value) into the basic feature code, which is unique and associated with the current timestamp, tender project identifier and other information, so that even if the basic feature code is leaked, a third party cannot directly use it for decryption operation. After salted processing, the tenderer calls the public key of the security management node to encrypt the salted feature code. Due to the characteristics of public key encryption, only the security management node holding the corresponding private key can decrypt the feature code, ensuring the security of the feature code in the transmission process. The final result formed after encryption is the winning feature code, which contains the key information of the positioning target feature vector and is secured by multiple encryption mechanisms, providing a unique and secure credential for subsequent targeted decryption of the security management node.

[0102] In an embodiment, the digital signature of the tenderer is embedded in the public key encryption process, and a signature verification request is attached when sent to the security management node.

[0103] In this embodiment, during the process of public key encryption of the salted feature code to generate the winning feature code, the tenderer will simultaneously perform digital signature embedding operation. Specifically, the tenderer first performs hash operation on the salted feature code based on its own private key to generate a fixed-length digest information, which is the digital signature of the tenderer. Then, the tenderer embeds this digital signature in the feature code data structure encrypted by the public key in a specific format, so that the digital signature and the encrypted feature code form an inseparable whole. When the winning feature code containing the digital signature is generated, the tenderer will attach a signature verification request when sending it to the security management node. This request instruction will explicitly inform the security management node that the received winning feature code contains the digital signature of the tenderer, and the authenticity and integrity of the signature need to be verified.

[0104] After receiving the winning feature code and the signature verification request, the security management node first extracts the digital signature embedded in it, and then calls the public key of the tenderer to decrypt the signature and obtain the corresponding digest information. At the same time, the security management node will perform independent hash operation on the main part of the encrypted feature code to generate a new digest. By comparing whether the two digests obtained are consistent, the signature verification can be completed. If they are consistent, it proves that the winning feature code is indeed generated by the tenderer and has not been tampered with in the transmission process; if they are not consistent, the feature code is invalid and the subsequent decryption operation is refused.

[0105] The mechanism of embedding the digital signature in the public key encryption and attaching the verification request not only guarantees the transmission security of the winning characteristic code through the public key encryption, but also ensures the authenticity of the source and the data integrity of the characteristic code by means of the digital signature and the verification process, effectively prevents the risk of falsification or tampering of the winning characteristic code by malicious third parties, and further strengthens the security and reliability of the entire bidding security control process.

[0106] In an embodiment, the security control node only partially decrypts the feature vector matching the winning characteristic code, generates result information containing the winning identifier but not leaking the bidding data, and pushes it to the corresponding bidder and tenderer, including:

[0107] The security control node extracts the core mapping relationship in the winning characteristic code by decryption;

[0108] The core mapping relationship is compared with the key pool to locate the corresponding feature vector, and the components of the bidding core data involved in the feature vector are shielded by the preset field mask rule;

[0109] The feature vector after shielding the components is executed directional decryption, and only the identifier field related to the winning state determination is decrypted;

[0110] Generate result information, the result information contains the winning identifier, the bidding party code and the encrypted compliance description; wherein the compliance description is encrypted by the public key of the bidding party, and only the corresponding bidding party can decrypt and view;

[0111] The result information containing the complete winning identifier is pushed to the tenderer, and the encrypted information containing only the winning state of the corresponding bidder is pushed to the corresponding bidder.

[0112] In this embodiment, after receiving the winning characteristic code containing the digital signature and the signature verification request, the security control node first decrypts the winning characteristic code by its private key. Since the winning characteristic code is generated by encrypting the public key of the security control node, only its private key can complete the decryption, ensuring the uniqueness and security of the decryption process. After decryption, the security control node extracts the core mapping relationship from the characteristic code. The core mapping relationship is the key information associated with the key pool elements extracted from the target feature vector by the tenderer when generating the basic characteristic code, which contains the core clues for locating the corresponding feature vector. By extracting the core mapping relationship, the security control node obtains the necessary basis for subsequent positioning of the target feature vector, laying the foundation for accurate decryption.

[0113] The security management node compares the extracted core mapping relationship with the elements in the dynamic key pool one by one. Since the core mapping relationship is directly related to the corresponding logic of the target feature vector and the key pool element, through comparison, the feature vector pointed to by the core mapping relationship, i.e. the encrypted bidding data carrier of the first bidder in the sorting sequence, can be accurately located. After positioning, the security management node starts the preset field mask rule: according to the pre-set security level and data sensitivity of the tenderer, the components of the feature vector related to the core data of the bid (such as specific bid values, technical parameter details, cost composition, etc.) are shielded. The shielding operation is realized by applying an encryption mask to these components, so that the shielded components cannot be decrypted in subsequent processing, thereby preventing the leakage of core data from the source.

[0114] After shielding the core data components, the security management node performs directional decryption on the remaining feature vector part. The directional decryption process strictly limits the decryption range and only operates on the identification fields directly related to the bid-winning status determination, including the unique code of the bidder, the submission timestamp of the bid file, the basic conclusion of compliance verification, etc. Such information does not involve the core business content of the bid data, but provides necessary basis for the confirmation of the bid-winning result. During the decryption process, the security management node calls the corresponding elements in the key pool to analyze the identification fields through the preset decryption algorithm, ensuring the accuracy of the decryption result. Since the core data components have been shielded, directional decryption will not cause any leakage of sensitive information.

[0115] After directional decryption, the security management node integrates the decrypted information to generate standardized result information. The result information contains three core parts: first, the bid-winning identification indicates the bid-winning status of the bidder in the form of symbols or codes; second, the bidder code uniquely identifies the bid-winning subject, facilitating the identity correspondence between the tenderer and the bidder; third, the compliance description records the compliance of the bidder in each evaluation dimension (such as which dimension fully meets the requirements and which dimension has acceptable deviations, etc.). It is worth noting that the compliance description is encrypted using the public key of the corresponding bidder. Due to the characteristics of public key encryption, only the private key of the bidder can decrypt and view the content, and any other participants (including the tenderer) cannot obtain the content, which not only ensures the bidder's right to know the details of its own compliance, but also avoids unnecessary diffusion of information among multiple parties.

[0116] The security management node adopts a differentiated pushing mechanism to ensure accurate transmission and security of the result information. For the tendering party, the pushed result information includes the complete bid-winning identifier, the tenderer code and the unencrypted basic compliance conclusion to meet the needs of the tendering party to confirm the bid-winning subject and carry out the subsequent process. For the bid-winning tenderer, the pushed information is encrypted and only includes the bid-winning identifier of the tenderer and the encrypted compliance description. The encryption key is a dynamically generated session key and is only valid in the current pushing process. This differentiated mechanism ensures that the tendering party obtains the necessary bid-winning information and the tenderer can only know the bid-winning status and exclusive compliance description, avoiding cross leakage of the bid-winning result among the tenderers.

[0117] In an embodiment, the tendering party generates an initial key matrix, and each tenderer inputs a random disturbance value into the initial key matrix to form a dynamically changing key pool, including:

[0118] The tendering party extracts the technical field feature vector of each tenderer based on the pre-uploaded summary information of the tender file, and dynamically determines the dimension parameter of the initial key matrix according to the Euclidean distance clustering result of the technical field feature vector. The matrix elements are generated by using encryption parameters to obtain the initial key matrix.

[0119] When the tenderer generates a random disturbance value, the distribution entropy of the numerical field in the tender data is analyzed, and the Lyapunov exponent of the chaotic mapping is adjusted based on the distribution entropy. The length parameter of the random disturbance value is set as the prime factor decomposition result of the word count of the tender file.

[0120] When the tendering party aggregates and generates the key pool, a weighted factor based on the historical bidding deviation rate of the tenderer is introduced, and the final value of the key pool element is calculated as the weighted product of the matrix element and the random disturbance value. The weight parameter is dynamically generated by the sigmoid function of the historical bidding deviation rate of the tenderer.

[0121] In this embodiment, after receiving the bid file summary information pre-uploaded by each bidder, the tenderer first parses the summary, extracts the key information reflecting the technical field of the bidder (such as technical classification, core technology keywords, past project technology labels, etc.), and converts these information into structured technical field feature vectors. Each component of the technical field feature vector corresponds to a quantitative value of a technical feature, and the vector dimension is determined by the total number of technical features. Subsequently, the tenderer calculates the Euclidean distance between all feature vectors, and classifies the bidders with similar technical fields into a class through a clustering algorithm (such as the K-means algorithm), obtaining several clustering groups. The dimension parameters of the initial key matrix are dynamically set according to the clustering results: generally, the number of rows and columns of the matrix is related to the number of clusters and the number of bidders in each cluster, for example, the more the number of clusters and the more the number of bidders in a single cluster, the larger the matrix dimension, to ensure that the matrix can cover the technical field feature differences of all bidders. After determining the dimension, the tenderer generates matrix elements using a pre-set encryption algorithm (such as lattice cryptography algorithm, elliptic curve encryption algorithm, etc.), and the element values are dynamically adjusted by encryption parameters (such as key length, random number seed) to ensure uniqueness and security, finally forming the initial key matrix, providing a basic framework for subsequent perturbation value injection.

[0122] Before generating a random perturbation value, the bidder first performs statistical analysis on the numerical fields (such as bid price, project period, technical parameter indicators, etc.) in its own bid data, and calculates the distribution entropy. The distribution entropy reflects the degree of dispersion and uncertainty of numerical data, and the higher the entropy value, the stronger the randomness and complexity of the data. According to the calculation result of the distribution entropy, the bidder dynamically adjusts the Lyapunov exponent of the chaotic mapping algorithm: when the distribution entropy is high, the Lyapunov exponent is increased to enhance the divergence of the chaotic mapping, so that the generated perturbation value has higher randomness; when the distribution entropy is low, the exponent is appropriately reduced to ensure the stability of the perturbation value. At the same time, the bidder counts the total number of words in the bid file, performs prime factor decomposition on the word count (i.e. decomposes the word count into the product of several prime numbers), and takes the number of prime factors or the maximum prime factor value obtained by decomposition as the length parameter of the random perturbation value, so that the length of the perturbation value is nonlinearly related to the size of the bid file. The random perturbation value generated in this way is closely bound to the intrinsic characteristics (distribution entropy) of the bid data, and its uniqueness is enhanced through the special setting of the length parameter, effectively avoiding the risk of predicting or forging the perturbation value.

[0123] After collecting all the random disturbance values injected by the bidders, the tenderer starts the aggregation process of the key pool. First, the tenderer retrieves the historical bidding records of each bidder and calculates the historical bidding deviation rate. The deviation rate is the statistical average of the deviation (such as bid price deviation, construction period deviation, etc.) between the historical bidding data and the actual execution result, reflecting the credibility of the bidder's past bidding information. Based on the deviation rate, the tenderer generates the corresponding weight parameter through the sigmoid function. The sigmoid function maps the deviation rate to the interval [0, 1]. When the deviation rate is low (high credibility), the weight parameter tends to 1, making the disturbance value of this bidder occupy a higher proportion in the aggregation. When the deviation rate is high (low credibility), the weight parameter tends to 0, reducing the influence of its disturbance value. After obtaining the weight parameter, the tenderer performs weighted multiplication operation on the elements of the initial key matrix and the random disturbance value of the corresponding bidder: each matrix element is multiplied by the product of the weight parameter and the disturbance value of the bidder corresponding to the element, obtaining the final key pool element value. Through this weighted aggregation method, the disturbance information of all bidders is integrated, and the weight adjustment based on historical credit realizes differentiated control, making the generated dynamic key pool not only have the distributed characteristics of multi-party participation, but also reflect the quantitative consideration of the credit of the bidder, enhancing the security and rationality of the key pool.

[0124] In an embodiment, the tenderer generates an initial key matrix, and each bidder inputs a random disturbance value into the initial key matrix to form a dynamically changing key pool, comprising:

[0125] The tenderer analyzes the technical complexity level of the bidding project and sets the size of the initial key matrix according to the complexity from low to high; the matrix elements are generated based on encryption parameters of lattice cryptography, and the encryption strength is adjusted according to the project budget amount;

[0126] When the bidder generates a random disturbance value, it first identifies the most frequently modified sensitive passage in its own bidding document, and based on the modification frequency, adjusts the randomness of the chaotic algorithm used to generate the random disturbance value; the length of the random disturbance value is determined according to the total number of pages of the bidding document;

[0127] During the process of injecting the random disturbance value into the matrix, the number of current bidders is monitored in real time, and based on the number of bidders, the encryption threshold used in the operation of the disturbance value and the matrix element and the length of the hash value of the operation result are determined;

[0128] When the tenderer aggregates to generate the key pool, the historical bidding compliance records of each bidder are combined to determine the weight proportion of the random disturbance value of each bidder in the final result of the key pool; the size of the sliding window used in the aggregation operation is adjusted according to the average length of time the bidders respond to the tender;

[0129] The key pool is generated, and its update interval gradually shortens as the bidding deadline approaches, and the anti-interference parameters of the encryption algorithm are automatically enhanced each time the update is performed.

[0130] In this embodiment, the tenderer first comprehensively evaluates the technical complexity of the bidding project, divides the technical complexity into several levels (such as low, medium, high, and extremely high) according to factors such as the breadth of the technical field involved, the difficulty of technical implementation, and the innovation requirements. The size (i.e., the number of rows and columns) of the initial key matrix is directly related to the technical complexity level: the lower the technical complexity level, the smaller the matrix size; as the complexity level increases, the matrix size increases accordingly. This setting ensures that the computational load of the key matrix matches the technical difficulty of the project, avoiding resource waste or insufficient security redundancy. When generating matrix elements, the tenderer uses encryption algorithms based on lattice cryptography to generate element values with high security by setting core encryption parameters (such as lattice dimension and modulus length). At the same time, the encryption strength is adjusted in segments according to the project budget amount: when the budget amount is in a lower range, basic encryption parameters are used; for every preset amount of budget increase, the security level of the encryption parameters is increased accordingly (such as increasing the lattice dimension or using a larger modulus length), so that the security strength of the key matrix matches the economic value of the project, forming a hierarchical security protection system.

[0131] Before generating the random disturbance value, the tenderer performs a backtracking analysis on the modification records of the bidding document, and identifies the most frequently modified sensitive paragraphs. Sensitive paragraphs usually contain key information such as quotes, core technical parameters, and service commitments, and high modification frequency indicates strong data dynamics and sensitivity. Based on this modification frequency, the tenderer dynamically adjusts the randomness of the chaotic algorithm: the higher the modification frequency, the more complex the randomness parameters (such as initial iteration value and disturbance coefficient) of the algorithm are set, making the generated random disturbance value more unpredictable; when the modification frequency is low, the complexity of the randomness parameters is appropriately reduced, ensuring security while improving computational efficiency. In addition, the tenderer counts the total number of pages of the bidding document, and uses the total number of pages as a basis for determining the length of the random disturbance value: the more pages, the longer the bit length of the disturbance value, to ensure that the disturbance value can cover more dimensions of information features in the document. In this way, the characteristics of the random disturbance value are associated with the dynamic modification characteristics and size of the bidding document, enhancing the binding of the disturbance value and the bidding data and reducing the risk of malicious imitation.

[0132] In the process of injecting random disturbance values into the initial key matrix by each bidder, the number of valid bidders currently participating in bidding is monitored in real time. The encryption threshold is positively related to the number of bidders: the more the number of bidders, the higher the encryption threshold (i.e. the minimum number of verification nodes required to trigger the encryption operation) is set, and more nodes are required to synchronize the verification operation process to resist the risk of coordinated attacks that may be brought about by the increase in the number of participants; when the number of bidders is small, the encryption threshold is correspondingly reduced, balancing security and operation efficiency. At the same time, the length of the hash value of the operation result is negatively related to the number of bidders: the more the number of bidders, the shorter the bit length of the hash value reserved, and by shortening the data length, the redundant load of multi-node transmission is reduced; when the number of bidders is small, a longer hash value is reserved to improve the accuracy of data integrity checking. This dynamic adjustment mechanism based on the number of participants enables the disturbance value injection process to adapt to different sizes of bidding scenarios, maintaining a balance between security and efficiency in complex network environments.

[0133] When the tenderer aggregates the random disturbance values of all bidders to generate the key pool, the tenderer first retrieves the historical bidding compliance records of each bidder, including whether there is false qualification, inaccurate bidding data, violation of bidding rules, etc. According to the degree of compliance record, the weight proportion of the random disturbance value of each bidder is allocated: the better the compliance record (such as no violation record, high accuracy of past bidding data), the higher the weight proportion, and the greater the influence of its disturbance value on the final result of the key pool; the weight proportion of the bidder with adverse compliance record is correspondingly reduced to weaken its potential negative impact on the security of the key pool. In the aggregation operation process, the size of the sliding window is dynamically adjusted according to the average length of time for the bidders to respond to the tender: the shorter the average response time (i.e. the more timely the feedback of the bidders to the tendering process), the smaller the window size is set, so that the operation can capture the real-time changes of the disturbance value faster; when the average response time is longer, the window size is appropriately increased to improve the stability of the aggregation result by expanding the data sampling range. This aggregation strategy combining historical compliance and real-time response efficiency makes the key pool reflect differentiated management of the credibility of bidders and adapt to data characteristics under different response speeds, enhancing the reliability and dynamic adaptability of the key pool.

[0134] The key pool is not static after generation, but is dynamically updated as the bidding process advances. The setting of the update interval is related to the remaining time of the bidding deadline: in the early stage of the bidding period, the deadline is far away, and the update interval is long (such as updating once a day), reducing unnecessary computational consumption; as the deadline approaches, the update interval gradually shortens, and in the critical period before the bidding deadline (such as the last 24 hours), it is shortened to updating once an hour or even once every half hour, in order to quickly respond to real-time changes in bidding data and resist high-frequency attacks that may occur near the deadline. At each update, the anti-interference parameters of the automatic enhancement encryption algorithm are increased (such as increasing the noise factor and increasing the number of key iterations), so that the anti-cracking ability of the key pool gradually increases over time. This dynamic evolution mechanism over time ensures that the key pool always maintains a security strength matching the risk level throughout the bidding period, providing stronger security protection for high-sensitivity operations in the bidding deadline phase.

[0135] With reference to Figure 2 In an embodiment of the present application, a multi-party secure computing-based bidding secret control device is also provided, comprising a tendering party, a plurality of bidding parties, and a security control node.

[0136] The tendering party generates an initial key matrix, and each bidding party inputs a random disturbance value into the initial key matrix to form a dynamically changing key pool.

[0137] The bidding party converts the bidding data into a feature vector associated with an element in the key pool and embeds it into a transmission stream for transmission to the security control node.

[0138] After receiving the transmission stream, the security control node generates a matching mask based on the tendering rules without decrypting the feature vector; and obtains a matching result by performing a modulo operation on the feature vector and the matching mask, and transmits the matching result to the tendering party.

[0139] After obtaining the matching result, the tendering party generates a winning feature code and sends it to the security control node.

[0140] The security control node only partially decrypts the feature vector matching the winning feature code, generates result information containing the winning identifier but not leaking the bidding data, and pushes it to the corresponding bidding party and tendering party.

[0141] In the present embodiment, the specific implementation of each component in the above-mentioned device embodiment is described in the above-mentioned method embodiment, and will not be described here.

[0142] With reference to Figure 3 In an embodiment of the present application, a computer device is also provided, which can be a server, and the internal structure thereof can be as follows: Figure 3The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store corresponding data in the embodiment. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement the above method.

[0143] Those skilled in the art can understand that, Figure 3 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied.

[0144] The computer readable storage medium in the embodiment of the present application can be a volatile readable storage medium or a non-volatile readable storage medium.

[0145] In summary, the bidding secret control method and device based on multi-party secure computation provided in the embodiment of the present application include: a tendering party generates an initial key matrix, each bidding party inputs a random disturbance value to the initial key matrix to form a dynamically changing key pool; the bidding party converts bidding data into a feature vector associated with an element in the key pool, and embeds the feature vector into a transmission stream and transmits the transmission stream to a secure control node; the secure control node receives the transmission stream, generates a matching mask based on the tendering rules without decrypting the feature vector, obtains a matching result by performing a modulo operation on the feature vector and the matching mask, and transmits the matching result to the tendering party; the tendering party obtains the matching result, generates a winning feature code, and sends the winning feature code to the secure control node; and the secure control node only partially decrypts the feature vector matching the winning feature code, generates result information containing a winning identifier without leaking bidding data, and pushes the result information to the corresponding bidding party and tendering party. In the present application, the bidding data is converted into a feature vector associated with an element in the key pool and embedded into a transmission stream for transmission, without using key encryption to transmit the bidding data, thereby overcoming the defect that the current bidding secret control method is prone to leaking bidding data.

[0146] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, storage, database or other medium provided by the present application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0147] It should be noted that in this document, the terms "comprising", "including", or any other variant thereof are intended to cover a non-exclusive inclusion, such that a process, device, article or method that comprises a list of elements does not only include those elements, but can also include other elements not expressly listed or inherent to such process, device, article or method. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, device, article or method that includes the element.

[0148] The above description is only the preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, based on the content of the present application specification and drawings, are also included in the patent protection scope of the present application.

Claims

1. A method for controlling the confidentiality of bids based on multi-party secure computation, characterized in that, Includes the following steps: The tendering party generates an initial key matrix, and each bidder inputs a random perturbation value into the initial key matrix to form a dynamically changing key pool; The bidder converts the bidding data into a feature vector associated with the elements in the key pool and embeds it into the transport stream for transmission to the security management node. After receiving the transmission stream, the security control node generates a matching mask based on the bidding rules without decrypting the feature vector; the matching result is obtained by modulo operation between the feature vector and the matching mask and transmitted to the bidding party. After obtaining the matching results, the bidding party generates a winning bid feature code and sends it to the security control node. The security control node only partially decrypts the feature vector that matches the winning bid feature code, generates result information containing the winning bid identifier but without disclosing the bidding data, and pushes it to the corresponding bidder and tenderer.

2. The bidding confidentiality control method based on multi-party secure computation according to claim 1, characterized in that, The key elements in the key pool are only temporarily activated based on real-time network characteristics at the moment the bidding data is submitted, and automatically expire after a single activation.

3. The bidding confidentiality control method based on multi-party secure computation according to claim 1, characterized in that, The bidder transforms the bidding data into a feature vector associated with elements in the key pool, including: The bidder splits the bid data according to preset dimensions to obtain multiple data sub-items; Each data item is assigned a unique identifier, which is mapped to an element selected from the key pool. The mapping rule is determined by the bidder based on a hash function generated by its own private key. Based on the mapping relationship, the data sub-item is XORed with the corresponding key pool element to generate a feature component; All feature components are combined according to the order in which each data item is split to form a feature vector associated with the elements in the key pool.

4. The bidding confidentiality control method based on multi-party secure computation according to claim 1, characterized in that, The matching result obtained by modulo operation between the feature vector and the matching mask is transmitted to the bidding party, including: The security control node decomposes the feature vector into sub-vectors that correspond one-to-one with each sub-item of the bidding rules, and each sub-vector corresponds to an evaluation dimension in the bidding data. To match the mask, a weight matrix is ​​set that is consistent with the dimensions of the subvector. The weight values ​​are dynamically generated based on the importance of each evaluation dimension. Perform a modulo operation on each subvector and the corresponding dimension submask in the matching mask to obtain the dimension matching value; The matching values ​​of each dimension are weighted and summed with the weight matrix to generate a comprehensive matching degree; If the overall matching degree is greater than the preset threshold, a matching result containing a successful matching indicator and key dimension compliance information will be generated.

5. The bidding confidentiality control method based on multi-party secure computation according to claim 4, characterized in that, The key dimension compliance information is obtained through modulo operation remainder range mapping and does not include the original bidding data.

6. The bidding confidentiality control method based on multi-party secure computation according to claim 4, characterized in that, After obtaining the matching results, the bidding party generates a winning bid feature code, including: The bidding party extracts the feature vectors corresponding to all bidders whose comprehensive matching degree is greater than a preset threshold from the matching results to form a candidate set. A ranking algorithm based on the priority of bidding rules is used to sort the candidate set by weight to generate a ranking sequence; wherein, the ranking parameters of the ranking algorithm are embedded in the ranking sequence after being encrypted with the private key of the bidding party. Select the feature vector corresponding to the first bidder in the sorted sequence, extract its core mapping relationship with the key pool elements, and generate the basic feature code; The basic feature code is salted, and the salted feature code is encrypted using the public key of the security control node to generate the winning feature code.

7. The bidding confidentiality control method based on multi-party secure computation according to claim 6, characterized in that, The digital signature of the bidding party is embedded in the public key encryption process, and a signature verification request is attached when the data is sent to the security control node.

8. The bidding confidentiality control method based on multi-party secure computation according to claim 6, characterized in that, The security control node only partially decrypts the feature vector matching the winning bid signature, generating result information containing the winning bid identifier but without disclosing the bidding data, and pushes it to the corresponding bidder and tendering party, including: The security control node extracts the core mapping relationship from the winning bid feature code through decryption; The core mapping relationship is compared with the key pool to locate the corresponding feature vector, and the components of the bidding core data involved in the feature vector are masked by a preset field masking rule. Perform targeted decryption on the feature vector after masking the components, decrypting only the identifier fields related to the determination of the winning bid status; The result information includes the winning bid identifier, the bidder code, and an encrypted compliance statement; wherein the compliance statement is encrypted using the bidder's public key and can only be decrypted and viewed by the corresponding bidder. The system pushes result information containing the complete winning bid identifier to the bidding party, and pushes encrypted information containing only its own winning bid status to the corresponding bidders.

9. A bid confidentiality control device based on multi-party secure computation, characterized in that, include: The tendering party, multiple bidders, and security control nodes; The tendering party generates an initial key matrix, and each bidder inputs a random perturbation value into the initial key matrix to form a dynamically changing key pool; The bidder converts the bidding data into a feature vector associated with the elements in the key pool and embeds it into the transport stream for transmission to the security management node. After receiving the transmission stream, the security control node generates a matching mask based on the bidding rules without decrypting the feature vector; the matching result is obtained by modulo operation between the feature vector and the matching mask and transmitted to the bidding party. After obtaining the matching results, the bidding party generates a winning bid feature code and sends it to the security control node. The security control node only partially decrypts the feature vector that matches the winning bid feature code, generates result information containing the winning bid identifier but without disclosing the bidding data, and pushes it to the corresponding bidder and tenderer.

Images