Privacy set intersection processing method in dynamic data scene
By combining VOLE and Silent OT technologies with key initialization and feature intersection calculation, the problems of low efficiency and security in privacy set intersection in dynamic data scenarios are solved. This achieves efficient privacy data intersection processing, reduces communication and computation overhead, and enhances the security and privacy protection of data transmission.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NAVAL UNIV OF ENG PLA
- Filing Date
- 2026-01-30
- Publication Date
- 2026-05-12
AI Technical Summary
Existing privacy set intersection techniques are inefficient and pose security risks in dynamic data scenarios, making it difficult to achieve efficient privacy protection and data matching in multi-party dynamic data scenarios.
By employing the Vector Unintentional Linear Evaluation (VOLE) protocol and Silent Unintentional Transmission (Silent OT) technology, and combining key initialization, query construction, privacy transmission, and result decryption steps, efficient privacy data intersection processing is achieved through encrypted matching and feature intersection calculation, reducing communication and computational overhead.
Without disclosing the original data, it improves query efficiency in dynamic data scenarios, reduces communication and computing costs, supports deduplication of large-scale datasets, and enhances data transmission security and privacy protection.
Smart Images
Figure CN122027129A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of privacy data sharing and processing technology, and in particular relates to a method for processing the intersection of privacy sets in dynamic data scenarios. Background Technology
[0002] Privacy-preserving data intersection processing has been widely applied in various fields such as academic integrity maintenance, data notarization, anti-plagiarism detection, and enterprise data deduplication, and has significant practical implications. With the advent of the big data era, the volume of data is exploding, and the need for privacy protection is becoming increasingly prominent. Therefore, exploring ways to achieve high efficiency and accuracy in intersection processing while ensuring data security is of paramount importance, both theoretically and practically. Currently, the relevant technologies or patents mainly include: "A method and system for finding intersections of FATE federated privacy sets based on vector inadvertent evaluation". This scheme generates vectors and scalars in the federated learning framework FATE through the vector inadvertent evaluation (VOLE) protocol, and constructs an inadvertent pseudo-random function (OPRF) protocol to achieve privacy set intersection (PSI). It is mainly used to solve the security and efficiency problems in malicious scenarios. "A privacy request method based on the intersection of unbalanced privacy sets" is mainly for unbalanced datasets. It uses hash value to specifically locate and filter the target dataset, combines cuckoo hash and ordinary hash bucketing, and transmits pseudo-random function values in stages to achieve privacy queries. The above schemes are mainly for centralized processing of static datasets and are difficult to apply well in dynamic environments with multiple dynamic data scenarios. Summary of the Invention
[0003] The purpose of this invention is to provide a method for efficient linear computation of batch data based on the Vector Unintentional Linear Evaluation (VOLE) protocol. By combining communication compression and proactive security verification through Silent Unintentional Transmission (Silent OT), this invention addresses the efficiency bottlenecks and security risks of traditional Privacy Set Intersection (PSI) technology in dynamic data scenarios. It enables matching through unintentional transmission and privacy intersection, protecting data privacy; it also supports deduplication of large-scale datasets, reduces communication and computational overhead, and improves query efficiency.
[0004] To achieve the above objectives, the present invention adopts the following technical solution.
[0005] A method for processing intersection of privacy sets in dynamic data scenarios includes: key initialization, query construction, privacy transmission, and result decryption;
[0006] Key initialization: This refers to the process where the requester and the server each generate the public and private keys needed for unintentional transmission, and then use the public key to encrypt all query data; the requester generates the key pair. , The requester's public key, The requester's private key; the server generates a key pair. ,in For the server's public key, The server's private key is used; the requester and the server encrypt their query and response data respectively, and establish a basic communication channel.
[0007] Query construction: refers to the process by which the requester calculates the query hash based on local hash characteristics. Construct a query request and encrypt it to obtain The requesting party then sends the query set. Give it to the server;
[0008] Privacy transmission refers to the process where the requester and the server implement a linear combination of privacy measures, and then the server performs encrypted matching with the requester to determine whether the query data matches its local data. The server calculates the intersection of the matching features and outputs the intersection of the matching features.
[0009] Decryption of the result: refers to the requester receiving the encrypted data returned by the server. The shared information and private key generated by VOLE are used for decryption, and finally the matching hash feature set is restored.
[0010] A further improvement or preferred implementation of the aforementioned method for processing intersection of privacy sets in dynamic data scenarios, wherein the privacy transmission specifically includes encrypted matching and feature intersection calculation;
[0011] Encryption matching refers to the server and the requester encrypting the hash feature separately. The requester uses its own public key to encrypt the hash feature, while the server uses its public key to encrypt the corresponding feature.
[0012] Feature intersection calculation: This refers to the server processing the encrypted hash features submitted by the requester and calculating the intersection with the corresponding data in its own database. Specifically, this includes: both the requester and the server hashing their features and encrypting them to prevent leakage of the original data during the calculation process; generating random vectors using VOLE technology, and the requester and the server calculating the intersection of matching features through linear combinations of these random vectors; the calculated intersection information is returned to the requester only in encrypted form, and the requester uses a private key to decrypt these encrypted results to obtain the final intersection features.
[0013] A further improvement or preferred implementation of the aforementioned method for intersecting privacy sets in dynamic data scenarios, wherein the encrypted matching specifically includes: the requesting party submitting its query data. and encrypted shared random vectors Combined, an encrypted query request is generated. ;in, It is an encrypted query request sent by the requester to the server. It is a random vector generated by the requester. It queries data; after receiving the query request, the server uses local data. Encryption request from the requester The server performs a matching process to determine if the query data matches. The server then encrypts and compresses the matching results before sending them back to the requester. The requester uses their private key to decrypt the results and obtain the final outcome.
[0014] A further improvement or preferred implementation of the aforementioned method for intersecting privacy sets in dynamic data scenarios includes the following method for generating the shared random vector: randomly generating a shared matrix. Construct a matrix relating the possible outputs of the requester. The matrix associated with the possible output of the server ,and ;in , ;Requester extracts matrix Several lines in Composition and Matrix Related submatrix And solve the random row vectors using a linear equation solver. row vector satisfy The requester and server perform a VOLE operation, distributing a vector to the requester. , Distribute vectors to the server and scalar ,in ; Request direction to send to server The server uses an unintentional pseudo-random function Compute key ; This refers to a random oracle; the requester solves the pseudo-random function by performing a linear operation of inner product.
[0015] A further improvement or preferred embodiment of the aforementioned method for processing the intersection of privacy sets in dynamic data scenarios includes a step for performing privacy data segmentation, specifically:
[0016] For the data interaction set consisting of the requester and the server Define the data interaction terminal. Corresponding privacy dataset ; Refers to privacy datasets The first in If there are 1 set of privacy data, then the intersection of the privacy data is represented as: ;
[0017] During key initialization, for each data sender and data receiving end Combined distribution of random key seeds ; Using pseudo-random functions Determine the index value of the data interaction terminal Corresponding privacy data share , represented as Obtain a random key , This refers to the party sending the request. and the receiving party The first key network constitutes A key fragment;
[0018] During communication, the sending end generates control coefficients. And a prime number much greater than 1 And send it to the receiving end; the sending end calculates... And send it to the receiving end, where It is a random number and ; Receiver calculation And send it to the sender, where It is a random number and ; Sending end calculation Obtain the session key Receiver calculation Obtain the session key An encrypted channel is established using this session key;
[0019] For each privacy dataset Privacy data fragments The sending end calculates the pseudo-random value corresponding to the privacy data fragment using the session key, and then concatenates them to obtain the complete data. Specifically, for the sending end... Computing and receiving end Shared the first A key fragment The corresponding privacy data share The complete data is obtained by splicing together all the privacy data segments.
[0020] A further improvement or preferred implementation of the aforementioned method for processing intersection of privacy sets in dynamic data scenarios includes a step for decrypting the results to achieve distributed, high-speed multi-party intersection processing. Specifically:
[0021] For any two requesters a and b, define their private datasets as follows: and ;
[0022] During the communication process, the server generates control coefficients. And a prime number much greater than 1 And send it to all requesters;
[0023] The requester A secretly selects a random number. and ,calculate Then use the private key right Digital signature obtained after encryption , i=1, 2, and send it to the requester b;
[0024] Requester B obtains digital signature Then, the public key sent by requester A The digital signature is verified; if verification fails, the transmission is terminated; if verification succeeds, the requester (b) secretly selects a random number. and , Then use the private key right Digital signature obtained after encryption i=1, 2, and digital signature. , , Send to requester A;
[0025] Requester A obtains digital signature Then, the public key sent by requester b The digital signature is verified. If the verification fails, the transmission is terminated; if the verification succeeds, a connection is established.
[0026] After the connection is successfully established, requester A calculates... , Then use Generate random number sets ,use Its private dataset Obtained by performing pseudo-random function calculation and will Send to the server; at the same time, requester b uses Generate random number sets Random number set After shuffling the order, we get , Then use Its private dataset Obtained by performing pseudo-random function calculation ,Will Selecting elements as key-value sets The first M elements are used to obtain the dataset. ,Will After unintentionally encoding the key-value pairs, the data is sent to the server.
[0027] The server obtains the data sent by requester A and requester B, in order to For unintentional key-value storage encoded objects, sent by requester a1 In Unintentionally decode the key-value store to obtain Then the decoding result Return to requester A;
[0028] Requester A obtains the decoding result Then, calculate Calculation results Simultaneously, they are private datasets respectively and The intersection of. Attached Figure Description
[0029] Figure 1 Technical architecture diagram for privacy set intersection processing method in dynamic data scenarios;
[0030] Figure 2 This is a flowchart of the matching calculation process in the embodiment;
[0031] Figure 3 This is a flowchart of the feature intersection calculation in the embodiment;
[0032] Figure 4 This is a diagram illustrating the system transmission resource consumption under different schemes. Detailed Implementation
[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.
[0034] like Figure 1 As shown, this application provides a method for processing privacy set intersection in dynamic data scenarios, which is mainly used to effectively transmit necessary data without disclosing the original data, enabling the requester to obtain the required matching features, while reducing the communication cost required for traditional OT interaction. It includes at least four basic steps: key initialization, query construction, privacy transmission, and result decryption.
[0035] Key initialization refers to the process where the requesting party and the server each generate the public and private keys needed for unintentional transmission, and use the public key to encrypt all query data, including:
[0036] The requester generates a key pair. ,in For public key, The private key; the server generates a key pair. ,in For public key, For private key;
[0037] Based on this, the requester and the server encrypt their query data and response data respectively, and establish a basic communication channel; at the same time, both parties use a shared random vector to support the privacy transmission of subsequent data;
[0038] Furthermore, to enhance data security, the generation method of the shared random vector is optimized. The generation method of the shared random vector is as follows:
[0039] Randomly generate shared matrix Construct a matrix relating the possible outputs of the requester. The matrix associated with the possible output of the server ,and ;in , ;Requester extracts matrix Several lines in Composition and Matrix Related submatrix And solve the random row vectors using a linear equation solver. row vector satisfy The requester and server perform a VOLE operation, distributing the vector v to the requester. Distribute vectors to the server and scalar ,in In subsequent processing, the requesting party sends a message to the server. The server uses an unintentional pseudo-random function Compute key ; This refers to a random oracle; the requester solves the pseudo-random function by performing a linear operation of inner product.
[0040] Query construction refers to the process by which the requester calculates the query hash based on local hash feature values. Construct a query request and encrypt it. The requesting party then sends the query set. Give it to the server;
[0041] Privacy transmission utilizes VOLE technology. After the requester and server achieve a linear combination of privacy, the server and requester perform matching calculations to determine whether the queried data matches their local data and use compressed VOLE to reduce communication complexity.
[0042] In multi-terminal matching calculations within dynamic data scenarios, due to the limitations of existing solutions, the requesting party and the server need to perform a large amount of data exchange during the matching process. This significantly increases data storage and transmission costs, while reducing the efficiency of matching data for privacy data transmission. To solve this problem, reduce communication costs, and improve transmission efficiency, such as... Figure 2 As shown, this invention proposes an improved scheme that encrypts the data exchanged between the requester and the server during the matching calculation process and reduces the exchange of redundant information, specifically including:
[0043] The requester will query the data. Combined with the encrypted shared random vector v, an encrypted query request is generated:
[0044] in, It is an encrypted query request sent by the requester to the server. It's for querying data.
[0045] After receiving a query request, the server performs matching calculations with the requester using Silent OT technology. The server uses local data. Encryption request from the requester Perform a match to determine if the query data matches:
[0046]
[0047] in, To match the indicator, indicating that data is being queried. Is it related to local server data? Match. If a match is found, return 1; otherwise, return 0.
[0048] By leveraging OT Extensions, the amount of data exchanged can be reduced to a constant level, and the communication overhead for each data interaction can be reduced to a constant level. The server encrypts the matching results and further reduces the data volume through VOLE compression technology, thereby reducing communication overhead. The server sends the final encrypted matching information back to the requester, who uses their private key to decrypt and obtain the final result.
[0049] Decryption refers to the process by which the requesting party receives the ciphertext data returned by the server. Decryption is performed using the shared information and private key generated by VOLE. Finally, the matching hash feature set is restored. This step is responsible for ensuring that the requester can calculate the intersection of matching features with the server without exposing the original data. The VOLE-PSI intersection calculation module takes the matching hash features as input and the intersection of matching features as output. It includes three steps: encrypted matching, feature intersection calculation, and privacy-preserving intersection calculation, which are explained in detail below.
[0050] Encryption matching refers to the process where both the server and the requester encrypt the hash feature. The requester uses its own public key to encrypt the hash feature, while the server uses its public key to encrypt the corresponding feature. This encryption ensures data privacy during transmission.
[0051] Feature intersection calculation refers to the process by which the server processes the cryptographic hash features submitted by the requester and calculates the intersection with the corresponding data in its own database.
[0052] In traditional feature intersection calculations, the efficiency of the computation process decreases significantly with increasing data volume. To address this issue, such as... Figure 3 As shown, this invention designs and proposes an improved feature intersection calculation scheme. This scheme effectively resolves the contradiction between privacy protection and computational efficiency by inadvertently exchanging randomly generated vectors and linearly combining these vectors. Specifically:
[0053] Both the requester and the server hash their respective features and then encrypt them to prevent the original data from being leaked during the computation process. The requester's encrypted feature is as follows:
[0054]
[0055] Characteristics of the encrypted server:
[0056]
[0057] Using VOLE technology, the requester and server exchange generated random vectors, performing an unintentional linear combination. This exchange process allows both parties to perform effective intersection calculations without directly comparing the original features; the requester generates random vectors. Generate random vectors with the server :
[0058]
[0059]
[0060] in, and It is a random vector generated by the requester and the server. and That is the corresponding private key.
[0061] Through linear combinations of these random vectors (e.g.) The requester and the server can compute the intersection of matching features under encrypted and randomized conditions.
[0062] The calculated intersection information is returned to the requester only in encrypted form to avoid revealing the original data. The requester can use their private key to decrypt these encrypted results and obtain the final intersection characteristics.
[0063] Privacy-preserving intersection is a method where, after calculating the intersection, the server only returns an encrypted value of the matching result to the requesting party to prevent data leakage. The requesting party uses its private key to decrypt the returned encrypted data to obtain the characteristics of the match.
[0064] In particular, in dynamic data scenarios, the total amount of privacy data varies greatly. Although current information data transmission systems already possess powerful data transmission capabilities, excessively large data volumes can lead to various uncertainties during the encryption and decryption of privacy data. To improve the flexible processing capabilities in dynamic scenarios, based on the aforementioned, this invention further eliminates the optimization steps used to implement privacy data segmentation processing. This step is mainly used for effective privacy transmission and data restoration in scenarios with large data volumes and data segmentation. Specifically:
[0065] For the data interaction set consisting of the requester and the server Define the data interaction terminal. Corresponding privacy dataset ; Refers to privacy datasets The first in If there are 1 set of privacy data, then the intersection of the privacy data is represented as: ;
[0066] During key initialization, for each data sender and data receiving end Combined distribution of random key seeds ; Using pseudo-random functions Determine the index value of the data interaction terminal Corresponding privacy data share , represented as Obtain a random key , This refers to the party sending the request. and the receiving party The first key network constitutes A key fragment;
[0067] During communication, the sending end generates control coefficients. And a prime number much greater than 1 And send it to the receiving end; the sending end calculates... And send it to the receiving end, where It is a random number and ; Receiver calculation And send it to the sender, where It is a random number and ; Sending end calculation Obtain the session key Receiver calculation Obtain the session key An encrypted channel is established using this session key;
[0068] For each privacy dataset Privacy data fragments The sending end calculates the pseudo-random value corresponding to the privacy data fragment using the session key, and then concatenates them to obtain the complete data. Specifically, for the sending end... Computing and receiving end Shared the first A key fragment The corresponding privacy data share The complete data is obtained by splicing together all the privacy data segments.
[0069] Furthermore, in some dynamic data scenarios, such as edge computing, most requesters primarily act as data sharing units, with only one or a few high-computing-power servers. Therefore, based on the aforementioned technical solutions, a further step is provided to achieve distributed, high-speed multi-party intersection processing. This step mainly aims to further improve intersection efficiency and enhance privacy protection between the two requesters in the aforementioned or similar scenarios. Specifically:
[0070] For any two requesters a and b, define their private datasets as follows: and ;
[0071] During the communication process, the server generates control coefficients. And a prime number much greater than 1 And send it to all requesters;
[0072] The requester A secretly selects a random number. and ,calculate Then use the private key right Digital signature obtained after encryption , i=1, 2, and send it to the requester b;
[0073] Requester B obtains digital signature Then, the public key sent by requester A The digital signature is verified; if verification fails, the transmission is terminated; if verification succeeds, the requester (b) secretly selects a random number. and , Then use the private key right Digital signature obtained after encryption i=1, 2, and digital signature. , , Send to requester A;
[0074] Requester A obtains digital signature Then, the public key sent by requester b The digital signature is verified. If the verification fails, the transmission is terminated; if the verification succeeds, a connection is established.
[0075] After the connection is successfully established, requester A calculates... , Then use Generate random number sets ,use Its private dataset Obtained by performing pseudo-random function calculation and will Send to the server;
[0076] At the same time, requester B uses Generate random number sets Random number set After shuffling the order, we get , Then use Its private dataset Obtained by performing pseudo-random function calculation ,Will Selecting elements as key-value sets The first M elements are used to obtain the dataset. ,Will After unintentionally encoding the key-value pairs, the data is sent to the server.
[0077] The server obtains the data sent by requester A and requester B, in order to For unintentional key-value storage encoded objects, sent by requester a1 In Unintentionally decode the key-value store to obtain Then the decoding result Return to requester A;
[0078] Requester A obtains the decoding result Then, calculate Calculation results Simultaneously, they are private datasets respectively and The intersection of.
[0079] For the aforementioned schemes, anonymization schemes a, c, and c based on UPSI-CA technology are selected for illustration. To simplify the analysis, while keeping the privacy data capacity of one party constant (level 2^20), the privacy data capacity of the other party is adjusted for intersection processing. The data system transmission resource consumption is shown in Table 1. Figure 4 As shown.
[0080] Table 1. Data system transmission resource consumption under different schemes with changes in unilateral privacy data capacity.
[0081]
[0082] From Table 1 and Figure 4 It is evident that as the total amount of privacy data increases, the system transmission resource consumption of each scheme increases synchronously. Scheme a has a relatively low system resource consumption growth rate under lower privacy data capacity, but as the privacy data capacity increases, its transmission resource consumption rate far exceeds that of other schemes. Scheme c performs well under high privacy data capacity. Considering the application scenario of this application, which faces the uncertainty of the total amount of privacy data in dynamic data scenarios, the latter being in a real-time changing situation, the scheme of this application can achieve a certain level of system resource balance capability in both low and high privacy data capacity scenarios. The above results take into account the change in privacy data capacity of one party. Under the dynamic change of privacy data of two or even multiple parties, this application can further leverage its flexible adjustment capability to reduce the overall system resource consumption. Through testing and analysis, in general scenarios, such as the real-time privacy data transmission process of operating ship and other mechanical systems or equipment, its comprehensive system resource consumption rate is reduced by about 15-30% compared with existing schemes, showing significant improvement. At the same time, the improved scheme achieves independence from platforms such as large computing service centers, providing a technical foundation for the application of the scheme on independent platforms or equipment with limited computing resources, such as ships.
[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.
Claims
1. A method for processing intersection of privacy sets in dynamic data scenarios, characterized in that, It includes four basic steps: key initialization, query construction, privacy transmission, and result decryption; Key initialization: This refers to the requester and the server generating public and private keys that are needed for unintentional transmission, and using the public key to encrypt all query data; The requester generates a key pair. , The requester's public key, The requester's private key; the server generates a key pair. ,in For the server's public key, This is the server's private key; The requester and the server encrypt their query and response data respectively, and establish a basic communication channel; Query construction: refers to the process by which the requester calculates the query hash based on local hash characteristics. Construct a query request and encrypt it to obtain The requesting party then sends the query set. Give it to the server; Privacy transmission refers to the process where the requester and the server implement a linear combination of privacy measures, and then the server performs encrypted matching with the requester to determine whether the query data matches its local data. The server calculates the intersection of the matching features and outputs the intersection of the matching features. Decryption of the result: refers to the requester receiving the encrypted data returned by the server. The shared information and private key generated by VOLE are used for decryption, and finally the matching hash feature set is restored.
2. The method for intersecting privacy sets in dynamic data scenarios according to claim 1, characterized in that, The privacy transmission specifically includes encrypted matching and feature intersection calculation; Encryption matching refers to the server and the requester encrypting the hash feature separately. The requester uses its own public key to encrypt the hash feature, while the server uses its public key to encrypt the corresponding feature. Feature intersection calculation: This refers to the server processing the encrypted hash features submitted by the requester and calculating the intersection with the corresponding data in its own database. Specifically, this includes: both the requester and the server hashing their features and encrypting them to prevent leakage of the original data during the calculation process; generating random vectors using VOLE technology, and the requester and the server calculating the intersection of matching features through linear combinations of these random vectors; the calculated intersection information is returned to the requester only in encrypted form, and the requester uses a private key to decrypt these encrypted results to obtain the final intersection features.
3. The method for intersecting privacy sets in dynamic data scenarios according to claim 2, characterized in that, The encrypted matching specifically includes: the requester submitting its query data. and encrypted shared random vectors Combined, an encrypted query request is generated. ;in, It is an encrypted query request sent by the requester to the server. It is a random vector generated by the requester. It queries data; after receiving the query request, the server uses local data. Encryption request from the requester The server performs a matching process to determine if the query data matches. The server then encrypts and compresses the matching results before sending them back to the requester. The requester uses their private key to decrypt the results and obtain the final outcome.
4. The method for intersecting privacy sets in dynamic data scenarios according to claim 1, characterized in that, The shared random vector is generated as follows: a shared matrix is generated randomly. Construct a matrix relating the possible outputs of the requester. The matrix associated with the possible output of the server ,and ;in , ;Requester extracts matrix Several lines in Composition and Matrix Related submatrix And solve the random row vectors using a linear equation solver. row vector satisfy ; The requester and server perform a VOLE operation, distributing a vector to the requester. , Distribute vectors to the server and scalar ,in ; Request direction to send to server The server uses an unintentional pseudo-random function Compute key ; This refers to a random oracle; the requester solves the pseudo-random function by performing a linear operation of inner product.
5. The method for intersecting privacy sets in dynamic data scenarios according to claim 2, characterized in that, The feature intersection calculation also includes steps for implementing privacy data segmentation processing, specifically: For the data interaction set consisting of the requester and the server ; Define the data interaction terminal Corresponding privacy dataset ; Refers to privacy datasets The first in If there are 1 set of privacy data, then the intersection of the privacy data is represented as: ; During key initialization, for each data sender and data receiving end Combined distribution of random key seeds ; Using pseudo-random functions Determine the index value of the data interaction terminal Corresponding privacy data share , represented as Obtain a random key , This refers to the party sending the request. and the receiving party The first key network constitutes A key fragment; During communication, the sending end generates control coefficients. And a prime number much greater than 1 And send it to the receiving end; the sending end calculates... And send it to the receiving end, where It is a random number and ; Receiver calculation And send it to the sender, where It is a random number and ; Sending end calculation Obtain the session key Receiver calculation Obtain the session key An encrypted channel is established using this session key; For each privacy dataset Privacy data fragments The sending end calculates the pseudo-random value corresponding to the privacy data fragment using the session key, and then concatenates them to obtain the complete data. Specifically, for the sending end... Computing and receiving end Shared A key fragment The corresponding privacy data share The complete data is obtained by splicing together all the privacy data segments.
6. The method for intersecting privacy sets in dynamic data scenarios according to claim 2, characterized in that, The result decryption also includes steps for implementing distributed multi-party high-speed intersection processing, specifically: For any two requesters a and b, define their private datasets as follows: and ; During the communication process, the server generates control coefficients. And a prime number much greater than 1 And send it to all requesters; The requester A secretly selects a random number. and ,calculate Then use the private key right Digital signature obtained after encryption , i=1, 2, and send it to the requester b; Requester B obtains digital signature Then, the public key sent by requester A The digital signature is verified; if verification fails, the transmission is terminated; if verification succeeds, the requester (b) secretly selects a random number. and , Then use the private key right Digital signature obtained after encryption i=1, 2, and digital signature. , , Send to requester A; Requester A obtains digital signature Then, the public key sent by requester b The digital signature is verified. If the verification fails, the transmission is terminated; if the verification succeeds, a connection is established. After the connection is successfully established, requester A calculates... , Then use Generate random number sets ,use Its private dataset Obtained by performing pseudo-random function calculation and will Send to the server; at the same time, requester b uses Generate random number sets Random number set After shuffling the order, we get , Then use Its private dataset Obtained by performing pseudo-random function calculation ,Will Selecting elements as key-value sets The first M elements are used to obtain the dataset. ,Will After unintentionally encoding the key-value pairs, the data is sent to the server. The server obtains the data sent by requester A and requester B, in order to For unintentional key-value storage encoded objects, sent by requester a1 In Unintentionally decode the key-value store to obtain Then the decoding result Returned to requester A; Requester A obtains the decoding result Then, calculate Calculation results Simultaneously, they are private datasets respectively and The intersection of.