CN116266180A - Data searching and inquiring method and system based on searchable encryption and homomorphic encryption - Google Patents

Abstract

The invention provides a data search and query method and system based on searchable encryption and homomorphic encryption. Retrieve the data entry index mapping table containing keywords corresponding to the search voucher from the entry index mapping table; decrypt the data entry index mapping table to obtain the data entry index, and download the data item to be obtained based on the data entry index; where, the data The entry index mapping table is constructed by segmenting the data in the form of text through the word segmentation technology, and extracting the keywords from the word segmentation. The present invention converts the keywords to be obtained into a data entry index mapping table through the search voucher, and downloads the required data items based on the data entry index mapping table, thereby realizing retrieval in an encrypted state and targeted downloading, avoiding It reduces the waste of download resources.

Description

Data search and query method and system based on searchable encryption and homomorphic encryption

technical field

The invention relates to the field of information security, in particular to a data search and query method and system based on searchable encryption and homomorphic encryption.

Background technique

With the continuous deployment of "digital new infrastructure" decisions, data sources, forms, and scales continue to increase, and the demand for data management capabilities is also increasing. At present, the storage and management tasks of massive power data can no longer be undertaken by local equipment, so the form of data center is widely used to store and manage data. At the same time, these data often contain some sensitive data, such as user personal privacy information, expert database data, etc. This kind of data has confidentiality requirements and is not suitable for storage in the data center directly operated and maintained by the operator. At present, data owners often use data encryption to ensure data security, but this will cause encrypted data to be used after decryption, calculation encryption, etc., which will increase time costs and system resource consumption, with high efficiency and low cost .

For the security processing technology of sensitive data, there are currently three mainstream solutions: Secure Multi-party Computation (SecureMulti-party Computation, SMPC), trusted hardware (Trusted ExecutionEnvironment, TEE), and fully homomorphic encryption (Fully- Homomorphic Encryption, FHE).

Secure multi-party computing refers to the joint calculation of a joint function by distributed multi-party participants. Each participant provides its own input information and obtains corresponding output results. The security of secure multi-party computation ensures that the input and output of each participant will not be leaked to other participants. This source language can be used to construct secure queries on data—database maintainers and customers can be participants in secure multi-party computations; joint functions can be designed to use customer input as query conditions and obtain output as query results ciphertext.

Fully homomorphic encryption is a special encryption scheme. The data encrypted by this scheme can directly perform arbitrary operations on the ciphertext without decryption, and obtain the ciphertext of the operation result——after decrypting it, it can be The expected calculation results can be obtained. Fully homomorphic encryption can also be used to construct query functions on encrypted data.

In recent years, the hardware-based trusted computing technology commonly used in new processors provides another way of thinking for encrypted data processing. It can provide three features to ensure the security and privacy of data processing: completely isolated execution, data encapsulation and remote attestation. The implementation of these features will vary according to the products of different manufacturers. Specifically, "isolated execution" divides a specific area in memory, and at the same time restricts only specific processes from accessing that memory space; other processes, even system processes, hypervisors, and system management modules Data in this part of the memory space cannot be accessed. "Data encapsulation" is through encryption and authentication, so that only a specific process can perform the decryption operation and get the original data. "Remote attestation" proves to users that all code is executing securely and without modification. These three features cooperate with each other, and the effect is equivalent to the user dividing a "hidden area" on the server. Any operation in this space-the server cannot observe, modify and control the data and operations in this area. This technology can also be used to build trusted database queries. The cloud server installs trusted hardware, encrypts the data, and puts the data required for specific operations into the "hidden area" (it is worth mentioning that even if the data is in the The "hidden area" is decrypted, and the decrypted plaintext is still invisible to the cloud server), which can realize data search and query functions.

At present, the protection and use of sensitive power data mainly adopts the idea of authority management and access control, stores sensitive data in the negative list space, and ensures the security of sensitive data through user authority approval and security zone operation. However, for highly sensitive data such as expert databases, this simple safeguard may still be broken, and once a data leak occurs, it will cause great harm and is not enough to meet the needs of the business department. Sensitive data is stored in encrypted form to fully guarantee its security.

After the data is encrypted and saved, if you want to use the data directly, you need to download and decrypt it, which is very expensive. In order to make it usable to a certain extent, technologies such as search and query for encrypted data need to be designed, but the existing encrypted data search technology cannot meet the actual needs.

Contents of the invention

In order to solve the problem that the traditional sensitive data protection method relies on authority management and access control, which has defects in security and wastes downloading resources in the case of encryption, the present invention proposes a data search and query method based on searchable encryption and homomorphic encryption. include:

generating a search voucher for the keyword based on the keyword of the data item that the data visitor needs to obtain;

Retrieving a data entry index mapping table corresponding to the search credential and including the keyword from a pre-built data entry index mapping table based on the search credential;

Decrypting the data entry index mapping table to obtain a data entry index, and downloading data items to be obtained based on the data entry index;

Wherein, the data entry index mapping table is constructed by segmenting data in the form of text through a word segmentation technology, and extracting keywords therein from the word segmentation.

Preferably, the construction of the data entry index mapping table includes:

Word segmentation technology is used to segment data in the form of text;

Extract keywords from the word segmentation;

A data entry index mapping table is constructed from the keywords and word segmentations corresponding to the keywords.

Preferably, generating the keyword search voucher based on the keyword of the data item that the data visitor needs to obtain includes:

Verify whether the data visitor has access rights based on the keyword query application sent by the data visitor, and generate a search voucher corresponding to the keyword for the data visitor when the data visitor has access rights .

Preferably, it also includes:

Numerical data is queried based on the query command issued by the data visitor.

Preferably, the querying of numerical data based on the query instruction issued by the data visitor includes:

Analyze the query command sent by the data visitor, and when the result of the analysis contains only addition operations, directly calculate the encrypted columns of the homomorphic encryption to obtain the ciphertext corresponding to the numerical data to be queried;

When the analysis result includes addition operation and non-addition operation, use the generated random number to blind the operation number, use the double trapdoor of the BCP cryptosystem to decrypt the blinded operation data to obtain the plaintext of the blinded data; and The plaintext of the blinded data is calculated, and the calculated result is encrypted at the same time to obtain the ciphertext with the blinding factor; the deblinding operation is performed on the ciphertext with the blinding factor to obtain the value to be queried The ciphertext corresponding to the type data.

Based on the same inventive concept, the present invention also provides a data search and query system based on searchable encryption and homomorphic encryption, including:

A voucher generation module, configured to generate a search voucher for the keyword based on the keyword of the data item that the data visitor needs to obtain;

An index mapping module, configured to retrieve a data entry index mapping table corresponding to the search credential and containing the keyword from a pre-built data entry index mapping table based on the search credential;

A data download module, configured to decrypt the data entry index mapping table to obtain a data entry index, and download data items to be obtained based on the data entry index;

Wherein, the data entry index mapping table is constructed by segmenting data in the form of text through a word segmentation technology, and extracting keywords therein from the word segmentation.

Preferably, the construction of the data entry index mapping table includes:

Word segmentation technology is used to segment data in the form of text;

Extract keywords from the word segmentation;

A data entry index mapping table is constructed from the keywords and word segmentations corresponding to the keywords.

Preferably, it also includes: a numerical data query module;

A numeric data query module, configured to query numeric data based on query instructions issued by the data visitor.

Preferably, the numerical data query module includes:

The addition operation sub-module is used to analyze the query instruction issued by the data visitor. When the analysis result contains only the addition operation, it directly calculates the encrypted columns of the homomorphic encryption to obtain the ciphertext corresponding to the numerical data to be queried;

The mixed operation sub-module is used to blind the operation number with the generated random number when the analysis result includes addition operation and non-addition operation, and use the double trapdoor of the BCP cryptosystem to decrypt the blinded operation data to obtain blindness the plaintext of the data after the blinding; and calculate the plaintext of the data after the blinding, and encrypt the result of the calculation at the same time to obtain the ciphertext with the blinding factor; deblind the ciphertext with the blinding factor The cipher text corresponding to the numerical data to be queried is obtained through the operation.

Compared with prior art, the beneficial effect of the present invention is:

The present invention provides a data search and query method based on searchable encryption and homomorphic encryption, including: generating a keyword search voucher based on the keyword of the data item that the data visitor needs to obtain; Retrieve the data entry index mapping table containing keywords corresponding to the search voucher from the entry index mapping table; decrypt the data entry index mapping table to obtain the data entry index, and download the data item to be obtained based on the data entry index; where, the data The entry index mapping table is constructed by segmenting the data in the form of text through the word segmentation technology, and extracting the keywords from the word segmentation. The present invention converts the keywords to be obtained into a data entry index mapping table through the search voucher, and downloads the required data items based on the data entry index mapping table, thereby realizing retrieval in an encrypted state and targeted downloading, avoiding It reduces the waste of download resources.

Description of drawings

Fig. 1 is the flow chart of the data search and query method based on searchable encryption and homomorphic encryption of the present invention;

Fig. 2 is a schematic diagram of the encrypted data search and query method of the present invention;

FIG. 3 is a schematic diagram of the deployment architecture and workflow of the present invention.

Detailed ways

In order to better understand the present invention, the content of the present invention will be further described below in conjunction with the accompanying drawings and examples. The present invention proposes a data search and query method based on searchable encryption and homomorphic encryption. By adopting efficient searchable encryption, a cryptographic protocol is designed—a query scheme for numerical data is designed by using efficient cryptographic source language. In addition to file keyword search, the present invention also supports SQL statement query on relational databases. Specifically, for query statements that can be expressed only by addition, the present invention uses homomorphic encryption technology to obtain query results directly through calculations on encrypted data; for complex queries that require both addition and multiplication, the present invention uses auxiliary The way of the server, so as to construct a complete database query mechanism.

Example 1:

The present invention proposes a data search and query method based on searchable encryption and homomorphic encryption, as shown in Figure 1: including:

Step 1: Generate a search voucher for the keyword based on the keyword of the data item that the data visitor needs to obtain;

Step 2: Retrieving a data entry index mapping table corresponding to the search voucher and including the keyword from a pre-built data entry index mapping table based on the search voucher;

Step 3: Decrypt the data entry index mapping table to obtain a data entry index, and download the data items to be obtained based on the data entry index;

Wherein, the data entry index mapping table is constructed by segmenting data in the form of text through a word segmentation technology, and extracting keywords therein from the word segmentation.

Before step 1 also include:

The construction of the data entry index mapping table includes:

Word segmentation technology is used to segment data in the form of text;

Extract keywords from the word segmentation;

A data entry index mapping table is constructed from the keywords and word segmentations corresponding to the keywords.

The construction of the data entry index mapping table is described in detail below:

For confidential data in the form of text, the data owner first uses the word segmentation technology to segment the multiple fields in the data table to extract the keywords. Keywords are usually important content words representing the information of this data item.

Based on the keyword set extracted above, the data owner builds a keyword-based search index for the text file. The form of the index is a mapping table of [keyword A, the index of the data item containing the keyword A in all items of the file].

The data owner uses the mainstream encryption algorithm (such as AES) to encrypt all the data in the text file line by line, and at the same time encrypts each extracted keyword separately. For the established index structure, it is replaced by cryptographic methods (such as bilinear pairing), so as to prevent others from knowing the corresponding relationship between keywords and data items according to the index when they have the index; It can be used to obtain the corresponding part of the information in the index.

The present invention designs a secret search scheme for character-type fields, so that when a user wants to find a data item containing a certain keyword w, he can generate a search voucher [w] for this keyword, and send [w] to the data item The storage server, the data storage server performs a search using [w], and sends the data item (in an encrypted state) containing the keyword w to the user.

In step 1, a search voucher for the keyword is generated based on the keyword of the data item that the data visitor needs to obtain, specifically including:

Verify whether the data visitor has access rights based on the keyword query application sent by the data visitor, and generate a search voucher corresponding to the keyword for the data visitor when the data visitor has access rights . The rights management part in this embodiment is used to judge whether the data visitor has access rights, and the data storage part is used to store and manage the data.

According to the authority separation management mechanism, when a user (data visitor) wants to search and obtain data items containing a certain keyword, he needs to first send a query application to the authority management party. After verifying that the user has the query authority, the authority management party generates a search voucher corresponding to the keyword and sends it to the data storage party.

In step 2, based on the search credential, the data entry index mapping table containing the keyword corresponding to the search credential is retrieved from the pre-built data entry index mapping table, specifically including:

After receiving the search voucher for a certain keyword, the data storage party retrieves the corresponding data entry index mapping table containing the keyword in the encrypted index according to the search voucher, and returns the encrypted mapping table to the visitor.

In step 3, decrypt the data entry index mapping table to obtain a data entry index, and download the data items to be obtained based on the data entry index, specifically including:

The visitor decrypts the obtained result, downloads the corresponding encrypted data according to the obtained data entry index and decrypts it.

The present invention also includes querying numerical data based on query instructions issued by the data visitor.

Data format and privacy protection requirements:

双陷门是指对于任意公钥pk_i加密的数据/>

除可用其对应的私钥sk_i解密之外，还存在一个“特权密钥”mk能够解密任何私钥加密的密文即第二陷门；线性和非线性运算是SQL查询语句的重要组成部分。Data is stored in the form of two-dimensional tables. In a database table, each column is a field, also known as an attribute, and each row is a record, also known as a tuple. There are two types of attributes in the database table: numeric and character; at the same time, each attribute includes numeric and character, which may be sensitive or non-sensitive data. Sensitive character data is encrypted using a searchable encryption scheme to ensure the search function for character data; sensitive numerical data is encrypted by BCP encryption algorithm, where BCP encryption algorithm is a State and double trapdoor encryption algorithms to ensure linear and nonlinear operations on numerical data. Among them, adding homomorphism refers to the ciphertext of two data encrypted with the same key, and a certain operation on the ciphertext can directly obtain the ciphertext of the sum of the two data without decryption, that is, Enc _pk (m ₁ )+Enc _pk (m ₂ )=Enc _pk (m ₁ +m ₂ ); Key homomorphism means that the sum of two public keys is used as the data encrypted by the public key, and a private key corresponding to the two public keys can be used decryption, ie

The double trapdoor refers to the data encrypted for any public key pk _i />

In addition to being decrypted with its corresponding private key sk _i , there is also a "privileged key" mk that can decrypt any ciphertext encrypted by the private key, that is, the second trapdoor; linear and nonlinear operations are an important part of SQL query statements .

In particular, for query operations on numerical data, three types of situations may be involved: 1. The query statement only involves non-sensitive data; 2. The query statement only involves sensitive data; 3. The query statement involves both sensitive and non-sensitive data. For case 1, the query function in the existing database management system can be completed.

For cases 2 and 3, in the present invention, we use a structured database, also known as a data table, for storage in view of the characteristics of power data. Each record contains several fields, also known as attributes, where a record refers to a row of a database table , also becomes a tuple, some of these fields are character type, and some are numeric type, where the numeric type includes integer or floating-point data. Some of these attributes are sensitive data and some are general data. In order to protect the security and privacy of the data, and ensure that the data storage service provider can perform necessary operations on the data, the necessary operations here include search and query, the present invention uses a homomorphic encryption scheme to encrypt sensitive data; for non-sensitive data use Plaintext storage to improve the efficiency of data query operations. At the same time, the present invention designs corresponding schemes to complete specific operations on data. In particular, considering that in the actual application scenarios of power grid data, multi-column data query operations often occur, the present invention makes a special design for the data query scheme for the execution of cross-column query statements.

Querying numerical data based on query instructions sent by the data visitor includes:

Analyze the query command sent by the data visitor, and when the result of the analysis contains only addition operations, directly calculate the encrypted columns of the homomorphic encryption to obtain the ciphertext corresponding to the numerical data to be queried;

When the analysis result includes addition operation and non-addition operation, use the generated random number to blind the operation number, use the double trapdoor of the BCP cryptosystem to decrypt the blinded operation data to obtain the plaintext of the blinded data; and The plaintext of the blinded data is calculated, and the calculated result is encrypted at the same time to obtain the ciphertext with the blinding factor; the deblinding operation is performed on the ciphertext with the blinding factor to obtain the value to be queried The ciphertext corresponding to the type data.

For query statements aimed at numerical data, the present invention designs a query analysis scheme, which resolves user query instructions into basic algebraic operations between various data items in the database table.

For the analyzed basic operations, the present invention designs different operation execution schemes according to the types of operations and the data involved. First of all, for the linear operation involving only a single encrypted column, the linear operation here includes addition and multiplication, the present invention uses the addition homomorphic property of the BCP encryption system, that is, the ciphertext encrypted with the same key can be encrypted without decryption. In this case, the ciphertext of the sum of two plaintexts after addition is directly obtained, and the result of the operation is directly obtained. Secondly, for the addition and multiplication operations involving two encrypted columns, the present invention designs an interactive protocol between the storage server and the auxiliary operation server to complete the operation. Specifically, the storage server first generates a random number, uses the random number to blind the operand, and sends the blinded operand to the auxiliary computing server; the auxiliary computing server uses the double trapdoor property of the BCP cryptosystem to obtain the blind After completing the calculation and encryption, the ciphertext of the operation result with the blinding factor is sent to the storage server; the storage server deblinds it to obtain the final operation result. Finally, for the numerical comparison operation involving multiple encrypted columns, the comparison operation here is to judge the size of two numerical values or determine whether the two numerical values are equal. The present invention designs a similar The interactive protocol completes the operation, as shown in Figure 2.

Considering the currently commonly used deployment scenarios, the present invention adopts the deployment architecture shown in Figure 3: the data center is responsible for data storage and maintenance, and at the same time, we use computing cloud servers that belong to different service providers from the data storage cloud. In this embodiment, a storage cloud and an auxiliary computing cloud server are used, and an interaction protocol between the storage cloud and the auxiliary computing cloud server is designed to complete nonlinear operations. This design only requires the computing cloud server to be responsible for extremely light calculations, so economical deployment can be achieved.

If the parsed result only contains addition operations, directly use the properties of homomorphic encryption to operate on the ciphertext, complete the query and return the query results to the user, where the query results refer to the numerical results that meet the conditions and the results that meet the conditions data item.

Example 2:

The query operation for numeric data is as follows:

In order to be compatible with other database systems and user habits, the query of numerical data is performed in the form of SQL statements. When executing a SQL statement, it is necessary to first calculate the query condition statement, which is usually an algebraic operation statement and a logic operation statement. Here, the kangaroo operation statement includes addition, subtraction, multiplication, and power operation, etc., and the logic operation statement is such as "=" operation Determine whether the values at both ends of the operator are equal; the ">" operation determines whether the value before the operator is greater than the value of the data after the operator; then output the result that meets the query condition statement in the query. Wherein because part data is already encrypted, so the present invention has designed special-purpose calculation scheme under cipher text; And output query result after obtaining the calculation result of query condition sentence, the present invention directly adopts the ripe scheme in the existing database management system namely Can.

[Scenario 1] When an operation such as SELECT*WHERE A+B=6 is executed, two operations need to be performed—the addition operation of A+B, and the equivalent judgment operation of the result of A+B operation and c. The addition operation is a linear operation, and the equivalence comparison is a nonlinear operation.

[Scenario 2] When performing operations such as SELECT*WHERE A>5; or SELECT*WHERE A=B, it is necessary to perform comparison and equivalence test operations, both of which are non-linear operations.

[Scenario 3] When performing operations such as SELECT COUNT(*)WHERE A>5 or SELECT SUM(*)WHERE2*B>C, COUNT (counting), SUM (summing) and multiplication operations are required. All are linear operations.

[Scenario 4] When performing an operation such as SELECT*WHERE A X B>15, it is necessary to perform an A×B multiplication operation.

Processing type 1: operations involving only a single attribute addition:

For A+B, A×B, etc. involving algebraic operations on multiple attributes, when attribute A is data stored in ciphertext, key pk _A is used, and attribute B is data stored in plaintext, according to the nature of BCP homomorphic encryption , the data center first encrypts each element of attribute B with pk _A , and obtains the ciphertext of the operation result under the ciphertext. For operations involving multiple data items of the same attribute, such as SUM, AVE, etc., the data center can also directly complete the operation by using the nature of homomorphic encryption. For multiplication operations such as α×A, where α is an integer, the ciphertext of the multiplication result can be obtained by using the mathematical properties of BCP encryption.

Processing type two: Algebraic operations involving multiple columns of encrypted data:

For A+B, A×B, etc. that involve algebraic operations on multiple attributes, when attributes A and B are data stored in ciphertext, using the keys pk _A and pk _B respectively, the data center cannot independently Calculated on the ciphertext. Therefore, the present invention designs a dedicated interactive protocol to complete the calculation process with the cooperation of the auxiliary calculation server.

和/>

并将它们发送给辅助计算服务器。辅助计算服务器掌握mk，可将其解密得到a+r₁和b+r₂，随后进行求和运算得到(a+b)+(r₁+r₂)，并将其加密得到/>

并将其发回给数据中台；其中加密密钥pk_c是根据BCP密码系统的密钥同态性质，利用pk_a和pk_b计算而得。数据中台得到/>

后，因其保留了r₁和r₂，因此根据密码系统同态性质可在密文中减去(r₁+r₂)，得到

其他涉及到多列加密数据的代数运算，诸如乘法运算等，也通过类似的交互协议的思路可以得到计算结果的密文。Taking the addition operation as an example, when calculating A+B, it is necessary to calculate the sum of each group of data items in the attribute, that is, a+b, and the operation result is recorded as c. First, the data center generates two random numbers r ₁ and r ₂ , and obtains the ciphertext of a+r ₁ and b+r ₂ according to the homomorphism of the BCP encryption algorithm

and />

and send them to the secondary computing server. The auxiliary computing server masters mk, can decrypt it to get a+r ₁ and b+r ₂ , then perform summation to get (a+b)+(r ₁ +r ₂ ), and encrypt it to get />

And send it back to the data center; the encryption key pk _c is calculated by using pk _a and pk _b according to the key homomorphic property of the BCP cryptosystem. Obtained in the data center />

Afterwards, because r ₁ and r ₂ are retained, according to the homomorphic property of the cryptographic system, (r ₁ +r ₂ ) can be subtracted from the ciphertext to obtain

Other algebraic operations involving multi-column encrypted data, such as multiplication, can also obtain the ciphertext of the calculation results through similar ideas of interactive protocols.

Processing type three: non-algebraic operations involving multiple columns of encrypted data:

Non-algebraic operations are mainly comparison operations, including judging the magnitude of two values and judging whether two values are equal. According to different inputs in the database management system, X types need to be processed: (1) compare a ciphertext data and a plaintext data; (2) compare two data encrypted with the same key; (3) compare two different encrypted data key-encrypted data. Because of the homomorphism of the BCP encryption system, the case (1) can be converted to (2) for processing.

Similarly, the comparison result, also known as output, should also have two types—plaintext output and ciphertext output. For example, when executing a query that outputs calculation results such as SELECT COUNT(*)WHERE A>B, the result of each comparison is an intermediate data for executing the query, which needs to be kept secret in the middle of the data, so the comparison result should be in ciphertext form; if Execute a query such as SELECT*WHEREA>B, because when the statement is executed, it is directly determined whether to output the tuple after each comparison, so there is no need to keep the data center secret, and the comparison result is output in plain text. Therefore, the present invention designs a comparison protocol as shown in Table 1. In particular, according to the current research progress, it is easier to design the protocol to obtain the comparison result of the Goldwasser-Micali cryptosystem, also known as QR encryption. The present invention continues to use the current achievement, that is, utilizes the thinking design of the existing scheme to obtain the result encrypted with the QR encryption scheme, and designs a protocol to change the encryption mode, and converts the QR encrypted data into BCP encrypted data, as shown in Table 1.

Table 1

So far, the present invention has designed a method for solving SQL query conditional statements under ciphertext, and then adopts a method in a mature database management system to output the corresponding content.

Example 3:

Based on the same inventive concept, the present invention also provides a data search and query system based on searchable encryption and homomorphic encryption, including:

A voucher generation module, configured to generate a search voucher for the keyword based on the keyword of the data item that the data visitor needs to obtain;

An index mapping module, configured to retrieve a data entry index mapping table corresponding to the search credential and containing the keyword from a pre-built data entry index mapping table based on the search credential;

A data download module, configured to decrypt the data entry index mapping table to obtain a data entry index, and download data items to be obtained based on the data entry index;

Wherein, the data entry index mapping table is constructed by segmenting data in the form of text through a word segmentation technology, and extracting keywords therein from the word segmentation.

A data search and query system based on searchable encryption and homomorphic encryption further includes: a table construction module for constructing a data entry index mapping table.

The form building blocks are specifically used for:

Word segmentation technology is used to segment data in the form of text;

Extract keywords from the word segmentation;

A data entry index mapping table is constructed from the keywords and word segmentations corresponding to the keywords.

The table building blocks are described in detail below:

For confidential data in the form of text, the data owner first uses the word segmentation technology to segment the multiple fields in the data table to extract the keywords. Keywords are usually important content words representing the information of this data item.

Based on the keyword set extracted above, the data owner builds a keyword-based search index for the text file. The form of the index is a mapping table of [keyword A, the index of the data item containing the keyword A in all items of the file].

The data owner uses the mainstream encryption algorithm (such as AES) to encrypt all the data in the text file line by line, and at the same time encrypts each extracted keyword separately. For the previously established index structure, it is replaced by cryptographic methods (such as bilinear pairing), so as to prevent others from knowing the corresponding relationship between keywords and data items according to the index when they have the index; The voucher can be used to obtain the corresponding part of the information in the index.

The present invention designs a secret search scheme for character-type fields, so that when a user wants to find a data item containing a certain keyword w, he can generate a search voucher [w] for this keyword, and send [w] to the data item The storage server, the data storage server uses [w] to perform a search, and sends the data item containing the keyword w to the user in an encrypted state.

The credential generation module is specifically used for:

Verify whether the data visitor has access rights based on the keyword query application sent by the data visitor, and generate a search voucher corresponding to the keyword for the data visitor when the data visitor has access rights . The rights management part in this embodiment is used to judge whether the data visitor has access rights, and the data storage part is used to store and manage the data.

According to the authority separation management mechanism, when a user, also known as a data visitor, wants to search and obtain data items containing a certain keyword, he needs to first send a query application to the authority management party. After verifying that the user has the query authority, the authority management party generates a search voucher corresponding to the keyword and sends it to the data storage party.

In the index mapping module, the data entry index mapping table containing the keyword corresponding to the search voucher is retrieved from the pre-built data entry index mapping table based on the search voucher, specifically including:

After receiving the search voucher for a certain keyword, the data storage party retrieves the corresponding data entry index mapping table containing the keyword in the encrypted index according to the search voucher, and returns the encrypted mapping table to the visitor.

The data download module is specifically used for:

The visitor decrypts the obtained result, downloads the corresponding encrypted data according to the obtained data entry index and decrypts it.

The present invention also includes a numerical data query module;

A numeric data query module, configured to query numeric data based on query instructions issued by the data visitor.

Numerical data query modules include:

The addition operation sub-module is used to analyze the query instruction issued by the data visitor. When the analysis result contains only the addition operation, it directly calculates the encrypted columns of the homomorphic encryption to obtain the ciphertext corresponding to the numerical data to be queried;

The mixed operation sub-module is used to blind the operation number with the generated random number when the analysis result includes addition operation and non-addition operation, and use the double trapdoor of the BCP cryptosystem to decrypt the blinded operation data to obtain blindness the plaintext of the data after the blinding; and calculate the plaintext of the data after the blinding, and encrypt the result of the calculation at the same time to obtain the ciphertext with the blinding factor; deblind the ciphertext with the blinding factor The cipher text corresponding to the numerical data to be queried is obtained through the operation.

For query statements aimed at numerical data, the present invention designs a query analysis scheme, which resolves user query instructions into basic algebraic operations between various data items in the database table.

For the analyzed basic operations, the present invention designs different operation execution schemes according to the types of operations and the data involved. First of all, for linear (addition and multiplication) operations involving only a single encrypted column, the addition operation submodule of the present invention uses the addition homomorphic property of the BCP encryption system, that is, the ciphertext encrypted with the same key can be encrypted without decryption. In this case, the ciphertext of the sum of two plaintexts after addition is directly obtained, and the result of the operation is directly obtained. Secondly, for the addition and multiplication operations involving two encrypted columns, the hybrid operation sub-module of the present invention designs an interactive protocol between the storage server and the auxiliary operation server to complete the operation. Specifically, the storage server first generates a random number, uses this random number to blind the operand, and sends the blinded operand to the auxiliary computing server; the auxiliary computing server uses the double trapdoor property of the BCP cryptosystem to obtain the blinded After completing the calculation and encryption, the ciphertext of the operation result with the blinding factor is sent to the storage server; the storage server deblinds it to obtain the final operation result. Finally, for the numerical comparison operation involving multiple encrypted columns, the comparison operation here includes judging the size of two numerical values or judging whether the two numerical values are equal. The present invention designs a similar The interactive protocol completes the operation.

We use computing cloud servers belonging to different service providers from the data storage cloud. Here we use the storage cloud and auxiliary computing cloud servers, and design the interaction protocol between the storage cloud and auxiliary computing cloud servers to complete nonlinear operations. This design only requires the computing cloud server to be responsible for extremely light calculations, so economical deployment can be achieved.

If the parsed result only contains addition operations, the addition operation submodule directly uses the property of homomorphic encryption to operate on the ciphertext, completes the query and returns the query results to the user. The query results here are numerical results that meet the conditions and A data item that satisfies the condition.

Apparently, the described embodiments are some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

The above is only an embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the present invention pending approval. within the scope of the claims.

Claims (9)

1. The data searching and inquiring method based on the searchable encryption and homomorphic encryption is characterized by comprising the following steps:

generating a search certificate of a keyword based on the keyword of the data item required to be acquired by a data visitor;

retrieving a data entry index mapping table containing the key words corresponding to the search credentials from a pre-constructed data entry index mapping table based on the search credentials;

decrypting the data entry index mapping table to obtain a data entry index, and downloading a data item to be acquired based on the data entry index;

the data entry index mapping table is constructed by segmenting text form data through a word segmentation technology and extracting keywords from the segmented data.

2. The method of claim 1, wherein the constructing of the data entry index mapping table comprises:

word segmentation is carried out on the text-form data by adopting a word segmentation technology;

extracting real words representing the item information of the data from the data obtained by word segmentation as keywords;

and constructing a data entry index mapping table from the indexes of the key words and the data items containing the key words in all the items of the file.

3. The method of claim 1, wherein the generating the search credentials for the key based on the key for the data item that the data visitor needs to obtain comprises:

and based on the keyword query application sent by the data visitor, verifying whether the data visitor has access rights, and generating a search credential corresponding to the keyword for the data visitor when the data visitor has the access rights.

4. The method as recited in claim 1, further comprising:

and inquiring the numerical data based on the inquiry instruction sent by the data visitor.

5. The method of claim 4, wherein querying the numeric data based on the query instruction issued by the data visitor comprises:

analyzing a query instruction sent by a data visitor, and directly calculating an encrypted column encrypted in the same state to obtain a ciphertext corresponding to numerical data to be queried when an analysis result only comprises an addition operation;

when the analysis result comprises addition operation and non-addition operation, blinding the operand by using the generated random number, and decrypting the blinded operand by using a double trapdoor of the BCP cryptographic system to obtain a plaintext of the blinded data; calculating the plaintext of the blinded data, and encrypting the calculated result to obtain the ciphertext with the blinding factor; and performing blind removal operation on the ciphertext with the blind factors to obtain ciphertext corresponding to the numerical data to be queried.

6. A searchable and homomorphic encryption based data search and query system comprising:

the credential generation module is used for generating a search credential of the keyword based on the keyword of the data item required to be acquired by the data visitor;

an index mapping module, configured to retrieve a data entry index mapping table corresponding to the search credential and containing the keyword from pre-constructed data entry index mapping tables based on the search credential;

the data downloading module is used for decrypting the data entry index mapping table to obtain a data entry index and downloading the data item to be acquired based on the data entry index;

the data entry index mapping table is constructed by segmenting text form data through a word segmentation technology and extracting keywords from the segmented words.

7. The system of claim 6, wherein the construction of the data entry index mapping table comprises:

word segmentation is carried out on the text-form data by adopting a word segmentation technology;

extracting real words representing the item information of the data from the data obtained by word segmentation as keywords;

and constructing a data entry index mapping table from the indexes of the key words and the data items containing the key words in all the items of the file.

8. The system as recited in claim 6, further comprising: a numerical data query module;

and the numerical data query module is used for querying the numerical data based on the query instruction sent by the data visitor.

9. The system of claim 8, wherein the numeric data query module comprises:

the addition operation sub-module is used for analyzing the query instruction sent by the data visitor, and directly calculating the homomorphic encryption column to obtain the ciphertext corresponding to the numerical data to be queried when the analysis result is that the query instruction only comprises addition operation;

the mixed operation sub-module is used for blinding the operand by using the generated random number when the analysis result comprises the addition operation and the non-addition operation, and decrypting the blinded operation data by using the double trapdoor of the BCP cryptographic system to obtain the plaintext of the blinded data; calculating the plaintext of the blinded data, and encrypting the calculated result to obtain the ciphertext with the blinding factor; and performing blind removal operation on the ciphertext with the blind factors to obtain ciphertext corresponding to the numerical data to be queried.

Images