CN111526155A - System for protecting user privacy in social network and optimal path matching method - Google Patents

Abstract

The invention relates to a system for protecting user privacy in a social network and an optimal path matching method. The method includes: the trusted center generates system parameters and keys, and sends the system parameters to the server and the user; the user registers; the server registers; the user constructs vertex information and weight information, and will configure the vertex information and weight information in the construction of the vertex information and weight information. The ciphertext information generated in the information process is sent to the server; the server is used to construct a social graph according to the vertex information and weight information; the query user in the user uploads the origin and destination identifiers to the server; the server generates a path sequence according to the social graph and the origin and destination identifiers and weight sequence; the server sends the weight sequence to the query user, so that the query user can determine the ciphertext position; the query user determines the optimal path from the path sequence by inadvertent transmission according to the ciphertext position. The invention can ensure that the user's privacy is not leaked during the route query, and meanwhile, the query speed is fast, and the processing is accurate and efficient.

Description

A system and optimal path matching method for protecting user privacy in social network

technical field

The invention relates to the technical field of social networks, in particular to a system for protecting user privacy in a social network and an optimal path matching method.

Background technique

In order to realize various intelligent services, such as recommendation services, sharing services, and query services, online social networks need to use the user's personal attribute information to perform information correlation query and matching to obtain the optimal recommendation in the social network. However, these user attribute information contains many users' personal privacy, such as user age, gender, work unit, place of residence and other sensitive information. Malicious attackers can obtain and analyze these data and use them for illegal operations through monitoring and other means. Therefore, how to protect the privacy of users in social networks is an important topic.

In a social network, all users upload their private information to a cloud server. The user can query the propagation path between the two coordinates by providing the source point information and the end point information. The optimal propagation path means that the user's propagation cost is smaller and the propagation is more effective. Through the optimal propagation path, users can quickly and efficiently propagate a message to target users. To protect privacy, user information is usually encrypted before uploading, and the propagation cost (defined as weight) between users needs to be compared in an outsourcing environment. Therefore, the whole process should ensure that the server cannot know the user's sensitive data, and the system user cannot obtain any information except the propagation path. There are also some methods to solve the above problems, such as the path query method based on the homomorphic encryption algorithm; ciphertext. However, in the existing path query schemes, firstly, the definition of weight information between vertices is not accurate enough, and the interaction between vertices is not considered, so the purpose of path query cannot be clearly expressed; secondly, the query speed is positively correlated with the number of vertices, which cannot guarantee large-scale social networks. The utility of the query may require a trade-off between accuracy and efficiency.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a system for protecting user privacy in a social network and an optimal path matching method, which can ensure that the propagation paths of any two vertices of a large social network can be quickly queried and the query speed does not depend on the number of users , to improve privacy protection security.

For achieving the above object, the present invention provides the following scheme:

A system for protecting user privacy in a social network, comprising: a trusted center, a server and a user, the trusted center is used to generate system parameters and keys, and send the parameters to the server; the user registers Post-construct vertex information and weight information, and send the ciphertext information generated in the process of constructing the vertex information and the weight information to the server, and the server is used for registering according to the vertex information and The weight information constructs a social graph, the query user among the users is used to provide the identities of himself and the target user, the server is used to query all the propagation paths and the corresponding weights, and through inadvertent transmission, it is The query user provides the optimal path.

An optimal path matching method for protecting user privacy in a social network, comprising:

The trusted center generates system parameters and keys, and sends the system parameters to the server and the user;

the user registers;

the server is registered;

The user constructs vertex information and weight information, and sends the ciphertext information generated in the process of constructing the vertex information and the weight information to the server;

The server is configured to construct a social graph according to the vertex information and the weight information;

The query user among the users uploads the origination destination identifier to the server;

The server generates a path sequence and a weight sequence according to the social graph and the starting and ending point identifiers;

The server sends the weight sequence to the query user, so that the query user determines the location of the ciphertext;

The query user determines the optimal path from the path sequence in an inadvertent transmission mode according to the ciphertext position.

Optionally, the trusted center generates system parameters and keys, and sends the system parameters to the server and the user, specifically including:

Obtain the safety parameter κ;

p是大素数且满足|p|＝κ，

是[1,p-1]内的任意整数，g是

的一个生成元；选择随机数R，其中|R|＜|p|/3；According to the security parameters, calculate the master public key mpk and master key msk of the encryption algorithm, where mpk=(g ^msk , g, p),

p is a large prime number satisfying |p|=κ,

is any integer in [1,p-1], and g is

A generator of ; choose a random number R, where |R|<|p|/3;

sk₁+sk₂＝msk mod(p-1)；Obtain a key pair (sk ₁ , sk ₂ ), the key pair includes a first key sk ₁ and a second key sk ₂ , wherein,

sk ₁ +sk ₂ =msk mod(p-1);

其中，

Randomly obtain n user key vectors

in,

Publish parameter information, where the parameter information includes the master public key, the generator, the prime number and the random number.

Optionally, the user performs registration, which specifically includes:

send the first registration request;

返回至所述当前用户；The trusted center randomly selects an integer from the integer sequence as the current user's identity ID _i according to the first registration request, and uses the key vector

return to the current user;

Send the identification ID _i to all users connected with the current user;

After the inquiring user obtains the identity identifier ID _i , a second registration request is sent;

The trusted center returns the first key sk ₁ to the query user according to the second registration request.

Optionally, the server performs registration, which specifically includes:

send a third registration request;

The trusted center returns the second key sk ₂ to the server according to the third registration request.

Optionally, the user constructs vertex information and weight information, and sends the ciphertext information generated in the process of constructing the vertex information and the weight information to the server, specifically including:

Get the attribute information of the current user USER _i ;

The attribute information is binarized by one-hot encoding, so that only one bit in the binary value corresponding to each dimension attribute has a value of 1; the user converts all discrete attributes of the individual into attribute vectors of length w

加密所述属性向量

得到在社交网络中的顶点信息v_i；The user USER _i according to the key vector

encrypt the attribute vector

get the vertex information _vi in the social network;

For each connected user USER _j , the current user USER _i sends an application to the trusted center;

The trusted center returns the weight key according to the application;

Using the encryption homomorphism property and the weight key, calculate the weight ciphertext, and send the weight ciphertext to the server;

The server integrates the ciphertext to obtain weight information.

Optionally, the server generates a path sequence and a weight sequence according to the social graph and the starting and ending point identifiers, specifically including:

In the social graph, determine all the connected vertices of the vertex v _s corresponding to the starting point identifier, and define it as the first layer starting vertex set S ₁ ; find all the connecting vertices of the vertex v _t corresponding to the end point identifier and define it as the first layer A layer of termination vertex set T ₁ ;

中，同时查询顶点v_u对应的身份标识ID_u，将ID_u加入到路径序列

中；Determine the common vertex of the first layer start vertex set S ₁ and the first layer end vertex set T ₁ , denoted as v _u , where v _u ∈ {S ₁ ∩ T ₁ }, query according to the social graph The weights between vertices v _s and v _u and v _u and v _t are denoted as Es _,u and _Eu,t respectively. Add Es _,u ·E _u,t to the weight sequence

, at the same time query the ID _{u corresponding to the vertex v u ,} and add the ID _u to the path sequence

middle;

Determine all connected vertices v _j of each vertex v _i of the first-layer starting vertex set S ₁ according to the social graph, and define it as the second-layer starting vertex set S ₂ , while changing v _i →v _j Add to the preceding vertex set P ₁ ;

中，同时查询顶点v_i和v_u对应的身份标识ID_i和ID_u，将ID_i·R+ID_u加入到路径序列

中；Determine the common vertex of the second layer start vertex set S ₂ and the first layer end vertex set T ₁ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₁ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and record it as v _i , query the weights between vertices v _s and v _i , v _i and v _u , and v _u and v _t in turn according to the social graph and denoted as E _s,i , E _i,u and E _u,t respectively, add Es _,i ·E _i,u ·E _u,t to the weight sequence

, simultaneously query the identities ID _i and ID _u corresponding to the vertices v _i and v _u , and add ID _i R+ID _u to the path sequence

middle;

Determine all connected vertices v _j' of each vertex v _i' of the first-level termination vertex set T ₁ according to the social graph, and define it as the second-level termination vertex set T ₂ , while changing v _i' →v _j' is added to the post-vertex set P ₂ ;

中，同时查询顶点v_i、v_i′以及v_u对应的身份标识ID_i、ID_i′以及ID_u，将ID_i·R²+ID_u·R+ID_i′加入到

中，得到路径序列

和权值序列

Determine the common vertex of the second-level starting vertex set S ₂ and the second-level ending vertex set T ₂ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₂ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and denote it as v _i , find the post-vertex of v _u in the post-vertex set P ₂ and denote it as vi _' , according to the social graph Query the weights between vertices v _s and v _i , v _i and v _u , v _u and v _i ', and v _i ' and v _u in turn, and denote them as E _s,i , E _i,u , E _{u, i′} and E _i′,t , add Es _,i ·E _i,u ·E _u,i′ ·E _i′,t to

, simultaneously query the identities ID _i , ID _i ' and ID _u corresponding to the vertices v _i , v _i ' and v _u , and add ID _i · R ² +ID _u · R+ID _i' to the

, get the path sequence

and weight sequence

Optionally, the server sends the weight sequence to the querying user, so that the querying user determines the location of the ciphertext, specifically including:

的每个元素得到第一解密权值序列

并发送给所述查询用户；Decrypt the weight sequence according to the second key sk ₂

Each element of gets the first decryption weight sequence

and send it to the query user;

的每个元素，得到第二解密权值序列

The query user decrypts the first weight sequence according to the first key sk ₁

For each element of , the second decryption weight sequence is obtained

得到密文位置。According to the second decryption weight sequence

Get the ciphertext location.

Optionally, the query user determines the optimal path from the path sequence in an inadvertent transmission mode according to the position of the ciphertext, specifically including:

i＝1,2,…,q，其中q为序列

中元素的个数，

The server sends q random integers C _i to the query user,

i=1,2,...,q, where q is the sequence

the number of elements in ,

计算q个公钥，并将各所述公钥发送至所述服务器，其中密文位置对应的所述公钥由密钥k生成，其余各所述公钥根据此所述公钥与所述整数C_i生成；The query user generates a key

Calculate q public keys, and send each public key to the server, where the public key corresponding to the ciphertext position is generated by the key k, and the other public keys are based on this public key and the Integer C _i is generated;

The server checks each of the public keys, and obtains a check result;

The server encrypts the path sequence with the public key according to the check result, and sends the path sequence to the query user;

The query user decrypts the path ciphertext corresponding to the ciphertext position in the path sequence according to the key k, and the query user can determine the optimal path of the social network by inadvertent transmission.

According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

(1) The present invention has high security, and all processes are realized by using the ElGamal encryption system. Even if the server has complete graph information, such as vertex information and weight information, as long as there is no collusion between the user and the server, the server and external attackers cannot obtain any private information. At the same time, the user cannot recover the user key of the connected user from the weight key obtained by the inner product of the vector. Therefore, the present invention has high privacy protection security.

(2) Using the inadvertent transmission method, due to the difficult assumption of discrete logarithms, the server does not know the specific content of the path the user wants to query, that is, the specific location of the ciphertext b cannot be determined; because the server verifies the correctness of the public key in the transmission , the user can only get one path, that is, the key required for other ciphertexts cannot be forged or determined except for k.

(3) The present invention proposes an optimal path matching method for protecting user privacy in a social network. On the premise of ensuring security, by using the data structure of the bidirectional breadth-first search of the source point and the destination point, the server operation speed is fast, The processing is efficient, and the parameters in the process can be pre-generated to reduce the actual running time.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the present invention. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative labor.

1 is a schematic diagram of a system composition for protecting user privacy in a social network of the present invention;

2 is a schematic diagram of data transmission within the system for protecting user privacy in the social network of the present invention;

3 is a flowchart of an optimal path matching method for protecting user privacy in the social network of the present invention;

4 is a flowchart of entity registration in a method according to an embodiment of the present invention;

FIG. 5 is a flowchart of inadvertent transmission in a method according to an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The purpose of the present invention is to provide a system for protecting user privacy in a social network and an optimal path matching method, which can ensure that the propagation paths of any two vertices of a large social network can be quickly queried and the query speed does not depend on the number of users , to improve privacy protection security.

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

查询密钥(sk₁,sk₂)、权值密钥f、查询密钥(sk₁,sk₂)，USER_i将个人的敏感信息通过加密上传到服务器2，服务器2通过计算并整合构建出一张由密文组成的社交图G＝(V,E)，用户3中的查询用户RU提供自身和目标用户的身份标识(ID_s,ID_t)，服务器2查询出所有的传播路径

和对应的权值

查询用户解密权值序列得到最优传播路径的下标，通过多对一不经意传输OT₁ ^q的方式，查询用户可以获取最优传播路径。FIG. 1 is a schematic diagram of the composition of a system for protecting user privacy in a social network according to the present invention. As shown in FIG. 1 , a system for protecting user privacy in a social network includes: a trusted center 1 , a server 2 , and a user 3 . FIG. 2 is a schematic diagram of data transmission within the system for protecting user privacy in the social network of the present invention. As shown in Figure 2, the system includes a trusted center (TA), a server (CS) and a user (USER). Trusted center 1 is used to generate system parameters and keys, and send the parameters to server 2; after user 3 registers, it constructs vertex information and weight information, and uses the ciphertext information generated in the process of constructing vertex information and weight information. Send to server 2. After registration, server 2 is used to construct a social graph based on vertex information and weight information. The query user in user 3 is used to provide the identity of himself and the target user. Server 2 is used to query all propagation paths and The corresponding weights are transmitted inadvertently to provide the optimal path for query users. Suppose there are n users USER _i (i=1,2,...,n) in the system. Trusted center 1 allocates and calculates system parameters: user 3's identity ID, user key

Query key (sk ₁ , sk ₂ ), weight key f, and query key (sk ₁ , sk ₂ ), USER _i uploads personal sensitive information to server 2 through encryption, and server 2 constructs a A social graph G=(V, E) composed of ciphertext, the query user RU in user 3 provides the identities (ID _s , ID _t ) of itself and the target user, and server 2 queries all propagation paths

and the corresponding weights

The query user decrypts the weight sequence to obtain the subscript of the optimal propagation path, and the query user can obtain the optimal propagation path by inadvertently transmitting OT ₁ ^q by many-to-one.

FIG. 3 is a flowchart of an optimal path matching method for protecting user privacy in a social network according to the present invention. As shown in Figure 3, an optimal path matching method for protecting user privacy in a social network includes:

Step 101: The trusted center generates system parameters and keys, and sends the system parameters to the server and the user, specifically including:

Obtain the safety parameter κ.

p是大素数且满足|p|＝κ，

代表[1,p-1]内的任意整数，g是

的一个生成元，选择随机数R，其中|R|＜|p|/3。According to the security parameters, calculate the master public key mpk and master key msk of the encryption algorithm, where mpk=(g ^msk , g, p),

p is a large prime number satisfying |p|=κ,

represents any integer in [1,p-1], g is

A generator of , choose a random number R, where |R|<|p|/3.

sk₁+sk₂＝msk mod(p-1)。Obtain a key pair (sk ₁ , sk ₂ ), the key pair includes a first key sk ₁ and a second key sk ₂ , wherein,

sk ₁ +sk ₂ =msk mod(p-1).

其中，

Randomly obtain n user key vectors

in,

Publish parameters (mpk, g, p, R), that is, the public key used for encryption, the generator element, the prime number modulus information, and the system random number information used for the integration path.

Step 102: The user performs registration, which specifically includes:

Send the first registration request.

返回至所述当前用户。The trusted center randomly selects an integer from the integer sequence as the current user's identity ID _i according to the first registration request, and uses the key vector

Return to the current user.

The identity identifier ID _i is sent to all users connected to the current user.

The inquiring user sends a second registration request after obtaining the identity identifier ID _i .

The trusted center returns the first key sk ₁ to the query user according to the second registration request.

Step 103: The server performs registration, which specifically includes:

A third registration request is sent.

The trusted center returns the second key sk ₂ to the server according to the third registration request.

Step 104: The user constructs vertex information and weight information, and sends the ciphertext information generated in the process of constructing the vertex information and the weight information to the server, specifically including:

Get the attribute information of the current user USER _i .

The attribute information is binarized by one-hot encoding, so that only one bit in the binary value corresponding to each dimension attribute has a value of 1; the user can convert all discrete attributes of the individual into attribute vectors of length w

加密所述属性向量

得到在社交网络中的顶点信息v_i；其中：The user USER _i according to the key vector

encrypt the attribute vector

Get the vertex information _vi in the social network; where:

For each connected user USER _j , the current user USER _i sends an application to the trusted center.

The trusted center returns the weight key according to the application, specifically:

采用下式计算：Using the encryption homomorphism property and the weight key, calculate the weight ciphertext, and send the weight ciphertext to the server; wherein the weight ciphertext

Calculated using the following formula:

where rank _i,j is a predefined integer intimacy value between USER _i and USER _j .

其中权值e_i,j定义为：

The server integrates the ciphertext to obtain weight information. The specific weight information is:

The weights e _i,j are defined as:

Step 105: The server is configured to construct a social graph according to the vertex information and the weight information; construct a social graph G=(V, E) according to the vertex information and the weight information, where: V={ v _i |i∈[1,n]}, E={E _i,j |i,j∈[1,n], USER _i and USER _j are connected}.

Step 106: The query user among the users uploads the origin and destination identifiers to the server.

Step 107: The server generates a path sequence and a weight sequence according to the social graph and the starting and ending point identifiers, specifically including:

In the social graph, determine all the connected vertices of the vertex v _s corresponding to the starting point identifier, and define it as the first layer starting vertex set S ₁ ; find all the connecting vertices of the vertex v _t corresponding to the end point identifier and define it as the first layer A layer terminates vertex set T ₁ .

中，同时查询顶点v_u对应的身份标识ID_u，将ID_u加入到路径序列

中。Determine the common vertex of the first layer start vertex set S ₁ and the first layer end vertex set T ₁ as v _u , where v _u ∈{S ₁ ∩T ₁ }, and query the vertex according to the social graph The weights between v _s and v _u and v _u and v _t are denoted as Es _,u and E _u,t respectively. Add Es _,u ·E _u,t to the weight sequence

, at the same time query the ID _{u corresponding to the vertex v u ,} and add the ID _u to the path sequence

middle.

Determine all connected vertices v _j of each vertex v _i of the first-layer starting vertex set S ₁ according to the social graph, and define it as the second-layer starting vertex set S ₂ , while changing v _i →v _j Add to the preceding vertex set P ₁ .

中，同时查询顶点v_i和v_u对应的身份标识ID_i和ID_u，将ID_i·R+ID_u加入到路径序列

中。Determine the common vertex of the second layer start vertex set S ₂ and the first layer end vertex set T ₁ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₁ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and record it as v _i , query the weights between vertices v _s and v _i , v _i and v _u , and v _u and v _t in turn according to the social graph and denoted as E _s,i , E _i,u and E _u,t respectively, add Es _,i ·E _i,u ·E _u,t to the weight sequence

, simultaneously query the identities ID _i and ID _u corresponding to the vertices v _i and v _u , and add ID _i R+ID _u to the path sequence

middle.

Determine all connected vertices v _j' of each vertex v _i' of the first-level termination vertex set T ₁ according to the social graph, and define it as the second-level termination vertex set T ₂ , while changing v _i' →v _j' is added to the post-vertex set P ₂ .

中，同时查询顶点v_i、v_i′以及v_u对应的身份标识ID_i、ID_i′以及ID_u，将ID_i·R²+ID_u·R+ID_i′加入到

中，得到路径序列

和权值序列

Determine the common vertex of the second-level starting vertex set S ₂ and the second-level ending vertex set T ₂ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₂ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and denote it as v _i , find the post-vertex of v _u in the post-vertex set P ₂ and denote it as vi _' , according to the social graph Query the weights between vertices v _s and v _i , v _i and v _u , v _u and v _i ', and v _i ' and v _u in turn, and denote them as E _s,i , E _i,u , E _{u, i′} and E _i′,t , add Es _,i ·E _i,u ·E _u,i′ ·E _i′,t to

, simultaneously query the identities ID _i , ID _{i '} and ID _u corresponding to the vertices v _i , v _i ' and v _u , and add ID _i · R ² +ID _u · R+ID _i' to the

, get the path sequence

and weight sequence

Step 108: The server sends the weight sequence to the querying user, so that the querying user can determine the location of the ciphertext, which specifically includes:

的每个元素得到第一解密权值序列

并发送给所述查询用户。Decrypt the weight sequence according to the second key sk ₂

Each element of gets the first decryption weight sequence

and sent to the query user.

的每个元素，得到第二解密权值序列

The query user decrypts the first weight sequence according to the first key sk ₁

For each element of , the second decryption weight sequence is obtained

得到密文位置。According to the second decryption weight sequence

Get the ciphertext location.

的每个元素得到

并发送给查询用户：Decrypt the sequence of weights with the second key sk ₂

Each element of gets

and send to the querying user:

的每个元素：The query user decrypts the weight sequence with the first key sk ₁

Each element of :

That is, the final decrypted w" _i is the total weight of each path. Through a sorting algorithm, such as a bubble algorithm, RU can select the subscript of the minimum value, that is, the position of the ciphertext.

Step 109: The query user determines the optimal path from the path sequence in an inadvertent transmission mode according to the ciphertext position, which specifically includes:

其中q为序列

中元素的个数，即

The server sends q random integers C _i to the query user,

where q is the sequence

the number of elements in

计算出第b个所述公钥β_b，而其他所述公钥分两个方向逐步由相邻下标的所述公钥与所述随机整数计算而得，形成链式结构：The querying user calculates q public keys, and sends each of the public keys β ₁ , β ₂ , . . . , β _q to the server; wherein a random number is selected

The b-th public key β _b is calculated, and the other public keys are gradually calculated in two directions from the public keys of adjacent subscripts and the random integers, forming a chain structure:

β _i =C _i /β _i+1 mod p,i=1,2,...,b-1

β _b =g ^k mod p

β _j =C _j-1 /β _j-1 mod p,j=b+1,b+2,...,q

The server checks each of the public keys, that is, checks C _i =β _i ·β _i+1 mod p, and obtains a check result.

The server encrypts the path sequence with the public key according to the inspection result, and sends the path sequence to the query user; the server encrypts m _i with β _i and sends it to the RU:

并通过迭代计算最终得到最优路径中的每个顶点标识ID_i＝(m_b-(m_b mod R))/R。The query user decrypts the path ciphertext corresponding to the ciphertext position in the path sequence according to the key k, and the query user can determine the optimal path of the social network by inadvertent transmission. That is, the query user uses the key k to decrypt the b-th ciphertext c _b to obtain the optimal path

And through the iterative calculation, the ID _i =(m _b -(m _b mod R))/R of each vertex in the optimal path is finally obtained.

Example 1:

The present invention provides an optimal path matching method for protecting user privacy in a social network, comprising the following steps:

Step 1: System parameter generation.

其中p是个大素数且满足|p|＝κ，g是

的一个生成元。选择随机数R，其中|R|＜|p|/3。Step 1.1: The trusted center (TA) selects the security parameter κ, and calculates the master public key (mpk=g ^msk , g, p) and master key of the ElGamal encryption algorithm

where p is a large prime number satisfying |p|=κ, and g is

a generator of . Choose a random number R, where |R| < |p|/3.

使得sk₁+sk₂＝msk mod(p-1)。Step 1.2: TA chooses a pair of keys (sk ₁ , sk ₂ ), where

Let sk ₁ +sk ₂ =msk mod(p-1).

Step 1.3: TA randomly selects n user key vectors

Step 1.4: TA publishes parameters (mpk, g, p, R).

The entity registration process refers to Figure 4.

Step 2: Entity registration.

Step 2.1: User USER _i (i=1,2,...,n) registers.

Step 2.1.1: User USER _i (i=1,2,...,n) sends a registration request.

返回给此用户。Step 2.1.2: TA randomly selects an integer from the integer sequence {1,2,...,n} as the user's identity ID _i , and uses

Return to this user.

Step 2.1.3: User USER _i sends his identity ID _i to all his connected users.

Step 2.1.4: The query user RU sends a registration request.

Step 2.1.5: The TA returns the key sk ₁ to the RU.

Step 2.2: Server CS registration.

Step 2.2.1: The server CS sends a registration request.

Step 2.2.2: The TA returns the key sk ₂ to the CS.

Step 3: Social graph construction.

Step 3.1: Vertex information construction.

Step 3.1.1: The user USER _i binarizes his attribute information through one-hot encoding, that is, only one bit in the binary value corresponding to each dimension attribute has a value of 1. in:

When the gender is male: sex _i =10; when the gender is female: sex _i =01.

When the age is 0-20: age _i = 100; when the age is 21-50: age _i = 010; when the age is 50 or more: age _i = 001.

其中w表示用户的属性向量的长度。So that all discrete attributes of the user are converted into attribute vectors with each element as a binary value

where w represents the length of the user's attribute vector.

加密属性向量

形成顶点

Step 3.1.2: USER _i with the key

encrypted attribute vector

form vertex

Step 3.2: Construction of weight information.

Step 3.2.1: For each connected user USER _j , user USER _i sends an application to TA

计算并返回权值密钥f_j,i：Step 3.2.2: TA queries the user key vector corresponding to ID _j

Compute and return the weight key f _j,i :

并发送给服务器CS：Step 3.2.3: USER _i uses the ElGamal homomorphism to calculate the weight ciphertext

and send to server CS:

where rank _i,j is a predefined integer intimacy value between USER _i and USER _j .

Step 3.2.4: CS integrates the ciphertext and forms the weights as:

The weights e _i,j are defined as:

This shows that the closer the two users are, the faster the message spreads, making the weight between users smaller.

Step 3.3: According to the vertices and edges, the social graph G=(V, E) can be constructed.

Where: V={v _i |i∈[1,n]}, E={E _i,j |i,j∈[1,n], USER _i and USER _j are connected}.

Step 4: Path query.

Step 4.1: query the user RU to upload the identifiers (ID _s , ID _t ) corresponding to the origin and destination to the server CS.

和权值序列

Step 4.2: CS Generate Path Sequence

and weight sequence

Step 4.2.1: CS finds all connected vertices of the starting point v _s in the graph G, and defines it as the first-level starting vertex set S ₁ ; finds all the connecting vertices of the end point v _t and defines it as the first-level termination vertex Set _T1 . which excludes v _s and v _t .

中，将v_u对应的身份标识ID_u加入到

中。Step 4.2.2: CS picks out the common vertex v _u ∈ {S ₁ ∩ T ₁ } in S ₁ and T ₁ , queries the weights between vertices v _s and v _u and v _u and v _t , and assigns E _s,u ·E _{u,t is} added to

, add the ID _u corresponding to v _u to the

middle.

in:

The weights of the two connected edges are aggregated together using the homomorphic property of the ElGamal encryption system.

Step 4.2.3: CS finds all connected vertices v _j of each vertex v _i of S ₁ in graph G, and defines it as the second-level starting vertex set S ₂ while adding v _i → v _j to the preceding vertices Set P ₁ ; which excludes v _s and v _t .

中，同时查询顶点v_i和v_u对应的身份标识ID_i和ID_u，将ID_i·R+ID_u加入到

中。Step 4.2.4: CS picks out the common vertex v _u ∈ {S ₂ ∩ T ₁ } in S ₂ and T ₁ , finds the preceding vertex of v _u in P ₁ as v _i , and queries vertices v _s and v weights between _i , vi and v _{u , and v u and} v _t , and add Es _,i ·E _i,u ·E _u,t to

, simultaneously query the identities ID _i and ID _u corresponding to the vertices v _i and v _u , and add ID _i R+ID _u to the

middle.

in:

Step 4.2.5: CS finds all connected vertices v _j' of each vertex v _i' of T ₁ in the graph G, and defines it as the second-level termination vertex set T ₂ while adding v _i' → v _j' to Post-vertex set P ₂ ; which excludes v _s and v _t .

中，同时查询顶点v_i、v_i′以及v_u对应的身份标识ID_i、ID_i′以及ID_u，将ID_i·R²+ID_u·R+ID_i′加入到

中。Step 4.2.6: CS picks out the common vertex v _u ∈ {S ₂ ∩ T ₂ } in S ₂ and T ₂ , finds the pre-vertex of v _u in P ₁ as v _i , and finds out in P ₂ The post vertex of v _u is v _i′ , according to the social graph, query the weights between vertices v _s and v _i , v _i and v _u , v _u and v _i ′, and v _i ′ and v _u in turn, Add E _s,i ·E _i,u ·E _u,i′ ·E _i′,t to

, simultaneously query the identities ID _i , ID _{i '} and ID _u corresponding to the vertices v _i , v _i ' and v _u , and add ID _i · R ² +ID _u · R+ID _i' to the

middle.

in:

Step 4.3: RU selects the subscript b of the optimal path in the path sequence.

的每个元素得到

并发送给RU：Step 4.3.1: CS decrypts weight sequence with key sk ₂

Each element of gets

and send to RU:

的每个元素：Step 4.3.2: RU decrypts weight sequence with key sk ₁

Each element of :

That is, the final decrypted w" _i is the total weight of each path. Through a sorting algorithm, such as the bubble algorithm, RU can select the subscript corresponding to the minimum value as

The inadvertent transmission process refers to Figure 5.

Step 4.4: RU inadvertently obtains the optimal path m _b .

其中q为序列

中元素的个数，即

Step 4.4.1: CS sends q random integers to RU

where q is the sequence

the number of elements in

计算β₁,β₂,…,β_q并发送给CS：Step 4.4.2: RU chooses random number

Calculate β ₁ , β ₂ ,…,β _q and send to CS:

β _i =C _i /β _i+1 mod p,i=1,2,...,b-1

β _b =g ^k mod p

β _j =C _j-1 /β _j-1 mod p,j=b+1,b+2,...,q

Step 4.4.3: CS checks C _i =β _i ·β _i+1 mod p.

Step 4.4.4: CS encrypts m _i with β _i and sends to RU:

并通过迭代计算最终得到最优路径中的每个顶点标识ID_i＝(m_b-(m_b mod R))/R。Step 4.4.5: RU uses the key k to decrypt the b-th ciphertext c _b to obtain the optimal path

And through the iterative calculation, the ID _i =(m _b -(m _b mod R))/R of each vertex in the optimal path is finally obtained.

Based on the ElGamal homomorphic encryption and the inadvertent transmission OT ₁ ^q method, the invention realizes an optimal path matching scheme for protecting user privacy in a social network, and the scheme realizes resistance to external attacks and internal attacks.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other.

In this paper, specific examples are used to illustrate the principles and implementations of the present invention. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present invention; meanwhile, for those skilled in the art, according to the present invention There will be changes in the specific implementation and application scope. In conclusion, the contents of this specification should not be construed as limiting the present invention.

Claims (9)

1. a system for protecting user privacy in a social network, comprising: a trusted center, a server and a user, the trusted center is used to generate system parameters and keys, and send the parameters to the server; the user constructs vertex information and weight information after registration, and sends the ciphertext information generated in the process of constructing the vertex information and the weight information to the server, and the server is used for A social graph is constructed according to the vertex information and the weight information, the query user among the users is used to provide the identities of himself and the target user, the server is used to query all propagation paths and corresponding weights, The optimal path is provided for the querying user in an inadvertently transmitted manner. 2. An optimal path matching method for protecting user privacy in a social network, characterized in that, comprising: The trusted center generates system parameters and keys, and sends the system parameters to the server and the user; the user registers; the server is registered; The user constructs vertex information and weight information, and sends the ciphertext information generated in the process of constructing the vertex information and the weight information to the server; The server is configured to construct a social graph according to the vertex information and the weight information; The query user among the users uploads the origination destination identifier to the server; The server generates a path sequence and a weight sequence according to the social graph and the starting and ending point identifiers; The server sends the weight sequence to the query user, so that the query user determines the location of the ciphertext; The query user determines the optimal path from the path sequence in an inadvertent transmission mode according to the ciphertext position. 3. The optimal path matching method for protecting user privacy in the social network according to claim 1, wherein the trusted center generates system parameters and a key, and sends the system parameters to the server and the user, Specifically include: Obtain the safety parameter κ; According to the security parameters, calculate the master public key mpk and master key msk of the encryption algorithm, where mpk=(g ^msk , g, p),

p is a large prime number satisfying |p|=κ,

is any integer in [1,p-1], and g is

A generator of , choose a random number R, where |R|<|p|/3; Obtain a key pair (sk ₁ , sk ₂ ), the key pair includes a first key sk ₁ and a second key sk ₂ , where sk ₁ ,

sk ₁ +sk ₂ =msk mod(p-1); Randomly obtain n user key vectors

in,

Publish parameter information, where the parameter information includes the master public key, the generator, the prime number and the random number. 4. The optimal path matching method for protecting user privacy in a social network according to claim 3, wherein the user registers, specifically comprising: send the first registration request; The trusted center randomly selects an integer from the integer sequence as the current user's identity ID _i according to the first registration request, and uses the key vector

return to the current user; Send the identification ID _i to all users connected with the current user; After the inquiring user obtains the identity identifier ID _i , a second registration request is sent; The trusted center returns the first key sk ₁ to the query user according to the second registration request. 5. The optimal path matching method for protecting user privacy in a social network according to claim 3, wherein the server performs registration, specifically comprising: send a third registration request; The trusted center returns the second key sk ₂ to the server according to the third registration request. 6. The optimal path matching method for protecting user privacy in a social network according to claim 3, wherein the user constructs vertex information and weight information, and will construct the vertex information and the weight information when constructing the vertex information and the weight information. The ciphertext information generated in the information process is sent to the server, which specifically includes: Get the attribute information of the current user USER _i ; The attribute information is binarized by one-hot encoding, so that only one bit in the binary value corresponding to each dimension attribute has a value of 1; the user can convert all discrete attributes of the individual into attribute vectors of length w

The user USER _i according to the key vector

encrypt the attribute vector

get the vertex information _vi in the social network; For each connected user USER _j , the current user USER _i sends an application to the trusted center; The trusted center returns the weight key according to the application; Using the encryption homomorphism property and the weight key, calculate the weight ciphertext, and send the weight ciphertext to the server; The server integrates the ciphertext to obtain weight information. 7. The optimal path matching method for protecting user privacy in a social network according to claim 3, wherein the server generates a path sequence according to the social graph and the starting and ending point identifiers

and weight sequence

Specifically include: In the social graph, determine all the connected vertices of the vertex v _s corresponding to the starting point identifier, and define it as the first layer starting vertex set S ₁ ; find all the connecting vertices of the vertex v _t corresponding to the end point identifier and define it as the first layer A layer of termination vertex set T ₁ ; Determine the common vertex of the first layer start vertex set S ₁ and the first layer end vertex set T ₁ , denoted as v _u , where v _u ∈ {S ₁ ∩ T ₁ }, query according to the social graph The weights between vertices v _s and v _u and v _u and v _t are denoted as E _s,u and E _u,t , respectively. Es _,u ·E _u,t will be added to the

, at the same time query the ID _u corresponding to the vertex v _u , and add the ID _u to the

middle; Determine all connected vertices v _j of each vertex v _i of the first-layer starting vertex set S ₁ according to the social graph, and define it as the second-layer starting vertex set S ₂ , while changing v _i →v _j Add to the preceding vertex set P ₁ ; Determine the common vertex of the second layer start vertex set S ₂ and the first layer end vertex set T ₁ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₁ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and record it as v _i , query the weights between vertices v _s and v _i , v _i and v _u , and v _u and v _t in turn according to the social graph and denoted as E _s,i , E _i,u and E _u,t respectively, add E _s,i ·E _i,u ·E _u,t to

, simultaneously query the identities ID _i and ID _u corresponding to the vertices v _i and v _u , and add ID _i R+ID _u to the

middle; Determine all connected vertices v _j' of each vertex v _i' of the first-level termination vertex set T ₁ according to the social graph, and define it as the second-level termination vertex set T ₂ , while changing v _i' →v _j' is added to the post-vertex set P ₂ ; Determine the common vertex of the second-level starting vertex set S ₂ and the second-level ending vertex set T ₂ , and update it as v _u , where v _u ∈ {S ₂ ∩ T ₂ }, and at the same time before the Find the pre-vertex of v _u in the vertex set P ₁ and denote it as v _i , find the post-vertex of v _u in the post-vertex set P ₂ and denote it as vi _' , according to the social graph Query the weights between vertices v _s and v _i , v _i and v _u , v _u and v _i' , and v _i' and v _u in turn, and denote them as Es _,i , E _i,u , E _{u, i′} and E _i′,t , add Es _,i ·E _i,u ·E _u,i′ ·E _i′,t to

, simultaneously query the identities ID _i , ID _{i '} and ID _u corresponding to the vertices v _i , v _i' and v _u , and add ID _i · R ² +ID _u · R+ID _i' to the

, get the path sequence

and weight sequence

8 . The optimal path matching method for protecting user privacy in a social network according to claim 3 , wherein the server sends the weight sequence to the query user, so that the query user can determine the ciphertext. 9 . location, including: Decrypt the weight sequence according to the second key sk ₂

Each element of gets the first decryption weight sequence

and send it to the query user; The query user decrypts the first weight sequence according to the first key sk ₁

For each element of , the second decryption weight sequence is obtained

According to the second decryption weight sequence

Get the ciphertext location. 9 . The optimal path matching method for protecting user privacy in a social network according to claim 3 , wherein the query user determines the optimal path from the path sequence by inadvertent transmission according to the ciphertext position. 10 . , including: The server sends q random integers C _i to the query user,

where q is the sequence

the number of elements in ,

The query user generates a key

Calculate q public keys, and send each public key to the server, where the public key corresponding to the ciphertext position is generated by the key k, and the other public keys are based on this public key and the Integer C _i is generated; The server checks each of the public keys, and obtains a check result; The server encrypts the path sequence with the public key according to the check result, and sends the path sequence to the query user; The querying user decrypts the path ciphertext corresponding to the ciphertext position in the path sequence according to the key k, and the querying user can determine the optimal path of the social network by inadvertent transmission.

Images