Mobile social network privacy protection routing method based on commutative encryption

Abstract

This invention discloses a privacy-preserving routing method for mobile social networks based on exchangeable encryption, comprising: (1) system setup: setting up a mobile social network including trusted organizations (TAs), fixed networks, and mobile nodes; (2) degree centrality calculation: determining whether the identifier sets of the nodes that meet have an intersection. If there is no intersection, the nodes perform bidirectional authentication. If the authentication is successful, a global degree centrality calculation is performed. Then, it is determined whether the two nodes are in the same group. If they are in the same group, a local degree centrality calculation is performed; (3) routing efficiency value calculation: after the degree centrality of the two nodes meets anonymously and completes the calculation, the routing efficiency value is calculated for the messages in the cache of both nodes. m (3) Calculate the routing efficiency value of the node; (4) Message forwarding: When two nodes meet, their efficiency values ​​are compared. If the efficiency value of the other party is greater than that of the node itself, the message is sent to the other party. This method can protect the privacy information of all nodes without relying on a third-party trusted institution (TA) and simultaneously calculate the social metric value of node degree centrality and perform message forwarding.

Description

Technical Field

[0001] This invention relates to the field of network security technology, specifically a privacy-preserving routing method for mobile social networks based on exchangeable encryption. Background Technology

[0002] Mobile Social Networks (MSNs) are a combination of social networks and mobile communication networks, consisting of multiple users with social attributes interconnected through various mobile devices. When mobile nodes in an MSN meet, they transmit messages using high-speed, short-range, low-cost wireless communication technologies such as Wi-Fi and Bluetooth, employing a "store-carry-forward" routing method. This allows mobile users to effectively share and transmit data such as news, pictures, music, videos, and traffic information. MSNs have important applications in Pocket Switched Networks (PSNs), vehicular networks, and wireless sensor networks.

[0003] To effectively deliver messages, some MSN routing protocols calculate social metrics such as degree centrality, betweenness centrality, and social similarity when nodes meet, and then calculate routing utility values ​​based on these metrics for message forwarding (see: Pan H, Crowcroft J, Yoneki E. BUBBLE Rap: social-based forwarding in delay-tolerant networks. IEEE Transactions on Mobile Computing, 2011, 10(11): 1576-1589 and Dalyd E, Haahr M. Social network analysis for information flow in disconnected delay-tolerant MANETs. IEEE Transactions on Mobile Computing, 2009, 8(5): 606-621). These routing protocols use real identifiers for communication when nodes meet, which does not consider privacy issues and could leak private information such as the number of encounters, time, and location.

[0004] To protect node privacy, the first anonymous communication solution was proposed in the paper (Kate A, Zaverucha G, Hengartner U. Anonymity and security in delay-tolerant networks. Proc of IEEE SecureComm, Nice, France, 2007:504-513), which proposed a security framework and ensured effective secure communication through an identity-based encryption method. The paper (Ahmad N, Cruickshank H, Cao Y, et al. Privacy by architecture pseudonym framework for delay-tolerant network. Future Generation Computer Systems, 2019, 93:979-992) proposed a new secure and private communication method that uses pseudonyms to hide the real identity of nodes, preventing privacy attacks and providing a high degree of anonymity. Using real identities would reveal various private information, while anonymity would prevent nodes from collecting various encounter information to calculate metrics. The above privacy protection methods do not consider how to anonymously calculate node metrics for message forwarding while protecting node privacy. To address this, the paper (Chen K, Shen HY. FaceChange: holding neighbor node anonymity in mobile opportunity-oriented social networks with fine-grained control. IEEE / ACM Transactions on Networking, 2017, 25(2): 1176-1189) proposes a FaceChange strategy. In FaceChange, nodes communicate anonymously when connecting to prevent other nodes from recognizing the true identity of the node they intend to attack. When the network disconnects, the encounter information is forwarded to the other node through other intermediate nodes, and then the metric is calculated based on the encounter information. Although FaceChange can prevent nodes from being identified and attacked when they meet, the encounter information obtained by the intermediate nodes after the node leaves is still based on the node's true identity, thus leaking private information such as the encounter time and location.The literature (Jiang QF, Deng K, Zhang L, et al. A privacy-preserving protocol for utility-based routing in DTNs. Information, 2019, 10(4): 1-11) proposes a privacy-preserving PPUR protocol to solve the above problems. This protocol can anonymously calculate the node routing metric for message forwarding while protecting the node's privacy information. However, PPUR requires a third-party trusted authority (TA) to calculate the metric. Over-reliance on the TA is not suitable for MSNs in many cases and has a single point of failure problem. The literature (Jiang Q FA privacy-preserving exchange-based routing protocol for opportunistic networks. Security and Communication Networks, 2022, Article ID 6815911, 1-13) proposes a privacy-preserving routing protocol PPERP. In this protocol, nodes can effectively collect node encounter record information to calculate the metric while protecting the privacy information of non-transferring nodes. PPERP does not rely on a third-party trusted authority (TA) and is suitable for MSNs where nodes are typically sparsely distributed. However, it can only protect the privacy information of non-transmitting nodes and cannot protect the privacy of all nodes at the same time.

[0005] In summary, regarding the privacy protection issue of MSNs routing, the above routing method cannot protect the privacy information of all nodes and simultaneously calculate social metrics without relying on a third-party trusted authority (TA). Summary of the Invention

[0006] The purpose of this invention is to address the privacy protection issue in MSNs routing by providing a privacy-preserving MSNs routing method. This method can protect the privacy information of all nodes without relying on a third-party trusted authority (TA), and simultaneously calculate the social metric value of node degree centrality (the number of different nodes encountered within a certain period of time) for message forwarding.

[0007] The technical solution to achieve the purpose of this invention is: a privacy-preserving routing method for mobile social networks based on exchangeable encryption, comprising four steps: system setup, degree centrality calculation, routing efficiency value calculation, and message forwarding. The specific steps are as follows:

[0008] (1) The system settings are as follows:

[0009] The MSNs system network model consists of three parts: Trusted Organization (TA), fixed network, and mobile nodes.

[0010] The Authorization Center (TA) is responsible for generating the real and pseudo-random identifiers, as well as public and private keys, for mobile nodes. The fixed network, including the wired Internet and wireless access points (APs), connects the TA and the mobile nodes. Mobile nodes represent pedestrians or vehicles carrying mobile devices with high-speed, short-range wireless communication capabilities such as Wi-Fi and Bluetooth. Mobile nodes use pseudo-random identifiers for communication and authenticate themselves using the public and private keys issued by the TA.

[0011] (2) The degree centrality calculation is specifically as follows:

[0012] Assume two nodes N i and N j Encounter, N i and N j The commutative encryption function is used to determine if there is an intersection between the set of identifiers of the nodes that have already met and the set of pseudo-random identifiers of the other node. If there is an intersection, degree centrality is not calculated; otherwise, node N has no intersection. i and N j Two-way authentication will be performed. If authentication is successful, global degree centrality will be calculated, the global degree centrality value will be incremented by one, and the pseudo-random identifier of the other party will be added to its own set of encountered node identifiers, MeetPIDSet. Then, a commutative encryption function will be used to determine whether the two nodes are in the same group. If they are in the same group, local degree centrality will be calculated, and the local degree centrality value will be incremented by one.

[0013] (3) The routing efficiency value is calculated as follows:

[0014] When two nodes meet, for a message m in each other's caches, each node first determines whether it is the destination node of message m. If a node is the destination node of m, its efficiency value for message m is 2; otherwise, it uses a commutative encryption function to determine whether it and the destination node of m are in the same group. If they are in the same group, the efficiency value is set to 1 + LocalDegree / NodeNumber; otherwise, the efficiency value is set to GlobalDegree / NodeNumber, where LocalDegree is the local degree centrality value, GlobalDegree is the global degree centrality value, and NodeNumber is the total number of nodes in the network.

[0015] (4) The message forwarding specifically refers to:

[0016] When two nodes meet, their efficiency values ​​are compared. If the other node's efficiency value is greater than its own, the message is sent to the other node.

[0017] Compared with existing technologies, this invention has significant advantages: it uses a commutative encryption function to determine whether there is an intersection between the set of identifiers of encountered nodes and the pseudo-random identifier set of the other node, and to determine whether the two nodes are in the same group, thereby calculating the global and local degree centrality of the nodes. This method can protect the privacy information of all nodes without relying on a third-party trusted institution (TA), and simultaneously calculate the social metric value of node degree centrality for message forwarding. Attached Figure Description

[0018] Figure 1 This is the main flowchart of the privacy protection routing method for mobile social networks based on exchangeable encryption, as described in this invention.

[0019] Figure 2 This is a system network model diagram.

[0020] Figure 3 This is a flowchart of degree centrality calculation.

[0021] Figure 4 This is a flowchart for determining the intersection of identifier sets.

[0022] Figure 5 This is a flowchart of the node identity authentication process.

[0023] Figure 6 It is the same set of decision flowcharts.

[0024] Figure 7 This is a flowchart of the routing efficiency value calculation process.

[0025] Figure 8 This is a message forwarding flowchart. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] This invention is based on a privacy-preserving routing method for exchangeable encryption in mobile social networks, such as... Figure 1 As shown, the process includes four steps: system setup, degree centrality calculation, routing efficiency value calculation, and message forwarding.

[0028] (1) System Settings

[0029] The system network model diagram of the present invention is as follows: Figure 2As shown, the system comprises three parts: a Trusted Authority (TA), a fixed network, and mobile nodes. The TA, acting as a trusted authorization center, is responsible for generating genuine and pseudo-random identifiers for mobile nodes, and issuing public and private key certificates. The fixed network includes wired Internet access and wireless access points (APs), connecting the TA and mobile nodes. Mobile nodes N1 to N8 represent pedestrians or vehicles carrying mobile devices with high-speed short-range wireless communication capabilities such as Wi-Fi and Bluetooth. Mobile nodes obtain keys and node identifiers from the TA through the fixed network, communicate using pseudo-random identifiers, and authenticate their identities using the public and private keys issued by the TA.

[0030] TA uses bilinear mapping technology to generate bilinear parameters based on the safety parameter k. Where G1 and G2 are groups of prime order q, and P is a generator. For a non-degenerate, efficiently computable bilinear mapping that satisfies the following conditions:

[0031] ① Bilinear

[0032] If for any

[0033] ② Non-degenerative

[0034] There exist P, Q∈G1 such that

[0035] ③ Computability

[0036] For any P, Q ∈ G1, there exists an efficient algorithm to compute

[0037] Choose the symmetric encryption algorithm AES() and the hash function H1:{0,1} * ×G2→(Z / qZ) × H2:{0,1} * →G1*, G1*:=G\{0}, select a random number s∈(Z / qZ) × Used as the master key to generate the public key P pub =sP, publish system parameters

[0038] When moving node N i When registering with the system, TA generates a set containing multiple pseudo-random identifiers. Each pseudo-random identifier Pid i =AES s (N i ||r) is generated by a random number r, the symmetric encryption algorithm AES(), and the master key s. Mobile node N i Pseudo-random identifier Pid iThe corresponding public key is PK. i =H1(Pid) i The private key is SK. i =sH1(Pid i Mobile nodes periodically use new pseudo-random identifiers for node communication, thereby hiding their true identifiers. TA also assigns a group identifier, GroupID, to mobile nodes; nodes in the same group share the same group identifier.

[0039] (2) Degree centrality calculation

[0040] Assume two nodes N i and N j The degree centrality calculation process in the case of meeting and anonymity is as follows: Figure 3 As shown, it includes several parts: determining the intersection of identifier sets, identity authentication, calculating global degree centrality, determining whether they belong to the same group, and calculating local degree centrality.

[0041] 1) Determine the intersection of the sets of identifiers

[0042] Assume node N i The set of encountered node identifiers is MeetPIDSet i N j The set of pseudo-random identifiers is PID. j If the set MeetPIDSet i and PID j If there is an intersection, it means the two nodes have met; otherwise, it means they have not met. During the process of finding the intersection, the node identifiers in the set must be protected; otherwise, node N... i By PID j The middle element knows that all pseudo-random identifiers in the set correspond to an N. j Real identifiers, thus obtaining N j The time, location, and other movement trajectory privacy information. To protect the privacy of the node identifier sets of both parties, N i and N j The existence of intersections is determined using the commutative encryption functions and methods described in the literature (Agrawal R, Evfimievski A, Srikant R. Information sharing across private databases. Proc of SIGMOD 2003, San Diego, USA, 2003: 86-97). The commutative encryption function is defined as x∈Dom,k∈Key={1,2…,(p-1) / 2-1},f k (x)≡x kmod p, where p is a 1024-bit secure prime number, meaning both p and (p-1) / 2 are prime numbers, Dom is the set of quadratic residues modulo p, and Key is the key set. A commutative cryptographic function has the property: x∈Dom,k i ,k j ∈Key,

[0043] like Figure 4 For node N i and N j The process for determining whether there is an intersection is as follows:

[0044] ①N i MeetPIDSet is processed using the hash function H:V→Dom. i All elements in the dataset are hashed and mapped to the DOM, generating a new set HashMeetPIDSet. i =H(MeetPIDSet) i ), N j PID is processed by hash function H j A new set HashPID is generated by hashing all elements in the set. j =H(PID) j ).

[0045] ②N i A private key k is randomly selected from the key set Key. i ∈Key, through a commutative cryptographic function on the set HashMeetPIDSet i Encryption is performed to generate a new encrypted set. N j Randomly select a private key k from the key set. j ∈Key, through a commutative encryption function and k j For the set HashPID j Encryption is performed to generate an encrypted set.

[0046] ③N i right The order of elements in the set is shuffled using a shuffle algorithm, and then distributed to N. j N j right The order of elements in the set is shuffled using the Shuffle algorithm, and then sent to N. i .

[0047] ④N i For received Encryption to generate a set N j For received Encryption to generate a set

[0048] ⑤ Node N i right The order of elements in the set is shuffled using the Shuffle algorithm, and then sent to N. j Node N j right The order of elements in the set is shuffled using the Shuffle algorithm, and then sent to N. i .

[0049] ⑥ Node N i Traversal All elements in N j Traversal All elements in, if If stx == y, then node N i and N j If the sets of identifiers have an intersection, stop traversing the elements; otherwise, it means that the sets of identifiers have no intersection.

[0050] Because node N i and N j Since the other party's key is unknown, it is impossible to know the elements of the other party's node set and the elements in the intersection when determining whether there is an intersection.

[0051] 2) Identity Authentication

[0052] Node N i and N j If the sets of identifiers intersect, it means they have met before, and degree centrality calculation is unnecessary. If they do not intersect, it means they have not met before, and both parties must authenticate each other to prevent spoofing attacks.

[0053] The identity authentication process uses, for example Figure 5 The challenge / response mechanism shown assumes N j For N i Identity verification includes the following process:

[0054] ①N i To N j Send a request for authentication information.

[0055] ②N j Received N i After sending the request, generate a random number R. j As a challenge.

[0056] ③N i Use its own private key SK i =sH1(Pid iThe R signature was obtained using the Hess signature method (see reference: Hess F. Efficient identity based signature schemes based on pairings. In: International Workshop on Selected Areas in Cryptography 2002, pp. 310-324). j Digital signature generation Then send to N j In response. N i Choose any P1∈G1* and k∈(Z / qZ). × Then, the digital signature is generated by calculating according to equations (1)-(4).

[0057]

[0058] v = h(R) j ,r) (2)

[0059] u=vSK i +kP1 (3)

[0060]

[0061] ④ Receive digital signature After that, node N j Use N i public key PK i =H1(Pid) i ) through equation (5) Perform authentication.

[0062] r = e(u,P)·e(H1(Pid) i ),-P pub ) v (5)

[0063] If v = h(R) j If r), then accept N. i Node N must be authenticated; otherwise, authentication will be rejected. i For N j The process of identity verification is similar.

[0064] 3) Calculate global degree centrality

[0065] If node N i and N j If the sets of identifiers have no intersection and have passed identity authentication, then the global degree centrality value is incremented by 1, and the other party's identifier is added to its own set of encountered node identifiers, MeetPIDSet.

[0066] 4) Determine if they belong to the same group

[0067] Figure 6 The diagram shows node N. i and N j To determine whether processes belong to the same group, follow these steps:

[0068] ① Determine the intersection of the sets of identifiers in the same group. N i The GroupID is determined using the hash function H:V→Dom. i Perform hash operations to map the hash data to the DOM and generate a HashGroupID. i =H(GroupID) i ), N j The GroupID is determined by the hash function H. j Perform hash operations to generate HashGroupID j =H(GroupID) j ).

[0069] ②N i Randomly select private key k i ∈Key, using a commutative cryptographic function to define its own group identifier HashGroupID i Encrypt and generate encrypted data. N j Randomly select private key k j ∈Key, through a commutative encryption function and k j HashGroupID is the identifier for its own group. j Encrypt and generate encrypted data.

[0070] ③ Node N i Will Send to N j N j Will Send to N i .

[0071] ④ Node N i For received Encryption to generate a set N j For received Encryption to generate a set

[0072] ⑤ Node N i Will Send to N j Node N j Will Send to Ni .

[0073] ⑥If This indicates that node N i and N j Groups with the same identifier belong to the same group; otherwise, they do not belong to the same group.

[0074] Because node N i and N j Since the other party's key is unknown, if the two nodes are not in the same group, it is impossible to know the group identifier information of the other party's node.

[0075] 5) Calculate locality centrality

[0076] If node N i and N j If they are in the same group, the locality centrality value is increased by 1.

[0077] (3) Calculation of routing efficiency value

[0078] Figure 7 The diagram shows the routing efficiency value calculation process. After the degree centrality of nodes meet anonymously and is calculated, the routing efficiency value of the node is calculated according to equation (6) for the message m in the buffers of both parties.

[0079]

[0080] Among them: U i For node N i The efficiency value is represented by LocalDegree, which is the local degree centrality value, GlobalDegree, which is the global degree centrality value, and NodeNumber, which is the total number of nodes in the network.

[0081] A node determines whether it is the destination node of a message. If it is the destination node of message m, its efficiency value for message m is the maximum value of 2; otherwise, it uses a commutative encryption function to determine whether it and the destination node of m are in the same group. If it and the destination node of m are in the same group, the efficiency value is set to 2. Otherwise, let the efficiency value be... If the node itself is the destination node of the message, the efficiency value is at most 2. If the node itself and the destination node of m are in the same group, the efficiency value is greater than or equal to 1 and less than 2. If the node itself and the destination node of m are not in the same group, the efficiency value is less than 1.

[0082] (4) Message forwarding

[0083] Figure 8 The diagram illustrates the message forwarding process. When two nodes meet, they compare the efficiency values ​​of messages in their caches. If the efficiency value of the other node is greater than that of the first node, the message is forwarded to the other node.

Claims

1. A privacy-preserving routing method for mobile social networks based on exchangeable encryption, characterized in that, include: (1) System settings: Configure the mobile social network, including Trusted Organizations (TAs), fixed networks, and mobile nodes; (2) Degree centrality calculation: For nodes that meet, determine whether their identifier sets have an intersection. If there is no intersection, the nodes perform two-way identity authentication. If the authentication is successful, perform global degree centrality calculation. Then, determine whether the two nodes are in the same group. If they are in the same group, perform local degree centrality calculation. (3) Calculation of routing efficiency value: After the degree centrality of two nodes meet anonymously, the routing efficiency value of the nodes is calculated for the message m in the cache of both nodes. The degree centrality calculation step (2) includes: Assume two nodes N i and N j Encounter, N i and N j The system uses a commutative encryption function to determine if there is an intersection between its own set of encountered node identifiers and the pseudo-random identifier set of the other node. If there is an intersection, degree centrality is not calculated; if there is no intersection, node N... i and N j Perform two-way authentication; if authentication is successful, calculate the global degree centrality, increment the global degree centrality value by one, and add the other party's pseudo-random identifier to its own set of encountered node identifiers, MeetPIDSet. Then, use a commutative encryption function to determine whether the two nodes are in the same group. If they are in the same group, calculate the local degree centrality value by one. (4) Message forwarding: When two nodes meet, their efficiency values ​​are compared. If the efficiency value of the other party is greater than that of the other party, the message is sent to the other party.

2. The privacy-preserving routing method for mobile social networks based on exchangeable encryption according to claim 1, characterized in that, The system setup steps (1) include: The MSNs system network model consists of three parts: Trusted Authority (TA), fixed network, and mobile nodes. The TA is the authorization center, responsible for generating real and pseudo-random identifiers and public and private keys for mobile nodes. The fixed network includes the wired Internet and wireless access points (APs), which are responsible for connecting the TA and mobile nodes. Mobile nodes represent pedestrians or vehicles carrying mobile devices with high-speed short-range wireless communication capabilities such as Wi-Fi and Bluetooth. Mobile nodes use pseudo-random identifiers for communication and authenticate their identities using public and private keys issued by the TA.

3. The privacy-preserving routing method for mobile social networks based on exchangeable encryption according to claim 1, characterized in that, The routing efficiency value calculation step (3) includes: when two nodes meet, for message m in both parties' caches, each node first determines whether it is the destination node of message m; if a node is the destination node of m, its efficiency value for message m is 2; otherwise, it determines whether it and the destination node of m are in the same group through a commutative encryption function; if it is in the same group as the destination node of m, the efficiency value is set to 1 + LocalDegree / NodeNumber; otherwise, the efficiency value is set to GlobalDegree / NodeNumber, where LocalDegree is the local degree centrality value, GlobalDegree is the global degree centrality value, and NodeNumber is the total number of nodes in the network.

Images