A privacy protection method for LBS in sparse user environment of Internet of Vehicles

Abstract

The present invention provides a privacy protection method for LBS in a sparse user environment of an Internet of Vehicles, belonging to the field of Internet of Vehicles application technology. It solves the security problem that the correlation between vehicle location information and identity information is too high and the anonymity is reduced in a sparse user environment; its technical solution is: including the following steps: S1, system initialization; S2, mixed request set generation; S3, LBS service privacy protection. The beneficial effects of the present invention are: the present invention designs a privacy protection method for LBS in a sparse user environment of an Internet of Vehicles, which uses the homomorphic encryption property to dynamically generate vehicle pseudonyms to achieve anonymous authentication, and defines a new virtual address generation algorithm to ensure the non-clustering of virtual addresses; by using a mixture of multiple pseudonyms and K anonymity technology, the indistinguishability of pseudonyms is achieved from the data source, which greatly reduces the correlation between the real identity and location of users in a sparse user environment, and cuts off the possibility of user privacy leakage.

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking application technology, and in particular to an LBS privacy protection method in a vehicle networking environment with sparse users. Background Art

[0002] Vehicular ad-hoc network (VANET) is a self-organized intelligent network that uses wireless communication technology to connect entities such as vehicles, roadside traffic facilities and traffic networks. With the rapid development of VANET technology, intelligent vehicle services have become popular, especially location-based services (LBS), which have been widely integrated into people's daily lives and brought great convenience to them. For example, during the communication between vehicles, vehicles need to broadcast their own positions and speeds to surrounding vehicles in real time in order to form real-time road conditions and achieve beyond-line-of-sight traffic safety control; during the communication between vehicles and service platforms, vehicle users can easily obtain content-based search services from LBS providers by submitting points of interest (POI) and current locations, such as querying local weather conditions, surrounding shopping malls and hotels, etc. However, location-based applications require users to send their location information to other vehicles or location service application providers, and in this process, their sensitive information may be lost.

[0003] Internet of Vehicles service privacy includes location privacy and query privacy. Location privacy leakage refers to the user's location information being inferred by attackers based on time, space and other information. Internet of Vehicles service privacy attackers can be other malicious vehicles or location service providers, among which location service providers are more threatening because they collect massive service data during the service provision process and will leak data (user request service content and location coordinate data) to others for the purpose of profit. Therefore, in the protection of Internet of Vehicles service privacy, it must be assumed that the location service provider is an untrustworthy party.

[0004] At present, experts and scholars at home and abroad have made a lot of efforts in the research of privacy and security protection of Internet of Vehicles. The research methods are diverse and can be mainly divided into encryption mechanism, caching strategy, anonymous technology, etc. The privacy protection scheme based on encryption generally processes the user's location information through a one-way irreversible encryption function, which can ensure the availability of services without leaking the user's identity and location information. However, it is expensive and computationally complex, and cannot be used in the Internet of Vehicles environment. The scheme based on the cache mechanism reduces the number of interactions between users and LBS providers to reduce the possibility of user privacy information leakage. However, these caching strategies rarely take into account the needs of the complexity of Internet of Vehicles users. Anonymous technology uses generalization theory to protect service privacy. It does not rely on any background knowledge of the attacker. It has strong privacy protection and can completely cut off the possibility of user privacy leakage from the data source. Therefore, it has been increasingly introduced into the Internet of Vehicles in recent years to protect the service privacy of vehicle users. Anonymous technology mainly involves pseudonym (anonymity) technology and k-anonymity technology.

[0005] 1. Pseudonym technology is to hide the real identity of the user by using a pseudonym, cut off the connection between the vehicle location information and the identity information, so that the attacker cannot find the user identity through the location information, and then decouple the query content from the user identity. In pseudonym technology, the degree of location privacy protection depends on the degree of association between the location information and the user identity. The higher the degree of association, the lower the degree of protection. The most common methods are the mix-zone pseudonym scheme that allows vehicles to change pseudonyms in a specified interval and the group (ring) signature scheme that treats a single vehicle as a member of a group. The basic idea of ​​the mix-zone-based pseudonym scheme is to equip the vehicle with a large number of pseudonyms that do not reveal its true identity. Each pseudonym is only used for a period of time before being replaced, and the operation of replacing the pseudonym is performed within a specific geographical area (mix-zone). Pseudonyms are short-term identity identifiers that change over time (or speed). Using pseudonyms can protect the real identity of the vehicle. Scholar capkun gave a time-based pseudonym definition:

[0006] Among them, the HMAC key hash function uses the key K vi Vehicle ID vi Hash operation is performed with time t, P is the vehicle V iThe pseudonym generated at time t. The pseudonym update of the scheme is measured in time. In 2003, scholar Eresford proposed a pseudonym replacement method based on mixed zones. Mixed zones refer to the geographical areas where pseudonyms are replaced. The scheme coordinates multiple users to synchronously replace pseudonyms in a specific spatial area. Scholar Buttyán et al. first used the mix-zone method for vehicle networks in 2007. The scheme allocates a large number of pseudonyms to vehicles in advance. Vehicles replace pseudonyms in mixed zones and choose different paths to leave after the replacement is completed.

[0007] 2. k-anonymity technology combines the user's real location with K-1 fake query locations that are very similar to the user's location to form an anonymous set containing k locations. The user uses the anonymous set instead of the real location to initiate a query, so that the user's location information cannot be distinguished from the other k-1 users in the set. The LBS provider responds to the query and returns a list of services required for each query location. The user filters the query results according to his real location and obtains his own services. Generally, the larger the k value, the more effective the privacy protection.

[0008] The anonymous technology used in Internet of Vehicles service privacy, whether it is pseudonym technology or k-anonymity technology, allows users to obtain services by cutting off the association between identity and location, that is, when submitting location information, personal identity information is not leaked, so that other entities cannot know the user's real personal information. However, consider the following situation: Since the LBS server processes location information, when the amount of this location information is small, and the number of users who initiate access to the LBS service is also very small, in such an extreme case (called sparse user state), the user identity and user location can be matched only by observing the user's location information and performing some simple analysis, so the anonymity of the user identity will be poor. When the number of users in the application space is small, that is, the space is a sparse environment, attackers can discover the user's identity by analyzing the user's location information, thereby reducing the anonymity of the user identity, and even in extreme cases, the user's identity anonymity may be zero.

[0009] Obviously, in the most extreme case, if only one user initiates access to the LBS, even with the use of pseudonym technology, the anonymity of the user's personal identity information is 0. Therefore, fewer elements in the user identity information set or fewer elements in the location information set will lead to a reduction in the complexity of association mapping in a sparse user environment, resulting in reduced anonymity.

[0010] In the sparse network user environment, some scholars have proposed solutions to the problem of reduced privacy protection capabilities of anonymous technologies (pseudonym technology and k-anonymity technology). Representative ones include: Liu Heng proposed in the article "Research on Several Technologies for Privacy Protection Based on Location Services in Ubiquitous Computing Environments" that the correlation between user identity information and user location in sparse environments will increase, and the anonymity will be significantly reduced, and pointed out that this problem can be solved by adding virtual users. In the article "Privacy Protection Method Based on Location Services in Sparse Environments", Wu Xu et al. simulated and generated real user location information and its distribution based on the time delay characteristics of the spatial region to ensure the authenticity of virtual information. By adding virtual user information or location information, the complexity of association mapping is increased to achieve user privacy protection. However, the above solutions focus on the definition of sparse environments and the analysis of the correlation between virtual users and locations. No specific solutions are given for the specific generation methods of virtual users and virtual locations. At the same time, the above solutions only consider the location privacy problem in the sparse environment of general mobile networks, and do not design effective solutions for the specific architecture of the Internet of Vehicles. Summary of the invention

[0011] The purpose of the present invention is to provide an LBS privacy protection method in a sparse user environment of a vehicle network, and designs an LBS privacy protection algorithm in a sparse user environment of a vehicle network. The algorithm uses the homomorphic encryption property to dynamically generate vehicle pseudonyms to achieve anonymous authentication, and defines a new virtual address generation algorithm to ensure the non-clustering of virtual addresses. The present invention uses a mixture of multiple pseudonyms and K-anonymity technology to achieve the indistinguishability of pseudonyms from the data source, greatly reducing the correlation between the real identity and location of users in a sparse user environment, and cutting off the possibility of user privacy leakage.

[0012] The inventive concept of the present invention is: the present invention proposes an LBS privacy protection method in a sparse user environment of a vehicle network, which comprehensively uses pseudonym technology and k-anonymity technology, but does not need to generate batches of pseudonyms in advance for use. Instead, an algorithm for dynamically generating virtual users and virtual addresses using a homomorphic encryption mechanism in a sparse user environment is designed. The method uses the properties of homomorphic encryption to complete the legitimacy verification of virtual users, realizes the indistinguishability of pseudonyms from the data source, increases the complexity of mapping users and locations, confuses real location information, cuts off the possibility of user privacy leakage, and efficiently solves the security problem of excessive correlation between vehicle location information and identity information and reduced anonymity in a sparse user environment.

[0013] In order to achieve the above-mentioned invention object, the present invention is implemented by the following technical solution: a LBS privacy protection method in a sparse user environment of an Internet of Vehicles, comprising the following steps:

[0014] S1, system initialization: complete system parameter initialization, RSU initialization and vehicle offline registration;

[0015] S2, hybrid request set generation: In a sparse user environment, the vehicle generates a sufficient number of virtual users to form an LBS hybrid request set and send it to the RSU;

[0016] S3. LBS service privacy protection: RSU forwards the LBS request of the legitimate virtual user and returns the service result. The vehicle filters out the LBS results that match its real location.

[0017] Furthermore, the S1 step includes:

[0018] S11, system parameter initialization. 1) Initialization of homomorphic encryption algorithm: TA runs the algorithms paillier.Setup(.) and paillier.KeyGen(.) to generate the public and private key pair PH = {n, g}, SH = λ used by the homomorphic encryption algorithm; 2) Initialization of TA public and private keys: In order to generate the TA public and private key pair, select the RSA algorithm. The initialization process is as follows: Select large prime numbers p and q, and calculate: m = p × q, φ(m) = (p-1)(q-1). Select a random number e that satisfies that e and φ(m) are relatively prime; calculate d = e -1 modφ(m). The private key is SK TA ={d,m}, public key is PK TA = {e, m}, this key pair is used to ensure the confidentiality and integrity of data; 3) TA publishes system parameters {PK TA ,PH}, where the homomorphic public key PH is used to encrypt the real identity of the vehicle, and the public key PK TA Used to verify the legitimacy of the certificate issued by TA.

[0019] S12, RSU initialization. TA uses the RSA algorithm to assign a public-private key pair to the RSU: select a large prime number p R ,q R , calculate n R =p R ×q R ,φ(n R )=(p R -1)(q R -1). Choose the same as φ(n R ) Coprime random numbers e R , and calculate d R =e R -1 modφ(n R ), get the RSU public-private key pair: PK R ={e R ,n R}, SK R ={d R ,n R}. TA assigns a unique ID to RSUR , issue a certificate Cert for RSU R , the certificate contains the ID R ,PK R , timestamp, CA signature and other information;

[0020] S13, TA according to vehicle V i The submitted information (license plate number, driver's license, etc.) is used to assign an identity to the vehicle. To protect the privacy of the vehicle's true identity, the identity is homomorphically encrypted using the paillier.Enc(.) algorithm:

[0021] S131, TA is a legal vehicle V i Generate a unique identifier Calculate the hash value of the identity Create an identity data table with As a foreign key, establish a mapping relationship between it and the unique identity identifier and timestamp; at the same time, establish a foreign key A vehicle information table corresponding to the vehicle information (license plate number, driver's license number, etc.) to enable the identity of malicious vehicles to be traced;

[0022] S132, TA uses the paillier.Enc(.) encryption algorithm to generate vehicle V i Encrypted identity

[0023]

[0024] The encrypted identity information generated Stored in the encrypted identity table. Fields and Encrypted Identities Then, TA stores the random number r1 used for encryption in the vehicle's tamper-proof memory.

[0025] Furthermore, the step S2 specifically includes the following steps:

[0026] S21, RSU i Broadcast Hello message within its coverage area:

[0027] The Hello message contains:

[0028] ① Indicates that TA is RSU i Certificate issued;

[0029] ②TS is the current timestamp;

[0030] ③ For RSU iDigitally sign the message using your own private key;

[0031] S22, Vehicle Verification RSU i identity:

[0032] Vehicle V i Receive RSU i Broadcast Hello message to verify RSU i Legitimacy of identity: Vehicle V i First, check the timeliness of the Hello packet. If the time difference |T-TS|≤Δt is verified, decrypt the signature field.

[0033]

[0034] If the decryption result is consistent with the hash value of the data in the Hello message, it means that the Hello message is not forged. Then, the TA's public key is used to verify the RSU i Certificate, obtain RSU from the certificate i 's public key;

[0035] S23, sparse environment detection

[0036] Vehicle V i Completion of RSU i After the authentication, the vehicle triggers the sparse environment detection algorithm. Assume that V i There are N vehicles within the communication coverage area, V i The privacy protection requirement is K. Input parameter K into the sparse environment detection algorithm:

[0037] 1) Determine the spatial factor A. The vehicle obtains the two-dimensional coordinates of its current location through GPS: u =(l ux ,l uy ). With this point as the center, the maximum distance r that a vehicle can reach within a unit time Δt max is the radius, forming the largest anonymous area A max , A max Generate a region for the virtual address, A∈A max .

[0038] 2) Determine the number of virtual users to be generated. max ) has a total of N vehicle users, and the privacy factor K represents the privacy requirement of the vehicle requesting the service. By comparing the parameters K and N, it is determined whether it is necessary to generate virtual users and the number of virtual users to be generated.

[0039] ① When K≤N, it indicates that the current number of existing users can meet the privacy protection requirements. There is no need to generate virtual users, and the k-anonymity algorithm can be directly executed.

[0040] When K>N, it indicates that the vehicle is in a sparse user environment and needs to generate P=KN virtual users, and S24 is executed.

[0041] S24. Vehicle pseudonym generation

[0042] The vehicle first selects a set of random numbers for generating a pseudonym

[0043]

[0044] Where P = KN. Using homomorphic encryption public key PH and parameter n, Random numbers in Execute the homomorphic encryption paillier.Enc(.) algorithm:

[0045]

[0046] And the random number set r used for this encryption 2j Stored in tamper-proof memory in the vehicle's internal memory.

[0047] The vehicle executes the paillier.Enc(.) algorithm, r1、r 2j As input, compute the pseudonym:

[0048]

[0049] Vehicle pair encryption results Perform hash operation to get pseudonym

[0050] Use the same method to generate the other P-1 pseudonyms. During the pseudonym generation process, the encrypted A set of random numbers {r 21 ,...,r 2p} is stored in the vehicle's tamper-proof memory to obtain a pseudonym set

[0051] S25. Virtual address generation

[0052] The vehicle calls the virtual address generation algorithm and inputs the number of virtual addresses to be generated P-1,r min ,r max , is V i Generate P-1 virtual addresses and store them in the virtual address set The specific process is as follows:

[0053] S251. According to the sparse user environment judgment algorithm, it is known that the vehicle user is in a sparse environment and P-1=K-N-1 virtual positions need to be generated.

[0054] S252, with the vehicle coordinates loc u =(l ux ,l uy ) is the center of the circle, and the maximum distance r that a vehicle can reach within a unit time Δt max is the radius, constituting the maximum range A max Generates a region for a virtual address.

[0055] S253, in order to evenly distribute P virtual positions in A max In the area, select the minimum radius r min , forming the minimum anonymous space A min , A max The area is divided into A min and A min ∩A max Two parts. Set the distribution coefficient d so that the generated virtual addresses are distributed in A min and annular area A min ∩A max middle.

[0056] If the i-th position satisfies

[0057] i%d==0

[0058] Then execute S254, in area A of the area min Generate a virtual position in the annular area A; otherwise, execute S255. min ∩A max Generate a virtual address in

[0059] S254, in A min Generate virtual positions internally and calculate:

[0060] θ=2π×random(0,1)

[0061] x=r min ·radom(0,1)·cosθ

[0062] y=r min ·radom(0,1)·sinθ

[0063] Where random(0,1) represents a random number selected at (0,1) to ensure that the coordinate points x and y are within the minimum radius r min In the circle formed, the virtual address loc is obtained i (x,y). Calculate

[0064]

[0065] If the result satisfies 0<distance(loc i ,locu )<r min , it means that the virtual address is within a reasonable range and is stored in the virtual address set PseuLoc.

[0066] S255. Generate a virtual address in the ring area and calculate:

[0067] a=r min / r max

[0068] x=r max ·radom(0,a)·cosθ

[0069] y=r max ·radom(0,a)·sinθ

[0070] where r max ·random(0,a) ensures that points x and y are in A min ∩A max The virtual address loc is formed in the ring i (x,y). Calculate

[0071]

[0072] If the result satisfies the condition r min ≤distance(loc i ,loc u )≤r max , it indicates that the virtual address is within the circular range and is stored in the virtual address set PseuLoc.

[0073] S26. Generation of LBS mixed request set

[0074] The vehicle confuses its real address with the P-1 virtual addresses generated in the previous step: In the set, randomly select the jth (1≤j≤P) pseudonymous element and construct it with the real address. Binding, forming LBS request

[0075]

[0076] The identity authentication request

[0077] ① The hash value of the vehicle's unique identity, used to query the vehicle's encrypted identity;

[0078] ②r1 is the random number returned when the unique identity of the vehicle is encrypted using homomorphic encryption;

[0079] ③ Indicates the pseudonym used to generate The random number to use.

[0080] ④r 2j Represents homomorphic encryption The random number returned when

[0081] ⑤ represents the jth pseudonymous element generated using the homomorphic encryption algorithm;

[0082] ⑥ Indicates the real address of the vehicle;

[0083] ⑦Search LBS Indicates the location service query content requested by the vehicle.

[0084] The vehicle generates a symmetric session key for this pseudonym Encrypt the above fields to form an LBS request To ensure the confidentiality of the message. At the same time, the same method is used to bind the remaining P-1 pseudonyms and the generated address to form an LBS mixed request set. The vehicle needs to record the pseudonym corresponding to the real address In order to finally filter out the LBS results that you really need;

[0085] S27,

[0086] The vehicle sends an LBS request to the RSU and session key Timestamp TS. Session key and timestamp TS using RSU i Public key encryption. Encrypted by a symmetric key. The vehicle uses the same method to send another P-1 LBS requests to the RSU.

[0087] Furthermore, the step S3 is:

[0088] S31, RSU i Get LBS request content

[0089] RSU i After receiving the request message, use the private key to decrypt the message:

[0090]

[0091] RSU i Get the temporary session key, timestamp, and check the validity of the timestamp. i Using Session Keys Decryption Message, get RSU i Transmitted to TA via wired media Apply for legality certification of the pseudonym;

[0092] S32. Vehicle pseudonym legitimacy verification

[0093] TA uses the private key to decrypt get According to the value of , the encrypted identity tag is retrieved to obtain the encrypted identity of the vehicle Call the pseudonym authentication algorithm to verify the pseudonym:

[0094] Enter a random number set Elements in r1、r 2j

[0095] ① Execute the paillier.Enc(.) algorithm to get the random number set Nonce pseu Perform homomorphic encryption:

[0096]

[0097] ②Query the encrypted identity obtained With the encryption result Perform homomorphic multiplication operations and take hash values ​​respectively:

[0098]

[0099] The calculation results and For comparison:

[0100] a) If the two are equal, according to the properties of homomorphic encryption, and because the random numbers r1, r 2i Vehicle V only i Knowing, therefore vehicle V i The vehicle is a legitimate one and the pseudonym authentication is successful. i Returns the result of successful authentication.

[0101] b) If the results are not equal, TA returns the authentication failure result, RSU i Discard the LBS request packet and send a i Return failure message:

[0102]

[0103] Among them, "ERROR" indicates that the pseudonym authentication failed; TS is the timestamp, which is encrypted with a shared key. After the vehicle receives the feedback message, it decrypts it with the session key and needs to send the request packet again to request the location service again;

[0104] S33, RSUi Forwarding LBS requests

[0105] RSU i After receiving the authentication success message, the mapping relationship between the vehicle symmetric key and the pseudonym and address is cached. From the mapping relationship, it can be seen that for the return value of the TA verification result, the RSU i We only know that the pseudonym and address come from a legitimate user, but we cannot associate the pseudonym with the real identity of the vehicle. This creates a scenario where multiple different users request LBS at the same time, which confuses users and addresses and achieves a double anonymity effect for LBS. At the same time, to further weaken the relevance of the pseudonym, a timestamp is generated for each group of connections in the table to invalidate the pseudonym timing when it times out.

[0106] After confusion, RSU i It is impossible to distinguish between real and virtual addresses in the request set, and forward all service requests to the LBS through the wired medium. This request only contains two fields: address and service query content, namely Sensitive identity information is hidden. Because the current environment is sparsely populated, the increase in service requests to achieve identity and address confusion will not lead to excessive use of communication resources;

[0107] S34,

[0108] LBS receives the service request set, obtains the corresponding LBS query content, makes the corresponding service response, and returns the service result RSU i According to the address information in the service result, the corresponding pseudonym and symmetric key are found through the mapping relationship between the key, pseudonym and address to form the service result:

[0109]

[0110] Encrypted with the symmetric key and returned to the vehicle V i ;

[0111] S35: Vehicle obtains LBS service

[0112] Vehicle decryption received RSU i The returned LBS service result set verifies the timeliness and integrity of the message, excludes the confusing pseudonyms, and selects the real pseudonyms Get the LBS service result corresponding to the real address and discard the rest of the returned results.

[0113] Compared with the prior art, the present invention has the following beneficial effects:

[0114] (1) The present invention provides a LBS privacy protection method in a sparse user environment of the Internet of Vehicles. It innovatively designs a virtual address generation algorithm, which aims to generate a set of virtual addresses in a sparse user environment and confuse the virtual addresses with the real addresses of the vehicles to achieve location privacy protection. Compared with other schemes, the distribution of virtual user locations in this method is more uniform, and there are fewer open areas, which is more conducive to the confusion of vehicle locations. At the same time, when the privacy demand is high, the attacker can hardly narrow the locking range of the real location of the vehicle according to the location distribution and obtain the location privacy information of the vehicle.

[0115] (2) The present invention provides a privacy protection method for LBS in a sparse user environment of the Internet of Vehicles, which creatively uses a homomorphic encryption algorithm to realize the generation of a vehicle pseudonym set, legitimacy authentication, and identity tracing. The homomorphic encryption algorithm has the same property as the result obtained by directly performing a specific algebraic operation on the ciphertext, which is the same as the result of performing the same operation on the plaintext and then encrypting the result. At the same time, since each authentication of the vehicle uses a temporarily generated pseudonym and generates unique random numbers r1 and r2, the random numbers are only known to the vehicle itself, and anonymous pseudonym authentication can be completed. In addition, the vehicle identity authentication information is encrypted using the TA's public key, and the identity information contains a random number. The encryption result is random and can only be decrypted by the TA. Therefore, even the RSU cannot determine whether the multiple pseudonyms in the LBS request are from the same vehicle, which effectively avoids the correlation between the vehicle's true identity and the LBS request service.

[0116] (3) The present invention provides a privacy protection method for LBS in a sparse user environment of the Internet of Vehicles, and innovatively designs a sparse user environment judgment algorithm. The algorithm combines the privacy factor and the space factor, and constrains the judgment value from the two aspects of the number of users and space, respectively, solving the problem that it is difficult to determine the judgment standard of the sparse user environment and the number of generated virtual users. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0118] Figure 1 This is a flow chart of the LBS privacy protection method in a sparse user environment of an Internet of Vehicles according to an embodiment of the present invention;

[0119] Figure 2 2 is a system architecture diagram in an embodiment of the present invention.

[0120] Figure 3 A flow chart for generating a hybrid request set in an embodiment of the present invention.

[0121] Figure 4This is a diagram of a sparse environment analysis algorithm in an embodiment of the present invention.

[0122] Figure 5 The figure is a diagram of a virtual address generation algorithm in an embodiment of the present invention.

[0123] Figure 6 This is a diagram of a pseudonym dynamic generation and authentication algorithm in an embodiment of the present invention.

[0124] Figure 7 4 is a diagram showing the virtual address distribution effect in an embodiment of the present invention.

[0125] Figure 8 This is a comparison diagram of the virtual position generation effect in an embodiment of the present invention.

[0126] Fig. 9 4 is a comparison diagram of the uniformity of virtual position distribution in an embodiment of the present invention. DETAILED DESCRIPTION

[0127] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. Of course, the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0128] Example

[0129] This embodiment provides a method for protecting LBS privacy in a sparse user environment in an Internet of Vehicles. Figure 1 As shown, the following steps are included:

[0130] S1, system initialization: complete system parameter initialization, RSU initialization and vehicle offline registration;

[0131] S2, hybrid request set generation: In a sparse user environment, the vehicle generates a sufficient number of virtual users to form an LBS hybrid request set and send it to the RSU;

[0132] S3. LBS service privacy protection: RSU forwards the LBS request of the legitimate virtual user and returns the service result. The vehicle filters out the LBS results that match its real location.

[0133] like Figure 2 ,The system architecture of the LBS privacy protection method in the sparse user environment of the Internet of Vehicles includes:

[0134] 1) Trust Authentication (TA): The Trust Authentication Center is an authoritative organization in the Internet of Vehicles system. It is responsible for managing all nodes connected to the Internet of Vehicles, distributing public and private keys to nodes, and issuing certificates.

[0135] 2) Location Based Server (LBS): The LBS server is responsible for providing location-based services for vehicles. It is wired to the roadside unit and receives and responds to vehicle location service requests forwarded by the roadside unit.

[0136] 3) Road Side Unit (RSU): RSU is evenly distributed on both sides of the road. It is both a relay node for vehicle message distribution and a message receiving node. With a certain computing power, it can complete the identity authentication of the vehicle with the cooperation of TA. At the same time, it forwards LBS requests and results as a relay node.

[0137] 4) On Broad Unit (OBU): OBU is embedded inside the vehicle and is responsible for communicating with external entities and storing basic vehicle information (public and private keys, certificates, etc.). OBU is both the sender of messages and the receiver and forwarder of messages, and is extremely vulnerable to network security threats such as masquerade attacks and Sybil attacks.

[0138] The S1 step includes:

[0139] S11, system parameter initialization. 1) Initialization of homomorphic encryption algorithm: TA runs the algorithms paillier.Setup(.) and paillier.KeyGen(.) to generate the public and private key pair PH = {n, g}, SH = λ used by the homomorphic encryption algorithm; 2) Initialization of TA public and private keys: In order to generate the TA public and private key pair, select the RSA algorithm. The initialization process is as follows: Select large prime numbers p and q, and calculate: m = p × q, φ(m) = (p-1)(q-1). Select a random number e that satisfies that e and φ(m) are relatively prime; calculate d = e -1 modφ(m). The private key is SK TA ={d,m}, public key is PK TA = {e, m}, this key pair is used to ensure the confidentiality and integrity of data; 3) TA publishes system parameters {PK TA ,PH}, where the homomorphic public key PH is used to encrypt the real identity of the vehicle, and the public key PK TA Used to verify the legitimacy of the certificate issued by TA.

[0140] S12, RSU initialization. TA uses the RSA algorithm to assign a public-private key pair to the RSU: select a large prime number p R ,q R , calculate n R =p R ×q R ,φ(n R )=(p R -1)(q R-1). Choose the same as φ(n R ) Coprime random numbers e R , and calculate d R =e R -1 modφ(n R ), get the RSU public and private key pair: PK R ={e R ,n R}, SK R ={d R ,n R}. TA assigns a unique ID to RSU R , issue a certificate Cert for RSU R , the certificate contains the ID R ,PK R , timestamp, CA signature and other information;

[0141] S13, vehicle offline registration. TA according to the vehicle V i The submitted information (license plate number, driver's license, etc.) is used to assign an identity to the vehicle. To protect the privacy of the vehicle's true identity, the identity is homomorphically encrypted using the paillier.Enc(.) algorithm:

[0142] S131, TA is a legal vehicle V i Generate a unique identifier: Calculate the hash value of the identity Establish the structure as shown in Table 1 identity data table, the table is As a foreign key, establish a mapping relationship between it and the unique identity identifier and timestamp; at the same time, establish a foreign key The vehicle information table corresponding to the vehicle information (license plate number, driver's license number, etc.) has a structure as shown in Table 2 to achieve identity tracing of malicious vehicles;

[0143] Table 1 Identity data table

[0144]

[0145] Table 2 Vehicle information table

[0146]

[0147] S132, TA uses the paillier.Enc(.) encryption algorithm to generate vehicle V i Encrypted identity

[0148]

[0149] The encrypted identity information generated Stored in the encrypted identity table, the structure is shown in Table 3. Fields and Encrypted Identities Then, TA stores the random number r1=2383985447 used for encryption in the vehicle's tamper-proof memory.

[0150] Table 3 Encrypted identity table

[0151]

[0152] like Figure 3 The specific contents of step S2 include the following steps:

[0153] S21, RSU i Broadcast Hello message within its coverage area:

[0154] The Hello message contains:

[0155] ④ Indicates that TA is RSU i Certificate issued;

[0156] ⑤TS is the current timestamp;

[0157] ⑥ For RSU i Digitally sign the message using your own private key;

[0158] S22, Vehicle Verification RSU i identity:

[0159] Vehicle V i Receive RSU i Broadcast Hello message to verify RSU i Legitimacy of identity: Vehicle V i First, check the timeliness of the Hello packet. If the time difference |T-TS|≤Δt is verified, decrypt the signature field.

[0160]

[0161] If the decryption result is consistent with the hash value of the data in the Hello message, it means that the Hello message is not forged. Then, the TA's public key is used to verify the RSU i Certificate, obtain RSU from the certificate i 's public key;

[0162] S23, sparse environment detection

[0163] Vehicle V i Completion of RSU iAfter authentication, the vehicle triggers the sparse environment detection algorithm, which is as follows Figure 4 Assume that V i There are N = 80 vehicles within the communication coverage area, V i The privacy protection requirement is K = 120. Input parameter K into the sparse environment detection algorithm:

[0164] 1) Determine the spatial factor A. The vehicle obtains the two-dimensional coordinates of its current location through GPS: u =(l ux ,l uy ). With this point as the center, the maximum distance r that a vehicle can reach within a unit time Δt max is the radius, forming the largest anonymous area A max , A max Generate a region for the virtual address, A∈A max .

[0165] 2) Determine the number of virtual users to be generated. When K>N, it indicates that the vehicle is in a sparse user environment and needs to generate P=K–N=40 virtual users.

[0166] S24. Vehicle pseudonym generation

[0167] The vehicle first selects a set of random numbers for generating a pseudonym

[0168]

[0169] Where P = KN. Using homomorphic encryption public key PH and parameter n, Random numbers in Execute the homomorphic encryption paillier.Enc(.) algorithm:

[0170]

[0171] And the random number set r used for this encryption 2j Stored in tamper-proof memory in the vehicle's internal memory.

[0172] The vehicle executes the paillier.Enc(.) algorithm, r1、r 2j As input, compute the pseudonym:

[0173]

[0174] Vehicle pair encryption results Perform hash operation to get pseudonym

[0175] Use the same method to generate the other P-1 pseudonyms. During the pseudonym generation process, the encrypted A set of random numbers {r 21 ,...,r 2p} is stored in the vehicle's tamper-proof memory to obtain a pseudonym set

[0176] S25. Virtual address generation

[0177] The vehicle calls the virtual address generation algorithm and inputs the number of virtual addresses to be generated P-1,r min ,r max , is V i Generate P-1 virtual addresses and store them in the virtual address set Algorithms such as Figure 5 As shown, the specific process is as follows:

[0178] S251. According to the sparse user environment judgment algorithm, it is known that the vehicle user is in a sparse environment and P-1=K-N-1 virtual positions need to be generated.

[0179] S252, with the vehicle coordinates loc u =(l ux ,l uy ) is the center of the circle, and the maximum distance r that a vehicle can reach within a unit time Δt max is the radius, constituting the maximum range A max Generates a region for a virtual address.

[0180] S253, in order to evenly distribute P virtual positions in A max In the area, select the minimum radius r min , forming the minimum anonymous space A min , A max The area is divided into A min and A min ∩A max Two parts. Set the distribution coefficient d so that the generated virtual addresses are distributed in A min and annular area A min ∩A max middle.

[0181] If the i-th position satisfies

[0182] i%d==0

[0183] Then execute S254, in area A of the area min Generate a virtual position in the annular area A; otherwise, execute S255. min ∩A max Generate a virtual address in

[0184] S254, in A min Generate virtual positions internally and calculate:

[0185] θ=2π×random(0,1)

[0186] x=r min ·radom(0,1)·cosθ

[0187] y=r min ·radom(0,1)·sinθ

[0188] Where random(0,1) represents a random number selected at (0,1) to ensure that the coordinate points x and y are within the minimum radius r min In the circle formed, the virtual address loc is obtained i (x,y). Calculate

[0189]

[0190] If the result satisfies 0<distance(loc i ,loc u )<r min , it means that the virtual address is within a reasonable range and is stored in the virtual address set PseuLoc.

[0191] S255. Generate a virtual address in the ring area and calculate:

[0192] a=r min / r max

[0193] x=r max ·radom(0,a)·cosθ

[0194] y=r max ·radom(0,a)·sinθ

[0195] where r max ·random(0,a) ensures that points x and y are in A min ∩A max The virtual address loc is formed in the ring i (x,y). Calculate

[0196]

[0197] If the result satisfies the condition r min ≤distance(loc i ,loc u )≤r max , it indicates that the virtual address is within the circular range and is stored in the virtual address set PseuLoc.

[0198] S26. Generation of LBS mixed request set

[0199] The vehicle confuses its real address with the P-1 virtual addresses generated in the previous step: In the set, randomly select the jth (1≤j≤P) pseudonymous element and construct it with the real address. Binding, forming LBS request

[0200]

[0201] The identity authentication request

[0202] ① The hash value of the vehicle's unique identity, used to query the vehicle's encrypted identity;

[0203] ②r1 is the random number returned when the unique identity of the vehicle is encrypted using homomorphic encryption;

[0204] ③ Indicates the pseudonym used to generate The random number to use.

[0205] ④r 2j Represents homomorphic encryption The random number returned when

[0206] ⑤ represents the jth pseudonymous element generated using the homomorphic encryption algorithm;

[0207] ⑥ Indicates the real address of the vehicle;

[0208] ⑦Search LBS Indicates the location service query content requested by the vehicle.

[0209] The vehicle generates a symmetric session key for this pseudonym Encrypt the above fields to form an LBS request To ensure the confidentiality of the message. At the same time, the same method is used to bind the remaining P-1 pseudonyms and the generated address to form the LBS mixed request set shown in Table 4. The vehicle needs to record the pseudonym corresponding to the real address In order to finally filter out the LBS results that you really need;

[0210] Table 4 LBS mixed request set

[0211]

[0212] S27,

[0213] The vehicle sends an LBS request to the RSU and session key Timestamp TS. Session key and timestamp TS using RSU i Public key encryption. Encrypted by a symmetric key. The vehicle uses the same method to send another P-1 LBS requests to the RSU.

[0214] The specific contents of step S3 include the following steps:

[0215] S31, RSU i Get LBS request content

[0216] RSU i After receiving the request message, use the private key to decrypt the message:

[0217]

[0218] RSU i Get the temporary session key, timestamp, and check the validity of the timestamp. i Using Session Keys Decryption Message, get RSU i Transmitted to TA via wired media Apply for legality certification of the pseudonym;

[0219] S32. Vehicle pseudonym legitimacy verification

[0220] TA uses the private key to decrypt get According to the value of , the encrypted identity tag is retrieved to obtain the encrypted identity of the vehicle Call the pseudonym authentication algorithm, such as Figure 6 As shown, the pseudonym is verified:

[0221] Enter a random number set Elements in r1、r 2j

[0222] ① Execute the paillier.Enc(.) algorithm to get the random number set Nonce pseu Perform homomorphic encryption:

[0223]

[0224] ②Query the encrypted identity obtained With the encryption result Perform homomorphic multiplication operations and take hash values ​​respectively:

[0225]

[0226] The calculation results and For comparison:

[0227] a) If the two are equal, according to the properties of homomorphic encryption, and because the random numbers r1, r 2i Vehicle V only i Knowing that, vehicle V i The vehicle is a legitimate one and the pseudonym authentication is successful. i Returns the result of successful authentication.

[0228] b) If the results are not equal, TA returns the authentication failure result, RSU i Discard the LBS request packet and send a i Return failure message:

[0229]

[0230] Among them, "ERROR" indicates that the pseudonym authentication failed; TS is the timestamp, which is encrypted with a shared key. After the vehicle receives the feedback message, it decrypts it with the session key and needs to send the request packet again to request the location service again;

[0231] S33, RSU i Forwarding LBS requests

[0232] RSU i After receiving the authentication success message, the mapping relationship between the vehicle symmetric key and the pseudonym and address is cached. The structure is shown in Table 5.

[0233] Table 5 Symmetric key and pseudonym, address mapping table

[0234]

[0235] From the mapping relationship, we can see that for the return value of the TA verification result, RSU i We only know that the pseudonym and address come from a legitimate user, but we cannot associate the pseudonym with the real identity of the vehicle. This creates a scenario where multiple different users request LBS at the same time, which confuses users and addresses and achieves a double anonymity effect for LBS. At the same time, to further weaken the relevance of the pseudonym, a timestamp is generated for each group of connections in the table to invalidate the pseudonym timing when it times out.

[0236] After confusion, RSU i It is impossible to distinguish between real and virtual addresses in the request set, and forward all service requests to the LBS through the wired medium. This request only contains two fields: address and service query content, namely Sensitive identity information is hidden. Because the current environment is sparsely populated, the increase in service requests to achieve identity and address confusion will not lead to excessive use of communication resources;

[0237] S34,

[0238] LBS receives the service request set, obtains the corresponding LBS query content, makes the corresponding service response, and returns the service result RSU i According to the address information in the service result, the corresponding pseudonym and symmetric key are found through the mapping relationship between the key, pseudonym and address to form the service result:

[0239]

[0240] Encrypted with the symmetric key and returned to the vehicle V i ;

[0241] S35: Vehicle obtains LBS service

[0242] Vehicle decryption received RSU i The returned LBS service result set verifies the timeliness and integrity of the message, excludes the confusing pseudonyms, and selects the real pseudonyms Get the LBS service result corresponding to the real address and discard the rest of the returned results.

[0243] In order to verify the feasibility of this embodiment, the correctness and feasibility of the method of the present invention are analyzed.

[0244] 1. Security Analysis

[0245] 1) Analysis of the effectiveness of service privacy protection

[0246] The present invention utilizes The real identity is encrypted, and the pseudonym update and the real identity are linked through the homomorphic encryption mechanism, so that the pseudonym can be used for identity authentication by the vehicle, preventing the leakage of the real identity. At the same time, each LBS service request of the vehicle temporarily generates a set of pseudonyms, and generates a unique random number r1 and a set {r 21 ,...,r 2p}, making the authentication message random, even if the attacker has the ability to collect and guess information, it is impossible to link one car to multiple authentication messages.

[0247] In addition, in the LBS request uploaded by the vehicle, the elements required to verify the pseudonym are encrypted using the TA's public key to ensure that only the TA can know the vehicle's true identity information. At the same time, the vehicle confuses its real address with the virtual address. The above measures make it impossible for the RSU to locate the vehicle that generates the pseudonym, nor to identify the real address used by the vehicle in the LBS request. It only plays a relay role, further improving the privacy of the vehicle's location. In summary, the RSU and the LBS server cannot associate the vehicle's true identity and true location information, thus achieving location service privacy protection in sparse environments.

[0248] 2) Malicious vehicle identity traceability analysis

[0249] To trace malicious vehicles, the trusted center stores the mapping between the real identity and the encrypted identity of the vehicle. TA is calculated based on the encrypted identity Enc of the malicious vehicle using the Paillier decryption algorithm:

[0250]

[0251] TA uses homomorphically encrypted private keys to restore the real identity of malicious vehicles, meeting the traceability of pseudonym management. At the same time, TA does not need to save the mapping between each pseudonym and the real identity of the vehicle, reducing pseudonym management overhead and privacy leakage risks.

[0252] 3) Analysis of anti-denial of service attacks

[0253] In the present invention, legitimate vehicles generate multiple pseudonyms and false addresses to confuse their real identities and addresses, but attackers cannot use this technology to implement denial of service attacks. Because any pseudonyms generated by legitimate vehicles need to be verified by TA before location-based services can be obtained. The present invention uses TA public key to encrypt random numbers and other information to ensure the security of vehicle authentication. Due to the problem of discrete logarithms and large prime number factorization, attackers cannot obtain TA's private key, and thus cannot disguise multiple legitimate vehicle pseudonyms to carry out denial of service attacks, as shown below:

[0254] ① Discrete logarithm problem: Take a prime number p, and an integer a is the primitive root of p. Its power can generate all integers between 1 and p-1. Then a mod p, a 2 mod p,…,a p-1 mod p is different. For any integer b and prime number p primitive root a, only a unique exponent i can be found to obtain

[0255] b≡a i modp, where 0≤i≤(p-1)

[0256] It is easy to raise prime numbers to powers, but it is very difficult to calculate discrete logarithms; for large prime numbers, calculating discrete logarithms is considered infeasible.

[0257] ②Difficulty in factoring large prime numbers: In the RSA algorithm, m=p·q, the lengths of p and q are both more than 1024 bits, and the length of m is more than 2048 bits. Currently, there is no efficient way to factorize large integers, and it is extremely difficult to crack them using brute force alone.

[0258] 4) Analysis of the irrelevance of pseudonyms

[0259] The core issue of vehicle service privacy protection in a sparse user environment is the irrelevance of pseudonyms and their locations, to ensure that neither the RSU nor the attacker can associate the vehicle's real identity and location information. In the LBS request message designed by the present invention, the vehicle does not directly send real identity information for authentication, but sends challenge information: the hash value of the encrypted identity and the encrypted random number, and the challenge information is encrypted with the TA public key. Due to the existence of the encrypted random number in the challenge information, even if it comes from different pseudonyms of the same vehicle, the encrypted challenge information is different. This ensures that the RSU cannot distinguish the sources of different pseudonyms, and only the TA can verify the legitimacy of the vehicle pseudonym.

[0260] When TA authenticates the pseudonym, it does not need to know the real identity information of the vehicle. As long as the challenge information is homomorphically calculated using the homomorphic encryption property, pseudonym authentication can be achieved. Therefore, what TA returns to RSU is only the validity of the pseudonym, and does not contain any other information related to the real identity of the vehicle. Neither the attacker nor the RSU can distinguish which vehicles a group of pseudonyms in the LBS service request process belongs to. At the same time, the vehicle pseudonyms stored by the RSU are attached with a timestamp, stipulating that the pseudonyms must be replaced after the expiration date, so as to further reduce the correlation between pseudonyms and protect the identity and trajectory privacy of the vehicle.

[0261] 2. Experimental Analysis

[0262] The present invention uses tools such as IntelliJ IDEA 2021.1 and PyCharm 2021.3.3 to simulate vehicle V i In a sparse user environment of 5km×5km, a virtual user pseudonym and corresponding virtual address are generated to protect the privacy of service requests. To prove the feasibility and advantages of the present invention, the virtual address generated by the algorithm is confused with the real address and compared with the anonymity effect of other schemes. The experimental results show that the pseudonyms generated by the algorithm of the present invention have less correlation, more uniform distribution of virtual addresses, and stronger anonymity. The experimental analysis is as follows:

[0263] 1) Pseudonym generation and authentication verification

[0264] Assume that the vehicle V i There are N = 80 vehicle users within the coverage area of ​​the RSU, V iSet the privacy factor K = 120. Since N < K, it is determined that V i is in a sparse user environment and needs to generate K - N = 40 virtual users. In the initialization phase, TA assigns a unique identity identifier i to vehicle V as 337767699453702274400866080810937938320, and uses the Paillier homomorphic encryption algorithm to encrypt the real identity ID:

[0265]

[0266] Calculate the encrypted identity Store the generated random number r1 = 2383985447.

[0267] In the pseudonym generation phase, the vehicle randomly selects a value pseu - nonce and calculates a new pseudonym:

[0268] E(ID·pseu - nonce, r1, r2) = (g ID+pseu-nonce ·(r1·r2) n ) mod n 2

[0269] Store the generated random number r2.

[0270] In the pseudonym authentication phase, TA inputs the encrypted identity E(ID, r1) and the encrypted random number E(pseu - nonce, r2), verifies whether the pseudonym conforms to the homomorphic multiplication property, and conducts authentication:

[0271] E(ID, r1)·E(pseu - nonce, r2)

[0272] =[(g ID ·r1 n ) mod n 2 ·[(g pseu-nonce ·r2 n ) mod n 2

[0273] In the pseudonym authentication process, the real vehicle identity ID is not directly used, only the encrypted identity values and the random numbers r1, r2 are used.

[0274] Table 6 lists the authentication results of 5 pseudonym elements in the pseudonym set. The values of the last two columns of data E(ID·pseu - nonce), E(ID)·E(pseu - nonce) are equal, indicating that according to the multiplication property of homomorphic encryption, the pseudonym authentication is successful.

[0275] Table 6 Pseudonym Authentication Results ​

[0276]

[0277] 2) Virtual address generation verification

[0278] Assume the real address coordinates of the vehicle are the center of the circle (0,0) and the maximum anonymous area radius is r max The minimum anonymous area radius is 500m. min When the distribution coefficient d=4, the virtual addresses can be evenly distributed in the largest anonymous area and the smallest anonymous area. According to the virtual address generation algorithm, the i-th virtual address is calculated:

[0279] First determine the offset angle of the virtual address relative to the center of the circle:

[0280] θ=2π×random(0,1)

[0281] If i%4 is 0, then in A min Generate a virtual address in the area and calculate the coordinates loc i (x,y):

[0282] x=r min ·radom(0,1)·cosθ

[0283] y=r min ·radom(0,1)·sinθ

[0284] If the virtual address loc i If the distance between (x, y) and the center coordinates does not exceed the minimum radius, the virtual location is generated within the specified range and stored in the virtual location set PseuLoc. Table 7 lists 5 groups of virtual addresses generated in the minimum anonymous area.

[0285] Table 7 Minimum anonymous area virtual address

[0286]

[0287]

[0288] If i%4 is not 0, then in A min ∩A max The virtual address is generated in the annular area. Calculate a to control the generated virtual address coordinates to be within the annular area:

[0289] a=r min / r max

[0290] Calculate the coordinates loc i (x,y):

[0291] x=r max·radom(0,a)·cosθ

[0292] y=r max ·radom(0,a)·sinθ

[0293] If the virtual address loc i The distance from (x,y) to the center of the circle is (r min ,r max ), the virtual address is generated successfully. Table 8 lists 5 groups of virtual addresses generated in the ring area.

[0294] Table 8 Ring area virtual address

[0295]

[0296] Finally, the vehicle V i The generated virtual address distribution effect is as follows Figure 7 shown.

[0297] 3) Comparison of privacy protection effects

[0298] The present invention is compared with the virtual position generation algorithm proposed by the existing scheme, such as the pseudo position generation algorithm. The results are as follows Figure 8 As shown, the distribution of the virtual user positions of the present invention is more uniform, and there are fewer open areas, which is more conducive to the confusion of vehicle positions.

[0299] At the same time, as the privacy factor K increases, the randomness of the virtual location distribution generated by this algorithm becomes more uniform. Fig. 9 As shown in Figure 2, the uniformity of virtual location distribution shows a gradually increasing trend. Therefore, when privacy requirements are high, attackers can hardly narrow the locking range of the vehicle's real location based on the location distribution and obtain the vehicle's location privacy information.

[0300] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A LBS privacy protection method in a sparse user environment of an Internet of Vehicles, characterized in that: The steps include: S1, system initialization, complete system parameter initialization, RSU initialization and vehicle offline registration. In this stage, TA generates relevant system parameters, uses RSA algorithm and Paillier algorithm to generate system public and private keys and homomorphic keys respectively, and publishes system parameters; registers the identity of RSU and issues a public key certificate; uses homomorphic algorithm to encrypt the real identity of the vehicle, stores the real identity information of the vehicle, so as to realize the identity tracing of malicious vehicles; S2, vehicle V i Drive into RSU i Within the coverage area, first receive RSU i The Hello message broadcast within its coverage area verifies the RSU through the certificate carried in the Hello message i Legitimacy of identity and obtaining RSUs i Public key, in preparation for subsequent secure data transmission. When requesting location services, the vehicle first triggers the sparse environment detection algorithm to determine whether it is in a sparse environment. If the vehicle is in a sparse environment, it needs to generate a group of virtual users. The virtual user includes a set of pseudonyms and their corresponding virtual addresses. The vehicle mixes the virtual user with the real user LBS request and sends it to the RSU. i , RSU i Submit the virtual user identity to TA for legitimacy verification and decide whether to forward the corresponding LBS request to the vehicle; The S2 step specifically includes the following steps: S21, RSU i Broadcast Hello message within its coverage area: The Hello message contains: ① Indicates that TA is RSU i Certificate issued; ②TS is the current timestamp; ③ For RSU i Digitally sign the message using your own private key; S22, Vehicle Verification RSU i identity Vehicle V i Receive RSU i Broadcast Hello message to verify RSU i Legitimacy of identity: Vehicle V i First, check the timeliness of the Hello packet. If the time difference |T-TS|≤Δt is verified, decrypt the signature field. If the decryption result is consistent with the hash value of the data in the Hello message, it means that the Hello message is not forged. Then, the TA's public key is used to verify the RSU i Certificate, obtain RSU from the certificate i 's public key; S23, sparse environment detection Vehicle V i Completion of RSU i After authentication, the vehicle triggers the sparse environment detection algorithm, assuming that V i There are N vehicles within the communication coverage area, V i The privacy protection requirement is K, and the parameter K is input into the sparse environment detection algorithm: 1) Determine the spatial factor A. The vehicle obtains the two-dimensional coordinates of its current location through GPS. u =(l ux ,l uy ), with this point as the center, and the maximum distance r of the vehicle per unit time Δt max is the radius, forming the largest anonymous area A max , A max Generate a region for the virtual address, A∈A max ; 2) Determine the number of virtual users to be generated, the current area A (A∈A max ) has a total of N vehicle users, and the privacy factor K represents the privacy requirements of the vehicle requesting the service. By comparing the parameters K and N, it is determined whether it is necessary to generate virtual users and the number of virtual users to be generated; ① When K≤N, it indicates that the current number of existing users meets the privacy protection requirements, and there is no need to generate virtual users, and the k-anonymity algorithm is directly executed; ② When K>N, it indicates that the vehicle is in a sparse user environment and needs to generate P=KN virtual users, and execute S24; S24. Vehicle pseudonym generation The vehicle first selects a set of random numbers for generating a pseudonym Where P = K – N, using the homomorphic encryption public key PH and parameter n, Random numbers in Execute the homomorphic encryption paillier.Enc(.) algorithm: And the random number set r used for this encryption 2j stored in tamper-proof memory in the vehicle's internal memory; The vehicle executes the paillier.Enc(.) algorithm, r1、r 2j As input, compute the pseudonym: Vehicle pair encryption results Perform hash operation to get pseudonym Use the same method to generate other P-1 pseudonyms. During the pseudonym generation process, the encrypted A set of random numbers {r 21 ,...,r 2p } is stored in the vehicle's tamper-proof memory to obtain a pseudonym set S25. Virtual address generation The vehicle calls the virtual address generation algorithm and inputs the number of virtual addresses to be generated P-1,r min ,r max , is V i Generate P-1 virtual addresses and store them in the virtual address set S26. LBS mixed request set generation The vehicle confuses its real address with the P-1 virtual addresses generated in the previous step: In the set, randomly select the jth (1≤j≤P) pseudonymous element and construct it with the real address. Binding, forming LBS request The identity authentication request ① The hash value of the vehicle's unique identity, used to query the vehicle's encrypted identity; ②r1 is the random number returned when the unique identity of the vehicle is encrypted using homomorphic encryption; ③ Indicates the pseudonym used to generate The random number used; ④r 2j Represents homomorphic encryption The random number returned when ⑤ represents the jth pseudonymous element generated using the homomorphic encryption algorithm; ⑥ Indicates the real address of the vehicle; ⑦Search LBS Indicates the location service query content requested by the vehicle; The vehicle generates a symmetric session key for this pseudonym Encrypt the above fields to form an LBS request To ensure the confidentiality of the message, the same method is used to bind the remaining P-1 pseudonyms and the generated address to form an LBS mixed request set. The vehicle needs to record the pseudonym corresponding to the real address. It is used to finally filter out the LBS results that you really need; S27、V i →RSU i : The vehicle sends an LBS request to the RSU and session key Timestamp TS, session key and timestamp TS using RSU i Public key encryption, Encrypted by a symmetric key, the vehicle uses the same method to send another P-1 LBS requests to the RSU; S3, during the LBS service privacy protection process, the vehicle V i RSUs to Agents i Send LBS request message, RSU i After receiving this request message, the identity authentication information Forwarded to TA to verify the legitimacy of the vehicle pseudonym generation and resist denial of service attacks by illegal vehicles. TA receives RSU i Forwarded Call the pseudonym authentication algorithm to verify the legitimacy of the vehicle pseudonym and return the authentication result. If the authentication is successful, RSU i The address and service type field values ​​are selected from the LBS mixed request set uploaded by the vehicle, and the service request is forwarded to the LBS server. The LBS server sends the service request result set to the RSU. i , RSU i The result set is forwarded to the vehicle, and the vehicle filters out the location service results it needs from the result set and discards other location service results.

2. The LBS privacy protection method in a sparse user environment of an Internet of Vehicles according to claim 1 is characterized in that: The step S1 specifically includes the following steps: S11, system parameter initialization 1) Initialization of homomorphic encryption algorithm TA runs the algorithms paillier.Setup(.) and paillier.KeyGen(.) to generate the public and private key pairs PH={n,g}, SH=λ used by the homomorphic encryption algorithm; 2) TA public and private key initialization In order to generate the TA public-private key pair, the RSA algorithm is selected. The initialization process is as follows: select large prime numbers p and q, calculate: m = p × q, φ(m) = (p-1)(q-1), select a random number e, satisfying that e and φ(m) are relatively prime; calculate d = e -1 modφ(m), the private key is SK TA ={d,m|, public key is PK TA ={e,m}, this key pair is used to ensure the confidentiality and integrity of data; 3)TA announces system parameters {PK TA ,PH}, where the homomorphic public key PH is used to encrypt the real identity of the vehicle, and the public key PK TA Used to verify the legitimacy of the certificate issued by TA; S12, RSU initialization TA uses the RSA algorithm to assign a public-private key pair to RSU: select a large prime number p R ,q R , calculate n R =p R ×q R ,φ(n R )=(p R -1)(q R -1), choose the same as φ(n R ) Coprime random numbers e R , and calculate d R =e R -1 modφ(n R ), get the RSU public and private key pair: PK R ={e R ,n R }, SK R ={d R ,n R }, TA assigns a unique ID to RSU R , issue a certificate Cert for RSU R , the certificate contains the ID R ,PK R , timestamp, CA signature information; S13, TA according to vehicle V i The submitted information is used to assign an identity to the vehicle. To protect the privacy of the vehicle's true identity, the identity is homomorphically encrypted using the paillier.Enc(.) algorithm.

3. The LBS privacy protection method in a sparse user environment of a vehicle network according to claim 2 is characterized in that: In step S13, homomorphic encryption of the real identity of the vehicle specifically includes the following steps: S131, TA is a legal vehicle V i Generate a unique identifier Calculate the hash value of the identity Create an identity data table with As a foreign key, establish a mapping relationship between it and the unique identity identifier and timestamp; at the same time, establish a foreign key A vehicle information table corresponding to the vehicle information to enable identity tracing of malicious vehicles; S132, TA uses the paillier.Enc(.) encryption algorithm to generate vehicle V i Encrypted identity The encrypted identity information generated Stored in the encrypted identity table, forming Fields and Encrypted Identities The TA then stores the random number r1 used for this encryption into the vehicle's tamper-proof memory.

4. The LBS privacy protection method in a sparse user environment of an Internet of Vehicles according to claim 1 is characterized in that: In step S25, the vehicle V i The virtual address generation process specifically includes the following steps: S251. According to the sparse user environment judgment algorithm, it is known that the vehicle user is in a sparse environment and P-1=K-N-1 virtual positions need to be generated; S252, using the vehicle coordinates loc u =(l ux ,l uy ) is the center of the circle, and the maximum distance r of the vehicle per unit time Δt max is the radius, constituting the maximum range A max Generate regions for virtual addresses; S253, in order to evenly distribute P virtual positions in A max In the area, select the minimum radius r min , forming the minimum anonymous space A min , A max The area is divided into A min and A min ∩A max Two parts, set the distribution coefficient d so that the generated virtual addresses are distributed in A min and annular area A min ∩A max middle; If the i-th position satisfies i%d==0 Then execute S254, in area A of the area min Generate a virtual position in the annular area A; otherwise, execute S255. min ∩A max Generate a virtual address in S254, in A min Generate virtual positions internally and calculate: θ=2π×random(0,1) x=r min ·radom(0,1)·cosθ y=r min ·radom(0,1)·sinθ Where random(0,1) represents a random number selected at (0,1) to ensure that the coordinate points x and y are within the minimum radius r min In the circle formed, the virtual address loc is obtained i (x,y), calculate If the result satisfies 0 <distance(loc i ,loc u )<r min , it means that the virtual address is within a reasonable range and is stored in the virtual address set PseuLoc; S255. Generate a virtual address in the ring area and calculate: a＝r min / r max x=r max ·radom(0,a)·cosθ y=r max ·radom(0,a)·sinθ where r max ·random(0,a) ensures that points x and y are in A min ∩A max In the ring formed, a virtual address loc is formed i (x,y), calculate If the result satisfies the condition r min ≤distance(loc i ,loc u )≤r max , it indicates that the virtual address is within the circular range and is stored in the virtual address set PseuLoc.

5. The LBS privacy protection method in a sparse user environment of an Internet of Vehicles according to claim 1 is characterized in that: The S3 step specifically includes the following steps: S31, RSU i Get LBS request content RSU i After receiving the request message, use the private key to decrypt the message: RSU i Get the temporary session key and timestamp, and check the validity of the timestamp, RSU i Using Session Keys Decryption Message, get RSU i Transmitted to TA via wired media Apply for legality certification of the pseudonym; S32. Vehicle pseudonym legitimacy verification TA uses the private key to decrypt get According to the value of , the encrypted identity tag is retrieved to obtain the encrypted identity of the vehicle Call the pseudonym authentication algorithm to verify the pseudonym: Enter a random number set Elements in r1、r 2j ① Execute the paillier.Enc(.) algorithm to get the random number set Nonce pseu Perform homomorphic encryption: ②Query the encrypted identity obtained With the encryption result Perform homomorphic multiplication operations and take hash values ​​respectively: The calculation results and For comparison: a) If the two are equal, according to the properties of homomorphic encryption, and because the random numbers r1, r 2i Vehicle V only i Knowing, therefore vehicle V i It is a legitimate vehicle, the pseudonym authentication is successful, TA pays RSU i Returns the result of successful authentication; b) If the results are not equal, TA returns the authentication failure result, RSU i Discard the LBS request packet and send a i Return failure message: Among them, "ERROR" indicates that the pseudonym authentication fails; TS is the timestamp, which is encrypted with a shared key. After the vehicle receives the feedback message, it uses the session key to decrypt it and needs to send the request package again to re-request the location service; S33, RSU i Forwarding LBS requests RSU i After receiving the authentication success message, the mapping relationship between the vehicle symmetric key and the pseudonym and address is cached. From the mapping relationship, it can be seen that for the return value of the TA verification result, the RSU i We only know that the pseudonym and address come from a legitimate user, but we cannot associate the pseudonym with the real identity of the vehicle. This creates a scenario where multiple different users request LBS at the same time, which confuses the user and address and achieves a double anonymity effect for LBS. At the same time, in order to further weaken the relevance of the pseudonym, a timestamp is generated for each group of connections in the table to invalidate the pseudonym timing when it times out. After confusion, RSU i It is impossible to distinguish between real and virtual addresses in the request set, and forward all service requests to the LBS through the wired medium. This request only contains two fields: address and service query content, namely Sensitive identity information is hidden because in the current sparse user environment, the increased service requests to achieve identity and address confusion will not lead to excessive occupation of communication resources; S34、RSU i →V i : LBS receives the service request set, obtains the corresponding LBS query content, makes the corresponding service response, and returns the service result RSU i According to the address information in the service result, the corresponding pseudonym and symmetric key are found through the mapping relationship between the key, pseudonym and address to form the service result: Encrypted with the symmetric key and returned to the vehicle V i ; S35: Vehicle obtains LBS service Vehicle decryption received RSU i The returned LBS service result set verifies the timeliness and integrity of the message, excludes the confusing pseudonyms, and selects the real pseudonyms Get the LBS service result corresponding to the real address and discard the rest of the returned results.

Images