Internet of vehicles privacy protection query method based on inner product encryption

Abstract

The invention discloses an Internet of Vehicles privacy protection query method based on inner product encryption, and relates to the field of Internet of Vehicles information security. A current Internet of Vehicles security query method mainly aims at preventing user data from being illegally accessed, but with diversification of Internet of Vehicles services, the contradiction between user privacy protection and service individuation requirements is increasingly remarkable. According to the method, an inner product encryption technology and a BFV homomorphic encryption technology are combined, and a query scheme for balancing privacy and service requirements is provided. The encryption index is generated through inner product encryption, and a service provider can complete retrieval by calculating the inner product of the encryption index and the pre-stored encryption keyword without directly obtaining query content. The BFV homomorphic encryption is adopted to allow necessary calculation on the ciphertext, so that the personalized service is realized while the query privacy is ensured. According to the method, the anonymity and privacy of user query are improved, a service provider is prevented from obtaining real query content of the user, meanwhile, the encryption algorithm design is optimized, and the query efficiency and the actual application value are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle networking, and in particular to a data security query method between a vehicle and an LBS in a vehicle networking. Background Art

[0002] With the rapid development of positioning technology, more and more people are beginning to use location-based services (LBS). At the same time, in order to provide personalized recommendation services based on user preferences, location service providers (LSPs) often collect their privacy data without the user's knowledge. This behavior may not only lead to the leakage of users' private information such as contact information and hobbies, but may also further threaten the user's personal safety. Therefore, how to provide users with efficient and reliable services while ensuring user privacy is an important issue that needs to be solved urgently.

[0003] Data range query is one of the most important outsourced services in Vehicular Ad-Hoc Network (VANET). For example, vehicle users may query the traffic congestion of a certain road section at a certain time, or query hotel and catering services within a one-kilometer radius. However, VANET services are latency-sensitive and have high reliability requirements, which makes it extremely challenging to achieve efficient and secure queries while ensuring data privacy and query privacy.

[0004] The unique characteristics of the Internet of Vehicles, such as self-organization, high-speed mobility, unfamiliarity between neighbors, and open communication channels, make it a potential target for attackers. Attackers can easily capture, change, replay, or delete information transmitted in the Internet of Vehicles, which will cause a series of security issues. These problems not only affect the reliability of the system, but also undermine users' trust in Internet of Vehicles services. Specifically, there are two main reasons: Security issues: Due to the use of inappropriate encoding methods or range query token generation algorithms, data privacy and query privacy may be leaked. Query performance issues: Traditional methods require a lot of pre-computation and matching operations, resulting in low query efficiency. When faced with large-scale Internet of Vehicles data sets and high-concurrency queries, serious query latency problems will occur, which in turn affects the quality of service.

[0005] To solve the above problems, this paper proposes a low information leakage secure outsourcing range query scheme. This scheme significantly improves query efficiency while protecting data privacy and query privacy, and is suitable for high-concurrency query requirements in large-scale Internet of Vehicles environments.

[0006] In addition to meeting the needs of range queries, the Internet of Vehicles also supports local communication through self-organizing networks, and can connect to the backup network through roadside units (RSU) as access points, thereby providing users with richer services such as entertainment and in-car office. However, when enjoying these services, users usually want to protect their privacy, while service providers want to obtain more information to provide more accurate personalized services. This contradiction leads to the problem of balancing the need for privacy protection and the need for service quality. In this context, controllable link technology becomes a feasible way to solve this problem, which helps to coordinate the privacy needs of users and service providers.

[0007] However, as a semi-trusted third party, LSP may passively or actively leak users' location privacy and other sensitive data, such as hobbies, age, and income level. This privacy leakage not only reduces users' acceptance of positioning services, but also restricts the promotion and application of Internet of Vehicles security query services in cloud environments. Therefore, this paper introduces the Private Information Retrieval (PIR) method to solve the privacy protection problem of key-value pair (KV-Private) query in the Internet of Vehicles, so as to better protect user privacy security. Summary of the invention

[0008] In order to overcome the above-mentioned shortcomings of the prior art, a privacy protection query method for Internet of Vehicles based on inner product encryption is proposed. The technical solution of the present invention is as follows:

[0009] A privacy protection query method for Internet of Vehicles based on inner product encryption includes four entities: vehicle-side OBU, trusted third-party TTP, roadside unit RSU and location server (Location Based Severs, LBS); including initialization stage, registration stage, matching stage and security query stage;

[0010] In the initialization phase, OBU first generates an efficient index vector for the database, and further constructs a retrieval structure for multi-dimensional hierarchical data, which is encrypted and protected using the inner product function encryption mechanism. At the same time, the index location information is encrypted using an efficient inner product encryption algorithm, and the encrypted retrieval structure is uploaded to the LBS.

[0011] In the online stage, the vehicle-side OBU generates a query trapdoor based on the real-time collected RSU information through the same mapping method, and uses the inner product function encryption mechanism to encrypt the query trapdoor and send it to the LBS. The LBS executes the retrieval algorithm of the query trapdoor on the encrypted retrieval structure, obtains the index vector that best matches the query trapdoor, and returns the corresponding index position information in the database. The vehicle-side OBU then decrypts the returned result to obtain the inner product result of the target location.

[0012] Since the inner product value generated by the constructed index vector during matching is fixed, it is efficient to use the mapped index vector for secure query. At the same time, the constructed efficient retrieval structure can significantly reduce the computational overhead of multi-keyword query under the keyword PIR retrieval framework. Therefore, this scheme has both fewer communication rounds and higher computational efficiency while achieving privacy protection.

[0013] The initialization module is used to complete the registration operation of the on-board unit (OBU) and the roadside unit (RSU) to the trusted third party (TTP);

[0014] The registration module is used to retrieve point of interest (POI) information from the RSU and complete a preliminary connection with the RSU when the OBU enters the communication range of the new RSU;

[0015] The matching module is used to complete the rapid matching and positioning of target information in the RSU database through the index mechanism when the OBU initiates a multi-keyword query request;

[0016] The security query module is used to implement multi-keyword security query of the vehicle to the RSU, which specifically includes the following three stages: index encoding stage, query stage and response stage, so as to ensure the privacy and integrity of the data during the query process.

[0017] Furthermore, the system initialization module is specifically used to complete the initialization configuration of the entire Internet of Vehicles system and generate security parameters of a trusted third party (TTP), which specifically includes the following contents:

[0018] Step 1.1 The trusted third party TTP generates a key pair for the public parameters of the entire system and itself, including selecting three cyclic groups G1, G2, G T And the group generators g1∈G1, g2∈G2, choose three hash functions Select a bilinear pair e:G1×G2→G T ; and select a random value And calculate Then for a vector of length n, randomly generate s i , And calculate Output master public key mpk = (G1, G2, G T ,q,e,g1,g2,h,(h1,…,h n )). For the security query module, given the security parameter λ, randomly sample s←χ, let sk=s as the query private key, and sample a←χ, calculate the query public key pk=([-(a·s)+e] q,a), TTP sends (pk,sk) to the on-board unit OBU through a secure channel. And at the same time, it sends pk to the roadside unit RSU.

[0019] Step 1.2: Further, the registration module is used for the location server LBS and the roadside unit RSU to register with the TTP, specifically including:

[0020] First, TTP selects the corresponding private key as the inner product decryption master private key msk=(v,(s1,…,s n ),(t1,…,t n )). And send msk to RSU through a secure channel.

[0021] When an OBU enters the communication range of a new RSU, the OBU first sends a request message to obtain the master private key. After receiving the request message, the RSU sends the msk and token RI to the OBU.

[0022] The encryption algorithm run by RSU takes the master public key mpk and keyword W as input and outputs the ciphertext ct. Then, RSU randomly selects a random number of length n The encryption process is to calculate and C2 = h r , and construct the following query vector X=(x1,x2,...,x 2n ), where each item satisfies x 2i-1 =w i ·r·r i , x 2i =-r·r i For each i∈[2n], calculate Ciphertext ct=(C1,C2,(E1,…,E n )).

[0023] RSU needs to encrypt all keywords to be uploaded according to the above steps. Finally, RSU sends these indexes generated by the files to the location server (LBS).

[0024] Step 1.3 After OBU receives the master private key msk, use the private key encryption Get feature key It may contain wildcard characters such as *. In addition, OBU needs to According to the following rules, it is expanded into a vector Y=(y1,...,y 2n ):

[0025]

[0026] The OBU then selects a random number And calculate the encrypted ciphertext: Function Key Then OBU Finally, the vehicle sends the function key generated by the multiple keywords it wants to search for to the cloud server and waits for the matching result.

[0027] Step 1.4 LBS receives search trap Then, use the master public key mpk and Decrypt the ciphertext ct to obtain the inner product result e(g1,g2) of the search z·＜X,Y＞ , verify the equation If so, the LBS completes the matching process and returns the corresponding encrypted search index to the OBU; otherwise, the LBS cannot match the corresponding keyword information in the database.

[0028] Furthermore, the security query module is used to complete the retrieval of the vehicle's message index, which specifically includes the index encoding stage, the query stage, and the decryption stage, a total of three stages:

[0029] The index encoding phase described in step 1.5 is used to split the database stored in the RSU and complete the correct decryption of the information. (n,m,k,b)-batch code means encoding the database DB of n elements into m codewords, distributed in b bins, and querying k elements at a time. Specifically, it includes:

[0030] RSU splits the stored database and uses SIMD technology to correctly decrypt the information through appropriate parameter selection. Batch processing technology (n,m,k,b)-batch code is expressed as encoding the database DB of n elements into m code words distributed in b bins. RSU splits the database DB into b buckets (C0,...,C b-1 ), contains 0 or more codes, and satisfies For each index in the DB, calculate the three hashes of cuckooHash respectively, get three indexes, and insert (i, data) into the three corresponding indexes.

[0031] Step 1.6 If a selection vector L containing multiple query indices is executed simultaneously, the RSU maps each index in L to a random bucket through CucooHash, and calculates k candidate bins for each query by applying k hash functions. First, determine whether the codeword exists. For i∈L, one or more bin indices are given, from which the codeword that can be used to reconstruct the element DB[i] is retrieved. If so, the query can obtain the corresponding element DB[i] in the database. Otherwise, the RSU outputs ⊥. Then, for each query index Li , L i Put it into the corresponding candidate bucket. At this time, the number of candidate buckets is about 1.5k. Then, perform PIR for each bin, a total of b PIRs. The index of each PIR is the index of the database actually queried by the client in the bin. Only when the index is correct and the bucket mapping is passed at the same time, RSU considers this query to be valid.

[0032] Step 1.7: Query phase: The vehicle OBU needs to generate a query trap T based on the index. q , and then retrieve the data. First, OBU selects the plaintext domain m∈Z[X] p / (X N +1), where the chosen plaintext modulus p is an integer, the order of the polynomial is a power of 2, and the chosen ciphertext domain ct∈Z[X] q / (X N +1), and the ciphertext modulus q>>p. In order to realize information retrieval, the OBU first embeds all bits in the query vector into a plaintext polynomial p(x)=∑ i∈{1,2,...,N} L i x i ; Then, OBU generates an encrypted query trapdoor T q , encode the index selection vector L into Z[X] q / (X N +1) domain; then, u, e1, e2 are randomly and uniformly sampled, where parameters u and e are coefficient polynomials. The vehicle OBU sends a query trap door T q =BFV(L,pk)=([pk1·u+e1+Δ·L] q ,[pk2·u+e2] q ) to RSU. For a database DB[n] containing n elements d, RSU obtains Enc(d i ·L i ).

[0033] Step 1.8: Response phase: For any record (i, d i ), i<=n, RSU outputs the query result R and sends it back to OBU. Denote Enc(1), using represents Enc(0); then, select the value at that index position in the vector and use express, On the contrary; finally, OBU obtains decryptable ciphertext

[0034] A privacy protection query method for Internet of Vehicles based on inner product encryption includes four parts: system initialization module, registration module, inner product matching module and security query module; wherein,

[0035] System initialization module, completes the initialization of the Internet of Vehicles system and the trusted third-party TTP, trusted third-party TTP public parameter set:

[0036] mpk=(G1,G2,G T ,q,e,g1,g2,h,(h1,…,h n ))

[0037] The registration module completes the registration of the on-board unit OBU and the roadside unit RSU. The on-board unit OBU and the roadside unit RSU obtain the master private key and the master public key from the TTP respectively, and encrypt their respective indexes and send them to the LBS through a secure channel.

[0038] The inner product matching module completes the index matching of the location server LBS to the vehicle-mounted unit query keyword and obtains the corresponding database index within the RSU range.

[0039] The security query module, when the on-board unit OBU receives the database encrypted index, completes the security query of the database, including the index encoding stage, query stage, and response stage.

[0040] The advantages and beneficial effects of the present invention are as follows:

[0041] Compared with the prior art, the invention has the following three advantages:

[0042] (1) The proposed method achieves anonymity in the matching stage. By encoding multiple keyword information, other users except the LBS cannot obtain the user's query keywords based on the transmitted information.

[0043] (2) The proposed method achieves efficient acquisition of database indexes. The OBU can query information with multiple keywords by simply using the public and private keys of a trusted third party. It also balances the privacy requirements between the OBU and the RSU while protecting user privacy information.

[0044] (3) The proposed method achieves adaptive security in the matching phase and can solve certain security issues. Compared with other solutions for achieving semantic security, this method can resist active attacks from malicious adversaries. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A general model diagram provided for the present invention;

[0046] Figure 2This is a schematic diagram of CucooHash bucketing of the present invention;

[0047] Figure 3 A structural diagram of the vehicle networking query system provided by the present invention;

[0048] Figure 4 A diagram showing the definition of symbols used in the present invention; DETAILED DESCRIPTION

[0049] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0050] The technical solution of the present invention to solve the above technical problems is:

[0051] Reference Figure 1-Figure 4 , the specific implementation methods of the present invention are as follows:

[0052] 1. In the initial case, each element of the Internet of Vehicles system is initialized, and the trusted third-party organization TTP generates a key pair for the public parameters of the entire system and itself, including selecting three cyclic groups G1, G2, G T And the group generators g1∈G1, g2∈G2, choose three hash functions Select a bilinear pair e:G1×G2→G T ; and select a random value And calculate Then for a vector of length n, randomly generate And calculate Output master public key mpk = (G1, G2, G T ,q,e,g1,g2,h,(h1,…,h n )). For the security query module, given the security parameter λ, randomly sample s←χ, let sk=s as the query private key, and sample a←χ, calculate the query public key pk=([-(a·s)+e] q ,a), TTP sends (pk,sk) to the on-board unit OBU through a secure channel. And at the same time, it sends pk to the roadside unit RSU.

[0053] 2. The participants in the Internet of Vehicles, the onboard unit OBU and the roadside unit RSU, register with the TTP. First, the TTP selects the corresponding private key as the inner product decryption master private key msk=(v,(s1,…,s n ),(t1,…,t n )). And send msk to RSU through a secure channel.

[0054] When an OBU enters the communication range of a new RSU, the OBU first sends a request message to obtain the master private key. After receiving the request message, the RSU sends the msk and token RI to the OBU.

[0055] The encryption algorithm run by RSU takes the master public key mpk and keyword W as input and outputs the ciphertext ct. Then, RSU randomly selects a random number of length n The encryption process is to calculate and C2 = h r , and construct the following query vector X=(x1,x2,...,x 2n ), where each item satisfies x 2i-1 =w i ·r·r i , x 2i =-r·r i For each i∈[2n], calculate Ciphertext ct=(C1,C2,(E1,…,E n )).

[0056] 3. RSU needs to encrypt all keywords to be uploaded according to the above steps. Finally, RSU sends these indexes generated by the files and the attached keywords to the cloud server.

[0057] After OBU receives the master private key msk, it uses the private key encryption Get feature key It may contain wildcard characters such as *. In addition, OBU needs to According to the following rules, it is expanded into a vector Y=(y1,...,y 2n ):

[0058]

[0059] The OBU then selects a random number And calculate the encrypted ciphertext, Function Key Then OBU Finally, the OBU sends the generated function key containing multiple keywords he wants to search to the cloud server and waits for the matching result.

[0060] 4. LBS receives search trap Then, use the master public key mpk and Decrypt the ciphertext ct to obtain the inner product result e(g1,g2) of the search z·＜X,Y＞ , verify the equation If so, the LBS completes the matching process and returns the corresponding encrypted search index to the OBU; otherwise, the LBS cannot match the corresponding keyword information in the database.

[0061] 5. Furthermore, the safety query module is used to complete the retrieval of the vehicle's message index, specifically including:

[0062] In order to split the database stored in the RSU and complete the correct decryption of the information, batch coding is used to encode the database. (n,m,k,b)-batch code means that the database DB of n elements is encoded into m codewords, distributed in b bins, and k elements are queried at the same time. Specifically, it includes:

[0063] RSU splits the stored database and uses SIMD technology to correctly decrypt the information through appropriate parameter selection. Batch processing technology (n,m,k,b)-batch code is expressed as encoding the database DB of n elements into m code words distributed in b bins. RSU splits the database DB into b buckets (C0,...,C b-1 ), contains 0 or more codes, and satisfies For each index in the DB, calculate the three hashes of cuckooHash respectively, get three indexes, and insert (i, data) into the three corresponding indexes.

[0064] 6. If a selection vector L containing multiple query indices is executed simultaneously, RSU maps each index in L to a random bucket through CucooHash, and calculates k candidate bins for each query by applying k hash functions. First, determine whether the codeword exists. For i∈L, one or more bin indices are given, from which the codeword that can be used to reconstruct the element DB[i] is retrieved. If so, the query can obtain the corresponding element DB[i] in the database. Otherwise, RSU outputs ⊥. Then, for each query index L i , L i Put it into the corresponding candidate bucket. At this time, the number of candidate buckets is about 1.5k. Then, perform PIR for each bin, a total of b PIRs. The index of each PIR is the index of the database actually queried by the client in the bin. Only when the index is correct and the bucket mapping is passed at the same time, RSU considers this query to be valid.

[0065] 7. The vehicle OBU needs to generate a query trap T based on the index q , and then retrieve the data. First, OBU selects the plaintext domain m∈Z[X] p / (X N+1), where the chosen plaintext modulus p is an integer, the order of the polynomial is a power of 2, and the chosen ciphertext domain ct∈Z[X] q / (X N +1), and the ciphertext modulus q>>p. In order to achieve information retrieval, the OBU first embeds all bits in the query vector into a plaintext polynomial p(x)=Σ i∈{1,2,...,N} L i x i ; Then, OBU generates an encrypted query trapdoor T q , encode the index selection vector L into Z[X] q / (X N +1) domain; then, u, e1, e2 are randomly and uniformly sampled, where parameters u and e are coefficient polynomials. The vehicle OBU sends a query trap door T q =BFV(L,pk)=([pk1·u+e1+Δ·L] q ,[pk2·u+e2] q ) to RSU. For a database DB[n] containing n elements d, RSU obtains Enc(d i ·L i ).

[0066] 8. For any record in the database (i, d i ), i<=n, RSU outputs the query result R and sends it back to OBU. Denote Enc(1), using represents Enc(0); then, select the value at that index position in the vector and use express, On the contrary; finally, OBU obtains decryptable ciphertext

[0067] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A privacy protection query method for Internet of Vehicles based on inner product encryption, characterized in that: It includes the following four parts: system initialization module, registration module, matching module and security query module; wherein, the system initialization module is used for initialization of the Internet of Vehicles system and the trusted third-party organization TTP; the registration module is used for registration of the on-board unit OBU and the roadside unit RSU; the matching module is used for obtaining the database index of a new RSU from the LBS when the OBU enters the communication range of the RSU; the security query module is used for completing the vehicle security query operation based on the keyword index after the OBU receives the database index.

2. According to claim 1, a privacy protection query method for Internet of Vehicles based on inner product encryption is characterized in that: The system initialization module is used for initialization of the Internet of Vehicles system and the trusted third-party organization TTP, specifically including: The trusted third party TTP generates a key pair for the public parameters of the entire system and itself, including selecting three cyclic groups G1, G2, G T And the group generators g1∈G1, g2∈G2, choose three hash functions Select a bilinear pair e:G1×G2→G T ; and select a random value And calculate Then for a vector of length n, randomly generate And calculate Output master public key mpk = (G1, G2, G T ,q,e,g1,g2,h,(h1,…,h n )). For the security query module, given the security parameter λ, randomly sample s←χ, let sk=s as the query private key, and sample a←χ, calculate the query public key pk=([-(a·s)+e] q ,a), TTP sends (pk,sk) to the on-board unit OBU through a secure channel. And at the same time, it sends pk to the roadside unit RSU.

3. According to claim 2, a privacy protection query method for Internet of Vehicles based on inner product encryption is characterized in that: The registration module is used for the on-board unit OBU and the roadside unit RSU to register with the TTP, specifically including: First, TTP selects the corresponding private key as the inner product decryption master private key msk=(v,(s1,…,s n ),(t1,…,t n )). And send msk to RSU through a secure channel. When an OBU enters the communication range of a new RSU, the OBU first sends a request message to obtain the master private key. After receiving the request message, the RSU sends the msk and token RI (RSUID) to the OBU. The encryption algorithm run by RSU takes the master public key mpk and keyword W as input and outputs the ciphertext ct. Then, RSU randomly selects a random number of length n The encryption process is to calculate and At the same time, the following query vector X=(x1,x2,...,x 2n ), where each item satisfies x 2i-1 =w i ·r·r i , x 2i =-r·r i For each i∈[2n], calculate Ciphertext ct=(C1,C2,(E1,…,E n )). RSU needs to encrypt all keywords to be uploaded according to the above steps. Finally, RSU sends these indexes generated by the files to the location server (LBS). OBU registration includes: After OBU receives the master private key msk, it uses the private key encryption Get feature key It may contain wildcard characters such as *. In addition, OBU needs to According to the following rules, it is expanded into a vector Y=(y1,...,y 2n ): The OBU then selects a random number And calculate the encrypted ciphertext: OBU gets the function key Then send Finally, the vehicle sends the function key generated by the multiple keywords it wants to search for to the cloud server and waits for the matching result.

4. According to claim 3, a privacy protection query method for Internet of Vehicles based on inner product encryption is characterized in that: The matching module is used to obtain the database index of a new RSU from the LBS when the OBU enters the communication range of the RSU, specifically including: LBS receives search trap Then, use the master public key mpk and Decrypt the ciphertext ct to obtain the inner product result e(g1,g2) of the search z·＜X,Y＞ , verify the equation If so, LBS completes the match and returns the corresponding encrypted search index to OBU; otherwise, LBS cannot match the corresponding keyword information in the database.

5. According to claim 4, a privacy protection query method for Internet of Vehicles based on inner product encryption is characterized in that: The safety query module, after the vehicle-mounted unit receives the database index, is used to complete the vehicle's query on the keyword index, which specifically includes three stages: The index encoding stage is used to split the database stored in the RSU to complete the correct decryption of the information. (n,m,k,b)-batch code means encoding the database DB of n elements into m codewords, distributed in b buckets (bins), and querying k elements at a time. Specifically, it includes: RSU splits the stored database and uses SIMD technology to correctly decrypt the information through appropriate parameter selection. Batch processing technology (n,m,k,b)-batch code is expressed as encoding the database DB of n elements into m code words distributed in b bins. RSU splits the database DB into b buckets (C0,...,C b-1 ), contains 0 or more codes, and satisfies For each index in the DB, calculate the three hashes of cuckooHash respectively, get three indexes, and insert (i, data) into the three corresponding indexes.

6. For an index set L that executes multiple queries simultaneously, RSU maps each index in L to a random bucket through CucooHash, and calculates k candidate bins for each query by applying k hash functions. First, determine whether the codeword exists. For i∈L, one or more bin indices are given, from which the codeword that can be used to reconstruct the element DB[i] is retrieved. If so, the query can obtain the corresponding element DB[i] in the database. Otherwise, RSU outputs ⊥. Then, for each query index q i , q i Put it into the corresponding candidate bucket. At this time, the number of candidate buckets is about 1.5k. Then, perform PIR for each bin, a total of b PIRs. The index of each PIR is the index of the database actually queried by the client in the bin. Only when the index is correct and the bucket mapping is passed at the same time, RSU considers this query to be valid.

7. If a selection vector L containing multiple query indices is executed simultaneously, RSU maps each index in L to a random bucket through CucooHash, and calculates k candidate bins for each query by applying k hash functions. First, determine whether the codeword exists. For i∈L, one or more bin indices are given, from which the codeword that can be used to reconstruct the element DB[i] is retrieved. If so, the query can obtain the corresponding element DB[i] in the database. Otherwise, RSU outputs ⊥. Then, for each query index L i , L i Put it into the corresponding candidate bucket. At this time, the number of candidate buckets is about 1.5k. Then, perform PIR for each bin, a total of b PIRs. The index of each PIR is the index of the database actually queried by the client in the bin. Only when the index is correct and the bucket mapping is passed at the same time, RSU considers this query to be valid.

8. The query phase: the vehicle OBU needs to generate a query trap T according to the index q , and then retrieve the data. First, OBU selects the plaintext domain m∈Z[X] p / (X N +1), where the chosen plaintext modulus p is an integer, the order of the polynomial is a power of 2, and the chosen ciphertext domain ct∈Z[X] q / (X N +1), and the ciphertext modulus q>>p. In order to realize information retrieval, the OBU first embeds all bits in the query vector into a plaintext polynomial p(x)=∑ i∈{1,2,...,N} L i x i ; Then, OBU generates an encrypted query trapdoor T q , encode the index selection vector L into Z[X] q / (X N +1) domain; then, u, e1, e2 are randomly and uniformly sampled, where parameters u and e are coefficient polynomials. The vehicle OBU sends a query trap door T q =BFV(L,pk)=([pk1·u+e1+Δ·L] q ,[pk2·u+e2] q ) to RSU. For a database DB[n] containing n elements d, RSU obtains Enc(d i ·L i ).

9. The response phase: For any record (i, d i ), i<=n, RSU outputs the query result R and sends it back to OBU. Denote Enc(1), using Indicates Enc(0); then, the value of the index position in the selection vector is represented by 1, and 0 is the opposite; finally, OBU obtains the decryptable ciphertext in, System initialization module, completes the initialization of the Internet of Vehicles system and the trusted third-party TTP, trusted third-party TTP master private key set: mpk=(G1,G2,G T ,q,e,g1,g2,h,(h1,…,h n )) The registration module completes the registration of the on-board unit OBU and the roadside unit RSU. The on-board unit OBU and the roadside unit RSU obtain the master private key and the master public key from the TTP respectively, and encrypt their respective indexes and send them to the LBS through a secure channel. The matching module completes the index matching of the location server LBS to the vehicle-mounted unit query keyword and obtains the corresponding database index within the RSU range. The security query module, when the on-board unit OBU receives the database encrypted index, completes the security query of the database, including the index encoding stage, query stage, and response stage.

Images