A blockchain-based trust management method for the Internet of Vehicles

Abstract

A blockchain-based trust management method for the Internet of Vehicles (IoV), which relates to the technical field of IoV. The present invention is to solve the problem that the existing trust management methods for IoV still have low security. The present invention includes: allocating registration certificates to vehicles in the IoV network and uploading them to the blockchain; vehicle Vi broadcasts traffic event information, and the RSU closest to Vi obtains the trust value of Vi based on the evaluations of other vehicles on the broadcast traffic event information, and updates the trust threshold of Vi; the Raft consensus algorithm is used to select an RSU as the leader, and the leader packs the trust values sent by the vehicles into a block and broadcasts the block to all RSUs. The RSUs verify the integrity of the fields in the block. If the verification passes, the RSUs transmit the audit results to each other. When the leader receives the audit results of the quantity threshold A, the block is submitted to the blockchain, thereby updating the vehicle trust value on the blockchain. The present invention is used to ensure the security of vehicle information in IoV.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle networking, and particularly relates to a vehicle networking trust management method based on blockchain. Background Art

[0002] As the infrastructure of the intelligent transportation system, vehicle networking has attracted extensive attention from the industrial and academic circles. Due to the high mobility and variability of vehicles, adjacent vehicles are unfamiliar with each other and do not trust each other. Once a malicious vehicle leaks the location of other vehicles or reports false information to disrupt the normal driving of other vehicles, it will cause unpredictable losses. Therefore, evaluating the credibility of vehicles has become an important research content in vehicle networks.

[0003] Many studies have proposed that trust management is an effective way to solve the trust problem in vehicle networking. It can calculate the credibility of vehicle networking information, thereby improving the accuracy of identifying false information. It can also calculate and update the trust value of vehicles, effectively solving the problems of false information and malicious vehicles in vehicle networking. Therefore, the credibility of vehicles in vehicle networking can be used as a basis for service operators to punish malicious vehicle behaviors. Existing trust management mainly adopts a centralized architecture. The centralized architecture records all ratings and results calculated by the trust model on a centralized server. However, the decision-making of vehicles within an extremely short time delay will affect the service quality of vehicles. In addition, if the centralized server is attacked by an attacker, the trust information of vehicles will be leaked, causing unpredictable losses. Therefore, the existing vehicle networking trust management method based on a centralized architecture still has the problem of low security. Summary of the Invention

[0004] The purpose of the present invention is to solve the problem of low security existing in the existing vehicle networking trust management method, and a vehicle networking trust management method based on blockchain is proposed.

[0005] A vehicle networking trust management method based on blockchain includes:

[0006] Step 1: Allocate registration certificates to vehicles in the vehicle networking and upload the registration certificates to the blockchain;

[0007] Step 2: Vehicle Vi in the vehicle networking uses the registration certificate as a pseudonym. Vehicle Vi broadcasts traffic event information in the vehicle networking using the pseudonym. The RSU closest to vehicle Vi obtains the trust value of vehicle Vi according to the evaluations of other vehicles on the broadcast traffic event information. The RSU closest to vehicle Vi updates the trust threshold of vehicle Vi using the trust value of vehicle Vi;

[0008] Step 3: Use the Raft consensus algorithm to select an RSU as the leader. Determine whether the RSU closest to vehicle Vi is the leader. If the RSU closest to vehicle Vi is not the leader, no operation is performed. If the RSU closest to vehicle Vi is the leader, the leader packs the trust value sent by the vehicle into a block and broadcasts the block to all RSUs. The RSUs verify the integrity of the fields in the block. If the verification passes, the RSUs exchange the audit results with each other. If the number of audit results received by the leader is greater than the quantity threshold A, the block is submitted to the blockchain, thereby updating the vehicle trust value on the blockchain. If the number of audit results received by the leader from the RSUs is less than the quantity threshold A, a new leader is re-elected.

[0009] Further, the process of allocating a registration certificate to a vehicle in the vehicle network and uploading the registration certificate to the blockchain in Step 1 is specifically as follows:

[0010] Step 1-1: After the vehicle enters the vehicle network, the trusted third party TA allocates a registration certificate to the vehicle;

[0011] Step 1-2: Upload the vehicle registration certificate to the blockchain:

[0012] The vehicle that has been allocated the registration certificate generates a public-private key pair and sends a message containing real identity information to the TA. The TA determines whether the current vehicle identity is legal. If the current vehicle identity is illegal, the registration certificate of the current vehicle is revoked. If the current vehicle identity is legal, the current vehicle uses the registration certificate as a pseudonym and uploads the pseudonym to the certificate blockchain.

[0013] If the difference between the expiration time of the current vehicle registration certificate and the current time is less than or equal to the preset time interval, the vehicle regenerates a new public-private key pair and sends the real identity information to the TA again. The TA updates the registration certificate for the current vehicle and uploads the updated registration certificate to the certificate blockchain.

[0014] Further, the process by which the TA determines whether the current vehicle identity is legal is specifically as follows: Determine whether the current vehicle registration certificate has expired. If it has not expired, it is legal; otherwise, it is illegal.

[0015] Further, in Step 2, vehicle Vi uses the pseudonym to broadcast traffic event information in the vehicle network. The RSU closest to vehicle Vi obtains the trust value of vehicle Vi based on the evaluations of other vehicles on the broadcast traffic event information. The RSU closest to vehicle Vi updates the trust threshold of vehicle Vi using the trust value of vehicle Vi. Specifically:

[0016] Step 2-1: Vehicle Vi broadcasts traffic event information to other vehicles in the vehicle network. Other vehicles use the label number of the type of the broadcast traffic event information as the observation value, thereby obtaining an observation sequence;

[0017] The types of the broadcast traffic event information include: 1. accident information; 2. road anomaly information; 3. traffic information;

[0018] Step 22: The RSU closest to vehicle Vi trains the CRF using the historical hidden sequence and historical observation sequence of the vehicle until the recall rate, precision rate, and F1 score of the trained CRF model all reach the preset thresholds, and obtains the optimal parameters of the CRF model:

[0019] The hidden sequence is the evaluation sequence of other vehicles on the broadcast traffic event information;

[0020] The hidden sequence has true and false. True indicates that the vehicle approves the broadcast traffic event information, and false represents that the vehicle does not approve the broadcast traffic event information;

[0021] Step 23: Input the optimal parameters of the CRF model and the observation sequence of vehicle Vi into the CRF model to obtain the hidden sequence of other vehicles on the broadcast traffic event information, and obtain the trust value of vehicle Vi using the hidden sequence; Step 24: The RSU closest to vehicle Vi updates the trust threshold of vehicle Vi using the trust value of vehicle Vi.

[0022] Further, the CRF model is specifically:

[0023]

[0024] where x is the observation sequence, y is the hidden sequence, Z(x) is the intermediate variable, i is the time label, k is the edge label, t k is the eigenvalue defined on the edge, s l is the eigenvalue defined on the node, l is the node label, λ k is the edge weight, μ l is the node weight, y i is the hidden value at the i-th time, y i-1 is the hidden value at the (i - 1)-th time.

[0025] Further, the obtaining the trust value of vehicle Vi using the hidden sequence in Step 23 is specifically:

[0026] f(y) = 1 / (1 + e -wy )

[0027] where y is the hidden sequence, w is the trust value weight, and f(y) is the trust value of vehicle Vi.

[0028] Further, the RSU closest to vehicle Vi updating the trust threshold of vehicle Vi in Step 24 is specifically:

[0029] First, set the vehicle trust threshold group (δ1, δ2,.. δ j ... δ n );

[0030] Among them, (δ1, δ2,.. δ j , δ n ) is an increasing arithmetic sequence, δ j is the j-th trust threshold, and n is the total number of trust thresholds;

[0031] Then, the RSU closest to vehicle Vi compares the trust value of vehicle Vi with the current trust threshold. If the trust value of vehicle Vi is greater than or equal to the current trust threshold δ j , it means that the time information sent by the current vehicle is true information, and the current trust threshold is saved; if the vehicle trust value is less than the current trust threshold δ j , then it is judged whether the current trust threshold δ j is equal to δ n . If δ j =δ n , then the current vehicle is excluded from the vehicle networking network. If the current trust threshold δ j is not equal to δ n , the current trust threshold is updated to δ j+1 ;

[0032] Among them, the initial trust threshold of the vehicle is δ1.

[0033] Furthermore, the quantity threshold in step three

[0034] The beneficial effects of the present invention are as follows:

[0035] The present invention ensures a safe and trustworthy trust environment in the vehicle networking network by calculating the trust value of the vehicle, uploads the calculated trust data of the vehicle to the blockchain for data management, and uses blockchain technology to ensure the anonymity and immutability of the vehicle data, thereby enhancing the security of the vehicle networking trust management method. Based on the training and prediction of the CRF model, the present invention proposes a method for calculating the trust value of a vehicle, predicts the hidden sequence of the vehicle according to the observed sequence of the vehicle, thereby obtaining the trust value of the vehicle, determines whether the current vehicle is trustworthy according to the trust threshold of the vehicle, and ensures the security of the vehicle information through the blockchain, greatly reducing the generation of malicious vehicles, ensuring the immutability and anonymity of the vehicle data, and enhancing the privacy security of the vehicle data. The present invention can judge whether the behavior of a vehicle is trustworthy in an untrusted vehicle networking environment, thereby enhancing the security of the vehicle networking trust management. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is the system framework diagram of the present invention;

[0037] Figure 2 It is a diagram of the CRF model. Specific implementation manners

[0038] In the field of blockchain, the emergence of blockchain provides a new direction for the distributed storage and management of vehicle networking data. Blockchain is a decentralized ledger that integrates technologies such as smart contracts, consensus algorithms, and cryptography, and has characteristics such as immutability and anonymity. Each transaction stored in the blockchain is not easily tampered with, and the trust values of vehicles recorded on the blockchain have good anonymity and traceability. Moreover, due to the decentralized nature of the blockchain, the present invention conducts trust management among distributed RSUs, effectively avoiding the problems of a centralized server. In addition, the blockchain benefits from distributed consensus algorithms, enabling RSUs to work together to maintain a consistent database to ensure the synchronization of trust value messages. First, the vehicles use pseudonyms to communicate with each other to protect their real identities. Second, the RSUs verify whether the identities of the vehicles are legal. If legal, the trust values of the vehicles are calculated through a designed trust evaluation algorithm based on CRF, and malicious vehicles are identified through trust thresholds to restrict and regulate the behaviors of the vehicles. Finally, the RSUs package the trust values into blocks and run a consensus mechanism to reach a consistency agreement, and run a smart contract to update the trust values of the vehicles. Next, the present invention will be described in combination with specific implementation manners.

[0039] Specific implementation manner one: As Figure 1 shown, the specific process of a vehicle networking trust management method based on blockchain in this implementation manner is as follows:

[0040] Step 1: Allocate registration certificates for the vehicles in the vehicle networking network and upload the registration certificates to the blockchain;

[0041] Step 11: After a vehicle enters the vehicle networking network, a trusted third party TA allocates a registration certificate for the vehicle;

[0042] Step 12: The vehicle that has been allocated a registration certificate generates a public-private key pair and sends a message containing real identity information to the TA. The TA determines whether the current vehicle identity is legal. If the current vehicle identity is not legal, the registration certificate of the current vehicle is revoked; if the current vehicle identity is legal, the current vehicle uses the registration certificate as a pseudonym and uploads the pseudonym to the certificate blockchain;

[0043] If the difference between the expiration time of the current vehicle's registration certificate and the current time is less than or equal to a preset time interval, the vehicle regenerates a new public-private key pair and sends the real identity information to the TA again. The TA updates the registration certificate for the current vehicle and uploads the updated registration certificate to the certificate blockchain;

[0044] The determination of whether the current vehicle identity is legal is specifically as follows: Determine whether the current vehicle registration certificate has expired. If it has not expired, it indicates legality; otherwise, it indicates illegality.

[0045] The real information includes: location and behavior.

[0046] Step 2: Vehicle Vi in the vehicle networking network uses the registration certificate as a pseudonym. Vehicle Vi broadcasts traffic event information in the vehicle networking network using the pseudonym. The RSU closest to vehicle Vi obtains the trust value of vehicle Vi based on the evaluations of other vehicles on the broadcast traffic event information. The RSU closest to vehicle Vi updates the trust threshold of vehicle Vi using the trust value of vehicle Vi, specifically as follows:

[0047] Step 2-1: Vehicle Vi broadcasts traffic event information to other vehicles in the vehicle networking network. Other vehicles use the type label of the broadcast traffic event information as the observation value to obtain an observation sequence.

[0048] The types of the traffic event information include: 1) Accident information: For example, emergencies such as vehicle crashes. 2) Road anomaly information: Such as road icing, waterlogging, etc., to avoid accidents due to road anomalies. 3) Traffic information: Such as ordinary situations like traffic congestion on the road;

[0049] Step 2-2: The RSU closest to vehicle Vi trains a CRF model using the vehicle historical hidden sequence and historical observation sequence until the recall rate, precision, and F1 score of the trained CRF model all reach the preset thresholds to obtain the optimal parameters of the CRF model:

[0050] As Figure 2 shown, the conditional random field (CRF) combines the characteristics of the maximum entropy model and the hidden Markov model. It is an undirected graph model. The CRF model obtains the conditional probability distribution of the hidden sequence under the given observation sequence:

[0051]

[0052] Among them, x is the observation sequence, y is the hidden sequence, Z(x) is the intermediate variable, i is the time label, k is the edge label, t k is the eigenvalue defined on the edge, s l is the eigenvalue defined on the node, l is the node label, λ k is the edge weight, μ l is the node weight, y i is the hidden value at the i-th time, y i-1 is the hidden value at the (i - 1)-th time;

[0053] Train the CRF model until the recall rate, precision rate, and F1 score all reach the preset thresholds, and take the currently trained CRF parameters as the optimal parameters;

[0054] Step 2.3: Input the optimal parameters of the CRF model and the observation sequence of vehicle Vi into the CRF model to obtain the hidden sequence of other vehicles' broadcast information, and use the hidden sequence to obtain the trust value of vehicle Vi, and send the vehicle trust value to the RSU:

[0055] Use the hidden sequence to obtain the trust value of vehicle Vi, specifically:

[0056] f(y) = 1 / (1 + e -wy )

[0057] where y is the hidden sequence and w is the trust value weight;

[0058] The hidden sequence consists of true and false, where true represents that the vehicle approves of this broadcast information, and false represents that the vehicle does not approve of this broadcast information.

[0059] Step 2.4: The RSU closest to vehicle Vi uses the vehicle trust value to obtain and update the trust threshold of vehicle Vi, specifically:

[0060] First, set the vehicle trust threshold group (δ1, δ2,..δ j ...δ n );

[0061] where (δ1, δ2,..δ j ...δ n ) is an increasing arithmetic sequence;

[0062] Then, the RSU compares the vehicle trust value with the current trust threshold. If the vehicle trust value is greater than or equal to the current trust threshold δ j , it means that the time information sent by the current vehicle is true information, and the current trust threshold is saved. If the vehicle trust value is less than the current trust threshold δ j , it is determined whether the current trust threshold δ j is equal to δ n . If δ j =δ n , the current vehicle is excluded from the vehicle networking. If the current trust threshold δ j is not equal to δ n , the current trust threshold is updated to δ j+1 ;

[0063] where the initial trust threshold of the vehicle is δ1.

[0064] Step 3: Use the Raft consensus algorithm to select an RSU as the leader. Determine whether the RSU closest to vehicle Vi is the leader. If the RSU closest to vehicle Vi is not the leader, no operation is performed; if the RSU closest to vehicle Vi is the leader, the leader packs the trust value sent by the vehicle into a block and broadcasts the block to all authorized RSUs. The authorized RSUs verify the integrity of the fields in the block. If the verification passes, the authorized RSUs transmit the audit results to each other. When the number of audit results received by the leader from the RSUs is greater than the quantity threshold the block is submitted to the blockchain, thereby updating the vehicle trust value on the blockchain. If the number of audit results received by the leader from the RSUs is less than the quantity threshold A, a new leader is elected.

[0065] The Raft consensus algorithm includes three types of roles, namely the leader, candidate, and follower. It uses a heartbeat mechanism to trigger leader elections, and the leader periodically sends heartbeat packets to all followers to maintain its authority.

Claims

1. A blockchain-based trust management method for vehicle networking, characterized in that The specific process of the method is as follows: Step 1: Assign a registration certificate to the vehicles in the vehicle networking and upload the registration certificate to the blockchain. Step 2: The vehicle Vi in the vehicle networking uses the registration certificate as a pseudonym. The vehicle Vi broadcasts traffic event information in the vehicle networking using the pseudonym. The RSU closest to the vehicle Vi obtains the trust value of the vehicle Vi based on the evaluations of other vehicles on the broadcast traffic event information. The RSU closest to the vehicle Vi updates the trust threshold of the vehicle Vi using the trust value of the vehicle Vi. Specifically: Step 2-1: The vehicle Vi broadcasts traffic event information to other vehicles in the vehicle networking. Other vehicles use the label of the type of the broadcast traffic event information as the observation value, thereby obtaining an observation sequence. The types of the broadcast traffic event information include:

1. Accident information; 2. Road anomaly information; 3. Traffic information. Step 2-2: The RSU closest to the vehicle Vi trains a CRF using the historical hidden sequence and historical observation sequence of the vehicle until the recall rate, precision rate, and F1 score of the trained CRF model all reach the preset threshold, obtaining the optimal parameters of the CRF model. The hidden sequence is the evaluation sequence of other vehicles on the broadcast traffic event information. The hidden sequence has true and false. True indicates that the vehicle approves the broadcast traffic event information, and false represents that the vehicle does not approve the broadcast traffic event information. Step 2-3: Input the optimal parameters of the CRF model and the observation sequence of the vehicle Vi into the CRF model to obtain the hidden sequence of other vehicles on the broadcast traffic event information, and obtain the trust value of the vehicle Vi using the hidden sequence. Step 2-4: The RSU closest to the vehicle Vi updates the trust threshold of the vehicle Vi using the trust value of the vehicle Vi. Specifically: First, set the vehicle trust threshold group (δ1, δ2,..δ j ...δ n ); Among them, (δ1, δ2,..δ j ., δ n ) is an increasing arithmetic progression, δ j is the j-th trust threshold, and n is the total number of trust thresholds; Then, the RSU closest to vehicle Vi compares the trust value of vehicle Vi with the current trust threshold. If the trust value of vehicle Vi is greater than or equal to the current trust threshold δ j , it indicates that the time information sent by the current vehicle is true information, and the current trust threshold is saved; if the vehicle trust value is less than the current trust threshold δ j , then it is judged whether the current trust threshold δ j is equal to δ n . If δ j =δ n , then the current vehicle is excluded from the vehicle networking. If the current trust threshold δ j is not equal to δ n , the current trust threshold is updated to δ j+1 ; Among them, the initial trust threshold of the vehicle is δ1. Step 3: Use the Raft consensus algorithm to select an RSU as the leader Leader. Determine whether the RSU closest to the vehicle Vi is the leader. If the RSU closest to the vehicle Vi is not the leader, no operation is performed. If the RSU closest to the vehicle Vi is the leader, the leader packs the trust value sent by the vehicle into a block and broadcasts the block to all RSUs. The RSUs verify the integrity of the fields in the block. If the verification passes, the RSUs transmit the audit results to each other. If the number of RSU audit results received by the leader is greater than the quantity threshold A, the block is submitted to the blockchain, thereby updating the vehicle trust value on the blockchain. If the number of RSU audit results received by the leader is less than the quantity threshold A, a new leader is re-elected.

2. The method for trust management of an Internet of Vehicles based on blockchain according to claim 1, wherein: The process of assigning a registration certificate to the vehicles in the vehicle networking and uploading the registration certificate to the blockchain in Step 1 is specifically: Step 1-1: After the vehicle enters the vehicle networking, a trusted third party TA assigns a registration certificate to the vehicle. Step 1-2: Upload the vehicle registration certificate to the blockchain: The vehicle assigned with the registration certificate generates a public-private key pair and sends a message containing real identity information to the TA. The TA determines whether the current vehicle identity is legal. If the current vehicle identity is illegal, the registration certificate of the current vehicle is revoked. If the current vehicle identity is legal, the current vehicle uses the registration certificate as a pseudonym and uploads the pseudonym to the certificate blockchain; If the difference between the expiration time of the current vehicle registration certificate and the current time is less than or equal to the preset time interval, the vehicle regenerates a new public-private key pair and sends the real identity information to the TA again. The TA updates the registration certificate for the current vehicle and uploads the updated registration certificate to the certificate blockchain.

3. The method for vehicle networking trust management based on blockchain according to claim 2, characterized in that: The TA determines whether the current vehicle identity is legal. Specifically, it determines whether the current vehicle registration certificate has expired. If it has not expired, it means it is legal; otherwise, it means it is illegal.

4. A vehicle networking trust management method based on blockchain according to claim 3, characterized in that: The CRF model, specifically: Among them, x is the observation sequence, y is the hidden sequence, Z(x) is the intermediate variable, i is the time label, k is the edge label, t k is the eigenvalue defined on the edge, s l is the eigenvalue defined on the node, l is the node label, λ k is the weight of the edge, μ l is the weight of the node, y i is the hidden value at the i-th time, y i-1 is the hidden value at the (i - 1)-th time.

5. The method for trust management of an Internet of Vehicles based on blockchain according to claim 4, wherein: In step 23, obtaining the trust value of vehicle Vi using the hidden sequence is specifically: f(y) = 1 / (1 + e -wy ) Among them, y is the hidden sequence, w is the trust value weight, and f(y) is the trust value of vehicle Vi.

6. The method for trust management of an Internet of Vehicles based on blockchain according to claim 5, wherein: The quantity threshold in the third step

Images