A vehicle credibility measurement method based on blockchain and recommendation trust
The vehicle credibility measurement method using blockchain and recommended trust addresses the credibility challenges in VANETs by authenticating vehicles with pseudonyms and updating trust values on a blockchain, enhancing information authenticity and privacy in VANETs.
Patent Information
- Application Number
- CN202211581607.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-08
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2042-12-08
AI Technical Summary
The challenge in Vehicular Ad Hoc Networks (VANETs) is the difficulty in quickly judging the credibility of interactive information due to high-speed vehicle movement, openness, and vulnerability, leading to the spread of false information and privacy breaches, which existing centralized and decentralized trust management models struggle to address effectively.
A vehicle credibility measurement method based on blockchain and recommended trust, involving direct and indirect trust calculations, where vehicles are authenticated with pseudonyms, and trust values are maintained and updated on a blockchain network through direct and recommended trust aggregation.
This method accurately identifies malicious vehicles, ensuring the authenticity of traffic information and privacy, thereby improving road network conditions and reducing human intervention in unmanned driving.
Smart Images

Figure CN115941332B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of intelligent transportation and blockchain technology, and specifically relates to a vehicle credibility measurement method based on blockchain and recommended trust. Background Art
[0002] The maturity of 5G technology has promoted the vigorous development of the Internet of Things and greatly facilitated people's lives. As a special type of the Internet of Things, Vehicular Ad Hoc Networks (VANETs) are expected to become an important data sensing and processing platform for intelligent transportation systems. In addition to providing drivers with a pleasant driving experience, VANETs play an important role in sharing real-time road information to alleviate traffic pressure and improve traffic efficiency. In VANETs, each vehicle is equipped with an on-board unit (OBU) for communicating with other entities, which also generates two types of communication, namely, vehicle-to-vehicle (V2V) communication and vehicle-to-infrastructure (V2I) communication.
[0003] Due to the high-speed movement of vehicles in the road network, the communication time between vehicles is too short to quickly judge the credibility of the interactive information. Due to the openness and vulnerability of VANETs, the authenticity of traffic information cannot be guaranteed, so it is difficult for vehicles on the road to determine the trust relationship. In addition, the low transmission cost further allows malicious or selfish vehicles to spread false information, which is very likely to affect normal road conditions and even cause traffic accidents. In addition, malicious vehicles on the road may track other vehicles to obtain private data for profit. At present, the Internet of Vehicles mainly relies on trust management models to solve the above problems. Existing trust management models are divided into two categories: centralized trust models and decentralized trust models.
[0004] A common feature of all centralized trust management models is that they all establish a central server to manage the trust value of vehicles. However, in addition to considering the high latency and inevitable single point failure in the multi-request state, the growing areas covered by intelligent transportation systems also mean that centralized trust management models are no longer applicable. Decentralized trust models often deploy the system in roadside units (RSUs), but there are situations where RSUs are attacked or hacked.
[0005] The emergence of blockchain technology provides a new idea for decentralized trust management models. In order to ensure the legitimacy and privacy security of vehicle identities in the road network, the design of the trust model also needs to include a mature identity authentication mechanism. Summary of the invention
[0006] The present invention proposes a vehicle credibility measurement method based on blockchain and recommended trust, which is used to solve the problems of vehicle identity authentication and mutual trust between vehicles in the Internet of Vehicles.
[0007] In order to achieve the above object, the present invention adopts the following technical scheme:
[0008] A vehicle credibility measurement method based on blockchain and recommendation trust includes the following steps:
[0009] Step 1. Initialize the information of each vehicle node registering to enter the Internet of Vehicles, including a pair of public and private keys and an initial trust value; the initial trust value is uploaded to the blockchain built by the trusted institution, roadside unit and vehicle node;
[0010] Step 2. The source vehicle node conducts trust evaluation on the nearby target vehicle nodes during the road network driving process, obtains the combined trust of the target vehicle nodes based on the aggregation of direct trust and recommended trust, and uploads the combined trust value to the roadside unit;
[0011] Step 3. Each roadside unit aggregates the received combined trust, packages it into a transaction and uploads it to the blockchain. The blockchain completes the trust value update of the vehicle node. The size of the trust value on the blockchain is used as a standard to measure the credibility of the vehicle node.
[0012] The present invention has the following advantages:
[0013] As described above, the present invention relates to a method for measuring the credibility of a vehicle based on blockchain and recommended trust, which involves the calculation of the trust value of a vehicle node in the Internet of Vehicles. Considering the openness and fragility of the Internet of Vehicles, the authenticity of traffic information on the road and the reliability of information transmission cannot be guaranteed, and the unmanned driving technology continues to mature, and human intervention in vehicles will be reduced. The present invention proposes a method for measuring the credibility of vehicle nodes. The trust value of a vehicle is based on direct trust and indirect trust. Direct interaction between vehicles can be used as a means of calculating direct trust values, including the authenticity verification of messages. Direct trust can be passed as second-hand information to surrounding vehicles for reference, which is the source of recommended trust. Since there are malicious vehicles that deliberately spread false recommended trust, the present invention provides a method for filtering recommended information. The final trust value of a vehicle node obtained by weighting the comprehensive direct trust and recommended trust is uploaded to the blockchain, and is continuously maintained and updated. The present invention can identify malicious vehicles on the road with high accuracy, which is conducive to improving the environment of the road network. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 This is the overall architecture diagram of the blockchain network built in the vehicle credibility measurement method based on blockchain and recommended trust.
[0015] Figure 2 Schematic diagram of obtaining recommended trust in the vehicle credibility measurement method based on blockchain and recommended trust.
[0016] Figure 3 Schematic diagram of the blockchain structure for storing trust values in the vehicle credibility measurement method based on blockchain and recommended trust.
[0017] Figure 4 Flowchart of the vehicle credibility measurement method based on blockchain and recommendation trust. DETAILED DESCRIPTION
[0018] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:
[0019] This embodiment describes a vehicle credibility measurement method based on blockchain and recommended trust to solve the technical problems of vehicle identity authentication and mutual trust between vehicles in the Internet of Vehicles. The general steps are as follows:
[0020] First, the identity of the vehicle entering the road network is authenticated and maintained at all times; according to the proposed method, the information transmission behavior of the vehicle on the road is cumulatively judged, and the trust value of the vehicle is continuously calculated and updated and maintained through the blockchain.
[0021] like Figure 1 As shown in the figure, a trust management system based on the blockchain network is first built. The blockchain network consists of four parts, namely the trusted institution TA, the roadside unit RSU, the vehicle node and the blockchain. The following are the introductions of each part:
[0022] Trusted Agency TA:
[0023] As a fully trusted entity in the trust management system, TA combines the functions of a registration authority and a certificate authority in the public key infrastructure. It can not only review and authenticate the identity of vehicle nodes applying for certificates, but is also responsible for the registration, renewal and revocation of all certificates.
[0024] When a vehicle node requests a certificate, it needs to provide real identity information to TA. TA records the mapping between the real identity of the vehicle node and the pseudonym represented by its certificate, which can provide assistance when the vehicle node commits illegal acts.
[0025] The present invention realizes identity authentication by introducing a public key infrastructure (PKI) certificate as a pseudonym (public key address), and uses the pseudonym to hide the communication ID during communication, thereby well protecting the privacy of the vehicle user.
[0026] When a vehicle first enters the network, it generates a pair of public and private keys and registers a digital certificate as its own identity and pseudonym. Due to the frequent use of certificates in message delivery and identity authentication, they need to be updated regularly to protect the unlinkability between the certificate and the real identity of the attacker. In addition, when the certificate is about to expire or the vehicle's private key is lost or leaked, a certificate update request must also be made. When a vehicle is determined to be a malicious vehicle by the trust management model, its certificate will be revoked and removed from the system.
[0027] The process of registering, renewing and revoking a certificate is as follows:
[0028] Certificate of Registration:
[0029] When a vehicle v enters the vehicle network for the first time, the system generates a pair of public and private keys for it. The public key and personal information need to be uploaded to the trusted agency TA to apply for a certificate
[0030] Among them, ID CA Indicates certificate ID, PU CA and Represent the public keys of the certificate authority (CA) and vehicle v, respectively, Sig CA and Sig alg They represent the CA’s digital signature on the certificate and the corresponding signature algorithm, T val Indicates the validity period of the certificate.
[0031] Certificate Update:
[0032] Due to the frequent use in messaging and identity authentication, digital certificates need to be updated frequently to protect the unlinkability between the certificate and the real identity of the attacker. When the certificate is about to expire or the private key is lost, a certificate update request must be submitted to the TA. After the identity authentication is passed, when the vehicle chooses to revoke the original certificate, the old certificate will be added to the certificate revocation list.
[0033] Assuming that the certificate of vehicle v needs to be updated, it needs to generate a new pair of public-private keys, and then request the trusted authority TA to update the certificate through a secure channel; the trusted authority TA first confirms that vehicle v has not been judged as a malicious vehicle by the system and the certificate has been revoked, and then obtains the length of time the certificate has been used based on the generation time of the current certificate to prevent malicious vehicles from sending too many update requests in a short period of time and causing waste of resources. If the above conditions are met, a new certificate will be issued.
[0034] Certificate Revocation:
[0035] The trust management system will continuously update the trust value of each vehicle on the blockchain based on the vehicle credibility measurement method based on blockchain and recommended trust. TVThe vehicle will be regarded as a malicious vehicle and its certificate will be revoked and recorded in the certificate revocation list.
[0036] Roadside Unit RSU:
[0037] In VANETs, roadside units (RSUs) or roadside units (RSUs) and vehicle nodes frequently communicate with each other to exchange real-time road information and expand the communication range.
[0038] In the present invention, RSU is responsible for information transmission between vehicles and collecting road information, such as trust evaluation information between vehicles, speed and location information of vehicles, etc. The specific computing and storage capabilities of roadside units RSU support them to perform a series of edge computing. In addition, roadside units RSU jointly maintain a blockchain that records the trust values of vehicle nodes.
[0039] Vehicle Node:
[0040] Each vehicle node in the environment is equipped with an on-board unit that can process real-time traffic information collected by sensors and share information with nearby RSUs or other vehicle nodes. When a vehicle node first enters the network, it needs to apply for a certificate from a trusted authority TA to obtain a legitimate identity. The certificate can be used as a pseudonym to protect identity privacy.
[0041] Blockchain:
[0042] The blockchain is maintained by the roadside unit RSU and records the global trust value of the vehicle node.
[0043] The vehicle node only has read access to the blockchain, and the consensus process is completed by the roadside unit RSU. Due to the immutability of the blockchain, the trust value of the vehicle node can be obtained by querying the recent transaction records under a pseudonym.
[0044] The blockchain structure for storing trust value is as follows Figure 3 As shown, each block contains a block header and a block body.
[0045] Among them, the block header encapsulates information such as version number, timestamp, hash value of the previous block, etc., while the block body records detailed transaction data and stores the trust value of the vehicle node in the form of key-value values.
[0046] like Figure 4 As shown, the vehicle credibility measurement method based on blockchain and recommendation trust is described in detail below.
[0047] A vehicle credibility measurement method based on blockchain and recommendation trust includes the following steps:
[0048] Step 1. Initialize the information for each vehicle node to register and enter the Internet of Vehicles, including a pair of public and private keys and an initial trust value; the initial trust value is uploaded to the blockchain network built by the trusted organization, roadside units and vehicle nodes.
[0049] Step 2. The source vehicle node conducts a trust evaluation on the nearby target vehicle nodes while traveling on the road network, obtains the combined trust of the target vehicle nodes based on the aggregation of direct trust and recommended trust, and uploads the combined trust value to the roadside unit.
[0050] The interaction between vehicles in the road network will be used as the basis for calculating trust values. Each vehicle will be evaluated by nearby vehicles, and these evaluations will be uploaded to the roadside unit RSU and eventually updated to the blockchain in the form of transactions.
[0051] First, the source vehicle node interacts directly with the target vehicle, and the direct trust value of the source vehicle node to the target vehicle is obtained based on the result of the interaction, which includes the authenticity judgment of the transmitted information.
[0052] Then, the source vehicle node collects the direct trust values of the nearby intermediate vehicles for the target vehicle to form the recommended trust value of the source vehicle node for the target vehicle node. Since there may be deliberate low-evaluation behaviors or high-evaluation behaviors of accomplices among the intermediate vehicles, the present invention proposes a recommended trust filtering algorithm to filter improper evaluations, and the vehicles to which the screened evaluations belong form a credible cluster. These evaluations are then corrected for similarity to finally obtain the recommended trust value.
[0053] Finally, the source vehicle node obtains the combined trust by combining the direct trust and recommended trust of the target vehicle node, and transmits the combined trust value to the roadside unit RSU for aggregation. After the blockchain consensus protocol, the evaluation results will be packaged into transactions and uploaded to the blockchain to complete the update of the global trust of the target vehicle.
[0054] In step 2, the process of trust evaluation of the source vehicle node on the target vehicle node is as follows:
[0055] Step 2.1. The source vehicle node verifies the authenticity of the information sent by the target vehicle node, and the verification result is used to calculate the direct trust value of the source vehicle node to the target vehicle node.
[0056] In VANETs, vehicles broadcast messages to each other to share road information such as traffic conditions, leading to direct interactions between them. The authenticity of the messages determines the outcome of the direct interactions.
[0057] Definition (α ij ,β ij ) represents the result of direct interaction between source vehicle node i and target vehicle node j at time t. ij and β ijThey represent the accumulation of positive and negative interactions experienced by source vehicle node i with respect to target vehicle node j, respectively.
[0058] Considering the cold start problem where the source vehicle node i and the target vehicle node j have never interacted, the initial trust value is set to 0.5. The direct trust value of the source vehicle node i to the target vehicle node j based on Beta distribution
[0059] The result of the message authenticity determination determines the result of the direct interaction between the source vehicle node i and the target vehicle node j.
[0060] When the target vehicle node j observes the occurrence of event x somewhere, it broadcasts the message to nearby vehicles including the source vehicle node i. At the same time, the source vehicle node i may receive other messages about event x.
[0061] First, receiver i adds the received message about event x to the message set Among them, Represents the message about event x reported by target vehicle node j.
[0062] Considering the trust value of the target vehicle node j and its distance from event x, the credibility of event x conveyed by j is for:
[0063]
[0064] Among them, TV j is the trust value of the target vehicle node j on the blockchain, e represents a natural constant, represents the distance between j and event x, γ and k are two parameters that control the lower limit of the rate of change and the authenticity of the message respectively.
[0065] Next, we get the credibility set of all messages about event x
[0066] The truth or falsity of event x is determined by the formula Make a judgment.
[0067] Among them, x and are complementary events, P(x) is the prior probability of event x, N is the message set M x The number of elements in .
[0068] If P(x|C x )>Thr occur , then the event is considered true, otherwise the event is considered false, where Thr occuris the preset threshold; if the time is real, it is considered that the source vehicle node i and the target vehicle node j have a positive interaction, otherwise it is a negative interaction.
[0069] Considering that a vehicle with a high trust value in the past does not mean that its credibility is also high now, the past experience should decrease over time, so the time decay function f(x) = e is introduced -λx .
[0070] In addition, in order to save the interaction records between vehicles, each vehicle node maintains a table IR, which only stores the results of the most recent M interactions. After combining the time decay function and the records in IR, at time t (α ij ,β ij ) is calculated as:
[0071]
[0072]
[0073] in, and They represent the accumulation of positive and negative interactions between the source vehicle node i and the target vehicle node j at time t, respectively.
[0074] M represents the number of recent interaction results between node i and node j stored in source vehicle node i, q∈[1,M], λ represents the time decay factor, and controls and The rate of change of attenuation; Represents the introduced time decay function.
[0075] and They represent the source vehicle node i and the target vehicle node j at time t q The direct trust of source vehicle node i to target vehicle node j at time t is defined as The calculation formula is:
[0076] Step 2.2. The source vehicle node collects the direct evaluations of the target vehicle node from nearby neighbor nodes, and filters the evaluations using a recommended trust filtering algorithm to form a recommended trust of the source vehicle node for the target vehicle node.
[0077] Source vehicle node i obtains the recommendation trust for vehicle node j by collecting the direct trust of nearby intermediate vehicles on target vehicle node j. The evaluations given by intermediate vehicle nodes need to be filtered because the information they provide is not always credible.
[0078] If the target vehicle node j behaves maliciously, the intermediate vehicle node may deliberately discredit the target node’s behavior. On the contrary, if the recommended vehicle node has a cooperative relationship with the target node, it will provide a high rating.
[0079] like Figure 2 As shown, the source vehicle node i at time t is selected from the recommendation set R ij = {r1, r2, ...} collects the recommended information about the target vehicle node j, where the recommended set R ij The elements in represent the common neighboring vehicle nodes of the source vehicle node i and the target vehicle node j. Next, the recommended trust filtering algorithm is used to obtain the trusted cluster. The specific process is as follows:
[0080] First, for the recommendation set R ij Each element in constructs a vector Among them, r f represents the fth common neighbor vehicle between source vehicle node i and target vehicle node j, f = 1, 2, ...; Represents the source vehicle node i to the neighbor vehicle r f The direct trust value of Represents the neighbor vehicle node r f The direct trust value of the target vehicle node j.
[0081] Then, from the recommended set R ij Randomly select k vectors as the initial centroid of the cluster, for each vector Calculate the Euclidean distance of each vector to each centroid separately and add the vector to the nearest cluster.
[0082] The centroid of each cluster is iteratively updated according to the newly divided clusters until all centroids are no longer updated or the number of iterations reaches the preset maximum value, then all vectors will be divided into k clusters, and finally the cluster with the largest horizontal coordinate is selected as the credible cluster.
[0083] The recommended trust provided by the recommended node in the trusted cluster will be retained and adopted. Considering that in a complex environment, vehicles will inevitably produce inaccurate trust evaluations. Specifically, the source vehicle node may generate inaccurate direct trust for the target node. In addition, due to the small number of interactions, the recommended vehicle node may also generate inaccurate direct trust for the target vehicle node. Therefore, the embodiment of the present invention introduces two factors combined with the trusted cluster to improve the accuracy of the recommended trust.
[0084] First, we consider that the recommendation trust provided by vehicles with high trust values recorded on the blockchain is more credible, and then we introduce similarity. Define Sim(i,r′ F ) represents the similarity of the direct evaluation of the common target vehicle by two vehicles, and the calculation formula is:
[0085]
[0086] Among them, r′ F is an element in a credible cluster, Set(i,r′ F ) is the source vehicle node i and vehicle node r′ F The set of target vehicle nodes that have been evaluated, x is the set of target vehicle nodes Set(i,r′ F ) in the target vehicle.
[0087] represents the direct trust value of the source vehicle node i to the target vehicle node x, Represents the vehicle node r′ F The direct trust value of the target vehicle node x, Num(Set(i,r′ F )) represents the number of elements in the collection.
[0088] After obtaining the trusted cluster using the recommendation trust filtering algorithm, the evaluation provided by the vehicles in the trusted cluster is integrated with the above factors to obtain the recommended trust of the source vehicle node i for the target vehicle node j. The calculation formula is as follows:
[0089]
[0090] Where N represents the number of elements in the credible cluster, F∈[1,N], is the vehicle node r′ F The trust value recorded on the blockchain, Denotes the recommended vehicle node r′ F Direct trust in the target vehicle node j.
[0091] Step 2.3. The source vehicle node aggregates the direct trust and the recommended trust to obtain the combined trust of the target vehicle node.
[0092] Both direct trust and recommended trust calculation methods have certain limitations.
[0093] For direct trust, the source node needs to interact efficiently and directly with the destination node to obtain more accurate values, otherwise it is difficult to reflect the current credibility of the target node only through historical information.
[0094] For recommendation trust, the negative impact of malicious nodes such as inaccurate evaluations is inevitable.
[0095] Therefore, the present invention combines direct trust with recommended trust to form a combined trust on the target node, thereby improving the accuracy of the evaluation. The combined trust of the source vehicle node i on the target vehicle node j is The calculation formula is as follows:
[0096] Among them, ω is the weight, ω∈(0,1).
[0097] Step 3. Each roadside unit aggregates the recommended trusts received, packages them into transactions and uploads them to the blockchain. The blockchain completes the trust value update of the vehicle node. The size of the trust value on the blockchain is used as a standard to measure the credibility of the vehicle node.
[0098] Assume that at time t l and t l+1 Between, the set I={i1,i1,…,i p The vehicle in} uploads the combined trust of the target vehicle node j to the roadside unit, and the roadside unit aggregates all the combined trust values to obtain the final trust value:
[0099]
[0100] in, Represents the final trust value of vehicle node j at time l+1. represents the trust value of the target vehicle node j that has recently appeared on the blockchain, p represents the number of vehicles in set I, and i n represents the nth vehicle node in set I, n∈[1,p], represents vehicle node i at time l+1 n Combined trust in vehicle node j.
[0101] Each time the final trust value of the target vehicle node j is updated, it is necessary to determine whether the final trust value of the target vehicle node j is less than the preset trust value threshold. If the final trust value of the target vehicle node j is less than the preset trust value threshold, it will be judged as a malicious vehicle. The roadside unit aggregates the final trust value calculated Upload to the blockchain.
[0102] The present invention solves the problem of mutual trust between vehicles through the above-mentioned vehicle credibility calculation method. Users judged as malicious vehicles will be kicked out of the entire network to reduce the impact of these vehicles on the normal operation of road traffic.
[0103] Of course, the above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiments. It should be noted that all equivalent substitutions and obvious deformation forms made by any technician familiar with the field under the guidance of this specification fall within the essential scope of this specification and should be protected by the present invention.
Claims
1. A vehicle credibility measurement method based on blockchain and recommendation trust, characterized in that: The steps include: Step 1. Initialize the information of each vehicle node registering to enter the Internet of Vehicles, including a pair of public and private keys and an initial trust value; the initial trust value is uploaded to the blockchain network built by the trusted institution, roadside units and vehicle nodes; Step 2. The source vehicle node performs trust evaluation on the nearby target vehicle node during the road network driving process, obtains the combined trust value of the target vehicle node based on direct trust and recommended trust aggregation, and uploads the combined trust value to the roadside unit; Step 3. Each roadside unit aggregates the received combined trust values, packages them into transactions and uploads them to the blockchain. The blockchain completes the trust value update of the vehicle node. The size of the trust value on the blockchain is used as a standard to measure the credibility of the vehicle node. In step 2, the process of trust evaluation of the source vehicle node on the target vehicle node is as follows: Step 2.
1. The source vehicle node verifies the authenticity of the information sent by the target vehicle node, and the verification result is used to calculate the direct trust value of the source vehicle node to the target vehicle node; Step 2.
2. The source vehicle node collects the direct evaluations of the nearby neighbor nodes on the target vehicle node, and filters the evaluations using the recommended trust filtering algorithm to form the recommended trust of the source vehicle node on the target vehicle node; Step 2.
3. The source vehicle node aggregates the direct trust and the recommended trust to obtain a combined trust value for the target vehicle node; In step 2.1, the calculation process of direct trust is as follows: definition It represents the result of direct interaction between source vehicle node i and target vehicle node j at time t. The calculation formulas are as follows: in, and They represent the accumulation of positive and negative interactions between source vehicle node i and target vehicle node j at time t, respectively; M represents the number of recent interaction results between node i and node j stored in source vehicle node i, q∈[1,M]; λ represents the time decay factor, which is used to control and The rate of change of attenuation; represents the introduced time decay function; and They represent the source vehicle node i and the target vehicle node j at time t q The accumulation of positive and negative interactions at time t; the direct trust of source vehicle node i on target vehicle node j at time t is defined as The calculation formula is: In step 2.2, the process of recommending the trust filtering algorithm is as follows: The source vehicle node i at time t is selected from the recommendation set R ij = {r1, r2, ...} collects the recommended information about the target vehicle node j, where the recommended set R ij The elements in represent the common neighbor vehicles of the source vehicle node i and the target vehicle node j; Next, we use the recommended trust filtering algorithm to obtain the trusted cluster. The specific process is as follows: First, for the recommendation set R ij Each element in constructs a vector Among them, r f represents the fth common neighbor vehicle between source vehicle node i and target vehicle node j, f = 1, 2, ...; Represents the source vehicle node i to the neighboring vehicle node r f The direct trust value of Represents the neighbor vehicle node r f Direct trust value of target vehicle node j; Then, from the recommended set R ij Randomly select k vectors as the initial centroid of the cluster, for each vector Calculate the Euclidean distance of each vector to each centroid separately and add the vector to the nearest cluster; The centroid of each cluster is iteratively updated according to the newly divided clusters until all centroids are no longer updated or the number of iterations reaches the preset maximum value, then all vectors will be divided into k clusters, and finally the cluster with the largest horizontal coordinate is selected as the credible cluster; Define Sim(i,r′ F ) represents the similarity of the direct evaluation of the common target vehicle node by two vehicle nodes, and the calculation formula is: Among them, r′ F is an element in a credible cluster, Set(i,r′ F ) is the source vehicle node i and vehicle node r′ F The set of target vehicles that have been evaluated, x is the set of target vehicle nodes Set(i,r′ F ) in the target vehicle node; represents the direct trust value of the source vehicle node i to the target vehicle node x, Represents the vehicle node r′ F The direct trust value of the target vehicle node x, Num(Set(i,r′ F )) represents the number of elements in the collection; After obtaining the trusted cluster using the recommended trust filtering algorithm, the recommended trust of the source vehicle node i for the target vehicle node j is obtained. The calculation formula is as follows: Where N represents the number of elements in the credible cluster, F∈[1,N], is the vehicle r′ F The trust value recorded on the blockchain, Denotes the recommended vehicle node r′ F Direct trust in the target vehicle node j.
2. The vehicle credibility measurement method based on blockchain and recommendation trust according to claim 1 is characterized in that: In step 2.3, the calculation process of the combined trust value is as follows: The combined trust value of source vehicle node i to target vehicle node j is The calculation formula is as follows: Among them, ω is the weight, ω∈(0,1).
3. The vehicle credibility measurement method based on blockchain and recommendation trust according to claim 1 is characterized in that: In step 3, the trust value update process of the vehicle node is as follows: Assume that at time t l and t l+1 Between, the set I={i1,i1,…,i p The vehicle in} uploads the combined trust value of the target vehicle j to the roadside unit, and the roadside unit aggregates all the combined trust values to obtain the final trust value: in, represents the final trust value of vehicle node j at time l+1; represents the trust value of the target vehicle j that has recently appeared on the blockchain, p represents the number of vehicles in set I, and i n represents the nth vehicle in set I, n∈[1,p], represents vehicle node i at time l+1 n The combined trust value of vehicle node j; Each time the final trust value of the target vehicle j is updated, it is necessary to determine whether the final trust value of the target vehicle j is less than the preset trust value threshold. If the final trust value of the target vehicle j is less than the preset trust value threshold, it will be judged as a malicious vehicle; The roadside unit calculates the final trust value of the target vehicle j through aggregation Upload to the blockchain.
4. The vehicle credibility measurement method based on blockchain and recommendation trust according to claim 1 is characterized in that: In the blockchain network, the trusted institutions, roadside units and vehicle nodes are as follows: As a fully trusted entity in the blockchain network, the trusted authority has the functions of both the registration authority and the certificate authority in the public key infrastructure. It reviews and authenticates the identity of the vehicle node applying for the certificate, as well as registers, updates, and revokes all certificates. When a vehicle node requests a certificate from a trusted institution, it needs to provide real identity information to the trusted institution. The trusted institution records the mapping between the real identity of the vehicle node and the pseudonym represented by its certificate. Roadside units are responsible for information transmission between vehicles and collecting road information, including trust evaluation information between vehicles, vehicle speed and location information; roadside units jointly maintain a blockchain that records the trust values of vehicle nodes; Each vehicle node is equipped with an on-board unit that shares information with nearby roadside units or other vehicle nodes.
5. The vehicle credibility measurement method based on blockchain and recommendation trust according to claim 4 is characterized in that: The process of registering, renewing and revoking a certificate is as follows: Certificate of Registration: When a vehicle v enters the blockchain network for the first time, the system generates a pair of public and private keys for it. Public key and personal information need to be uploaded to a trusted organization to apply for a certificate Among them, ID CA Indicates certificate ID, PU CA and Represent the public keys of the certificate authority CA and vehicle v, respectively, Sig CA and Sig alg They represent the CA’s digital signature on the certificate and the corresponding signature algorithm, T val Indicates the validity period of the certificate; Certificate Update: When the certificate is about to expire or the private key is lost, submit a certificate update request to the trusted authority. After the identity authentication is passed, when the vehicle chooses to revoke the original certificate, the old certificate will be added to the certificate revocation list; Certificate Revocation: The system will continuously update the trust value of each vehicle on the blockchain based on the vehicle credibility measurement method based on blockchain and recommended trust. When the trust value is less than the preset trust value threshold, the vehicle will be regarded as a malicious vehicle and its certificate will be revoked.
Citation Information
Patent Citations
Internet of vehicles trust management method for resisting collusion attack based on block chain
CN115022883A
Distributed Internet of Vehicles dynamic trust management method based on origin information
CN115174615A
Privacy blockchain-based internet of vehicles protection method, and mobile terminal
WO2022082893A1
