A data sharing method based on blockchain and incentive mechanism in Internet of Vehicles

Abstract

The present invention belongs to the technical field of vehicle network data transmission, and specifically relates to a data sharing method based on blockchain and incentive mechanism in the vehicle network; the method comprises: constructing an in-vehicle social network system architecture, in which vehicles share data according to the incentive mechanism, and the process comprises: constructing a benefit function of cluster members; constructing a benefit function of cluster heads; constructing a benefit function of base stations; using a three-party four-stage Stackelberg game iterative algorithm to solve the benefit functions of cluster members, cluster heads and base stations, and obtain a data sharing decision that maximizes the system benefit; vehicles share data according to the data sharing decision; the present invention makes the user experience quality of cluster member vehicles gradually tend to be stable, and at the same time considers content popularity and node selfishness for dynamic pricing, has a good incentive effect on cluster head vehicles, and can provide better service quality for cluster member vehicles.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle networking data transmission, and specifically relates to a data sharing method based on blockchain and incentive mechanism in a vehicle networking. Background Art

[0002] With the rapid development of autonomous driving technology, the vehicular ad hoc network (VANET) in the intelligent transportation system can significantly improve system security and user satisfaction, and has therefore received extensive attention and research. In VANET, vehicles equipped with wireless interfaces can directly communicate with nearby vehicles through vehicle-to-vehicle (V2V) communication; at the same time, they can also communicate with fixed roadside units (RSU), which is called vehicle-to-infrastructure (V2I) communication. When vehicles move at high speeds, the topology continues to change. In V2V and V2I communication modes, there is a problem of unstable or even interrupted transmission links, which has become a bottleneck problem that needs to be solved urgently for application-oriented VANET. At the same time, with the gradual realization of smart cities, the study and analysis of human behavior and preferences in urban communication networks is particularly important, and the same is true for the development of the Internet of Vehicles.

[0003] Online social networks are the most popular application scenarios of social networks. The most common application in social networks is behavioral analysis and prediction based on social data collection. When users log in to applications such as social software or search engines, the application that provides data services will collect information such as login account username, login time, and real-time location through the social network data integration module. After that, this data will be authorized for use by social media software. These applications will classify and analyze user data by adding tags and other methods, and make "personalized recommendations" for different types of users. Considering only the above classification and analysis methods, it is no longer possible to meet the personalized recommendations and security guarantees for the target groups of data sharing in the future Internet of Vehicles.

[0004] Nowadays, blockchain technology is evolving towards high efficiency, security and convenience. Cross-chain technology, extension technology and core technology are strong supports for development. While core technologies represented by Blockchain Transmission Network (BTN), consensus algorithm and smart contract framework are developing steadily, blockchain extension technologies such as cloud chain integration, privacy computing, Internet of Things and digital identity are developing rapidly. Therefore, in recent years, domestic and foreign scholars have introduced blockchain technology into the Internet of Vehicles.

[0005] There are problems such as node privacy information leakage and unstable data links between nodes in the Internet of Vehicles system. Using incentive mechanisms to optimize the system can solve the above problems to a certain extent. Incentive mechanism refers to the process of maximizing the interests of all parties through specific methods and management systems. In the traditional incentive mechanism, the use of a unified average price for incentives is the most common way. However, this scheme has the disadvantage of insufficient incentives. A dynamic pricing incentive mechanism takes into account the popularity of shared content, content size, link quality and link interruption probability, and combines caching strategies to improve the content delivery rate and hit rate of the system. A method that combines caching strategies with incentive mechanisms and establishes a dynamic game model solves the problem of conflict of interests between operators and mobile users. However, none of the above incentive mechanisms takes selfishness into account in the game, only evaluates the participation of vehicle users from the side, and cannot solve problems such as node privacy leakage and unstable data links between nodes.

[0006] To sum up, in order to solve the above problems, an in-vehicle social network fusion architecture assisted by blockchain technology is adopted. In the process of data sharing, the selfishness of vehicles is taken into consideration, and a more practical incentive mechanism is designed to better meet the rapid development of the current Internet of Vehicles. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention proposes a data sharing method based on blockchain and incentive mechanism in the Internet of Vehicles, the method comprising: constructing an in-vehicle social network system architecture, including a physical layer, a social relationship layer and a blockchain layer; in the in-vehicle social network system architecture, vehicles share data according to the incentive mechanism; wherein the vehicles include cluster head vehicles and cluster member vehicles;

[0008] The process of vehicles sharing data according to the incentive mechanism includes:

[0009] Cluster members construct their benefit function based on the user experience quality and the unit price of purchased content;

[0010] The cluster head constructs the benefit function of the cluster head according to the reward amount issued by the base station to the cluster head, the energy consumption of the cluster head and the computing power loss of the cluster head;

[0011] The base station constructs the benefit function of the base station based on the amount of reward issued by the base station to the cluster head, the unit price of the content purchased by the cluster member and the energy consumption of the base station;

[0012] The three-party four-stage Stackelberg game iterative algorithm is used to solve the benefit functions of cluster members, cluster heads and base stations, and obtain the data sharing decision that maximizes the system benefit.

[0013] Vehicles share data based on data sharing decisions.

[0014] Preferably, the process of constructing the benefit function of cluster members includes: calculating the user experience quality of cluster members based on content popularity, vehicle selfishness and data packet transmission rate; calculating the unit price of cluster members purchasing content based on content cost, content popularity and vehicle selfishness; and calculating the benefit function of cluster members based on user experience quality and the unit price of purchasing content.

[0015] Furthermore, the formula for calculating the benefit function of cluster members is:

[0016]

[0017] Us CM ≥0

[0018]

[0019] Among them, U CM represents the benefit of cluster members, QoE represents the user experience quality of cluster members, m(x) represents the unit price of cluster members to purchase content x, ω x represents the size of the content x, k represents the unit size of the vehicle transmission data packet, PDR x,i Indicates vehicle v i Packet transmission rate when transmitting content x, s x,i Indicates vehicle v i The selfishness when forwarding content x, r(x) represents the popularity ranking of content x, X represents the number of contents, c(x) represents the cost of content x, r min represents the minimum content ranking, r max Indicates the maximum content ranking, Represents a collection of content.

[0020] Preferably, the process of constructing the benefit function of the cluster head includes: calculating the amount of reward issued by the base station to the cluster head according to the unit price of the content purchased by the cluster members and the selfishness of the vehicle; calculating the energy consumption overhead of the cluster head according to the downlink transmission rate of the vehicle, the transmission power of the vehicle and the size of the content; calculating the computing power loss of the cluster head according to the computing power of the vehicle, the selfishness of the vehicle and the amount of computing resources required for transmitting the content; calculating the benefit function of the cluster head according to the amount of reward issued by the base station to the cluster head, the energy consumption overhead of the cluster head and the computing power loss of the cluster head.

[0021] Furthermore, the formula for calculating the benefit function of the cluster head is:

[0022]

[0023] Us CH ≥0

[0024]

[0025] Among them, U CHrepresents the benefit of the cluster head, p(x) represents the reward amount issued by the base station to the cluster head, and E CH represents the energy consumption of the cluster head, E cost represents the computing power loss of the cluster head, m(x) represents the unit price of the content x purchased by the cluster member, and s x,i Indicates vehicle v i The selfishness when forwarding content x, u represents the selfishness influencing factor, p i Indicates vehicle v i The transmission power, ω x Indicates the size of content x, R i Indicates vehicle v i The downlink transmission rate, θ represents the sparse energy consumption, F i Indicates vehicle v i The computing power of cy x represents the amount of computing resources required to process the content ranked r(x), s x,imin Indicates vehicle v i The minimum selfishness when forwarding content x, s x,imax Indicates vehicle v i The maximum selfishness when forwarding content x, Represents a collection of content.

[0026] Preferably, the process of constructing the benefit function of the base station includes: calculating the energy consumption overhead of the base station based on the base station transmission energy consumption rate and the base station downlink transmission rate; calculating the benefit function of the base station based on the reward amount issued by the base station to the cluster head, the unit price of the content purchased by the cluster members and the energy consumption overhead of the base station.

[0027] Furthermore, the formula for calculating the benefit function of the base station is:

[0028] U BS =m(x)-p(x)-E BS

[0029] =(1-ln(s x,i +u))·m(x)·-R s,d ·l

[0030] =(1-ln(s x,i +u))·c(x)·s x,i ·(r(x) / X)-R s,d ·l

[0031] Us BS >0

[0032]

[0033] Among them, U BSrepresents the efficiency of the base station, m(x) represents the unit price of the content x purchased by the cluster member, p(x) represents the reward amount issued by the base station to the cluster head, and E BS represents the energy consumption of the base station, s x,i Indicates vehicle v i The selfishness when forwarding content x, u represents the selfishness influencing factor, R s,d represents the downlink transmission rate of the base station, l represents the transmission energy consumption rate of the base station, c(x) represents the cost of content x, r(x) represents the popularity ranking of content x, X represents the number of contents, c min represents the minimum cost of content x, c max represents the maximum cost of content x, Represents a collection of content.

[0034] Preferably, the process of vehicles sharing data according to the incentive mechanism also includes: before data sharing, the vehicle generates a metadata index and packages the index and sends it to the base station; the base station receives the metadata index, starts the information sharing smart contract and reaches a consensus; wherein the metadata index includes a timestamp, data information description, data owner information, storage address, information index and historical sharing record.

[0035] The beneficial effects of the present invention are as follows: based on the blockchain-assisted VSN architecture, the present invention aims at the problem of selfish behavior of rational data forwarders in the process of social vehicle data sharing, and designs a price incentive method to realize the data sharing process of social vehicles; under the premise of ensuring the stability of the cluster, the security of the data is effectively guaranteed, and the personalization of data sharing is correspondingly improved, so that the user experience quality QoE of the cluster member vehicle CM gradually tends to be stable, and at the same time, the content popularity and node selfishness are considered for dynamic pricing, which has a better incentive effect on the cluster head vehicle CH, and can provide CM with better service quality, with high practicality and good application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic diagram of the corresponding relationship between the social network and the vehicle network in the present invention;

[0037] Figure 2 This is a schematic diagram of a common in-vehicle social network architecture;

[0038] Figure 3 This is a schematic diagram of the in-vehicle social network system architecture of the present invention;

[0039] Figure 4 It is a schematic diagram of the game model in the present invention;

[0040] Figure 5 This is a simulation diagram of the impact of the content ranking strategy in the present invention on the user experience quality;

[0041] Figure 6 This is a simulation diagram of the impact of the content ranking strategy in the present invention on the effect function of cluster members. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] For the in-vehicle social network architecture, social attributes are very important for establishing stable links. Therefore, mining the corresponding relationship between the Internet of Vehicles and social networks is an important part of establishing the in-vehicle social network architecture. Figure 1 As shown in the figure, social relations, interests and activities are the links that connect the two network architectures. First, the topological structure in the social network is maintained and extended by social relations, and the topological structure can also be predicted based on social relations. Then, the system can perform operations such as information recommendation based on the interests of users in the social network. The diverse interests of different users have also spawned various user-centric applications, which have promoted the development of network applications. Finally, by mining the social behavior patterns of mobile users, communities with the same interests can be discovered. These data with specific labels can also help people model and study different modes of mobility.

[0044] Based on the above correspondence, the architecture of the in-vehicle social network is as follows: Figure 2 As shown in the figure, from bottom to top, the physical vehicles in the Internet of Vehicles are mapped to nodes in the social network layer and the relationship between nodes. In the Internet of Vehicles, vehicles with similar locations are divided into the same group framed by dotted lines. However, as can be seen from the figure, vehicles with similar physical locations may have different interests. As shown in the top-level social network, people with the same interests or social relationships are divided into 3 groups. From this architecture diagram, we can clearly see the advantage of the in-vehicle social network, that is, the physical location attributes of the vehicle entity and the social attributes between the vehicle driver or passengers can be reflected in one network at the same time.

[0045] In order to realize humanized clustering of social vehicles, incentivize selfish vehicle data transmission, and ensure secure and reliable storage and sharing of vehicle privacy information, the present invention proposes a data sharing method based on blockchain and incentive mechanism in the Internet of Vehicles, the method comprising: constructing an in-vehicle social network architecture, in which vehicles share data according to the incentive mechanism.

[0046] like Figure 3As shown, the in-vehicle social network architecture of the present invention is divided into three layers, namely: physical layer, social relationship layer and blockchain layer, involving four types of roles, namely Trusted Authority (TA), vehicle user (VU), base station and blockchain.

[0047] The description of the in-vehicle social network system architecture in the present invention is as follows:

[0048] (1) Physical layer: a) Vehicles act as communication nodes and interact with vehicles or base stations by establishing V2V and V2I links; b) Base stations act as roadside communication units to provide information resources for vehicles; and they also act as distributed databases to store data such as the privacy status of vehicles. In the in-vehicle social network, two social attributes of the vehicle, points of interest and activity, are taken into account in the system. In order to achieve global queryability of vehicle information, a blockchain layer maintained by the base station is constructed. Among them, vehicles include cluster head vehicles and cluster member vehicles.

[0049] (2) Social relationship layer: It is composed of virtual nodes mapped by vehicles. After the vehicles in the physical layer trigger the clustering process, the nodes in the social relationship layer will be mapped dynamically and in real time, intuitively showing the influence and role of social similarity and activity in the cluster head election process.

[0050] (3) Blockchain layer: It is composed of blockchain nodes, which are maintained by the base station. The private information of the vehicle, such as clustering and status, is encrypted by the cluster head and sent to the base station. The base station stores it in the local database and adds an index. In order to achieve global queryability of vehicle information, the base station reaches a node consensus through storage proof, stores the index of the encrypted content in a new block, and then forms a blockchain. The implementation of global queryability is divided into three steps: first, the vehicle sends a query request to the base station; then, the base station accesses the index stored in the blockchain through a smart contract; finally, the base station queries and obtains data according to the obtained index information.

[0051] The roles of the in-vehicle social network architecture are described as follows:

[0052] (1) Trusted Center: An absolutely trusted institution. During the system initialization phase, TA will register the identities of vehicle users and base stations and issue keys. After initialization, TA will remain offline to avoid affecting the decentralization of the blockchain. When malicious behavior occurs (for example, a malicious driver sends forged data), TA will track the sender of the malicious data and revoke its identity authentication.

[0053] (2) Vehicle users: The smallest communication unit in the in-vehicle social network, which can be divided into three states: cluster head, cluster member, and undefined vehicle. Multimedia data such as text, pictures, audio and video are exchanged between vehicle users and base stations. Cluster heads, cluster members and base stations trade data through a pricing plan.

[0054] (3) Base stations: Distributed at intervals along the roadside, they are complete nodes that store all block data. The information index during the transaction process will be stored on the blockchain by the base station, and a large amount of transaction data will be stored in the equipped Mobile Edge Computing (MEC) server.

[0055] (4) Blockchain: Blockchain can be viewed as a decentralized and reliable platform that ensures data integrity. Registration information of various entities will be published on the blockchain to ensure that vehicle users and base stations can easily query the information stored on the blockchain. Transaction index information is stored in the form of a chain to ensure the authenticity, validity and immutability of shared data.

[0056] In the in-vehicle social network architecture, the process of vehicle data sharing based on the incentive mechanism includes the following:

[0057] The system is first initialized, and various entities (base stations, vehicle users) will register their identities with TA to become legal entities, and distribute keys, including public and private keys for identity authentication;

[0058] Due to the limited storage space, the vehicle user will first transmit some road condition information, entertainment information around the road, and data of interest to the nearby base station. At this time, the data is anonymous and encrypted, and the digital signature of the data provider is attached. The same data provider will use different pseudonyms for different raw data, and achieve data security and privacy protection for vehicle users by reducing the correlation between data.

[0059] As data providers, vehicles will give priority to sharing data with nearby base stations on the road. Under the in-vehicle social network architecture, the will of the vehicle reflects the will of the person, so vehicles with similar interests and hobbies will actively approach. Using the activity-aware social vehicle clustering algorithm, we get the cluster head vehicle CH responsible for communicating with the base station BS, the cluster member vehicles CM that must forward data from CH, and some undefined vehicles UD that have not been clustered. When cluster C i When there are similar data requirements, the CH will initiate a request to the BS as a representative.

[0060] According to the incentive mechanism, the vehicle sends the original data, i.e., shared information, to the base station. After receiving the original data from the vehicle, the base station integrates the data and packages it into blocks, triggering the Smart Contract for Data Storage (SCDS), and reaching a node consensus through the storage space proof mechanism. The packaged data blocks will be stored in the local MEC server under the blockchain, and a list of original data storage addresses will be generated and stored on the chain.

[0061] Before data sharing, the vehicle generates a metadata index and packages the index and sends it to the base station; the base station receives the metadata index and starts the Smart Contract for Information Sharing (SCIS), the base station verifies the element index, and then reaches a consensus through the proof-of-work mechanism; among them, the metadata index includes timestamp, data information description, data owner information, storage address, information index and historical sharing record.

[0062] The incentive mechanism of the present invention includes the following contents:

[0063] Assume that there are M BSs deployed on the road, expressed as Each BS is equipped with a MEC server. Assume that there are N vehicles on the road, represented as In the in-vehicle social network, different base stations store content about different points of interest x = {ω x ,cy x},ω x Represents the size of content x, cy x represents the amount of computing resources required to transmit content x, i.e., the number of CPU cycles. The content request transaction process between the base station, cluster head, and cluster members is as follows: Figure 4 As shown in the figure, since the vehicles are moving at high speed on the road, the data transmission link is unstable. According to the clustering algorithm, a vehicle cluster with long duration, high interest similarity and close distance can be obtained. Furthermore, data is forwarded and shared within a stable cluster. Before CH sends shared information, it is first necessary to price the data between BS and CH. In order to overcome the selfishness of CH and the inertia of forwarding data, BS needs to provide price incentives for it. By obtaining the optimal price strategy, that is, sharing decision, the system benefit is maximized, and data sharing decision is adopted for data sharing. The specific process includes the following:

[0064] Cluster members construct the benefit function of cluster members based on the user experience quality QoE and the unit price of purchasing content; specifically, the user experience quality of cluster members is calculated based on the content popularity, vehicle selfishness and data packet transmission rate. The calculation formula is:

[0065] QoE=ln[ω x / k·(1-PDR x,i ·s x,i ·r(x) / X)]

[0066] Among them, QoE represents the user experience quality of cluster members, ω x represents the size of the content x, k represents the unit size of the vehicle transmission data packet, PDR x,i Indicates vehicle v i Packet transmission rate when transmitting content x, s x,i Indicates vehicle v i The selfishness when forwarding content x, r(x) represents the popularity ranking of content x, and X represents the number of contents.

[0067] The unit price of content purchased by cluster members is calculated based on content cost, content popularity and vehicle selfishness. The calculation formula is:

[0068]

[0069] Where m(x) represents the unit price of content x purchased by cluster members; c(x) represents the cost of content x, and c(x) is the pricing of the base station; Represents a collection of content.

[0070] The benefit function of cluster members is calculated based on the user experience quality and the unit price of purchased content. The calculation formula is:

[0071]

[0072] Us CM ≥0

[0073]

[0074] Among them, U CM represents the benefit of cluster members, r min represents the minimum content ranking, r max Indicates the maximum content rank.

[0075] The cluster head constructs the benefit function of the cluster head according to the reward amount issued by the base station to the cluster head, the energy consumption of the cluster head and the computing power loss of the cluster head; specifically, the reward amount issued by the base station to the cluster head is calculated according to the unit price of the content purchased by the cluster members and the selfishness of the vehicle; the calculation formula is:

[0076] p(x)=m(x)·ln(s x,i +u)

[0077] Among them, p(x) represents the reward amount issued by the base station to the cluster head, and u represents the selfishness influencing factor.

[0078] The energy consumption of the cluster head is calculated based on the vehicle downlink transmission rate, the vehicle's transmission power and the size of the content; the calculation formula is:

[0079]

[0080] R i =B log2(1+SNR i,j )

[0081] SNR i,j =p i h i,j / N0

[0082] Among them, E CH represents the energy consumption of the cluster head, p i Indicates vehicle v i The transmission power, R i Indicates vehicle v i The downlink transmission rate, B represents the channel bandwidth, h i,j represents the transmission channel gain, and N0 represents the Gaussian white noise power.

[0083] The computing power loss of the cluster head is calculated based on the vehicle computing power, vehicle selfishness and the amount of computing resources required for transmitting content; the calculation formula is:

[0084] E cost =θ·F i 2 ·(1+s x,i )·cy x

[0085] Among them, E cost represents the computing power loss of the cluster head, θ represents the sparse energy consumption, and F i Indicates vehicle v i The computing power of cy x Represents the amount of computing resources required to transmit content x.

[0086] The benefit function of the cluster head is calculated based on the reward amount issued by the base station to the cluster head, the energy consumption of the cluster head, and the computing power loss of the cluster head; the calculation formula is:

[0087]

[0088] Us CH ≥0

[0089]

[0090] Among them, s x,imin Indicates vehicle v i The minimum selfishness when forwarding content x, s x,imax Indicates vehicle vi The maximum selfishness when forwarding content x.

[0091] The base station constructs the benefit function of the base station according to the reward amount issued by the base station to the cluster head, the unit price of the content purchased by the cluster member and the energy consumption of the base station; specifically, the energy consumption of the base station is calculated according to the base station transmission energy consumption rate and the base station downlink transmission rate. The calculation formula is:

[0092] E BS =R s,d ·l

[0093] R s,d =B log2(1+SNR i,j )

[0094] SNR i,j =P s,d h i,j / N0

[0095] Among them, E BS represents the energy consumption of the base station, R s,d represents the downlink transmission rate of the base station, l represents the transmission energy consumption rate of the base station, P s,d Indicates the downlink transmission power of the base station.

[0096] The benefit function of the base station is calculated based on the amount of reward issued by the base station to the cluster head, the unit price of the content purchased by the cluster member, and the energy consumption of the base station. The calculation formula is:

[0097] U BS =m(x)-p(x)-E BS

[0098] =(1-ln(s x,i +u))·m(x)·-R s,d ·l

[0099] =(1-ln(s x,i +u))·c(x)·s x,i ·(r(x) / X)-R s,d ·l

[0100] Us BS >0

[0101]

[0102] Among them, c min represents the minimum cost of content x, c max represents the maximum cost of content x.

[0103] The three-party four-stage Stackelberg game iterative algorithm is used to solve the benefit functions of cluster members, cluster heads and base stations, and obtain the data sharing decision that maximizes the system benefit. The specific process includes the following contents:

[0104] Since the strategies of the three parties will affect each other, a three-party four-stage Stackelberg master-slave game model is established. The reverse induction method is used to analyze and prove the Nash equilibrium of two pairs of master-slave relationships, the CH-CM master-slave relationship and the BS-CH master-slave relationship, and the three-party strategy set that maximizes the system benefit is The solution is obtained according to the four-stage game, where r * (x) represents the popularity ranking of content x that maximizes the system benefit, represents the vehicle v that maximizes the system benefit i Selfishness when forwarding content x, c * (x) represents the cost of content x to maximize the system benefit.

[0105] As a follower in the master-slave relationship between CH and CM, CM needs to make an optimal decision first. The benefit equation of CM is a multivariate function, and solving its extreme value requires judging and proving its second-order partial derivative Hessian matrix.

[0106] When the matrix rank det(H)>0 and trace trace(H)<0, the real symmetric Hessian matrix is ​​negative definite. The Hessian matrix of the CM benefit equation is:

[0107]

[0108] The off-diagonal elements in the matrix are symmetrical, and the sum of the main diagonal elements is negative, so trace(H) < 0. The rank of the matrix can be calculated by the following formula:

[0109]

[0110] According to the system initialization parameter setting, the first term is much larger than the second term, so det(H)>0. Therefore, the Hessian matrix is ​​negative definite.

[0111] The negative definite function is a convex function, and its critical point can be obtained by calculating its first-order partial derivative Jacobian matrix. The Jacobian matrix of the CM benefit function is as follows: Let J = 0, and the three game variables r(x), s x,i The fixed relationship between and c(x) gives the Nash equilibrium strategy of CM:

[0112]

[0113]

[0114] Among them, r * (s,c) represents the popularity ranking of content x under the constraints of content cost and vehicle selfishness.

[0115] The goal of CM is to select * So far, the first stage of the four-stage game has ended.

[0116] Since there is a master-slave relationship between CH and both CM and BS, the two relationships will be analyzed separately here.

[0117] First, CH, as the leader of CM, makes the optimal decision r based on the follower CM. * , CH will make corresponding decisions. Substituting the CM Nash equilibrium expression obtained from the first stage game into the benefit function of CH, we can get the simplified benefit function of CH:

[0118]

[0119] Us CH ≥0

[0120]

[0121] At this point, the second phase of the game is completed.

[0122] Then, as a follower of BS, CH needs to give the optimal decision s * , that is, to find the extreme points of a multivariate function through mathematical methods.

[0123] A continuous real variable function in a closed interval is a convex function. When the cost-optimal strategy c is given, it can be solved by letting the CH benefit function U CH The first-order derivative is zero to obtain the extreme point. The first-order derivative and second-order derivative of the CH benefit function are:

[0124]

[0125]

[0126] It can be observed that if the second-order derivative is negative, the benefit function is a convex function. If the first-order derivative is set to zero, the extreme point can be obtained. In summary, the Nash equilibrium expression of CH is:

[0127]

[0128] Among them, s * (c) represents the selfishness of CH under the constraint of content cost, that is, the optimal strategy made by CH.

[0129] At this point, the third stage of the game is completed and the optimal strategy for CH has been given.

[0130] As the leader of CH, BS will give its own optimal cost strategy according to the optimal strategy of CH in the fourth stage; Substitute the Nash equilibrium strategy expression of CM and the Nash equilibrium expression of CH into the BS benefit function to obtain the simplified BS benefit function:

[0131]

[0132] Us BS >0

[0133] From the first-order derivative of the CH Nash equilibrium expression, we know that if its value is positive, the function is monotonically increasing. Substituting the constraints into c min <c * <c max The minimum value c under the condition min and the maximum value c max .

[0134] c min =θ·F i 2 ·cy x ·PDR x,i (s min +u)-PDR

[0135] c max =θ·F i 2 ·cy x ·PDR x,i (s max +u)-PDR

[0136] Finally, the optimal strategy c of BS is obtained * , which is the cost of content x that maximizes the system benefit. So far, the fourth stage of the game ends.

[0137] The vehicle shares data according to the data sharing decision, i.e., the CM selects content x according to the data sharing decision and makes content selection decision r. * (x) and pays BS. After the transaction is completed, the information sharing smart contract SCIS will be automatically triggered, and BS will save the transaction record to the blockchain for global sharing.

[0138] By simulating and analyzing the influence of the selfishness incentive factor on the game process and determining its value, the present invention simulates the influence of the BS, CH and CM tripartite benefit functions of the game process as the three-party strategies change, and compares and analyzes the incentive mechanism proposed in the present invention with the participation incentive algorithm, forwarding rate algorithm and average price algorithm to prove the advantages of the proposed incentive mechanism.

[0139] The changing trend of CM user service quality QoE under different incentive mechanisms as the content ranking strategy increases is as follows: Figure 5 As shown. Under the four incentive mechanisms, QoE shows an upward trend. According to the above analysis, the price of content is proportional to the ranking. The higher the ranking, the lower the selfishness of CH forwarding, and the higher the QoE obtained by CM. As the content ranking decreases, the selfishness of CH increases, and the QoE of CM gradually stabilizes. It can be observed that the algorithm of the present invention performs better. This is because the incentive mechanism proposed in the present invention considers both the popularity of the content and the selfishness of the node for dynamic pricing, which has a better incentive effect on CH, and thus can provide CM with better service quality. Correspondingly, the average price incentive algorithm performs the worst. In the participation incentive algorithm, although the participation of CH is incentivized through dynamic pricing, the influence of user selfishness is not considered. The forwarding rate incentive algorithm considers the influence of content popularity when determining the incentive value, but does not consider the selfishness of the node, so the QoE is lower. The trend of the impact of the CH benefit function during the change of content ranking from high to low is shown in the figure. Figure 6 As shown. The benefit function curves under the four incentive algorithms all show an upward trend. Among them, the average price incentive algorithm performs the worst. This is because in the game process of the present invention, the three strategies check and balance each other. If the incentives received by CH are always the same each time, its selfishness will become uncontrollable, resulting in poor profit effect. Both the forwarding rate incentive algorithm and the participation incentive algorithm only consider the content popularity and node selfishness, and do not consider the content ranking, content cost and node selfishness at the same time like the algorithm proposed in the present invention, resulting in suboptimal results.

[0140] The simulation results show that the method proposed in this invention can better overcome the selfishness of vehicles and achieve more reliable data forwarding.

[0141] The present invention is based on the blockchain-assisted VSN architecture. Aiming at the problem of selfish behavior of rational data forwarders in the process of social vehicle data sharing, a price incentive method is designed to realize the data sharing process of social vehicles. Under the premise of ensuring the stability of the cluster, the security of the data is effectively guaranteed, and the personalization of data sharing is correspondingly improved, so that the user experience quality QoE of the cluster member vehicle CM gradually tends to be stable. At the same time, dynamic pricing is performed by considering the popularity of content and the selfishness of nodes, which has a better incentive effect on the cluster head vehicle CH, and can provide CM with better service quality. The present invention is highly practical and has a good application prospect.

[0142] The above embodiments further illustrate the purpose, technical solutions and advantages of the present invention in detail. It should be understood that the above embodiments are only preferred implementation modes of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made to the present invention within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data sharing method based on blockchain and incentive mechanism in Internet of Vehicles, characterized in that: include: Construct an in-vehicle social network system architecture, including a physical layer, a social relationship layer, and a blockchain layer; in the in-vehicle social network system architecture, vehicles share data based on an incentive mechanism; wherein the vehicles include cluster head vehicles and cluster member vehicles; The process of vehicles sharing data according to the incentive mechanism includes: The cluster members construct the benefit function of the cluster members according to the user experience quality and the unit price of the purchased content; the process of constructing the benefit function of the cluster members includes: calculating the user experience quality of the cluster members according to the content popularity, vehicle selfishness and data packet transmission rate; calculating the unit price of the cluster members to purchase content according to the content cost, content popularity and vehicle selfishness; calculating the benefit function of the cluster members according to the user experience quality and the unit price of the purchased content; The formula for calculating the benefit function of cluster members is: s.t.U CM ≥0 Among them, U CM represents the benefit of cluster members, QoE represents the user experience quality of cluster members, m(x) represents the unit price of cluster members to purchase content x, ω x represents the size of the content x, k represents the unit size of the vehicle transmission data packet, PDR x,i Indicates vehicle v i Packet transmission rate when transmitting content x, s x,i Indicates vehicle v i The selfishness when forwarding content x, r(x) represents the popularity ranking of content x, X represents the number of contents, c(x) represents the cost of content x, r min represents the minimum content ranking, r max represents the maximum content ranking, χ represents the content set; The cluster head constructs a benefit function of the cluster head according to the amount of reward issued by the base station to the cluster head, the energy consumption of the cluster head and the computing power loss of the cluster head; the process of constructing the benefit function of the cluster head includes: calculating the amount of reward issued by the base station to the cluster head according to the unit price of the content purchased by the cluster members and the selfishness of the vehicle; calculating the energy consumption of the cluster head according to the downlink transmission rate of the vehicle, the transmission power of the vehicle and the size of the content; calculating the computing power loss of the cluster head according to the computing power of the vehicle, the selfishness of the vehicle and the amount of computing resources required for transmitting the content; calculating the benefit function of the cluster head according to the amount of reward issued by the base station to the cluster head, the energy consumption of the cluster head and the computing power loss of the cluster head; The formula for calculating the benefit function of the cluster head is: s.t.U CH ≥0 Among them, U CH represents the benefit of the cluster head, p(x) represents the reward amount issued by the base station to the cluster head, and E CH represents the energy consumption of the cluster head, E cost represents the computing power loss of the cluster head, s x,i Indicates vehicle v i The selfishness when forwarding content x, u represents the selfishness influencing factor, p i Indicates vehicle v i The transmission power, R i Indicates vehicle v i The downlink transmission rate, θ represents the sparse energy consumption, F i Indicates vehicle v i The computing power of cy x represents the amount of computing resources required to process the content ranked r(x), s x,imin Indicates vehicle v i The minimum selfishness when forwarding content x, s x,imax Indicates vehicle v i The maximum selfishness when forwarding content x; The base station constructs a benefit function of the base station according to the amount of reward issued by the base station to the cluster head, the unit price of the content purchased by the cluster member and the energy consumption cost of the base station; the process of constructing the benefit function of the base station includes: calculating the energy consumption cost of the base station according to the transmission energy consumption rate of the base station and the downlink transmission rate of the base station; calculating the benefit function of the base station according to the amount of reward issued by the base station to the cluster head, the unit price of the content purchased by the cluster member and the energy consumption cost of the base station; The formula for calculating the benefit function of the base station is: U BS =m(x)-p(x)-E BS =(1-ln(s x,i +u))·m(x)·-R s,d ·l =(1-ln(s x,i +u))·c(x)·s x,i ·(r(x) / X)-R s,d ·l s.t.U BS >0 Among them, U BS represents the benefit of the base station, p(x) represents the reward amount issued by the base station to the cluster, E BS represents the energy consumption of the base station, R s,d represents the downlink transmission rate of the base station, l represents the transmission energy consumption rate of the base station, and c min represents the minimum cost of content x, c max represents the maximum cost of content x; The three-party four-stage Stackelberg game iterative algorithm is used to solve the benefit functions of cluster members, cluster heads and base stations, and obtain the data sharing decision that maximizes the system benefit. Vehicles share data based on data sharing decisions.

2. According to claim 1, a data sharing method based on blockchain and incentive mechanism in the Internet of Vehicles is characterized in that: The process of vehicles sharing data according to the incentive mechanism also includes: before sharing data, the vehicle generates a metadata index and packages the index and sends it to the base station; the base station receives the metadata index, starts the information sharing smart contract and reaches a consensus; among which, the metadata index includes timestamp, data information description, data owner information, storage address, information index and historical sharing record.

Images