Node trust evaluation method, system and device in 6G heterogeneous network scene and medium

Abstract

The invention discloses a node trust evaluation method, system and equipment in a 6G heterogeneous network scene and a medium. The method comprises the following steps: constructing a 6G heterogeneous network scene and a core role and a participation role in a trust domain formed by different heterogeneous networks; constructing a trust evaluation model based on the block chain, wherein the trust evaluation model comprises trust definition, trust property and node behavior division; respectively designing a direct trust evaluation scheme, a recommended trust evaluation scheme and a comprehensive trust evaluation scheme; the system, the device and the medium are used for bearing and implementing the method. According to the method, the characteristics of the 6G heterogeneous network are analyzed, trust among nodes in the heterogeneous network and related properties are defined, a novel trust evaluation model based on the block chain is designed, time decay factors, penalty factors and historical window parameters of different trust domains are introduced, and the trust evaluation efficiency is improved. A direct trust evaluation scheme, a recommended trust evaluation scheme and a comprehensive trust evaluation scheme are designed, and the method can be effectively applied to a scene of 6G complex heterogeneous network convergence.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to a method, system, device and medium for node trust evaluation in a 6G heterogeneous network scenario. Background Art

[0002] In recent years, after the establishment of the IMT-2030 (6G) Promotion Group, more and more experts and scholars have begun to explore and promote the development of China's sixth-generation mobile communication technology. In the future 6G scenario, there will be a heterogeneous network in which multiple different structured networks coexist and integrate with each other. However, there are problems such as non-uniform standards and inability to transfer trust among users and devices in the heterogeneous network, which easily result in information and trust islands.

[0003] In the existing network scenario, researchers usually abstract entities such as devices and users as nodes, and endow the nodes with corresponding characteristics and behaviors, so that a complex and changeable relationship network is woven among the nodes in the network. When the number of nodes increases and the interactions between nodes become more and more frequent, new problems arise, that is, the nodes are no longer pure ideal nodes. There will inevitably be malicious nodes in the nodes, which try to disrupt the normal operation of the network and even cause damage to other honest nodes. In order to establish a secure and fair 6G heterogeneous network environment and reshape the trust among nodes, it is a very effective method to introduce a node trust evaluation model based on blockchain technology.

[0004] Blockchain technology integrates technologies such as cryptography, consensus algorithms, and P2P networks, enabling the traceability of data on the chain and preventing malicious tampering, which can provide strong support for the establishment of a trust evaluation model. As a security authentication mechanism in the field of network security, the trust evaluation model demonstrates excellent flexibility and adaptability in areas such as distributed networks, wireless sensor networks, and peer-to-peer (P2P) networks. The network node trust evaluation model is different from traditional network security authentication mechanisms. The trust model can flexibly design matching trust evaluation methods and trust incentive mechanisms according to the actual application scenario, and endow the network with the ability to perceive dynamic behavior by recording and updating the changes in trust in real time. This change from passive to active can effectively deal with malicious attackers in the network. By evaluating the behavior of network nodes, the trust model can quickly identify malicious nodes in the network and provide new solutions for the identity authentication link in the network. In existing related research (G. Du et al., "A Blockchain-Based Trust-Value Management Approach for Secure Information Sharing in Internet of Vehicles," IEEE Internet of Things Journal, vol. 11, no. 1, pp. 333-344, 2024.), for the information sharing scenario in the Internet of Vehicles, trust is described as the message credibility of information sharing between vehicles, and a trust value management method is proposed to establish a trusted environment for information sharing between vehicles. Other research focuses on the Internet of Things data exchange scenario. Aiming at the problem that the existence of malicious devices threatens the integrity and reliability of the exchanged data, a semi-centralized single-domain and multi-domain trust management system architecture is proposed to calculate the trust value of dynamic malicious devices. Some other research (L. Shi, T. Wang, Z. Xiong, Z. Wang, Y. Liu and J. Li, "Blockchain-Aided Decentralized Trust Management of Edge Computing: Toward Reliable Off-Chain and On-Chain Trust," IEEE Network, vol. 38, no. 5, pp. 182-188, Sept. 2024.) believes that trust management is an effective way to improve the network security, efficiency, and scalability of the industrial Internet, and a dynamic trust management model for edge devices in the industrial Internet based on feedback is proposed under the edge computing architecture.There are also studies (B. Veith, D. Krummacker and H. D. Schotten, "The Road to Trustworthy 6G: A Survey on Trust Anchor Technologies," IEEE Open Journal of the Communications Society, vol. 4, pp. 581-595, 2023.) that analyze the definition and description of trust in future 6G networks from three aspects: human-to-human, human-to-machine, and machine-to-machine.

[0005] In summary, (1) The existing technologies design trust evaluation models for a single network structure and do not consider the scenarios of heterogeneous networks. In the future 6G network, there will be a fusion of multiple heterogeneous networks, and a large number of users and devices will access the network in a heterogeneous mode, forming different trust domains. The existing trust evaluation models for a single network structure cannot be effectively applied to heterogeneous networks. (2) The existing technologies only consider two results of node interaction, namely success and failure or normal and malicious. However, in the complex 6G heterogeneous network scenario, this cannot comprehensively reflect the true performance of nodes in the heterogeneous network, reducing the objectivity and accuracy of the trust evaluation model. (3) The existing technologies manage trust-related data in a centralized manner, which is prone to single-point failures, causing the complete paralysis of the trust system. The centralized data storage method is also prone to leakage, tampering, or destruction of trust data. In addition, trust centralization makes users and devices have to trust the centralized management party. The overly centralized architecture lacks transparency, and improper behaviors such as abuse of power by the management party may occur. (4) The existing trust evaluation models usually adopt a real-time update method. In the 6G network, the number of users and devices (UEs) is huge. If it is required to update the trust of each node in real time, this will greatly increase the burden on the system and cause system instability. Summary of the Invention

[0006] In order to overcome the above deficiencies of the existing technologies, the purpose of the present invention is to provide a node trust evaluation method, system, device, and medium in a 6G heterogeneous network scenario. By analyzing the characteristics of the 6G heterogeneous network, defining the trust between nodes in the heterogeneous network and related properties, designing a new blockchain-based trust evaluation model, introducing time decay factors, penalty factors, and historical window parameters for different trust domains, and designing direct trust evaluation schemes, recommended trust evaluation schemes, and comprehensive trust evaluation schemes, it can be effectively applied to the scenario of the integration of complex 6G heterogeneous networks.

[0007] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0008] A node trust evaluation method in a 6G heterogeneous network scenario, comprising the following steps:

[0009] Step 1: Construct a 6G heterogeneous network scenario and the core roles and participating roles in the trust domains formed by different heterogeneous networks;

[0010] Step 2: Construct a blockchain-based trust evaluation model, including trust definition, trust nature, and node behavior classification;

[0011] Step 3: Design a direct trust evaluation scheme between nodes in the 6G heterogeneous network scenario constructed in Step 1 according to the trust nature defined in the blockchain-based trust evaluation model constructed in Step 2, as the trust update measurement standard after direct interaction between nodes;

[0012] Step 4: Design a recommended trust evaluation scheme for nodes according to the direct trust evaluation scheme between nodes designed in Step 3 to improve the heterogeneous network node trust evaluation scheme;

[0013] Step 5: Design a comprehensive trust evaluation scheme for nodes based on the direct trust evaluation scheme between nodes designed in Step 3 and the recommended trust evaluation scheme for nodes designed in Step 4, and finally complete the 6G heterogeneous network node trust evaluation.

[0014] Further, the process of Step 1 is as follows:

[0015] Construct a 6G heterogeneous network scenario, with the cellular network, Internet of Things, and wireless local area network as the underlying communication layer. The roles in each heterogeneous network are divided into two categories, namely core roles and participating roles. In the cellular network, the core role is the base station device responsible for network access and data transmission, and the participating roles are mobile subscribers (MSUs) and devices; in the Internet of Things, the core role is the Internet of Things core network device, and the participating role is the Internet of Things sensor device; in the wireless local area network, the core role is the wireless router device, and the participating role is the wireless terminal device; in the blockchain-based trust evaluation model, the nodes participating in trust evaluation are the participating roles in each heterogeneous network. The core roles are considered to be secure and trustworthy and serve as the basis for the normal operation of the entire system without participating in trust evaluation. After the core roles between heterogeneous networks jointly discuss and confirm the trust evaluation parameters and schemes, trust is automatically updated through the smart contract algorithm at the blockchain layer.

[0016] Further, the process of Step 2 is as follows:

[0017] 2.1 Trust Definition and Trust Nature

[0018] The definition of trust is as follows:

[0019] Trust refers to the possibility that a node believes that another node will behave responsibly in a heterogeneous network, which is measured by the trust value;

[0020] Trust is defined with the following properties:

[0021] (1) Peer-to-peer: In a peer-to-peer trust model, trust is established on a one-to-one relationship. In a heterogeneous network, each node independently establishes a trust relationship with other nodes.

[0022] (2) Unidirectionality: In the one-way trust model, trust is one-way, that is, the degree of trust one node has in another node is independent of the trust the other node has in it. This trust relationship is determined by the behavior and performance of the trusted node. The relationship between two nodes can be one-way trust or one-way distrust, or one node trusts the other node while the other node distrusts it.

[0023] (3) Dynamicity: If a node abides by the consensus mechanism and operating rules of the network, does not arbitrarily discard, tamper with or damage data during data transmission and storage, and completes the assigned tasks on time, then the trust of other nodes interacting with it will increase. On the contrary, if the node violates the rules, its trust will decrease accordingly. As time goes by, the historical behavior records of the node will serve as an important basis for evaluating its trust. In the heterogeneous network trust model, nodes can exchange their experiences and evaluations with each other, and this feedback information will affect the trust evaluation of a node by the entire network.

[0024] (4) Vulnerability: This is manifested as the trust value decreasing to a level lower than the initial trust value within a short unit time, and requiring at least ten times the unit time to restore the trust value to the level before the decrease. The unit time is in the order of seconds, which is determined by the hardware device on which the trust assessment model is deployed.

[0025] 2.2 Heterogeneous Network Node Behavior Division

[0026] Based on the classification of the behaviors between individual nodes in heterogeneous network scenarios, five behavior modes are divided from the two aspects of interaction information and transmission delay. The specific definitions are as follows:

[0027] Normal behavior 1 : The node transmitted the information without tampering and damage, and the transmission delay was higher than twice the theoretical delay standard value of the network, but lower than three times the standard value, and the information transmission was finally completed;

[0028] Normal behavior 2 : The node transmits information without tampering or damage, and the transmission delay is less than twice the theoretical delay standard value of the network, and finally completes the information transmission;

[0029] Malicious behavior δ 1 : The node not only transmitted tampered and damaged information, but also had a transmission delay three times higher than the theoretical delay standard value of the network where it was located, and finally completed the information transmission;

[0030] Malicious behavior δ 2 : The node transmitted tampered and damaged information, but the transmission delay was less than three times the theoretical delay standard value of the network where it was located, and finally completed the information transmission;

[0031] Malicious behavior δ 3 : The receiving node did not receive the information from the sending node, and finally did not complete the information transmission;

[0032] As long as the interaction delay between nodes does not exceed three times the theoretical delay standard value of the network where they are located, it can also be regarded as normal behavior and an incremental incentive of direct trust value will be given.

[0033] Furthermore, the process of the third step is as follows:

[0034] Use the Beta distribution as the probability modeling tool for node trust evaluation. The Beta distribution is defined on the interval (0, 1), and any point on it represents the probability of a certain event occurring. Its probability density function (PDF) is:

[0035]

[0036] where θ is the random variable of the Beta distribution, and α and β are the key shape parameters. α and β determine the shape and characteristics of the Beta distribution; when α is small, the left side of the distribution drops rapidly and the probability of a small value appearing is high; when α is large, the left side of the distribution is relatively flat and the probability of a small value appearing is low; map α - 1 to the normal behavior γ of the node, map β - 1 to the malicious behavior δ of the node, the direct trust value of node i for node j in the x domain at time t follows the Beta distribution, and based on the mean value E(B(γ, δ)) of the Beta distribution, the direct trust value The expression is:

[0037]

[0038] where x is the heterogeneous network trust domain where node i is located, The value range of is [0, 1], and in the formula and respectively represent the positive and negative feedback degrees, and θ x is the penalty factor for the malicious behavior of the node in different trust domains, which is determined by the characteristics of the heterogeneous network where node i is located; The calculation comprehensively considers the influence degree of historical trust on current direct trust, the attenuation degree of the trust value within the trust domain where the node is located over time, and the incentive coefficient of the normal behavior γ of the node; in summary, the degree of positive feedback of node i to node j The expression is as follows:

[0039]

[0040] In the formula, represents the influence of historical trust on current direct trust, where t is the current moment, and t q is the historical moment, and λ x is the time decay factor. If a node does not interact for a long time, its direct trust value will gradually decrease over time and will be reduced to 0.5 at most under the influence of time decay; the size of the time decay factor is inversely proportional to the coverage range of the trust domain. The smaller the coverage range of the trust domain, the higher the attenuation degree of the node's direct trust value over time; when calculating the direct trust value, the parameter n x is introduced to control how long ago the historical data needs to be referred to. The larger the value of n x , the earlier node interaction historical data will be included in the calculation scope; μγ 1 +vγ 2 in the formula represents the trust incentive brought by the node's normal behavior in the latest time period, where μ is defined as the incentive coefficient of the normal behavior γ 1 , and ν is defined as the incentive coefficient of the normal behavior γ 2 ;

[0041] According to the trust nature defined in step two, trust is vulnerable. For the degree of negative feedback of the node, a more severe punishment mechanism for the malicious behavior of the node is designed as follows:

[0042]

[0043] In the formula, represents the punishment brought by the node's malicious behavior in the latest time period. If a certain node makes the worst behavior δ 1 , then the degree of negative feedback of this node will increase exponentially; when the node makes a malicious behavior δ 1 that is second only to δ 2 in terms of severity, the rising trend of the degree of negative feedback is also much greater than that of making the same number of normal behaviors;

[0044] In the blockchain-based trust evaluation model, the comprehensive trust value between nodes is collected by the trust management blockchain nodes for the interaction information between nodes, and is uniformly calculated and updated through the trust evaluation model. The update timing is determined according to whether a node actively queries, makes malicious behaviors or for other specific purposes. If a node has never actively queried the trust request, and other nodes do not include this node in the alternative nodes when calculating the recommended trust;

[0045] The process of uploading the data directly interacted by nodes to the blockchain and updating the direct trust value is as follows:

[0046] (1) Node A initiates an interaction request to node B, where A and B are any nodes in the heterogeneous network. In the blockchain-based trust evaluation model, the information sent from the requester to the receiver can be accurately delivered, and the trust management blockchain network nodes are the core devices in the heterogeneous network and are trustworthy and secure;

[0047] (2) After sending the interaction request information, node A packages the digest and timestamp of this information to generate a block and uploads it to the blockchain;

[0048] (3) Node B responds to the interaction request of node A;

[0049] (4) After the response, node B also packages the digest and timestamp of the information to generate a block and uploads it to the blockchain;

[0050] (5) Node A judges the rating of node B's behavior, which is divided into five types according to whether the information is malicious and the size of the transmission delay. Among them, correct information and large delay are classified as normal behavior γ 1 , correct information and small delay are classified as normal behavior γ 2 , malicious information and large delay are classified as malicious behavior δ 1 , malicious information and small delay are classified as malicious behavior δ 2 , if node A does not receive a response from node B within the specified time, node A determines that node B's behavior is malicious behavior δ 3 ;

[0051] (6) Node A feeds back the behavior rating of node B to the trust management blockchain node;

[0052] (7) After a specified time, the blockchain node verifies the behavior rating according to the collected behavior ratings and the timestamps of the information uploaded to the blockchain in the second and fourth steps, and writes the current node interaction record into the blockchain. When the condition for calculating the node trust is triggered, the direct trust value is automatically updated through the trust evaluation model smart contract deployed in the blockchain.

[0053] Furthermore, in the fourth step, the process of calculating the node recommended trust is as follows:

[0054] (1) Determine whether the interaction times between node i and node j meet the threshold TH. If it has exceeded the threshold TH, it means that the two nodes already have sufficient historical direct interaction behaviors, and there is no need to improve the accuracy and objectivity of the comprehensive trust by calculating the recommended trust.

[0055] (2) Select the node zi with the most interaction times with node j from the node list. Each time an alternative recommended node is selected, query dataNum trust values within the historical time period through the smart contract of the blockchain, and the interval of each historical time period is Δt.

[0056] (3) Repeat the second step until nodeNum alternative recommended nodes are selected to obtain the matrix Tru, where the rows represent the number of alternative recommended nodes, and the columns represent the number of trust values obtained by querying each alternative recommended node.

[0057] (4) Calculate the correlation coefficient pairwise for the nodeNum groups of trust data, screen out the largest one, select the corresponding two nodes as the finally determined recommended nodes, and calculate the average value of the two groups of trust values of the two recommended nodes to finally obtain the recommended trust value of node j.

[0058]

[0059] Among them, TT p,j is the comprehensive trust value of node p for node j, TT q,j is the comprehensive trust value of node q for node j, Col(p, q, j) is the correlation coefficient between TT p,j and TT q,j , and n is the number of TT p,j and TT q,j .

[0060] Furthermore, the process of the fifth step is as follows:

[0061] The direct trust value obtained from step three and the recommended trust value obtained from step four are weighted and calculated to obtain the comprehensive trust of the node The specific expression is as follows:

[0062]

[0063] Among them, is the weight of the direct trust evaluation in the comprehensive trust evaluation, and the weight of the recommended trust is The expression of

[0064]

[0065] Among them, ζ is a parameter that controls the change trend of the curve, and Count(i, j, Δt) represents the total number of interactions between nodes i and j within a specified time period, which is obtained by calculation. When there is no interaction history between two nodes, the direct trust weight takes a value of 0, and the comprehensive trust value is determined by the recommended trust value. When the number of interactions between two nodes gradually increases, the proportion of the direct trust value gradually increases, and the proportion of the recommended trust value gradually decreases. The degree of trust of a node in other nodes is gradually determined by its own interaction experience. When the number of interactions between two nodes reaches the system-set threshold TH, there is already a rich interaction history between the nodes, and there is no need to assist other recommended nodes to participate in the calculation of the comprehensive trust value. At this time, the comprehensive trust between nodes completely depends on the direct trust. The number of node interactions Count(i, j, Δt) will change dynamically over time. If two nodes have no interaction for a long time, the number of node interactions Count(i, j, Δt) gradually decreases to 0.

[0066] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0067] 1. By analyzing the characteristics of the 6G heterogeneous network in step two, the present invention defines the trust between nodes in the heterogeneous network and related properties, designs a new blockchain-based trust evaluation model, and the direct trust evaluation scheme designed in step three, the recommended trust evaluation scheme designed in step four, and the comprehensive trust evaluation scheme designed in step five. By introducing the time decay factor and historical window parameters of different trust domains, it can be effectively applied to the scenario of complex heterogeneous network fusion in 6G.

[0068] 2. Through the direct trust evaluation scheme designed in step three, the present invention analyzes the behavior characteristics of nodes in the 6G heterogeneous network scenario, designs a more fine-grained node behavior classification method, and divides two normal behaviors and three malicious behaviors based on whether the behavior is malicious and the size of the transmission delay. The trust evaluation criteria and update methods are improved for five different node behaviors, which more truly reflects the real performance of nodes in the heterogeneous network.

[0069] 3. Through the calculation methods of the positive feedback degree and negative feedback degree designed in step three, the present invention gives trust incentives to nodes that perform normal behaviors. For nodes that perform malicious behaviors, the trust evaluation mechanism will immediately significantly reduce their comprehensive trust values. The system can accurately screen out malicious nodes from numerous nodes in a short time, ensuring the normal order and data security of the heterogeneous network system.

[0070] 4. Through the method of direct interaction and data uploading of nodes designed in Step 3, the present invention stores the interaction data between nodes through the blockchain, and completes the update of the direct trust value of nodes through the smart contract deployed in the blockchain, constructing a trustworthy distributed system architecture for node trust evaluation in heterogeneous network scenarios, greatly reducing the probability of single-point failure and enhancing the reliability of the heterogeneous network trust evaluation model. The blockchain technology builds a solid defense line for the security and immutability of trust data, ensuring the accuracy and credibility of trust evaluation results in the complex and ever-changing 6G heterogeneous network environment.

[0071] 5. Through the method of direct trust value update of nodes designed in Step 3, the present invention optimizes the timing of node trust update in the trust evaluation system. The system only needs to perform trust update on a node when the node actively initiates a trust query, makes malicious behavior, or serves as an alternative node for calculating the recommended trust value, which can reduce the trust update calculation amount and improve the stability of the trust evaluation system.

[0072] 6. Through the recommended trust evaluation scheme designed in Step 4, the present invention can calculate the comprehensive trust value of a node through the historical interaction data of the recommended node with the highest correlation coefficient before the number of interactions between nodes reaches the interaction threshold, improving the accuracy of calculating the comprehensive trust value of nodes when the trust evaluation system lacks direct interaction data between nodes.

[0073] 7. Through the comprehensive trust evaluation scheme designed in Step 5, the present invention can dynamically adjust the proportion weights of the direct trust value and the recommended trust value according to the number of direct interactions between nodes, improving the flexibility and rationality of the trust evaluation model.

[0074] In summary, by constructing a 6G heterogeneous network node trust evaluation model integrating the blockchain, introducing time decay factors and historical window parameters in different trust domains, designing a node incentive and punishment mechanism, and proposing a higher fine-grained node behavior division and "update when using and punishing" strategy, the present invention has the advantages of cross-domain trust intercommunication, efficient screening of malicious nodes, secure and trustworthy storage, and system load optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 is a flowchart of the 6G heterogeneous network trust evaluation scheme provided by an embodiment of the present invention.

[0076] Figure 2 is a scenario diagram of the 6G heterogeneous network trust evaluation model provided by an embodiment of the present invention.

[0077] Figure 3 is a flowchart of direct trust update provided by an embodiment of the present invention.

[0078] Figure 4It is a simulation result diagram of the influence of different recommended trust values of the Internet of Things node in the embodiment of the present invention on the comprehensive trust value.

[0079] Figure 5 It is a simulation result diagram of the relationship between the direct trust, recommended trust and comprehensive trust of the Internet of Things node in the embodiment of the present invention.

[0080] Figure 6 It is a simulation result diagram of the comprehensive trust value of the Internet of Things node when maintaining good behavior in the embodiment of the present invention.

[0081] Figure 7 It is a simulation result diagram of the comprehensive trust value of the nodes in three different heterogeneous networks without interaction for a long time in the embodiment of the present invention.

[0082] Figure 8 It is a simulation result diagram of the comprehensive trust value of the malicious behavior and ordinary binary behavior divided in the embodiment of the present invention for the Internet of Things node.

[0083] Figure 9 It is a simulation result diagram of the comparison of the punishment intensities of three different heterogeneous networks in the embodiment of the present invention. Specific implementation manners

[0084] The technical solutions adopted by the present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0085] In view of the problems of non-uniform heterogeneous network standards and inability to transfer trust in the 6G scenario, the present invention designs a trust evaluation scheme for heterogeneous network nodes based on blockchain. This scheme considers a typical 6G heterogeneous wireless communication network scenario. The underlying communication layer is composed of three heterogeneous networks, namely cellular network, wireless local area network, and Internet of Things. The top layer is a blockchain network composed of core nodes of the three heterogeneous networks, which is responsible for unified trust management of the nodes in the communication layer. Further, see Figure 1 , the node trust evaluation method in the 6G heterogeneous network scenario includes the following steps:

[0086] Step 1: Construct the core roles and participating roles in the 6G heterogeneous network scenario and the trust domains formed by different heterogeneous networks, so as to design a specific trust evaluation scheme subsequently;

[0087] Further, the process of Step 1 is as follows:

[0088] As Figure 2As shown in the figure, the present invention constructs a typical scenario in a 6G heterogeneous network, which consists of three heterogeneous networks: a cellular network, an Internet of Things (IoT), and a wireless local area network (WLAN), forming the underlying communication layer. In the real world, such scenarios are prevalent in schools, hospitals, and industrial parks, where different users and devices in different networks exchange data information. The roles in each heterogeneous network can be divided into two major categories, namely core roles and participating roles. For example, in the cellular network, the core role is the base station device responsible for network access and data transmission, and the participating roles are mobile subscribers (MSUs) and devices; in the IoT, the core role is the IoT core network device, and the participating role is the IoT sensor device; in the WLAN, the core role is the wireless router device, and the participating role is the wireless terminal device; in the trust evaluation model based on blockchain proposed in the present invention, the nodes participating in the trust evaluation are the participating roles in each heterogeneous network. The core roles are considered to be secure and trustworthy and, as the basis for the normal operation of the entire system, do not need to participate in the trust evaluation. After the core roles between heterogeneous networks jointly discuss and confirm the trust evaluation parameters and schemes, the trust is automatically updated through the smart contract algorithm at the blockchain layer.

[0089] Step 2: Construct a trust evaluation model based on blockchain, including trust definition, trust nature, and node behavior classification, laying a foundation for the trust evaluation method designed later in the present invention;

[0090] Furthermore, the process of Step 2 is as follows:

[0091] 2.1 Trust Definition and Trust Nature

[0092] The present invention defines trust as follows:

[0093] Trust refers to the possibility that in a heterogeneous network, one node believes that another node will act responsibly, which is measured by a trust value;

[0094] The trust defined by the present invention has the following properties:

[0095] (1) Point-to-point nature: Point-to-point nature is a basic characteristic of trust in the 6G heterogeneous network trust model. This characteristic emphasizes the directness and individuality of the trust relationship. In a point-to-point trust model, trust is established on a one-to-one relationship, rather than a many-to-one or one-to-many relationship. This means that in a heterogeneous network, each node independently establishes a trust relationship with other nodes, rather than relying on the trust evaluation of the entire network or a certain central node.

[0096] (2) Unidirectionality: Unidirectionality is a notable feature of the trust model proposed in the present invention, which is significantly different from traditional two-way trust. In the one-way trust model, trust is one-way, that is, the degree of trust one node has in another node is independent of the trust the other node has in it. This trust relationship is determined by the behavior and performance of the trusted node, rather than based on mutual trust between the two parties. The relationship between two nodes can be one-way trust or one-way distrust, or one node trusts the other node while the other node does not trust it. This diversity provides a more flexible way of interaction for nodes in the network. Secondly, the flexibility of one-way trust is reflected in the fact that nodes can independently determine the degree of trust in other nodes based on their own experience and observations. This independence means that nodes can conduct trust assessments on other nodes based on specific criteria and scenarios without considering whether the other party has the same trust in themselves.

[0097] (3) Dynamicity: Dynamicity is a core feature of the 6G heterogeneous network trust model, which is reflected in the degree of trust between nodes in the network. This degree of trust is not static, but presents significant dynamic changes. First, the degree of trust between nodes will be affected by the behavior of the nodes themselves. If a node strictly abides by the consensus mechanism and operating rules of the network, does not arbitrarily discard, tamper with or damage data during data transmission and storage, and completes the assigned tasks on time regardless of difficulty, such as completing data transmission on time and maintaining network security, then the trust of other nodes interacting with it will increase. On the contrary, if the node violates regulations, such as deception, refusal to provide services or spreading malicious content, its trust will decrease accordingly. Secondly, the historical performance of the node plays an important role in the dynamics of trust. As time goes by, the historical behavior records of nodes will serve as an important basis for evaluating their trustworthiness. If a node maintains consistent good performance over a long period of time, its trustworthiness will be more solid; if it has a bad history, even if it has improved recently, the recovery of its trustworthiness may be affected to a certain extent; in addition, feedback from other nodes is also a key factor affecting the dynamics of trust; in a heterogeneous network trust model, nodes can communicate with each other about their respective experiences and evaluations, and this feedback information will affect the entire network's trust assessment of a certain node; if a node receives a lot of positive feedback, its recommendation trust will be enhanced; and negative feedback may lead to a decline in recommendation trust.

[0098] (4) Vulnerability: This is manifested as the trust value decreasing to a level lower than the initial trust value within a relatively short unit time, and requiring at least dozens of times the unit time to restore the trust value to the level before the decrease. The unit time is in the order of seconds, which is determined by the hardware device on which the trust assessment model is deployed.

[0099] Therefore, even if the node attempts to repair the trust through positive actions subsequently, it is difficult to restore the trust value to the level before attenuation in a short period of time.

[0100] 2.2 Heterogeneous Network Node Behavior Classification

[0101] The present invention classifies the behaviors between individual nodes in a heterogeneous network scenario, and divides five behavior patterns from two aspects of interaction information and transmission delay. The specific definitions are as follows:

[0102] Normal behavior γ 1 : The node transmits information without tampering and damage, the transmission delay is higher than twice the theoretical delay standard value of the network where it is located, but lower than three times the standard value, and finally completes the information transmission;

[0103] Normal behavior γ 2 : The node transmits information without tampering and damage, and the transmission delay is lower than twice the theoretical delay standard value of the network where it is located, and finally completes the information transmission;

[0104] Malicious behavior δ 1 : The node not only transmits tampered and damaged information, but also the transmission delay is higher than three times the theoretical delay standard value of the network where it is located, and finally completes the information transmission;

[0105] Malicious behavior δ 2 : The node transmits tampered and damaged information, but the transmission delay is lower than three times the theoretical delay standard value of the network where it is located, and finally completes the information transmission;

[0106] Malicious behavior δ 3 : The receiving node does not receive the information from the sending node, and finally does not complete the information transmission;

[0107] The present invention designs the corresponding judgment criteria for the normal behavior between nodes. As long as the interaction delay between nodes does not exceed three times the theoretical delay standard value of the network where it is located, it can also be regarded as normal behavior, and an incremental incentive of direct trust value is given. The purpose of this design is to encourage nodes to interact with other nodes in a complex heterogeneous network.

[0108] Step 3: Design a direct trust evaluation scheme between nodes in the 6G heterogeneous network scenario constructed in Step 1 according to the trust nature defined in the blockchain-based trust evaluation model constructed in Step 2, as the trust update measurement standard after direct interaction between nodes, laying a foundation for the subsequent design of the node comprehensive trust value calculation method of the present invention;

[0109] Furthermore, the process of Step 3 is as follows:

[0110] The direct trust value of a node in a heterogeneous network can be expressed as the probability that a node considers the other party to be trustworthy when interacting with other nodes. The Beta distribution is the conjugate prior of the binomial distribution parameters. When we observe the behavior of a node multiple times, it can be regarded as a series of binomial trials. Temporarily, the behavior of the node is simply divided into normal and malicious. The Beta distribution can effectively combine prior information and observed data to update the direct trust value of the node. Using the Beta distribution as a probability modeling tool for node trust evaluation, the Beta distribution is defined on the interval (0, 1), and any point on it can be interpreted as the probability of an event occurring. Its probability density function (PDF) is:

[0111]

[0112] where θ is the random variable of the Beta distribution, and α and β are the key shape parameters. α and β determine the shape and characteristics of the Beta distribution. When α is small, the left side of the distribution drops rapidly, and the probability of smaller values is higher. When α is large, the left side of the distribution is relatively flat, and the probability of smaller values is lower. Map α - 1 to the normal behavior γ of the node, and map β - 1 to the malicious behavior δ of the node. The direct trust value of node i in the x domain of node j at time t follows the Beta distribution. Based on the mean E(B(γ, δ)) of the Beta distribution, the direct trust value The expression is:

[0113]

[0114] where x is the heterogeneous network trust domain where node i is located, The value range of is [0, 1]. In the formula, and respectively represent the positive and negative feedback degrees. θ x is the penalty factor for the malicious behavior of the node in different trust domains, which is determined by the characteristics of the heterogeneous network where node i is located. In addition, The calculation of also needs to comprehensively consider the influence degree of historical trust on the current direct trust, the attenuation degree of the trust value over time in the trust domain where the node is located, and the incentive coefficient of the normal behavior γ of the node. To sum up, the positive feedback degree The expression of is:

[0115]

[0116] In the formula, represents the influence of historical trust on the current direct trust, where t is the current time, t q is the historical time, and λ xis the time decay factor. If a node does not interact for a long time, its direct trust value will gradually decrease over time and will be reduced to 0.5 at most under the influence of time decay. The time decay factors in different heterogeneous network trust domains are different, which is determined by the characteristics of the domain. Generally, the size of the time decay factor is inversely proportional to the coverage range of the trust domain. The smaller the coverage range of the trust domain, the higher the degree of decay of the node's direct trust value over time. When calculating the direct trust value, the parameter n x is introduced to control how far back in history the data needs to be referenced. The larger the value of n x , the earlier node interaction historical data will be included in the calculation scope; μγ 1 +νγ 2 in the formula represents the trust incentive brought by the node's normal behavior in the latest time period, where μ is defined as the incentive coefficient for the normal behavior γ 1 , and ν is defined as the incentive coefficient for the normal behavior γ 2 .

[0117] According to the trust nature defined in step two of the present invention, trust is vulnerable. Therefore, for the degree of negative feedback of nodes, the present invention designs a more severe penalty mechanism for node malicious behaviors, which is specifically as follows:

[0118]

[0119] In the formula, represents the penalty brought by the node's malicious behavior in the latest time period. If a certain node performs the worst behavior δ 1 , then the degree of negative feedback of this node will increase exponentially; when the node performs a malicious behavior δ 1 that is second only to δ 2 in terms of severity, the rising trend of the degree of negative feedback is also much greater than that of performing the same number of normal behaviors.

[0120] In the trust evaluation model based on blockchain proposed by the present invention, the comprehensive trust value between nodes is not updated immediately after the interaction is completed. Instead, the trust management blockchain nodes collect the interaction information between nodes and uniformly calculate and update it through the trust evaluation model. The specific update timing is determined according to whether the node actively queries, performs malicious behaviors or other specific purposes. If a certain node has never actively queried the trust request and other nodes do not include this node in the candidate nodes when calculating the recommended trust, therefore, the comprehensive trust value of this node does not need to be updated in real time in the trust evaluation model based on blockchain of the present invention. The process of the data of node direct interaction being uploaded to the chain and the update of the direct trust value is as Figure 3 shown, and the specific steps are as follows:

[0121] (1) Node A initiates an interaction request to Node B. A and B can be any node in a heterogeneous network. In the trust evaluation model based on blockchain, the present invention believes that the information sent by the requesting party to the receiving party can be accurately delivered, and the nodes of the trust management blockchain network are the core devices in the heterogeneous network and are trusted and secure;

[0122] (2) After Node A sends the interaction request information, it packages the digest and timestamp of this information to generate a block and uploads it to the blockchain;

[0123] (3) Node B responds to the interaction request of Node A;

[0124] (4) Similar to step 2, after Node B responds, it also packages the digest and timestamp of the information to generate a block and uploads it to the blockchain;

[0125] (5) Node A judges the rating of Node B's behavior, which can be divided into five types according to whether the information is malicious and the size of the transmission delay. Among them, correct information and large delay are classified as normal behavior γ 1 , correct information and small delay are classified as normal behavior γ 2 , malicious information and large delay are classified as malicious behavior δ 1 , malicious information and small delay are classified as malicious behavior δ 2 , the above four types belong to the successful completion of information interaction between Node A and Node B. Another type is that Node A does not receive a response from Node B within the specified time, and Node A determines that Node B's behavior is malicious behavior δ 3 ;

[0126] (6) Node A feeds back the behavior rating of Node B to the trust management blockchain node;

[0127] (7) After a specified time, the blockchain node checks the behavior rating according to the collected behavior ratings and the timestamps of the information uploaded to the blockchain in the second and fourth steps, and writes the record of this node interaction into the blockchain; when the condition for calculating the node trust is triggered, the direct trust value is automatically updated through the trust evaluation model smart contract deployed in the blockchain.

[0128] Step Four: According to the direct trust evaluation scheme designed in Step Three for nodes, further design a recommended trust evaluation scheme for nodes to improve the heterogeneous network node trust evaluation scheme;

[0129] Furthermore, the process of Step Four is as follows:

[0130] The direct trust between nodes is calculated by collecting the behavior ratings of point-to-point interactions. When the number of interactions between two nodes is small, or even when they have not interacted yet, the direct trust value cannot be calculated. To improve the accuracy of the trust evaluation scheme for heterogeneous networks, the recommended trust parameter is introduced as a supplement to the comprehensive trust calculation of nodes. The specific algorithm is as follows:

[0131]

[0132] The process of calculating the recommended trust of a node is as follows:

[0133] (1) Determine whether the number of interactions between node i and node j meets the threshold TH. If it has exceeded the threshold TH, it means that the two nodes already have sufficient historical direct interaction behaviors, and there is no need to calculate the recommended trust to improve the accuracy and objectivity of the comprehensive trust;

[0134] (2) Select the node zi that has interacted with node j the most times from the node list. Each time, select an alternative recommended node, and query dataNum trust values within the historical time period through the smart contract of the blockchain. The interval of each historical time period is Δt;

[0135] (3) Repeat the second step until nodeNum alternative recommended nodes are selected to obtain the matrix Tru. The rows represent the number of alternative recommended nodes, and the columns represent the number of trust values obtained by querying each alternative recommended node;

[0136] (4) Calculate the correlation coefficient for each pair of the nodeNum groups of trust data, select the largest one among them, choose the corresponding two nodes as the finally determined recommended nodes, and calculate the average value of the two groups of trust values of the two recommended nodes to finally obtain the recommended trust value of node j

[0137]

[0138] Among them, TT p,j is the comprehensive trust value of node p for node j, TT q,j is the comprehensive trust value of node q for node j, Col(p, q, j) is the correlation coefficient between TT p,j and TT q,j , and n is the number of TT p,j and TT q,j .

[0139] The recommended trust of a node completely depends on the interaction history of other nodes. Therefore, it is necessary to extract effective information from the comprehensive trust value of other nodes for this node j as the key basis for selecting recommended nodes. To better explain the above formula, assume there are two alternative recommended nodes p and node q. Nodes p and q have a relatively large number of historical interaction times with the node j to be evaluated, and a set of corresponding comprehensive trust values within different historical time periods can be obtained respectively. If the correlation coefficients calculated from these two sets of comprehensive trust values are closer to 1, it indicates that the change trends of the comprehensive trust values of these two nodes for node j in the same multiple historical time intervals are closer. The present invention believes that the two nodes with the closest change trends of comprehensive trust values are most qualified to be recommended nodes.

[0140] Step Five: Based on the direct trust evaluation scheme designed in Step Three and the recommended trust evaluation scheme designed in Step Four for nodes, design a comprehensive trust evaluation scheme for nodes, and finally complete the trust evaluation of 6G heterogeneous network nodes.

[0141] Furthermore, the process of the above Step Five is as follows:

[0142] The comprehensive trust of a node is a metric for point-to-point trust in the trust evaluation model proposed by the present invention, and is obtained from the direct trust value obtained in Step Three and the recommended trust value obtained in Step Four, and the comprehensive trust of the node is calculated by weighted calculation

[0143]

[0144] wherein, is the weight of the direct trust evaluation in the comprehensive trust evaluation, and the weight of the recommended trust is The expression of

[0145]

[0146] wherein, ζ is a parameter controlling the change trend of the curve, Count(i,j,Δt) represents the total number of interactions between node i and node j within a specified time period, and is obtained by calculating ; when there is no interaction history between the two nodes, the direct trust weight When the value is 0, the comprehensive trust value is determined by the recommended trust value. As the interaction times between two nodes gradually increase, the proportion of the direct trust value gradually increases, and the proportion of the recommended trust value gradually decreases. The degree of trust of a node in other nodes can gradually be determined by its own interaction experience. When the interaction times between two nodes reach the system-set threshold TH, there are rich interaction experiences between the nodes, and there is no need to assist in calculating the comprehensive trust value through other recommended nodes. At this time, the comprehensive trust between nodes completely depends on the direct trust. It should be noted that the node interaction times Count(i, j, Δt) will change dynamically with time. If there is no interaction between two nodes for a long time, the node interaction times Count(i, j, Δt) will gradually decrease to 0.

[0147] As Figure 4 shown, it shows the influence of different recommended trust values on the total trust value under the influence of the same node behavior before reaching the interaction threshold TH. When the interaction times are less, the recommended trust value has a greater influence on the total trust value. In the figure, the red line represents the recommended trust value of 0.3, and the blue line represents the recommended trust value of 0.7. It can be seen that the comprehensive trust value corresponding to the red line is lower than the comprehensive trust value corresponding to the blue line. As the interaction times increase, the total trust value tends to be consistent, and the difference in the comprehensive trust value between the red line and the blue line also gradually decreases. This is because the proportion of the direct trust value in the total trust value gradually increases, and the proportion of the recommended trust value gradually decreases. Thus, it can be seen that when calculating the comprehensive trust value when the direct interaction times between nodes are less, the designed recommended trust evaluation scheme of the present invention can be used as an important reference basis.

[0148] As Figure 5 shown, it shows the change trends of the total trust value, direct trust value, and recommended trust value as the node interaction times increase. Before the system time of 150 seconds, when the node interaction times are less, the change trend of the total trust value represented by the red line is similar to the change trend of the recommended trust value represented by the green line. When the interaction times increase, the change trend of the total trust value is closer to the change trend of the direct trust value. Thus, it can be seen that the designed trust evaluation model of the present invention can statistically count the node interaction times in real time and dynamically adjust the proportion parameters of the direct trust value and the recommended trust value.

[0149] As Figure 6 shown, between 300 seconds and 550 seconds of the system time after reaching the interaction threshold TH, the nodes continuously maintain normal interactions, and the comprehensive trust value of the nodes rises steadily. However, after 550 seconds, even if the nodes maintain normal interactions, the trust value still decreases slightly. Thus, it can be seen that the designed trust evaluation model of the present invention has the dynamic characteristic of time decay.

[0150] As Figure 7As shown, when there is no interaction between nodes for a long time, the trust value will decay over time until it reaches 0.5. The decay degree varies in different domains. In the trust evaluation model designed in the present invention, the trust value of nodes in the wireless local area network with the highest decay degree decays to 0.5 first. It can be seen that the trust evaluation model designed in the present invention can flexibly adjust the decay degree of the trust value over time according to different heterogeneous network characteristics, and is more suitable for the 6G heterogeneous network scenario.

[0151] As Figure 8 shown, it shows the comparison of the malicious degrees of three malicious behaviors divided in the present invention and the traditional binary behavior. Nodes perform the same number of malicious behaviors between 450 - 500 seconds. It can be seen that among the three distinguished malicious behaviors, the malicious degrees of behavior δ 1 and behavior δ 2 are higher than that of the binary behavior, and the malicious degree of behavior δ 3 is lower than that of the binary behavior. It can be seen that the trust evaluation model designed in the present invention has a higher fine-grained division of node behaviors and richer trust evaluation criteria.

[0152] As Figure 9 shown, it shows the comparison of the decline degrees of the comprehensive trust values of nodes in three heterogeneous networks when they perform the same number and type of malicious behaviors between 450 - 500 seconds. In the design of the present invention, the penalty factor of the wireless local area network is the largest, the penalty factor of the cellular network is the smallest, and the penalty factor of the Internet of Things is between the penalty factor of the wireless local area network and the penalty factor of the cellular network. Figure 9 Among them, between 450 seconds and 500 seconds of the system time, the trust value of the malicious node rapidly drops from 0.8 to below 0.3. Among them, the trust value of the wireless local area network node drops the most, and the trust value of the cellular network node drops the least. It can be seen that the trust evaluation model designed in the present invention can quickly reduce the comprehensive trust value of malicious nodes and flexibly adjust the penalty intensity according to different trust domain characteristics.

[0153] The key points and protected points of the present invention include but are not limited to:

[0154] 1. The present invention constructs a node trust evaluation method that can cross trust domains by introducing time decay factors, penalty factors, historical window parameters in different heterogeneous networks, and making a more detailed division of the behaviors between nodes. The blockchain technology is added as the top layer of the trust evaluation model, responsible for trust management and reliable storage, providing a reliable environment for 6G heterogeneous network trust evaluation. That is the content in step three.

[0155] 2. Aiming at the problem that there may be fewer interactions between nodes in heterogeneous networks and the trust relationship cannot be accurately described by the direct trust value, the present invention further proposes an algorithm for recommended trust. That is the content in step four.

[0156] 3. The node comprehensive trust calculation method based on direct trust and referral trust of the present invention, as well as the dynamic algorithm for the weights of the two. That is, the content of Step Five.

[0157] Existing trust evaluation models are all designed for single-structured network models, such as scenarios of vehicle-to-everything (V2X), Internet of Things (IoT), distributed wireless networks, etc. There is no unified trust evaluation solution designed for the 6G heterogeneous network scenario yet. In the 6G scenario, there will be a large number of heterogeneous networks coexisting. Designing a secure, efficient and fair trust evaluation solution will provide a solid foundation for subsequent 6G network identity authentication. Therefore, there is no other alternative solution that can fully achieve the purpose of the present invention.

[0158] The present invention also provides a node trust evaluation system in a 6G heterogeneous network scenario, including:

[0159] A 6G heterogeneous network scenario construction module, which is used to implement constructing the 6G heterogeneous network scenario and the core roles and participating roles in the trust domains composed of different heterogeneous networks in Step One;

[0160] A blockchain-based trust evaluation model construction module, which is used to implement constructing a blockchain-based trust evaluation model in Step Two, including trust definition, trust nature, and node behavior division;

[0161] A direct trust evaluation scheme design module, which is used to implement designing a direct trust evaluation scheme between nodes in the 6G heterogeneous network scenario constructed in Step One according to the trust nature defined in the blockchain-based trust evaluation model constructed in Step Two, as the trust update measurement standard after direct interaction between nodes;

[0162] A referral trust evaluation scheme design module, which is used to implement designing a referral trust evaluation scheme for nodes according to the direct trust evaluation scheme between nodes designed in Step Three, and perfecting the heterogeneous network node trust evaluation scheme;

[0163] A comprehensive trust evaluation scheme design module, which is used to implement designing a comprehensive trust evaluation scheme for nodes based on the direct trust evaluation scheme between nodes designed in Step Three and the referral trust evaluation scheme for nodes designed in Step Four, and finally complete the 6G heterogeneous network node trust evaluation.

[0164] The present invention also provides a node trust evaluation device in a 6G heterogeneous network scenario, including:

[0165] A memory: storing the computer program of the above-mentioned node trust evaluation method in a 6G heterogeneous network scenario, which is a computer-readable device;

[0166] A processor: used to implement the above-mentioned node trust evaluation method in a 6G heterogeneous network scenario when executing the computer program.

[0167] The present invention also provides a computer-readable storage medium storing a computer program, which can implement the node trust evaluation method in a 6G heterogeneous network scenario when executed by a processor.

Claims

1. A node trust evaluation method in a 6G heterogeneous network scenario, characterized in that: The following steps are involved: Step 1: Construct 6G heterogeneous network scenarios and the core roles and participating roles in the trust domains formed by different heterogeneous networks; Step 2: Build a trust assessment model based on blockchain, including trust definition, trust nature, and node behavior division; Step 3: Based on the trust properties defined in the blockchain-based trust evaluation model constructed in Step 2, a direct trust evaluation scheme between nodes is designed in the 6G heterogeneous network scenario constructed in Step 1 as a trust update measurement standard after direct interaction between nodes; Step 4: Based on the direct trust evaluation scheme between nodes designed in step 3, design a recommended trust evaluation scheme for nodes and improve the trust evaluation scheme for heterogeneous network nodes; Step 5: Based on the direct trust evaluation scheme between nodes designed in step 3 and the recommended trust evaluation scheme for nodes designed in step 4, design a comprehensive trust evaluation scheme for nodes, and finally complete the trust evaluation of 6G heterogeneous network nodes.

2. According to a node trust evaluation method in a 6G heterogeneous network scenario according to claim 1, it is characterized in that: The process of step one is as follows: Constructing a 6G heterogeneous network scenario, the underlying communication layer is composed of three heterogeneous networks: cellular network, Internet of Things, and wireless local area network. The roles in each heterogeneous network are divided into two categories, namely core roles and participating roles. In cellular networks, the core roles are base station devices responsible for network access and data transmission, and the participating roles are mobile users (MSU) and devices; in the Internet of Things, the core roles are IoT core network devices, and the participating roles are IoT sensor devices; in wireless local area networks, the core roles are wireless router devices, and the participating roles are wireless terminal devices; in the trust evaluation model based on blockchain, the nodes participating in the trust evaluation are the participating roles in each heterogeneous network. The core roles are considered to be safe and reliable. As the basis for the normal operation of the entire system, they do not need to participate in the trust evaluation. After the core roles between heterogeneous networks jointly discuss and confirm the trust evaluation parameters and plans, the automatic update of trust is realized through the smart contract algorithm at the blockchain layer.

3. According to a node trust evaluation method in a 6G heterogeneous network scenario according to claim 1, it is characterized in that: The process of step 2 is as follows: 2.1 Trust Definition and Nature Trust is defined as follows: Trust refers to the possibility that a node believes that another node will behave responsibly in a heterogeneous network, which is measured by the trust value; Trust is defined with the following properties: (1) Peer-to-peer: In a peer-to-peer trust model, trust is established on a one-to-one relationship. In a heterogeneous network, each node independently establishes a trust relationship with other nodes. (2) Unidirectionality: In the one-way trust model, trust is one-way, that is, the degree of trust one node has in another node is independent of the trust the other node has in it. This trust relationship is determined by the behavior and performance of the trusted node. The relationship between two nodes can be one-way trust or one-way distrust, or one node trusts the other node while the other node distrusts it. (3) Dynamicity: If a node abides by the consensus mechanism and operating rules of the network, does not arbitrarily discard, tamper with or damage data during data transmission and storage, and completes the assigned tasks on time, then the trust of other nodes interacting with it will increase. On the contrary, if the node violates the rules, its trust will decrease accordingly. As time goes by, the historical behavior records of the node will serve as an important basis for evaluating its trust. In the heterogeneous network trust model, nodes can exchange their experiences and evaluations with each other, and this feedback information will affect the trust evaluation of a node by the entire network. (4) Vulnerability: This is manifested as the trust value decreasing to a level lower than the initial trust value within a relatively short unit time, and requiring at least ten times the unit time to restore the trust value to the level before the decrease. The unit time is in the order of seconds, which is determined by the hardware device on which the trust assessment model is deployed. 2.2 Heterogeneous Network Node Behavior Division Based on the behavior between individual nodes in heterogeneous network scenarios, five behavior modes are divided from the two aspects of interaction information and transmission delay. The specific definitions are as follows: Normal behavior γ1: The node transmits untampered and undamaged information, and the transmission delay is higher than twice the theoretical delay standard value of the network, but lower than three times the standard value, and the information transmission is finally completed; Normal behavior γ2: The node transmits information without tampering or damage, and the transmission delay is less than twice the theoretical delay standard value of the network, and the information transmission is finally completed; Malicious behavior δ1: The node not only transmits tampered and damaged information, but also transmits information with a delay three times higher than the theoretical delay standard value of the network, and finally completes the information transmission; Malicious behavior δ2: The node transmits tampered and damaged information, but the transmission delay is less than three times the theoretical delay standard value of the network, and the information is finally transmitted; Malicious behavior δ3: The receiving node does not receive the information from the sending node, and ultimately fails to complete the information transmission; As long as the interaction delay between nodes does not exceed three times the theoretical delay standard value of the network, it can be regarded as normal behavior and given an incremental incentive of direct trust value.

4. According to a node trust evaluation method in a 6G heterogeneous network scenario according to claim 1, it is characterized in that: The process of step three is as follows: Beta distribution is used as a probability modeling tool for node trust evaluation. Beta distribution is defined on the interval (0, 1). Any point is the probability of an event occurring, and its probability density function (PDF) is: Among them, θ is a random variable of Beta distribution, α and β are key shape parameters, which determine the shape and characteristics of Beta distribution; when α is small, the left side of the distribution decreases faster, and the probability of small values ​​is higher; when α is large, the left side of the distribution is relatively flat, and the probability of small values ​​is lower; α-1 is mapped to the normal behavior of the node γ, and β-1 is mapped to the malicious behavior of the node δ. The direct trust value of node i to node j in the x domain at time t obeys Beta distribution. Based on the mean value E(B(γ,δ)) of the Beta distribution, the direct trust value The expression is: Where x is the heterogeneous network trust domain where node i is located, The value range of is [0,1], and the and Represent the degree of positive and negative feedback, θ x is the penalty factor for malicious behavior of nodes in different trust domains, which is determined by the characteristics of the heterogeneous network where node i is located; The calculation of comprehensively considers the influence of historical trust on current direct trust, the decay of trust value over time in the trust domain where the node is located, and the incentive coefficient of the node's normal behavior γ; in summary, the degree of positive feedback from node i to node j The expression is: In the formula represents the impact of historical trust on current direct trust, where t is the current moment, t q It's a historical moment. x is the time decay factor. If a node does not interact for a long time, its direct trust value will gradually decrease over time, and under the influence of time decay, it will be reduced to 0.5 at most. The size of the time decay factor is inversely proportional to the coverage of the trust domain. The smaller the trust domain coverage, the higher the degree of decay of the node's direct trust value over time. When calculating the direct trust value, the parameter n is introduced. x To control how long the historical data needs to be referenced, n x The larger the value of , the earlier node interaction history data will be included in the calculation range; μγ1+vγ2 in the formula represents the trust incentive brought by the normal behavior of the node in the latest time period, where μ is defined as the incentive coefficient of normal behavior γ1, and ν is defined as the incentive coefficient of normal behavior γ2; According to the trust nature defined in step 2, trust is vulnerable. For the degree of negative feedback of nodes, a more severe node malicious behavior penalty mechanism is designed, as shown below: In the formula It represents the penalty for malicious behavior of a node in the latest time period. If a node performs the worst behavior δ1, the negative feedback of the node increases exponentially. When a node performs malicious behavior δ2, which is second only to δ1 in terms of severity, the negative feedback increases much more than the same number of normal behaviors. In the trust evaluation model based on blockchain, the comprehensive trust value between nodes is collected by the trust management blockchain node through the interaction information between nodes, and is uniformly calculated and updated through the trust evaluation model; the update timing is determined by whether the node actively queries, makes malicious behavior or other specific purposes. If a node has not actively queried trust requests, and other nodes have not included the node as a candidate node when calculating recommended trust; The specific steps of the data upload and trust value update process of direct node interaction are as follows: (1) Node A initiates an interaction request to node B. A and B are any nodes in a heterogeneous network. In the blockchain-based trust evaluation model, the information sent by the requester to the receiver can be accurately delivered, and the trust management blockchain network node is the core device in the heterogeneous network and is trustworthy and secure. (2) After node A sends the interaction request information, it packages the summary and timestamp of this information into a block and uploads it to the blockchain; (3) Node B responds to Node A’s interaction request; (4) After responding, Node B also packages the summary and timestamp of the information into a block and uploads it to the blockchain; (5) Node A judges the rating of node B's behavior. It is divided into five types according to whether the information is malicious and the transmission delay. Among them, correct information with a large delay is classified as normal behavior γ1, correct information with a small delay is classified as normal behavior γ2, malicious information with a large delay is classified as malicious behavior δ1, and malicious information with a small delay is classified as malicious behavior δ2. If node A does not receive a response from node B within the specified time, node A determines that node B's behavior is malicious behavior δ3; (6) Node A feeds back the behavior rating of Node B to the trust management blockchain node; (7) After a specified period of time, the blockchain node verifies the behavior rating based on the collected behavior rating and the timestamp of the on-chain information in the second and fourth steps, and writes the node interaction record into the blockchain; when the conditions for computing node trust are triggered, the direct trust value is automatically updated through the trust assessment model smart contract deployed in the blockchain.

5. According to a node trust evaluation method in a 6G heterogeneous network scenario according to claim 1, it is characterized in that: In step 4, the node recommended trust calculation process is: (1) Determine whether the number of interactions between node i and node j meets the threshold TH. If it exceeds the threshold TH, it means that the two nodes have sufficient historical direct interaction behaviors, and there is no need to improve the accuracy and objectivity of comprehensive trust by calculating the recommended trust; (2) Select the node zi that has the most interactions with node j from the node list, select a candidate recommended node each time, and query the dataNum trust values ​​in the historical time period through the smart contract of the blockchain. The interval of each historical time period is Δt; (3) Repeat the second step until nodeNum candidate recommendation nodes are selected, and obtain the matrix Tru, where the rows represent the number of candidate recommendation nodes and the columns represent the number of trust values ​​obtained by querying each candidate recommendation node; (4) Calculate the correlation coefficients for the trust data of the nodeNum group in pairs, select the largest correlation coefficient, select the corresponding two nodes as the final recommended nodes, and average the two groups of trust values ​​of the two recommended nodes to finally obtain the recommended trust value of node j. Among them, TT p,j is the comprehensive trust value of node p to node j, TT q,j is the comprehensive trust value of node q to node j, Col(p,q,j) is TT p,j With TT q,j The correlation coefficient between them, n is TT p,j and TT q,j The number of 6. The node trust evaluation method in a 6G heterogeneous network scenario according to claim 1 is characterized in that: The process of step five is as follows: The direct trust value obtained from step 3 And the recommendation trust value obtained in step 4 Weighted calculation to get the comprehensive trust of the node The specific expression is as follows: in, is the weight of direct trust evaluation in comprehensive trust evaluation, and the weight of recommended trust is The expression is as follows: Among them, ζ is the parameter that controls the trend of the curve, Count(i,j,Δt) represents the total number of interactions between nodes i and j in the specified time period, and is calculated by Get; When the two nodes have no interaction history, the direct trust weight The value is 0, and the comprehensive trust value is determined by the recommended trust value. When the number of interactions between two nodes gradually increases, the proportion of direct trust value gradually increases, and the proportion of recommended trust value gradually decreases. The degree of trust of nodes in other nodes is gradually determined by their own interaction experience. When the number of interactions between two nodes reaches the threshold TH set by the system, the nodes have rich interaction experience and do not need to assist in calculating the comprehensive trust value through other recommended nodes. At this time, the comprehensive trust between nodes depends entirely on direct trust. The number of node interactions Count(i,j,Δt) will change dynamically over time. If the two nodes have no interaction for a long time, the number of node interactions Count(i,j,Δt) will gradually decrease to 0.

7. A node trust evaluation system in a 6G heterogeneous network scenario based on the method according to any one of claims 1 to 6, characterized in that: include: 6G heterogeneous network scenario construction module, used to build 6G heterogeneous network scenarios and the core roles and participating roles in the trust domain formed by different heterogeneous networks; The blockchain-based trust assessment model construction module is used to build a blockchain-based trust assessment model, including trust definition, trust properties, and node behavior division; A direct trust evaluation scheme design module is used to design a direct trust evaluation scheme between nodes in a 6G heterogeneous network scenario according to the trust properties defined in the blockchain-based trust evaluation model, as a trust update measurement standard after direct interaction between nodes; The recommended trust evaluation scheme design module is used to design the recommended trust evaluation scheme of nodes based on the direct trust evaluation scheme between nodes, and improve the trust evaluation scheme of heterogeneous network nodes; The comprehensive trust evaluation scheme design module is used to design a comprehensive trust evaluation scheme for nodes based on the direct trust evaluation scheme between nodes and the recommended trust evaluation scheme for nodes, and finally complete the trust evaluation of 6G heterogeneous network nodes.

8. A node trust evaluation device in a 6G heterogeneous network scenario, characterized in that: include: Memory: a computer program storing a node trust evaluation method in a 6G heterogeneous network scenario according to any one of claims 1 to 6, which is a computer-readable device; Processor: used to implement a node trust evaluation method in a 6G heterogeneous network scenario as described in any one of claims 1-6 when executing the computer program.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it can implement a node trust evaluation method in a 6G heterogeneous network scenario as described in any one of claims 1-6.

Images