Trust evaluation method, device and equipment

Abstract

The invention provides a trust evaluation method, device and equipment. In the method, a first device requests a trust evaluation node to perform trust evaluation on a plurality of candidate devices, and the trust evaluation node obtains an initial trust value of each candidate device and divides the plurality of candidate devices into at least two trust types. And taking the target equipment as an example, the trust evaluation node determines a weight value corresponding to the direct trust value according to each type of candidate equipment, and determines a plurality of recommended equipment and a weight value corresponding to each recommended equipment. And the trust evaluation node determines the trust value of the target device according to the weight value corresponding to the direct trust value, the plurality of recommended devices and the weight value corresponding to each recommended device. And after the trust evaluation node determines the trust value of each candidate device, the trust evaluation node determines a trusted device and feeds back the trusted device to the first device. Through the scheme, the trust evaluation node can determine the parameter value required for calculating the trust value based on the candidate devices of different trust types, and the accuracy of trust evaluation is improved.

Description

[0001] This application is a divisional application of the original application with the application number 202210426054.9 and the original filing date of April 21, 2022, and the entire contents of the original application are incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the field of communication technology, in particular to a trust evaluation method, device and equipment. BACKGROUND

[0003] With the gradual commercialization of the 5th generation mobile communication technology (5G) network, more and more research has begun to focus on the future network vision. At present, the industry generally believes that the integration of heterogeneous networks is an inevitable trend of future network development, and the future network will move from independent and closed networks to heterogeneous interconnection, such as the integration of cellular networks, internet protocol (IP) networks, and satellite networks. The integration of heterogeneous networks has the following advantages: network integration can expand the coverage of the network, making the network more scalable; network integration can make full use of existing network resources, reduce operating costs, and enhance competitiveness; network integration can provide various services to different users and better meet the needs of future network users for diversity; network integration can improve network reliability, attack resistance, and other capabilities. However, different heterogeneous networks have different hardware network devices, and how to solve the mutual trust problem between different network devices is one of the great challenges faced by network integration.

[0004] Trust modeling based on data theory is currently a method for determining whether a device is a trusted device. In this method, a model of dynamic trust relationship is established by various mathematical methods and tools, the direct trust and indirect trust values corresponding to the device are combined together to form a final trust value. However, this method analyzes the trust relationship in a specific scenario, and then uses mathematical methods to combine the trust relationship to form a complete trust model, so this method has the problem of serious scene customization and poor portability. In addition, the trust model in this method involves many weight values and parameters, and these weight values or parameter values are the experience values of technical personnel, and setting different weight values or parameters has a great impact on the performance of the trust model, therefore, the robustness of the trust modeling method based on data theory is not high. SUMMARY

[0005] The present application provides a trust evaluation method, device and equipment to provide an accurate trust evaluation method suitable for various scenarios.

[0006] In a first aspect, the present application provides a trust evaluation method, which can be applied to a trust evaluation node, and the method comprises the following steps: receiving a trust evaluation request sent by a first device, the trust evaluation request comprising the identities of a plurality of candidate devices to be evaluated; obtaining an initial trust value of each candidate device in the plurality of candidate devices, and dividing the plurality of candidate devices into at least two trust types according to the initial trust values of the plurality of candidate devices; performing the following operations on each candidate device in the plurality of candidate devices to obtain the trust values of the plurality of candidate devices: determining a weight value corresponding to a direct trust value of a target device according to the candidate devices corresponding to each trust type; and determining a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device according to the candidate devices corresponding to each trust type; determining the trust value of the target device according to the weight value corresponding to the direct trust value of the target device, the plurality of recommended devices corresponding to the target device, and the weight value corresponding to each recommended device; wherein the target device is any device in the plurality of candidate devices; the direct trust value of the target device is used to indicate the trust degree of the target device relative to the first device; and determining a trusted device in the plurality of candidate devices according to the trust values of the plurality of candidate devices, and sending the identity of the trusted device to the first device.

[0007] In the above method, after receiving the trust evaluation request sent by the first device, the trust evaluation node can obtain the initial trust values of the plurality of candidate devices requested by the first device to perform trust evaluation, and divide the plurality of candidate devices into at least two trust types. The trust evaluation node determines the weight value corresponding to the direct trust value of each candidate device according to the candidate devices corresponding to each trust type, and determines the plurality of recommended devices corresponding to each candidate device and the weight value corresponding to each recommended device, so as to obtain the parameter values required when calculating the trust value of the candidate device, so as to obtain an accurate trust value and improve the accuracy of trust evaluation. When the trust evaluation of the candidate device is performed by the above scheme, the analysis of whether the device is trusted in a specific scenario is not performed, but the candidate devices are classified according to the initial trust values of the devices, and the parameters required in the trust evaluation process are determined according to the classified candidate devices of each type. Therefore, the trust evaluation method provided by the present application can be applied to various trust evaluation scenarios, has wide applicability, and can solve the device trust evaluation problem in various communication networks.

[0008] In one possible design, the determination of the weight value corresponding to the direct trust value of the target device according to the candidate devices corresponding to each trust type comprises the following steps: performing a plurality of rounds of weight calculation processes on the target device based on the first reinforcement learning model to obtain a plurality of candidate weight values; and determining the weight value corresponding to the direct trust value of the target device according to the plurality of candidate weight values. wherein any round of weight calculation process comprises: determining an interaction device selection strategy of the round of weight calculation process as a first strategy or a second strategy, determining at least one interaction device according to the interaction device selection strategy of the round of weight calculation process; wherein the first strategy is to determine a target trust type from the at least two trust types with a preset probability, and select one candidate device corresponding to the target trust type as an interaction device; the second strategy is to select a node with the highest trust value in a neighbor node of the target device as an interaction device, wherein the trust evaluation node stores the trust values of the neighbor nodes of the target device; obtaining interaction data of communication interaction between the at least one interaction device and the target device; determining a first reward value and a first state value of the round of weight calculation process based on the first reinforcement learning model according to the interaction data, and storing the first reward value of the round of weight calculation process; calculating a candidate weight value corresponding to the round of weight calculation process according to the first reward value and the first state value of the round of weight calculation process; wherein the first reward value is used to indicate the performance of the first reinforcement learning model in the round of weight calculation process; and the first state value is used to indicate the communication state of the target device and the at least one interaction device.

[0009] Through the design, the trust evaluation node can learn the calculation method of the weight value corresponding to the direct trust value based on the first reinforcement learning model, so as to determine the weight value corresponding to the direct trust value matched with the target device, and improve the accuracy of trust evaluation.

[0010] In one possible design, the determination of the interaction device selection strategy of the round of weight calculation as the first strategy or the second strategy comprises: determining a probability value of the interaction device selection strategy as the first strategy according to a preset decision function, and determining the interaction device selection strategy according to a random function corresponding to the probability value.

[0011] Through the design, when the first reinforcement learning model determines the weight value corresponding to the direct trust value of the target device, the interaction device selection strategy of each round of weight calculation process is determined with different probability values as the multiple rounds of weight calculation are performed, so as to accelerate the optimization speed of the reinforcement learning model, and solve the problems of slow convergence speed and unstable reward value of the existing reinforcement learning model.

[0012] In one possible design, the determination of the weight value corresponding to the direct trust value of the target device according to the multiple candidate weight values comprises: selecting a candidate weight value calculated in a round of weight calculation with the highest first reward value from the multiple candidate weight values; and taking the selected candidate weight value as the weight value corresponding to the direct trust value of the target device.

[0013] Through the design, the trust evaluation node can take the result of the weight calculation of the round with the highest reward value as the weight value corresponding to the direct trust value of the target device, so as to take the result of the weight calculation of the round with the best model performance in the multiple rounds of weight calculation of the first reinforcement learning model as the final result, thereby improving the accuracy of weight calculation.

[0014] In a possible design, the method further includes: after the M rounds of weight calculation for the target device, and / or, if the difference between the first reward values of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, stopping the multiple rounds of weight calculation; where M and N are positive integers, and M is greater than N.

[0015] Through the design, if the difference between the reward values of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than the preset threshold, it can be considered that the first reward value converges, indicating that the performance of the first reinforcement learning model tends to be stable, and at this time, the multiple rounds of weight calculation can be stopped; or the number of rounds of weight calculation of the first reinforcement learning model can be preset, and after the current number of rounds reaches the preset number of rounds, the multiple rounds of weight calculation are stopped, thereby preventing the first reinforcement learning model from entering an infinite loop and reducing efficiency.

[0016] In a possible design, the interaction data includes the number of successful communications and the number of failed communications between the target device and the at least one interaction device.

[0017] Through the design, the interaction data obtained by the first reinforcement learning model can be the number of successful communications and the number of failed communications between the target device and the at least one interaction device, so that the state value of the environment corresponding to the first reinforcement learning model can be determined according to the interaction data, and then the strategy of the first reinforcement learning model can be adjusted to obtain a first reinforcement learning model with better performance.

[0018] In a possible design, the determining, according to the candidate devices corresponding to each trust type, of the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device includes: obtaining multiple groups of candidate results by performing, by the second reinforcement learning model, multiple rounds of recommended device selection processes for the target device, each group of candidate results including multiple candidate recommended devices and the weight value corresponding to each candidate recommended device; and determining a target result from the multiple groups of candidate results, the target result including the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device. Any round of recommended device selection process includes: determining that the recommended device selection strategy of the current round of recommended device selection process is the third strategy or the fourth strategy, determining a plurality of first candidate recommended devices according to the recommended device selection strategy; wherein the third strategy is to select a candidate device that has communicated with the target device from each trust type of candidate device as a recommended device in a preset proportion; the fourth strategy is to select a preset number of nodes that have communicated with the target device from the neighbor nodes of the target device as recommended devices; determining a second state value of the current round of recommended device selection process based on the second reinforcement learning model according to the plurality of first candidate recommended devices; the second state value is used to indicate the communication state of the target device and the plurality of first candidate recommended devices; performing a plurality of rounds of weight calculation process based on the second reinforcement learning model to determine the weight value corresponding to each first candidate recommended device; determining a second reward value of the current round of recommended device selection process according to the plurality of first candidate recommended devices and the weight value corresponding to each first candidate recommended device, and storing the second reward value, the second reward value is used to indicate the performance of the second reinforcement learning model in the current round of recommended device selection process.

[0019] Through the design, the trust evaluation node can learn the selection method of the recommended device and the method of calculating the weight value corresponding to the recommended device based on the second reinforcement learning model, so as to determine the recommended device matched with the target device and the weight value corresponding to each recommended device, and improve the accuracy of trust evaluation.

[0020] In one possible design, the determination of the recommended device selection strategy of the current round of recommended device selection as the third strategy or the fourth strategy includes: determining a probability value of the recommended device selection strategy as the third strategy according to a preset decision function, and determining the recommended device selection strategy according to a random function corresponding to the probability value.

[0021] In one possible design, the determination of the target result from the plurality of groups of candidate results includes: selecting a candidate result determined by a round of recommended device selection with the highest second reward value from the plurality of groups of candidate results; and taking the selected candidate result as the target result.

[0022] In one possible design, the method further includes: stopping the plurality of rounds of recommended device selection when the target device is subjected to Q rounds of recommended device selection, and / or the difference between the reward values of any two adjacent rounds of recommended device selection in the last P rounds of recommended device selection for the target device is less than a preset threshold; wherein Q and P are positive integers, and Q is greater than P.

[0023] In a possible design, the obtaining of the initial trust value of each candidate device in the plurality of candidate devices comprises: receiving trust-related data of the plurality of candidate devices sent by the first device, and determining the initial trust value of each candidate device according to the trust-related data of each candidate device; wherein the trust-related data comprises at least one of a device manufacturer, a device usage range, a device importance, and a device deployment location; or receiving the initial trust value of the plurality of candidate devices sent by the first device; the initial trust value of the plurality of candidate devices is obtained by the first device from a blockchain or an interstellar file storage system.

[0024] Through this design, the trust evaluation node can receive the trust-related data of the plurality of candidate devices sent by the first device, and determine the initial trust value of each candidate device according to the trust-related data of each candidate device; or the first device can obtain the initial trust value of each candidate device from the blockchain or the interstellar file storage system, and then send the initial trust value of each candidate device to the trust evaluation node, so as to flexibly obtain the initial trust value to start the trust evaluation process of the candidate device.

[0025] In a possible design, after sending the identification of the trusted device to the first device, the method further comprises: receiving interaction data between the first device and the trusted device sent by the first device, and updating at least one of the following according to the interaction data between the first device and the trusted device: the model parameter of the first reinforcement learning model, the model parameter of the second reinforcement learning model, and the trust value of the trusted device.

[0026] Through this design, the trust evaluation node can optimize the model parameters of the first reinforcement learning model and the second reinforcement learning model according to the real interaction data between the first device and the trusted device, and further improve the performance of the two reinforcement learning models. Moreover, the first reinforcement learning model can determine the weight value corresponding to the direct trust value of the trusted device based on the interaction data between the first device and the trusted device, the second reinforcement learning model can determine the plurality of recommended devices corresponding to the trusted device and the weight value corresponding to each recommended device based on the interaction data between the first device and the trusted device, and then the trust evaluation node can redetermine the trust value of the trusted device, so as to update the trust value of the trusted device according to the real interaction data and further improve the accuracy of the trust evaluation.

[0027] In one possible design, determining the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value corresponding to each recommended device includes: determining the indirect trust value of the target device based on the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device, wherein the indirect trust value is used to indicate the degree of trust of the target device relative to the multiple recommended devices corresponding to the target device; determining the direct trust value of the target device; and determining the trust value of the target device based on the direct trust value of the target device, the indirect trust value, and the weight value corresponding to the direct trust value.

[0028] Through this design, the trust evaluation node can determine the indirect trust value based on the multiple recommended devices corresponding to the target device determined by the second reinforcement learning model and the weight value of each recommended device, and determine the trust value of the target device based on the weight value, direct trust value and indirect trust value corresponding to the direct trust value of the target device determined by the first reinforcement learning model, thereby improving the accuracy of trust evaluation and ensuring communication security.

[0029] Secondly, embodiments of this application provide a trust assessment device, which includes multiple functional modules. These multiple functional modules interact to implement the methods described in the first aspect and its various embodiments. The multiple functional modules can be implemented based on software, hardware, or a combination of both, and can be arbitrarily combined or divided based on specific implementations.

[0030] Thirdly, embodiments of this application provide an electronic device, including a processor and a memory, wherein the memory stores computer program instructions, and when the electronic device is running, the processor executes the method provided in the first aspect above.

[0031] Fourthly, embodiments of this application also provide a computer program that, when run on a computer, causes the computer to perform the methods provided in any of the above aspects.

[0032] Fifthly, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a computer, causes the computer to perform the methods provided in any of the above aspects.

[0033] Sixthly, embodiments of this application also provide a chip for reading a computer program stored in a memory and executing the method provided in any of the above aspects.

[0034] Seventhly, embodiments of this application also provide a chip system including a processor for supporting a computer device in implementing the methods provided in any of the above aspects. In one possible design, the chip system further includes a memory for storing programs and data necessary for the computer device. The chip system may be composed of chips or may include chips and other discrete devices. Attached Figure Description

[0035] Figure 1 A schematic diagram of a trust mathematical model provided in an embodiment of this application; Figure 2 This is a schematic diagram of two machine learning-based classification models; Figure 3 A schematic diagram illustrating a scenario to which the trust assessment method provided in the embodiments of this application is applicable; Figure 4 This is a schematic diagram of the structure of a trust evaluation node provided in an embodiment of this application; Figure 5 This application provides a functional diagram of each module within a trust assessment node, as illustrated in an embodiment of the present application. Figure 6 A flowchart illustrating a trust assessment method provided in this application embodiment; Figure 7 A flowchart illustrating a trust assessment method provided in this application embodiment; Figure 8 A flowchart illustrating a trust assessment method provided in this application embodiment; Figure 9 A flowchart illustrating a trust assessment method provided in this application embodiment; Figure 10 This is a schematic diagram of the structure of a trust assessment device provided in an embodiment of this application; Figure 11 This is a schematic diagram of the structure of a trust assessment device provided in an embodiment of this application. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the embodiments of this application will be further described in detail below with reference to the accompanying drawings. In the description of the embodiments of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first" and "second" may explicitly or implicitly include one or more of that feature.

[0037] It should be understood that in the embodiments of this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, a and b, a and c, b and c, or a, b, and c, where a, b, and c can be single or multiple.

[0038] The following explains the terms used in the embodiments of this application: (1) Reinforcement learning (RL) is an important branch of machine learning. Reinforcement learning learns from the target through an evaluation feedback mechanism. The difference between reinforcement learning and supervised learning is that the target in reinforcement learning does not have a label. Reinforcement learning is usually described using a Markov decision process (MDP), which includes two objects: an agent and an environment. In reinforcement learning, the environment is in a specific state, and the agent can change the state of the environment by performing specific actions. After the environment state changes, it returns an observation to the agent, which can be a state value indicating the current state of the environment. Simultaneously, the environment also returns a reward value to the agent, which can be positive or negative. A negative reward can be considered a penalty. The agent takes a new action based on the state value and reward value returned by the environment; how the agent chooses an action is called the policy. Repeating the above process, the task of reinforcement learning is to find a policy to maximize the reward value.

[0039] The Actor-Critic (AC) algorithm is a reinforcement learning method that combines policy gradient and temporal difference learning. The Actor network is the policy function, typically implemented using a neural network. Its input is the current state, and its output is an action. The Critic network is the value function, evaluating the performance of the current Actor network. The Actor network updates its policy based on the Critic's evaluation, and the goal of training the Actor network is to maximize the expected cumulative reward.

[0040] (2) Deep belief network (DBN) is a type of deep neural network based on a Boltzmann pre-trained structure. DBN can be used for both unsupervised and supervised learning. In the embodiments of this application, the DBN model can classify the input data by training the weights between its neurons.

[0041] (3) A blockchain is a chain composed of blocks. Each block stores a certain amount of information, and multiple blocks are linked together in the order they were generated. This chain is stored on all servers, and as long as one server in the entire system is working, the entire blockchain is secure. These servers are called nodes in the blockchain system, and they provide storage space and computing power support for the entire blockchain system. To modify information in the blockchain, the consent of more than half of the nodes must be obtained and the information in all nodes must be modified. These nodes are usually controlled by different entities, so tampering with information in the blockchain is extremely difficult. Compared with traditional networks, blockchain has two core characteristics: data is difficult to tamper with and it is decentralized. Based on these two characteristics, the information recorded by the blockchain is more authentic and reliable, which can help solve the problem of mutual distrust among people. In the embodiments of this application, blockchain can refer to a storage system that stores data based on the blockchain data structure, such as one or more servers that store data based on the blockchain data structure.

[0042] The embodiments of this application are described below: With the gradual commercialization of 5G networks, more and more research is focusing on the vision of future networks. Currently, the industry generally believes that the convergence of heterogeneous networks is an inevitable trend in future network development. Future networks will move from independent and closed systems to heterogeneous interconnection, such as the integration of cellular networks, Internet Protocol (IP) networks, and satellite networks. The convergence of heterogeneous networks has the following advantages: it can expand network coverage, making the network more scalable; it can fully utilize existing network resources, reduce operating costs, and enhance competitiveness; it can provide various services to different users, better meeting the diverse needs of future network users; and it can improve network reliability and resistance to attacks. However, different heterogeneous networks use different hardware network devices. To ensure communication efficiency and security, solving the mutual trust problem between different network devices is one of the major challenges facing network convergence.

[0043] In one alternative implementation, trust modeling can be performed using data theory to establish a mathematical model of trust that can dynamically assess the trust value of devices. For example, Figure 1 This is a schematic diagram of a trust mathematical model provided in an embodiment of this application. (Reference) Figure 1 This trust mathematical model is used to determine the trust value of a trusted device relative to a trusting device. For ease of description, the trusting device is referred to as device i, and the trusted device as device j. When calculating the trust value of device j relative to device i, the direct trust value and the indirect trust value of device j can be calculated separately. The direct trust value of device j can be determined based on the interaction data between device j and device i, which can be the number of successful and failed communications between device i and device j. The indirect trust value of device j depends on whether the recommending device corresponding to device j is trustworthy and the degree of trust the recommending device has in device j; we refer to the recommending device of device j as device k. (Reference) Figure 1 In this trust mathematical model, the trust value of device j can satisfy the following formula 1: Formula 1 in, Let be the trust value of device j relative to device i. Let be the direct trust value of device j relative to device i. The weight value corresponding to the direct trust value. This is the indirect trust value for device j.

[0044] Optionally, the direct trust value of device j relative to device i can satisfy either Formula 2 or Formula 3: Formula 2 Formula 3 in, Let be the number of normal data packets in the data exchange between device i and device j. Let be the number of malicious data packets in the interaction data between device i and device j. The decay exponent, This is a constant. It should be noted that normal data packets can be those indicating successful communication, while malicious data packets can be those indicating failed communication.

[0045] Optionally, the indirect trust value of device B can satisfy the following formula 4: Formula 4 in, Let j be the set of devices that have had communication interactions with device j. Let k be the trust value of device k relative to device i. Let k be the trust value of device k relative to device j.

[0046] It should be noted that in Formula 4 This is one way to set the weight value corresponding to device k. The weight value corresponding to device k can also be set to the value based on the experience of technical personnel.

[0047] Optionally, the weight value corresponding to the direct trust value can satisfy the following formula 5 or formula 6: Formula 5 Formula 6 in, Let c be the total number of data packets in the interaction data between device i and device j, where c is a constant. This is the attenuation coefficient.

[0048] As can be seen from Formulas 5 and 6, as the number of communication interactions between device i and device j increases, the weight value corresponding to the direct trust value increases. At this time, the direct trust value between device i and device j has a greater impact on the trust value of device j.

[0049] As can be seen from the above formulas, the trust mathematical model includes many weighting factors and parameters. For example, the weight value corresponding to the direct trust value in Formula 1. The parameters such as c in Formula 5, and other weighting factors and parameters, are generally empirical values ​​obtained by technicians based on simulation data. Setting different weighting factors and parameters has a significant impact on the performance of the trust mathematical model. Therefore, the robustness of trust modeling methods based on data theory is not high. In addition, trust relationships will change in different application scenarios. The trust mathematical model can only analyze trust relationships in specific scenarios, resulting in scenario-customized verification of the trust mathematical model, poor transferability, and difficulty in applying it to trust assessment of actual network systems.

[0050] In another alternative implementation, a machine learning-based model for device trust assessment can be trained. Since machine learning-based trust modeling is less dependent on trust relationships and generally does not require mathematical modeling of trust relationships for specific scenarios, it is applicable to a wider range of scenarios compared to the aforementioned mathematical trust models. When performing trust modeling based on machine learning, machine learning classification algorithms can be used to classify devices based on device data to determine their trustworthiness. However, this process requires labeled sample data to train the classification model. For trust scenarios, it is difficult to obtain labeled trust data. For example, a device's trust type may differ in different scenarios, making it difficult to determine whether the device's data corresponds to a trustworthy or untrustworthy label. Therefore, machine learning-based trust modeling struggles to obtain sample data for model training, making its practical application in device trust assessment difficult. Furthermore, when assessing device trust, some devices may lack interaction data; that is, these devices have not yet communicated with other devices. Machine learning-based classification models cannot determine the trustworthiness of devices without existing interaction data. Thus, cold start is also a problem that machine learning-based classification models struggle to solve.

[0051] For example, Figure 2 This is a schematic diagram illustrating two machine learning-based classification models. (Reference) Figure 2 (a) in the middle Figure 2 Figure (a) shows the distribution of sample data with different trust labels. When classifying the device data of the device to be evaluated, a circular area is obtained with the location of the device data of the device to be evaluated as the center and a classification radius K. In this area, if the number of sample data with the trust label is large, the SVM model determines that the device to be evaluated is trustworthy; if the number of sample data with the trust label is untrustworthy, the SVM model determines that the device to be evaluated is untrustworthy. Based on Figure 2 As can be seen in (a), this classification model requires a large amount of sample data with trust labels, and different values ​​of the classification radius K have a significant impact on the performance of the SVM model. For example, as... Figure 2 As shown in (a), when K is 3, the SVM determines the device to be evaluated to be untrustworthy because the number of samples labeled as untrustworthy is greater than the number of samples labeled as trustworthy. When K is 7, the SVM determines the device to be evaluated to be trustworthy because the number of samples labeled as trustworthy is greater than the number of samples labeled as untrustworthy. Since the value of K is generally an empirical value used by technical personnel, the performance of the SVM classification model in assessing device trust is not stable.

[0052] Figure 2(b) shows the distribution of sample data with different trust labels. The classification model needs to determine a hyperplane based on this distribution using a regression algorithm. This hyperplane can classify the trust types of the sample data to achieve trust assessment of the device. Therefore, this classification model also requires a large amount of sample data with trust labels to obtain a relatively accurate hyperplane.

[0053] In summary, current solutions for trust assessment of devices suffer from problems such as severe scenario customization, difficulty in obtaining large amounts of sample data with trust labels, low robustness, and difficulty in cold start, making them unsuitable for trust assessment of devices in actual network systems.

[0054] To address the aforementioned issues, this application provides a trust assessment method that is accurate and applicable to various scenarios. Figure 3 This is a schematic diagram illustrating a scenario to which the trust assessment method provided in this application is applicable. (Reference) Figure 3 This scenario includes a primary device, multiple candidate devices, a trust evaluation node, and a storage system. Furthermore, the storage system can be a blockchain or the Interplanetary File System (IPFS). The primary device initiates the trust evaluation, and the multiple candidate devices are those for which the primary device requests trust evaluation. For example, before communicating with the multiple candidate devices, the primary device requests the trust evaluation node to perform trust evaluations on the candidate devices. The trust evaluation node can be used to perform trust evaluations on the multiple candidate devices to determine whether each candidate device is trustworthy. The blockchain can be used to store and transmit data from each device. IPFS can also be used to store data from each device to alleviate the storage pressure on the blockchain.

[0055] In some embodiments of this application, reference is made to Figure 3In the scenario shown, a first device can send a trust assessment request to a trust assessment node. This request may include the identifiers of multiple candidate devices to be assessed. Upon receiving the trust assessment request from the first device, the trust assessment node can obtain the initial trust value of each candidate device and, based on these initial trust values, categorize the candidate devices into at least two trust types. The trust assessment node can then perform a trust assessment on each candidate device to determine its trust value. Taking the trust assessment node's trust assessment of a target device as an example, the target device is any one of the multiple candidate devices. The trust assessment node can determine the weight value corresponding to the direct trust value of the target device based on a first reinforcement learning model, and, based on the candidate devices corresponding to each trust type, determine the recommended devices and their corresponding weight values ​​based on a second reinforcement learning model. The trust assessment node can then determine the target device's trust value based on the weight value corresponding to its direct trust value, the recommended devices, and their corresponding weight values. In this way, the trust assessment node can perform a trust assessment on each candidate device, thereby determining its trust value. Then, the trust evaluation node can determine the trustworthy device from the multiple candidate devices based on their trust values, and send the identifier of the trustworthy device to the first device. After receiving the identifier of the trustworthy device sent by the trust evaluation node, the first device can communicate and interact with the trustworthy device to ensure communication efficiency and security.

[0056] The functions performed by the trust evaluation node in the trust evaluation method provided in this application embodiment are further described below: Figure 4 This is a schematic diagram of the structure of a trust evaluation node provided in an embodiment of this application. (Reference) Figure 4 The trust assessment node may include a data collection module, a classification module, a reinforcement learning module, a basic mathematical model (BMM) module, and a trust decision module. It is understood that these multiple modules can be integrated into the same device or deployed on different devices; this application embodiment does not limit this. The following... Figure 4 The functions of each module in the trust assessment node shown are introduced below: Figure 5 This is a functional diagram of each module within a trust assessment node provided in an embodiment of this application.

[0057] refer to Figure 5The data collection module is used to collect trust-related data about the device. Optionally, the trust-related data may include at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location. This trust-related data can be used to determine the device's initial trust value.

[0058] The classification module can be based on a deep belief network (DBN). This module categorizes devices based on their initial trust level. For example, a DBN-based classification module can classify devices according to their initial trust level. Figure 3 In the scenario shown, the initial trust value of multiple candidate devices determines the trust type of each candidate device. Optionally, the DBN-based classification module can classify multiple candidate devices into at least two trust types. For example, if the DBN-based classification module performs binary classification on multiple candidate devices, the trust type of the candidate devices can be trusted or untrustworthy; if the DBN-based classification module performs ternary classification on multiple candidate devices, the trust type of the candidate devices can be trusted, generally trusted, or untrustworthy. Similarly, the classification module can also classify multiple candidate devices into more types, and in specific implementations, the number of trust types that the classification module can classify can be dynamically adjusted according to the scenario.

[0059] The reinforcement learning module can be based on a double actor-critic (DAC) model. This module learns the weights corresponding to direct trust values ​​and how to select the recommended device and the weights for each device. Optionally, the DAC-based reinforcement learning module can include two reinforcement learning models, such as a first model and a second model. The first model can learn either the weights corresponding to direct or indirect trust values, while the second model can learn how to select the recommended device and the weights for each device. It's important to note that the sum of the weights corresponding to direct and indirect trust values ​​is 1. The first model can learn either the weights corresponding to direct or indirect trust values ​​to determine the other.

[0060] The BMM module is used to calculate the direct trust value and indirect trust value of a device, as well as to calculate the device's trust value by weighting the direct and indirect trust values.

[0061] The trust decision module is used to determine the trust assessment result of a device based on its trust value. For example, it can determine the trusted device among multiple devices based on their trust values.

[0062] based on Figure 4 and Figure 5The following describes the steps of the trust evaluation node in the trust evaluation method provided in this application embodiment to perform trust evaluation on multiple candidate devices. Figure 6 A flowchart illustrating a trust assessment method provided in an embodiment of this application. (Reference) Figure 6 The method includes the following steps: S601: The data collection module acquires trust-related data for each of the multiple candidate devices.

[0063] The trust-related data for each candidate device may include at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location.

[0064] S602: The data collection module sends the trust-related data of each candidate device to the classification module.

[0065] S603: The classification module determines the initial trust value of each candidate device based on the trust-related data of each candidate device, and classifies multiple candidate devices into at least two trust types based on the initial trust values ​​of multiple candidate devices.

[0066] Optionally, candidate devices of the same trust type can be grouped into a set. For example, assuming the classification module uses a binary classification DBN model to classify candidate devices, multiple candidate devices can be divided into a set of trusted devices and a set of untrusted devices based on the trust type determined by the classification module.

[0067] S604: The classification module sends the trust type of each candidate device among multiple candidate devices to the first reinforcement learning model and the second reinforcement learning model in the reinforcement learning module.

[0068] The first reinforcement learning model is used to determine the weight value corresponding to the direct trust value of each candidate device; the second reinforcement learning model is used to determine the multiple recommended devices corresponding to each candidate device, and the weight value corresponding to each of the multiple recommended devices.

[0069] S605: The first reinforcement learning model determines whether the interaction device selection strategy in the current round of weight calculation for the target device is the first strategy. If yes, proceed to S606; otherwise, determine that the interaction device selection strategy in the current round of weight calculation is the second strategy, and proceed to S607.

[0070] The target device is any one of multiple candidate devices.

[0071] In this embodiment, the first reinforcement learning model can perform multiple rounds of weight calculations on the target device to determine the weight value corresponding to the direct trust value of the target device. During each round of weight calculation, at least one interactive device can be selected to interact with the target device and obtain interaction data. The interaction device selection strategy in any round of weight calculation can be a first strategy or a second strategy. The first strategy can also be called an exploration strategy, and the second strategy can also be called an exploitation strategy. Specifically, the first strategy involves determining the target trust type from at least two trust types with a preset probability, and selecting a candidate device from the candidate devices corresponding to the target trust type as the interactive device. The second strategy involves selecting the node with the highest trust value among the target device's neighboring nodes as the interactive device.

[0072] In one optional implementation, the first reinforcement learning model can determine the interaction device selection strategy for the current round of weight calculation based on a preset decision function, such as the preset decision function being the following formula 7: Formula 7 in, , , It is a constant. The probability value for selecting the exploration strategy as the interaction device in this round of weight calculation process. t is the time between the start of weight calculation for the target device and the current round of weight calculation, or t can be the number of rounds in the weight calculation for the target device. For example, if this round of weight calculation is the 3rd round of weight calculation for the target device, then t can be 3.

[0073] It should be noted that the prediction function shown in Formula 7 yields the probability value that the interaction device selection strategy in this round of weight calculation is the exploration strategy. After determining the probability value of the interaction device selection strategy in this round of weight calculation according to Formula 7, the first reinforcement learning model can determine whether the interaction device selection strategy in this round of weight calculation is the exploration strategy or the exploit strategy based on the random function corresponding to the determined probability value. For example, the first reinforcement learning model can determine the interaction device selection strategy in this round of weight calculation through the random function Random(X), where X is the probability value. For example, when X=0.7, Random(0.7) means that the probability of the interaction device selection strategy in this round of weight calculation being the exploration strategy is 0.7, and the first reinforcement learning model uses this probability to randomly select and determine the interaction device selection strategy in this round of weight calculation.

[0074] In some embodiments of this application, during the multi-round weight calculation process for the target device, the first reinforcement learning model can learn how to determine the weight value corresponding to the direct trust value. The strategy for determining the weight value corresponding to the direct trust value is adjusted through multiple rounds of weight calculation, and the output of the optimized first reinforcement learning model is used as the weight value corresponding to the direct trust value of the target device. The interaction device selection strategy for each round of weight calculation can be an exploration strategy or an exploitation strategy. Specifically, when the interaction device selection strategy for a round of weight calculation is an exploration strategy, it indicates that there is relatively little interaction data between the current target device and other devices. In this round of weight calculation, candidate devices can be randomly selected to interact with the target device to obtain interaction data. When the interaction device selection strategy for a round of weight calculation is an exploitation strategy, it indicates that the current target device has already communicated with other devices multiple times. In this round of weight calculation, neighboring nodes that have communicated with the target device can be selected to interact with the target device to obtain interaction data. Therefore, referring to Formula 7, according to the decision function provided in the embodiments of this application, when the value of t corresponding to a round of weight calculation is larger, it indicates that multiple rounds of weight calculation have been performed before this round of weight calculation, and the target device has communicated and interacted with multiple candidate devices or neighboring nodes. At this time, the probability value of the interaction device selection strategy for this round of weight calculation determined by the first reinforcement learning module as an exploration strategy is smaller, and the interaction device selection strategy for this round of weight calculation determined by the random function corresponding to the probability value is more likely to be an exploit strategy.

[0075] For example, Formula 7 can be set as:

[0076] Assuming that t in the above formula represents the number of rounds in the weight calculation targeting the target device, when t=1, the probability value calculated by this formula for the interaction device selection strategy in this round of weight calculation to be an exploration strategy is approximately 0.956. Therefore, when determining the interaction device selection strategy in this round of weight calculation based on a random function with a probability value of 0.956, the more likely result is that the interaction device selection strategy in this round of weight calculation is an exploration strategy. When t=30, the probability value calculated by this formula for the interaction device selection strategy in this round of weight calculation to be an exploration strategy is approximately 0.301. Therefore, when determining the interaction device selection strategy in this round of weight calculation based on a random function with a probability value of 0.301, the more likely result is that the interaction device selection strategy in this round of weight calculation is an exploitation strategy.

[0077] Through the above design, the trust evaluation method provided in this application embodiment can determine the interaction device selection strategy for each round of weight calculation with different probability values ​​when determining the weight value corresponding to the direct trust value of the target device based on the reinforcement learning model, as multiple rounds of weight calculation proceed. This accelerates the optimization speed of the reinforcement learning model and solves the problems of slow convergence speed and unstable reward value in the existing reinforcement learning model.

[0078] S606: The first reinforcement learning model determines the target trust type from at least two trust types with a preset probability, and selects a candidate device from the candidate devices corresponding to the target trust type as the interaction device.

[0079] In one optional implementation, when the interaction device selection strategy in this round of weight calculation is an exploration strategy, the first reinforcement learning model can determine the target trust type from at least two trust types classified by the classification model with a preset probability, and select a candidate device from the candidate devices corresponding to the target trust type as the interaction device. For example, assuming that the classification module in S603 classifies multiple candidate devices into two trust types: trustworthy and untrustworthy, the multiple candidate devices can be divided into a set of trustworthy devices and a set of untrustworthy devices according to the trust type determined by the classification module. The first reinforcement learning model can then select a candidate device from the candidate devices corresponding to the target trust type. The target trust type is determined by a random function. If the target trust type is determined to be trustworthy, a candidate device is randomly selected from the set of trustworthy devices as the interaction device. If the target trust type is determined to be untrustworthy, a candidate device is randomly selected from the set of untrustworthy devices as the interaction device.

[0080] S607: The first reinforcement learning model selects the node with the highest trust value among the target device's neighboring nodes as the interaction device.

[0081] In one optional implementation, when the interaction device selection strategy in this round of weight calculation is the exploitation strategy, the first reinforcement learning model can select a node as the interaction device from the neighboring nodes of the target device, for example, selecting the node with the highest trust value among the neighboring nodes as the interaction device.

[0082] S608: The first reinforcement learning model acquires interaction data of at least one interactive device communicating with the target device, determines the first reward value and the first state value of the current weight calculation process based on the acquired interaction data, and determines the candidate weight value corresponding to the current weight calculation process based on the first reward value and the first state value of the current weight calculation process.

[0083] Optionally, the interaction data can be the number of successful communication and the number of communication failures when the target device communicates with at least one interactive device.

[0084] In this embodiment, the first reinforcement learning model can determine a first reward value and a first state value in the current round of weight calculation based on the acquired interaction data. The first reward value in the current round of weight calculation can indicate the performance of the first reinforcement learning model in this round of weight calculation, and the first state value can indicate the communication status between the target device and at least one interactive device. The first reinforcement learning model can determine candidate weight values ​​obtained from the current round of weight calculation based on the first reward value and the first state value. The determination of candidate weight values ​​by the first reinforcement learning module can be seen as a new action taken by the first reinforcement learning model based on the first reward value and the first state value. Through multiple rounds of weight calculation, the first reinforcement learning model can optimize the strategy for determining candidate weight values, thereby determining the weight value corresponding to the direct trust value of the target device.

[0085] S609: The first reinforcement learning model determines whether it can end the multi-round weight calculation. If yes, proceed to S610; otherwise, return to S605.

[0086] In one optional implementation, the first reinforcement learning model may stop multi-round weight calculation when the reward value converges and / or the current round reaches a preset number of rounds. For example, after performing M rounds of weight calculation for the target device, and / or when the difference between the reward values ​​of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, multi-round weight calculation is stopped; where M and N are positive integers, and M is greater than N.

[0087] It is understandable that the convergence of reward values ​​indicates that the performance of the first reinforcement learning model is stabilizing, at which point multiple rounds of weight calculation can be stopped; or the number of rounds for weight calculation of the first reinforcement learning model can be preset, and after the current round reaches the preset number of rounds, multiple rounds of weight calculation can be stopped to prevent the first reinforcement learning model from entering an infinite loop and reducing efficiency.

[0088] S610: The first reinforcement learning model determines the weight value corresponding to the direct trust value of the target device.

[0089] Optionally, the first reinforcement learning model can use the candidate weight value calculated in the round with the highest reward value in the multi-round weight calculation as the weight value corresponding to the direct trust value of the target device.

[0090] S611: The second reinforcement learning model determines whether the recommended device selection strategy for the current round of recommended device selection process for the target device is the third strategy. If yes, proceed to S612; otherwise, determine that the recommended device selection strategy for the current round of recommended device selection process is the fourth strategy, and proceed to S613.

[0091] In this embodiment, the second reinforcement learning model can perform multiple rounds of device recommendation selection for the target device to determine multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device. In each round of device recommendation selection, multiple candidate recommended devices can be selected, and the weight value corresponding to each candidate recommended device is determined. The device recommendation strategy in any round of device recommendation selection can be a third strategy or a fourth strategy. The third strategy can also be called an exploration strategy, and the fourth strategy can also be called an exploitation strategy. Specifically, the third strategy selects candidate devices that have communicated with the target device from each type of trust candidate device at a preset ratio as recommended devices; the fourth strategy selects a preset number of nodes that have communicated with the target device from the target device's neighbor nodes as recommended devices.

[0092] Optionally, when determining the recommended device selection strategy for this round of recommended device selection, the second reinforcement learning model can determine the probability value of the recommended device selection strategy being an exploration strategy based on the decision function provided in the embodiments of this application (such as Formula 7), and then determine whether the recommended device selection strategy for this round is an exploration strategy or an exploitation strategy based on the random function corresponding to the determined probability value. For specific implementation details, please refer to S605; repeated details will not be elaborated further.

[0093] S612: The second reinforcement learning model selects candidate devices that have had communication interactions with the target device from candidate devices of each trust type as the first alternative recommended devices according to a preset ratio.

[0094] In some optional implementations, when the recommendation device selection strategy in this round of device selection is an exploratory strategy, the second reinforcement learning model can select candidate devices that have interacted with the target device as first-option recommended devices from candidate devices of each trust type, and the number of candidate devices selected as first-option recommended devices from candidate devices of each trust type meets a preset ratio. For example, assuming that the classification module in S603 divides multiple candidate devices into two trust types: trusted and untrusted, the multiple candidate devices can be divided into a set of trusted devices and a set of untrusted devices according to the trust type determined by the classification module. The second reinforcement learning model can select recommended devices from the set of trusted devices and the set of untrusted devices at a preset ratio of X:Y. Specifically, the second reinforcement learning model can select X candidate devices from the set of trusted devices as first-option recommended devices and Y devices from the set of untrusted devices as first-option recommended devices, where X and Y are positive integers, for example, X can be 1 and Y can be 2.

[0095] S613: The second reinforcement learning model selects a preset number of nodes that have had communication interactions with the target device from the target device's neighboring nodes as the first candidate recommended devices.

[0096] In some optional implementations, when the recommended device selection strategy in this round of recommended device selection process is an exploitation strategy, the second reinforcement learning model can select a preset number of nodes from the target device's neighboring nodes as the first candidate recommended devices, such as selecting Q nodes from the target device's neighboring nodes that have had communication interactions with the target device as the first candidate recommended devices, where Q is a positive integer.

[0097] S614: The second reinforcement learning module determines the second reward value and the second state value for the current round of recommended device selection process based on the selected first candidate recommended devices.

[0098] The second reward value for the device selection in this round can be used to indicate the performance of the second reinforcement learning model in the device selection in this round, and the second state value can be used to indicate the communication status between the target device and the multiple first candidate recommended devices obtained in this round.

[0099] In one optional implementation, the second reinforcement learning model determines multiple first candidate recommended devices in this round of device selection. This can be seen as a new action taken by the second reinforcement learning model. The second reinforcement learning model can determine the second state value of the device selection process in this round based on the determined multiple recommended devices. For the determined multiple first candidate recommended devices, the second reinforcement learning model can determine the weight value corresponding to each first candidate recommended device through multiple rounds of weight calculation. In specific implementation, the second reinforcement learning model can determine the type of each round of weight calculation process according to the decision function (such as Formula 7) provided in the embodiments of this application, and then determine the weight value corresponding to each first candidate recommended device through multiple rounds of weight calculation. The second reinforcement learning model can determine the second reward value of the device selection process in this round based on the multiple first candidate recommended devices determined in each round of device selection and the weight value corresponding to each first candidate recommended device. Repeating the above process, the second reinforcement learning performs multiple rounds of device selection to obtain multiple sets of candidate results. Each set of candidate results includes multiple candidate recommended devices and the weight value corresponding to each candidate recommended device. By selecting recommended devices in multiple rounds, the strategy for selecting recommended devices and determining the weight value of each recommended device in the second reinforcement learning model can be optimized. Then, based on the optimized second reinforcement learning model, multiple recommended devices corresponding to the target device and the weight value of each recommended device can be determined.

[0100] S615: The second reinforcement learning model determines whether the multi-round recommendation device selection can be terminated; if yes, proceed to S616; otherwise, return to S611.

[0101] In one optional implementation, the second reinforcement learning model may stop multi-round device selection when the reward value converges and / or the current round reaches a preset number of rounds. For example, after Q rounds of device selection for the target device, and / or when the difference between the reward values ​​of any two adjacent rounds of device selection in the last P rounds of device selection for the target device is less than a preset threshold, multi-round device selection is stopped; where Q and P are positive integers, and Q is greater than P.

[0102] S616: The second reinforcement learning model determines multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device.

[0103] In one optional implementation, after the second reinforcement learning model performs multiple rounds of device recommendation selection for the target device, it can obtain multiple sets of candidate results. Each set of candidate results includes multiple candidate recommended devices and a weight value corresponding to each candidate recommended device. After stopping the multiple rounds of device recommendation selection, the second reinforcement learning model determines the target result from the multiple sets of candidate results. The target result includes multiple recommended devices corresponding to the target device and a weight value corresponding to each recommended device.

[0104] Optionally, the first reinforcement learning model can select the candidate result determined in the second round of the recommended device selection process with the highest second reward value from multiple candidate results, and use the selected candidate result as the target result.

[0105] S617: The reinforcement learning module sends the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value of each recommended device to the BMM module.

[0106] S618: The BMM module determines the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value of each recommended device.

[0107] In some embodiments of this application, the BMM module can determine the indirect trust value of the target device based on multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device. The indirect trust value of the target device can be used to indicate the degree of trust between the target device and its multiple recommended devices. For example, the BMM module can use the value in Formula 4... The indirect trust value of the target device is calculated by replacing the weight value corresponding to each recommended device determined by the second reinforcement model in this embodiment. The BMM module can also determine the direct trust value of the target device. The direct trust value of the target device can be used to indicate the degree of trust of the target device relative to the first device. For example, the BMM module can calculate the direct trust value of the target device according to Formula 2 or Formula 3. After determining the direct and indirect trust values ​​of the target device, the BMM module can determine the trust value of the target device based on the direct trust value, the indirect trust value, and the weight value corresponding to the direct trust value. For example, the BMM module can calculate the trust value of the target device according to Formula 1.

[0108] Repeat steps S605-S618 until the trust value of each candidate device among multiple candidate devices is determined.

[0109] S619: The BMM module sends the trust value of each candidate device among multiple candidate devices to the trust decision module.

[0110] S620: The trust decision module determines the trust assessment result based on the trust value of each candidate device among multiple candidate devices.

[0111] For example, the trust decision module determines the device identifier of a trustworthy device among multiple candidate devices based on the trust value of each candidate device among multiple candidate devices.

[0112] Using the above method, the trust evaluation node can perform trust evaluations on multiple candidate devices requested by the first device to determine whether the multiple candidate devices are trustworthy. The trust evaluation method provided in this application embodiment can be applied to various trust evaluation scenarios and can solve the cold start trust evaluation problem where the device to be evaluated does not have interactive data, thereby improving the accuracy of trust evaluation and ensuring communication security.

[0113] The trust assessment method provided in this application is further illustrated below with two examples: Example 1 Figure 7 A flowchart illustrating a trust assessment method provided in an embodiment of this application. This method can be... Figure 3 The scenario shown involves the execution of a first device, a trust evaluation node, and a storage system. In this example, the storage system can be a blockchain.

[0114] S701: The trust assessment node acquires trust-related data from multiple devices.

[0115] The trust-related data for each device may include at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location.

[0116] S702: The trust assessment node determines the initial trust value of each device based on the trust-related data of each device.

[0117] S703: The trust assessment node sends the initial trust value of each device to the blockchain.

[0118] In some embodiments of this application, the blockchain can store the trust value of each device. Optionally, the trust value of each device can be an initial trust value determined by a trust evaluation node based on the trust-related data of the device, or it can be a trust value obtained by a trust evaluation node after performing a trust evaluation on the device based on the trust evaluation method provided in the embodiments of this application.

[0119] S704: The first device sends a request message to the blockchain, which requests the initial trust value of multiple candidate devices to be evaluated.

[0120] S705: The blockchain sends the initial trust value of each of the multiple candidate devices to the first device.

[0121] S706: The first device sends a trust assessment request to the trust assessment node.

[0122] Optionally, the trust assessment request may include the identifiers of multiple candidate devices to be assessed and the initial trust value of each candidate device.

[0123] S707: The trust assessment node classifies multiple candidate devices into at least two trust types based on their initial trust values.

[0124] S708: The trust evaluation node determines the weight value corresponding to the direct trust value of each candidate device based on the first reinforcement learning model, and determines the recommended device and the weight value corresponding to each recommended device based on the second reinforcement learning model, according to the candidate devices corresponding to each trust type; and determines the trust value of each candidate device based on the weight value corresponding to the direct trust value of each candidate device, the recommended device corresponding to each candidate device, and the weight value corresponding to each recommended device.

[0125] It should be noted that the method by which the trust evaluation node determines the trust value of each candidate device can be found in [reference needed]. Figure 6 The embodiments shown are not described in detail here.

[0126] S709: The trust evaluation node determines the trustworthy device among multiple candidate devices based on the trust value of multiple candidate devices.

[0127] S710: The trust assessment node sends the identifier of the trusted device to the first device.

[0128] S711: The first device communicates and interacts with trusted devices, and records the interaction data between the first device and each trusted device.

[0129] S712: The first device sends the interaction data between the first device and each trusted device to the trust evaluation node.

[0130] S713: The trust evaluation node updates the model parameters of the first reinforcement learning model, the model parameters of the second reinforcement learning model, and the trust value of the trusted device based on the interaction data between the first device and the trusted device.

[0131] In one optional implementation, the first reinforcement learning model can determine a state value based on the interaction data between the first device and the trusted device, and determine the weight value corresponding to the direct trust value of the trusted device based on the state value. Through this process, the model parameters of the first reinforcement learning model can be optimized based on real interaction data, thereby further improving the performance of the first reinforcement learning model.

[0132] Similarly, the second reinforcement learning model can determine the state value based on the interaction data between the first device and the trusted device, and determine multiple recommended devices corresponding to the trusted device and the weight value corresponding to each recommended device based on the state value. Through this process, the model parameters of the second reinforcement learning model can be optimized based on real interaction data, thereby further improving the performance of the second reinforcement learning model.

[0133] In some embodiments, the first reinforcement learning model can determine the weight value corresponding to the direct trust value of the trusted device based on the interaction data between the first device and the trusted device, and the second reinforcement learning model can determine multiple recommended devices corresponding to the trusted device and the weight value corresponding to each recommended device based on the interaction data between the first device and the trusted device. In this way, the trust evaluation node can redetermine the trust value of the trusted device, so as to update the trust value of the trusted device according to the real interaction data and further improve the accuracy of trust evaluation.

[0134] S714: The trust evaluation node sends the updated trust value of the trusted device to the blockchain.

[0135] Example 2

[0136] Figure 8 A flowchart illustrating a trust assessment method provided in an embodiment of this application. This method can be... Figure 3 The scenario depicts the execution of a first device, a trust evaluation node, and a storage system. In this example, the storage system can be a blockchain or IPFS.

[0137] S801: The trust assessment node acquires trust-related data from multiple devices.

[0138] The trust-related data for each device may include at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location.

[0139] S802: The trust assessment node determines the initial trust value of each device based on the trust-related data of each device.

[0140] S803: The trust evaluation node sends the initial trust value of each device and the trust-related data of each device to IPFS.

[0141] S804: The trust assessment node sends the hash of the initial trust value of each device to the blockchain.

[0142] In one optional implementation, the trust evaluation node can send the device's trust value and related trust data to IPFS. IPFS can store the device's trust value and related trust data to alleviate the data storage pressure on the blockchain. The trust evaluation node can also send the hash value of the device's trust value to the blockchain. Storing the hash value of the device's trust value in the blockchain can achieve data integrity protection, preventing the device's trust value stored in IPFS from being tampered with and ensuring data security. The device's trust value stored in IPFS can be an initial trust value determined by the trust evaluation node based on the device's related trust data, or it can be a trust value obtained by the trust evaluation node after performing a trust evaluation on the device using the trust evaluation method provided in this application.

[0143] S805: The first device sends a request message to IPFS, which requests the initial trust values ​​of multiple candidate devices to be evaluated.

[0144] S806: IPFS sends the initial trust value of each of the multiple candidate devices to the first device.

[0145] In one alternative implementation, the first device may request initial trust values ​​for a plurality of candidate devices to be evaluated from IPFS, and IPFS may send the initial trust values ​​of the plurality of candidate devices to be evaluated to the first device.

[0146] In another optional implementation, the first device can also request the hash values ​​of the initial trust values ​​of the multiple candidate devices to be evaluated from the blockchain. After the blockchain sends the hash values ​​of the initial trust values ​​of the multiple candidate devices to the first device, the first device can send the hash values ​​of the initial trust values ​​of the multiple candidate devices to IPFS to obtain the initial trust values ​​of the multiple candidate devices stored in IPFS. This design ensures that the initial trust values ​​of the multiple candidate devices obtained by the first device are tamper-proof data, guaranteeing data security.

[0147] S807: The first device sends a trust assessment request to the trust assessment node.

[0148] Optionally, the trust assessment request may include the identifiers of multiple candidate devices to be assessed and the initial trust value of each candidate device.

[0149] S808: The trust evaluation node classifies multiple candidate devices into at least two trust types based on their initial trust values.

[0150] S809: The trust evaluation node determines the weight value corresponding to the direct trust value of each candidate device based on the first reinforcement learning model, and determines the multiple recommended devices and the weight value corresponding to each recommended device based on the second reinforcement learning model, according to the candidate devices corresponding to each trust type; and determines the trust value of each candidate device based on the weight value corresponding to the direct trust value of each candidate device, the multiple recommended devices corresponding to each candidate device, and the weight value corresponding to each recommended device.

[0151] It should be noted that the method by which the trust evaluation node determines the trust value of each candidate device can be found in [reference needed]. Figure 6 The embodiments shown are not described in detail here.

[0152] S810: The trust evaluation node determines the trustworthy device among multiple candidate devices based on the trust value of multiple candidate devices.

[0153] S811: The trust evaluation node sends the identifier of the trusted device to the first device.

[0154] S812: The first device communicates and interacts with trusted devices, and records the interaction data between the first device and each trusted device.

[0155] S813: The first device sends the interaction data between the first device and each trusted device to the trust evaluation node.

[0156] S814: The trust evaluation node updates the model parameters of the first reinforcement learning model, the model parameters of the second reinforcement learning model, and the trust value of the trusted device based on the interaction data between the first device and the trusted device.

[0157] S815: The trust evaluation node sends the updated trust value of the trusted device to IPFS.

[0158] S816: The trust evaluation node sends the hash value of the updated trust value of the trusted device to the blockchain.

[0159] Based on the same technical concept, this application also provides a trust assessment method, which can be used by... Figure 3 The trust assessment node and the first device are executed in the scenario shown. Figure 9 A flowchart illustrating a trust assessment method provided in an embodiment of this application. (Reference) Figure 9 The method includes the following steps: S901: The first device sends a trust assessment request to the trust assessment node.

[0160] The trust assessment request includes the identifiers of multiple candidate devices to be assessed.

[0161] S902: The trust evaluation node obtains the initial trust value of each candidate device among multiple candidate devices, and classifies the multiple candidate devices into at least two trust types based on the initial trust values ​​of the multiple candidate devices.

[0162] S903: The trust evaluation node performs the following operations on each of the multiple candidate devices to obtain the trust value of the multiple candidate devices: Based on the candidate devices corresponding to each trust type, it determines the weight value corresponding to the direct trust value of the target device; and based on the candidate devices corresponding to each trust type, it determines the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device. The trust evaluation node determines the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value corresponding to each recommended device.

[0163] The target device is any one of multiple candidate devices; the direct trust value of the target device is used to indicate the degree of trust of the target device relative to the first device. S904: The trust evaluation node determines the trustworthy device from among multiple candidate devices based on the trust values ​​of multiple candidate devices.

[0164] S905: The trust assessment node sends the identifier of the trusted device to the first device.

[0165] It should be noted that this application Figure 9 The trust assessment method shown can be referred to in the above embodiments of this application for specific implementation, and repeated parts will not be described again.

[0166] Based on the same technical concept, this application also provides a trust assessment device 1000, which can be applied to... Figure 3 The trust evaluation node in the scenario shown is used to achieve... Figure 6- Figure 9 The functions performed by the trust assessment node in the trust assessment method shown. Figure 10This is a schematic diagram of the structure of a trust assessment device 1000 provided in an embodiment of this application. The trust assessment device 1000 includes a receiving unit 1001, a processing unit 1002, and a sending unit 1003. The processing unit 1002 can be further divided into... Figure 4 The multiple modules shown below perform different functions as follows: Figure 5 As shown, it will not be elaborated further here.

[0167] The functions of each unit in the trust assessment device 1000 are described below.

[0168] The receiving unit 1001 is used to receive a trust assessment request sent by the first device, wherein the trust assessment request includes the identifiers of multiple candidate devices to be assessed. Processing unit 1002 is configured to obtain an initial trust value for each of the plurality of candidate devices; classify the plurality of candidate devices into at least two trust types based on their initial trust values; and perform the following operations on each of the plurality of candidate devices to obtain their trust values: determining a weight value corresponding to the direct trust value of the target device based on the candidate devices corresponding to each trust type; determining a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device based on the candidate devices corresponding to each trust type; determining the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the plurality of recommended devices corresponding to the target device, and the weight value corresponding to each recommended device; wherein, the target device is any device among the plurality of candidate devices; the direct trust value of the target device is used to indicate the degree of trust of the target device relative to the first device; and determining a trustworthy device among the plurality of candidate devices based on their trust values. The sending unit 1003 is used to send the identifier of the trusted device to the first device.

[0169] In one embodiment, the processing unit 1002 is specifically used to: perform a multi-round weight calculation process on the target device based on a first reinforcement learning model to obtain multiple candidate weight values; and determine the weight value corresponding to the direct trust value of the target device based on the multiple candidate weight values. The weight calculation process in any round includes: determining the interaction device selection strategy for this round of weight calculation as either a first strategy or a second strategy; determining at least one interaction device according to the interaction device selection strategy for this round of weight calculation; wherein, the first strategy is to determine a target trust type from the at least two trust types with a preset probability, and select a candidate device as the interaction device from the candidate devices corresponding to the target trust type; the second strategy is to select the node with the highest trust value among the neighbor nodes of the target device as the interaction device, wherein the trust evaluation node stores the trust values ​​of the neighbor nodes of the target device; acquiring interaction data of communication between the at least one interaction device and the target device; determining a first reward value and a first state value for this round of weight calculation based on the interaction data and the first reinforcement learning model, and storing the first reward value for this round of weight calculation; calculating the candidate weight value corresponding to this round of weight calculation based on the first reward value and the first state value for this round of weight calculation; wherein, the first reward value is used to indicate the performance of the first reinforcement learning model in this round of weight calculation; the first state value is used to indicate the communication status between the target device and the at least one interaction device.

[0170] In one embodiment, the processing unit 1002 is specifically used to: determine the probability value of the interaction device selection strategy being the first strategy according to a preset decision function, and determine the interaction device selection strategy according to the random function corresponding to the probability value.

[0171] In one embodiment, the processing unit 1002 is specifically used to: select, from the plurality of candidate weight values, the candidate weight value calculated in the weight calculation process with the highest first reward value in the first round of weight calculation; and use the selected candidate weight value as the weight value corresponding to the direct trust value of the target device.

[0172] In one embodiment, the processing unit 1002 is further configured to: after performing M rounds of weight calculation for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, stop the multi-round weight calculation; wherein M and N are positive integers, and M is greater than N.

[0173] In one embodiment, the processing unit 1002 is specifically used to: include the number of successful communication interactions and the number of communication failures between the target device and the at least one interactive device.

[0174] In one embodiment, the processing unit 1002 is specifically used to: perform a multi-round recommended device selection process for the target device based on a second reinforcement learning model to obtain multiple sets of candidate results, each set of candidate results including multiple candidate recommended devices and a weight value corresponding to each candidate recommended device; determine the target result from the multiple sets of candidate results, the target result including: multiple recommended devices corresponding to the target device and a weight value corresponding to each recommended device. The process of selecting a recommended device in any round includes: determining the recommended device selection strategy for this round as either a third strategy or a fourth strategy; determining multiple first candidate recommended devices based on the recommended device selection strategy; wherein the third strategy is to select candidate devices that have communicated with the target device from candidate devices of each trust type at a preset ratio as recommended devices; the fourth strategy is to select a preset number of nodes that have communicated with the target device from the target device's neighbor nodes as recommended devices; determining a second state value for this round of recommended device selection based on the multiple first candidate recommended devices and the second reinforcement learning model; the second state value is used to indicate the communication status between the target device and the multiple first candidate recommended devices; performing a multi-round weight calculation process based on the second reinforcement learning model to determine the weight value corresponding to each first candidate recommended device; determining a second reward value for this round of recommended device selection based on the multiple first candidate recommended devices and the weight value corresponding to each first candidate recommended device, and storing the second reward value, which is used to indicate the performance of the second reinforcement learning model in this round of recommended device selection.

[0175] In one embodiment, the processing unit 1002 is specifically used to: determine the probability value of the recommended device selection strategy being the third strategy according to a preset decision function, and determine the recommended device selection strategy according to the random function corresponding to the probability value.

[0176] In one embodiment, the processing unit 1002 is specifically used to: select the candidate result determined in the round of recommended device selection process with the highest second reward value from the multiple sets of candidate results; and use the selected candidate result as the target result.

[0177] In one embodiment, the processing unit 1002 is further configured to: after performing Q rounds of recommended device selection for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of recommended device selection in the last P rounds of recommended device selection for the target device is less than a preset threshold, stop the multi-round recommended device selection; wherein Q and P are positive integers, and Q is greater than P.

[0178] In one embodiment, the processing unit 1002 is specifically configured to: receive trust-related data of the plurality of candidate devices sent by the first device through the receiving unit 1001, and determine an initial trust value for each candidate device based on the trust-related data of each candidate device; wherein the trust-related data includes at least one of device manufacturer, device usage scope, device importance, and device deployment location; or receive the initial trust values ​​of the plurality of candidate devices sent by the first device through the receiving unit 1001; the initial trust values ​​of the plurality of candidate devices are obtained by the first device from a blockchain or interplanetary file storage system.

[0179] In one embodiment, the processing unit 1002 is further configured to: receive interaction data between the first device and the trusted device sent by the first device through the receiving unit 1001, and update at least one of the following based on the interaction data between the first device and the trusted device: model parameters of the first reinforcement learning model, model parameters of the second reinforcement learning model, and trust value of the trusted device. The sending unit 1003 is also used to send the updated trust value of the trusted device to the blockchain or interplanetary file storage system.

[0180] In one embodiment, the processing unit 1002 is specifically configured to: determine an indirect trust value for the target device based on a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device, wherein the indirect trust value is used to indicate the degree of trust of the target device relative to the plurality of recommended devices corresponding to the target device; determine a direct trust value for the target device; and determine a trust value for the target device based on the direct trust value, the indirect trust value, and the weight value corresponding to the direct trust value.

[0181] Based on the same technical concept, this application also provides a trust assessment device 1100. Figure 11 This is a schematic diagram of the structure of a trust assessment device 1100 provided in an embodiment of this application. The trust assessment device 1100 can be used to implement... Figure 6- Figure 9 The trust assessment node in the illustrated trust assessment method performs certain functions. (See also...) Figure 11 As shown, the trust assessment device 1100 includes: a communication interface 1101, a processor 1102, and a memory 1103.

[0182] Optionally, the communication interface 1101, the processor 1102, and the memory 1103 are interconnected via a bus 1104. The bus 1104 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 11 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0183] The communication interface 1101 is used to receive and send data, enabling communication and interaction with other devices. For example, the communication interface 1101 can be a transceiver, circuit, bus, module, or other type of communication interface.

[0184] In one embodiment of this application, the trust assessment device 1100 can implement this application. Figure 6- Figure 9 The functions of the trust evaluation node in the trust evaluation method shown, and the functions of processor 1102 are described below: Processor 1102 is configured to receive a trust assessment request sent by a first device via communication interface 1101, the trust assessment request including the identifiers of a plurality of candidate devices to be assessed; obtain an initial trust value for each of the plurality of candidate devices; classify the plurality of candidate devices into at least two trust types based on the initial trust values; perform the following operations on each of the plurality of candidate devices to obtain a trust value for the plurality of candidate devices: determine a weight value corresponding to the direct trust value of the target device based on the candidate devices corresponding to each trust type; determine a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device based on the candidate devices corresponding to each trust type; determine the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the plurality of recommended devices corresponding to the target device, and the weight value corresponding to each recommended device; wherein, the target device is any device among the plurality of candidate devices; the direct trust value of the target device is used to indicate the degree of trust of the target device relative to the first device; determine a trustworthy device among the plurality of candidate devices based on the trust values ​​of the plurality of candidate devices; and send the identifier of the trustworthy device to the first device via communication interface 1101.

[0185] In one embodiment, the processor 1102 is specifically configured to: perform a multi-round weight calculation process on the target device based on a first reinforcement learning model to obtain multiple candidate weight values; and determine the weight value corresponding to the direct trust value of the target device based on the multiple candidate weight values. The weight calculation process in any round includes: determining the interaction device selection strategy for this round of weight calculation as either a first strategy or a second strategy; determining at least one interaction device according to the interaction device selection strategy for this round of weight calculation; wherein, the first strategy is to determine a target trust type from the at least two trust types with a preset probability, and select a candidate device as the interaction device from the candidate devices corresponding to the target trust type; the second strategy is to select the node with the highest trust value among the neighbor nodes of the target device as the interaction device, wherein the trust evaluation node stores the trust values ​​of the neighbor nodes of the target device; acquiring interaction data of communication between the at least one interaction device and the target device; determining a first reward value and a first state value for this round of weight calculation based on the interaction data and the first reinforcement learning model, and storing the first reward value for this round of weight calculation; calculating the candidate weight value corresponding to this round of weight calculation based on the first reward value and the first state value for this round of weight calculation; wherein, the first reward value is used to indicate the performance of the first reinforcement learning model in this round of weight calculation; the first state value is used to indicate the communication status between the target device and the at least one interaction device.

[0186] In one embodiment, the processor 1102 is specifically used to: determine the probability value of the interaction device selection strategy being the first strategy according to a preset decision function, and determine the interaction device selection strategy according to the random function corresponding to the probability value.

[0187] In one embodiment, the processor 1102 is specifically used to: select, from the plurality of candidate weight values, the candidate weight value calculated in the weight calculation process with the highest first reward value in the first round of weight calculation; and use the selected candidate weight value as the weight value corresponding to the direct trust value of the target device.

[0188] In one embodiment, the processor 1102 is further configured to: after performing M rounds of weight calculation for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, stop the multi-round weight calculation; wherein M and N are positive integers, and M is greater than N.

[0189] In one embodiment, the processor 1102 is specifically used to: include the number of successful communication interactions and the number of communication failures between the target device and the at least one interactive device.

[0190] In one embodiment, the processor 1102 is specifically configured to: perform a multi-round recommended device selection process for the target device based on a second reinforcement learning model to obtain multiple sets of candidate results, each set of candidate results including multiple candidate recommended devices and a weight value corresponding to each candidate recommended device; determine the target result from the multiple sets of candidate results, the target result including: multiple recommended devices corresponding to the target device and a weight value corresponding to each recommended device. The process of selecting a recommended device in any round includes: determining the recommended device selection strategy for this round as either a third strategy or a fourth strategy; determining multiple first candidate recommended devices based on the recommended device selection strategy; wherein the third strategy is to select candidate devices that have communicated with the target device from candidate devices of each trust type at a preset ratio as recommended devices; the fourth strategy is to select a preset number of nodes that have communicated with the target device from the target device's neighbor nodes as recommended devices; determining a second state value for this round of recommended device selection based on the multiple first candidate recommended devices and the second reinforcement learning model; the second state value is used to indicate the communication status between the target device and the multiple first candidate recommended devices; performing a multi-round weight calculation process based on the second reinforcement learning model to determine the weight value corresponding to each first candidate recommended device; determining a second reward value for this round of recommended device selection based on the multiple first candidate recommended devices and the weight value corresponding to each first candidate recommended device, and storing the second reward value, which is used to indicate the performance of the second reinforcement learning model in this round of recommended device selection.

[0191] In one embodiment, the processor 1102 is specifically used to: determine the probability value of the recommended device selection strategy being the third strategy according to a preset decision function, and determine the recommended device selection strategy according to the random function corresponding to the probability value.

[0192] In one embodiment, the processor 1102 is specifically used to: select the candidate result determined in the round of recommended device selection process with the highest second reward value from the multiple sets of candidate results; and use the selected candidate result as the target result.

[0193] In one embodiment, the processor 1102 is further configured to: after performing Q rounds of recommended device selection for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of recommended device selection in the last P rounds of recommended device selection for the target device is less than a preset threshold, stop the multi-round recommended device selection; wherein Q and P are positive integers, and Q is greater than P.

[0194] In one embodiment, the processor 1102 is specifically configured to: receive trust-related data of the plurality of candidate devices sent by the first device through the communication interface 1101, and determine an initial trust value for each candidate device based on the trust-related data of each candidate device; wherein the trust-related data includes at least one of device manufacturer, device usage scope, device importance, and device deployment location; or receive the initial trust values ​​of the plurality of candidate devices sent by the first device through the communication interface 1101; the initial trust values ​​of the plurality of candidate devices are obtained by the first device from a blockchain or interplanetary file storage system.

[0195] In one embodiment, the processor 1102 is further configured to: receive interaction data between the first device and the trusted device sent by the first device through the communication interface 1101, and update at least one of the following based on the interaction data between the first device and the trusted device: model parameters of the first reinforcement learning model, model parameters of the second reinforcement learning model, and trust value of the trusted device. The communication interface 1101 is also used to send the updated trust value of the trusted device to a blockchain or interplanetary file storage system.

[0196] In one embodiment, the processor 1102 is specifically configured to: determine an indirect trust value for the target device based on a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device, wherein the indirect trust value is used to indicate the degree of trust of the target device relative to the plurality of recommended devices corresponding to the target device; determine a direct trust value for the target device; and determine a trust value for the target device based on the direct trust value, the indirect trust value, and the weight value corresponding to the direct trust value.

[0197] It is understood that this application Figure 11The memory 1103 can be volatile memory or non-volatile memory, or may include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM). It should be noted that the memory used in the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0198] Based on the above embodiments, this application also provides a computer program that, when run on a computer, causes the computer to execute the methods described in the embodiments of this application.

[0199] Based on the above embodiments, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a computer, causes the computer to perform the methods described in the embodiments of this application.

[0200] Based on the above embodiments, this application also provides a chip for reading computer programs stored in a memory to implement the methods described in the embodiments of this application.

[0201] Based on the above embodiments, this application provides a chip system including a processor for supporting a computer device in implementing the methods described in the embodiments of this application. In one possible design, the chip system further includes a memory for storing necessary programs and data of the computer device. This chip system may be composed of chips or may include chips and other discrete devices.

[0202] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0203] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure One One or more processes and / or boxes Figure One A device that provides the functions specified in one or more boxes.

[0204] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure One One or more processes and / or boxes Figure One The function specified in one or more boxes.

[0205] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure One One or more processes and / or boxes Figure One The steps of the function specified in one or more boxes.

[0206] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of protection of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A trust assessment method, characterized in that, Applied to trust evaluation nodes, the method includes: Receive a trust assessment request sent by a first device, the trust assessment request including the identifiers of multiple candidate devices to be assessed; Obtain the initial trust value of each candidate device among the plurality of candidate devices, and classify the plurality of candidate devices into at least two trust types based on the initial trust values ​​of the plurality of candidate devices; For each of the plurality of candidate devices, the following operations are performed to obtain the trust value of the plurality of candidate devices: Based on the candidate devices corresponding to each trust type, a weight value corresponding to the direct trust value of the target device is determined; and based on the candidate devices corresponding to each trust type, a plurality of recommended devices corresponding to the target device and a weight value corresponding to each recommended device are determined; based on the weight value corresponding to the direct trust value of the target device, the plurality of recommended devices corresponding to the target device, and the weight value corresponding to each recommended device, the trust value of the target device is determined; wherein, the target device is any device among the plurality of candidate devices; the direct trust value of the target device is used to indicate the degree of trust of the target device relative to the first device; Based on the trust values ​​of the multiple candidate devices, a trustworthy device is determined from the multiple candidate devices, and the identifier of the trustworthy device is sent to the first device.

2. The method as described in claim 1, characterized in that, The step of determining the weight value corresponding to the direct trust value of the target device based on the candidate devices corresponding to each trust type includes: Based on the first reinforcement learning model, a multi-round weight calculation process is performed on the target device to obtain multiple candidate weight values; based on the multiple candidate weight values, the weight value corresponding to the direct trust value of the target device is determined; The weight calculation process in any round includes: The interaction device selection strategy for this round of weight calculation is determined to be either a first strategy or a second strategy. At least one interaction device is determined according to the interaction device selection strategy for this round of weight calculation. The first strategy is to determine a target trust type from the at least two trust types with a preset probability, and select a candidate device from the candidate devices corresponding to the target trust type as the interaction device. The second strategy is to select the node with the highest trust value among the neighboring nodes of the target device as the interaction device. The trust evaluation node stores the trust values ​​of the neighboring nodes of the target device. The system acquires interaction data between the at least one interactive device and the target device; based on the interaction data, it determines a first reward value and a first state value for the current round of weight calculation based on the first reinforcement learning model, and stores the first reward value for the current round of weight calculation; based on the first reward value and the first state value for the current round of weight calculation, it calculates the candidate weight value corresponding to the current round of weight calculation; wherein, the first reward value is used to indicate the performance of the first reinforcement learning model in the current round of weight calculation; the first state value is used to indicate the communication status between the target device and the at least one interactive device.

3. The method as described in claim 2, characterized in that, The strategy for selecting the interactive device for this round of weight calculation is either a first strategy or a second strategy, including: The probability value of the interaction device selection strategy being the first strategy is determined according to a preset decision function, and the interaction device selection strategy is determined according to the random function corresponding to the probability value.

4. The method as described in claim 2 or 3, characterized in that, The step of determining the weight value corresponding to the direct trust value of the target device based on the plurality of candidate weight values ​​includes: Among the multiple candidate weight values, the candidate weight value calculated in the round of weight calculation with the highest first reward value is selected; the selected candidate weight value is used as the weight value corresponding to the direct trust value of the target device.

5. The method according to any one of claims 2-4, characterized in that, The method further includes: After performing M rounds of weight calculation for the target device, and / or, if the difference between the first reward values ​​of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, the multi-round weight calculation is stopped; where M and N are positive integers, and M is greater than N.

6. The method according to any one of claims 2-5, characterized in that, The interaction data includes the number of successful communication interactions and the number of communication failures between the target device and the at least one interactive device.

7. The method according to any one of claims 1-6, characterized in that, The step of determining multiple recommended devices corresponding to the target device and a weight value for each recommended device based on the candidate devices corresponding to each trust type includes: Based on the second reinforcement learning model, a multi-round device recommendation process is performed on the target device to obtain multiple sets of candidate results. Each set of candidate results includes multiple candidate recommended devices and a weight value corresponding to each candidate recommended device. From the multiple sets of candidate results, the target result is determined. The target result includes: multiple recommended devices corresponding to the target device and a weight value corresponding to each recommended device. The process of selecting recommended equipment in any round includes: The recommended device selection strategy for this round of recommended device selection process is determined to be either the third strategy or the fourth strategy. Based on the recommended device selection strategy, multiple first candidate recommended devices are determined. The third strategy is to select candidate devices that have communicated with the target device from candidate devices of each trust type at a preset ratio as recommended devices. The fourth strategy is to select a preset number of nodes that have communicated with the target device from the target device's neighbor nodes as recommended devices. Based on the plurality of first candidate recommended devices, a second state value is determined for the current round of recommended device selection process based on the second reinforcement learning model; the second state value is used to indicate the communication status between the target device and the plurality of first candidate recommended devices; a multi-round weight calculation process is performed based on the second reinforcement learning model to determine the weight value corresponding to each first candidate recommended device; a second reward value is determined for the current round of recommended device selection process based on the plurality of first candidate recommended devices and the weight value corresponding to each first candidate recommended device, and the second reward value is stored; the second reward value is used to indicate the performance of the second reinforcement learning model in the current round of recommended device selection process.

8. The method as described in claim 7, characterized in that, The recommended device selection strategy for determining the recommended device selection in this round is either the third strategy or the fourth strategy, including: The probability value of the recommended device selection strategy being the third strategy is determined according to a preset decision function, and the recommended device selection strategy is determined according to the random function corresponding to the probability value.

9. The method as described in claim 7 or 8, characterized in that, Determining the target result from the plurality of candidate results includes: Among the multiple sets of alternative results, the alternative result determined in the round of recommended equipment selection process with the highest second reward value is selected; the selected alternative result is taken as the target result.

10. The method according to any one of claims 7-9, characterized in that, The method further includes: After Q rounds of recommended device selection for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of recommended device selection in the final P rounds of recommended device selection for the target device is less than a preset threshold, the multi-round recommended device selection is stopped; where Q and P are positive integers, and Q is greater than P.

11. The method according to any one of claims 1-10, characterized in that, The step of obtaining the initial trust value of each candidate device among the plurality of candidate devices includes: The system receives trust-related data of the plurality of candidate devices sent by the first device, and determines an initial trust value for each candidate device based on the trust-related data of each candidate device; wherein the trust-related data includes at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location; or The system receives the initial trust values ​​of the plurality of candidate devices sent by the first device; the initial trust values ​​of the plurality of candidate devices are obtained by the first device from a blockchain or an interplanetary file storage system.

12. The method according to any one of claims 2-11, characterized in that, After sending the identifier of the trusted device to the first device, the method further includes: The system receives interaction data between the first device and the trusted device sent by the first device, and updates at least one of the following based on the interaction data between the first device and the trusted device: the model parameters of the first reinforcement learning model, the model parameters of the second reinforcement learning model, and the trust value of the trusted device.

13. The method according to any one of claims 1-12, characterized in that, The step of determining the trust value of the target device based on the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value corresponding to each recommended device includes: Based on the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device, the indirect trust value of the target device is determined, wherein the indirect trust value is used to indicate the degree of trust of the target device relative to the multiple recommended devices corresponding to the target device; Determine the direct trust value of the target device; The trust value of the target device is determined based on the direct trust value of the target device, the indirect trust value, and the weight value corresponding to the direct trust value.

14. A trust assessment device, characterized in that, The trust assessment device includes a receiving unit, a processing unit, and a sending unit. The receiving unit is configured to receive a trust assessment request sent by the first device, the trust assessment request including the identifiers of multiple candidate devices to be assessed. The processing unit is configured to obtain the initial trust value of each candidate device among the plurality of candidate devices, and classify the plurality of candidate devices into at least two trust types based on the initial trust values ​​of the plurality of candidate devices; For each of the plurality of candidate devices, the following operations are performed to obtain the trust value of the plurality of candidate devices: based on the candidate devices corresponding to each trust type, the weight value corresponding to the direct trust value of the target device is determined; and based on the candidate devices corresponding to each trust type, a plurality of recommended devices corresponding to the target device and the weight value corresponding to each recommended device are determined; The trust value of the target device is determined based on the weight value corresponding to the direct trust value of the target device, the multiple recommended devices corresponding to the target device, and the weight value corresponding to each recommended device; wherein, the target device is any one of the multiple candidate devices; the direct trust value of the target device is used to indicate the degree of trust of the target device relative to the first device; and a trustworthy device is determined among the multiple candidate devices based on the trust values ​​of the multiple candidate devices. The sending unit is used to send the identifier of the trusted device to the first device.

15. The apparatus as claimed in claim 14, characterized in that, The processing unit is specifically used for: Based on the first reinforcement learning model, a multi-round weight calculation process is performed on the target device to obtain multiple candidate weight values; based on the multiple candidate weight values, the weight value corresponding to the direct trust value of the target device is determined; The weight calculation process in any round includes: The interaction device selection strategy for this round of weight calculation is determined to be either a first strategy or a second strategy. At least one interaction device is determined according to the interaction device selection strategy for this round of weight calculation. The first strategy is to determine a target trust type from the at least two trust types with a preset probability, and select a candidate device from the candidate devices corresponding to the target trust type as the interaction device. The second strategy is to select the node with the highest trust value among the neighboring nodes of the target device as the interaction device. The trust evaluation node stores the trust values ​​of the neighboring nodes of the target device. The system acquires interaction data between the at least one interactive device and the target device; based on the interaction data, it determines a first reward value and a first state value for the current round of weight calculation based on the first reinforcement learning model, and stores the first reward value for the current round of weight calculation; based on the first reward value and the first state value for the current round of weight calculation, it calculates the candidate weight value corresponding to the current round of weight calculation; wherein, the first reward value is used to indicate the performance of the first reinforcement learning model in the current round of weight calculation; the first state value is used to indicate the communication status between the target device and the at least one interactive device.

16. The apparatus as claimed in claim 15, characterized in that, The processing unit is specifically used for: The probability value of the interaction device selection strategy being the first strategy is determined according to a preset decision function, and the interaction device selection strategy is determined according to the random function corresponding to the probability value.

17. The apparatus as claimed in claim 15 or 16, characterized in that, The processing unit is specifically used for: Among the multiple candidate weight values, the candidate weight value calculated in the round of weight calculation with the highest first reward value is selected; the selected candidate weight value is used as the weight value corresponding to the direct trust value of the target device.

18. The apparatus according to any one of claims 15-17, characterized in that, The processing unit is also used for: After performing M rounds of weight calculation for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of weight calculation in the last N rounds of weight calculation for the target device is less than a preset threshold, the multi-round weight calculation is stopped; where M and N are positive integers, and M is greater than N.

19. The apparatus according to any one of claims 15-18, characterized in that, The interaction data includes the number of successful communication interactions and the number of communication failures between the target device and the at least one interactive device.

20. The apparatus according to any one of claims 14-19, characterized in that, The processing unit is specifically used for: Based on the second reinforcement learning model, a multi-round device recommendation process is performed on the target device to obtain multiple sets of candidate results. Each set of candidate results includes multiple candidate recommended devices and the weight value corresponding to each candidate recommended device. From the multiple sets of candidate results, a target result is determined, which includes: multiple recommended devices corresponding to the target device and a weight value corresponding to each recommended device; The process of selecting recommended equipment in any round includes: The recommended device selection strategy for this round of recommended device selection process is determined to be either the third strategy or the fourth strategy. Based on the recommended device selection strategy, multiple first candidate recommended devices are determined. The third strategy is to select candidate devices that have communicated with the target device from candidate devices of each trust type at a preset ratio as recommended devices. The fourth strategy is to select a preset number of nodes that have communicated with the target device from the target device's neighbor nodes as recommended devices. Based on the plurality of first candidate recommended devices, a second state value is determined for the current round of recommended device selection process based on the second reinforcement learning model; the second state value is used to indicate the communication status between the target device and the plurality of first candidate recommended devices; a multi-round weight calculation process is performed based on the second reinforcement learning model to determine the weight value corresponding to each first candidate recommended device; a second reward value is determined for the current round of recommended device selection process based on the plurality of first candidate recommended devices and the weight value corresponding to each first candidate recommended device, and the second reward value is stored; the second reward value is used to indicate the performance of the second reinforcement learning model in the current round of recommended device selection process.

21. The apparatus as claimed in claim 20, characterized in that, The processing unit is specifically used for: The probability value of the recommended device selection strategy being the third strategy is determined according to a preset decision function, and the recommended device selection strategy is determined according to the random function corresponding to the probability value.

22. The apparatus as claimed in claim 20 or 21, characterized in that, The processing unit is specifically used for: Among the multiple sets of alternative results, the alternative result determined in the round of recommended equipment selection process with the highest second reward value is selected; the selected alternative result is taken as the target result.

23. The apparatus according to any one of claims 20-22, characterized in that, The processing unit is also used for: After Q rounds of recommended device selection for the target device, and / or, if the difference between the reward values ​​of any two adjacent rounds of recommended device selection in the final P rounds of recommended device selection for the target device is less than a preset threshold, the multi-round recommended device selection is stopped; where Q and P are positive integers, and Q is greater than P.

24. The apparatus according to any one of claims 14-23, characterized in that, The processing unit is specifically used for: The receiving unit receives trust-related data of the plurality of candidate devices sent by the first device, and determines an initial trust value for each candidate device based on the trust-related data of each candidate device; wherein, the trust-related data includes at least one of the following: device manufacturer, device usage scope, device importance, and device deployment location; or The receiving unit receives the initial trust values ​​of the plurality of candidate devices sent by the first device; the initial trust values ​​of the plurality of candidate devices are obtained by the first device from a blockchain or interplanetary file storage system.

25. The apparatus according to any one of claims 15-24, characterized in that, The processing unit is also used for: The receiving unit receives the interaction data between the first device and the trusted device sent by the first device, and updates at least one of the following based on the interaction data between the first device and the trusted device: the model parameters of the first reinforcement learning model, the model parameters of the second reinforcement learning model, and the trust value of the trusted device.

26. The apparatus according to any one of claims 14-25, characterized in that, The processing unit is specifically used for: Based on the multiple recommended devices corresponding to the target device and the weight value corresponding to each recommended device, the indirect trust value of the target device is determined, wherein the indirect trust value is used to indicate the degree of trust of the target device relative to the multiple recommended devices corresponding to the target device; Determine the direct trust value of the target device; The trust value of the target device is determined based on the direct trust value of the target device, the indirect trust value, and the weight value corresponding to the direct trust value.

27. A trust assessment device, characterized in that, It includes a processor and a memory; the memory stores computer program instructions, and when the trust assessment device is running, the processor executes the computer program instructions stored in the memory to implement the operational steps of the method described in any one of claims 1 to 13.

28. A computer-readable storage medium, characterized in that, Includes computer instructions that, when executed by a processor, cause the trust evaluation device to perform the method as described in any one of claims 1 to 13.

29. A computer program product, characterized in that, When the computer program product is run on the processor, it causes the trust evaluation device to perform the method as described in any one of claims 1 to 13.

Images