Individual mutual assistance utility evaluation method and system for Internet of Things networks

Abstract

A method and system for evaluating the utility of individual mutual assistance in an Internet of Things network, wherein the method comprises the following steps: in the Internet of Things network, a first networking unit and a second networking unit are included, the second networking unit obtains the third reputation evaluation of the first networking unit when providing services to the first networking unit, and the second networking unit provides differentiated hierarchical services according to the distribution of the third reputation evaluation value of the first networking unit relative to the third reputation evaluations of all networking units in the Internet of Things network. The above technical solution provides differentiated services by comparing the level of the third reputation of the second networking unit with the first networking unit in the entire Internet of Things network when establishing interaction, which can save the computing resources and network resources of the second networking unit, thereby comprehensively improving the cooperation efficiency of each networking unit in the Internet of Things network and improving the information interaction level of the Internet of Things network.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things communication technology, and in particular to a method capable of quantifying the cooperation status of units in the Internet of Things. Background Art

[0002] The Internet of Things technology is gradually developing towards the direction of multi-terminal interconnection and intelligence. For example, the Internet of Things can provide a development foundation for transportation management, vehicle-road information services, and intelligent vehicle control. With the development of the Internet of Things technology, the number of intelligent control systems, intelligent network systems, and audio-visual entertainment systems that may be installed on each terminal is increasing, and the security issues exposed by the Internet of Things are gradually becoming more prominent. For example, various sensors installed in intelligent networked vehicles are responsible for data collection and environmental perception of roads and driving conditions. The intelligent transportation system relies on driving information, vehicle status information, and road condition information collected by intelligent networked vehicles.

[0003] The intelligent entities participating in the Internet of Things have similar characteristics, are connected to each other through the network, and have similar values, and belong to the network group. The network group behavior belongs to cluster behavior, which has social, distributed, robust and self-organizing characteristics. Individuals between nodes in the network frequently interact and influence each other. Therefore, traditional statistical analysis research methods cannot solve the complexity and dynamics of network group behavior. Therefore, it is necessary to manage each networking unit in the Internet of Things network intelligently. Summary of the invention

[0004] Therefore, it is necessary to provide a new method for evaluating the interactive utility between individuals in an IoT network, so as to achieve the technical effect of quantifying the effect of mutual assistance among individuals in an IoT network.

[0005] To achieve the above purpose, the inventor provides a method for evaluating the mutual assistance utility of individuals in an Internet of Things network, comprising the following steps:

[0006] In an Internet of Things network, a first networking unit and a second networking unit are included. When providing services to the first networking unit, the second networking unit obtains a third reputation evaluation of the first networking unit. The third reputation evaluation includes a weighted average of a second reputation score and a reputation age score of the first networking unit. The reputation age score is positively correlated to the number of consecutive update cycles in which the second reputation score of the first networking unit exceeds a second reputation threshold.

[0007] The second reputation score is: a normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the last update cycle, wherein the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's requests to interact with other networking units;

[0008] The second networking unit provides differentiated hierarchical services according to the distribution of the third reputation evaluation value of the first networking unit relative to the third reputation evaluations of all networking units in the Internet of Things network.

[0009] In some embodiments of the present application, the differentiated service includes information transmission rate difference, response rate difference, response rate difference or response priority difference.

[0010] In some embodiments of the present application, the third reputation evaluation value of the first networking unit is divided into groups relative to the distribution of the second reputation scores of all networking units in the Internet of Things network, as shown below:

[0011]

[0012] Among them, the trust level A={a1,a2,…,a N-1 ,a N}, is the ordered segmentation class, the credit cutoff points of trust classification are b1, b2,…, b N-1 ,b N The second networking units provide the first networking unit with a percentile of second reputation scores of all networking units, and the second networking units provide the first networking unit with a response priority, a response rate, a response rate, and a transmission rate that decreases in sequence.

[0013] In some embodiments of the present application, the third reputation evaluation

[0014]

[0015] where ω i For different evaluation dimensions F i (u i ) weights, different evaluation dimensions F i (u i )include:

[0016] Historical credit rating

[0017]

[0018] m is the current cycle and the second historical credit score Υ(u m ),

[0019]

[0020] k is the decay constant;

[0021] Credit Age Score

[0022]

[0023] Among them, Y(u) is the credit age, Y max is the maximum age limit, Υ(u) is the current global credit rating, and g is the adjustment constant.

[0024] In some embodiments of the present application, when the update period T arrives, the reputation age is updated according to the comparison between the third reputation evaluation of the current individual and its historical value.

[0025] Specifically:

[0026] When the change in the third reputation evaluation of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise:

[0027] Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows:

[0028]

[0029] n is the number of cycles.

[0030] An individual mutual aid utility evaluation system of an Internet of Things network, the system comprising a first networking unit and a second networking unit in the Internet of Things network,

[0031] The second networking unit, when providing services for the first networking unit, obtains a third reputation evaluation of the first networking unit, the third reputation evaluation comprising a weighted average of the second reputation score and the reputation age score of the first networking unit, the reputation age score being positively correlated to the number of consecutive update cycles in which the second reputation score of the first networking unit exceeds the second reputation threshold, and the second networking unit is further configured to provide differentiated hierarchical services to the first networking unit according to the distribution of the third reputation evaluation value of the first networking unit relative to the third reputation evaluations of all networking units in the Internet of Things network;

[0032] The second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the previous update cycle, and the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's requests for interaction with other networking units.

[0033] In some embodiments of the present application, the differentiated service includes information transmission rate difference, response rate difference, response rate difference or response priority difference.

[0034] In some embodiments of the present application, the second networking unit is used to divide the distribution of the third reputation evaluation value of the first networking unit relative to the second reputation scores of all networking units in the Internet of Things network into groups, as shown below:

[0035]

[0036] Among them, the trust level A={a1,a2,…,a N-1 ,a N}, is the ordered segmentation class, the credit cutoff points of trust classification are b1, b2,…, b N-1 ,b N The first networking unit is provided with a response priority, a response rate, a response speed, and a transmission speed in descending order for the percentile points of the second reputation score of all networking units.

[0037] In some embodiments of the present application, the third reputation evaluation

[0038]

[0039] where ω i For different evaluation dimensions F i (u i ) weights, different evaluation dimensions F i (u i )include:

[0040] Historical credit rating

[0041]

[0042] m is the current cycle and the second historical credit score Υ(u m ),

[0043]

[0044] k is the decay constant;

[0045] Credit Age Score

[0046]

[0047] Among them, Y(u) is the credit age, Y max is the maximum age limit, Υ(u) is the current global credit rating, and g is the adjustment constant.

[0048] In some embodiments of the present application, a central node is included, and the central node is used to update the reputation age according to the comparison between the third reputation evaluation of the current individual and its historical value when the update period T arrives, and broadcast it in the Internet of Things network.

[0049] Specifically:

[0050] When the change in the third reputation evaluation of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise:

[0051] Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows:

[0052]

[0053] n is the number of cycles.

[0054] Different from the existing technology, the above technical solution provides differentiated services by comparing the level of the third credibility of the second networking unit in the entire Internet of Things network when the second networking unit establishes interaction with the first networking unit, which can save the computing resources and network resources of the second networking unit, thereby comprehensively improving the cooperation efficiency of each networking unit in the Internet of Things network and improving the level of information interaction in the Internet of Things network. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of a reputation evaluation method based on the Internet of Things according to a specific embodiment of the present invention;

[0056] Figure 2 This is a unit diagram of a reputation evaluation system based on the Internet of Things according to another embodiment of the present invention;

[0057] Figure 3 This is a flow chart of a global reputation evaluation method based on the Internet of Things according to a specific implementation mode of the present invention;

[0058] Figure 4 This is a unit diagram of a global reputation evaluation system based on the Internet of Things according to a specific implementation mode of the present invention;

[0059] Figure 5 It is a flow chart of a mutual trust evaluation method for an Internet of Things network according to a specific embodiment of the present invention;

[0060] Figure 6 A unit diagram of a mutual trust evaluation system of an Internet of Things network according to a specific embodiment of the present invention;

[0061] Figure 7 It is a flow chart of the method for evaluating the individual mutual assistance utility of the Internet of Things network according to the specific implementation mode of the present invention;

[0062] Figure 8 It is a unit diagram of the individual mutual assistance utility evaluation system of the Internet of Things network according to the specific implementation mode of the present invention;

[0063] Fig. 9 A schematic diagram of a two-way interactive mode scenario according to a specific embodiment of the present invention;

[0064] Fig.10 This is a diagram for establishing a dynamic social trust network according to a specific implementation manner of the present invention. DETAILED DESCRIPTION

[0065] In order to explain the technical content, structural features, achieved objectives and effects of the technical solution in detail, the following is a detailed description in conjunction with specific embodiments and accompanying drawings.

[0066] The Internet of Things refers to a network that connects any object to the Internet through information sensing devices and in accordance with agreed protocols to exchange information and communicate in order to achieve intelligent identification, positioning, tracking, monitoring and management. In layman's terms, the Internet of Things is "the Internet of Things connected to everything". It can achieve information exchange and cooperative sharing between objects.

[0067] However, it is found in real life that after the development of the Internet of Things network, with the improvement of individual intelligence, for individuals in the Internet of Things, it is the most resource-saving to not interact but only receive information. The spread of this selfishness may eventually lead to the functional paralysis of the Internet of Things network. The inventor of this application has noticed the above problems and hopes to propose a concept of altruism and construct an evaluation and scoring system for individuals in the Internet of Things network.

[0068] See also Figure 1 , is a flow chart of a reputation evaluation method based on the Internet of Things. The scheme includes the following steps: in the Internet of Things networking network, including the second networking unit and the first networking unit, please refer to Figure 1 ,

[0069] S11 The second networking unit performs a first reputation score on the first networking unit based on the interaction result between the second networking unit and the first networking unit. If S12 The first networking unit chooses to cooperate with the interaction request of the second networking unit, the first reputation is positively correlated with the interaction time and interaction quality of the first networking unit's cooperative behavior. If S13 The first networking unit chooses to reject the interaction request of the second networking unit, the first reputation is negatively correlated with the punitive reputation adjustment value.

[0070] In some embodiments, the first networking unit and the second networking unit are individuals participating in the Internet of Things network. The networking units may include smart home appliances, such as smart refrigerators, smart air conditioners, smart TVs, smart washing machines, etc., or smart wearable devices carried by the human body, such as mobile phones, smart watches, smart glasses, etc. In the case where the Internet of Things is the Internet of Vehicles, the first networking unit and the second networking unit may also be vehicle-mounted units, roadside units or cloud servers.

[0071] Take the second networking unit and the first networking unit as an example, both of which are devices of the same level, such as vehicle-mounted units. In a first reputation evaluation cycle, the second networking unit has sent an interaction request to the first networking unit. The interaction request here can be a request to share information, vehicle formation driving, road emergency broadcast, efficient speed guidance, priority vehicle traffic guarantee management, road traffic information sharing, multimedia and entertainment information push, etc. The first networking unit can choose to cooperate or refuse, which will affect the first reputation evaluation of the first networking unit by the second networking unit in the evaluation cycle. The first reputation is positively correlated with the interaction time and interaction quality of the first networking unit's cooperative behavior, that is, the higher the interaction time and interaction quality of the first networking unit's cooperative behavior, the higher the first reputation. If the first networking unit chooses to refuse the interaction request of the second networking unit, the first reputation is negatively correlated with the punitive reputation adjustment value, and a penalty mechanism is required. This punitive reputation adjustment value can be determined as needed. The result of refusing to cooperate will lead to a reduction in the first reputation.

[0072] Through the above scheme, the sender of the interaction request evaluates the interaction with the receiver in the Internet of Things network, and obtains the first reputation, which can achieve the technical effect of quantifying the interactive cooperation quality of the first networking unit in the Internet of Things network. The first reputation represents the response of the first networking unit to the external interaction request.

[0073] In some embodiments of the present application, a utility function η(u) is included. If the first networking unit chooses to cooperate with the interaction request of the second networking unit, the first reputation is positively correlated with the interaction time and interaction quality of the first networking unit's cooperative behavior, and the positive correlation coefficient is If the first networking unit chooses to reject the interaction request of the second networking unit, the first reputation is negatively correlated with the punitive reputation adjustment value, and the negative correlation coefficient is η(u). By designing the positive correlation coefficient and the negative correlation coefficient, the growth and penalty rules of the first reputation can be further refined, so that the growth rate and the reduction rate can be determined, and the technical effect of quantifying the interactive cooperation quality of the first networking unit in the Internet of Things network can be better achieved.

[0074] In some embodiments of the present application, if the first networking unit chooses to cooperate with the interaction request of the second networking unit, the first reputation:

[0075]

[0076] Where Υ(u) is the weighted result of the first reputation evaluation of the first networking unit by all networking units of the entire Internet of Things network in the previous evaluation cycle, t D is the standard interaction time of the standard interaction behavior and q D is the standard interaction quality value of the standard interaction behavior, Υ D1 is the cooperation adjustment value of cooperative behavior, t d ,q d The interaction time and interaction quality value of the interaction behavior between the first networking unit and the second networking unit. By introducing the evaluation results of the previous evaluation cycle, the historical reputation performance of the first networking unit can be comprehensively considered to prevent abnormal individual interaction behaviors from causing sudden changes in reputation. By introducing standard interaction time and standard interaction quality, the interaction time and interaction quality of the first networking unit have a comparison benchmark. The above steps better achieve the technical effect of quantifying the interactive cooperation quality of the first networking unit in the Internet of Things network.

[0077] In order to better determine the benchmark of the standard interaction time and the standard interaction quality value, in some embodiments of the present application, the standard interaction time and the standard interaction quality value are determined according to the interaction quality of the entire Internet of Things network in the previous evaluation cycle. This determination can be completed by the centralized node in the Internet of Things network. The centralized node undertakes more calculation tasks and can also strengthen the authority of the centralized node and reduce the resource burden of the networking unit.

[0078] In some embodiments of the present application, if the first networking unit chooses to reject the interaction request of the second networking unit, the first reputation:

[0079] R D (u) = Υ(u) - η(u)Υ D2 ,

[0080] Wherein Υ(u) is the weighted result of the first reputation evaluation of the first networking unit by all networking units of the entire Internet of Things networking network in the previous evaluation cycle, Υ D2 By introducing the evaluation results of the previous cycle, the fluctuation of the first reputation evaluation of the first networking unit is reduced and more stable. The above steps better achieve the technical effect of quantifying the interactive cooperation quality of the first networking unit in the Internet of Things network.

[0081] In some embodiments of the present application, the utility function η(u) is an exponential function, a logarithmic function, or a binomial function, and the range is [0,1]. The utility function can be set according to actual conditions and can be used as a proportional coefficient of the first credit addition or subtraction, and this solution does not limit this.

[0082] In some embodiments of the present application, the utility function:

[0083]

[0084] Where δ is the adjustment constant of the utility function, Υ)u( is the weighted result of the first reputation evaluation of the first networking unit by all networking units of the entire Internet of Things network in the previous evaluation cycle, and γ max is the maximum score of all networked units in the previous evaluation cycle. By setting the specific form of the utility function, the change of the utility function can be based on the ratio of the weighted evaluation result of the previous cycle to the global maximum, so that the reputation evaluation conforms to the reality of IoT applications. When a networked unit with high reputation performs well in the interaction, the change of its reputation value is smaller; on the other hand, when a networked unit with high reputation does not maintain cooperative behavior, that is, when the interaction strategy mutates, its reputation evaluation is adjusted significantly.

[0085] In some embodiments of the present application, an evaluation period T is included, and the evaluation period is determined according to the data transmission delay in the Internet of Things network and the variance of the first reputation value of each unit. When the data transmission delay in the Internet of Things network is large, the evaluation period can be relatively extended, so that the update frequency in the Internet of Things network is not too fast, resulting in the judgment results of many other networked units not being taken into account. When the first reputation of each networked unit tends to be stable, the evaluation period can also be extended to save the computing resource consumption of the evaluation action in the entire network. The specific time of the evaluation period can also be published through a centralized node or an authoritative node or the cloud.

[0086] exist Figure 2 In the embodiment shown, a reputation evaluation system based on the Internet of Things is also disclosed, including a second networking unit 22, a first networking unit 21,

[0087] The second networking unit is used to perform a first reputation score on the first networking unit according to the interaction result between the second networking unit and the first networking unit. If the first networking unit chooses to cooperate with the interaction request of the second networking unit, the first reputation is positively correlated with the interaction time and interaction quality of the first networking unit's cooperative behavior. If the first networking unit chooses to reject the interaction request of the second networking unit, the first reputation is negatively correlated with the punitive reputation adjustment value. The above technical solution obtains the first reputation by evaluating the interaction of the sender of the interaction request with the recipient in the Internet of Things network, which can achieve the technical effect of quantifying the interactive cooperation quality of the first networking unit in the Internet of Things network.

[0088] In some embodiments of the present application, if the first networking unit chooses to cooperate with the interaction request of the second networking unit, the first reputation:

[0089]

[0090] Where Υ(u) is the weighted result of the first reputation evaluation of the first networking unit by all networking units of the entire Internet of Things network in the previous evaluation cycle, t D is the standard interaction time of the standard interaction behavior and q D is the standard interaction quality value of the standard interaction behavior, Υ D1 is the cooperation adjustment value of cooperative behavior, t d ,q d is the interaction time and interaction quality value of the interaction behavior between the first networking unit and the second networking unit;

[0091] If the first networking unit chooses to reject the interaction request of the second networking unit, the first reputation:

[0092] R D (u) = Υ(u) - η(u)Υ D2 ,

[0093] Wherein Υ(u) is the weighted result of the first reputation evaluation of the first networking unit by all networking units of the entire Internet of Things networking network in the previous evaluation cycle, Υ D2 The reputation adjustment value for the interaction behavior;

[0094] The utility function:

[0095]

[0096] Where δ is the adjustment constant of the utility function, γ max It is the maximum score of all networked units in the previous evaluation cycle.

[0097] The above system solves the problem of first credit evaluation in a better and more complete way.

[0098] Others such as Figure 3 In the illustrated embodiment, a global reputation evaluation method based on the Internet of Things is also introduced. In the Internet of Things networking network, a first networking unit, a plurality of second networking units and a plurality of third networking units are included. In some embodiments of this solution, the third networking unit can be a smart home appliance, such as a smart refrigerator, a smart air conditioner, a smart TV, a smart washing machine, etc., or a smart wearable device carried by the human body, such as a mobile phone, a smart watch, and smart glasses. In the case where the Internet of Things is the Internet of Vehicles, the third networking unit can also be a vehicle-mounted unit, a roadside unit or a cloud server. In some preferred embodiments, the functions of the second networking unit and the third networking unit in the Internet of Vehicles network are distinguished, or the permissions and authority levels of the second networking unit and the third networking unit are different. For example, the third networking unit can be a centralized node, and the centralized node can be a roadside terminal that undertakes the overall planning and networking functions, or a cloud server. The second networking unit is a vehicle-mounted unit with the same functions and permissions as the first networking unit.

[0099] The global reputation evaluation method comprises the following steps: in an update cycle, S31 all second networking units and third networking units that request to interact with the first networking unit independently perform a first reputation score on the first networking unit, and the first reputation score performed by each is positively correlated with the interaction time and interaction quality of the first networking unit's request to interact with other networking units;

[0100] S33 calculates a second reputation score, where the second reputation score is: a normalized weighted average score of the first reputation score of the first networking unit by the third networking unit and the first reputation score of the first networking unit by the second networking unit in the update cycle;

[0101] S35 broadcasts the second reputation score in the Internet of Things network, so that other networked units can perform the first reputation score on the first networked unit in the next update cycle.

[0102] The second reputation is a comprehensive weighting of the first reputation score of the first networking unit by all networking units, which can also be called a global reputation score or global reputation. The second reputation can evaluate the interaction between the first networking unit and the entire Internet of Things network. Normalized weighting means that the sum of all weights of the second reputation score is 1. There can be multiple second networking units, and there can also be multiple third networking units. All second networking units and third networking units that request to interact with the first networking unit independently perform the first reputation score on the first networking unit, which can better achieve the global reputation evaluation in the weighted calculation network and solve the technical effect of quantifying the quality of interaction between the first networking unit in the Internet of Things network and other global networking units.

[0103] In some embodiments of the present application, the third networking unit is a centralized node, which can be a roadside terminal that undertakes the coordination and networking functions, or a cloud server. The authority of the third networking unit in the Internet of Things should be set higher, and the third networking unit generally undertakes more computing functions. The third networking unit is not easy to be tampered with, so the weight of the third networking unit can be set greater than the second networking unit. By making the weight of the first reputation score evaluated by the third unit greater than that of the second networking unit, the evaluation result of the second reputation can be made more reliable and more authoritative.

[0104] In some embodiments of the present application, the evaluation period is determined according to the data transmission delay in the Internet of Things networking network, the variance of the first reputation value of each unit, or the global reputation distribution. When the data transmission delay between the first networking unit and the second networking unit or the third networking unit in the Internet of Things networking network is large, the evaluation period can be relatively extended, so that the update frequency in the Internet of Things networking network is not too fast, resulting in the judgment results of many other networking units not being taken into account. When the first reputation of each networking unit tends to be stable, the evaluation period can also be extended to save the computing resource consumption of the evaluation action in the entire network. The specific time of the evaluation period can also be published through a centralized node or an authoritative node or the cloud.

[0105] In some embodiments of the present application, the second reputation score:

[0106]

[0107] Among them, β is the global reputation evaluation weight, R D (u) is the first reputation evaluation of the third networking unit on the first networking unit, J is the total number of second networking units that effectively interact with the first networking unit, R Dj The first reputation evaluation made by the second networking unit on the first networking unit according to the interaction between the second networking unit and the first networking unit within the update period T. In this embodiment, assuming that there is only one third networking unit in the Internet of Things network, an evaluation R can be obtained. D (u), on the other hand, it can also be assumed that there are multiple third networking units in the Internet of Things network, but only one unified evaluation R is made D (u), or there are multiple third networking units in the Internet of Things network, and each of them makes a weighted average of the first reputation evaluation to obtain R D (u). Through the above scheme, the global reputation of the first networking unit can be obtained, and an objective quantitative result is obtained for the cooperative interaction of the first networking unit with all other units in the Internet of Things network.

[0108] In some embodiments of the present application, the credibility of the first reputation evaluation made by the jth second networking unit is also included. The second reputation score

[0109]

[0110] Among them, β is the global reputation evaluation weight, R D (u) is the first reputation evaluation of the third networking unit on the first networking unit, J is the total number of second networking units that effectively interact with the first networking unit, R Dj The first reputation evaluation made by the second networking unit on the first networking unit within the update period T according to the interaction between the second networking unit and the first networking unit. The third networking unit is a centralized node or a cloud server. The credibility of the reputation evaluation of the second networking unit.

[0111] The second networking unit is generally a decentralized networking unit, so it is necessary to distinguish the credibility of the evaluation results of different second networking units to prevent malicious units from mixing into the Internet of Things network and causing damage.

[0112] In a further embodiment, in order to distinguish malicious units, the first evaluation result typed out by the second networking unit may be related to its own cooperative interaction level. Determined according to the second reputation score of the jth second networking unit in the previous cycle. In order to normalize the evaluation results, you can set

[0113] In some other further embodiments, in order to punish or better avoid the impact of the malicious second networking unit on the IoT network and increase the penalty for the malicious unit, the step of setting a threshold may also be performed: All second reputation scores are lower than The first credit rating R made by the networking unit Dj Through the above scheme, the present application can better achieve the result of evaluating the global reputation of the first unit, and the global reputation can also be uploaded to the cloud or uploaded to the roadside, and broadcast to the entire network, so as to meet the needs of the reputation calculation of the next cycle.

[0114] In such Figure 4 In the embodiment shown, a global reputation evaluation system based on the Internet of Things is also introduced.

[0115] It includes a first networking unit 41, a plurality of second networking units 42 and a plurality of third networking units 43, wherein the first networking unit forms an Internet of Things network with the second networking unit and the third networking unit.

[0116] In an update cycle, all second networking units and third networking units that request to interact with the first networking unit independently perform a first reputation score on the first networking unit, and the first reputation score performed by each is positively correlated with the interaction time and interaction quality of the first networking unit's request to interact with other networking units;

[0117] The third networking unit is used to calculate a second reputation score, where the second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by the third networking unit and the first reputation score of the first networking unit by the second networking unit during the update cycle; the third networking unit is also used to broadcast the second reputation score in the Internet of Things networking network, so that other networking units can perform the first reputation score on the first networking unit in the next update cycle.

[0118] Through the above system design, all second networking units and third networking units that request to interact with the first networking unit independently perform a first reputation score on the first networking unit, which can better achieve the global reputation evaluation in the weighted computing network and solve the technical effect of quantifying the quality of interaction between the first networking unit in the Internet of Things network and other global networking units.

[0119] In some embodiments of the present application, the second reputation score:

[0120]

[0121] Among them, β is the global reputation evaluation weight, R D (u) is the first reputation evaluation of the third networking unit on the first networking unit, J is the total number of second networking units that effectively interact with the first networking unit, R Dj The first reputation evaluation made by the second networking unit on the first networking unit within the update period T according to the interaction between the second networking unit and the first networking unit. The third networking unit is a centralized node or a cloud server. The credibility of the reputation evaluation of the second networking unit.

[0122] Different from the existing technology, the above technical solution obtains the first reputation by evaluating the interaction between the sender of the interaction request and the recipient in the Internet of Things network, and then obtains the global reputation evaluation in the network by weighting the first reputation scores of more than two networking units. This can solve the technical effect of quantifying the interaction quality of the first networking unit in the Internet of Things network with other global networking units.

[0123] exist Figure 5 In the embodiment shown, this solution also introduces a mutual trust evaluation method for an Internet of Things network, comprising the following steps:

[0124] S51 Before establishing the interaction request, the first networking unit and the second networking unit obtain the third reputation evaluation of the first networking unit, the third reputation evaluation includes the weighted average of the second reputation evaluation and reputation age score of the first networking unit, the second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request interaction with the first networking unit in the last update cycle, the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's request for interaction with other networking units; the reputation age score is positively correlated with the number of consecutive update cycles in which the second reputation score of the first networking unit exceeds the second reputation threshold. The foregoing content has explained that the second reputation can be announced to the Internet of Things network by the third networking unit or the cloud server, and the second reputation can reflect the situation of the first networking unit in the Internet of Things network with other networking units. The reputation age is accumulated by continuously providing a time higher than the second reputation threshold in the Internet of Things. The second reputation threshold can be automatically set according to the actual situation. The above technical solution can have corresponding expectations for the quality of cooperation between other networking units sending requests to the first networking unit by calling the second reputation information of the first networking unit, and calculate the third reputation evaluation, which can also be called expected reputation. This achieves the effect of quantifying the expected value of service interaction between each networking unit in the Internet of Things network.

[0125] In order to better punish the networking units with unsatisfactory interaction results, step S53 is further executed: when the third reputation evaluation of the first networking unit is less than the third threshold, the first networking unit and the second networking unit do not establish an interaction request. At this time, although the first networking unit may still belong to the same IoT network as the second networking unit through other means, the interaction requests between the two will not be able to interact due to the failure to meet the expected reputation. This is the penalty mechanism for the first networking unit. Through the above-mentioned penalty mechanism, the consumption of computing resources in the IoT network can be reduced. At the same time, the networking units with low reputation evaluation are prompted to change their own interaction strategies, and the generation and dissemination of cooperative behaviors within the IoT network are enhanced.

[0126] In some embodiments of the present application, a time decay function ψ(m) is also included.

[0127] The third reputation evaluation is further adjusted by the time decay function to obtain a historical reputation evaluation

[0128]

[0129] m is the current cycle and the second historical credit score Υ(u m), the number of update cycles at intervals of m increases as ψ(m) decreases. By designing a decay function, the utility weight of the second reputation score of the latest cycle can be made larger, so that some punished first networked units can "reform" faster and participate in mutual assistance services.

[0130] The value of the attenuation function can generally be 1 when m=0. The attenuation function only needs to be designed so that ψ(m) decreases as m increases. This application does not limit this. In some embodiments of this application,

[0131]

[0132] k is the decay constant. Designed as an exponential function, it can reduce the impact of historical evaluation on current credit evaluation.

[0133] In some embodiments of the present application, the reputation age score

[0134]

[0135] Among them, Y max is the maximum age limit, Υ(u) is the current global reputation evaluation, and g is the adjustment constant. The reputation age score is designed to be accumulated by continuously providing a time higher than the second reputation threshold in the Internet of Things. The above setting can enable individuals who continuously and stably provide interactive services to obtain better and more stable reputation age evaluations, thereby obtaining better and more stable expected reputations.

[0136] In some embodiments of the present application, the third reputation evaluation also includes a weighted average of the IoT network attribute correlation scores of the first networking unit and the second networking unit. IoT network attributes include whether they belong to the same IoT subnetwork, whether they belong to the same network usage attributes, such as whether the network is mainly used for entertainment purposes, road condition information sharing purposes, autonomous driving purposes, etc. The more identical attributes they have, the more likely it is that the two belong to the same social network. The higher the network attribute correlation score is set. By setting the IoT network attribute correlation, the expected reputation of two networking units in the same social network before requesting interaction can be improved, thereby better establishing interaction.

[0137] In some embodiments of the present application, the IoT network attribute correlation score

[0138]

[0139] Among them, C1 The value of the association degree of the IoT network attribute is adjusted. The above design can improve the expected reputation of two networked units in the same social network before requesting interaction, so as to better establish interaction.

[0140] In some embodiments of the present application, the third reputation evaluation further includes a weighted average of the social attribute tag association scores of the first networking unit and the second networking unit, wherein the social attribute tag association score

[0141]

[0142] Among them, C2 is the adjustment value of the social attribute label association. SP is the social attribute label association cluster. The social attribute label is a label mark made by the reputation authority organization after the interaction behavior of all individuals in the group with the individual, and the service behavior with high contribution level of the individual. Different service categories correspond to different label descriptions. For example: traffic safety is SVR1, traffic management is SVR2, traffic information is SVR3, traffic efficiency is SVR4, commercial entertainment is SVR5, collaborative service is SVR6, and so on. For example: when the cumulative number of traffic management and collaborative service provided by the individual in the entire interaction process within the evaluation period exceeds the assignable threshold, the social attribute label SP = {SVR2, SVR6) is assigned to it. When the service request type of the service requester is consistent with its social attribute label, it means that the expected return level of providing this type of service is high, that is, the service requester is likely to provide this type of service to other individuals in the Internet of Vehicles group in the future. The above design can improve the expected reputation of two networked units with the same social attribute label before requesting interaction, so as to better establish interaction.

[0143] In such Figure 6 In the embodiment shown, a mutual trust evaluation system for an Internet of Things network is also introduced.

[0144] The system includes a first networking unit 61 and a second networking unit 62 in an Internet of Things network. Before establishing an interaction request, the first networking unit and the second networking unit obtain a third reputation evaluation of the first networking unit. The third reputation evaluation includes a weighted average of the second reputation evaluation and the reputation age score of the first networking unit. The second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by all networking units in the Internet of Things network that request interaction with the first networking unit in the last update cycle. The first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's request for interaction with other networking units; the reputation age score is positively correlated with the number of consecutive update cycles in which the second reputation score of the first networking unit exceeds the second reputation threshold. When the third reputation evaluation of the first networking unit is less than the third threshold, the first networking unit and the second networking unit do not establish an interaction request. The system can have a corresponding expectation of the quality of cooperation between other networking units before sending requests to the first networking unit by calling the second reputation information of the first networking unit, and calculate the third reputation evaluation, which can also be called expected reputation. Thus, the effect of quantifying the expected value of service interaction between each networking unit in the Internet of Things network is achieved.

[0145] In some embodiments of the present application, the third reputation evaluation

[0146]

[0147] where ω i For different evaluation dimensions F i (u i ) weights, different evaluation dimensions F i (u i )include:

[0148] Historical credit rating

[0149]

[0150] m is the current cycle and the second historical credit score Υ(u m ),

[0151]

[0152] k is the decay constant;

[0153] Credit Age Score

[0154]

[0155] Among them, Y max is the maximum age limit, Υ(u) is the current global credit rating, and g is the adjustment constant;

[0156] IoT network attribute correlation score

[0157]

[0158] Among them, C1 Adjust the value of the IoT network attribute correlation;

[0159] Social attribute label relevance score

[0160]

[0161] Among them, C2 Adjust the value of social attribute tag relevance.

[0162] Different from the prior art, the above system can have corresponding expectations for the quality of cooperation between other networking units when sending requests to the first networking unit by calling the second reputation information of the first networking unit, and quantify it through comprehensive consideration of four aspects, namely historical reputation evaluation, reputation age, IoT network attributes, and social attribute labels, so as to achieve the effect of quantifying the expected value of service interaction between each networking unit in the IoT network. This saves the interaction cost and computing resources in the IoT network.

[0163] In other Figure 7 In the flowchart of the method for evaluating the individual mutual assistance utility of an Internet of Things network shown in the figure, the method is applicable to the Internet of Things network, including a first networking unit and a second networking unit, and the method includes the following steps: S71, when the second networking unit provides services for the first networking unit, obtains a third reputation evaluation of the first networking unit, the third reputation evaluation includes a weighted average of the second reputation evaluation and the reputation age score of the first networking unit, and the reputation age score is positively correlated to the number of consecutive update cycles in which the second reputation score of the first networking unit exceeds the second reputation threshold,

[0164] The second reputation score is: a normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the last update cycle, wherein the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's requests to interact with other networking units;

[0165] S73: The second networking unit provides differentiated grading services according to the distribution of the third reputation evaluation value of the first networking unit relative to the second reputation evaluations of all networking units in the Internet of Things network.

[0166] According to some of the introductions above, the second reputation can be calculated by a centralized node or cloud server in the previous cycle and broadcast to the Internet of Things network, and each unit participating in the networking has its own second reputation value. Therefore, the distribution of the second reputation evaluation of all networking units in the Internet of Things network can be obtained. Therefore, the present application can provide differentiated services by comparing the second reputation level of the second networking unit with the first networking unit in the entire Internet of Things network when establishing interaction, which can save the computing resources and network resources of the second networking unit, thereby comprehensively improving the cooperation efficiency of each networking unit in the Internet of Things network and improving the information interaction level of the Internet of Things network.

[0167] In some embodiments of the present application, the differentiated services include information transmission rate differences, response rate differences, response rate differences, or response priority differences. The transmission rate difference may refer to the high and low bandwidth capacity of information transmission, the response rate difference may refer to the delay from receiving information to processing information, the response rate difference may refer to the probability of returning valid information after receiving an interaction request, that is, even if an interaction request is received, it may be ignored by the recipient, and the response priority refers to the preset resources giving priority to responding to interaction requests of higher priority networking units. Designing these differentiated services can ensure that high-quality networking units can obtain higher-quality services, thereby improving the overall interaction level of the Internet of Things network.

[0168] The specific difference grading method can be set according to the actual situation by the technicians. In some embodiments of the present application, when the third reputation evaluation value of the first networking unit is in the top 25%, 25%-50%, 50%-75%, 75%-100% relative to the distribution percentage of the second reputation evaluation of all networking units in the Internet of Things network, the second networking unit provides the first networking unit with a successively decreasing response priority, response rate, response rate and transmission rate. The method of designing percentage ladder grading is simpler than the gradually decreasing grading method.

[0169] In order to better calculate the reputation age, some embodiments of the present application also propose the following scheme: when the update period T arrives, the reputation age is updated according to the comparison between the third reputation evaluation of the current individual and its historical value,

[0170] Specifically, when the change in the third credit rating of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise:

[0171] Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows:

[0172]

[0173] n is the number of cycles. Through the above method, the networked units that continuously provide effective interactive services can obtain better credit age evaluation, which helps to maintain the stability and effectiveness of the Internet of Things network.

[0174] Others such as Figure 8 In the illustrated embodiment, a system for evaluating the individual mutual assistance utility of an Internet of Things network is shown. The system includes a first networking unit 81 and a second networking unit 82 in the Internet of Things network.

[0175] The second networking unit, when providing services to the first networking unit, obtains a third reputation evaluation of the first networking unit, the third reputation evaluation comprising a weighted average of the second reputation evaluation and a reputation age score of the first networking unit, the reputation age score being positively correlated to the number of consecutive update cycles during which the second reputation score of the first networking unit exceeds a second reputation threshold, and the second networking unit is further configured to provide differentiated hierarchical services to the first networking unit according to the distribution of the third reputation evaluation value of the first networking unit relative to the second reputation evaluations of all networking units in the Internet of Things network;

[0176] The second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the previous update cycle, and the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's requests for interaction with other networking units.

[0177] The above system can provide differentiated services by comparing the level of the third credibility of the second networking unit in the entire Internet of Things network when the second networking unit establishes interaction with the first networking unit, which can save the computing resources and network resources of the second networking unit, thereby comprehensively improving the cooperation efficiency of each networking unit in the Internet of Things network and improving the level of information interaction in the Internet of Things network.

[0178] In some embodiments of the present application, the differentiated service includes information transmission rate difference, response rate difference, response rate difference or response priority difference.

[0179] In some embodiments of the present application, the second networking unit is used to provide the first networking unit with a response priority, response rate, response rate and transmission rate that decreases successively when the distribution percentage of the third reputation evaluation value of the first networking unit relative to the third reputation evaluations of all networking units in the Internet of Things network is in the top 25%, 25%-50%, 50%-75%, and 75%-100%.

[0180] In some embodiments of the present application, Figure 8 As shown, it includes a central node 83, which is used to update the reputation age according to the comparison between the third reputation evaluation of the current individual and its historical value when the update period T arrives, and broadcast it in the Internet of Things network.

[0181] Specifically:

[0182] When the change in the third reputation evaluation of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise:

[0183] Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows:

[0184]

[0185] n is the number of cycles.

[0186] In other comprehensive embodiments, this application takes the Internet of Things, which is the Internet of Vehicles, as an example, and the Internet of Vehicles reputation management mechanism includes three parts: participants, behavioral norms, and social paradigms. Participants in Internet of Vehicles reputation management: including reputation authority organizations and reputation participating individuals, forming a reputation hierarchical management framework. The behavioral norm refers to the principle that the reputation participating individuals decide their own behavior based on the reputation of the interacting party, and the principles of different participating individuals are differentiated. The social paradigm refers to the standard for comprehensive evaluation of the reputation of different participating individuals in the interaction of the Internet of Vehicles.

[0187] The behavior of participating individuals is to make behavioral decisions based on their own behavioral codes. Participants can be divided into three types based on their individual behavioral code strategies: cooperative individuals, reciprocal individuals, and selfish individuals. The specific behavioral codes of the three are as follows: cooperative individuals respond to the needs of all requesting individuals, adopt an unconditional cooperation strategy, and do not take different decision-making behaviors based on the reputation of the other party; reciprocal individuals adopt a conditional cooperation strategy and take different decision-making behaviors based on the reputation of the other party; selfish individuals do not consider the reputation of the other party, refuse to provide services to the other party, and only enjoy services from others.

[0188] The reputation of the Internet of Vehicles is a judgment made in the Internet of Vehicles based on the trust shown by the interactive behavior of an individual, and is a comprehensive judgment of the trust reflected by all the interactive behaviors of the individual. In the present invention, an incentive mechanism based on social indirect reciprocity is adopted, so the reputation of the Internet of Vehicles reflects the cooperation level of the participating individuals and reflects the level of participation and contribution to the activities of the Internet of Vehicles system. In the present invention, reputation is divided into three categories: global reputation, direct reputation, and expected reputation. Global reputation is a unique value at a certain moment, which is a comprehensive evaluation value based on all the interactive behaviors of the individual at that moment. The global reputation is made by the reputation authority organization. The direct reputation evaluation is the evaluation value made by other individuals who directly interact with the individual based on the performance of the individual's interactive process. Direct reputation is aggregated according to the global reputation evaluation rules to obtain global reputation. Expected reputation is the evaluation made by individuals on the expected reputation shown by the service interaction object, because even individuals with the same global reputation at present may have different expected performances after interaction due to different environments, individual attributes, etc., that is, the expected contribution level of the interactive object to the Internet of Vehicles may be different. Expected reputation reflects the expected benefit judgment of the interaction. When the expected benefit is insufficient, the responder will reject the service request and punish selfish individuals with low reputation in the Internet of Vehicles, promote selfish individuals to change their own interaction strategies, and encourage the generation and spread of cooperative behavior.

[0189] Reputation hierarchical management is mainly divided into three levels, namely, reputation authority center, reputation authority node and reputation participant. The reputation authority institution is the reputation authority center, responsible for the member registration, identity acquisition, global reputation evaluation, trust level adjustment, Internet of Vehicles service control, individual social attribute label assignment, etc. of the Internet of Vehicles participant. It is in a core position in the reputation management architecture and is trustworthy. Reputation authority nodes, such as the intelligent terminal RSU on the roadside of the Internet of Vehicles, are responsible for the collection of interactive behaviors, the collection of reputation evidence, the edge computing of reputation, the distribution of Internet of Vehicles services, etc. They are distributed authority nodes in the reputation management architecture and are trustworthy. Reputation participant individuals are intelligent entities participating in the Internet of Vehicles, mainly intelligent vehicles composed of vehicles, on-board units and systems, responsible for the generation of reputation evidence, uploading of reputation evidence, direct reputation evaluation, expected reputation evaluation and trust decision-making, etc. The reputation of a trustworthy individual is the maximum value of the reputation, and the trust is 1.

[0190] Services in the Internet of Vehicles are divided into many categories, such as traffic safety, traffic management, traffic information, traffic efficiency, commercial entertainment, collaborative services, etc. Specific services include: vehicle sensor information sharing, vehicle platoon driving, road emergency broadcast, efficient speed guidance, priority vehicle traffic security management, road traffic information sharing, multimedia and entertainment information push, etc. Service modes include: one-way service mode, request-response mode, and publish-subscribe mode. The request-response mode and the publish-subscribe mode are both two-way interaction modes. After the interactive requester initiates the interaction request, the responder provides the corresponding service to the responder according to the service decision result and the corresponding service requirements. Request-response mode service, for example: road traffic information sharing service, the interactive requester requests the surrounding vehicles to share sensor information, and the responder shares the road condition information according to the service requirements after making a service decision. In the request-response mode, the responder immediately provides information sharing within the required time period; in the subscription-publish mode, the responder pushes the road condition information to the subscription requester when the road condition information is updated and published. In the one-way service mode, the one-way service provider actively provides services to the service receiver without the need for the other party to request. For example: road emergency broadcast service, for example, when a vehicle is about to collide with an animal that has strayed into the lane, the vehicle immediately applies emergency brakes and broadcasts the emergency to surrounding vehicles through V2V (Vehicle to Vehicle) technology to avoid rear-end collisions with other vehicles and ensure road safety. Since the authoritative node is a node for data collection and aggregation, it is responsible for the provision and management of services such as traffic safety, traffic efficiency, and collaborative services in the Internet of Vehicles. Individuals can obtain comprehensive road information services, green wave speed guidance services, priority vehicle passage guarantee services, etc. from the node. At the same time, the authoritative node assumes a management role in service scenarios such as priority vehicle passage guarantee, vehicle convoy driving, and traffic management. When the authoritative node assumes a management role and requests individuals to participate and provide corresponding services, if the individual refuses, he will receive a reputation penalty.

[0191] like Fig. 9 The architecture shown is a schematic diagram of a two-way interaction mode scenario. After the interaction requester initiates an interaction request to the interaction responder, the responder makes an expected reputation evaluation of the requester based on the multi-factor evaluation method, and makes a trust decision based on the expected reputation evaluation, and provides or refuses the service according to the decision result. After receiving the response of service provision or service rejection, the requester makes a direct reputation evaluation of the responder in this interaction process. Behavior 2 is a one-way service mode. The service provider actively pushes the service, and the service receiver makes a direct reputation evaluation of the service provider based on the behavior after receiving the service. The global reputation evaluation is a comprehensive judgment made by the reputation authority organization on the reputation evaluation of individuals based on data information such as direct reputation evaluation between reputation participants, reputation evaluation evidence, and individual age attributes.

[0192] The expected reputation evaluation adopts a multi-factor evaluation method. Each factor is assigned a different weight according to its impact on the reputation, as shown below: Among them, ω i are the weights corresponding to different reputation influencing factors, and F i (u i ) is the mathematical expression of each influencing factor. The main influencing factors include: global reputation, historical reputation, reputation age, social network connection attributes, social attribute label description, etc.

[0193] Historical reputation mainly considers the impact of historical global reputation. Therefore, the calculation of the factors affecting historical reputation is as follows:

[0194]

[0195] Among them, Υ(u m ) is the historical global reputation evaluation value. ψ(m) is the time decay function, which can be taken as k is the decay constant, which is defined by the system itself, controls the decay rate, and adjusts the impact of the historical global reputation value on the expected reputation. m is the number of update cycles between the current global reputation evaluation and the historical reputation evaluation. For example, the m value of the global reputation evaluation in the current update week is 0, the m value of the global reputation evaluation in the previous update cycle is 1, the m value of the global reputation evaluation in the previous 2 update cycles is 2, and so on. The larger the m value, the smaller the time decay function value, that is, the earlier the historical reputation evaluation is made, the smaller the impact on the current expected reputation.

[0196] The impact of reputation age on expected reputation is that the older the individual is, the longer the individual maintains the current contribution level. The influencing factors of age are calculated as follows:

[0197]

[0198] Among them, Y max is the maximum age limit, Υ(u) is the current global reputation evaluation, and g is an adjustment constant, which is defined by the system to adjust the impact of age factors on expected reputation.

[0199] The social network connection attribute records the dynamic social trust network to which the interactive opponent is connected. The impact on the expected reputation is that when the interactive opponent belongs to the same dynamic social trust network as oneself, the probability of receiving corresponding feedback for providing services is very high, so the expected reputation will be improved. The influencing factors of the social network connection attribute are calculated as follows:

[0200]

[0201] Among them, C1Adjust the reputation value for the social network connection attribute.

[0202] The social attribute label is a label that a reputable authority organization makes to an individual's service behavior with a high contribution level after comprehensively analyzing the interaction behaviors of all individuals in the group with the individual. Different service categories correspond to different label descriptions. For example: traffic safety is SVP1, traffic management is SVP2, traffic information is SVP3, traffic efficiency is SVP4, commercial entertainment is SVR5, collaborative service is SVR6, and so on. For example: during the entire interaction process of the individual within the evaluation period, when the cumulative number of times the traffic management and collaborative service categories are provided exceeds the assignable threshold, the social attribute label SP = {SVR2, SVR6) is assigned to the individual. When the service request type of the service requester is consistent with its social attribute label, it means that the expected return level of providing this type of service is high, that is, the service requester is likely to provide this type of service to other individuals in the Internet of Vehicles group in the future. The influencing factors of the social attribute label are calculated as follows:

[0203]

[0204] Among them, C2 Adjust the value for the social attribute tag reputation.

[0205] Direct reputation evaluation is made by the service recipient to the service provider in direct interaction. Direct reputation evaluation adjusts the reputation by the degree of deviation of the interactive behavior, that is, whether the performance of the behavior can be rewarded with reputation compared with the standard interactive behavior defined by the system. After a single interactive behavior is completed, the evaluation is made immediately. Considering the practical significance of the reputation evaluation of the Internet of Vehicles, when the Internet of Vehicles individuals with high reputation perform well, their reputation value changes less; on the other hand, when individuals with high reputation do not maintain cooperative behavior, that is, the interaction strategy mutation node, their reputation evaluation should be adjusted significantly. Therefore, the reputation evaluation utility function η(u) is introduced. The specific expression is as follows:

[0206]

[0207] Among them, t D ,q D The interaction time and interaction quality values ​​of standard interaction behaviors are published by reputable authoritative centers based on the distribution of overall group interaction behaviors. D1 , Υ D2 is the reputation adjustment value of the interactive behavior, is a constant value, and is adjusted according to the evolution of the group. d ,q dThe interaction time and interaction quality value of the interaction behavior. The interaction quality value is obtained by normalizing the fuzzy description of the interaction quality evaluation. For example, the five-point method is used to divide the interaction quality from meeting the requirements to not meeting the requirements into five levels, corresponding to 1-5 points. After obtaining the evaluation, the score is converted into a numerical value as follows.

[0208]

[0209]

[0210] η(u) is the utility function, which can be an exponential utility function, a logarithmic utility function, a binomial utility function, etc. For example: the exponential utility function is as follows

[0211]

[0212] Among them, δ is the adjustment constant of the utility function. When it increases, the impact of interactive behavior on reputation will be aggravated. max It is the current maximum reputation value published by the reputation center based on the global reputation distribution of the group.

[0213] The global reputation evaluation of the reputation authority center is a comprehensive evaluation of the participating individuals, that is, a global judgment is made considering all the interaction behaviors of the individuals in the Internet of Vehicles group. Since the reputation authority center is also responsible for the management of Internet of Vehicles related services, and regards the interaction of the reputation authority nodes as the direct interaction of the center, it is necessary to consider direct reputation evaluation and recommended reputation evaluation when evaluating reputation. Due to the global reputation evaluation, it is necessary to integrate the mutual evaluation data between individuals in the group, and take into account the delay differences in data collection and transmission. Therefore, the global reputation update cycle T can be customized by the system and adjusted according to the evolution state of the current system. The update cycle is short and the consistency convergence of the reputation evaluation is accelerated. When the system evolution approaches the global equilibrium state, the reputation of each individual in the system tends to be stable, and the update cycle T can be appropriately increased to save system resource overhead.

[0214] The reputation authority center's global reputation evaluation of participating individuals is expressed as follows:

[0215]

[0216] Among them, β is the global reputation evaluation weight. D (u) is the direct reputation evaluation of the individual by the reputation center. J is the total number of effective participants in the direct reputation evaluation of the individual, that is, the total number of individuals who interacted with the individual and made effective direct reputation evaluations within the update period T. Dj It is the direct reputation evaluation made by the participating individuals in the update period T on other individuals based on their interactions with other individuals, with reference to the individual reputation evaluation. The credibility of direct reputation evaluation for individuals is determined by the reputation authority node, which assesses the reputation of each participating individual when collecting reputation evaluation evidence. When there are individuals with low credibility in the group, the rejection behavior of high credibility individuals should not be punished by credibility. However, according to the direct credibility evaluation rule, the rejection behavior will uniformly lower the credibility evaluation. Therefore, the credibility evaluation of individuals with low credibility should be filtered out to prevent selfish individuals from affecting the evolution of the group. Refer to the trust level classification to determine the value and set the threshold at the same time All direct reputation evaluations made by individuals whose credibility is lower than the threshold are considered invalid and removed. Since global reputation evaluation affects the reputation update of individuals, it has a wide impact on the reputation judgment and service decision of the Internet of Vehicles group. Therefore, global reputation evaluation supports asynchronous update, that is, when the change of individual reputation ΔR(u) exceeds the update threshold V R When , the global reputation evaluation of the individual is updated immediately, otherwise wait for the update period T. When a new individual joins, the reputation authority center assigns it a default initial reputation value γ0, which is the median of the current reputation distribution in the system.

[0217] The trust judgment of participating individuals is based on behavioral rules. Based on behavioral strategies, participants in the Internet of Vehicles can be divided into three categories: cooperative individuals, reciprocal individuals, and selfish individuals. Cooperative individuals adopt an unconditional cooperation strategy and respond to service requests from all requesters. Therefore, the trust of cooperative individuals in all participating individuals is unconditional trust, that is, the trust in the interactive opponents is 1. Selfish individuals adopt a selfish rejection strategy, ignoring the credibility of the other party, rejecting all service requests, and only obtaining resources and services from the Internet of Vehicles system, that is, the trust in the interactive opponents is 0.

[0218] In order to realize the incentive mechanism of social indirect reciprocity, suppress the refusal of selfish individuals in the Internet of Vehicles group to cooperate, and encourage the emergence, maintenance and dissemination of cooperative behavior based on indirect reciprocity in the group, the reciprocal individuals use the reputation of the interactive counterpart as the basis for distinguishing services, judge the contribution of the counterpart to the evolution of the Internet of Vehicles group and decide the corresponding services to provide to the counterpart. The service decision of the reciprocal individual is based on the global reputation distribution of the participating individuals in the Internet of Vehicles group, and the trust of the interactive counterpart is graded, which is expressed as follows:

[0219]

[0220] Among them, the trust level A={a1,a2,…,a N-1 ,a N}, is an ordered segmentation class. The credit cutoff points of trust grading are b1, b2, …, b N-1 ,b NThe reputable authority center is determined based on the overall distribution of reputation within the current group, that is, the dividing point is the corresponding quantile. For example, when the trust level is divided into 4 levels, b1, b2, and b3 correspond to the quantiles of 25%, 50%, and 75% of the reputation distribution respectively.

[0221] In order to accelerate the cooperative evolution of the Internet of Vehicles group, the evolution rules of individual reputation age are established. When the update cycle T arrives, the reputation authority center confirms whether the individual's reputation age Y is updated based on the comparison between the current individual's third reputation evaluation and its historical value. n -Υ n-1 |, whose value is greater than the fourth growth change threshold V Υ When the age value returns to the initial value 1, otherwise the credit age Y increases by 1 naturally; when the age value increases to the upper limit value Y max After that, it is considered as an evolutionary mature individual, and its age will not increase any more, as shown below:

[0222]

[0223] Based on the theory of spatial reciprocity, the network structure between individuals participating in the Internet of Vehicles is conducive to the maintenance and dissemination of cooperative behavior. Reciprocal individuals have the ability to independently choose social connections to adjust the network structure. By establishing a dynamic social trust network to isolate selfish individuals, it is effective to resist selfish individuals from occupying resources and malicious behaviors. At the same time, in some services of the Internet of Vehicles, such as vehicle platooning, efficient traffic speed guidance, priority vehicle traffic security management and other services rely on continuous and effective trust connection relationships. Therefore, establishing a dynamic social trust network will help improve the service benefits of the Internet of Vehicles group and resist the invasion of selfish individuals. The connection establishment in the social trust network is based on the evaluation of expected reputation, that is, considering the expected contribution level that the Internet of Vehicles group may obtain after establishing a connection with the individual, when the expected reputation is greater than the threshold, the connection is established. After the connection is established, the expected reputation of the connection is the lower value of the expected reputation evaluated by both parties, that is, R e(i,j) =min(R ei ,R ej ). The expected reputation evaluation refers to the aforementioned multi-factor evaluation method. After the connection is established, the individual's social network connection attributes and other information are updated, propagated within the group, and uploaded to the reputation evaluation authority node. When the expected reputation of a connection is lower than the connection threshold, the connection will be disconnected. The trust between individuals in the connected social network will be enhanced to encourage the maintenance and dissemination of cooperative behavior.

[0224] This application also continues to introduce specific process examples of individual interactions. After the service requester initiates a service request to the service responder, the service responder makes an individual reputation evaluation of the service requester based on the multi-factor individual reputation evaluation method and related reputation evidence data, and then graded and judged the trust of the requester based on its own service behavior decision criteria and the overall reputation distribution published by the authority center, and provided corresponding grade services. After providing the corresponding service, the requester uploads the reputation data generated this time and the reputation evidence such as the interaction process to the reputation authority organization. After receiving the service response, the service requester makes an individual reputation evaluation of the responder's interaction, and then uploads the updated individual reputation evaluation and reputation evidence to the reputation authority organization. Reputation data refers to the individual reputation evaluation value made after this interaction. Reputation evidence refers to the behavior information of both parties during this interaction, including the specific service request information of the service requester, the rejection or trust decision made by the service responder, the behavior information of the service provided, the quality evaluation of the service interaction, the corresponding time of the service interaction, etc.

[0225] When the participating individuals interact directly with the reputation authority node, the authority node will make an individual reputation evaluation for the interacting individuals by referring to the direct reputation evaluation method. The participating individuals upload the reputation data and evidence generated during the interaction process to the reputation authority node. After collecting the relevant reputation data and evidence, the node makes a judgment on it. When the reputation evaluation of some individuals mutates and exceeds the threshold, it is submitted to the reputation authority center to update the global reputation evaluation of the mutated individuals in a timely manner. The authority center makes a comprehensive judgment on the collected reputation data within the update cycle T, makes an evaluation of the global reputation within the group, and publishes the reputation distribution so that individuals in the group can update the service grading decision. At the same time, it evaluates the changes in the reputation of the individual, updates the reputation age of the individual and publishes it.

[0226] like Fig.10 In the illustrated embodiment, the process of establishing a dynamic social trust network is also shown. After participants 1, 2, 3, and 4 form a dynamic social network, participant 5 only needs to establish a social connection with any of the members to join the network. Participant 5 initiates a connection request to participant 1, a member of the network. Participant 1 evaluates the expected reputation of the individual based on the reputation evidence such as the individual's reputation, age, and social attribute label description. When the expected reputation is greater than the connection establishment threshold, the connection is allowed to be established. After the connection is established, the expected reputation of the connection is the lower value of the expected reputation evaluated by both parties, that is, R e(i,j) =min(R ei ,R ej). The expected reputation evaluation refers to the aforementioned multi-factor evaluation method. At the same time, both parties update the social network connection attribute information of the individual and share the connection establishment information with other members in the network. When the network connection structure changes, the corresponding information is uploaded to the reputation authority as reputation evidence. At the same time, after the change information is propagated among the members within the network, each member adjusts the expected reputation of the connection between each member according to the connection situation to monitor whether the connection is valid. In the process of network structure change, if an individual's individual reputation evaluation mutates due to the establishment of a new connection and exceeds the update threshold, the reputation authority organization updates the corresponding global reputation evaluation. When the reputation of participating individual 5 is low, it is difficult to join the network, it is impossible to establish a strong social relationship connection, and selfish behavior is difficult to obtain a better service return. Therefore, the selfish behavior of selfish individuals in the group will be suppressed, and the formation and maintenance of cooperative behavior will be promoted. At the same time, after the update cycle T, network members will evaluate the expected reputation of each connection established internally, and when the expected reputation value is lower than the connection threshold, the connection will be disconnected.

[0227] After the dynamic social trust network is established, some participating individuals will occupy an important position in the network. Fig.10 Participating individual 3 can forward the information of the authoritative center node, or because the number of individuals who have established interactive relationships is the highest. After individual 3 obtains an important position, if the service strategy is changed, it will become a selfish individual. Since the social network has been established, the individual may still enjoy high-level services during the update cycle. Therefore, selfish behavior may spread within the network, prompting members in the network to choose to "free ride" to enjoy services and resources and refuse services. At this time, the reputation authority node should focus on the reputation mutation of members in the network. After their service behavior decision criteria change, their interactive behavior and reputation evaluation will mutate in a short period of time. The reputation authority node immediately submits individual reputation changes to the authority center, timely and asynchronously updates the global reputation evaluation of some individuals, and suppresses the generation and spread of selfish behavior.

[0228] It should be understood by those skilled in the art that the above embodiments can be provided as methods, devices, or computer program products. These embodiments can take the form of complete hardware embodiments, complete software embodiments, or embodiments combining software and hardware. Although the above embodiments have been described, once the basic creative concepts are known to those skilled in the art, additional changes and modifications can be made to these embodiments. Therefore, the above description is only an embodiment of the present invention, and does not limit the scope of patent protection of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the scope of patent protection of the present invention.

Claims

1. A method for evaluating the mutual assistance utility of individuals in an Internet of Things network, characterized in that: The following steps are included: In an Internet of Things network, a first networking unit and a second networking unit are included. When providing services to the first networking unit, the second networking unit obtains a third reputation evaluation of the first networking unit. The third reputation evaluation includes a weighted average of a second reputation score and a reputation age score of the first networking unit. The second reputation score is: a normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the last update cycle, wherein the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's requests to interact with other networking units; The second networking unit provides differentiated grading services according to the distribution of the third reputation evaluation value of the first networking unit relative to the third reputation evaluation values ​​of all networking units in the Internet of Things network, The third reputation evaluation where ω i For different evaluation dimensions F i (u i ) weights, different evaluation dimensions F i (u i )include: Historical credit rating m is the current cycle and the second historical credit score Υ(u m ), k is the decay constant; Credit Age Score Among them, Y(u) is the credit age, Y max is the maximum age limit, Υ(u) is the current second credit score, g is the adjustment constant, When the update period T arrives, the reputation age is updated based on the comparison between the current individual's third reputation evaluation and its historical value. Specifically: When the change in the third reputation evaluation of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise: Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows: n is the number of cycles.

2. The method for evaluating the mutual assistance utility of individuals in an Internet of Things network according to claim 1, characterized in that: The differentiated hierarchical services include information transmission rate differences, response rate differences, response rate differences, or response priority differences.

3. The method for evaluating the mutual assistance utility of individuals in an Internet of Things network according to claim 2, characterized in that: The distribution of the third reputation evaluation value of the first networking unit relative to the second reputation scores of all networking units in the Internet of Things network is divided into groups, as shown below: Among them, the trust level A={a1,a2,…,a N-1 ,a N }, is the ordered segmentation class, the credit cutoff points of trust classification are b1, b2,…, b N-1 ,b N The second networking units provide the first networking unit with a percentile of second reputation scores of all networking units, and the second networking units provide the first networking unit with a response priority, a response rate, a response rate, and a transmission rate that decreases in sequence.

4. An individual mutual assistance utility evaluation system for an Internet of Things network, characterized in that: The system comprises a first networking unit and a second networking unit in an Internet of Things network, The second networking unit, when providing services to the first networking unit, obtains a third reputation evaluation of the first networking unit, the third reputation evaluation comprising a weighted average of the second reputation score and the reputation age score of the first networking unit, and the second networking unit is further configured to provide differentiated hierarchical services to the first networking unit according to the distribution of the third reputation evaluation value of the first networking unit relative to the third reputation evaluations of all networking units in the Internet of Things network; The second reputation score is: the normalized weighted average score of the first reputation score of the first networking unit by all networking units of the Internet of Things network that request to interact with the first networking unit in the previous update cycle, the first reputation score is positively correlated with the interaction time and interaction quality of the first networking unit's request to interact with other networking units, and the third reputation score is where ω i For different evaluation dimensions F i (u i ) weights, different evaluation dimensions F i (u i )include: Historical credit rating m is the current cycle and the second historical credit score Υ(u m ), k is the decay constant; Credit Age Score Among them, Y(u) is the credit age, Y max is the maximum age limit, Υ(u) is the current second credit score, g is the adjustment constant, When the update period T arrives, the reputation age is updated based on the comparison between the current individual's third reputation evaluation and its historical value. Specifically: When the change in the third reputation evaluation of an individual Δγ=| n -Υ n-1 |, greater than the fourth threshold V Υ When , the age value returns to the initial value 1, otherwise: Determine whether the credit age is equal to the upper limit value Y max If it is, it remains unchanged. If it is judged to be otherwise, the credit age Y will increase by 1 naturally. The specific expression is as follows: n is the number of cycles.

5. The individual mutual assistance utility evaluation system of the Internet of Things network according to claim 4 is characterized in that: The differentiated hierarchical services include information transmission rate differences, response rate differences, response rate differences, or response priority differences.

6. The individual mutual assistance utility evaluation system of the Internet of Things network according to claim 5 is characterized in that: The second networking unit is used to divide the distribution of the third reputation evaluation value of the first networking unit relative to the second reputation scores of all networking units in the Internet of Things network into groups, which is expressed as follows: Among them, the trust level A={a1,a2,…,a N-1 ,a N }, is the ordered segmentation class, the credit cutoff points of trust classification are b1, b2,…, b N-1 ,b N The first networking unit is provided with a response priority, a response rate, a response speed, and a transmission speed in descending order for the percentile points of the second reputation score of all networking units.

Images