A trust evaluation method for wireless network entities

Abstract

The present invention discloses a trust assessment method for wireless network entities, comprising: establishing a trust entity for entities participating in the trust assessment in the wireless network; a trust subject initiating a trust assessment on a trust object to obtain an initial trust value, performing a trust assessment on the trust object based on historical interaction records, and calculating a direct trust degree; quantifying and analyzing wireless network environmental factors and adjusting the direct trust degree; performing a trust assessment on the trust object based on recommendations about the trust object provided by a trust recommendation entity to the trust subject, and calculating an indirect trust degree; and aggregating the initial trust value, the adjusted direct trust degree, and the indirect trust degree to continuously quantify and assess the trust of the trust object to form a trust representation. The present invention can continuously quantify and assess the trust of network entities, improving the practicality and robustness of trust assessment.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a trust evaluation method for a wireless network entity. Background Art

[0002] With the widespread use of wireless networks and the rapid increase in the number of devices, wireless network environments are becoming increasingly complex and dynamic, increasing security risks. Traditional security mechanisms are often unable to cope with the changing and uncertain trust relationships in wireless networks. Therefore, trust assessment is particularly important. By accurately assessing the trustworthiness of wireless network entities, potential security threats can be effectively identified and prevented, ensuring the security of data transmission.

[0003] In the current evaluation scheme, the invention with publication number CN114357455A discloses a trust method based on multi-dimensional attribute trust evaluation, which characterizes the trustworthiness of nodes from multiple dimensions, proposes a node multi-dimensional attribute trust evaluation factor selection scheme based on the IoT application environment in edge computing, and dynamically measures the trustworthiness of the node's resource capability status and interactive service results in combination with the node computing environment, behavior and other attributes; performs trust modeling based on the correlation between trust and interactive behavior; and relies on the evaluation of multi-dimensional trust decision attributes to more comprehensively predict the node's identity and interaction results based on its performance and status. Through multi-dimensional trust modeling, a numerical evaluation basis can be provided for improving the success rate of interaction and building a trusted group. However, this invention does not consider the problem of internal attacks caused by the presence of malicious nodes in the network, and security is difficult to guarantee.

[0004] The invention with publication number CN116017469A discloses a trust assessment method suitable for wireless sensor networks. From the perspective of trust value accuracy and reflection of attack behavior, it summarizes the multi-dimensional trust assessment factor selection scheme, more comprehensively characterizes the trust relationship, and assists in trust measurement. The subjective weighted calculation of the trust factor is realized by using a feedback information fusion algorithm based on objective information entropy theory, which can effectively deal with malicious feedback. Trust is evaluated using reliability to reduce the impact of unreliable data on trust calculations. Risk control is also adopted to quantify the risk of the network in trust assessment. However, the trust assessment index designed by this invention ignores the problem of data packet loss caused by environmental factors such as high mobility and unstable wireless link transmission.

[0005] The invention, publication number CN117972301A, discloses a trust assessment method, device, and medium for accurately assessing the trustworthiness of intelligent agents. The method comprises: obtaining a collaborative environment information vector between a first and second intelligent agent, corresponding to a plurality of collaborative environment information; obtaining a regression coefficient vector based on historical collaboration records between the first and second intelligent agents, corresponding to the plurality of collaborative environment information; and obtaining a first trust value between the first and second intelligent agents based on the collaborative environment information vector and the regression coefficient vector. However, this method, which only analyzes historical collaboration records through a regression model, cannot fully characterize the complex and ever-changing trust relationships between intelligent agents. Therefore, it is necessary to introduce more multi-dimensional trust indicators and quantification methods. Summary of the Invention

[0006] Technical problem to be solved: Although current technical solutions have proposed various ideas for trust assessment, they often ignore malicious behavior in the assessment process, the assessment indicators are relatively simple, and the dynamic changes of environmental factors and their impact on trust assessment are not considered. To address the accuracy problem of trust assessment of wireless network entities, the present invention proposes a trust assessment method for wireless network entities. By adopting multi-party data integration and multi-dimensional indicator factors, and introducing a trust adjustment strategy based on wireless environment perception, the trust of network entities is continuously quantified and evaluated, thereby improving the practicality and soundness of trust assessment.

[0007] Technical solution:

[0008] A trust evaluation method for a wireless network entity, the trust evaluation method comprising the following steps:

[0009] S1, builds a trust body for entities involved in trust evaluation in wireless networks, including trust subjects, trust objects, and trust recommendation entities;

[0010] S2: The trust subject initiates a trust evaluation on the trust object to determine whether this is the first interaction between the trust subject and the trust object. If it is not the first interaction, the trust subject uses the most recent historical trust evaluation result as the initial trust value, and performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree. If it is the first interaction, the trust subject queries the trust object information, assigns an initial trust value, and then performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree.

[0011] S3, quantifies and analyzes wireless network environment factors and adjusts direct trust;

[0012] S4, based on the recommendation opinions on the trust object provided by the trust recommendation entity to the trust subject, perform trust evaluation on the trust object and calculate the indirect trust degree;

[0013] S5 aggregates the initial trust value of step S2, the adjusted direct trust degree of step 3, and the indirect trust degree of step S4, continuously quantifies and evaluates the trust of the trust object, and forms a trust representation.

[0014] Furthermore, in step S2, trust evaluation is performed on the trust object based on historical interaction records, and calculation of direct trust degree includes the following steps:

[0015] S21, calculating a direct trust representation result based on data interaction by the trust object in the wireless network data interaction behavior performance;

[0016] S22, calculating a direct trust characterization result based on the service quality of the business provided by the trust object in the wireless network;

[0017] S23, calculating the direct trust representation result based on the recommendation consistency by determining whether the trust recommendation provided by the trust object in the wireless network is honest and trustworthy;

[0018] S24, aggregate the direct trust characterization results based on data interaction, the direct trust characterization results based on service quality, and the direct trust characterization results based on recommendation consistency to obtain the direct trust degree of the trust subject to the trust object; the direct trust degree DT of the trust subject to the trust object i,j Characterized by:

[0019] DT i,j =α1f1(FR,AR,SR)+α2f2(SD,SS,SA)+α3f3(RC);

[0020] Where α1+α2+α3=1, α1, α2, and α3 are the weights corresponding to each indicator, f1(·), f2(·), and f3(·) are variable functions, which are customized according to the scenario, the trust subject's preferences, and the cognition of trust; FR, AR, and SR are the data forwarding rate factor, data transmission anomaly degree factor, and data interaction success rate factor used to calculate the direct trust representation results based on data interaction, respectively; SD, SS, and SA are the service delay factor, service satisfaction factor, and reliability factor used to calculate the direct trust representation results based on service quality, respectively; RC is the recommendation consistency factor used to calculate the direct trust representation results based on recommendation consistency.

[0021] Furthermore, in step S21, the process of calculating the direct trust representation result based on data interaction by the trust object in the wireless network data interaction behavior performance includes the following steps:

[0022] By detecting data forwarding behavior, it is determined whether the trusted object is negligent in forwarding data packets and maliciously intercepting data, and the data forwarding rate factor FR is calculated;

[0023] Assume that the data transmission rate of trusted object j in the most recent time window is tr new , whose data transmission rate at the time of last update was tr old , calculate the data transmission abnormality factor AR as:

[0024]

[0025] Assume that in the interaction records between trust subject i and trust object j, there are s successes and f failures. According to Bayesian inference, the data interaction success rate factor SR is calculated as follows:

[0026]

[0027] Furthermore, in step S22, the process of calculating the direct trust characterization result based on the service quality of the service provided by the trust object in the wireless network includes the following steps:

[0028] Assume that after trust subject i sends a request to trust object j, the average time required for trust object j to provide a response is t ij , the expected response time of trust subject i to trust object j is Calculate the service delay factor SD:

[0029]

[0030] Trust subject i evaluates the service of trust object j, judging whether it provides the required services according to the wireless network protocol or provides the subscribed information in a timely manner, and analyzes to obtain the service satisfaction factor SS;

[0031] The reliability factor SA of the trusted object j is evaluated by its failure rate m1, restart frequency m2, and average recovery time m3:

[0032]

[0033] Where max(m1), max(m2), and max(m3) represent the maximum values ​​of the failure rate, restart frequency, and average recovery time of all trust entities, respectively.

[0034] Furthermore, in step S23, it is assumed that the trust object j has a historical interaction entity set F j ={q1,q2,...,q M} has been evaluated, and the direct trust characterization results based on recommendation consistency are performed according to the following steps:

[0035] Set multiple trust attitudes, trust object j to historical interaction entity q x Direct trust Discrete processing, direct trust in different value ranges Representing different trust attitudes, we can get the trust object j towards the historical interaction entity q x Discrete recommended values ​​of

[0036] Forming a recommendation vector

[0037] For the entity set F j ={q1,q2,...,q M}, calculate the average recommendation value of each entity in the network

[0038]

[0039] Where P is the cumulative number of entities evaluated;

[0040] Forming the average recommendation vector

[0041] Calculate the recommended consistency factor RC:

[0042]

[0043] Furthermore, in step S3, wireless network environment factors are quantified and analyzed to adjust the direct trust

[0044] Calculate the environmental attenuation factor EDF:

[0045] EDF=w1E psr +w2E snr +w3E cu +w4E sd

[0046] Where w1+w2+w3+w4=1, w1, w2, w3, w4 are the weights of each indicator; E psr is the data packet loss rate factor used to reflect the stability of wireless link transmission, E snr E is the signal-to-noise ratio factor used to measure the ratio between the signal strength and the noise interference. cu is the channel utilization factor used to reflect the degree of channel congestion and communication efficiency, E sd is the moving speed difference factor used to reflect the difference in moving speed between trust subject i and trust object j;

[0047] The direct trust calculation result DT obtained in step S2 i,j Make adjustments to get the adjusted direct trust value

[0048]

[0049] Furthermore, in step S4, based on the recommendation opinion on the trust object provided by the trust recommendation entity to the trust subject, the trust object is evaluated and the process of calculating the indirect trust degree includes the following steps:

[0050] The trust subject uses the common interaction object with the trust object as the trust recommendation entity; the trust subject requests the trust recommendation entity for evaluation opinions on the trust object; the trust recommendation entity responds to the trust subject's request and returns the trust recommendation opinion in the format of: <recommender's unique identity, recommendee's unique identity, recommender's direct trust in the recommendee>; the trust subject calculates the credibility of the recommendation opinion from the trust recommendation entity based on multi-dimensional indicators; the trust subject uses the credibility to aggregate the recommendations from multiple recommendation entities to obtain the indirect trust in the trust object;

[0051] The unique identifier includes but is not limited to a MAC address, an IP address, an International Mobile Equipment Identity (IMEI), and a Universally Unique Identifier (UUID).

[0052] Furthermore, combined with the trust of trust subject i on recommended entity k, the recommendation credibility C i,k Calculated as:

[0053] C i,k =Intimacy i,k *CoF i,k *CoS i,k *DT i,k

[0054] In the formula, Intimacy i,k is the intimacy factor determined by the historical interaction between trust subject i and trust recommendation entity k, CoF i,k is the co-connection factor determined by the number of interacting entities between trust subject i and trust recommendation entity k, CoS i,k is the service correlation factor determined by the service dependency between the trust subject i and the trust recommendation entity k; DT i,k is the direct trust degree of trust subject i to trust recommendation entity k.

[0055] Furthermore, the normalized recommendation credibility is used as the weight to weight and normalize the direct trust provided by N trust recommendation entities to the trust object j, and the indirect trust RT i,j Characterized by:

[0056]

[0057] in, is the normalized trust subject i’s recommendation entity k lRecommended credibility, It is the direct trust of the trust recommendation entity on the trust object j.

[0058] Furthermore, in step S5, t h Trust evaluation result T of trust subject i on trust object j at any moment i,j (t h )for:

[0059]

[0060] Among them, μ is a variable weight factor, RT i,j is the indirect trust degree of trust subject i to trust object j, is the adjusted direct trust degree of trust subject i to trust object j, T i,j (t h-1 ) is t h-1 The degree of trust of trust subject i in trust object j at any moment.

[0061] Beneficial effects:

[0062] The present invention's wireless network entity trust assessment method integrates data from multiple sources, including direct trust, indirect trust, and historical assessment records, leveraging multi-dimensional indicators of entities' performance in data interaction, wireless network services, and trust recommendations to continuously quantify and assess the trust of wireless network entities. Furthermore, the method considers the impact of malicious insider attacks on assessment indicators and utilizes environmental attenuation factors for trust readjustment. This method accurately reflects the trustworthiness of wireless network entities, improving the practicality and robustness of trust assessments. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A schematic diagram of a scenario of a trust evaluation method for a wireless network entity according to the present invention;

[0064] Figure 2 This is a flow chart of the trust evaluation method for a wireless network entity of the present invention. DETAILED DESCRIPTION

[0065] The following examples may enable those skilled in the art to more fully understand the present invention, but are not intended to limit the present invention in any way.

[0066] The present invention discloses a trust evaluation method for a wireless network entity, the trust evaluation method comprising the following steps:

[0067] S1, builds a trust body for entities involved in trust evaluation in wireless networks, including trust subjects, trust objects, and trust recommendation entities;

[0068] S2: The trust subject initiates a trust evaluation on the trust object to determine whether this is the first interaction between the trust subject and the trust object. If it is not the first interaction, the trust subject uses the most recent historical trust evaluation result as the initial trust value, and performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree. If it is the first interaction, the trust subject queries the trust object information, assigns an initial trust value, and then performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree.

[0069] S3, quantifies and analyzes wireless network environment factors and adjusts direct trust;

[0070] S4, based on the recommendation opinions on the trust object provided by the trust recommendation entity to the trust subject, perform trust evaluation on the trust object and calculate the indirect trust degree;

[0071] S5 aggregates the initial trust value of step S2, the adjusted direct trust degree of step 3, and the indirect trust degree of step S4, continuously quantifies and evaluates the trust of the trust object, and forms a trust representation.

[0072] The present invention provides a trust evaluation method for a wireless network entity, comprising:

[0073] Wireless network trust entity: used to describe the entities involved in trust evaluation in wireless networks, including trust subjects, trust objects, and trust recommendation entities;

[0074] Wireless network trust recommendation mechanism: used to transfer trust between wireless network trust entities and assist in building trust relationships.

[0075] Direct trust representation for wireless networks: used to evaluate the trust of trusted objects based on historical interaction records and obtain direct trust;

[0076] Indirect trust representation for wireless networks: used to evaluate the trust of trusted objects based on multi-party recommendations and obtain indirect trust;

[0077] Direct trust adjustment based on wireless environment perception: used to quantify and analyze wireless network environment factors and adjust direct trust.

[0078] Trust evaluation model based on multi-dimensional indicators: used to continuously quantify and evaluate the trust of the trust object by integrating multi-dimensional indicators to form a trust representation.

[0079] Wireless network trust entities are used to describe the entities involved in trust evaluation in wireless networks, including trust subjects, trust objects, and trust recommending entities. The trust subject is the wireless network entity that initiates the trust evaluation, i.e., the entity making the trust judgment; the trust object is the wireless network entity undergoing the trust evaluation, i.e., the entity being subject to the trust judgment; and the trust recommending entity is the wireless network entity that provides recommendations regarding the trust object to the trust subject. Trust subjects, trust objects, and trust recommending entities can all be various entity types in the wireless network, such as UEs and base stations.

[0080] The wireless network trust recommendation mechanism refers to the trust transfer by means of trust recommendation, which includes the following steps:

[0081] (1) Selecting a trust recommendation entity: The trust subject selects a trust recommendation entity according to a selection strategy, wherein the selection strategy includes but is not limited to selecting a common interaction object with the trust object as a trust recommendation entity;

[0082] (2) Sending a trust recommendation request: The trust subject requests the trust recommendation entity to provide an evaluation opinion on the trust object;

[0083] (3) Feedback of trust recommendation opinion: The trust recommendation entity responds to the trust subject's request and returns the trust recommendation opinion, which is optional and has the format of: <recommender's unique identity, recommendee's unique identity, recommender's direct trust in recommendee>;

[0084] Unique identifiers in wireless networks include but are not limited to MAC addresses, IP addresses, IMEI (International Mobile Equipment Identity), and UUID (Universally Unique Identifier).

[0085] Direct trust representation for wireless networks refers to calculating the trustworthiness of a trusted object through its historical interaction records in the wireless network.

[0086] Optionally, based on the main behaviors of wireless network entities in the network, such as data flow transmission, wireless network services, and recommendations and evaluations made to other entities, direct trust representations can be divided into data interaction-based, service quality-based, and recommendation consistency-based. Among them, direct trust representations based on data interaction include but are not limited to the following three indicator factors:

[0087] (1) Data forwarding rate factor FR: Data forwarding is a basic data flow behavior. By detecting data forwarding, it can be determined whether the trust body has neglected to forward data packets, maliciously intercepted data, and so on.

[0088] (2) Data transmission anomaly factor AR: When conducting an attack or data theft, the trust body will usually show abnormal traffic behavior. However, it may also be caused by the network itself, such as network failure, network attack, network topology change, etc. Therefore, the stability of data transmission volume can be detected, and the relative difference in data transmission volume can be used to reflect its trustworthiness.

[0089] Optionally, let the data transmission rate of object j in the most recent time window be tr new , whose data transmission rate at the time of last update was tr old , then the AR value can be calculated as:

[0090]

[0091] (3) Data interaction success rate factor SR: i and j interact, with s successes and f failures. Optionally, SR can be calculated based on Bayesian inference as follows:

[0092]

[0093] The indicator factors for direct trust representation based on service quality include but are not limited to the following three:

[0094] (1) Service Delay Factor (SD): Fast response time is an important indicator for evaluating efficiency and performance. The service delay factor (SD) is defined as the difference between the average time required for trusted object j to provide a response after trusted subject i sends a request to trusted object j and the expected response time of trusted subject i to trusted object j.

[0095] Suppose that after i sends a request to j, the average time required for j to provide a response is t ij , the expected response time of subject i to object j is The service delay factor SD can be calculated as:

[0096]

[0097] (2) Service satisfaction factor SS: i evaluates j's service, such as whether the required service is provided according to the wireless network protocol or the subscribed information is provided in a timely manner. The wireless network protocol includes but is not limited to the service level agreement (SLA) and quality of service agreement (QoS).

[0098] (3) Service reliability factor SA: During the interaction between i and j, the service provided by a trusted entity should be highly reliable, that is, the service should be able to operate normally at any time and will not be interrupted or degraded due to single point failures or abnormal situations. Optionally, the reliability factor SA can be evaluated by using indicators such as j's failure rate m1, restart frequency m2, and average recovery time m3.

[0099]

[0100] Direct trust representation based on recommendation consistency refers to whether the evaluation opinions of trusted entity j on other trusted entities are within the permissible deviation range. In wireless networks, trusted entity j can lower the trustworthiness of other wireless network entities by making dishonest recommendations, or collude with other trusted entities to forge trust recommendations for itself, rapidly increasing its own trustworthiness. Therefore, for the same evaluation object, by calculating the degree of similarity between j's recommendations and those of other trusted entities, it is possible to determine whether the entity has malicious tendencies, thereby reflecting its trustworthiness in providing recommendations.

[0101] Optionally, let the trust object j have a set of historical interaction entities F j ={q1,q2,...,q M} has made an evaluation, the credibility of the trust recommendation can be calculated according to the following steps:

[0102] Set multiple trust attitudes, trust object j to historical interaction entity q x Direct trust Discrete processing, direct trust in different value ranges Representing different trust attitudes, we can get the trust object j towards the historical interaction entity q x Discrete recommended values The types of trust attitudes and their corresponding value ranges are determined according to actual needs. For example, For untrustworthy, For a neutral attitude, For trust, let 1 represent trust, -1 represent distrust, and 0 represent neutral attitude, that is:

[0103]

[0104] Forming a recommendation vector

[0105] For the entity set F j ={q1,q2,...,q M}, calculate the average recommendation value of each entity in the network as follows:

[0106]

[0107] Where P is the cumulative number of entities evaluated;

[0108] Forming the average recommendation vector

[0109] The recommended calculation method for consistency factor RC is:

[0110]

[0111] Therefore, based on the above trust indicators, we can directly trust DT i,j It can be characterized as:

[0112] DT i,j =α1f1(FR,AR,SR)+α2f2(SD,SS,SA)+α3f3(RC)

[0113] Where α1+α2+α3=1, α1, α2, and α3 are the weights corresponding to each indicator, and f(·) is a variable function. All of the above can be customized according to the scenario, the trust subject's preferences, and their cognition of trust.

[0114] Indirect trust representation for wireless networks refers to calculating the trustworthiness of a trust object through the recommendations provided by a trust recommendation entity.

[0115] Optionally, according to the design of the trust recommendation mechanism in wireless networks, let the trust entity sets that trust subject i and trust object j directly interact with be F i , F j , the common interactive trust entity set of the two can be selected as F i ∩F j ={k1,k2,...,k N}, as a set of trusted recommended entities. Any recommended entity k l , l=1,2...,N, after receiving the query request from the trust subject i, reply to the trust subject i about the evaluation of the trust object j Optional, trust recommendation format is: in represents k l Direct trust in j. Trust subject i receives trust recommendation from trust entity k l , l=1,2...,N recommendations, it is necessary to judge the credibility of the evaluation given by the trust subject i based on the information it has about the trust recommendation entity k, and use this as the weight to aggregate the recommendations from the above N trust entities, and finally obtain the indirect trust value of the trust subject i to the trust object j.

[0116] Alternatively, indirect trust representation includes three steps: obtaining trust recommendations, calculating recommendation credibility, and aggregating trust recommendations from multiple parties:

[0117] Getting trusted recommendations can be defined as:

[0118] The present invention fully considers the association between the trust subject i and the trust recommendation entity k (for ease of description, the subscript l is omitted in the following description), and the calculation of its recommendation credibility should include but not be limited to the following three indicators:

[0119] (1) Intimacy factor: optional, calculated by the proportion of the communication between trust subject i and trust recommendation entity k in the cumulative number of interactions of trust subject i;

[0120] (2) Co-connection factor CoF: optional, the number of common interaction entities between trust subject i and trust recommendation entity k, normalized by the maximum value;

[0121] (3) Service correlation CoS: The service interaction dependency between trust subject i and trust recommendation entity k. Optionally, it is calculated by normalizing the maximum value of the number of wireless services in which both parties participate.

[0122] Based on the above indicators, combined with the direct trust of trust subject i on recommended entity k, the recommendation credibility C i,k Calculated as:

[0123] C i,k =Intimacy i,k *CoF i,k *CoS i,k *DT i,k .

[0124] The recommendation credibility is normalized and used as the weight, and the N trust recommendation entities are provided After weighting and normalization, indirect trust can be characterized as follows:

[0125]

[0126] in, is the normalized value.

[0127] A direct trust adjustment strategy based on wireless environment awareness quantifies and analyzes wireless network environmental factors to adjust direct trust. Wireless environmental factors can affect network performance and stability, leading to problems such as signal attenuation, data transmission delays, and connection interruptions, which in turn impact trust relationships between trusted entities in the wireless network. To accurately assess the trustworthiness of trusted entity j, the node's trust value must be adjusted based on wireless environmental factors. This more accurately reflects the performance of trusted entity j under different environmental conditions, improving the accuracy of trust assessments and the overall security of the network. Furthermore, as previously discussed, direct trust is related to the interactions between trusted entities, while indirect trust stems from the evaluation and recommendations of other trusted entities.

[0128] Optionally, the present invention considers the impact of the wireless environment on direct trust and defines an environmental degradation factor (EDF). Its indicators include but are not limited to the following four:

[0129] (1) Data packet loss rate factor E psr : The instability of wireless links exacerbates the problem of packet loss in data transmission, leading to data transmission delays and other problems.

[0130] (2) Signal-to-noise ratio factor E snr : Measures the ratio of signal strength to noise interference, reflecting signal quality.

[0131] (3) Channel utilization factor E cu : Reflects the degree of channel congestion and affects the efficiency of communication.

[0132] (4) Moving speed difference factor E sd : The difference in mobile speed between trust subject i and trust object j will affect the stability of the connection. The greater the speed difference, the easier it is for the connection to be interrupted. sd =|Speed i -Speed j |.

[0133] The environmental attenuation factor is calculated as:

[0134] EDF=w1E psr +w2E snr +w3E cu +w4E sd

[0135] Among them, w1+w2+w3+w4=1, w1, w2, w3, w4 are the weights corresponding to each indicator. Similarly, the design can be customized according to the scenario.

[0136] The trust evaluation model based on multi-dimensional indicators refers to formulating appropriate aggregation rules based on direct trust representation for wireless networks, indirect trust representation for wireless networks, and direct trust adjustment based on wireless environment perception. It uses the process steps of trust initial value acquisition, direct trust calculation, direct trust adjustment based on wireless environment perception, indirect trust calculation, and trust aggregation update to continuously quantify and evaluate the trust object. That is:

[0137] Assume that trust subject i conducts trust evaluation on trust object j, and the time of initiating the evaluation is t h , then:

[0138] Step 1: Get the initial trust value T i,j (t h-1 ), that is, if the trust subject i and the trust object j are interacting for the first time, the trust subject i will query the identity, capabilities, resources, configuration and other information of the trust object j and customize its initial trust value. If it is not the first interaction, the historical trust evaluation result at the most recent moment will be used as the initial trust value;

[0139] Step 2: Calculate direct trust DT i,j ;

[0140] Step 3: Calculate the environmental attenuation factor (EDF) and adjust the direct trust calculation result in step 2 to obtain the adjusted direct trust value.

[0141]

[0142] Step 4: Calculate indirect trust RT i,j ;

[0143] Step 5: Aggregate the initial trust in step 1, the direct trust adjusted in step 3, and the indirect trust in step 4 as follows to obtain the trust evaluation result T of the trust subject i to the trust object j i,j (t h ):

[0144]

[0145] Among them, μ is a variable weight factor;

[0146] At this point, the trust assessment process ends.

[0147] Examples

[0148] The trust evaluation scenario of this example is shown in the attached Figure 1As shown in Figure 1, there are a total of 4 trust entities in a network, and data interaction exists between any two trust entities. Trust entity 1 performs trust evaluation on trust entity 3, that is, trust entity 1 is the trust subject, trust entity 3 is the trust object, and trust entities 2 and 4 are trust recommendation entities. Assume that trust entity 1 is not interacting with trust entity 3 for the first time, and the time difference between the most recent trust evaluation and the current moment is Δt, and the trust evaluation value is

[0149] This example adopts the following Figure 2 The trust evaluation process shown:

[0150] 1. Get the initial trust value, that is

[0151] 2. Calculate the direct trust degree DT of trustee 1 to trustee 3 1,3 , perform the following steps:

[0152] Calculate partial direct trust based on data interaction, such as calculating the following indicator factors:

[0153] (1) Obtain the data forwarding rate factor FR of the trusted entity 3, such as FR = 0.98;

[0154] (2) Calculate the data transmission anomaly factor AR of trust body 3: Assume that the data transmission rate of trust body 3 in the most recent time window is tr new , whose data transmission rate at the time of last update was tr old , then:

[0155]

[0156] (3) Calculate the data interaction success rate factor SR of trust body 3: Assume that trust body 1 and trust body 3 interact with each other, with s successes and f failures. According to Bayesian inference, it can be calculated as:

[0157]

[0158] Calculate partial direct trust based on service quality, such as using the following indicator factors:

[0159] (1) Calculate the service delay factor SD of trustee 3: Assume that after trustee 1 sends a request to trustee 3, the average time required for trustee 3 to provide a response is t 13 , the expected response time of trust subject i to trust object j is Then we have:

[0160]

[0161] (2) Service satisfaction factor SS: Trust entity 1 evaluates the service of trust entity 3, whether it provides the required services according to the wireless network protocol or provides the subscribed information in a timely manner, for example, SS = 0.95.

[0162] (3) Service reliability factor SA: Obtain the failure rate m1, restart frequency m2, average recovery time m3 and other indicators of trust body 3 during the interaction between trust body 1 and trust body 3 to evaluate its reliability, which is calculated as:

[0163]

[0164] (4) Calculate the partial direct trust based on recommendation consistency, that is, calculate the recommendation consistency index factor, as follows:

[0165] First, obtain the recommendation DT of trust body 3 to trust body 2 and trust body 4 3,2 , DT 3,4 ;

[0166] Then, for DT 3,2 , DT 3,4 Discrete processing is performed, for example, DT∈[0,0.5] is untrustworthy, DT∈(0.5,0.7] is neutral, and DT∈[0.7,1] is trustworthy. Let 1 represent trust, -1 represent distrust, and 0 represent neutral. 3,2 For example, there are:

[0167]

[0168] Similarly, we get R 3,4 , forming a recommendation vector

[0169] Next, calculate the average recommendation values ​​A2 and A4 of trust body 2 and trust body 4 in the network. Take A2 as an example:

[0170] A2=(R 1,2 +R 3,2 +R 4,2 ) / 3

[0171] It should be noted that in actual applications, the total number of network entities is usually very large, so the impact of a single trusted object on the average recommendation value is relatively small. This embodiment is only used for illustration.

[0172] This can form the average recommendation vector

[0173] Finally, the recommended consistency factor RC is calculated according to the following formula:

[0174]

[0175] Therefore, based on the above trust indicators, we can directly trust DT 1,3 It can be characterized as:

[0176] DT 1,3 =α1f1(FR,AR,SR)+α2f2(SD,SS,SA)+α3f3(RC)

[0177] Among them, α1+α2+α3=1, α1, α2, α3 are the weights corresponding to each indicator, such as α1=0.2, α2=0.35, α3=0.45, and f(·) is a variable function. The above can be customized according to the scenario, the trust subject's preferences, and the cognition of trust.

[0178] 3. Calculate the environmental attenuation factor (EDF), such as calculating the following index factors:

[0179] (1) Calculate the data packet loss rate factor E psr ;

[0180] (2) Calculate the signal-to-noise ratio factor E snr ;

[0181] (3) Calculate the channel utilization factor E cu ;

[0182] (4) Calculate the difference factor E between the moving speeds of trust body 1 and trust body 3 sd :

[0183] E sd =|Speed1-Speed3|

[0184] Therefore, the environmental attenuation factor is obtained as follows:

[0185] EDF=w1E psr +w2E snr +w3E cu +w4E sd

[0186] Among them, w1+w2+w3+w4=1, w1, w2, w3, w4 are the weights corresponding to each indicator. Similarly, the design can be customized according to the scenario.

[0187] 4. Readjust direct trust based on the environmental attenuation factor. The adjustment strategy is as follows:

[0188]

[0189] 5. Calculate the indirect trust RT of trust body 1 to trust body 3 1,3 , perform the following steps:

[0190] In this embodiment, trusted entities 2 and 4 are co-interacting entities of trusted entities 1 and 3 and can act as recommenders. Therefore, trusted entity 1 requests trusted entities 2 and 4 for recommendations about trusted entity 3, and trusted entities 2 and 4 accept the request and reply to trusted entity 1.

[0191] First, trust entity 1 obtains the recommendation, such as <2, 3, DT 2,3 > and <4, 3, DT 4,3 >

[0192] Next, trust entity 1 calculates the trustworthiness of trust entities 2 and 4, such as calculating the following index factors:

[0193] (1) Calculate the intimacy factor, such as the proportion of the communication between trust entity 1 and trust recommendation entities 2 and 4 in the cumulative number of interactions of trust entity 1;

[0194] (2) Calculate the co-connection factor (CoF): the number of common interaction entities between trust entity 1 and trust recommendation entities 2 and 4, respectively, and normalize the number by the maximum value;

[0195] (3) Service correlation CoS: The service interaction dependency between trust entity 1 and trust recommendation entities 2 and 4, respectively. The number of wireless services they participate in is calculated and normalized by the maximum value.

[0196] Based on the above indicators, combined with the direct trust of trust entity 1 on the trust recommendation entity, the recommendation credibility C can be obtained 1,2 , C 1,4 . With C 1,2 For example, we have:

[0197] C 1,2 =Intimacy 1,2 *CoF 1,2 *CoS 1,2 *DT 1,2

[0198] The multi-party recommendation trust aggregation is to normalize the recommendation credibility and use it as the weight, and then perform weighted and normalized processing on the recommendations provided. Finally, the indirect trust can be characterized as follows:

[0199] RT 1,3 =DT 2,3 *C ′ 1,2 +DT 4,3 *C ′ 1,4

[0200] Among them, C ′ 1,2 , C ′ 1,4is the normalized value.

[0201] 6. Initial Trust Adjusted direct trust and indirectly trust RT 1,3 The trust evaluation result T of trust entity 1 on trust entity 3 is obtained by aggregation as follows: 1,3 :

[0202]

[0203] Among them, μ is a variable weight factor;

[0204] This concludes the trust evaluation process. 1,3 The value is the trust degree of trust body 1 to trust body 3.

[0205] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A trust evaluation method for a wireless network entity, characterized in that: The trust evaluation method comprises the following steps: S1, builds a trust body for entities involved in trust evaluation in wireless networks, including trust subjects, trust objects, and trust recommendation entities; S2: The trust subject initiates a trust evaluation on the trust object to determine whether this is the first interaction between the trust subject and the trust object. If it is not the first interaction, the trust subject uses the most recent historical trust evaluation result as the initial trust value, and performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree. If it is the first interaction, the trust subject queries the trust object information, assigns an initial trust value, and then performs a trust evaluation on the trust object based on the historical interaction records to calculate the direct trust degree. S3, quantifies and analyzes wireless network environment factors and adjusts direct trust; S4, based on the recommendation opinions on the trust object provided by the trust recommendation entity to the trust subject, perform trust evaluation on the trust object and calculate the indirect trust degree; S5, aggregate the initial trust value of step S2, the adjusted direct trust degree of step 3, and the indirect trust degree of step S4, continuously quantify and evaluate the trust of the trust object, and form a trust representation; In step S2, trust evaluation is performed on the trusted object based on historical interaction records, and the calculation of direct trust degree includes the following steps: S21, calculating a direct trust representation result based on data interaction by the trust object in the wireless network data interaction behavior performance; S22, calculating a direct trust characterization result based on the service quality of the business provided by the trust object in the wireless network; S23, calculating the direct trust representation result based on the recommendation consistency by determining whether the trust recommendation provided by the trust object in the wireless network is honest and trustworthy; S24, aggregate the direct trust characterization results based on data interaction, the direct trust characterization results based on service quality, and the direct trust characterization results based on recommendation consistency to obtain the direct trust degree of the trust subject to the trust object; the direct trust degree DT of the trust subject to the trust object i,j Characterized by: DT i,j =α1f1(FR,AR,SR)+α2f2(SD,SS,SA)+α3f3(RC); Where α1+α2+α3=1, α1, α2, α3 are the weights of each indicator, f1(·), f2(·), and f3(·) are variable functions, customized according to the scenario, the trust subject's preferences, and the cognition of trust; FR, AR, and SR are the data forwarding rate factor, data transmission anomaly factor, and data interaction success rate factor used to calculate the direct trust representation results based on data interaction; SD, SS, and SA are the service delay factor, service satisfaction factor, and reliability factor used to calculate the direct trust representation results based on service quality, respectively; RC is the recommendation consistency factor used to calculate the direct trust representation results based on recommendation consistency; In step S3, the process of quantifying and analyzing wireless network environment factors and adjusting the direct trust degree includes the following steps: Calculate the environmental attenuation factor EDF: EDF=w1E psr +w2E snr +w3E cu +w4E sd Where w1+w2+w3+w4=1, w1, w2, w3, w4 are the weights of each indicator; E psr is the data packet loss rate factor used to reflect the stability of wireless link transmission, E snr E is the signal-to-noise ratio factor used to measure the ratio between the signal strength and the noise interference. cu is the channel utilization factor used to reflect the degree of channel congestion and communication efficiency, E sd is the moving speed difference factor used to reflect the difference in moving speed between trust subject i and trust object j; The direct trust calculation result DT obtained in step S2 i,j Make adjustments to get the adjusted direct trust value 2. The trust evaluation method for a wireless network entity according to claim 1, wherein: In step S21, the process of calculating the direct trust representation result based on data interaction by the trust object in the wireless network data interaction behavior performance includes the following steps: By detecting data forwarding behavior, it is determined whether the trusted object is negligent in forwarding data packets and maliciously intercepting data, and the data forwarding rate factor FR is calculated; Assume that the data transmission rate of trusted object j in the most recent time window is tr new , whose data transmission rate at the time of last update was tr old , calculate the data transmission abnormality factor AR as: Assume that in the interaction records between trust subject i and trust object j, there are s successes and f failures. According to Bayesian inference, the data interaction success rate factor SR is calculated as follows:

3. The trust evaluation method for a wireless network entity according to claim 1, wherein: In step S22, the process of calculating the direct trust characterization result based on the service quality provided by the trust object in the wireless network is as follows: The following steps are involved: set up After trust subject i sends a request to trust object j, the average time required for trust object j to provide a response is t ij , the expected response time of trust subject i to trust object j is Calculate the service delay factor SD: Trust subject i evaluates the service of trust object j, judging whether it provides the required services according to the wireless network protocol or provides the subscribed information in a timely manner, and analyzes to obtain the service satisfaction factor SS; The reliability factor SA of the trusted object j is evaluated by its failure rate m1, restart frequency m2, and average recovery time m3: Where max(m1), max(m2), and max(m3) represent the maximum values ​​of the failure rate, restart frequency, and average recovery time of all trust entities, respectively.

4. The trust evaluation method for a wireless network entity according to claim 1, wherein: In step S23, assume that the trust object j has a set of historical interaction entities F j ={q1,q2,...,q M } has been evaluated, and the direct trust characterization results based on recommendation consistency are performed according to the following steps: Set multiple trust attitudes, trust object j to historical interaction entity q x Direct trust Discrete processing, direct trust in different value ranges Representing different trust attitudes, we can get the trust object j towards the historical interaction entity q x Discrete recommended values Forming a recommendation vector For the entity set F j ={q1,q2,...,q M }, calculate the average recommendation value of each entity in the network Where P is the cumulative number of entities evaluated; Forming the average recommendation vector Calculate the recommended consistency factor RC:

5. The wireless network entity trust evaluation method according to claim 1, wherein: In step S4, based on the recommendation opinion on the trust object provided by the trust recommendation entity to the trust subject, the trust object is evaluated for trust, and the process of calculating the indirect trust degree includes the following steps: The trust subject uses the common interaction object with the trust object as the trust recommendation entity; the trust subject requests the trust recommendation entity for evaluation opinions on the trust object; the trust recommendation entity responds to the trust subject's request and returns the trust recommendation opinion in the format of: <recommender's unique identity, recommendee's unique identity, recommender's direct trust in the recommendee>; the trust subject calculates the credibility of the recommendation opinion from the trust recommendation entity based on multi-dimensional indicators; the trust subject uses the credibility to aggregate the recommendations from multiple recommendation entities to obtain the indirect trust in the trust object; The unique identifier includes but is not limited to a MAC address, an IP address, an International Mobile Equipment Identity (IMEI), and a Universally Unique Identifier (UUID).

6. The trust evaluation method for a wireless network entity according to claim 5, wherein: Combined with the direct trust of trust subject i on recommended entity k, the recommendation credibility C i,k Calculated as: C i,k =Intimacy i,k * CoF i,k *CoS i,k *DT i,k In the formula, Intimacy i,k is the intimacy factor determined by the historical interaction between trust subject i and trust recommendation entity k, CoF i,k is the co-connection factor determined by the number of interacting entities between trust subject i and trust recommendation entity k, CoS i,k is the service correlation factor determined by the service dependency between the trust subject i and the trust recommendation entity k; DT i,k is the direct trust degree of trust subject i to trust recommendation entity k.

7. The trust evaluation method for a wireless network entity according to claim 5, wherein: The normalized recommendation credibility is used as the weight, and the direct trust provided by N trust recommendation entities to the trust object j is weighted and normalized. The indirect trust RT i,j Characterized by: in, is the normalized trust subject i’s recommendation entity k l Recommended credibility, It is the direct trust of the trust recommendation entity on the trust object j.

8. The wireless network entity trust evaluation method according to claim 1, wherein: In step S5, t h Trust evaluation result T of trust subject i on trust object j at any moment i,j (t h )for: Among them, μ is a variable weight factor, RT i,j is the indirect trust degree of trust subject i to trust object j, is the adjusted direct trust degree of trust subject i to trust object j, T i,j (t h-1 ) is t h-1 The initial trust value of trust subject i to trust object j at each moment.

Images