A Multi-Factor Data Reliability Evaluation Method in Fog Computing

Abstract

The present invention relates to a multi-factor data reliability evaluation method in fog computing, belonging to the field of mobile communication technology. Data sharing enables users to obtain important information around them in a timely manner. However, due to the mobility and variability of users during the data sharing process, users do not fully trust each other. Due to sensor failures, virus infections, or even selfish reasons, devices may spread false information. To reduce the impact of these malicious messages on other users, the spread of false messages must be suppressed. The present invention designs a method to quantify the authenticity of data, enabling data requesters to obtain more reliable data during the data sharing process. The fog node makes a preliminary judgment on the authenticity of the data through Bayesian model inference to obtain the credibility value of the data itself. Based on the preliminary judgment result of the fog node, the requester comprehensively considers the trust value based on the provider's experience and historical interactions to obtain the satisfaction with the data, making the reliability evaluation of the data more comprehensive.

Description

Technical Field

[0001] The present invention belongs to the technical field of mobile communication, and relates to a method for evaluating the reliability of multi-factor data in fog computing. Background Art

[0002] Currently, "fog computing" is an emerging distributed computing paradigm that extends computing, communication, and caching to the network edge, with inherent characteristics such as low latency, mobility, and wireless access capabilities. Fog computing provides a platform for data sharing between mobile devices. Data sharing enables users to obtain important information around them in a timely manner, providing great convenience for users' lives. However, due to the mobility and variability of devices, devices do not fully trust each other. When there are malicious users in the network, they will deliberately spread false data, causing trouble for other users' judgment of the data. Therefore, how to effectively evaluate the credibility of data is an important issue in data sharing.

[0003] Based on the above problems, the present invention designs a method to quantify the authenticity of data. First, the fog node makes a preliminary judgment on the authenticity of the data through Bayesian inference to obtain the credibility value of the data itself. Secondly, based on the preliminary judgment result of the fog node, the requester comprehensively considers the trust value based on the provider's experience and historical interactions to obtain the satisfaction with the provided data, making the reliability evaluation of the data more comprehensive. Thus, the best data can be found for the data requester. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method for evaluating the reliability of multi-factor data in fog computing.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A method for evaluating the reliability of multi-factor data in fog computing, the method comprising the following steps:

[0007] S1: A two-layer data sharing model based on the fog network;

[0008] S2: A data preliminary judgment method based on Bayesian inference;

[0009] S3: A trust value update method based on user historical interactions and experience;

[0010] S4: A data reliability evaluation scheme based on the entropy weight method.

[0011] Further, in step S1, a data sharing model is established.

[0012] Further, in step S2, since the data is uncertain, in order to prevent users from sharing false information, the fog node will make a preliminary judgment on the content uploaded by the data provider. For a certain data el The credibility of

[0013]

[0014] where represents the data credibility of the data uploaded by provider j regarding event e l , represents the distance between user j and the location where the event occurs, represents the time t when user j learns about the event content j and the time difference between the time t when the event occurs b is the lower limit of data credibility, and α and β control the change rate of credibility, where α + β = 1. The shorter the distance of user j from the event occurrence and the earlier the time to learn about the event occurrence, the more trustworthy the data.

[0015] The fog node collects the data set within its communication range, and the credibility set C of data e can be obtained using the above formula (1) l l , Based on the obtained credibility set, the aggregated credibility P of event e is calculated using the Bayesian model inference: l

[0016]

[0017] where P(e / C) represents the aggregated credibility of event e, represents the complementary event of e, and P(c j / e) = c j ; P(e) represents the prior probability of event e. P(e / C) ∈ [0,1]. Once P(e / C) exceeds the preset threshold Thr, the fog node considers the data related to this event to be true and trustworthy; if P(e / C) does not exceed the set threshold, the data is considered unreliable. Users who upload unreliable data will be kicked out of the sharing list and will no longer participate in this round of data sharing.

[0018] Furthermore, in step S3, when users request data, they should not only consider the credibility of the data itself but also the behavior of data providers in data sharing. If the providers are malicious, they may have behaviors such as missing data transmission, forging data, and providing false data services. Therefore, data requesters should also combine the past behaviors of users and the satisfaction with the services provided by providers before to judge the authenticity of the data based on the pre-judgment results of the fog node on the data.

[0019] ​​The experience-based trust value updates the user's trust value by leveraging the user's past behaviors, indirectly judging the authenticity of data, and this value accumulates over time. After data sharing is completed, the data requester scores the provider based on data quality as T i,j ,T i,j ∈(-1,1). The fog node averages the scores given by the requesters L is the number of requesters interacting with data provider j this time. Use s j to represent the size of the experience-based trust of user j, s j ∈(-1,1), and update s on the basis of j .

[0020] When the requester initiates a sharing request to the provider, it measures the satisfaction with the services previously provided by the provider, and this satisfaction is related to the historical interactions between the two. p ij represents the satisfaction level with the current service, p ij ∈[0,1]. The cumulative value of historical interactions is:

[0021]

[0022] where p ij (t i ) represents the satisfaction level with the currently provided data; represents the accumulation of satisfaction, which is continuously updated The moment of requesting service is denoted as t n = t N >…> t2 > t1; N is the number of times of requesting data. A higher N means the requester has more prior knowledge about the provider, thus can judge the provider more accurately and obtain more complete and accurate data.

[0023] Furthermore, in step S4, the entropy weight method is used to evaluate the scoring weights of the three scoring indicators of data credibility, experience-based trust value, and historical interaction score, so as to comprehensively evaluate the reliability of the data provided by the user. In the stage of selecting a data provider, the requester R i evaluates the relevant providers based on the above three indicators and establishes an n×m provider scoring matrix W n×m , where n is the number of data providers and m = 3. First, to obtain a standardized evaluation matrix the normalization method is used to normalize the matrix elements. Secondly, calculate the weight of each evaluation indicator m of the provider and the information entropy H of the evaluation indicator m mNormalize the information entropy of the rating index m, and finally obtain the current satisfaction score of the requester i for each provider j regarding the data l. Next, the data requester initiates a data sharing request to the data provider with the highest score to obtain the most reliable data.

[0024] The beneficial effects of the present invention are as follows: According to the characteristics of the provided network scenario, the fog node uses Bayesian inference to pre-judge the data summary information uploaded by the data provider, and the data requester combines the trust value based on experience and historical interactions of the provider on the basis of the pre-judgment result to obtain the satisfaction with the provided data.

[0025] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, they will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following specification. Brief Description of the Drawings

[0026] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0027] Figure 1 It is a data sharing model based on fog computing. Detailed Embodiments

[0028] The following specific examples illustrate the embodiments of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0029] Among them, the accompanying drawings are only for illustrative purposes, showing only schematic diagrams, not physical diagrams, and cannot be understood as a limitation to the present invention; in order to better illustrate the embodiments of the present invention, some components in the accompanying drawings will be omitted, enlarged, or reduced, and do not represent the dimensions of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the accompanying drawings may be omitted.

[0030] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc. indicating the orientation or positional relationship, they are based on the orientation or positional relationship shown in the accompanying drawings. This is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the accompanying drawings are only for illustrative purposes and should not be construed as a limitation of the present invention. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0031] Figure 1 A data sharing network model composed of users and fog nodes is described. This model mainly includes users and interconnected fog nodes.

[0032] User: Equipped with advanced communication devices and having wireless communication capabilities, collects local data from sensing devices, uploads the summary information of the data (data size, time, address, etc.) to the fog node closest to its communication distance, and shares the data with data requesters. The user who collects and shares data acts as a data provider, denoted as The user who needs data acts as a data requester, denoted as Users play different roles according to their different needs.

[0033] Fog node: Has certain storage and computing capabilities, collects the set of data summary information uploaded by users within its communication range, evaluates the credibility of the data set, and obtains a preliminary judgment result for the reference of data requesters.

[0034] During the data sharing process, due to sensor failures, virus infections, or even selfish reasons, users may spread false information. To reduce the impact of these malicious messages on other users, the spread of false messages must be suppressed. For this purpose, the present invention designs a method to quantify the authenticity of data. First, the fog node makes a preliminary judgment on the data to obtain the credibility value of the data itself. Secondly, based on the preliminary judgment result of the fog node, the requester comprehensively considers the trust value based on the provider's experience and historical interactions to obtain the satisfaction with the provided data, making the evaluation of the authenticity of the data more comprehensive.

[0035] 1) Preliminary judgment of data

[0036] The preliminary judgment of data is equivalent to the evaluation of the credibility of the data itself, and this step is carried out on the fog node. First, the fog node groups all the data uploaded by users {E1, E2,..., E l,…}, where E l represents the event e lA data set, such as the event: "There is snow accumulation on a certain road section". However, the data in the same group do not have the same credibility. The credibility of the data uploaded by user j is defined as:

[0037]

[0038] where represents the credibility of the data related to event e uploaded by provider j l . represents the distance between user j and the location where the event occurred, represents the time t when user j learned about the event content j and the time difference between the time t when the event occurred b is the lower limit of data credibility, and α and β control the change rate of credibility, where α + β = 1. The shorter the distance of user j from the event occurrence and the earlier the time when user j learned about the event occurrence, the more trustworthy the data.

[0039] The fog node collects the data set within its communication range. Using the above formula (4), the credibility set C of data e l can be obtained. l , Based on the obtained credibility set, the aggregated credibility P of event e is calculated using the Bayesian model inference: l :

[0040]

[0041] where P(e / C) represents the aggregated credibility of event e, represents the complementary event of e, and P(c j / e) = c j ; P(e) represents the prior probability of event e. P(e / C) ∈ [0, 1]. Once P(e / C) exceeds the preset threshold Thr, the fog node considers the data related to this event to be true and trustworthy; if P(e / C) does not exceed the set threshold, the data is considered unreliable. The user who uploads unreliable data will be kicked out of the sharing list and will no longer participate in this round of data sharing.

[0042] 2) User's trust value based on experience

[0043] The trust value based on experience updates the user's trust value by using the user's past behavior, indirectly judging the authenticity of the data, and this value accumulates over time. After the data sharing is completed, the data requester will score the provider based on the data quality as T i,j , T i,j ∈ (-1, 1). The fog node averages the scores of the requesters L is the number of requesters that interact with data provider j this time. j represents the trust level of user j based on experience, s j ∈(-1,1), the update criteria are as follows:

[0044] like Trust j Increase to:

[0045]

[0046] like Trust j Reduced to:

[0047]

[0048] where s j represents the trust value based on experience at the current moment, s' k Represents the updated trust value. η is a positive increment factor 0<η<1; μ is a negative decay factor -1<μ<0. |μ|>|η|, once the user has cheating behavior, trust is easily destroyed and it is difficult to establish trust. λ represents the forgetting factor, 0<λ<1, t represents the time difference between the interaction moment between the current data provider and the data requester and the previous interaction moment. The user's behavior may change after a period of time. In order to reduce the impact of the trust value accumulated by the previous behavior on the current moment, the previous trust value is discounted by λ t or (λ -t ), which slows down the rate at which trust based on experience increases or decreases.

[0049] 3) User’s historical interactions

[0050] When a requester initiates a sharing request to a provider, the satisfaction of the provider’s previous services is measured, which is related to the historical interactions between the two. ij Indicates the level of satisfaction with the current service, p ij ∈[0,1]. The cumulative value of historical interactions is:

[0051]

[0052] where p ij (t i ) indicates the level of satisfaction with the data currently provided; Indicates the accumulation of satisfaction and continuous updating The instant of requesting service is denoted as t n =t N>…>t2>t1; N is the number of times of requesting data. A higher N means that the requester has more prior knowledge about the provider, thus can judge the provider more accurately and obtain more complete and accurate data.

[0053] In the stage of selecting a data provider, the requester R i evaluates relevant providers based on the above three metrics and establishes an n×m provider scoring matrix as:

[0054]

[0055] where n is the number of data providers, n ≤ j; c n represents the credibility score of the data l provided by Q n ; s n represents the experience-based trust value of the provider Q n ; h n represents the historical interaction score of the requester R i with the provider Q n .

[0056] After obtaining the scoring matrix W, the entropy weight method is used to evaluate the scoring weights of the three scoring metrics of data credibility, experience-based trust value, and historical interaction score, so as to evaluate the data provided by the user. First, to obtain a standardized evaluation matrix the normalization method is used to normalize the matrix elements:

[0057]

[0058] Calculate the weight of each evaluation metric m of the provider:

[0059]

[0060] The information entropy H of the evaluation metric m m is:

[0061]

[0062] Normalize the information entropy of the evaluation metric m:

[0063]

[0064] Obtain the current data reliability score of the requester i for each provider j regarding the data l:

[0065]

[0066] The data requester obtains the data reliability score G jAfter that, a data sharing request is sent to the data provider with the highest relevant score to obtain trustworthy data, preventing the spread of false messages.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A multi-factor data reliability evaluation method in fog computing, characterized in that: The method includes the following steps: S1: Establish a two-layer data sharing model based on the fog network; S2: A data pre-judgment method based on Bayesian inference; S3: Update the trust value based on the user's historical interactions and experience; S4: Evaluate the data reliability based on the entropy weight method; In the said S1, a data sharing model is established; the model includes users and interconnected fog nodes; users are equipped with communication devices and have wireless communication capabilities; users collect local data from sensing devices and upload the data summary to the fog node closest to their communication distance to share the content with data requesters; fog nodes have storage and computing capabilities, collect the data uploaded by users within their communication range, and pre-judge the credibility of the data; among them, the data summary includes data size, time, and address; In S2, the data is uncertain. To prevent users from spreading false information, the fog node pre-judges the data uploaded by users within its coverage area; the credibility of a certain data l is defined as This credibility is related to the distance and timeliness of the content occurrence; the fog node collects the data set within its communication range and the corresponding credibility set C. Based on the obtained credibility set, the fog node calculates the aggregated credibility P of the data l using Bayesian inference. When P is greater than the threshold Thr, the data itself is considered credible; In the said S3, when a user requests data, not only the credibility of the data itself is considered, but also the behavior of the data provider in data sharing is considered; if the provider is malicious and there are situations of missing transmission and forged data, providing false data services; the requester judges the authenticity of the data based on the data pre-judgment result of the fog node in combination with the user's past behavior and the satisfaction of the services provided by the provider before; the user's historical interaction is to measure the satisfaction of the services provided by the provider before when the requester initiates a sharing request to the provider, and this satisfaction is related to the historical interaction between the two; the trust value based on experience is to update the user's trust value using the user's past behavior and indirectly judge the authenticity of the data, and this trust value accumulates over time; In step S4, during the stage of selecting a data provider, the requester R i evaluates relevant providers based on three indicators: pre-judgment of data, the user's historical interactions, and the user's trust value based on experience, and establishes an n×m provider scoring matrix W; after obtaining the scoring matrix W, the entropy weight method is used to evaluate the scoring weights of the three scoring indicators of the data, and a comprehensive evaluation of the data provided by the user is conducted to finally obtain the satisfaction value of the data.

Images