An edge caching network user-to-user content sharing incentive method based on power perception

Abstract

The application discloses an edge cache network user inter-content sharing incentive method based on power perception, and comprises the following steps: establishing an edge cache network scene; establishing a content sharing management system in the edge cache network; and realizing the incentive of user inter-content sharing based on power perception, which comprises the following steps: calculating the evaluation of B-type users on content and the energy consumption cost of S-type users in transmitting content; B-type users submitting their request information to a small base station, and S-type users submitting their state information to the small base station; obtaining a matching success set through a user matching algorithm; determining the final integral deduction amount of B-type users and the final integral reward amount of S-type users by using a pricing scheme; and the users in the matching success set sharing content according to the information published by the small base station. Compared with the prior art, the application can effectively encourage the sharing of content among users, take into account the interests of both buyers and sellers, and maximize social welfare as much as possible.

Description

Technical Field

[0001] This invention belongs to the field of edge caching technology, and mainly relates to a method for incentivizing content sharing among users in an edge caching network based on power consumption awareness. Background Technology

[0002] With the continued proliferation of smartphones, tablets, and other mobile smart devices in recent years, their storage, computing, and communication capabilities have made significant progress. Simultaneously, due to the continuous emergence of multimedia applications for entertainment, social networking, and work, content delivery services, represented by video, now account for 69% of all data traffic on these user devices (UDs), and this dominant trend is expected to increase to 79% next year. Under the traditional mobile cellular network architecture, these content delivery services adopt an infrastructure-centric approach, where each UD downloads the source content it is interested in from nearby small base stations. However, this centralized architecture faces two major challenges in the face of the current continuous growth of mobile services: increasingly congested backhaul links and expensive data traffic costs.

[0003] To address the challenges of traditional architectures, Mobile Edge Caching (MEC) networks are considered a promising solution. They cache content such as popular video files through mobile edge nodes (e.g., UDs and small base stations) and allow neighboring UDs to share content via Device to Device (D2D) communication. Figure 1 The illustration shows a typical campus application scenario. The advantage of MEC networks lies in replacing backhaul links with the caching capabilities of UDs (User-Defined Objects). However, the performance of MEC systems still depends on the level of cooperation among participants. Unfortunately, due to concerns about resource consumption (such as UD's power consumption) caused by content sharing operations, users tend to download content from their neighbors while trying to avoid sharing with them. Clearly, without intervention, this greedy yet rational behavior of users will severely degrade the overall performance of MEC networks.

[0004] Therefore, designing an incentive method to encourage content sharing among users is a significant and widely applicable task. In this invention, user sharing behavior is incentivized through the deduction and reward of points. Users requesting content need to pay a certain number of points to users providing content as a reward for their sharing. Once a user receives content, they will provide points to the content provider according to a predetermined number, thus completing one content sharing session. Users earn points by participating in content sharing and use these points to obtain the content they need from other users. If a selfish user does not participate in content sharing for a long period, they will not be able to earn more points. When their points are exhausted, they will no longer be able to obtain content from the network and will have to choose to share content to earn more points. Summary of the Invention

[0005] Purpose of the invention: To address the above problems, this invention proposes a content sharing incentive method for users in an edge cache network based on power awareness. It establishes a content sharing management system to facilitate content sharing among users in the MEC network, realizing a content sharing mechanism for users and an incentive strategy for the MEC network, thereby maximizing social welfare.

[0006] Technical Solution: To achieve the objectives of this invention, the technical solution adopted is: a method for incentivizing content sharing among users in an edge caching network based on power consumption awareness, comprising the following steps:

[0007] An edge caching network scenario is established; the edge caching network consists of a small base station equipped with an edge server and N mobile user equipments randomly distributed around the small base station;

[0008] A content sharing management system is established in the edge caching network. The small base station acts as the system manager and classifies users into type B and type S according to their requests. Type B users are those who initiate content caching requests, while type S users are those who accept content caching requests and provide content data blocks.

[0009] Based on power consumption awareness, which takes into account the energy consumption of S-type users transmitting data, users in the edge buffer network share content. After B-type users obtain the content shared by S-type users, the small base station deducts a certain number of points from the B-type users and distributes these points to the S-type users as a reward for the S-type users' sharing behavior.

[0010] The number of points is determined based on the valuation of type B users and the energy consumption cost of type S users, and is determined through user matching algorithms and pricing schemes; small base stations aim to maximize social welfare, which refers to the sum of the utilities of type B users and type S users.

[0011] Furthermore, establish an edge caching network that allows D2D communication, including the following:

[0012] (1.1) Each user equipment has a limited device power Q and cache capacity D, and has two communication interfaces: a Wi-Fi or Bluetooth interface for direct D2D communication, and a cellular communication interface for direct connection to a small base station.

[0013] (1.2) User-requested content is influenced by content popularity, and its distribution follows a Zipf distribution; in the content library The probability that the content at the h-th position is requested is

[0014] ,

[0015] Where ε is the exponential constant of the Zipf distribution, This represents the number of times the content at position h was requested, where M represents the content library. Total amount of stored content, This represents the total number of times all content was requested;

[0016] (1.3) The probability that there are l neighboring users within a distance d around the user is:

[0017] ,

[0018] Here, λ represents that the user's neighboring users follow a Poisson distribution with density λ.

[0019] Furthermore, h-type content corresponds to the content library. The content data block at position h; The content sharing management system uses a points system issued to users by small base stations. Users can earn points by sharing content in the system. Obtaining content from other users requires deducting some points. These points have an expiration date and are reclaimed by the system after expiration.

[0020] Furthermore, the specific steps for each round of content sharing are as follows:

[0021] (2.1) Calculate the content valuation of type B users and the energy consumption cost of content transmission for type S users;

[0022] (2.2) Type B users submit their request information to the small base station, and Type S users submit their status information to the small base station;

[0023] (2.3) The small base station collects information from each user and uses a user matching algorithm to find the set of successfully matched users;

[0024] (2.4) Based on the successful matching set obtained in step (2.3), the final points deduction amount for type B users and the final points reward amount for type S users are determined using the pricing scheme;

[0025] (2.5) Type B users and Type S users share content based on the information published by the small base station.

[0026] Furthermore, step (2.1) calculates the content valuation for type B users and the energy consumption cost for content transmission for type S users, specifically including:

[0027] In a content sharing session, there are m type B users and n type S users, where m + n ≤ N, and a system administrator consisting of a small base station.

[0028] The B-type user set is defined as follows: The S-type user set is defined as follows: The current battery level and available cache capacity of any user's device are denoted as Q and D, respectively.

[0029] for It is calculated based on the probability of content being requested, the user's demand for the content, and the ease of obtaining the content. Valuation of H-type content , represents as

[0030] , and ;

[0031] in, Z represents the probability that content at position h in the content library will be requested; Z represents the average number of cached data blocks per user in the network, calculated as follows: M is the amount of content stored in the content library, and N is the number of mobile user devices; represent Demand for H-type content; yes The probability that there are l neighboring users within a distance d; express The number of data blocks that can provide content from surrounding neighbors; α and β represent the user's preference for the urgency of content demand and the ease of content acquisition, respectively, and their values ​​are set by the system;

[0032] for Calculate based on the power consumed by transmitting a data block of size p. Energy consumption cost , represents as

[0033] , and ;

[0034] in, express Power consumption per unit size of data block transmitted. express The device's current battery level, and the following conditions must be met. That is, the power consumption required for transmission must be less than The device's current power level; η represents the parameter coefficient set by the system.

[0035] Furthermore, in step (2.2), type B users submit their request information to the small base station, and type S users submit their status information to the small base station. Each user submits their information in a sealed manner over a discrete finite time {0,1,…,T}, specifically including:

[0036] At the start, type B users and type S users submit their respective information to the small base station within a specified time, and only the small base station can view this information.

[0037] The request information is represented as ,in, for The required content type for for The demand for content data blocks; The calculation method is to divide the size of the content data block by the unit size p, where the data block size is W. Type of content, Demand for it ; The points deduction per unit number of content data blocks, i.e. Points deducted when content is accessed; Valuation Related, represented as , , It's a correlation coefficient, referring to type B users sharing content in the same round. The values ​​are the same, and with As the number of failed content retrieval attempts increases, x gradually increases;

[0038] Provide status information based on your own available resources and equipment: ,in for supply Number of data blocks of type content The points reward per unit number of content data blocks, i.e. The number of points awarded after sharing content; Energy consumption cost Related, represented as , y is the correlation coefficient, representing S-shaped users in the same round of content sharing. The values ​​are the same, and with... As the number of failed content sharing attempts increases, y gradually decreases.

[0039] Furthermore, in step (2.3), the small base station collects information from each user and calculates the successfully matched set using a user matching algorithm, specifically including:

[0040] First, users participating in content sharing are divided into different sets according to content type. Type B users who need a certain type of content and Type S users who provide that type of content are counted separately. That is, users are divided into Type B user set and Type S user set according to h type content. Then, the points deduction or points reward of users in each set are sorted.

[0041] Suppose there are m type B users and n type S users, all targeting type h content. For the m type B users, based on... Sort the requests in descending order to obtain the sequence of requests from type B users. Then, for n S-type users, based on Sort the data in ascending order to obtain the state information sequence of S-type users. ;

[0042] Suppose that the result after the above sorting is The corresponding points deduction sequence is ,That

[0043] middle It is the largest in the sequence; The corresponding points reward sequence is ,in It is the smallest in the sequence;

[0044] There are two prerequisites for a successful match between a Type B user and a Type S user: First, the points deducted by the Type B user must be greater than or equal to the points rewarded by the Type S user; Second, the number of content data blocks that the Type S user can provide must be greater than or equal to the demand of the Type B user.

[0045] Based on the two matching prerequisites mentioned above, the last pair of users that can be successfully matched, namely type B and type S, is identified as follows: ; turn up The matching process then terminates, resulting in a set of successfully matched matches. .

[0046] Furthermore, in step (2.4), based on the successful matching set obtained in step (2.3), the final points deduction amount for type B users and the final points reward amount for type S users are determined using the pricing scheme, specifically including:

[0047] 1) Calculation Second points deduction ;

[0048] set up It is a user in the successfully matched set, specifying another user. It is its strongest competitor, if If no content request is made, then Alternative The position in the successfully matched set, and The request information is , but Represented as ; in, yes Points deduction amount;

[0049] 2) Calculation Final points deduction ;

[0050] set up It is a user in the successfully matched set, and its request information is: Matching The status information is , Final points deduction Represented as

[0051] ;

[0052] in, yes The base points deduction is equal to the matching points. The points reward per unit number of content data blocks is expressed as: ;

[0053] 3) Calculation Final points reward amount ;

[0054] The final points reward is equal to the amount matched with it. Final points deduction ,Right now Furthermore, the goal of maximizing user personal utility and social welfare in a content sharing management system is specifically:

[0055] set up Content successfully retrieved. This indicates the final points deduction amount, and the request information is as follows: , The utility function is ;

[0056] set up Match successful and sent to The quantity provided is Content data blocks, This indicates the final points reward amount. The utility function is ;

[0057] In the process of content sharing among users, the goal of designing user matching algorithms and pricing schemes is to maximize social welfare; the social welfare referred to here is the sum of the utilities of type B users and type S users.

[0058] Social welfare maximization is expressed as

[0059] ;

[0060] in, Indicating a successful match, the amount of content data blocks provided is not less than the required amount;

[0061] This indicates that the devices of all users participating in content sharing have a battery level greater than 0. This indicates that the cache size of all users participating in content sharing is greater than 0; express and The communication distance between them shall not exceed the maximum communication radius of D2D.

[0062] Beneficial Effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0063] This invention proposes a content sharing incentive method for users in an edge caching network based on power consumption awareness, and establishes a content sharing management system for... Figure 1The MEC network shown in the diagram enables content sharing among users, satisfying the following ideal properties: both Type B and Type S users meet the characteristics of individual rationality constraints; the content sharing system meets the characteristics of budget equilibrium; and Type B and Type S users meet the incentive compatibility (authenticity) requirement. This invention effectively incentivizes content sharing among users through point deduction and reward, balancing the interests of both Type B and Type S users, and maximizing social welfare as much as possible. Attached Figure Description

[0064] Figure 1 This is a diagram illustrating mobile edge caching network application scenarios;

[0065] Figure 2 This is a flowchart of the overall content sharing incentive method;

[0066] Figure 3 This is a flowchart of each round of content sharing. Detailed Implementation

[0067] The present invention will now be described in detail with reference to the accompanying drawings.

[0068] This invention discloses a content sharing incentive method for users in an edge caching network based on power consumption awareness, and establishes a content sharing management system for... Figure 1 The diagram illustrates content sharing among users in an MEC network. A flowchart of the content sharing incentive method is shown below. Figure 2 As shown, the edge caching network consists of a small base station equipped with an edge server and N mobile user equipments randomly distributed around the small base station. The small base station, acting as the system administrator, classifies users into Type B and Type S based on their requests. Type B users initiate content caching requests, while Type S users accept content caching requests and provide content data blocks. An edge caching network enabling D2D communication is established:

[0069] 1) Each user equipment has a limited device power Q and cache capacity D, and is equipped with two communication interfaces: a Wi-Fi or Bluetooth interface for direct D2D communication, and a cellular communication interface for direct connection to small base stations.

[0070] 2) User-requested content is influenced by content popularity, and its distribution follows a Zipf distribution; in the content library The probability that the content at the h-th position is requested is

[0071] ,

[0072] Where ε is the exponential constant of the Zipf distribution, This represents the number of times the content at position h was requested, and M represents the content library. Total amount of stored content, This represents the total number of times all content was requested;

[0073] 3) Within a distance d around the user, there exists The probability of each neighboring user is:

[0074] ,

[0075] Here, λ represents that the user's neighboring users follow a Poisson distribution with density λ.

[0076] In each round of content sharing in this system, multiple Type B users and multiple Type S users participate. Both Type B and Type S users have sufficient battery power, and Type B users have sufficient cache capacity to store content data blocks and enough points. The small base station will match multiple user pairs, meaning that Type B users and Type S users ultimately share content one-to-one.

[0077] In a single round of content sharing, there are m type B users and n type S users, where m + n ≤ N, and a system administrator acting as a small base station; the set of type B users is defined as follows: The S-type user set is defined as follows: Let Q and D be the current battery level and available cache capacity of any user's device, respectively. For example... Figure 3 As shown, each round of content sharing includes the following steps:

[0078] Step 1: Calculate the content valuation for Type B users and the energy cost for Type S users to transmit content. Valuation refers to the maximum number of points deducted for Type B users to acquire content, and energy cost refers to the minimum number of points awarded to Type S users to share content.

[0079] Because users play different roles, their purposes and intentions also differ. Different Type B users have different valuations of different content, and because this is also influenced by content popularity, different content data blocks also have different application values ​​on the network. Based on each... ( The probability of content being requested (∈ B), the user's demand for the content, and the ease of obtaining the content are all used to calculate the value. ( ∈ B) Estimation of content V i .

[0080] therefore, The valuation of h-type content can be expressed as: , and 0 < Vi < 100.

[0081] in, Z represents the probability that content at position h in the content repository will be requested; Z represents the average number of cached data blocks per user in the network, calculated as follows: ; represent For H-type content, the higher the demand, the higher the valuation of the content. yes Let d be the probability that there are l neighboring users within a distance of d. Since at least one neighboring user is required, l = 1 here, and d is taken as the maximum communication radius R of D2D. max That is, d = R max ; express The number of content data blocks available from surrounding neighbors reflects the ease of content acquisition; the higher the difficulty of acquisition, the higher the value of the content. α and β represent the user's preference for the urgency of content demand and the ease of content acquisition, respectively, and their values ​​are set by the system. Since all these variables are greater than 0 and less than 1, they are multiplied by 100 at the end to make... .

[0082] If an S-type user is successfully matched, transmitting content to the corresponding B-type user will earn points as a reward. Since transmitting content will consume the S-type user's device battery, the S-type user will incur certain energy costs. ( The energy cost of transmitting content of size p in (∈ S) can be expressed as: , and .

[0083] in, express Power consumption per unit size of data block transmitted. express The device's current battery level, and the following conditions must be met. That is, the power consumption required for transmission must be less than The device's current power level; η represents the parameter coefficient set by the system.

[0084] Step 2: Type B users submit their request information to the small base station, and Type S users submit their status information to the small base station;

[0085] Content sharing is discussed over a discrete finite time interval {0,1,…,T}, where users submit information in a sealed manner. At the start, type B users and type S users submit their respective information to the small base station within a specified time, and this information can only be viewed by the small base station.

[0086] for In other words, its request information can be represented as ,in for The content type required, when When it is h-type content, it corresponds to the content library. The content data block at position h; for for The demand for content of the same type is determined by the number of data blocks required. If different users require the same type of content, then their demand for data blocks of that content will also be the same. The calculation method is to divide the size of the content data block by the unit size p, for example, a block of size W. Type of content, Demand for it ; The points deduction per unit number of content data blocks, i.e. Points deducted when accessing content. Furthermore, Valuation Related, specifically means Where x ∈ [0.5,1], with an initial value of 0.5, the x value is the same for type B users in the same round of content sharing, and as... As the number of failed content retrieval attempts increases, x will gradually increase. It's important to note that no Type B user will assign a value higher than the estimated content value. The bid, that is .

[0087] Provide status information based on your own available resources and equipment: ,in for supply Number of data blocks of type content The points reward per unit number of content data blocks, i.e. The number of points awarded after sharing content. And, Energy consumption cost Related, represented as Where y ∈ [1, 1.5], with an initial value of 1.5, the y of S-type users in the same round of content sharing is the same, and as As the number of failed content sharing attempts increases, y will gradually decrease. To ensure stable personal utility, S-type users will consider the energy costs involved in content sharing. , make sure Not lower than energy consumption cost , Right now .

[0088] Step 3: The small base station collects information from each user and uses a user matching algorithm to find the set of successfully matched users, thus reasonably realizing the matching and pairing of buying and selling users.

[0089] First, users participating in content sharing are divided into different sets according to content type. Type B users who need a certain type of content and Type S users who provide that content are separately counted. That is, users are divided into Type B user sets and Type S user sets based on type h content. Then, the points deduction or points reward amount for each set is sorted to prepare for obtaining the successfully matched sets later.

[0090] Suppose there are m type B users and n type S users, all targeting type h content. For the m type B users, based on... Sort the requests in descending order to obtain the sequence of request information for type B users. Then, for n S-type users, based on Sort the data in ascending order to obtain the state information sequence of S-type users. ;

[0091] Suppose that the result after the above sorting is The corresponding points deduction sequence is ,That

[0092] middle It is the largest in the sequence. The corresponding points reward sequence is ,in It is the smallest in the sequence. There are two prerequisites for a successful match between a type B user and a type S user: First, the points deducted by the type B user is greater than or equal to the points rewarded by the type S user; second, the number of content data blocks that the type S user can provide is greater than or equal to the demand of the type B user.

[0093] Based on the two matching prerequisites mentioned above, the last pair of users that can be successfully matched, namely type B and type S, is identified as follows: ; turn up The matching process then terminates, resulting in a set of successfully matched matches. .

[0094] The time complexity of the above algorithm includes sorting and matching. The total time complexity for sorting the B-type user sequence and the S-type user sequence is O(mlog m) + O(nlog n). Since there are M popular content items in the network, the time complexity of content classification is O(M). The time complexity for matching the sorted sequences is O(mlog m) + O(nlog n). ) ,because ≤ m and ≤ n, and set , Therefore, the overall time complexity is O(mlog m) + O(nlog n).

[0095] Step 4: Based on the successful matching set obtained in Step 3, use the pricing scheme to determine the final points deduction amount for Type B users and the final points reward amount for Type S users.

[0096] After obtaining the successfully matched set, the small base station will determine the final points deduction for type B users and the final points reward for type S users through a pricing scheme. To ensure the authenticity of the system, this invention, based on step 3, uses a pricing scheme similar to Generalized Second Pricing (GSP) to determine the final points deduction for type B users and the final points reward for type S users. The specific process of this pricing scheme is as follows:

[0097] 1) Calculation Second points deduction .

[0098] set up It is a user in the successfully matched set, specifying another user. It is its strongest competitor, if If no content request is initiated, then Alternative The position in the successfully matched set, and The request information is ,but Represented as ; in, yes Points deduction amount;

[0099] 2) Calculation Final points deduction .

[0100] set up It is a user in the successfully matched set, and its request information is: Matching The status information is , Final points deduction Represented as

[0101] .

[0102] in, yes The base points deduction is equal to the matching points. The points reward per unit number of content data blocks is expressed as: .

[0103] 3) Calculation Final points reward amount .

[0104] The final points reward is equal to the amount matched with it. Final points deduction ,Right now .

[0105] In summary, the pricing scheme designed in this invention is based on the assumption that... Match successful. The second highest points deduction is that of its strongest competitor. , The base points deduction is its matching. Points reward amount per unit number of content data blocks , from and Choose the larger one as The final points reward amount.

[0106] Step 5: Type B users and Type S users share content based on the information published by the small base station.

[0107] After pricing is finalized, the small cell base station will release relevant information, including the successfully matched set and pricing results. The successfully matched set consists of type B and type S users. Finally, type B and type S users can use this information to share content data blocks between themselves.

[0108] Furthermore, the specific relationship between the user's personal utility and the goal of maximizing social welfare in the above system is as follows:

[0109] Within the MEC network, all users aim to maximize their personal utility from content sharing. Type B users seek content with minimal point deductions, while Type S users expect high point rewards for sharing content. Given this situation, assume... Content successfully retrieved, using This indicates the final points deduction amount, and the request information is... , So The utility function is expressed as .therefore, The utility function is The difference between the content's valuation and the final points deduction is multiplied by the number of content data blocks.

[0110] Similarly, assuming Match successful and sent to The number of shares is Content data blocks, using This indicates the final points reward amount. The utility function is Therefore, the utility obtained by Sj is the difference between the final points reward and the energy cost multiplied by the number of content databases.

[0111] In the content sharing process, to resolve the conflict between Type B and Type S users, maintain the stable operation of the content sharing system, and continuously attract user participation, the user matching algorithm and pricing scheme for content sharing must be designed with the goal of maximizing social welfare. Social welfare refers to the sum of the utilities of Type B and Type S users. Therefore, the social welfare maximization problem can be expressed as:

[0112]

[0113] in, Indicating a successful match, the amount of content data blocks provided is not less than the required amount;

[0114] This indicates that the devices of all users participating in content sharing have a battery level greater than 0. This indicates that the cache size of all users participating in content sharing is greater than 0; express and The communication distance between them shall not exceed the maximum communication radius of D2D.

[0115] However, both Type B and Type S users prioritize maximizing their own utility and are unwilling to sacrifice it for maximizing social welfare. Therefore, it is necessary to develop reasonable user matching algorithms and pricing schemes to incentivize users to truthfully submit information and participate in content sharing, thereby ensuring the maximization of social welfare. The following explains how the user matching algorithm and pricing scheme proposed in this invention achieve the maximization of social welfare.

[0116] The rules stipulate that all users participating in content sharing have met the constraints of the aforementioned social welfare maximization problem. Based on the successfully matched set obtained in step 3, only the first... If a user is successfully matched, and the remaining users are not successfully matched in this round of content sharing, their utility is 0, which can be represented as:

[0117]

[0118] Therefore, the problem of maximizing social welfare can be transformed into

[0119]

[0120] For each successfully matched pair of users in the set, the number of content data blocks obtained by user B is equal to the number of content data blocks shared by user S, and the points deducted from user B's content are equal to the points rewarded by user S's shared content. Therefore, taking B1 and S1 as examples, their individual utilities and the sum of their utilities can be expressed as follows:

[0121]

[0122] Therefore, the problem of maximizing social welfare can be transformed into

[0123]

[0124] As shown by the user matching algorithm in step 3, after sorting the points deductions of type B users in descending order, the first... The sum of points deducted by type B users is the largest. After sorting the points rewards of type S users in ascending order, the top... The sum of reward points for each S-type user is the minimum. At this point, the previous... The sum of the content valuations of the top B users is the largest. The sum of energy costs for all S-type users is minimized, and since B-type users have the same demand for the same type of content, a unified approach can be adopted. Therefore, The maximum sum of user utility is obtained by finding the maximum sum of user utility for this round of content sharing. Similarly, the maximum sum of user utility for other types of content can be obtained, thus providing an approximate solution for the maximum sum of user utility in the entire content sharing session, achieving maximum social welfare. Furthermore, the GSP-based pricing scheme in step 4 relatively reduces the final point deduction for each type B user and relatively increases the final point reward for each type S user, thereby incentivizing content sharing among users.

Claims

1. An electric quantity perception based edge cache network inter-user content sharing incentive method, characterized in that: The method comprises the following specific steps: (1) an edge cache network scenario is established; the edge cache network is composed of a small base station equipped with an edge server and N mobile user equipment randomly distributed around the small base station; (2) a content sharing management system is established in the edge cache network; The small base station serves as a system manager, and users are divided into B-type and S-type according to user requests, wherein the B-type users are users initiating content cache requests, and the S-type users are users accepting content cache requests and providing content data blocks; The specific steps of each round of content sharing are as follows: (2.1) the B-type users' evaluation of the content and the energy consumption cost of the S-type users in transmitting the content are calculated; (2.2) the B-type users submit their request information to the small base station, and the S-type users submit their state information to the small base station; (2.3) the small base station collects the information of each user, and obtains a matching success set through a user matching algorithm; (2.4) based on the matching success set obtained in step (2.3), a pricing scheme is used to determine the final integral deduction amount of the B-type users and the final integral reward amount of the S-type users; (2.5) the B-type users and the S-type users perform content sharing according to the information published by the small base station; (3) based on the power perception, that is, considering the energy consumption of the S-type users in transmitting data, after the B-type users obtain the content shared by the S-type users, the small base station deducts a certain amount of points from the B-type users and distributes the points to the S-type users as a reward for the sharing behavior of the S-type users; (4) the number of points is determined based on the B-type users' evaluation and the energy consumption cost of the S-type users, and through a user matching algorithm and a pricing scheme; the small base station aims to maximize social welfare, and the social welfare refers to the sum of the utilities of the B-type users and the S-type users. 2.The method of claim 1, wherein An edge cache network allowing D2D communication is established, comprising the following contents: (1.1) each user equipment has a limited device power Q and a cache capacity D, and has two communication interfaces, that is, a Wi-Fi or Bluetooth interface for directly performing D2D communication, and a cellular communication interface for directly connecting with a small base station; (1.2) User-requested content is influenced by content popularity, and its distribution follows a Zipf distribution; in the content library The probability that the content at the h-th position is requested is ， where ε is an exponent constant of Zipf distribution, denotes the number of requests for the content at the hth position, M denotes the total number of contents in the content library denotes the total number of contents stored, denotes the sum of the number of requests for all contents; (1.3) Within a distance d around the user, there exists The probability of each neighboring user is: ， Wherein, λ represents that the neighbor users around the user obey a Poisson distribution with a density λ. 3.The method of claim 2, wherein h-type content corresponds to the content library The hth position of the content data block; the content sharing management system uses a credit issued by the small base station to the user, the user obtains credit reward by completing content sharing in the system, and a certain credit needs to be deducted to obtain content from other users. The credit has a valid period of use, and is recycled by the system after expiration. 4.The method of claim 1, wherein: In step (2.1), the B-type users' evaluation of the content and the energy consumption cost of the S-type users in transmitting the content are calculated, which specifically comprises: In a content sharing, there are m B-type users and n S-type users, m + n ≤ N, and a system manager who is a small base station; The set of B-type users is defined as The set of S-type users is defined as The current power and the free buffer capacity of the device of any user are denoted as Q and D, respectively. for It is calculated based on the probability of content being requested, the user's demand for the content, and the ease of obtaining the content. Valuation of h-type content , represents , and ; where, f h is the probability that the content in the hth position of the content library is requested; Z represents the average number of cached data blocks of users in the network, which is calculated as , M is the number of contents stored in the content library, and N is the number of mobile user devices; represents the demand degree of the hth content; is the probability that there are neighbor users within the d distance of the hth content; represents the number of data blocks of the content that can be provided by the neighbor users; and α and β represent the preference of the user for the urgency of the content demand and the difficulty of the content acquisition, respectively, and the values thereof are set by the system. For , the energy cost of transmitting a content data block of size p is calculated as , where , and ; wherein, represents the power consumption of transmitting a unit size of content data block, represents the current power of the device, and needs to satisfy the condition , that is, the power consumption required for transmission is less than the current power of the device; η represents a parameter coefficient set by the system. 5.The method of claim 4, wherein: In step (2.2), the B-type users submit their request information to the small base station, and the S-type users submit their state information to the small base station; each user submits the respective information in a sealed form on a discrete finite time {0, 1, …, T}, which specifically comprises: At the beginning, the B-type users and the S-type users submit the respective information to the small base station within a specified time, and the information can only be viewed by the small base station; In step (2.3), the small base station collects the information of each user, and obtains a matching success set through a user matching algorithm, which specifically comprises: The request information is represented as ,in, for The required content type for for The demand for content data blocks; The calculation method is to divide the size of the content data block by the unit size p, where the data block size is W. Type of content, Demand for it ; The points deduction per unit number of content data blocks, i.e. Points deducted when content is accessed; Valuation Related, represented as , , It's a correlation coefficient, referring to type B users in the same round of content sharing. The values ​​are the same, and with... The increase in the number of failed content retrieval attempts Gradually increase; State information is given according to the resource and equipment conditions of the user himself: , wherein is provided with the number of data blocks of S-shaped content, is the integral reward amount of the unit number of content data blocks, that is the number of integral rewards after sharing content; is associated with the energy consumption cost , expressed as , , is the association coefficient, the S-shaped user's value of the same round of content sharing is the same, and gradually decreases with the increase of the number of failed content sharing times. ​ 6. The method of claim 5, wherein the method further comprises: determining a content sharing incentive for the user based on the content sharing information. ​ First, the users participating in content sharing are divided into different sets according to content types, and the B-type users needing certain content and the S-type users providing such content are counted respectively, that is, the users are divided into a B-type user set of h-type content and an S-type user set of h-type content according to the h-type content; then, the integral deduction amount or integral reward amount of each set of users is sorted; Let m B-type users and n S-type users all aim at h-type content, sort m B-type users in descending order according to to obtain the request information sequence of B-type users , then sort n S-type users in ascending order according to to obtain the state information sequence of S-type users ; Let the above ordering result in The corresponding sequence of integral deduction amounts is , In is the largest in the sequence; The corresponding integral reward sequence is , where is the smallest in the sequence; There are two prerequisite conditions for the successful matching of the B-type users and the S-type users: first, the integral deduction amount of the B-type users is greater than or equal to the integral reward amount of the S-type users; second, the number of content data blocks that the S-type users can provide is greater than or equal to the demand amount of the B-type users; According to the two matching prerequisites above, the last pair of successfully matched B-type user and S-type user is found, and the numbers of the two are ; the matching is terminated after finding , and the successfully matched set is obtained . 7.The method of claim 6, wherein: In step (2.4), based on the successfully matched set obtained in step (2.3), the final integral deduction amount of the B-type users and the final integral reward amount of the S-type users are determined using a pricing scheme, specifically including: 1) Calculate the second integral deduction amount ; Let be a user in the matching success set, and let be another user whose strongest competitor, if no content request is initiated, then is replaced by at the position in the matching success set, and the request information for is ; where is the integral deduction amount of 2) Calculate the final integral deduction amount ; Let is a user in the matching success set, whose request information is , the matching information of which is , the state information of which is , the final deduction amount of the score of which is represented as ； Wherein, is the basic integral deduction amount, equal to the integral reward amount of the content data block matched with it in the unit quantity, represented as ; 3) Calculate the final amount of the bonus ; The final amount of bonus points of the player is equivalent to the amount of points matched by the player The final amount of deduction points of the player i.e. . 8.The method of claim 7, wherein the method further comprises: determining a content sharing incentive for the user based on the content sharing information. The user personal utility and social welfare maximization target in the content sharing management system is specifically: Set Successfully acquired content, represents the final integral deduction amount, and the request information is , The utility function of ; Set a match is successful and provides a number of content data blocks, represents the final amount of points awarded, the utility function of ;​ In the process of content sharing among users, the user pairing algorithm and the pricing scheme are designed to achieve the maximization of social welfare; the social welfare refers to the sum of the utilities of the B-type users and the S-type users; The maximization of social welfare is represented as ； wherein, represents a user pair in which the matching is successful, and the amount of the content data block provided is not less than the demand amount; the device power of the user participating in the content sharing is greater than 0; the cache capacity of the user participating in the content sharing is greater than 0; the device power of the user participating in the content sharing is greater than 0; the communication distance between the user participating in the content sharing and the user participating in the content sharing is not more than the D2D maximum communication radius. the communication distance between the user participating in the content sharing and the user participating in the content sharing is not more than the D2D maximum communication radius.

Images