An edge computing resource sharing method, system and device based on coalition game

Abstract

The present invention discloses a method, system and device for edge computing resource sharing based on coalition game, which includes that an edge server collects all device nodes and their task information in the network to form an edge computing network; setting a coalition structure and establishing an initial single-device coalition structure C i , where i ∈ 1, 2, …, n; modeling the utility function of the device nodes; setting the coalition game and the preference operation relationship, and selecting an improved coalition structure C j , and calculating the coalition revenue V(C j ); obtaining the stable coalition structure set C * , until no coalition structure changes, stop the iteration; reaching the stable coalition structure set C * , obtaining the overall system offloading strategy, and the game reaches equilibrium. It can maximize the task completion rate of all device nodes, solve the problem that all tasks are submitted to the edge server resulting in task blocking and processing failure, relieve the computing pressure caused by its resource limitation, and at the same time enable all device nodes to obtain the maximum benefit, improving the overall efficiency and benefit of the system as well as the satisfaction of all parties.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to an edge computing resource sharing method, system, and device based on coalition game theory. Background Technology

[0002] In the era of the Internet of Things (IoT), the explosive growth in the number of IoT devices has generated massive amounts of data. Centralized resource processing centers like cloud computing can no longer efficiently meet the processing demands of this massive data. One reason is that the exponential growth rate of data far exceeds the growth rate of cloud server computing resources. Another reason is that the simultaneous uploading of large amounts of data to cloud centers puts enormous pressure on network transmission, leading to a sharp increase in bandwidth load and high network latency, which cannot meet the requirements of time-sensitive processing tasks. Therefore, edge big data processing, centered on the edge computing model and targeting the massive data generated by network edge devices, has emerged. Edge computing solves problems such as high network load, high latency, and insufficient bandwidth. It can transform edge devices with abundant computing resources into edge clouds, allowing resource-constrained terminal devices to offload tasks to edge cloud servers via wireless networks, achieving a distributed management model of cloud-edge-device.

[0003] Since edge server resources are often limited, they need to be allocated rationally. A method is needed to determine whether a task should be offloaded to a remote server, or to which edge server or central cloud, and ultimately how much computing resources should be allocated. Therefore, a reasonable computational offloading scheme and resource allocation scheme are particularly important. Currently, game theory-based resource allocation is widely used in edge computing. Game theory generally refers to the process by which multiple rational participants, with their own information and under mutual constraints, choose a strategy that maximizes their own interests. Many studies model the interaction between edge servers equipped with data processing centers and terminal devices with limited computing resources in edge systems as cooperative or non-cooperative games. Through a series of game decisions or more reasonable incentive mechanisms, the edge system can reach a Nash equilibrium point where all parties can obtain greater benefits.

[0004] However, with the ever-increasing volume of data and the improved computing power of edge devices, edge servers are gradually becoming unable to meet the demands of massive data volumes. Edge devices with certain computing capabilities can then serve as collaborative resources with servers to handle offloading tasks from other edge devices. Therefore, we consider establishing alliances among devices within the same edge computing network. Devices can communicate within their own communication range via D2D connections to offload tasks to other devices within the alliance, thereby alleviating MEC pressure. However, the different task requirements and distribution of each device node directly affect the formation of the alliance. Therefore, how to establish alliances based on device task conditions and select appropriate offloading strategies to achieve effective resource allocation and maximize task completion to meet the computing needs of each device is a problem that urgently needs to be solved. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] In view of the problems existing in the above-mentioned edge computing resource sharing methods based on alliance game theory, this invention is proposed.

[0007] Therefore, the purpose of this invention is to provide an edge computing resource sharing method based on alliance game theory.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an edge server collects information on all device nodes and their tasks in the network to form an edge computing network;

[0009] Set up the alliance structure and establish the initial single-device alliance structure. ;

[0010] Model the utility function of the device node;

[0011] Given a coalition game and preference calculation relationships, for an initial coalition structure, iterate through the device nodes in one of the coalitions. In the device node Selecting the improved alliance structure within the scope of D2D communication And calculate the alliance revenue. ;

[0012] The above steps are executed by traversing all device nodes in a predefined order. Task assignment is determined based on maximizing task completion rate and meeting time constraints, further building a stable alliance structure. To obtain a stable set of alliance structures The iteration continues until no changes occur in the alliance structure;

[0013] To achieve the set of stable alliance structures The overall system unloading strategy is obtained, and the game reaches equilibrium.

[0014] As a preferred embodiment of the edge computing resource sharing method based on alliance game theory described in this invention, the basic information and task information of the device nodes are collected and directly reported to the edge server through communication with the cellular network of the base station.

[0015] The basic information of the device node includes the CPU frequency f of the device node, and the task information includes the task computation amount W and the maximum tolerable time of the task. .

[0016] As a preferred embodiment of the edge computing resource sharing method based on alliance game theory described in this invention, wherein: initially, the device node itself acts as a single-device alliance, obtaining an initial alliance structure set. ;

[0017] Among them, the single-device alliance structure , 1≤i≤n, represents the i-th alliance structure;

[0018] Among them, the set of devices in the edge network system is defined. The device node The computational power representation of ) .

[0019] As a preferred embodiment of the edge computing resource sharing method based on coalition game theory described in this invention, the utility function is expressed as:

[0020]

[0021] in, For utility function, For the device node The utility that can be obtained from a unit of computational task; , For the device node The kth task is directed to the device node. The amount of unloaded tasks;

[0022] like Then the device node To the device The uninstallation relationship is valid if it is true, and vice versa.

[0023] As a preferred embodiment of the edge computing resource sharing method based on coalition game theory described in this invention, wherein: the coalition game theory is defined as follows: ;

[0024] in, The utility of a representative alliance is defined in Real-valued mappings on,

[0025] The preference operation is set as ;

[0026] like This indicates the device node. More willing to join the improved alliance structure Instead of the initial single-device alliance structure ,Right now

[0027]

[0028] in, Representing the current alliance The matching degree of tasks generated by all device nodes within the system;

[0029]

[0030] in, For the alliance Any device node within. Describe its task distribution status. Calculate the function for the matched value.

[0031] As a preferred embodiment of the edge computing resource sharing method based on coalition game theory described in this invention, wherein: the coalition revenue The formula is as follows:

[0032] .

[0033] As a preferred embodiment of the edge computing resource sharing method based on coalition game theory described in this invention, the following steps are taken: all device nodes in the coalition are traversed in a predetermined order, iterating continuously, and the coalition structure is improved based on a higher task distribution matching value within the coalition, until the game converges, thus obtaining the overall benefit of the edge network. , ;

[0034] And based on the constraint of maximizing overall benefits, a stable alliance structure is constructed. The constraints are as follows:

[0035]

[0036] st

[0037]

[0038]

[0039] All of the above constraints must be applied.

[0040] An edge computing resource sharing system based on coalition game theory, wherein:

[0041] The signal acquisition module and edge server collect information on all device nodes and their tasks in the network to form an edge computing network.

[0042] The module constructs a consortium structure, establishes an initial single-device consortium structure, models the utility function of the device nodes, and defines the consortium game and preference computation relationships; and...

[0043] The running module iterates through all device nodes in a predefined order, executing the above steps to further build a stable alliance structure. To obtain a stable set of alliance structures The iteration continues until no changes occur in the alliance structure;

[0044] To achieve the set of stable alliance structures The overall system unloading strategy is obtained, and the game reaches equilibrium.

[0045] A computer device, wherein:

[0046] It includes a memory and a processor, wherein the memory stores computer programs.

[0047] The characteristic is that the processor, when executing the computer program, implements the steps of the method described above.

[0048] A computer-readable storage medium, wherein:

[0049] It stores a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the method described above.

[0050] The beneficial effects of this invention are as follows: Based on the overall offloading strategy, the task completion rate of all device nodes is maximized, which solves the problem of task blocking and failure caused by all tasks being submitted to the edge server. It also alleviates the computing pressure caused by the limited resources of the edge server, and at the same time enables all device nodes to obtain the maximum benefit, thereby improving the overall efficiency and benefits of the system and the satisfaction of all parties. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0052] Figure 1 This is a diagram of the D2D offloading network model of the edge computing resource sharing method based on coalition game theory described in this invention.

[0053] Figure 2 This is a flowchart of the task unloading method of the edge computing resource sharing method based on alliance game theory of the present invention.

[0054] Figure 3 This is a flowchart illustrating the process of achieving a stable structure, i.e., Nash equilibrium, in the edge computing resource sharing method based on coalition games, as described in this invention.

[0055] Figure 4 This is a comparison chart showing the changes in overall system effectiveness and number of devices under different unloading schemes.

[0056] Figure 5 This is a graph showing the service time variation trend of four unloading schemes under different numbers of devices. Detailed Implementation

[0057] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0058] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0059] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0060] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0061] Example 1

[0062] Reference Figures 1-3 An edge computing resource sharing method based on coalition game theory includes:

[0063] S1: Edge servers collect information on all device nodes and their tasks in the network, forming an edge computing network. It should be noted that:

[0064] The basic information and task information of the device nodes are collected by communicating with the cellular network of the base station and reporting directly to the edge server.

[0065] The basic information of the device node includes the CPU frequency f, and the task information includes the task computation load W and the maximum tolerable time of the task. .

[0066] S2: Set up the alliance structure and establish the initial single-device alliance structure. It should be noted that:

[0067] Define the set of devices in an edge network system Among them, device nodes The computing power is expressed as (cycles / s).

[0068] Let set Indicates the k-th device The resulting set of tasks Represents a set of tasks The j-th task in , This represents the total amount of computing resources required to complete the task. This indicates the maximum delay requirement for the task.

[0069] Define Alliance Structure ,for , , Initially, the edge devices themselves form a single-device alliance, obtaining an initial alliance structure set. .

[0070] in , 1≤i≤n, represents the i-th alliance structure.

[0071] S3: The utility function for modeling device nodes. It should be noted that:

[0072] Based on device nodes For example, among which utility function Represented as:

[0073]

[0074] in, For utility function, For device nodes The utility that can be obtained from a unit of computational task; , For device nodes The kth task is directed to the device node. The amount of unloaded tasks;

[0075] like Then the device node To the equipment The uninstallation relationship is valid if it is true, and vice versa.

[0076] S4: Define the alliance game and preference calculation relationships. For the initial alliance structure, iterate through the device nodes in one of the alliances. At the device node Selecting an improved alliance structure within the scope of D2D communication And calculate the alliance revenue. It should be noted that:

[0077] Setting up alliance game .

[0078] in, The utility of a representative alliance is defined in Real-valued mappings on,

[0079] Set preference operation as .

[0080] like This indicates the device node. More willing to join improved alliance structures Instead of the initial single-device alliance structure ,Right now

[0081]

[0082] in, Representing the current alliance The matching degree of tasks generated by all device nodes within the system.

[0083]

[0084] in, For the alliance Any device node within. Describe its task distribution status. Calculate the function for the matched value.

[0085] Based on device nodes For example:

[0086] A1: The edge server selects a device node from a certain consortium structure. Find device nodes Within the scope of D2D communication, other alliances can be selected to join another alliance.

[0087] A2: For each optional alliance, the task distribution matching value and the corresponding offloading relationship among all devices within the new alliance structure can be calculated. .

[0088] A3: For device nodes The matching value q obtained by unloading the task to each optional alliance is compared, the maximum matching value is taken, and the device node utility value and device node value at this time are obtained. The corresponding uninstallation relationships.

[0089] A4: Based on the obtained unloading relationships, adjust and improve the alliance structure, and sum the unloading task utilities of all devices within the alliance to obtain the benefits of the alliance structure. .

[0090] Furthermore, alliance revenue It can be obtained from the following formula:

[0091]

[0092] S5: Traverse all device nodes in the set order and execute the above steps. Determine task assignment based on maximizing task completion rate and time limit requirements, and further build a stable alliance structure. To obtain a stable set of alliance structures The iteration continues until no changes occur in the alliance structure, at which point the iteration stops; a stable set of alliance structures is reached. The overall system unloading strategy is obtained, and the game reaches equilibrium. It should be noted that:

[0093] The process iterates through all device nodes in the alliance in a predetermined order, continuously improving the alliance structure based on higher task distribution matching values ​​within the alliance, until the game converges, yielding the overall benefit of the edge network. , And based on the constraint of maximizing overall benefits, a stable alliance structure is constructed. .

[0094] The constraints are as follows, and all of the following constraints must be satisfied:

[0095]

[0096] st

[0097]

[0098]

[0099] Example 2

[0100] This embodiment is the second embodiment of the present invention. Unlike the first embodiment, this embodiment provides a verification test of an edge computing resource sharing method based on alliance game theory, and verifies and explains the technical effects used in this method.

[0101] In this embodiment, the previous unified uninstallation based on cloud servers, task uninstallation based on non-cooperative game theory, and random uninstallation methods will be tested and compared with the resource sharing and task uninstallation scheme based on alliance game theory proposed in this paper.

[0102] Traditional offloading methods do not effectively achieve system load balancing and maximize overall utility. In contrast, this solution offers higher device utility and better performance.

[0103] Test environment: The system simulation was conducted using the EdgeCloudSim platform, and simulation data was obtained based on the experimental results. The number of devices was set to 100, 200, 300, 400, 500, and 600 respectively. Six sets of tests were conducted for each uninstallation scheme, and each set of tests was performed three times. The average value was taken.

[0104] The device node offloading tasks were randomly generated based on a Poisson distribution, with an arrival interval of 3 seconds. The device nodes were distributed using a nomadic model. The WLAN bandwidth was set to 300 Mbps and the WAN latency to 100 ms. The experimental results are as follows: Figure 4 and Figure 5 As shown.

[0105] Figure 4 The relationship between the overall system utility and the number of devices under different unloading schemes was compared. It can be seen that when the number of devices is low initially, all three algorithms show good performance. As the number of devices continues to increase, the utility growth shows a negative correlation with the failure rate. That is, the higher the device unloading task completion rate, the higher the device utility is obtained, and it gradually reaches the threshold.

[0106] Similarly, among the three algorithms, the alliance matching algorithm proposed in this paper shows the best performance and achieves the highest device utility value.

[0107] Figure 5 The data shows the service time trends of the four offloading schemes for different numbers of devices.

[0108] It can be seen that when the number of devices is small, the cloud processing method has strong processing power, so even if the transmission latency is slightly longer, the overall time is relatively shorter. When the number of devices is large, the consortium algorithm shows more advantages and achieves the best results among the four algorithms.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for edge computing resource sharing based on coalition game theory, characterized in that: include, Edge servers collect information on all device nodes and their tasks in the network to form an edge computing network; Set up the alliance structure and establish the initial single-device alliance structure. ; Model the utility function of the device node; Given a coalition game and preference calculation relationships, for an initial coalition structure, iterate through the device nodes in one of the coalitions. In the device node Selecting the improved alliance structure within the scope of D2D communication And calculate the alliance revenue. ; The above steps are executed by traversing all device nodes in a predefined order. Task assignment is determined based on maximizing task completion rate and meeting time constraints, further building a stable alliance structure. To obtain a stable set of alliance structures The iteration continues until no changes occur in the alliance structure; To achieve the set of stable alliance structures The overall system unloading strategy is obtained, and the game reaches equilibrium. The utility function is expressed as: ； in, For utility function, For the device node The utility that can be obtained from a unit of computational task; , For the device node The kth task is directed to the device node. The amount of unloaded tasks; like Then the device node To the device The uninstallation relationship is valid if it is true, and vice versa. The aforementioned coalition game ; in, The utility of a representative alliance is defined in Real-valued mappings on, The preference operation is set as ; like This indicates the device node. More willing to join the improved alliance structure Instead of an iterative alliance ,Right now ； in, Representing the current alliance The matching degree of tasks generated by all device nodes within the system; ； in, For the alliance Any device node within. Describe its task distribution status. Calculate the function for the matched value.

2. The edge computing resource sharing method based on coalition game theory as described in claim 1, characterized in that: The collection of basic information and task information of the device nodes is achieved by communicating with the cellular network of the base station and reporting directly to the edge server. The basic information of the device node includes the CPU frequency f of the device node, and the task information includes the task computation amount W and the maximum tolerable time of the task. .

3. The edge computing resource sharing method based on coalition game theory as described in claim 1, characterized in that: Initially, the device node itself acts as a single-device alliance, obtaining an initial alliance structure set. ; Among them, the single-device alliance structure , 1≤i≤n, represents the i-th alliance structure; Among them, the set of devices in the edge network system is defined. The device node The computational power representation of ) .

4. The edge computing resource sharing method based on coalition game theory as described in claim 1, characterized in that: The alliance benefits The formula is as follows: 。 5. The edge computing resource sharing method based on coalition game theory as described in claim 4, characterized in that: The process iterates through all device nodes in the alliance in a predetermined order, continuously improving the alliance structure based on higher task distribution matching values ​​within the alliance, until the game converges, yielding the overall benefit of the edge network. , ; And based on the constraint of maximizing overall benefits, a stable alliance structure is constructed. The constraints are as follows: ; ； s.t. ; ； ； All of the above constraints must be applied.

6. An edge computing resource sharing system based on coalition game theory, using the method described in any one of claims 1-5, characterized in that: The signal acquisition module and edge server collect information on all device nodes and their tasks in the network to form an edge computing network. The module is constructed, the alliance structure is set, and an initial single-device alliance structure is established. The utility function of the device node is modeled, and the alliance game and preference operation relationship are set. as well as, The running module iterates through all device nodes in a predefined order, executing the above steps to further build a stable alliance structure. To obtain a stable set of alliance structures The iteration continues until no changes occur in the alliance structure; To achieve the set of stable alliance structures The overall system unloading strategy is obtained, and the game reaches equilibrium.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, Its features are, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, The When a computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images