Demand prediction device, demand prediction method, and program

Abstract

A demand prediction device (104) is configured in a central management data center (100) in a wireless communication network (1000) constituted with the central management data center (100), a plurality of edge sites (201, 202, 203) deploying a container image of a multi-access edge computing application from the central management data center, and a network (300) connecting the central management data center and each of the plurality of edge sites, wherein the demand prediction device (104) has a generation unit (1041), a prediction unit (1042), and a configuration unit (1043).

Description

Technical Field

[0001] This invention relates to a demand forecasting device, a demand forecasting method, and a program. Background Technology

[0002] Multi-access edge computing (MEC) applications (hereinafter referred to as "MEC applications") are known (Patent Documents 1, 2, and 3). Users connect to MEC applications launched at an edge location close to their own location, thereby reducing latency. Users can thus benefit from improvements such as a better user experience or smoother gameplay in latency-critical applications.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2023-17762

[0006] Patent Document 2: Japanese Patent Application Publication No. 2022-54417

[0007] Patent Document 3: Japanese Patent Application Publication No. 2019-62510 Summary of the Invention

[0008] None of the patent documents 1, 2 and 3 discloses the following configuration: predicting the demand for MEC applications for each edge location based on past utilization, and configuring the container image of the MEC application to the corresponding edge location based on the predicted demand.

[0009] One aspect of the present invention aims to predict the demand for MEC applications for each edge location based on past utilization, and to configure the container images of MEC applications to the corresponding edge locations according to the predicted demand.

[0010] To address the aforementioned issues, one aspect of the present invention relates to a demand forecasting apparatus configured in a central management data center within a wireless communication network. This wireless communication network comprises the central management data center, a plurality of edge locations from which container images of multi-access edge computing applications are deployed, and a network connecting the central management data center to each of the plurality of edge locations. The demand forecasting apparatus includes: a generation unit that generates a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; a forecasting unit that, based on the table generated by the generation unit, forecasts the demand for multi-access edge computing applications at each of the plurality of edge locations; and a configuration unit that, based on the demand for multi-access edge computing applications predicted by the forecasting unit, configures the container image of the multi-access edge computing application to the corresponding edge location.

[0011] One aspect of the present invention relates to a demand forecasting method performed by a demand forecasting device configured in a central management data center in a wireless communication network. The wireless communication network comprises the central management data center, a plurality of edge locations from which container images of multi-access edge computing applications are deployed, and a network connecting the central management data center to each of the plurality of edge locations. The demand forecasting method includes the following steps: a generation step, generating a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; a forecasting step, based on the table generated in the generation step, forecasting the demand for multi-access edge computing applications at each of the plurality of edge locations; and a configuration step, configuring a container image of a multi-access edge computing application to a corresponding edge location based on the predicted demand for the multi-access edge computing application at each of the plurality of edge locations in the forecasting step.

[0012] The demand forecasting apparatus according to various embodiments of the present invention can be implemented by a computer. In this case, by running the computer as the various parts (software elements) of the demand forecasting apparatus, the demand forecasting apparatus program of the demand forecasting apparatus and the computer-readable recording medium on which the program is recorded are also within the scope of the present invention.

[0013] The effects of the invention

[0014] According to one aspect of the present invention, the demand for MEC applications can be predicted for each edge location based on past utilization, and the container images of MEC applications can be configured to the corresponding edge locations according to the predicted demand. Attached Figure Description

[0015] Figure 1 This is a diagram illustrating an example of a wireless communication network configuration according to an embodiment of the present invention.

[0016] Figure 2 This is a diagram illustrating an example of the functional configuration of the demand forecasting device involved in the above embodiments.

[0017] Figure 3 This is a flowchart illustrating the processing steps of the demand forecasting method involved in the above-described embodiments.

[0018] Figure 4 This is a diagram illustrating a specific example of the edge position table involved in the above-described embodiment.

[0019] Figure 5 This is a diagram illustrating a specific example of the MEC application table involved in the above-described implementation.

[0020] Figure 6 This is a diagram illustrating a specific example of the MEC application startup information table involved in the above-described implementation.

[0021] Figure 7 This is a diagram illustrating a specific example of the statistical results table related to the above-described implementation method.

[0022] Figure 8 This is a diagram illustrating a specific example of the MEC application configuration table involved in the above-described implementation. Detailed Implementation

[0023] The embodiments of the present invention will be described in detail below.

[0024] (Wireless communication network)

[0025] Figure 1 This is a schematic diagram illustrating the general configuration of the wireless communication network 1000 according to this embodiment. Figure 1 As shown, the wireless communication network 1000 includes a central management data center 100, a first edge location 201, a second edge location 202, a third edge location 203, and a network 300.

[0026] Here, in Figure 1In the diagram, three edge positions are shown as edge positions of the wireless communication network 1000: first edge position 201, second edge position 202, and third edge position 203. However, the number of edge positions is not limited to three; in addition to the first edge position 201, second edge position 202, and third edge position 203, the wireless communication network 1000 has a plurality of edge positions not shown. The first edge position 201, second edge position 202, and third edge position 203 are examples of edge positions of the wireless communication network 1000. It should be noted that, in the following text, when referring to the edge positions of the wireless communication network 1000 that include the first edge position 201, second edge position 202, and third edge position 203, they are referred to as "edge position X".

[0027] (Central Management Data Center)

[0028] The centrally managed data center 100 is the core location for cloud computing services, providing large-scale data processing, storage, and advanced application services. For example... Figure 1 As shown, the central management data center 100 has a 5G core 101, a user database (DB) 102, a container image storage unit 103, and a demand forecasting device 104.

[0029] The 5G Core 101 is a server designed for efficient processing of large amounts of data and complex calculations, featuring a high-performance processor, large-capacity random access memory (RAM), and a high-speed storage system.

[0030] User DB102 and container image storage unit 103 are each composed of large-capacity storage systems such as hard disk drives (HDDs) or solid-state drives (SSDs) to store large amounts of data. For example... Figure 1 As shown, the user DB102 and the container image storage unit 103 are illustrated as separate large-capacity storage systems. However, in this embodiment, the user DB102 and the container image storage unit 103 may also be constituted by the same large-capacity storage system. Furthermore, in the following description, this embodiment will be illustrated using the example of the user DB102 and the container image storage unit 103 being constituted by separate large-capacity storage systems.

[0031] User DB102 storage will be discussed later. Figures 4-8 The table shown includes the edge location table, MEC application table, MEC application startup information table, statistical results table, and MEC application configuration table. These five tables will be described later.

[0032] The container image storage unit 103 stores container images of MEC applications that can be launched at the edge location X. Container images are stored according to the number of MEC applications that can be executed at the edge location X. That is, one container image is prepared for each MEC application. Furthermore, the container image is a package of the code, libraries, configuration files, etc., required to execute the MEC application.

[0033] Based on past utilization, the demand forecasting device 104 forecasts the demand for MEC applications from edge location X for each edge location, and configures the container images of MEC applications to the corresponding edge locations according to the forecasted demand.

[0034] (Edge position)

[0035] Edge location X is a device that performs data processing in a location close to user equipment 500, capable of providing services with low latency. Hereinafter, edge location X will be described using the first edge location 201 as an example. Furthermore, in the following description of the first edge location 201, by replacing the component number or a portion thereof "201" attached to each constituent element with "202", this description can be understood as a description of the second edge location 202. Additionally, in the description of the first edge location 201, by replacing the component number or a portion thereof "201" attached to each constituent element with "203", this description can be understood as a description of the third edge location 203.

[0036] like Figure 1 As shown, the first edge location 201 has a MEC server 2011, a mirror warehouse 2012, and a base station 2013.

[0037] The MEC Server 2011 provides the execution environment for MEC applications. The MEC Server 2011 can perform data processing and service provision at locations 500 meters close to user equipment, significantly reducing latency and improving communication efficiency. Additionally, the MEC Server 2011 can also function as a RIC (Radio Access Network Intelligent Controller).

[0038] Image Repository 2012 is a system for storing and managing container images of MEC applications. MEC Server 2011 retrieves container images of MEC applications from Image Repository 2012, builds and executes instances of those MEC applications. This building and execution enables rapid deployment and updates of MEC applications.

[0039] Base station 2013 provides a communication interface with user equipment 500. Base station 2013 enables low-latency access to user equipment 500, making efficient data processing possible. Furthermore, Figure 1Only one user equipment 500 is illustrated, but the base station 2013 can provide communication interfaces with multiple user equipments, including user equipment 500. Additionally, as... Figure 1 As shown, a base station 2013 is illustrated as a base station present at the first edge location 201. However, the number of base stations present at the first edge location 201 is not limited to one; there may be multiple base stations in addition to base station 2013. In the following description, this embodiment will be illustrated using the example of a single base station 2013 present at the first edge location 201.

[0040] Network 300 connects the central management data center 100 to the edge location X. Network 300 is, for example, the Internet, a leased line independent of the Internet, or a cellular network such as 5G, or a combination thereof. Network 300 enables efficient and secure data communication between the central management data center 100 and the edge location X.

[0041] (Demand forecasting device)

[0042] Figure 2 This is a diagram illustrating an example of the functional configuration of the demand forecasting device 104. Furthermore, Figure 2 The diagram also shows other components of the wireless communication network 1000 required to illustrate the functional configuration of the demand forecasting device 104.

[0043] like Figure 2 As shown, the demand forecasting device 104 includes a generation unit 1041, a forecasting unit 1042, and a configuration unit 1043.

[0044] The generation unit 1041 generates a table showing the past utilization status of the MEC application from edge location X. Details of this table will be explained when describing the demand forecasting method performed by the demand forecasting device 104, which will be described later.

[0045] The prediction unit 1042 predicts the requirements of the MEC application for each edge location X based on the table generated by the generation unit 1041.

[0046] The configuration unit 1043 configures the container image of the MEC application to the corresponding edge location based on the needs of the MEC application at each edge location X as predicted by the prediction unit 1042.

[0047] (Demand forecasting methods)

[0048] Next, the demand forecasting method performed by the demand forecasting device 104 will be explained. Here, we will first explain the deployment step (hereinafter referred to as the "non-predictive deployment step") of the container image of the MEC application from the central management data center 100 to the first edge location 201 when the user equipment 500 connected to the first edge location 201 issues a startup request for the MEC application. This non-predictive deployment step is the deployment step of the container image of the MEC application from the central management data center 100 to the first edge location 201 without the demand forecasting device 104 performing the aforementioned demand forecasting method.

[0049] Initially, the first edge location 201 receives a startup request for the MEC application from the user equipment 500. Next, based on the received startup request, the first edge location 201 requests the required container image for the MEC application from the container image storage unit 103 within the central management data center 100. Then, after receiving the request from the first edge location 201, the central management data center 100 retrieves the specified MEC application container image from the container image storage unit 103 and deploys it to the image repository 2012 of the first edge location 201. Next, the first edge location 201 expands the container image received from the central management data center 100 and starts an instance of the MEC application. Finally, after the container image is expanded, the MEC application is started on the first edge location 201, completing the preparation for providing the MEC application to the user equipment 500. After the MEC application starts, the first edge location 201 begins providing the MEC application to the user equipment 500.

[0050] Through the above, the container image deployment of MEC applications from the central management data center 100 to the first edge location 201 can provide MEC applications based on the startup request from user equipment 500.

[0051] In the aforementioned unpredictable deployment steps, the first edge location 201 deploys a container image of the MEC application from the central management data center 100 when the MEC application starts, expands the container image, and then starts the MEC application. Because the container image is deployed at startup, the MEC application takes time to start, resulting in a slower startup time, which may impair the user experience when starting the MEC application.

[0052] Alternatively, it is also possible to always keep MEC applications running on the first edge location 201. However, if MEC applications are always running, the resources of the first edge location 201 will become strained as the variety of MEC applications deployed on the first edge location 201 increases.

[0053] In contrast, when the demand forecasting device 104 executes the aforementioned demand forecasting method, it forecasts the demand for MEC applications from the first edge location 201 based on past utilization data, and pre-configures the container images of MEC applications to the first edge location 201 according to the forecasted demand. Since the container images of MEC applications predicted to have demand at the first edge location 201 are pre-configured at the first edge location 201, it is no longer necessary to pre-start all MEC applications at the first edge location 201, thus reducing the resources required for the first edge location 201. Furthermore, it reduces the traffic generated by deploying container images from the central management data center 100 every time a MEC application startup request is received. Moreover, when starting MEC applications already stored in the image repository 2012 at the first edge location 201, startup can be achieved quickly.

[0054] The following is for reference Figures 3-8 This section provides a detailed explanation of the demand forecasting method performed by the demand forecasting device 104.

[0055] exist Figure 3 In S101 shown, Figure 2 The generation unit 1041 shown obtains the edge location table A, the MEC application table B, and the MEC application startup information table C from the user DB102. Figure 4 This is a diagram showing a specific example of edge position table A. Figure 5 This is a diagram showing a specific example of MEC application table B. Figure 6 This is a diagram showing a specific example of MEC application startup information table C.

[0056] like Figure 4 As shown, edge location table A is a table consisting of entries "Edge ID", "Region", and "Region Characteristics". Edge location table A links the contents of each entry ("Edge ID", "Region", and "Region Characteristics"). The entry "Edge ID" is an identification number pre-assigned to each edge location X. Furthermore, in the following text, edge ID "1" is set to... Figure 1 The edge ID of the first edge position 201 shown is given. Edge ID "2" is set to... Figure 1 The edge ID at the second edge position 202 shown. Edge ID "3" is set to... Figure 1 The edge ID at the third edge position 203 is shown. Additionally, edge ID "4" is set to... Figure 1 The edge ID of edge location X is not shown. The entry "Region" is the region configured for each edge location X. The entry "Region Characteristics" is the characteristic of the region configured for each edge location X.

[0057] like Figure 5As shown, MEC Application Table B consists of entries for "MEC Application ID" and "MEC Application Name". MEC Application Table B associates the contents of the "MEC Application ID" and "MEC Application Name" entries. The "MEC Application ID" entry is an identification number pre-assigned to each MEC application capable of deploying container images from the central management data center 100. The "MEC Application Name" entry is the name of the aforementioned MEC application.

[0058] like Figure 6 As shown, the MEC application startup information table C is a table consisting of entries "Startup Date and Time", "Startup User", "Startup Edge", and "Start MEC Application". Table C associates the contents of these entries. The "Startup Date and Time" entry is the actual date and time when each MEC application capable of deploying container images from the central management data center 100 was actually started in the past. The "Startup User" entry is the user who started the MEC application in the past. The "Startup Edge" entry is the edge location X where the MEC application was actually started in the past. The "Startup MEC Application" entry is the MEC application ID of the MEC application that was actually started in the past.

[0059] The generation unit 1041 uses the edge location table A, MEC application table B, and MEC application startup information table C obtained from user DB102 to generate a statistical result table D. Figure 7 This is a graph showing a specific example of the statistical results in Table D.

[0060] like Figure 7 As shown, Statistical Results Table D consists of entries for "Startup Date and Time", "Startup User", "Startup Edge", and "Startup MEC Application". Each entry in Statistical Results Table D is related to... Figure 6 The entries in MEC application startup information table C shown are identical, so their descriptions will not be repeated.

[0061] Statistical Results Table D and Figure 6 The difference in the MEC application startup information table C shown is that duplicate MEC application startup information that meets the criteria of being launched by the same user, being launched at the same edge location X, and being launched by the same MEC application has been excluded. Specifically, in Figure 6The MEC application startup information in Table C contains MEC application startup information that includes the startup date and time "2023 / 11 / 1 10:35", the startup user "B", the startup edge "Okinawa", and the startup MEC application "Application 3". This information is duplicated with the MEC application startup information that includes the startup date and time "2023 / 11 / 1 10:38", the startup user "B", the startup edge "Okinawa", and the startup MEC application "Application 3". Therefore, in Figure 7 The duplicate was excluded in the statistical results table D shown. Statistical results table D does not contain MEC application startup information consisting of the startup date and time "2023 / 11 / 1 10:38", the startup user "B", the startup edge "Okinawa", and the startup MEC application "Application 3". The generation unit 1041 refers to... Figure 4 The edge position table A shown below and Figure 5 The MEC application table B shown above processes the deduplication described above.

[0062] As described above, the generation unit 1041 generates the MEC application launch information table C into a form that can process the demand forecast of MEC applications from edge location X, namely, the statistical results table D.

[0063] Next, in Figure 3 In S102 shown, the prediction unit 1042 processes the demand forecast for MEC applications from edge location X using various statistical methods based on the statistical result table D generated by the generation unit 1041. Additionally, the prediction unit 1042 may also consider the regional characteristics of edge location X, information and data available on the Internet, such as information on external events held in the region of edge location X, when processing the demand forecast for MEC applications.

[0064] The forecasting unit 1042 creates the MEC application configuration table E as the processing result of the above demand forecast. Figure 8 This is a diagram showing a specific example of MEC application configuration table E.

[0065] like Figure 8 As shown, the MEC application configuration table E is a table consisting of entries "Edge Region," "Configure MEC Application," and "Configuration Period." Table E links the contents of these entries. The "Edge Region" entry specifies the region where each edge location X, predicted to have MEC application demand, is configured. The "Configure MEC Application" entry specifies the MEC application predicted to have demand from edge location X. The "Configuration Period" entry specifies the period during which the MEC application is configured to each edge location X, predicted to have MEC application demand.

[0066] Finally, Figure 3In S103 shown, the configuration unit 1043 determines the configuration of the container image at edge location X according to the MEC application configuration table E created by the prediction unit 1042. Figure 8 For example, in image repository 2012 configured at the first edge location 201 in the edge region "Hokkaido", the container image of the MEC application "Movement Detection Application" will be configured during the configuration period "2023 / 11 / 1-11 / 30".

[0067] As described above, the configuration unit 1043 can configure container images based on the predicted needs of MEC applications. (Summary)

[0068] The demand forecasting apparatus according to one embodiment of the present invention is configured in a central management data center in a wireless communication network. The wireless communication network comprises the central management data center, a plurality of edge locations from which container images of multi-access edge computing applications are deployed, and a network connecting the central management data center to each of the plurality of edge locations. The demand forecasting apparatus includes: a generation unit that generates a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; a forecasting unit that, based on the table generated by the generation unit, forecasts the demand for multi-access edge computing applications at each of the plurality of edge locations; and a configuration unit that, based on the demand for multi-access edge computing applications predicted by the forecasting unit, configures the container image of the multi-access edge computing application to the corresponding edge location.

[0069] Based on the above configuration, it is possible to predict the demand for multi-access edge computing applications for each edge location based on past utilization, and configure the container images of multi-access edge computing applications to the corresponding edge locations according to the predicted demand.

[0070] The demand forecasting device involved in the second method of the present invention is as follows: in the first method described above, the table generated by the generation unit is associated with the following: the date and time when each multi-access edge computing application capable of deploying container images from the central management data center was actually launched in the past, the launching user of the multi-access edge computing application that was actually launched in the past, the edge location where the multi-access edge computing application was actually launched in the past, and the multi-access edge computing application that was actually launched in the past. The table excludes duplicate information where the duplicate information is consistent with launching by the same user, launching at the same edge location, and launching the same multi-access edge computing application.

[0071] Based on the above configuration, a table in the form of a function can be generated to process demand forecasts for multi-access edge computing applications from edge locations.

[0072] The demand forecasting method of the third embodiment of the present invention is executed by a demand forecasting device configured in a central management data center in a wireless communication network. The wireless communication network comprises the central management data center, a plurality of edge locations from which container images of multi-access edge computing applications are deployed, and a network connecting the central management data center to each of the plurality of edge locations. The demand forecasting method includes the following steps: a generation step, generating a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; a forecasting step, based on the table generated in the generation step, forecasting the demand for multi-access edge computing applications at each of the plurality of edge locations; and a configuration step, configuring the container image of the multi-access edge computing application to the corresponding edge location according to the predicted demand of the multi-access edge computing application at each of the plurality of edge locations in the forecasting step.

[0073] Based on the above configuration, the same effect as the first method described above can be obtained.

[0074] This invention is not limited to the embodiments described above. Various modifications can be made within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included within the technical scope of this invention. Furthermore, new technical features can be formed by combining the technical means disclosed in each embodiment.

[0075] Explanation of reference numerals in the attached figures

[0076] 100 Central Management Data Center

[0077] 104 Demand Forecasting Device

[0078] 201 First edge position

[0079] 202 Second edge position

[0080] 203 Third edge position

[0081] 300 Network

[0082] 1000 wireless communication network

[0083] 1041 Generation Department

[0084] 1042 Forecasting Department

[0085] 1043 Configuration Department

Claims

1. A demand forecasting apparatus configured in a central management data center within a wireless communication network, the wireless communication network comprising the central management data center, a plurality of edge locations, and a network, wherein the plurality of edge locations deploy container images of multi-access edge computing applications from the central management data center, and the network connects the central management data center to each of the plurality of edge locations, wherein, The demand forecasting device has: The generation unit generates a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; The prediction unit, based on a table generated by the generation unit, predicts the multi-access edge computing application requirements for each of the plurality of edge locations; and The configuration unit configures the container image of the multi-access edge computing application to the corresponding edge location based on the requirements of the multi-access edge computing application at each of the plurality of edge locations predicted by the prediction unit.

2. The demand forecasting device according to claim 1, wherein, The table generated by the generation department is a table that establishes a relationship between the following: the date and time when each multi-access edge computing application capable of deploying container images from the central management data center was actually launched in the past; the user who launched the multi-access edge computing application in the past; the edge location where the multi-access edge computing application was actually launched in the past; and the multi-access edge computing application that was actually launched in the past. The table excludes duplicates of the following information: launches performed by the same user, launches at the same edge location, and launches of the same multi-access edge computing application.

3. A demand forecasting method, performed by a demand forecasting device configured in a centrally managed data center within a wireless communication network, the wireless communication network comprising the centrally managed data center, a plurality of edge locations, and a network, wherein the plurality of edge locations deploy container images of multi-access edge computing applications from the centrally managed data center, and the network connects the centrally managed data center to each of the plurality of edge locations, wherein, The demand forecasting method includes the following steps: The generation step involves generating a table representing the past utilization of multi-access edge computing applications from each of the plurality of edge locations; The prediction step, based on the table generated in the generation step, predicts the demand for multi-access edge computing applications at each of the plurality of edge locations. as well as The configuration step involves configuring the container image of the multi-access edge computing application to the corresponding edge location based on the needs of the multi-access edge computing application predicted in the prediction step for each of the plurality of edge locations.

4. A program for enabling a computer to function as a demand forecasting device as described in claims 1 and 2, wherein, The program is used to enable the computer to function as the generation unit, the prediction unit, and the configuration unit.

Images