Method for training learning models of analysis function servers located at each of multiple sites, analysis function server, and program

Abstract

To provide a learning model training method for an analysis function server capable of reducing a processing load of an analysis control server at an edge site and reducing the use capacity of network resources between the edge site and a central side, an analysis function server and a program.SOLUTION: When training a learning model from network data in a data source of an edge site, a first analysis function server at the edge site determines whether or not the capacity of the network data is equal to or less than the capacity of a learning model held by itself. In a case where it is determined to be true, the network data is transmitted to a second analysis function server at a central site, and the second analysis function server is caused to train a learning model in accordance with the network data. In a case where it is determined to be false, on the other hand, the learning model is trained by the network data, and the trained learning model is transmitted to the second analysis function server at the central site.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a technology for an analysis control server arranged in a core system of a mobile communication network, and in particular to a technology for a Network Data Analytics Function (NWDAF). [Background technology]

[0002] A mobile communication network, common to both 5G (5th Generation) and 4G (4th Generation) standards, is composed of user equipment (UE), a radio access network (RAN), and a core network system.

[0003] FIG. 1 is a functional configuration diagram of a core system in the prior art.

[0004] The core system has a group of physical user plane network devices arranged on a virtualization platform of a group of logical control plane network devices. The control plane network devices accommodate a large number of user terminals via the RAN and send and receive control signals such as for establishing communications. The user plane network devices send and receive user data from user terminals via the RAN. Furthermore, in the core system, various NF (Network Function) servers are arranged to control a group of control plane network devices.

[0005] The core system includes an NWDAF (analysis control server) that acquires and analyzes a wide variety of log data from various data sources, such as communication equipment. The NWDAF receives analysis requests from NF consumers and returns the analysis results. This is standardized by 3GPP (3rd Generation Partnership Project) (registered trademark) (see, for example, Non-Patent Document 1). NWDAF introduces federated learning, which allows it to use a learning (AI) model for inference while training it with network data (see, for example, Non-Patent Document 2).

[0006] According to Figure 1, it is composed of the following NF servers: Application Function Server (NF Consumer) 10 Repository Function Server (NRF) 11 Analysis Function Server (NWDAF) 12 The NWDAF 12 has a learning model corresponding to the analysis item, executes inference processing in response to an analysis request, and returns the analysis results. For this purpose, multiple NWDAFs 12 are deployed for each type of analysis item. The NF consumer 10 is an NF server that is the source of an analysis request. In response to a discovery request from an NF consumer, the NRF 11 selects one NWDAF from among multiple NWDAFs 12 and responds. The control plane network devices and the user plane network devices serve as data sources, and serve as input data for the NWDAF to perform inference processing using a learning model.

[0007] FIG. 2 is a sequence diagram of the prior art.

[0008] (s01) Assume that the NF consumer 10 wishes to obtain an analysis result for an analysis item α. At this time, the NF consumer 10 transmits a discover request to the NRF 11 to inquire about an NWDAF that can analyze the analysis item α. It is also preferable that the discovery request includes a Preferred Supported Analytics Delay, which is a parameter for the time required to obtain the analysis result for the analysis item α (see, for example, Non-Patent Document 2).

[0009] (s02) When the NRF 11 receives a discovery request from the NF consumer 10, it selects an NWDAF 12 that can perform analysis. It returns the address of the selected NWDAF 12 to the NF consumer 10. The NWDAF 12 selected here has a "learning model" for analyzing the analysis item α requested by the NF consumer 10.

[0010] (s03) The NF consumer 10 sends an "analysis request" to the address of the NWDAF returned from the NRF 11. The analysis request includes the following information: Analysis item (Analytics ID): Specify the analysis process Analytics target period: The period for which the analysis results are requested Expected waiting time: The time to wait for a response from the analysis process

[0011] When the NWDAF 12 receives an analysis request and is unable to return the analysis result within the desired waiting time, it returns an error response to the NF consumer 10. At this time, the NWDAF 12 may set an analysis processing waiting time (revised waiting time) in the error response.

[0012] (s04) The NWDAF 12 transmits a network data request to the data source to acquire input data for analysis on the analysis item α. In response to this, the data source transmits the network data to the NWDAF 12. This "data collection time" is delayed as the processing time of the NWDAF 12.

[0013] (s05) The NWDAF 121 inputs network data into the learning model, executes inference processing, and obtains the analysis results. This "inference time" is delayed as the processing time of the NWDAF 12.

[0014] (s06) Then, the NWDAF 121 returns the analysis result to the NF consumer 10. [Prior art documents] [Non-patent literature]

[0015] [Non-Patent Document 1] 3GPP TS 23.288 “Architecture enhancements for 5G System (5GS) to support network data analytics services”, [online], [searched on March 7, 2020], Internet <URL:https: / / portal.3gpp.org / desktopmodules / Specifications / SpecificationDetails.aspx?specificationId=3579> [Non-patent document 2] Hexa-X deliverables D4.2 “AI-driven communication & computation co-design: initial solutions”, [online], [Retrieved March 7, 2023], Internet<URL: https: / / hexa-x.eu / wp-content / uploads / 2022 / 07 / Hexa-X_D4.2_v1.0.pdf> Summary of the Invention [Problem to be solved by the invention]

[0016] As mentioned above, conventionally, various NF servers in a core system are managed centrally at one site. In contrast, in recent years, 5G and 6G have adopted an architecture based on distributed computing consisting of a central site and edge sites. For this reason, multiple NWDAFs are being deployed not only at central sites but also at edge sites. Deploying an NWDAF at an edge site where NF consumers are located has the advantage of shortening the transmission delay between sending an analysis request and receiving the analysis results.

[0017] However, when an NF consumer at an edge site sends an "analysis request" to an NWDAF located within that edge site, there is a problem that analysis requests generated within the edge site are concentrated at the NWDAF at that edge site. Many NF servers at edge sites generally have limited computing resources compared to those at central sites. As a result, NWDAFs at edge sites are also prone to resource depletion. When NF consumers send analysis requests to their own edge site NWDAFs, the expected waiting time cannot be met.

[0018] In addition, various NF servers in the core system will be distributed to central and edge sites, which will increase the usage of network resources between the sites. In particular, the more geographically dispersed the central site and multiple edge sites are, the more the network resource capacity will become an issue.

[0019] Therefore, the present invention aims to provide a learning model training method for an analysis function server, an analysis function server, and a program that can reduce the processing load of the analysis control server at the edge site and reduce the network resource usage capacity between the edge site and the central site. [Means for solving the problem]

[0020] According to the present invention, there is provided a method for training a learning model of a plurality of analysis function servers that analyzes network data using a learning model for a predetermined analysis item, the method comprising: a first analytics server is located at an edge site and a second analytics server is located at a central site; The first analytics server at the edge site is a first step of determining whether or not the capacity of the network data in a data source at an edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a 21st step of transmitting the network data to a second analysis function server at a central site when the first step determines that the answer is true, and causing the second analysis function server to train a learning model using the network data; a 22nd step of training a learning model using the network data if the first step determines that the result is false, and transmitting the trained learning model to a second analysis function server at the central site; The present invention is characterized by having the following.

[0021] According to another embodiment of the learning model training method of the analysis function server of the present invention, As a prelude to step 1, if the first analysis function server at the edge site determines that its own computing resource utilization rate is higher than a predetermined threshold, it proceeds to step 21. It is also preferable.

[0022] According to another embodiment of the learning model training method of the analysis function server of the present invention, The first analysis function server at the edge site performs the first step as a preliminary step. Receives an analysis request including an analysis item and an analysis request period from a consumer server, acquires network data corresponding to the analysis request period for the analysis item from a data source at an edge site, executes inference processing using a learning model, and returns the analysis results to the consumer server. 、 Or, An analysis request including an analysis item and an analysis request period is received from a consumer server, network data corresponding to the analysis request period for the analysis item is obtained from a data source at an edge site, inference processing is performed using a learning model, and the analysis results are returned to the consumer server. If the utilization rate of its own computing resources is higher than a predetermined threshold, the process proceeds to step 21. It is also preferable.

[0023] According to another embodiment of the learning model training method of the analysis function server of the present invention, The second analysis function server at the central site accumulates and shares the learning models of the multiple first analysis function servers at the multiple edge sites. It is also preferable.

[0024] According to another embodiment of the learning model training method of the analysis function server of the present invention, The analysis function server is NWDAF (Network Data Analytics Function), The learning model of the analysis function server is trained by MTLF (Model Training logical function), The inference process of the analytical function server is performed by AnLF (Analytics logical function) It is also preferable.

[0025] According to another embodiment of the learning model training method of the analysis function server of the present invention, Further, an analysis control function server is provided at the edge site, The consumer server sends an analysis request to the analysis control function server at the edge site where it is located, The analysis control function server selects an analysis control server at the edge site and transfers the analysis request to that analysis control server. It is also preferable.

[0026] According to the present invention, an analysis function server at an edge site is capable of analyzing network data using a learning model for a predetermined analysis item and communicating with other analysis function servers at a central site, a determination means for determining whether or not the capacity of the network data in the data source of the edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a network data transmission means for transmitting the network data to another analysis function server at the central site when the first step determines that the result is true, and causing the second analysis function server to train a learning model using the network data; a learning model transmission means for training a learning model using the network data when the result is determined to be false in the first step and transmitting the trained learning model to another analysis function server at the central site; The present invention is characterized by having the following.

[0027] According to the present invention, a program for causing a computer installed in an analysis function server at an edge site to function is provided, which analyzes network data using a learning model for a predetermined analysis item and is capable of communicating with other analysis function servers at a central site, and a determination means for determining whether or not the capacity of the network data in the data source of the edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a network data transmission means for transmitting the network data to another analysis function server at the central site when the first step determines that the result is true, and causing the second analysis function server to train a learning model using the network data; a learning model transmission means for training a learning model using the network data when the result is determined to be false in the first step and transmitting the trained learning model to another analysis function server at the central site; The present invention is characterized by the fact that it makes a computer function by using the above-mentioned method. [Effects of the Invention]

[0028] The present invention's learning model training method for an analysis function server, analysis function server, and program can reduce the processing load on the analysis control server at the edge site and reduce the network resource usage capacity between the edge site and the central site. [Brief explanation of the drawings]

[0029] [Figure 1] FIG. 1 is a diagram illustrating a system configuration according to a conventional technique. [Figure 2] FIG. 1 is a sequence diagram according to the prior art. [Figure 3] 1 is a system configuration diagram according to the present invention. [Figure 4] 10 is a flowchart of an analysis function server at an edge site in the present invention. [Figure 5]FIG. 10 is a system configuration diagram further including an analysis control function server. DETAILED DESCRIPTION OF THE INVENTION

[0030] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0031] FIG. 3 is a diagram showing the system configuration according to the present invention.

[0032] The core system shown in Figure 3 is composed of multiple edge sites and one central site. The multiple edge sites and the central site may be geographically distributed or logically distributed in consideration of computing resources.

[0033] The NRF 11 can discover various NF servers required by the NF consumer 10, as in the prior art. The NF consumer 10 can inquire of the NRF 11 arranged in the edge site to which the NF consumer 10 belongs about the address of the NWDAF 12 that can perform analysis of the analysis item in question.

[0034] At each of the edge sites and the central site, a plurality of NWDAFs (analysis function servers) 12 are arranged, which perform analysis on predetermined analysis items using a learning model from network data. NWDAF12 is an existing system that has the following two functions. AnLF(Analytics Logical Function) MTLF (Model Training Logical Function) The learning model of NWDAF12 is trained by MTLF, which retrains the learning model using log data obtained from the data source and makes the learning model available to AnLF. The inference process of the NWDAF 12 is executed by the AnLF, which receives analysis requests and returns the analysis results.

[0035] 3, the NWDAF 12 at the edge site X can communicate with the NWDAF 22 at the central site. The NWDAF 12 at the edge site X may query the NRF 21 at the central site to discover the NWDAF 22 at the central site. The NWDAF 22 at the central site may also store the training models of multiple NWDAFs 12 at multiple edge sites, allowing the NWDAF 22 at the central site to build a shared training model that can be used by multiple NWDAFs 12 at edge sites.

[0036] FIG. 4 is a flowchart of the analysis function server at the edge site in the present invention.

[0037] (s03) This is the same as in Fig. 2 described above. The NWDAF 12 receives an "analysis request" from the NF consumer 10. The analysis request includes at least an analysis item and an analysis request period. It is assumed that the NWDAF 12 possesses a learning model A corresponding to the analysis item α included in the analysis request. Here, the possessed learning model holds metadata such as the last update date and time and the period during which the learned network data can be analyzed.

[0038] (s04) This is the same as in FIG. 2. The NWDAF 12 transmits a "network data request" to the data source to request network data corresponding to the analysis request period in order to acquire input data for analysis of the analysis item α. As a result, the NWDAF 12 receives the network data from the data source.

[0039] (s05) This is the same as in Fig. 2. The NWDAF 12 inputs the network data into the learning model for the analysis item α, executes inference processing, and returns the analysis result to the NF consumer 10.

[0040] <Retraining the learning model> After returning the analysis results, the NWDAF 12 preferably retrains using the network data. This allows the learning model to be updated. The learning model is also shared between the NWDAF 12 at the edge site and the NWDAF 22 at the central site. The learning model held by the NWDAF 22 at the central site can also be used at other edge sites.

[0041] (S0) First, the NWDAF 12 determines whether its own computational resource utilization rate is equal to or lower than a predetermined threshold. If it determines that this is true (computational resource utilization rate≦predetermined threshold), it proceeds to S1. On the other hand, if it determines that this is false (computational resource utilization rate>predetermined threshold), it is determined that the NWDAF 12 is in an overloaded state, and it proceeds to S21. This allows the processing load of the NWDAF 12 at the edge site to be constantly reduced. The NWDAF 12 may use the processor usage rate or the reception rate of analysis requests per unit time as the computational resource usage rate that it can measure itself.

[0042] (S1) The NWDAF 12 determines whether the capacity of the network data is equal to or less than the capacity of the learning model held by the NWDAF 12. If it determines that this is true (the capacity of the network data is equal to or less than the capacity of the learning model), it proceeds to S21. On the other hand, if it determines that this is false (the capacity of the network data is greater than the capacity of the learning model), it proceeds to S22.

[0043] (S21) Here, the NWDAF 12 at the edge site determines that it is better to have the NWDAF 22 at the central site retrain the learning model A than to retrain it itself, from the viewpoint of computational resources and network resources. To this end, the NWDAF 12 transmits network data to the NWDAF 22 at the central site. This causes the NWDAF 22 to train the learning model A using the network data. However, if the network between the edge site and the central site is congested and the transmission delay exceeds a predetermined threshold, it is also preferable to proceed to S22.

[0044] In addition, the NWDAF12 at the edge site trains its learning model using only network data acquired from the data source that exceeds the last update date and time and the analyzable period held by the learning model.

[0045] (S22) Here, the NWDAF12 at the edge site trains the learning model A using network data in terms of computational resources and network resources, and transmits the trained learning model A to the NWDAF22 at the central site. By transmitting and receiving learning models between the edge site and the central site rather than transmitting and receiving network data, the amount of traffic is reduced and the network resource capacity is reduced.

[0046] Furthermore, the NWDAF12 at the central site trains its learning model using only the network data received from the NWDAF12 at the edge site that exceeds the last update date and time and the analyzable period held by the learning model.

[0047] According to Figure 4, it is possible to realize both inference that takes into account the computational resources of the NWDAF12 at the edge site and retraining of the learning model that takes into account the network resources between the edge site and the central site.

[0048] FIG. 5 is a diagram showing the configuration of a system further including an analysis control function server.

[0049] 5, in comparison with FIG. 3, analysis control function servers (NWDAF controllers) 13 and 23 are arranged at the edge site and the central site, respectively. The NWDAF controller 13 at the edge site functions as a producer server for the NF consumer 10.

[0050] The NF consumer 10 inquires about the address of the NWDAF controller 13 from the NRF 11 of the edge site where the NF consumer 10 is located. Next, the NF consumer 10 sends an "analysis request" to the NWDAF controller 13. Thus, by arranging the NWDAF controller 13, the NRF 11 does not need to discover the NWDAF 12 according to the analysis items of the analysis request, and also does not need to manage the Supported Analytics Delay for each NWDAF 12. The NRF 11 can reduce the processing load in the management of the NWDAF 12 and direct it to the processing load for the discovery of various other NF servers.

[0051] <Management of the current computing resource utilization rate for each NWDAF 12 in the NWDAF controller> The NWDAF controller 13 may manage the current computing resource utilization rate for each NWDAF 12. The NWDAF controller 13 may receive the computing resource utilization rate regularly from a plurality of NWDAF 12 under its management. The NWDAF controller 13 searches for a plurality of NWDAF 12 that satisfy a predetermined condition from the highest current computing resource utilization rate (the worst computing resource efficiency). Here, the "predetermined condition" is assumed to be below a predetermined rate. That is, if there is only NWDAF 12 with a computing resource utilization rate higher than the predetermined rate, no search candidate is selected. On the other hand, the NWDAF with the highest computing resource utilization rate that is below the predetermined rate may be used as the search candidate.

[0052] <Selection of NWDAF 12 in the NWDAF controller> The NWDAF controller 13 selects the NWDAF 12 that satisfies the expected waiting time of the analysis request from among the NWDAF 12 that are search candidates. The NWDAF controller 13 calculates the desired waiting time based on the inference time corresponding to the time length of the analysis request period (Analytics target period), and calculates the network data acquisition time corresponding to the time length of the analysis request period. Then, the NWDAF 12 for which the sum of the inference time and the data collection time is less than or equal to the desired waiting time is used as the search candidate. Inference time + data collection time ≦ desired latency If the NWDAF 12 cannot be found, it returns an error response to the NF consumer 10. At this time, the NWDAF controller 13 sets the inference time as the revised waiting time in the error response. This allows the NF consumer 10 to send the analysis request again, taking the revised waiting time into consideration. The NWDAF controller 13 may schedule and manage, for each NWDAF 12, the time slots for inference of analysis items to be executed.

[0053] Then, the NWDAF controller 13 forwards the “analysis request” received from the NF consumer 10 to the selected NWDAF 12.

[0054] As described in detail above, the learning model training method for an analysis function server, analysis function server, and program of the present invention can reduce the processing load on the analysis control server at the edge site and reduce the network resource usage capacity between the edge site and the central site.

[0055] Furthermore, this will, for example, "reduce the processing load on core systems of mobile communication networks," which will contribute to Goal 9 of the United Nations-led Sustainable Development Goals (SDGs), which is to "build resilient infrastructure, promote sustainable industrialization and foster innovation."

[0056] With respect to the various embodiments of the present invention described above, various changes, modifications, and omissions that fall within the scope of the technical spirit and aspects of the present invention may be easily made by those skilled in the art. The above description is merely illustrative and is not intended to be limiting in any way. The present invention is limited only by the claims and their equivalents. [Explanation of symbols]

[0057] 10 NF Consumer, Application Function Server 11 NRF, Repository Function Server 12 NWDAF, Analysis Function Server 13 NWDAF Controller, Analysis and Control Function Server 21 NRF, Repository Function Server 22 NWDAF, Analysis Function Server 23 NWDAF Controller, Analysis and Control Function Server

Claims

1. A method for training a learning model of a plurality of analysis function servers that analyzes network data using a learning model for a predetermined analysis item, comprising: A first analytics function server is located at the edge site and a second analytics function server is located at the central site; The first analysis function server at the edge site a first step of determining whether or not the capacity of the network data in a data source at the edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a 21st step of transmitting the network data to a second analysis function server at a central site when the first step determines that the answer is true, and causing the second analysis function server to train a learning model using the network data; a 22nd step of training a learning model using the network data if the first step determines that the result is false, and transmitting the trained learning model to a second analysis function server at the central site; A learning model training method for an analytical function server, comprising:

2. As a prelude to step 1, if the first analysis function server at the edge site determines that its own computing resource utilization rate is higher than a predetermined threshold, it proceeds to step 21.

2. The method for training a learning model of an analysis function server according to claim 1.

3. The first analysis function server at the edge site performs the following as a preliminary step of the first step: receive an analysis request including an analysis item and an analysis request period from the consumer server, acquire network data corresponding to the analysis request period for the analysis item from a data source at the edge site, execute inference processing using a learning model, and return the analysis results to the consumer server; Or, An analysis request including an analysis item and an analysis request period is received from the consumer server, network data corresponding to the analysis request period for the analysis item is obtained from the data source at the edge site, inference processing is performed using the learning model, and the analysis results are returned to the consumer server. If the utilization rate of its own computing resources is higher than a predetermined threshold, the process proceeds to step 21.

2. The method for training a learning model of an analysis function server according to claim 1.

4. The second analysis function server at the central site accumulates and shares the learning models of the multiple first analysis function servers at the multiple edge sites.

3. The method for training a learning model of an analysis function server according to claim 1 or 2.

5. The analysis function server is NWDAF (Network Data Analytics Function), The learning model of the analysis function server is trained by MTLF (Model Training logical function), The inference process of the analytical function server is performed by AnLF (Analytics logical function).

3. The method for training a learning model of an analysis function server according to claim 1 or 2.

6. Further, an analysis control function server is provided at the edge site, The consumer server sends an analysis request to the analysis control function server at the edge site where it is located, The analysis control function server selects an analysis control server at the edge site and transfers the analysis request to that analysis control server.

3. The method for training a learning model of an analysis function server according to claim 1 or 2.

7. An analysis function server at an edge site that analyzes network data using a learning model for a predetermined analysis item and can communicate with other analysis function servers at a central site, a determination means for determining whether or not the capacity of the network data in the data source of the edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a network data transmission means for transmitting the network data to another analysis function server at the central site when the first step determines that the answer is true, and causing the second analysis function server to train a learning model using the network data; a learning model transmission means for training a learning model using the network data when the result is determined to be false in the first step and transmitting the trained learning model to another analysis function server in the central site; An analysis function server at an edge site, comprising:

8. A program that causes a computer installed in an analysis function server at an edge site to function, the computer analyzing network data using a learning model and capable of communicating with other analysis function servers at a central site, a determination means for determining whether or not the capacity of the network data in the data source of the edge site is equal to or less than the capacity of the learning model held by the edge site when training the learning model from the network data; a network data transmission means for transmitting the network data to another analysis function server at the central site when the first step determines that the answer is true, and causing the second analysis function server to train a learning model using the network data; a learning model transmission means for training a learning model using the network data when the result is determined to be false in the first step and transmitting the trained learning model to another analysis function server in the central site; and a program for an analysis function server that causes a computer to function as described above.

Images