JP7646065B1 - COMMUNICATION CONTROL DEVICE AND COMMUNICATION CONTROL METHOD - Google Patents

Abstract

The present invention aims to suppress communication delays among multiple user terminals using a simpler configuration.
SOLUTION
The communication control device 1 includes a calculation unit 13 that provides information about multiple base stations 3 in which each of the multiple user terminals 2 is located as unknown input to a trained machine learning model, performs calculations on the trained machine learning model, and outputs network setting information regarding a destination cloud 5 among multiple clouds 5 that are final data processing nodes to which the multiple user terminals 2 can connect, which suppresses delays in communication with the multiple user terminals 2, and a communication control unit 14 that instructs the core network 4 to set up a data communication path between each of the multiple user terminals 2 and the destination cloud 5 based on the network setting information output by the calculation unit 13.
[Selected Figure] Figure 1

Description

The present invention relates to a communication control device and a communication control method.

Multi-access Edge Computing (MEC) is a method for providing cloud computing functions in locations physically close to user terminals in mobile communication networks (see Patent Document 1).

In the technology described in Patent Document 1, a user plane function (UPF) of the core network is set for each user terminal to suppress delays in data communication. Therefore, even when multiple user terminals connect to a cloud that provides the same website, or when multiple user terminals communicate with each other, edge computing that takes into account the overall transmission time delay of multiple user terminals has not been performed.

JP 2018-160813 A

Conventional technology has not been able to reduce communication delays between multiple user terminals with a simpler configuration.

The present invention has been made to solve the above-mentioned problems, and aims to reduce communication delays between multiple user terminals using a simpler configuration.

In order to solve the above-mentioned problems, the communication control device according to the present invention includes a calculation unit configured to provide information on a plurality of base stations in which each of a plurality of user terminals is located as an unknown input to a trained machine learning model, perform calculations on the trained machine learning model, and output network setting information on a destination data processing device that suppresses delays in communication with the plurality of user terminals among a plurality of data processing devices that are final data processing nodes to which the plurality of user terminals can be connected, and a communication control unit configured to instruct a core network to set up a data communication path between each of the plurality of user terminals and the destination data processing device based on the network setting information output by the calculation unit.

In addition, the communication control device according to the present invention further includes a learning unit configured to learn, using a machine learning model, the relationship between the combination of the plurality of base stations in which each of the plurality of user terminals is located and the network setting information regarding the connection destination data processing device among the plurality of data processing devices to which the plurality of user terminals can be connected, which suppresses delays in the communication with the plurality of user terminals, and a storage unit configured to store the learned machine learning model constructed by the learning unit, and the calculation unit may read out the learned machine learning model stored in the storage unit and perform calculations.

In addition, the communication control device according to the present invention may further include an acquisition unit configured to acquire information about the base stations in which each of the plurality of user terminals is located, based on a location registration request signal transmitted by each of the plurality of user terminals when the each of the plurality of user terminals is located in the range of the plurality of base stations, and the calculation unit may use the information about the plurality of base stations acquired by the acquisition unit as the unknown input.

In addition, in the communication control device according to the present invention, the location registration request signal is associated with a group identifier of a group to which the multiple user terminals belong, the acquisition unit acquires the group identifier associated with the location registration request signal in association with information on the multiple base stations, the storage unit stores the group identifier in association with the trained machine learning model, and the communication control unit may specify the multiple user terminals having the group identifier and instruct the core network to set up the data communication path between each of the multiple user terminals and the connected data processing device.

In addition, in the communication control device according to the present invention, the network setting information may include information on the destination data processing device that is located at a shorter physical distance from the base stations among the plurality of data processing devices, and a common user plane function for data communication of each of the plurality of user terminals may be set by the information on the destination data processing device.

In order to solve the above-mentioned problems, the communication control method according to the present invention includes a calculation step of providing information on a plurality of base stations in which each of a plurality of user terminals is located as an unknown input to a trained machine learning model, performing calculations on the trained machine learning model, and outputting network setting information on a destination data processing device that suppresses delays in communication with the plurality of user terminals, among a plurality of data processing devices that are final data processing nodes to which the plurality of user terminals can be connected, and a communication control step of instructing a core network to set up a data communication path between each of the plurality of user terminals and the destination data processing device, based on the network setting information output in the calculation step.

In addition, the communication control method according to the present invention further includes a learning step of learning, using a machine learning model, the relationship between the combination of the plurality of base stations in which each of the plurality of user terminals is located and the network setting information related to the connection destination data processing device among the plurality of data processing devices to which the plurality of user terminals can be connected, which suppresses delays in the communication with the plurality of user terminals, and a storage step of storing the trained machine learning model constructed in the learning step in a storage unit, and the calculation step may read out the trained machine learning model stored in the storage unit and perform calculations.

In addition, the communication control method according to the present invention further includes an acquisition step of acquiring information about the base stations in which each of the plurality of user terminals is located, based on a location registration request signal transmitted by each of the plurality of user terminals when the each of the plurality of user terminals is located in the range of the plurality of base stations, and the calculation step may use the information about the plurality of base stations acquired in the acquisition step as the unknown input.

In addition, in the communication control method according to the present invention, the location registration request signal is associated with a group identifier of a group to which the multiple user terminals belong, the acquisition step associates and acquires the group identifier associated with the location registration request signal with information on the multiple base stations, the storage step associates the group identifier with the trained machine learning model and stores it in the storage unit, and the communication control step may specify the multiple user terminals having the group identifier and instruct the core network to set up the data communication path between each of the multiple user terminals and the connected data processing device.

In addition, in the communication control method according to the present invention, the network setting information may include information on a destination data processing device that is located at a shorter physical distance from the base station among the plurality of data processing devices, and a common user plane function for data communication of each of the plurality of user terminals may be set by the information on the destination data processing device.

According to the present invention, information on multiple base stations in which multiple user terminals are located is provided as unknown input to a trained machine learning model, and the trained machine learning model is operated to output network setting information related to a destination data processing device that suppresses communication delays with multiple user terminals among multiple data processing devices that are final data processing nodes to which multiple user terminals can be connected. Therefore, communication delays between multiple user terminals can be suppressed with a simpler configuration.

FIG. 1 is a block diagram showing a configuration of a communication control system including a communication control device according to an embodiment of the present invention. FIG. 2 is a diagram for explaining the configuration of the first storage unit of the communication control device according to the present embodiment. FIG. 3 is a diagram for explaining the learning process performed by the learning unit included in the communication control device according to the present embodiment. FIG. 4 is a block diagram showing a hardware configuration of the communication control device according to the present embodiment. FIG. 5 is a sequence diagram for explaining the operation of a communication control system including a communication control device according to the present embodiment. FIG. 6 is a flowchart for explaining the operation of the communication control device according to the present embodiment. FIG. 7 is a flowchart for explaining the operation of the communication control device according to the present embodiment.

A preferred embodiment of the present invention will be described in detail below with reference to Figures 1 to 7.

[Configuration of communication control system]
First, with reference to FIG. 1, an overview of a communication control system including a communication control device 1 according to an embodiment of the present invention will be described.

The communication control system according to this embodiment includes a communication control device 1, a user terminal 2, a base station 3, a core network 4, and a cloud (data processing device) 5. The communication control system is provided in a 5G mobile communication network. The communication control system predicts a specific cloud 5 that further suppresses communication delays among multiple clouds 5 to which each of multiple user terminals 2 can connect via a base station 3 in which the user terminals 2 are located. Furthermore, the communication control system sets an optimal UPF 44 for connecting to the specific cloud 5 for each of the multiple user terminals 2. The communication control device 1 is connected to the core network 4 via a network NW.

The user terminal 2 is realized as a mobile communication terminal such as a smartphone, a tablet computer, a laptop computer, or the like. The user terminal 2 is equipped with a SIM, and the contract profile of the SIM stores the user's subscriber identification information, including identifier information such as the subscriber identification number (IMSI: International Mobile Subscriber Identity) assigned to the mobile phone line contract, the telephone number of the subscriber user (MSISDN: Mobile Subscriber International Subscriber Directory Number), and the SIM card number (ICCID: Integrated Circuit Card Identifier). The user terminal 2 is uniquely identified by the IMSI.

The user terminal 2 is also assigned a terminal IP address that uniquely identifies the terminal. The IP address is assigned to the user terminal 2 through the SMF 42 after the session is established. In this embodiment, there are N user terminals 2 (N is a positive integer equal to or greater than 2). The user terminals 2-1, 2-2, ..., 2-N are each present in the communication area of a different base station 3. The user terminals 2 access a specific cloud 5 from a specified UPF 44 via the base station 3 in which they are each present. In this embodiment, the multiple user terminals 2 are grouped in advance based on a predetermined attribute, and a group ID (group identifier) is assigned. The predetermined attribute can be determined based on any criteria, such as multiple user terminals 2 managed by the same organization or QoS requirements.

When the user terminal 2 moves within a communication area, as a periodic location update, or when the power is turned on, the user terminal 2 transmits a location registration request signal to the core network 4 via the base station 3. The location registration request signal transmitted by the user terminal 2 includes the IMSI.

The base station 3 is composed of a wireless base station compatible with the 5G system, and relays communication between the user terminal 2 present in the communication area and the core network 4. The base station 3 is connected to the core network 4 via a network such as a backhaul link. N base stations 3 (N is an integer equal to or greater than 2) are provided. Hereinafter, when the base stations 3-1, 3-2, ..., 3-N are not distinguished from one another, they are collectively referred to as "base station 3". The base station 3 is uniquely identified by a base station ID. The base station ID also makes it possible to ascertain the geographical location of the base station 3. That is, the base station ID is associated with location information consisting of the latitude, longitude, and altitude of the location where the base station 3 is located, and is stored in the auxiliary storage device 105 of the communication control device 1 described below.

The core network 4 is connected to the communication control device 1 via a network NW such as a LAN, WAN, or the Internet. The core network 4 includes an AMF (Access and Mobility Management Function) 40, a UDM (Unified Data Management)/UDR (Unified Data Repository) 41, a SMF (Session Management Function) 42, and a PCF (Policy Control Function) 43, which are nodes in the C-plane. The core network 4 also includes multiple UPFs (User Plane Functions) 44 in the U-plane. Functional nodes in the U-plane and C-plane other than those mentioned above that are provided in the core network 4 are not shown in the figure.

AMF40 is an access and mobility management device that manages the registration and wireless connections of user terminals 2 that move into each communication area.

The UDM/UDR 41 manages subscriber profiles, performs authentication, and manages mobility. In this embodiment, the UDM/UDR 41 stores a group ID assigned to each IMSI in the subscriber profile. The UDM/UDR 41 has a communication interface 41a for communicating with the communication control device 1.

In addition, in response to an instruction from the communication control device 1, the UDM/UDR 41 creates and stores a setting information table T1 that associates the IMSIs, group IDs, base station IDs, communication paths of the UPF 44 set in data communication of the user terminals 2, and the connected cloud 5, as shown in FIG. 2. The instruction from the communication control device 1 includes grouping information that associates multiple IMSIs and group IDs of control targets that have been specified in advance. The value of the IP address of the UPF 44 in the setting information table T1 is registered after the communication paths are set by the SMF 42 and PCF 43 described below. The IP address of the cloud 5 is registered when a predicted value is obtained by the calculation unit 13 of the communication control device 1 described below.

The UDM/UDR 41 can configure the setting information table T1 by adding a group ID field to the subscriber profile. The setting information table T1 managed by the UDM/UDR 41 is also synchronized with the communication control device 1, and the same contents are stored (second storage unit 15). Note that, in this embodiment, the UDM/UDR 41 is exemplified as a device in which the UDM and UDR are configured as a single device, but the UDM/UDR 41 may also be a device in which the UDM and UDR are arranged separately.

The SMF42 is a session management function that establishes, modifies, and releases PDU (Packet Data Unit) sessions between the user terminal 2 and a data network such as the Internet. The SMF42 sets an appropriate communication path for data communication between the user terminal 2 and the UPF44 based on a PCC (Policy and Charging Control) policy from the PCF43.

The PCF 43 determines the QoS and policies and provides them to the SMF 42. The PCF 43 applies PCC rules based on the 3GPP (registered trademark) specifications, and creates a PCC policy for setting the communication path of the UPF 44 with which the user terminal 2 communicates in response to instructions from the communication control device 1.

The UPF 44 is a user plane function that processes packets between the base station 3 and a data network such as the Internet. The UPF 44 functions as a gateway between the core network 4 and an external data network. A plurality of UPFs 44 are provided in the core network 4. As shown in FIG. 1, each of the UPFs 44_1 to UPF_N is directly connected to each of the base stations 3-1, 3-2, ..., 3-N, so-called full mesh connection. In this embodiment, each UPF 44 transmits packets from the user terminal 2 to the cloud 5 via a data network such as the Internet. Also, each UPF 44 forwards packets transmitted from the cloud 5 to the user terminal 2. As shown in FIG. 1, each UPF 44 has one cloud 5 as a connection destination. In this embodiment, multiple user terminals 2 belonging to the same group ID communicate with any of the optimal UPFs 44_1 to UPF_N designated according to the cloud 5 related to the specific base to which they are connected. The UPF 44 has an IP address, which allows the UPF 44 to be uniquely identified.

Cloud 5 provides a specific website, web application, etc. Cloud 5 is a cloud base that is a geographical location or area where physical devices that make up a website, etc. are located. Cloud 5 can be an edge server such as a server, a data center, or an MEC server. In this embodiment, multiple user terminals 2 having the same group ID can be connected to multiple clouds 5 that are data final processing nodes that provide the same website, etc.

The multiple clouds 5 are distributed and located at locations geographically distant from each other. Furthermore, the distance from each base station 3 to each of the multiple clouds 5 is different for each cloud 5. For example, the cloud 5 physically closest to base station 3-1 is "cloud 1-1", and the cloud 5 farthest is "cloud 1-N". Multiple user terminals 2 access the same cloud 5 from the base station 3 in which they are located via the same designated UPF 44. The cloud 5 can be uniquely identified by its IP address, and location information is associated with the IP address and stored in the auxiliary storage device 105 of the communication control device 1 described below.

[Functional block of communication control device]
1, the communication control device 1 includes an acquisition unit 10, a learning unit 11, a first storage unit 12 (storage unit), a calculation unit 13, a communication control unit 14, and a second storage unit 15. The communication control device 1 learns the relationship between a combination of a plurality of base stations 3 in which a plurality of user terminals 2 are located, and network setting information related to a connection destination cloud 5 that suppresses delays in data communication with the plurality of user terminals 2.

The acquisition unit 10 acquires information on the multiple base stations 3 in which each of the multiple user terminals 2 is located, based on a location registration request signal transmitted by each of the multiple user terminals 2 when it is within the range of the multiple base stations 3. The acquisition unit 10 also acquires a group ID associated with the location registration request signal transmitted by the multiple user terminals 2 and information on the multiple base stations 3 in association with each other. A base station ID is used as the information on the base station 3. When the location registration request signal transmitted by the user terminal 2 passes through the core network 4, the base station ID is added to the location registration request signal in the core network 4. As described above, the location registration request signal is transmitted when the user terminal 2 crosses the communication area of each base station 3 or at regular intervals.

The base station ID acquired by the acquisition unit 10 indicates information on the base station 3 in which multiple user terminals 2 with the same group ID are located during the same time period. The base station ID for each user terminal 2 acquired by the acquisition unit 10 is used as an unknown input when the calculation unit 13 described below performs calculations on a trained machine learning model. In addition, the base station ID acquired by the acquisition unit 10 can be used as part of the teacher data when the learning unit 11 learns the machine learning model.

The learning unit 11 uses a machine learning model to learn the relationship between the combination of the multiple base stations 3 in which each of the multiple user terminals 2 is located and the network setting information for the connection destination cloud 5 that suppresses delays in communication with the multiple user terminals 2, among the multiple clouds 5 that are final data processing nodes to which the multiple user terminals 2 can connect. The learning unit 11 can learn the machine learning model by supervised learning.

The communication delay is particularly the transmission delay of the multiple user terminals 2 as a whole in data communication between the multiple user terminals 2. In other words, the transmission delay of data communication on a group ID basis is the target of suppression. The transmission delay includes delay caused by the endpoint being physically far away when viewed from the multiple base stations 3 in which the multiple user terminals 2 are located. Specifically, the transmission delay in data transfer in the user plane from the multiple base stations 3 to the cloud 5 can be suppressed. In this case, the transmission delay of data communication for the entire group is suppressed by connecting to the cloud 5 that is the shortest or optimal physical distance overall from each of the base stations 3 in which the multiple user terminals 2 are located.

The network setting information includes information on the cloud 5 that is located at a shorter physical distance from the base stations 3 among the clouds 5 that are located at different geographical locations for each of the user terminals 2. That is, the network setting information identifies the cloud 5 to connect to, and the information on the identified cloud 5 makes it possible to set an optimal UPF 44 for data communication of each of the user terminals 2. For example, by identifying cloud 1 as the connection destination among clouds 1 to N, the optimal UPF_1 for connecting to cloud 1 is identified.

Specifically, the network setting information can be a DNN (Data Network Name) identifier that can identify the information of the cloud 5 to which the connection is made and the information of the UPF 44 to which the routing is made. Note that when optimizing the physical distance from multiple base stations 3 to a specific cloud 5, the distance from a certain user terminal 2 is not necessarily the shortest distance. A cloud 5 that is closer to the multiple base stations 3, that is, the optimal distance, is identified by the network setting information.

Figure 3 shows the structure of a neural network model adopted as an example of a machine learning model used by the learning unit 11. The neural network model has an input layer x, a hidden layer h, and an output layer y. The learning unit 11 provides the base station IDs of the base stations 3 in which the multiple user terminals 2 (IMSI) having the same group ID are located to the input layer of the neural network model, applies an activation function to the weighted sum of the inputs, and passes the output determined by threshold processing to the output layer. The output node of the output layer outputs a predicted output of the model of network setting information regarding the cloud 5 to which the multiple user terminals 2 connect for the combination of the base station IDs of the base stations 3 in which the multiple user terminals 2 are located.

By introducing an objective function, the learning unit 11 learns the parameters of the neural network model so that the predicted value of the network setting information from the neural network model for a combination of base stations 3 in which multiple user terminals 2 are located becomes the value of the network setting information of the correct label. The learning unit 11 adjusts the weight parameters of the neural network related to the machine learning model so that the objective function is minimized, that is, becomes 0. The learning unit 11 can optimize the objective function using a backpropagation method or the like.

Specifically, the learning unit 11 can train the neural network to predict the cloud 5 to be connected to as a class classification, and output the cloud 5 with the correct label with a high probability by taking the optimal cloud 5 as the correct answer. In this case, cross entropy loss is used as the objective function. Alternatively, the learning unit 11 can introduce an objective function that minimizes the squared error, which is the distance error of the cloud 5 corresponding to the multiple base station IDs, as a regression problem. When treating it as a regression problem, learning is performed to minimize the squared error between the actual distance between the positions of the multiple base stations 3 and the position of the cloud 5 and the predicted value of the distance output by the model, using the actual distance as the correct label.

The first storage unit 12 stores the trained machine learning model constructed by the training unit 11. The first storage unit 12 stores the trained machine learning model in association with a group ID.

The calculation unit 13 provides information on the multiple base stations 3 in which each of the multiple user terminals 2 is located as unknown input to the trained machine learning model, performs calculations on the trained machine learning model, and outputs network setting information regarding a destination cloud 5 that reduces communication delays with the multiple user terminals 2, among multiple clouds 5 of the data final processing node to which the multiple user terminals 2 can connect.

The calculation unit 13 uses the base station IDs of the multiple base stations 3 that are added to the location registration request signals transmitted by the multiple user terminals 2 and acquired by the acquisition unit 10 as unknown inputs. That is, the base station IDs for each IMSI of the same group ID are provided as unknown inputs to the trained machine learning model corresponding to the group ID stored in the first storage unit 12. The calculation unit 13 can output the identifier of the DNN as the network setting information. The IP address of the destination cloud 5 indicated by the network setting information output by the calculation unit 13 is registered in the setting information table T1.

The communication control unit 14 instructs the core network 4 to set a data communication path between each of the multiple user terminals 2 and the destination cloud 5 based on the network setting information output by the calculation unit 13. More specifically, the communication control unit 14 specifies the IMSIs of the multiple user terminals 2 having the same group ID, and instructs the core network 4 to set a UPF 44 common to each of the multiple user terminals 2. When the calculation unit 13 outputs a DNN identifier as the network setting information, it can acquire the identification information of the destination cloud 5 associated with the output DNN identifier by referring to the information stored in the auxiliary storage device 105.

The communication control unit 14 specifies the cloud 5 to connect to, identified by the DNN identifier, for the IMSI of each user terminal 2 of the same group ID, and instructs the setting of the UPF 44 of each user terminal 2. In detail, the communication control unit 14 refers to table T1 shown in FIG. 2 in response to receiving a location registration request signal from the user terminal 2 via the core network 4. Note that the location registration request signal includes the IMSI of the user terminal 2, as described above.

The communication control unit 14 requests the PCF 43 of the core network 4 to create a PCC policy, specifying the group ID, IMSI, IP address of the communication control device 1, and the cloud 5 to which the connection is to be made. In response to the creation request, the PCF 43, SMF 42, and UDM/UDR 41 of the core network 4 work together to set an appropriate communication path of the UPF 44 for data communication of the user terminal 2. When the communication path of the UPF 44 is set for data communication of the user terminal 2, the IP address of the UPF 44 related to the setting of the communication path is stored for each IMSI in table T1 of the second storage unit 15, as described in FIG. 2.

The second storage unit 15 stores a setting information table T1 that associates the IMSIs, group IDs, base station IDs, communication paths of the UPF 44 set in data communication of the user terminals 2, and the destination cloud 5 of the connection with each other. In the setting information table T1, the identification information of the destination cloud 5 ("cloud IP address") is information that is specified and stored by the identifier of the DNN output from the calculation unit 13. In addition, the identification information of the UPF 44 ("UPF IP address") is information that is stored in response to the UPF 44 being set in accordance with an instruction from the core network 4 of the communication control unit 14.

[Hardware configuration of the communication control device]
Next, an example of a hardware configuration for realizing the communication control device 1 having the above-mentioned functions will be described with reference to FIG.

As shown in FIG. 4, the communication control device 1 can be realized by, for example, a computer including a processor 102, a main memory device 103, a communication interface 104, an auxiliary memory device 105, and an input/output I/O 106 connected via a bus 101, and a program that controls these hardware resources. Furthermore, the communication control device 1 can include a display device 107 connected via the bus 101.

The processor 102 is realized by a CPU, GPU, FPGA, ASIC, etc.

The main memory device 103 stores in advance programs that allow the processor 102 to perform various controls and calculations. The processor 102 and the main memory device 103 realize the various functions of the communication control device 1, such as the acquisition unit 10, learning unit 11, calculation unit 13, and communication control unit 14 shown in FIG. 1.

The communication interface 104 is an interface circuit for network connection between the communication control device 1 and various external electronic devices.

The auxiliary storage device 105 is composed of a readable/writable storage medium and a drive for reading and writing various information such as programs and data from the storage medium. The auxiliary storage device 105 can use semiconductor memory such as a hard disk or flash memory as the storage medium.

The auxiliary storage device 105 has a program storage area for storing the communication control program executed by the communication control device 1. The auxiliary storage device 105 also has a program storage area for storing the machine learning program executed by the communication control device 1. The auxiliary storage device 105 also has an area for storing information in which the DNN identifier is associated with the identification information of the cloud 5 and the identification information of the UPF 44. The auxiliary storage device 105 also has an area for storing grouping information in which the IMSI of a plurality of user terminals 2 is associated with a group ID. The auxiliary storage device 105 also has an area for storing the base station ID, the identification information of the UPF 44, and the identification information of the cloud 5 in association with their respective location information. The auxiliary storage device 105 realizes the first storage unit 12 and the second storage unit 15 described in FIG. 1. In addition, for example, it may have a backup area for backing up the above-mentioned data, programs, etc.

The input/output I/O 106 is an input/output device that inputs signals from external devices and outputs signals to external devices.

The display device 107 is configured with an organic EL display, a liquid crystal display, or the like. The display device 107 can display the setting information of the cloud 5 and the UPF 44 connected to each group ID.

[Operation of communication control system]
Next, the operation of the communication control system including the communication control device 1 having the above-mentioned configuration will be described with reference to the sequence of Fig. 5. Fig. 5 shows the processing of the communication control system related to inference after the learning unit 11 of the communication control device 1 has constructed a trained machine learning model.

First, the communication control device 1 transmits grouping information associated with a group ID and an IMSI to the UDM/UDR 41 (step S100). The grouping information is stored in advance in the communication control device 1. Next, the UDM/UDR 41 creates a setting information table T1 (FIG. 2) based on the received grouping information (step S101). In step S101, the values of "group ID" and "IMSI" in the setting information table T1 are registered, and there are no entries yet for the values of the other items. The setting information table T1 is also stored in the second storage unit 15 of the communication control device 1.

After that, the multiple user terminals 2 of group ID "1" transmit location registration request signals to the core network 4 via the base stations 3 in which they are located (step S102). The location registration request signal includes the IMSI of the user terminal 2 and the base station ID of the base station 3 in which they are located. When the location registration request signal is received by the UDM/UDR 41 via the AMF 40, the UDM/UDR 41 refers to the grouping information in the setting information table T1 (Figure 2) created in step S101, associates the group ID with the base station ID and IMSI included in the received location registration request signal, and further transfers the location registration request signal to the communication control device 1 (step S103). Note that in step S103, the value of the base station ID in the setting information table T1 is set.

Next, the calculation unit 13 of the communication control device 1 reads out from the first storage unit 12 the trained machine learning model associated with the group ID received in step S103, and performs calculation processing (step S104). In step S104, the base station ID of the IMSI of the same group ID acquired by the acquisition unit 10 of the communication control device 1 in step S103 is used as an unknown input, and calculations are performed on the trained machine learning model. The calculation processing in step S104 outputs optimal network setting information, i.e., the identifier of the DNN. The IP address of the destination cloud 5 identified by the identifier of the DNN is registered in the setting information table T1 (Figure 2).

Then, the communication control unit 14 of the communication control device 1 requests the PCF 43 to create a PCC policy by specifying the IP address of the cloud 5 associated with the identifier of the DNN output in step S104, as well as the group ID, IMSI, and IP address of the communication control device 1 (step S105). In the example of FIG. 5, cloud_2 is specified. Next, the PCF 43 creates a PCC policy based on the specified requirements (step S106). The PCF 43 creates a PCC policy that includes information about which UPF 44 the data communication of the user terminal 2 will pass through. The PCC policy specifies the optimal UPF 44 as a communication path for data communication from the user terminal 2 of the same group ID to cloud_2.

Next, PCF43 transmits the created PCC policy to SMF42 (step S107). After that, SMF42 transmits data communication path setting information related to the user plane function, such as the setting of the communication path specified in the PCC policy, to UPF44 specified by the PCC policy (step S108). In the example of FIG. 5, UPF_2 is set. Next, UPF44 (UPF_2) registers the received data communication path setting information in memory (step S109).

Next, UPF44 (UPF_2) transmits an ACK to SMF42 to notify it of the IP address of UPF_2 (step S110). Furthermore, SMF42 transmits an ACK to the communication control device 1 to notify it of the IP address of UPF_2 (step S111). After that, the communication control device 1 notifies UDM/UDR41 of the IP address of UPF_2 and transmits an ACK to the user terminal 2 (step S112). At that time, the IP address of UPF_2 communicating with the user terminal 2 is set in the setting information table T1 (FIG. 2). In addition, the setting information table T1 updated by UDM/UDR41 is also stored in the second storage unit 15 of the communication control device 1. After that, a communication path for data communication between the user terminal 2, UPF44 (UPF_2), and cloud 5 (cloud_2) is established (step S113). The processes from step S102 to step S113 are also performed for other user terminals 2 with the same group ID "1".

By the above process, multiple user terminals 2 with the same group ID can connect to the same cloud_2 and perform data communication via a transfer route that minimizes overall data communication delays.

Next, the operation of the communication control device 1 having the above-mentioned configuration will be described with reference to the flowcharts of Figures 6 and 7. Figure 6 is a flowchart showing the learning process by the communication control device 1. Figure 7 is a flowchart showing the calculation process by the communication control device 1.

As shown in FIG. 6, first, the learning unit 11 prepares training data (step S1). Specifically, the base station IDs of the base stations 3 in which the respective user terminals 2 having the same group ID are located, which are acquired by the acquisition unit 10, can be used as training data. A DNN identifier is given in advance as a correct label to a set of combinations of base station IDs for each IMSI. Alternatively, the identification information of the cloud 5 may be given as the correct label.

Next, the learning unit 11 uses the teacher data prepared in step S1 to learn a machine learning model (step S2). More specifically, the learning unit 11 uses the machine learning model to learn the relationship between the combination of base station IDs of the multiple base stations 3 in which each of the multiple user terminals 2 is located and the network setting information related to the destination cloud 5 that suppresses delays in data communication with the multiple user terminals 2.

Then, the first storage unit 12 associates the group ID with the trained machine learning model constructed in step S2 and stores it (step S3).

Next, as shown in FIG. 7, the acquisition unit 10 acquires the base station IDs of the base stations 3 in which multiple user terminals 2 with a common group ID are located (step S10). Step S10 corresponds to step S103 described in FIG. 5. That is, in step S10, the base station IDs of the base stations 3 in which each user terminal 2 is located are acquired in response to location registration request signals transmitted by multiple user terminals 2 with the same group ID.

Next, the calculation unit 13 reads out the trained machine learning model corresponding to the group ID acquired in step S10 from the first storage unit 12, performs calculations on the trained machine learning model using the combination of base station IDs acquired in step S10 as unknown inputs, and outputs network setting information on the destination cloud 5 that suppresses transmission delays in data communication for the entire group (step S11). In step S11, an identifier of a DNN indicating a cloud 5 that is closer physically to the multiple base stations 3 is output. Step S11 corresponds to step S104 described in FIG. 5.

Then, the communication control unit 14 designates each of the multiple user terminals 2 having the same group ID based on the network setting information output in step S11, and instructs the core network 4 to configure the UPF 44 for the multiple user terminals 2 (step S12). Step S12 corresponds to step S105 described in FIG. 5.

As described above, according to the communication control device 1 of this embodiment, the combination of base stations 3 in which the multiple user terminals 2 are located and the network setting information regarding the connection destination cloud 5 that can suppress transmission delays in data communication with the multiple user terminals 2 are learned by a machine learning model, and the learned network setting information is stored in a database in advance. Therefore, communication delays between the multiple user terminals 2 can be suppressed with a simpler configuration.

Furthermore, according to the communication control device 1 of this embodiment, multiple user terminals 2 are grouped, and the optimal connection destination cloud 5 is selected for each group. Therefore, by selecting a cloud 5 that is physically close as the connection destination for each group of multiple user terminals 2, it is possible to more efficiently suppress transmission delays in data communication for the entire group with a simpler configuration.

In addition, according to the communication control device 1 of this embodiment, a location registration request signal transmitted by a base station 3 in which multiple user terminals 2 are located is used as a trigger to acquire a base station ID, calculate a trained machine learning model, and request the creation of a PCC policy. Therefore, with a simpler configuration, it is possible to suppress communication delays between multiple user terminals 2.

In the embodiment described above, the machine learning model used by the learning unit 11 is exemplified as a multi-layer neural network. However, the machine learning model may be a decision tree-based model such as a multi-layer perceptron or random forest, or a gradient boosting-based decision tree such as XGBoost or LightGBM, k-nearest neighbors, support vector machine, or linear regression.

In the embodiment described above, the base station ID of the base station 3 in which multiple user terminals 2 with the same group ID are located is input to the input layer of the multi-layer neural network used by the learning unit 11. However, if a model can be constructed that outputs information about the connected cloud 5 to suppress transmission delays in data communication across multiple user terminals 2, the feature values input to the input layer can be not only the base station ID for each IMSI, but also values such as the number of hops in the route taking into account logical distance, network delays such as the number of relay nodes and switching processing time, and QoS requirements such as bandwidth and reliability.

The above describes the embodiments of the communication control device and communication control method of the present invention, but the present invention is not limited to the described embodiments, and various modifications that a person skilled in the art can imagine are possible within the scope of the invention described in the claims.

1...communication control device, 10...acquisition unit, 11...learning unit, 12...first memory unit, 13...calculation unit, 14...communication control unit, 15...second memory unit, 2...user terminal, 3...base station, 4...core network, 40...AMF, 41...UDM/UDR, 42...SMF, 43...PCF, 44...UPF, 5...cloud, 101...bus, 102...processor, 103...main memory device, 41a, 104...communication interface, 105...auxiliary memory device, 106...input/output I/O, 107...display device, NW...network.

Claims (10)

a calculation unit configured to provide information on a plurality of base stations in which each of a plurality of user terminals is located as an unknown input to a trained machine learning model, perform calculations on the trained machine learning model, and output network setting information related to a destination data processing device that suppresses a delay in communication with the plurality of user terminals, among a plurality of data processing devices that are data final processing nodes to which the plurality of user terminals can be connected;
and a communication control unit configured to instruct a core network to set a data communication path between each of the plurality of user terminals and the connected data processing device based on the network setting information output by the calculation unit. 2. The communication control device according to claim 1,
a learning unit configured to learn, using a machine learning model, a relationship between a combination of the plurality of base stations in which each of the plurality of user terminals is located and the network setting information related to a data processing device that is a connection destination among the plurality of data processing devices to which the plurality of user terminals can be connected, the data processing device suppressing a delay in the communication with the plurality of user terminals;
A memory unit configured to store the trained machine learning model constructed by the training unit,
The communication control device, wherein the calculation unit reads out the trained machine learning model stored in the storage unit and performs calculations. 3. The communication control device according to claim 2,
Further, an acquisition unit configured to acquire information of the plurality of base stations in which each of the plurality of user terminals is located, based on a location registration request signal transmitted by each of the plurality of user terminals when the each of the plurality of user terminals is located in the plurality of base stations,
The communication control device, wherein the calculation unit uses the information of the plurality of base stations acquired by the acquisition unit as the unknown input. 4. The communication control device according to claim 3,
When the location registration request signal transmitted by each of the plurality of user terminals when the each of the plurality of user terminals is within the range of the plurality of base stations is transmitted to the communication control device via the core network, the core network associates a group identifier of a group to which the plurality of user terminals belong with a subscriber identification number included in the location registration request signal ,
the acquiring unit acquires the group identifier associated with the subscriber identification number included in the location registration request signal and information of the plurality of base stations in association with each other;
The storage unit stores the trained machine learning model in association with the group identifier,
The communication control device is characterized in that the communication control unit specifies the multiple user terminals having the group identifier and instructs the core network to set up the data communication path between each of the multiple user terminals and the connected data processing device. 2. The communication control device according to claim 1,
The network setting information includes information of a connection destination data processing device that is located at a position where a physical distance from the base stations is shorter among the plurality of data processing devices, for each of the plurality of user terminals, and
A communication control device comprising: a communication control unit configured to set a common user plane function for data communication of each of the plurality of user terminals based on information about the connected data processing device. a calculation step of providing information on a plurality of base stations in which each of a plurality of user terminals is located as an unknown input to a trained machine learning model, performing calculations on the trained machine learning model, and outputting network setting information related to a destination data processing device that suppresses a delay in communication with the plurality of user terminals, among a plurality of data processing devices that are data final processing nodes to which the plurality of user terminals can be connected;
and a communication control step of instructing a core network to set up a data communication path between each of the plurality of user terminals and the connected data processing device based on the network setting information output in the calculation step. 7. The communication control method according to claim 6,
Further, a learning step of learning, using a machine learning model, a relationship between a combination of the plurality of base stations in which each of the plurality of user terminals is located and the network setting information related to a data processing device that is a connection destination among the plurality of data processing devices to which the plurality of user terminals can be connected, the data processing device suppressing a delay in the communication with the plurality of user terminals;
A storage step of storing the trained machine learning model constructed in the learning step in a storage unit,
The communication control method, wherein the calculation step reads out the trained machine learning model stored in the storage unit and performs a calculation. 8. The communication control method according to claim 7,
Further, the method includes the step of acquiring information on the plurality of base stations in which each of the plurality of user terminals is located, based on a location registration request signal transmitted by each of the plurality of user terminals when the each of the plurality of user terminals is located in the plurality of base stations,
The communication control method according to claim 1, wherein the calculation step uses information about the plurality of base stations acquired in the acquisition step as the unknown input. 9. The communication control method according to claim 8,
when the location registration request signal transmitted by each of the plurality of user terminals when it is within the range of the plurality of base stations is transmitted to a communication control device that executes the communication control method via the core network, the core network associates a group identifier to which the plurality of user terminals belong with a subscriber identification number included in the location registration request signal ;
the acquiring step acquires the group identifier associated with the subscriber identification number included in the location registration request signal and information of the plurality of base stations in association with each other;
The storing step includes storing the trained machine learning model in the storage unit in association with the group identifier;
The communication control method is characterized in that the communication control step specifies the plurality of user terminals having the group identifier and instructs the core network to set up the data communication path between each of the plurality of user terminals and the connected data processing device. 7. The communication control method according to claim 6,
The network setting information includes information of a connection destination data processing device that is located at a position where a physical distance from the base stations is shorter among the plurality of data processing devices, for each of the plurality of user terminals, and
A communication control method, comprising: a common user plane function for data communication of each of the plurality of user terminals can be set based on information about the connected data processing device.