Network Slice Management Device and Management Method

Abstract

A network slice management device according to one aspect is a network slice management device that manages network slices in a cellular network. The network slice management device has a control unit. The control unit inputs packet information indicating the operation of packets of an application in a cellular network using a learned AI model for application identification and outputs an application name. Further, the control unit inputs the application name using a learned AI model for communication control and outputs communication control information indicating information for communication control of the packets in the cellular network. Furthermore, when receiving a packet including the name of a new application, the control unit retrains at least the learned AI model for application identification and the learned AI model for communication control using at least the name of the new application.

Description

Technical Field

[0001] The present invention relates to a network slice management apparatus and a management method.

Background Art

[0002] In a mobile communication system including a cellular network, network slicing has been introduced. Network slicing is a technology for creating a plurality of virtual networks by virtually dividing a physical network constructed by an operator. Each virtual network is called a network slice. By means of network slicing, a communication operator can create network slices according to service requirements of different service types such as enhanced Mobile Broadband (eMBB), Ultra-Reliable and Low Latency Communications (URLLC), and massive Machine Type Communications (mMTC). By means of network slicing, for example, optimization of network resources can be achieved.

[0003] On the other hand, a user may execute various applications using a terminal device such as a smartphone. In particular, these days, there are also use cases for executing new applications that have not been seen in the past, such as remote medical treatment or robot control.

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Non-Patent Document 2

Summary of the Invention

[0005] A network slice management device according to one aspect is a network slice management device that manages network slices in a cellular network. The network slice management device has a control unit. The control unit inputs packet information indicating the operation of packets of an application in the cellular network using a learned AI model for application identification and outputs the application name. Further, the control unit inputs the application name using a learned AI model for communication control and outputs communication control information indicating information for communication control of the packets in the cellular network. Furthermore, when receiving a packet including the name of a new application, the control unit re-learns at least the learned AI model for communication control and the learned AI model for application identification using at least the name of the new application.

[0006] Also, a management method according to one aspect is a management method in a network slice management device that manages network slices in a cellular network. The management method includes a step of inputting packet information indicating the operation of packets of an application in the cellular network using a learned AI model for application identification and outputting the application name. The management method also includes a step of inputting the application name using a learned AI model for communication control and outputting communication control information indicating information for communication control of the packets in the cellular network. Furthermore, the management method includes a step of re-learning at least the learned AI model for application identification and the learned AI model for communication control using at least the name of the new application when receiving a packet including the name of the new application.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

DETAILED DESCRIPTION OF THE INVENTION

[0008] [First Embodiment] As described above, network slicing has been introduced in mobile communication systems. On the other hand, in mobile communication systems, new applications that have not been seen before, such as telemedicine or robot control, may be executed. For example, in telemedicine, not only high-speed and large-capacity communication but also low-latency communication is required. Also, for example, in robot control, both low-latency communication and multi-connection communication are required.

[0009] An operator who operates the infrastructure of a mobile communication system manually determines for each application which network slice to apply among high-speed and large-capacity (eMBB), ultra-low latency (URLLC), and multi-connection (mMTC) according to the communication requirements for each application.

[0010] However, in the current situation where new applications are expected to increase in the future, it may take a huge amount of time and labor for the operator to manually determine which network slice to apply for each application.

[0011] Here, a network slice is, for example, each virtual network when a physical network constructed by an operator in a cellular network is virtually divided. A network slice may be a logical network in a cellular network and a network capable of dynamically allocating resources in a cellular network. Alternatively, a network slice may be a network obtained by virtually dividing a cellular network according to the service provided to a user.

[0012] A network slice includes a RAN slice, a transport slice, and a core slice. The RAN slice provides, for example, resource control and priority control for each network slice in a RAN (Radio Access Network) that performs radio access control. The RAN slice may be called a RAN slice subnet. Also, the transport slice provides network slice functions for each network, such as a fronthaul (the network between the RU (Radio Unit) and the DU (Distributed Unit) when the RAN is separated into the RU, the DU, and the CU (Central Unit)), a midhaul (the network between the DU and the CU), and a backhaul (the network between the CU and the core network (CN)). The transport slice may be called a transport slice subnet. Furthermore, the core slice provides core network functions, for example, in each network slice. The core slice may be called a core slice subnet.

[0013] The network slice manually applied by the operator is intended to be applied for each application to the entire network slice including the RAN slice, the transport slice, and the core slice.

[0014] Therefore, in the first embodiment, the aim is to perform optimal communication control settings within the network.

[0015] Hereinafter, the first embodiment will be specifically described with reference to the drawings. In the description of the drawings, the same or similar parts are denoted by the same or similar reference numerals.

[0016] (1) Configuration example of the communication system First, a configuration example of the communication system according to the first embodiment will be described.

[0017] FIG. 1 is a diagram showing a configuration example of the communication system 1.

[0018] As shown in FIG. 1, the communication system 1 includes a user equipment (UE) 100, a node 110, a core network (CN) device 120, and a network slice management device 130. The communication system 1 may include a cellular network 10. The cellular network 10 includes the UE 100, the node 110, the CN device 120, and the network slice management device 130.

[0019] The cellular network 10 is a network capable of wireless communication with the movable UE 100. The cellular network 10 is also a network to which a mobile communication system of 3GPP standards is applied. The cellular network 10 is, for example, a network compliant with the 5th generation system (5GS) of 3GPP standards. A network compliant with the LTE (Long Term Evolution) system of 3GPP standards may be partially applied to the cellular network 10, or in the future, a 6th generation (6G) system whose standardization is planned may be at least partially applied. The cellular network 10 may be a mobile communication system. A configuration example of the cellular network 10 will be described later.

[0020] The UE 100 is a movable wireless communication device. The UE 100 may be any device as long as it is a device used by a user. For example, the UE 100 may be a mobile phone terminal (including a smartphone), a tablet terminal, a notebook PC, a communication module (including a communication card or a chipset), a sensor or a device provided in the sensor, a vehicle or a device provided in the vehicle (Vehicle UE), an aircraft or a device provided in the aircraft (Aerial UE, or UAV (Unmanned Aerial Vehicle)). Alternatively, the UE 100 may be an IoT (Internet of Things) device, an IoT sensor, or the like.

[0021] Node 110 may be connected as one node or multiple nodes in the cellular network 10. Node 110 may function as a base station in the cellular network 10. Details of Node 110 will be described later. Node 110 may also be referred to as a network node.

[0022] The CN device 120 functions as a communication device in the cellular network 10. The CN device 120 may be a device that connects the cellular network 10 and an external network. Alternatively, the CN device 120 may be a device having a gateway function for connecting the cellular network 10 and an external network. The CN device 120 may be an access and mobility management function (AMF) device that manages access and mobility control of the UE 100. Alternatively, the CN device 120 may be a session management function (SMF) device that manages the communication session of the UE 100 in the cellular network 10. Alternatively, the CN device 120 may function as an end point of a communication session (for example, a PDU (Protocol Data Unit) session) between the UE 100 in the cellular network 10 and exchange user data with the UE 100, and may be a UPF (User Plane Function).

[0023] The network slice management device 130 manages one or more network slices in the cellular network 10. The network slice management device 130 manages the RAN slice, transport slice, and core slice included in the network slice as a network slice orchestrator. The network slice management device 130 may manage the generation of network slices. Further, the network slice management device 130 may set network slices in the node 110 and the CN device 120 by transmitting configuration information to the node 110 and the CN device 120. Note that "management" of a network slice may include "control" of the network slice. Alternatively, "management" of a network slice and "control" of a network slice may be used with the same meaning.

[0024] In the first embodiment, the network slice management device 130 can also manage network slices by using a learning function based on AI (Artificial Intelligence). Details will be described later. Hereinafter, the network slice management device 130 may be referred to as a "network slice orchestrator" (or simply an "orchestrator").

[0025] (1.1) Configuration example of a cellular network Next, a configuration example of the cellular network 10 will be described.

[0026] FIG. 2 is a diagram showing a configuration example of the cellular network 10 according to the first embodiment.

[0027] As shown in FIG. 2, the cellular network 10 includes a UE 100, a RAN (Radio Access Network) 30, a CN (Core Network) 40, and an orchestrator 130.

[0028] RAN 30 includes the one or more nodes 110 (in the example of FIG. 2, 110-1 to 110-3) described above. The nodes 110 are interconnected via an inter-node interface. The nodes 110 may be referred to as base stations. When the cellular network 10 is a 5GS, the nodes 110 are referred to as gNBs. The nodes 110 may be composed of (i.e., functionally split into) an RU, a CU, and a DU.

[0029] The node 110 manages one or more cells. The node 110 performs wireless communication with the UE 100 that has established a connection with its own cell. The node 110 has functions such as a radio resource management (RRM) function, a routing function for user data (hereinafter simply referred to as "data"), and a measurement control function for mobility control and scheduling. Note that "cell" is used as a term indicating the smallest unit of a wireless communication area. "Cell" is also used as a term indicating a function or resource for performing wireless communication with the UE 100. One cell belongs to one carrier frequency (hereinafter simply referred to as "frequency").

[0030] CN 40 includes the CN device 120 described above.

[0031] (1.1.1) Network slice The network slice according to the first embodiment is constructed in the cellular network 10. Here, the network slice will be described.

[0032] As described above, a network slice is a virtual network created by virtually dividing the physical network constructed by an operator. By constructing a network slice, services such as high-speed large-capacity (eMBB), ultra-low latency (URLLC), and massive machine type communication (mMTC) can be provided to the UE 100 by occupying resources within the cellular network 10.

[0033] FIG. 3 is a diagram showing an example of a network slice. The network slice is set on a network composed of RAN 30 and CN 40. One or more network slices can be set on the network. Each network slice is associated with one service type. Examples of service types include eMBB, URLLC, MIoT (Massive Internet of Things), V2X (Vehicle to Everything), HMTC (High Performance Machine Type Communication), etc. For example, as shown in FIG. 3, the service type eMBB is associated with network slice #1, the service type MIoT is associated with network slice #2, and the service type HMTC is associated with network slice #3, and so on.

[0034] Each network slice is provided with a slice identifier for identifying the network slice. An example of the slice identifier is S-NSSAI (Single Network Slicing Selection Assistance Information). S-NSSAI includes an 8-bit SST (slice / service type). S-NSSAI may further include a 24-bit SD (slice differentiator).

[0035] SST is an example of service type information indicating the service type with which the network slice is associated. For example, SST = 1 indicates eMBB, SST = 2 indicates URLLC, SST = 3 indicates MIoT (Massive Internet of Things), SST = 4 indicates V2X (Vehicle to everything), and SST = 5 indicates HMTC (High Performance Machine Type Communication). However, these are just examples, and the value of SST may be associated with other service types, or other service types may be associated with other SST values.

[0036] SD is information for differentiating a plurality of network slices associated with the same service type.

[0037] Information including a plurality of S-NSSAIs is called NSSAI (Network Slice Selection Assistance Information). Also, one or more network slices may be grouped to form a slice group. Also, a slice group is a group including one or more network slices, and a slice group identifier is assigned to the slice group.

[0038] As shown in FIG. 3, a plurality of slices may be set for one UE100. In this case, the UE100 can simultaneously receive the provision of a plurality of services via each slice.

[0039] Also, as shown in FIG. 3, a network slice includes hardware and functional blocks in RAN30 and CN40. In the example of FIG. 3, network slice #1 includes RU, DU, CU#1, and UPF#1. Also, network slice #2 includes RU, DU, CU#2, and UPF#2. Further, network slice #3 includes RU, DU, CU#3, and UPF#3. Each hardware and each logical functional block of RAN30 and CN40 may be shared by a plurality of network slices. Each network slice can provide each service such as eMBB and URLLC to the UE100 by using the hardware and functional blocks. Hereinafter, the hardware and functional blocks in the cellular network 10 may be simply referred to as "resources".

[0040] As described above, a network slice includes a RAN slice, a transport slice, and a core slice. FIG. 4 is a diagram showing an example of a network slice including a RAN slice, a transport slice, and a core slice. In the example of FIG. 4, network slice #1 includes RAN slice #1, transport slice #1, and core slice #1. Also, network slice #2 includes RAN slice #2, transport slice #2, and core slice #2. Further, network slice #3 includes RAN slice #3, transport slice #3, and core slice #3. By including a RAN slice, a transport slice, and a core slice in a network slice in this way, for example, the resources used in the RAN slice can be dynamically changed, and it becomes possible to flexibly respond to the requirements for each network slice for each of the RAN slice, the transport slice, and the core slice.

[0041] The orchestrator 130 (network slice management device 130) includes a RAN slice controller 131, a transport slice controller 132, and a core slice controller 133, respectively, in order to manage the RAN slice, the transport slice, and the core slice included in the network slice. And the orchestrator 130 includes a network slice controller 135 in order to manage the entire network slice.

[0042] The RAN slice controller 131 controls the RAN slice in the RAN 30. For example, the RAN slice controller 131 controls which RAN slice the radio resources used for wireless communication with the UE 100 and the resources (such as RUs, DUs, and CUs) in the RAN 30 used for communication with the UE 100 belong to. The RAN slice controller 131 may control each RAN slice according to an instruction from the network slice controller 135. Also, when a packet is transmitted in the RAN 30, the RAN slice controller 131 may acquire RAN packet information indicating how the packet has operated in the RAN 30. The RAN slice controller 131 outputs the RAN packet information to the network slice controller 135.

[0043] The transport slice controller 132 controls the transport slice in the fronthaul, midhaul, and backhaul. For example, the transport slice controller 132 controls the path information indicating the path from the fronthaul to the backhaul for each transport slice. The transport slice controller 132 may also control each transport slice according to an instruction from the network slice controller 135. Also, when a packet is transmitted in the backhaul, midhaul, and fronthaul, the transport slice controller 132 may acquire transport packet information indicating how the packet has operated in the backhaul, midhaul, and fronthaul. The transport slice controller 132 outputs the transport packet information to the network slice controller 135.

[0044] The core slice controller 133 controls which CN device 120 is to be applied for each network slice. The core slice controller 133 may control each core slice according to an instruction from the network slice controller 135. Further, the core slice controller 133 outputs core packet information indicating how the packets transmitted in the CN 40 have operated to the network slice controller 135.

[0045] The network slice controller 135 manages the entire network slice in the cellular network. The network slice controller 135 may respectively instruct the RAN slice controller 131, the transport slice controller 132, and the core slice controller 133 to set the RAN slice, the transport slice, and the core slice. The network slice controller 135 may preset a network slice as shown in FIG. 4 and transmit network slice setting information indicating the set network slice to the AMF (or SMF) which is the CN device 120. The AMF (or SMF) may select an allowed network slice (specifically, Allowed NSSAI) based on the network slice setting information for the network slice (specifically, Requested NSSAI) requested from the UE 100.

[0046] FIGS. 5(A) and 5(B) are diagrams showing examples of network slice setting information according to the first embodiment. The network slice setting information is managed in the orchestrator 130. As shown in FIGS. 5(A) and 5(B), each network slice is identified by an S-NSSAI, and the RAN slice, the transport slice, and the core slice included in the network slice are associated with the S-NSSAI.

[0047] In the example of Fig. 5(A), the network slice #1 shown in Fig. 4 is represented by S-NSSAI #1, and S-NSSAI #1 is associated with the identification information (ID) of RAN slice #1, the identification information (ID) of transport slice #1, and the identification information (ID) of core slice #1. Thereby, the network slice controller 135 can manage the RAN slice #1, the transport slice #1, and the core slice #1 in the network slice #1. And in the ID of RAN slice #1, the radio resources used in RAN slice #1 are associated with the identification information of RU, DU, and CU#1 in RAN 30. Thereby, the network slice controller 135 can manage the radio resources and hardware used in RAN slice #1.

[0048] Also, the network slice controller 135 can manage which backhaul, which middle-haul, and which front-haul are used in the transport slice #1 by associating and managing the path information #1 with respect to the ID of the transport slice #1. Furthermore, the network slice controller 135 can manage which core network functions are included in the core slice #1 (UPF#1 in Fig. 5(A)) by associating and managing UPF#1 with respect to the ID of the core slice #1. Regarding the path information, a path table (or routing table) storing path information representing each path from the backhaul to the front-haul may be prepared. The network slice controller 135 may acquire the path information using the path table.

[0049] Note that the network slice configuration information and the path table shown in Figs. 5(A) and 5(B) may be stored in the memory in the orchestrator 130.

[0050] Returning to FIG. 4, the network slice controller 135 may obtain packet information indicating the operation of the packets transmitted in the cellular network 10 based on the RAN packet information, the transport packet information, and the core packet information.

[0051] (1.1.2) Configuration example of each device in the cellular network Next, a configuration example of each device in the cellular network 10 will be described.

[0052] First, a configuration example of the UE 100 will be described.

[0053] FIG. 6 is a diagram showing a configuration example of the UE 100 (user equipment) according to the first embodiment. As shown in FIG. 6, the UE 100 includes a receiving unit 101, a transmitting unit 102, and a control unit 103. The receiving unit 101 and the transmitting unit 102 constitute a wireless communication unit that performs wireless communication with the node 110.

[0054] The receiving unit 101 performs various receptions under the control of the control unit 103. The receiving unit 101 includes an antenna and a receiver. The receiver converts the wireless signal received by the antenna into a baseband signal (received signal) and outputs it to the control unit 103.

[0055] The transmitting unit 102 performs various transmissions under the control of the control unit 103. The transmitting unit 102 includes an antenna and a transmitter. The transmitter converts the baseband signal (transmitted signal) output by the control unit 103 into a wireless signal and transmits it from the antenna.

[0056] The control unit 103 performs various controls and processes in the UE 100. Such processes include the processes of each layer described later. The control unit 103 includes at least one processor and at least one memory. The memory stores programs executed by the processor and information used for the processes by the processor. The processor may include a baseband processor and a CPU (Central Processing Unit). The baseband processor performs modulation / demodulation and encoding / decoding of baseband signals, etc. The CPU executes programs stored in the memory to perform various processes. Note that the processes or operations in the UE 100 described later may be performed by the control unit 103. Also, the transmission of messages in the UE 100 described later may be performed by the transmission unit 102, and the reception of messages in the UE 100 may be performed by the reception unit 101.

[0057] Next, a configuration example of the node 110 will be described.

[0058] FIG. 7 is a diagram showing the configuration of the node 110 (base station) according to the first embodiment. As shown in FIG. 7, the node 110 includes a transmission unit 111, a reception unit 112, a control unit 113, and a network communication unit 115. The transmission unit 111 and the reception unit 112 constitute a wireless communication unit that performs wireless communication with the UE 100. The network communication unit 115 constitutes a network communication unit that communicates with the CN device 120.

[0059] The transmission unit 111 performs various transmissions under the control of the control unit 113. The transmission unit 111 includes an antenna and a transmitter. The transmitter converts the baseband signal (transmission signal) output by the control unit 113 into a wireless signal and transmits it from the antenna.

[0060] The reception unit 112 performs various receptions under the control of the control unit 113. The reception unit 112 includes an antenna and a receiver. The receiver converts the wireless signal received by the antenna into a baseband signal (reception signal) and outputs it to the control unit 113.

[0061] The control unit 113 performs various controls and processes at the node 110. Such processes include the processes of each layer described later. The control unit 113 includes at least one processor and at least one memory. The memory stores programs executed by the processor and information used for the processes by the processor. The processor may include a baseband processor and a CPU. The baseband processor performs modulation / demodulation and encoding / decoding of baseband signals, etc. The CPU executes programs stored in the memory to perform various processes. Note that the processes or operations at the node 110 described later may be performed by the control unit 113. Also, the transmission of messages at the node 110 described later may be performed by the transmission unit 111 and the network communication unit 115, and the reception of messages at the node 110 may be performed by the reception unit 112 and the network communication unit 115.

[0062] The network communication unit 115 is connected to the CN device 120 via an NG interface which is an interface between the base station and the core network. Also, the network communication unit 115 is connected to other nodes via an Xn interface. Various processes may also be performed on the network communication unit 115 under the control of the control unit 113.

[0063] Next, a configuration example of the CN device 120 will be described.

[0064] FIG. 8 is a diagram showing a configuration example of the CN device 120 according to the first embodiment. As shown in FIG. 8, the CN device 120 includes a reception unit 121, a transmission unit 122, and a control unit 123.

[0065] The reception unit 121 performs various receptions under the control of the control unit 123. The reception unit 121 receives a message transmitted from the node 110 (for example, a message via the N2 interface). Also, the reception unit 121 may receive a message transmitted from another CN device (for example, a message via the N11 interface). The reception unit 121 outputs the received message to the control unit 123.

[0066] The transmission unit 122 performs various transmissions under the control of the control unit 123. The transmission unit 122 transmits the message received from the control unit 123 (for example, the message by the N2 interface) to the node 110 according to the instruction from the control unit 123. Also, the transmission unit 122 transmits the message received from the control unit 123 (for example, the message by the N11 interface) to other CN apparatuses according to the instruction from the control unit 123.

[0067] The control unit 123 performs various controls and processes in the CN apparatus 120. The control unit 123 includes at least one processor and at least one memory. The memory stores a program executed by the processor and information used for the process by the processor. The processor may include a CPU. The CPU executes the program stored in the memory to perform various processes. Note that the process or operation in the CN apparatus 120 described later may be performed by the control unit 123. Also, the transmission of the message in the CN apparatus 120 described later may be performed by the transmission unit 122, and the reception of the message in the CN apparatus 120 may be performed by the reception unit 121.

[0068] Next, a configuration example of the network slice management apparatus (or orchestrator) 130 will be described.

[0069] FIG. 9 is a diagram showing a configuration example of the orchestrator 130. As shown in FIG. 9, the orchestrator 130 includes a reception unit 136, a transmission unit 137, and a control unit 138.

[0070] The reception unit 136 performs various receptions under the control of the control unit 138. For example, the reception unit 136 receives RAN packet information from the RAN 30, core packet information from the CN 40, and transport packet information from the RAN 30 and the CN 40. The reception unit 136 outputs the RAN packet information, the core packet information, and the transport packet information to the control unit 138.

[0071] The transmission unit 137 performs various transmissions under the control of the control unit 138. For example, the transmission unit 137 transmits a message or the like to the RAN 30 and / or the CN 40 according to an instruction from the control unit 138.

[0072] The control unit 138 performs various controls and processes in the orchestrator 130. The control unit 138 may be a block that realizes the functions of the RAN slice controller 131, the transport slice controller 132, the core slice controller 133, and the network slice controller 135. The control unit 138 may be divided into four control units 138 in order to realize the functions of the RAN slice controller 131, the transport slice controller 132, the core slice controller 133, and the network slice controller 135, rather than just one. The control unit 138 includes at least one processor and at least one memory. The memory stores a program executed by the processor and information used for the processing by the processor. The processor may include a CPU. The CPU executes a program stored in the memory to perform various processes. Note that the processes or operations in the orchestrator 130 described later may be performed by the control unit 138.

[0073] (2) Operation example according to the first embodiment Next, an operation example according to the first embodiment will be described.

[0074] FIG. 10 is a diagram showing an operation example according to the first embodiment.

[0075] In the first embodiment, when an application is executed in the UE 100, in the orchestrator 130, a learning model (or an AI model. Hereinafter, it may be referred to as a "learning model" in some cases) is created based on the packet information.

[0076] Generally, when creating a learning model, learning for the learning model is performed by inputting learning data into the learning model. The state in which the learning model is being learned by inputting learning data into the learning model may be referred to as the "training mode". Also, the learning model being learned in the training mode may be referred to as the "training model" (or the training AI model). On the other hand, for the learning model after a certain amount of learning has been performed on the "training model", inference result data can be obtained by inputting inference data. The learning model after a certain amount of learning has been performed on the "training model" may be referred to as the "trained model" (or the trained AI model). The state of obtaining inference result data from inference data using the trained model may be referred to as the "inference mode". The inference mode may also be called the execution mode or the operation mode.

[0077] Note that generally, in the training mode, the learning performed by providing correct answer data (or teacher data) as learning data may be referred to as supervised learning. On the other hand, in the training mode, the learning performed without providing correct answer data as learning data may be referred to as unsupervised learning. In the first embodiment, mainly the case where supervised learning is used will be described, but it is not limited thereto, and unsupervised learning may be used.

[0078] FIG. 11 is a diagram showing an example of generation of a trained model according to the first embodiment. As shown in FIG. 11, the control unit 138 of the orchestrator 130 has a model learning unit 1380 and a model inference unit 1381.

[0079] The model learning unit 1380 generates a training model from the learning data and also generates a trained model by inputting the learning data into the training model. On the other hand, the model inference unit 1381 outputs inference result data from the inference data using the trained model.

[0080] Returning to FIG. 10, in the orchestrator 130, in the learning mode, in each phase from step S2 to step S4, learning is performed on the learning model, and a trained model is generated. Then, in the orchestrator 130, in the inference mode, in each phase from step S5 to step S8, inference result data is output using the trained model.

[0081] First, an operation example in the learning mode will be described below, and then an operation example in the inference mode will be described.

[0082] (2.1) Operation example in the learning mode In the learning mode, a trained model is generated in each of the following three phases.

[0083] (2.1.1) Application identification (step S2)

[0084] (2.1.2) Network slice design (step S3)

[0085] (2.1.3) Communication control (step S4) The following will be described in order. Note that the operation example shown in FIG. 10 is described as being performed under the following premises, for example.

[0086] That is, in UE100, an application is executed to generate packets. The packets include the application name. The application name may be included in the trailer part of the packet or in the payload part of the packet (e.g., Fig. 15). For example, the network slice controller 135 can obtain the application name from the packets transmitted by the cellular network 10 in RAN30 or CN40. Also, the network slice controller 135 can obtain the application name from an external network. Each slice controller (RAN slice controller 131, transport slice controller 132, and / or core slice controller 133) may obtain the application name from the packets transmitted by the cellular network 10 and output the obtained application name to the network slice controller 135.

[0087] (2.1.1) Application Identification (Step S2) In UE100, when an application is executed, packets related to the application are transmitted from UE100. The packets are input to RAN30 and CN40 of the cellular network 10. At this time, the orchestrator 130 collects the communication path information of the packets from the node 110 included in RAN30 and / or the CN device 120 included in CN40 (Step S1).

[0088] First, the communication path information may include packet information indicating the operation of the packet data in the cellular network. The packet information may be the packet information of one packet transmitted from the own UE100. Alternatively, the packet information may be the packet information of a plurality of packets transmitted from the own UE100. Alternatively, the packet information may be the packet information not only of the own UE100 but also of other UE100. The packet information includes, for example, at least any of the following information.

[0089] (2.1.1.1) The data length of the data included in the payload part of the packet data

[0090] (2.1.1.2) Transmission interval of the packet data

[0091] (2.1.1.3) The number of UEs 100 (or terminal devices) that connect to the cellular network when executing an application, and

[0092] (2.1.1.4) The amount of data accumulated in the transmission queue of the CN device 120 in the cellular network In the learning mode, the orchestrator 130 can obtain, from the packet information, what the actual operation of the packets in the cellular network 10 was like.

[0093] Second, the communication path information may include network slice setting information indicating information for setting a network slice. The network slice setting information is, for example, setting information regarding the network slice set in the cellular network 10 when an application is executed in the UE 100. The network slice setting information is shown, for example, in FIG. 5. The network slice setting information is set in the CN device 120 (for example, the AMF) when the UE 100 performs a registration procedure to the cellular network 10. The network slice setting information may be generated by the CN device 120 and transmitted from the CN device 120 to the orchestrator 130. Alternatively, the orchestrator 130 generates the network slice setting information and transmits it to the AMF, and the AMF may select any network slice setting information from the network slice setting information received from the orchestrator 130 during the registration procedure.

[0094] Thirdly, the communication path information may include communication control information. The communication control information indicates information for communication control in the cellular network 10 of the packet. That is, the communication control information indicates information for controlling the communication of the packet in the cellular network 10 when the packet is transmitted in the cellular network 10. The communication control information may include, for example, information related to a QoS flow set between the UE 100 and the UPF (an example of the CN device 120). As information related to the QoS flow, a QoS flow identifier (QFI: QoS Flow ID) for identifying each QoS flow from other QoS flows, a 5QI (5G QoS Identifier) representing the characteristics of each QoS flow, and / or an allocation and retention priority (ARP) of each QoS flow may be included. The information related to the QoS flow is set in the SMF (an example of the CN device 120) when the UE 100 establishes a PDU session with the UPF. The information related to the QoS flow (communication control information) may be generated by the SMF and transmitted from the SMF to the orchestrator 130. Further, the communication control information may include, for example, information related to the radio resources allocated by the node 110 to the UE 100. As information related to the radio resources, for example, time resources and frequency resources, and / or the modulation order may be included. The information related to the radio resources is set in the node 110 when the node 110 performs wireless communication with the UE 100. The information related to the radio resources (communication control information) may be generated by the node 110 and transmitted from the node 110 to the orchestrator 130. Alternatively, the communication control information may include information related to the computing resources in the node 110. The computing resources may be represented by the number of CPUs, the number of clock cycles per unit time of the CPU, and / or the processing speed per unit time of the CPU. Alternatively, the computing resources may be represented by the number of memories and / or the memory capacity. Alternatively, the communication control information may include information related to routing. The information related to routing is information related to the selection of the communication path of the packet.Alternatively, the information regarding routing may be information indicating through which node among the plurality of nodes 110 the packet is transmitted.

[0095] In application identification (step S2), the orchestrator 130 performs learning by the learning model in training using the packet information among the collected communication path information.

[0096] FIG. 12(A) is a diagram showing a learning example using the learning model in training for application identification according to the first embodiment.

[0097] As shown in FIG. 12(A), in the learning mode, the packet information and the application name are input to the learning model in training for application identification to train the learning model in training for application identification. The training data is the packet information and the application name. The application name may be information representing a specific application such as "video delivery service", "telemedicine", or "robot control".

[0098] There is a certain relationship between the application name and the packet information. For example, in the case of "video delivery service", since a large amount of data is transmitted, the data length of the user data included in the packet can be a certain length or more. Also, for example, in the case of "telemedicine", since video data etc. is transmitted in real time, the data length can be a certain length or more and the packet transmission interval can be a certain interval or less. By training the application name and the packet information having such a relationship with each other in the learning model, it is possible to generate a trained model that outputs the application name when the packet information is input.

[0099] In the control unit 138 (model learning unit 1380), packet information is acquired from the cellular network 10, the application name is acquired from the packet, and the learning model being trained for application identification is trained. After the learning is completed, the control unit 138 (model learning unit 1380) generates a learned model for application identification (or a learned AI model for application identification) as the learned model for application identification.

[0100] (2.1.2) Network Slice Design (Step S3) Next, the network slice design (Step S3) in FIG. 10 will be described.

[0101] In the network slice design (Step S3), among the communication path information collected by the orchestrator 130, the network slice setting information is used. Also, in the learning mode of the network slice design (Step S3), learning is performed using the learning model being trained for network slice design.

[0102] FIG. 13(A) is a diagram showing an example of learning using the learning model being trained for network slice design according to the first embodiment.

[0103] As shown in FIG. 13(A), in the learning mode, learning is performed on the learning model being trained for network slice design using the network slice setting information and the application name as learning data.

[0104] There is a certain relationship between the application name and the network slice configuration information. For example, in the case of a "video delivery service", since a large amount of data is transmitted, network slice configuration information corresponding to a network slice used as high-speed large-capacity (eMBB) may be used. Also, for example, in the case of "remote medical treatment", since video data etc. is transmitted in real time, network slice configuration information corresponding to a network slice used as ultra-low latency (URLLC) may be used. Furthermore, for example, in the case of "robot control", since control is performed using a large number of cameras, network slice configuration information corresponding to a network slice used as massive machine type communication (mMTC) may be used. By using the application name and the network slice configuration information that have a certain relationship with each other to train the model being trained, it is possible to generate a trained model that outputs network slice configuration information when an application name is input.

[0105] That is, the control unit 138 (model learning unit 1380) acquires the application name from the packet, acquires the network slice configuration information from the orchestrator 130 or the CN device 120 etc., and trains the model being trained for network slice design. After the training is completed, the control unit 138 (model learning unit 1380) will generate a trained model by using the trained model being trained for network slice design as a trained model for network slice design (or a trained AI model for network slice design).

[0106] As the application name input to the model being trained for network slice design, the application name obtained from the model being trained for application identification (or the trained AI model for application identification) may be used.

[0107] (2.1.3) Communication control (step S4) Next, the communication control (step S4) shown in FIG. 10 will be described.

[0108] In communication control (step S4), among the communication path information collected by the orchestrator 130, communication control information is used. Also, in communication control (step S4), in the learning mode, a learning model for communication control is learned. In the first embodiment, in a situation where a network slice is set between the RAN 30 and the CN 40 of the cellular network 10, information set in the communication path (for example, a QoS flow) between the UE 100 and the UPF will be described as the target of the communication control information.

[0109] FIG. 14(A) is a diagram showing a learning example using a learning model for communication control according to the first embodiment.

[0110] As shown in FIG. 14(A), in the learning mode, the application name and the communication control information are used as learning data to learn a learning model for communication control.

[0111] There is a certain relationship between the application name and the communication control information. For example, in the case of a "video delivery service", since a large amount of data is transmitted, communication control information with a bit rate guaranteed to be above a certain level may be set. Also, for example, in the case of "telemedicine", since video data etc. is transmitted in real time, communication control information with an upper limit of packet data delay (packet delay budget) below a certain level and a packet error rate also below a certain level may be set. Further, for example, in the case of "robot control", since many cameras are used to acquire video data, communication control information with a bit rate guaranteed to be above a certain level and a packet error rate also below a certain level may be set. By learning the application name and the communication control information, which have a certain relationship with each other, with a learning model for communication control, it is possible to generate a learned model that outputs communication control information when the application name is input.

[0112] That is, in the control unit 138 (model learning unit 1380), the application name is obtained from the packet data, and communication control information is obtained from the orchestrator 130 or the CN device 120 (or the orchestrator 130 or the node 110), etc., to train the learning model for communication control. After the training is completed, the control unit 138 (model learning unit 1380) will generate a trained model for communication control as a trained model (or a trained AI model for communication control) from the learning model for communication control.

[0113] The application name input to the learning model for communication control may be the application name obtained from the learning model for application identification (or the trained model for application identification).

[0114] Then, returning to FIG. 11, after the trained model is created in each phase (step S2 to step S4), in the inference mode, the processing using the trained model is executed in each phase from step S5 to step S7. Next, an operation example in the inference mode (step S5 to step S7) will be described. However, in the following operation example, it is described that the same application as the one performed in the learning mode is performed in the UE 100 in the inference mode.

[0115] (2.2) Operation example in the inference mode

[0116] (2.2.1) Application identification (step S5) FIG. 12(B) is a diagram showing an inference example using a learned model for application identification according to the first embodiment. As shown in FIG. 12(B), in the inference mode, packet information is input to the learned model for application identification to infer the application name. The inference data is the packet information, and the inference result data is the application name. The packet information to be input to the learned model for application identification is the operation information of the packet transmitted along with the application executed in the UE 100 in the inference mode. The packet information may be acquired by the orchestrator 130 from the node 110 and / or the CN device 120 as in the learning mode. Then, the control unit 138 (or the model inference unit 1381) of the orchestrator 130 infers the application name from the packet information using the learned model for application identification.

[0117] (2.2.2) Network slice design (step S6) FIG. 13(B) is a diagram showing an inference example using a learned model for network slice design according to the first embodiment. As shown in FIG. 13(B), in the inference mode, the application name is input to the learned model for network slice design to infer the network slice setting information. The inference data is the application name, and the inference result data is the network slice setting information. The application name to be input to the learned model for network slice design is the application name inferred from the learned model for application identification (step S5, FIG. 12(B)). In the control unit 138 (or the model inference unit 1381) of the orchestrator 130, the network slice setting information is inferred from the application name using the learned model for network slice design.

[0118] (2.2.3) Communication control (step S7) FIG. 14(B) is a diagram showing an inference example using a learned model for communication control according to the first embodiment. As shown in FIG. 14(B), in the inference mode, the application name is input to the learned model for communication control to infer communication control information. The inference data is the application name, and the inference result data is the communication control information. The application name input to the learned model for communication control is the application name inferred from the learned model for application identification (step S5, FIG. 12(B)). In the control unit 138 (or the model inference unit 1381) of the orchestrator 130, the communication control information is inferred from the application name using the learned model for communication control.

[0119] Returning to FIG. 10, in step S8, the orchestrator 130 sets the network slice of the cellular network 10 using the network slice setting information (step S6), and controls the packet communication (or communication path) in the cellular network 10 using the communication control information (step S7). Thereafter, the orchestrator 130 collects communication path information (step S1) using the packets transmitted via the communication path in the set cellular network 10, and repeats steps S5 to S8 described above.

[0120] (2.3) When a new application is executed Next, an operation example when an application that has not been executed so far, that is, an application that has not been generated as a learned model (hereinafter sometimes referred to as a "new application") is executed in the UE 100 will be described.

[0121] In the orchestrator 130, when a new application is executed in the UE 100, the learned models for application identification, network slice design, and communication control are relearned. Specifically, when the control unit 138 receives a packet (for example, the second packet) including the application name of the application that has not been executed in the UE 100 (that is, a new application) when the application is executed in the UE 100, at least using the application name, the learned model for application identification and the learned model for network slice design are relearned. Further, when the control unit 138 receives a packet including the application name of the new application, at least using the application name, the learned model for communication control is relearned.

[0122] By retraining the learned model for application identification, the learned model for network slice design, and the learned model for communication control, it is possible to generate a learned model for application identification, a learned model for network slice design, and a learned model for communication control that correspond to a new application, respectively. And in the control unit 138, even when packet data is received by executing a new application, by using the retrained learned model for application identification, the learned model for network slice design, and the learned model for communication control, the application name, network slice setting information, and communication control information that correspond to the new application can be obtained, respectively. And, for example, by applying the network slice setting information corresponding to the new application to the cellular network 10, it is also possible to automatically set the network slice corresponding to the new application in the cellular network 10. Therefore, it becomes possible to automatically apply the optimal network slice for an application (especially a new application). Also, for example, by applying the communication control information corresponding to the new application to the cellular network 10, it is also possible to optimize the transmission path of the packets transmitted by the new application. Therefore, it becomes possible to set optimal communication control within the network (for example, the cellular network 10).

[0123] Specifically, the retraining is as follows.

[0124] (2.3.1) Retraining of the Learned Model for Application Identification When a new application is executed, the control unit 138 obtains the name of the new application from the trailer part or the payload part of the packet (the second packet) received in the cellular network 10. Specifically, the control unit 138 obtains the name of the application from the node 110 or the CN device 120. Alternatively, the control unit 138 may obtain the application name from an external network. For example, the control unit 138 stores the application names obtained so far in the memory, and determines that it is the name of a new application when the obtained application name is different from the stored application names. Then, the control unit 138 (the model learning unit 1380) obtains the packet information of the packet from the CN device 120, and retrains the learned model for application identification using the packet information of the packet and the name of the new application (FIG. 12(A)). The control unit 138 (the model learning unit 1380) generates a learned model for application identification that also corresponds to the new application by retraining. Then, the control unit 138 (the model inference unit 1381) infers the name of the new application from the packet information using the learned model for application identification (FIG. 12(B)).

[0125] (2.3.2) Retraining of the learned model for network slice design Retraining of the learned model for network slice design is also performed based on the packets received in the cellular network 10 when a new application is executed. That is, in the control unit 138 (the model learning unit 1380), the learned model for network slice design is retrained using the network slice setting information (the second network slice setting information) set in the cellular network 10 when the new application is executed and the name of the new application included in the packet (FIG. 13(A)). The control unit 138 (the model learning unit 1380) generates a retrained learned model for network slice design.

[0126] Then, the control unit 138 (model inference unit 1381) uses the learned model for the re-learned network slice design to obtain the network slice setting information of the new application from the application name (Fig. 13(B)).

[0127] As the name of the new application to be input to the learned model for the re-learned network slice design, the application name inferred from the learned model for the re-learned application identification may be used.

[0128] (2.3.3) Re-learning of the learned model for communication control The re-learning of the learned model for communication control is also performed based on the packets received in the cellular network 10 by the execution of the new application. That is, in the control unit 138 (model learning unit 1380), when the new application is executed, the communication control information (second communication control information) set in the cellular network 10 and the name of the new application included in the packet are used to re-learn the learned model for communication control (Fig. 14(A)). The control unit 138 (model learning unit 1380) generates a re-learned learned model for communication control.

[0129] Then, the control unit 138 (model inference unit 1381) uses the re-learned learned model for communication control to obtain the communication control information of the new application from the application name (Fig. 14(B)).

[0130] As the name of the new application to be input to the re-learned learned model for communication control, the application name obtained from the re-learned learned model for application identification may be used.

[0131] [Other Embodiments] Each of the above operation flows can be implemented not only separately and independently, but also by combining two or more operation flows. For example, some steps of one operation flow may be added to another operation flow, or some steps of one operation flow may be replaced with some steps of another operation flow. In each flow, it is not necessarily required to execute all steps, and only some steps may be executed.

[0132] In the above embodiments and examples, an example where the base station is an NR base station (gNB) has been described, but the base station may be an LTE base station (eNB) or a 6G base station.

[0133] Also, the term "network node" or "node" mainly means a base station, but may also mean a core network device or a part of a base station (CU, DU, or RU).

[0134] A program may be provided to cause a computer to execute each process performed by UE100, node 110, CN device 120, or network slice management device 130. The program may be recorded on a computer-readable medium. By using a computer-readable medium, it is possible to install the program on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM. Also, circuits for executing each process performed by UE100, node 110, CN device 120, or network slice management device 130 may be integrated, and at least a part of UE100, node 110, CN device 120, or network slice management device 130 may be configured as a semiconductor integrated circuit (chipset, SoC).

[0135] As used in this disclosure, the terms "based on" and "depending on / in response to" do not mean "only based on" or "only in response to" unless otherwise specified. The term "based on" means both "only based on" and "at least partially based on". Similarly, the term "depending on" means both "only depending on" and "at least partially depending on". The terms "include", "comprise", and their variants do not mean only including the listed items, but mean that they may include only the listed items or may further include additional items in addition to the listed items. Also, the term "or" used in this disclosure is not intended to be an exclusive disjunction. Further, any reference to an element using designations such as "first", "second", etc. used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used herein as a convenient way to distinguish between two or more elements. Thus, a reference to a first and a second element does not mean that only two elements can be employed there or that the first element must precede the second element in some form. In this disclosure, for example, when articles are added by translation like a, an, and the in English, these articles shall be construed to include plural ones unless the context clearly indicates otherwise.

[0136] Also, the functions implemented by the UE 100 or the base station 200 (network node) may be implemented in circuitry or processing circuitry including a general-purpose processor, a specific-purpose processor, an integrated circuit, ASICs (Application Specific Integrated Circuits), a CPU (a Central Processing Unit), conventional circuitry, and / or combinations thereof, programmed to implement the described functions. The processor includes transistors and other circuitry and is regarded as circuitry or processing circuitry. The processor may be a programmed processor that executes a program stored in a memory.

[0137] In this specification, circuitry, unit, and means are hardware programmed to implement the described functions or hardware that executes. The hardware may be any hardware disclosed in this specification or any hardware known to be programmed or execute to implement the described functions.

[0138] When the hardware is a processor regarded as a type of circuitry, the circuitry, means, or unit is a combination of the hardware and software used to configure the hardware and / or the processor.

[0139] As described above, one embodiment has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the gist. Also, each embodiment, each operation example, or each process can be appropriately combined within a non - conflicting range.

[0140] (Appendix)

[0141] (Appendix 1) A network slice management device for managing network slices in a cellular network, using a learned AI model for application identification, inputs packet information indicating the operation of packets of an application in the cellular network and outputs an application name, and has a control unit that, using a learned AI model for communication control, inputs the application name and outputs communication control information indicating information for communication control of the packets in the cellular network. When the control unit receives a packet including the name of a new application, it retrains at least the learned AI model for application identification and the learned AI model for communication control using at least the name of the new application. Network slice management device.

[0142] (Appendix 2) When the control unit receives the packet, it retrains the learned AI model for application identification using the name of the new application and the packet information of the packet. The network slice management device according to Appendix 1.

[0143] (Appendix 3) When the control unit receives the packet, it retrains the learned AI model for communication control using the name of the new application and the communication control information set when the new application is executed. The network slice management device according to Appendix 1.

[0144] (Appendix 4) The communication control information includes information related to a QoS flow set between a user device and a core network device. The network slice management device according to Appendix 1. (Appendix 5) The packet information represents at least one of the data length of the data included in the payload part of the packet, the transmission interval of the packet, the number of user devices connected to the cellular network when the application is executed, and the amount of data accumulated in the transmission queue of the core network device in the cellular network. The network slice management device according to Supplementary Note 1.

[0145] (Supplementary Note 6) A management method in a network slice management device that manages network slices in a cellular network, a step of inputting packet information indicating the operation of the packets of the application in the cellular network using a learned AI model for application identification and outputting the application name; a step of inputting the application name using a learned AI model for communication control and outputting communication control information indicating information for communication control of the packets in the cellular network; when a packet including the name of a new application is received, at least using the name of the new application, retraining the learned AI model for application identification and the learned AI model for communication control. Management method.

Explanation of Signs

[0146] 1: Communication system 10: Cellular network 20: SDN 30: RAN 40: CN 100: UE 102: Transmission unit 103: Control unit 110: Node 111: Transmission unit 113: Control unit 120: CN device 121: Reception unit 122: Transmission unit 123: Control unit 130: Network slice management device 131: RAN Slice Controller 132: Transport Slice Controller 133: Core Slice Controller 136: Receiver 137: Transmitter 138: Control Unit 1380: Model Learning Unit 1381: Model Inference Unit

Claims

1. A network slice management device for managing network slices in a cellular network, comprising: Using a trained AI model for application identification, input packet information indicating the operation of a packet of an application in the cellular network, and output an application name; a control unit that inputs the application name and outputs communication control information indicating information for communication control in the cellular network of a packet using a trained AI model for communication control; When the control unit receives a packet including a name of a new application, the control unit re-learns the trained AI model for application identification and the trained AI model for communication control using at least the name of the new application. Network slice management device.

2. When the control unit receives the packet, the control unit re-trains the trained AI model for application identification using the name of the new application and packet information of the packet. The network slice management device according to claim 1.

3. When the control unit receives the packet, the control unit re-trains the trained AI model for communication control using the name of the new application and the communication control information set when the new application is executed. The network slice management device according to claim 1.

4. The communication control information includes information regarding a QoS flow established between the user equipment and the core network equipment. The network slice management device according to claim 1.

5. The packet information indicates at least any one of a data length of data included in a payload portion of the packet, a transmission interval of the packet, a number of user devices connected to the cellular network when executing the application, and an amount of data accumulated in a transmission queue of a core network device in the cellular network. The network slice management device according to claim 1.

6. A management method in a network slice management device that manages network slices in a cellular network, comprising: A step of inputting packet information indicating an operation of a packet of an application in the cellular network and outputting an application name using a trained AI model for application identification; A step of inputting the application name and outputting communication control information indicating information for communication control of a packet in the cellular network using a trained AI model for communication control; When a packet including a name of a new application is received, the trained AI model for application identification and the trained AI model for communication control are re-trained using at least the name of the new application. Management method.