Apparatus and method for controlling overload of entity by using load information

Abstract

The device of the first network function (NF) may comprise: a memory including instructions; a transceiver; and at least one processor. The instructions, when executed individually or collectively by the at least one processor, may cause the apparatus to: acquire load information for indicating a load of each NF in an NF group including a second NF and a third NF; determine whether each NF in the NF group is in a first state indicating an overload of the NF by using the load information; on the basis of determining that the second NF is in the first state and the third NF is in a second state, change a ratio between the second NF and the third NF for distributing a plurality of calls from the first ratio to a second ratio different from the first ratio; and on the basis of the changed second ratio, allocate, to the second NF, second calls less than first calls according to the first ratio among the plurality of calls, and allocate, to the third NF, fourth calls more than third calls according to the first ratio among the plurality of calls.

Description

Device and method for controlling overload of an entity using load information

[0001] The following descriptions relate to a device and method for controlling overload of an entity using load information.

[0002] In a communication system, a core network (or core network) may include multiple entities. For example, the multiple entities may include network functions (NFs). Some of the multiple entities may select other entities to provide services.

[0003] A device of a first network function (NF) may include a memory including instructions. The device may include a transceiver. The device may include a processor. The instructions, when executed by the processor, may be configured to cause the device to obtain load information indicating a load of each NF in a NF group including a second NF and a third NF. The instructions, when executed by the processor, may be configured to cause the device to determine, using the load information, whether each NF in the NF group is in a first state indicating an overload of the NF. The instructions, when executed by the processor, may be configured to cause the device to change a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio, based on determining that the second NF is in the first state and the third NF is in a second state different from the first state. The instructions, when executed by the processor, may be configured to cause the device to assign, based on the changed second ratio, a second number of calls less than a first number of calls according to the first ratio among the plurality of calls to the second NF, and to assign, based on the changed second ratio, a fourth number of calls more than a third number of calls according to the first ratio among the plurality of calls to the third NF.

[0004] A method performed by a device of a first NF (network function) may include an operation of obtaining load information for indicating a load of each NF in a NF group including a second NF and a third NF. The method may include an operation of determining, using the load information, whether each NF in the NF group is in a first state indicating an overload of the NF. The method may include an operation of changing a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio based on determining that the second NF is in the first state and the third NF is in a second state different from the first state. The method may include, based on the changed second ratio: an operation of allocating second calls, which are reduced from first calls to be distributed to the second NF among the plurality of calls, to the second NF, and an operation of allocating fourth calls, which are increased from third calls to be distributed to the third NF among the plurality of calls, to the third NF.

[0005] A non-transitory computer-readable storage medium may store one or more programs including instructions that, when executed by a processor of a device of a first network function (NF) including a transceiver, cause the device to obtain load information for indicating a load of each NF in a group of NFs including a second NF and a third NF. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when executed by the processor, cause the device to determine, using the load information, whether each NF in the group of NFs is in a first state indicating an overload of the NF. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when executed by the processor, cause the processor to change a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio based on determining that the second NF is in the first state and the third NF is in a second state different from the first state. The non-transitory computer-readable storage medium may store one or more programs including instructions that, when executed by the processor, cause the processor to: assign second calls, which are reduced from first calls to be distributed to the second NF among the plurality of calls, to the second NF; and assign fourth calls, which are increased from third calls to be distributed to the third NF among the plurality of calls, to the third NF.

[0006] Figure 1 illustrates an example of a communication system.

[0007] Figure 2a illustrates an example of the functional configuration of a base station in a communication system.

[0008] Figure 2b illustrates an example of the functional configuration of a terminal in a communication system.

[0009] Figure 2c illustrates an example of the functional configuration of a core network entity in a communication system.

[0010] Figure 3 illustrates an example of user plane functions (UPFs) deployed within multiple serving areas.

[0011] Figure 4 illustrates an example of an operational flow for a method of controlling overload of NF.

[0012] Figure 5a illustrates an example of how to control overload of NFs within one serving area.

[0013] Figure 5b illustrates an example of a method for controlling overload of NFs within multiple serving areas.

[0014] Figure 6a illustrates an example of a method for learning an artificial intelligence model (AI model) based on load information of multiple time intervals.

[0015] Figure 6b illustrates an example of a method for obtaining predictive load information using an artificial intelligence model.

[0016] Figures 7a and 7b illustrate examples of graphs representing predicted load information and collected load information over time.

[0017] Figure 8 illustrates an example of an operational flow for how a first NF (network function) controls overload of a second NF of an NF group.

[0018] The terms used in this disclosure are used only to describe specific embodiments and may not be intended to limit the scope of other embodiments. The singular expression may include plural expressions unless the context clearly indicates otherwise. Terms used herein, including technical or scientific terms, may have the same meaning as commonly understood by those of ordinary skill in the art described in this disclosure. Terms defined in general dictionaries among the terms used in this disclosure may be interpreted as having the same or similar meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined in this disclosure. In some cases, even if a term is defined in this disclosure, it cannot be interpreted to exclude embodiments of the present disclosure.

[0019] The various embodiments of the present disclosure described below illustrate a hardware-based approach as an example. However, since the various embodiments of the present disclosure include techniques utilizing both hardware and software, the various embodiments of the present disclosure do not exclude a software-based approach.

[0020] In the following description, terms referring to signals (e.g., packet, message, signal, information, signaling), terms referring to network entities (e.g., entity, NF (network function), node, NF group, entity group), terms for operational states (e.g., step, operation, procedure), terms referring to data (e.g., packet, message, user stream, information, bit, symbol, codeword), terms referring to channels, and terms referring to components of devices, etc., are examples for convenience of explanation. Therefore, the present disclosure is not limited to the terms described below, and other terms having equivalent technical meanings may be used. In addition, terms such as '... part', '... device', '... thing', '... body', etc. used below may mean at least one shape structure or a unit that processes a function.

[0021] In addition, in the present disclosure, expressions such as "more than" or "less than" may be used to determine whether a specific condition is satisfied or fulfilled, but this is merely a description for expressing an example and does not exclude descriptions such as "more than" or "less than." A condition described as "more than" may be replaced with "more than," a condition described as "less than" may be replaced with "less than," and a condition described as "more than and less than" may be replaced with "more than and less than." In addition, hereinafter, "A" to "B" mean at least one of elements from A (including A) to B (including B). hereinafter, "C" and / or "D" mean at least one of "C" or "D," that is, including {"C", "D", "C" and "D"}.

[0022] Although the present disclosure describes various embodiments using terms used in some communication standards (e.g., 3rd Generation Partnership Project (3GPP), extensible radio access network (xRAN), open radio access network (O-RAN)), these are merely examples for illustrative purposes. The various embodiments of the present disclosure can be easily modified and applied to other communication systems.

[0023] For example, in a communication system (e.g., LTE or 5G), the gateway-control plane (GW-C) (or session management function (SMF), network repository function (NRF)) can select the gateway-user plane (GW-U) (or user plane function (UPF)) for processing the user plane. At this time, the GW-C (or SMF, NRF) can select the GW-U (or UPF) by considering the location information, service, or capacity of the GW-U (or UPF). As edge computing technology is utilized, the GW-U (or UPF) can be gradually miniaturized and deployed in a distributed manner near the base station. A network with the above deployment can be referred to as a distributed network. The above-described distributed network has the effect of reducing delay time and improving service quality by reducing the distance between the GW-U (or UPF), which is an entity for processing the user plane, and the base station. However, the above-described distributed network has a complex structure in that it requires a large number of miniaturized entities (e.g., GW-U or UPF), and the capacity of the entities is small in a specific area (or region), which may cause overload of the entities. Hereinafter, embodiments of the present disclosure propose a method of detecting overload of entities using load information (or load data) of entities in a specific area, and reducing the overload through entities in the specific area or another area when overload of a specific entity among the entities is detected.

[0024] Hereinafter, in the present disclosure, the entity may be referred to as a component of a network (e.g., a radio access network (RAN) or a core network). For example, the entity may be referred to as a network function (NF) or a node. The overload may indicate a state in which traffic (or load) of the entity is above a certain level based on a comparison between at least some of the parameters in the load information of the entity and a reference value (or reference level) according to the at least some of the parameters. Hereinafter, the state indicating the overload may be referred to as a first state. The state that is not an overload (i.e., a state in which the traffic is below the certain level) may be referred to as a second state.

[0025] Figure 1 illustrates an example of a communication system.

[0026] Referring to FIG. 1, the communication system may include a radio access network (RAN) (102) and a core network (CN) (104).

[0027] The wireless access network (102) is a network directly connected to the terminal (120) and is an infrastructure that provides wireless access to the terminal (120). The wireless access network (102) includes a set of multiple base stations, including the base station (110), and the multiple base stations can communicate with each other through interfaces formed between them. At least some of the interfaces between the multiple base stations may be wired or wireless.

[0028] The base station (110) may have a structure divided into a central unit (CU) and a distributed unit (DU). In this case, one CU may control multiple DUs. In addition to the base station, the base station (110) may include an 'access point (AP)', a 'next generation node B (gNB)', and a '5G node (5 th The terminal (120) may be referred to as a 'generation node', a 'wireless point', a 'transmission / reception point (TRP)', or other terms having an equivalent technical meaning thereto. The terminal (120) connects to the wireless access network (102) and communicates with the base station (110) through a wireless channel. The terminal (120) may be referred to as a 'user equipment (UE)', a 'mobile station', a 'subscriber station', a 'remote terminal', a 'wireless terminal', a 'user device', or other terms having an equivalent technical meaning thereto.

[0029] The core network (104) is a network that manages the entire system, controls the wireless access network (102), and can process data and control signals for terminals (120) transmitted and received through the wireless access network (102). The core network (104) can perform various functions, such as control of the user plane and control plane, processing of mobility, management of subscriber information, billing, and interworking with other types of systems (e.g., long term evolution (LTE) system).

[0030] To perform the various functions described above, the core network (104) may include a number of functionally separated entities having different network functions (NFs). The entities may be referred to as NFs or nodes. For example, the core network (104) may include an access and mobility management function (AMF) (130a), a session management function (SMF) (130b), a user plane function (UPF) (130c), a policy and charging function (PCF) (130d), a network repository function (NRF) (130e), a user data management (UDM) (130f), a network exposure function (NEF) (130g), a unified data repository (UDR) (130h), or a network data analytics function (NWDAF) (130i). However, embodiments of the present disclosure are not limited thereto. For example, the core network (104) may include more NFs, or may include less than at least one of the NFs illustrated in FIG. 1.

[0031] For example, although each of the entities of the core network (104) of FIG. 1 is depicted as independently existing entities, the embodiments of the present disclosure are not limited thereto. For example, a specific NF may be included in another NF. For example, the SMF (130b) may include the NWDAF (130i). The NWDAF (130i) included in the SMF (130b) may be configured with hardware, software, or hardware and software that perform the functions of the NWDAF (130i).

[0032] The terminal (120) can be connected to the AMF (130a) that is connected to the wireless access network (102) and performs the mobility management function of the core network (104). The AMF (130a) can perform connection to the wireless access network (102) and mobility management of the terminal (120). The SMF (130b) can manage sessions. The AMF (130a) is connected to the SMF (130b), and the AMF (130a) can route session-related messages for the terminal (120) to the SMF (130b). The SMF (130b) can be connected to the UPF (130c) to allocate resources of the user plane to be provided to the terminal (120), and can establish a tunnel for transmitting data between the base station (110) and the UPF (130c). PCF (130d) can control information related to policies and charging for sessions used by terminal (120). NRF (130e) can store information on NFs installed in a mobile communication operator network and perform a function of notifying the stored information. NRF (130e) can be connected to all NFs. When each NF starts operating in the operator network, it can notify NRF (130e) that the corresponding NF is operating in the network by registering with NRF (130e). UDM (130f) is an NF that performs a role similar to HSS (home subscriber server) of a 4G network and can store subscription information of terminal (120) or context used by terminal (120) in the network. NEF (130g) can be connected to a third party (3 rd It can play a role in connecting a third-party server (or a third-party application (3)) and an NF within a 5G mobile communication system. For example, a third-party server (or a third-party application (3)) rdThe NEF (130g) may be an application function (AF). In addition, the NEF (130g) may provide data to the UDR (130h), update data, or acquire data. The UDR (130h) may store subscription information of the terminal (120), store policy information, store data exposed to the outside, or store data for a third-party application (3 rd The UDR (130h) can perform the function of storing information required for a party application. In addition, the UDR (130h) can also perform the function of providing the stored data to other NFs. The NWDAF (130i) can provide the function of collecting and analyzing network data. For example, the NWDAF (130i) can acquire data from other NFs and perform inference through analysis or learning based on the acquired data.

[0033] Figure 2a illustrates an example of the functional configuration of a base station in a communication system.

[0034] The configuration illustrated in Fig. 2a can be understood as the configuration of a base station (110). Terms such as '... unit', '... unit', etc. used hereinafter mean a unit that processes at least one function or operation, and this can be implemented by hardware, software, or a combination of hardware and software.

[0035] Referring to FIG. 2a, the base station (110) may include a wireless communication unit (211), a backhaul communication unit (212), a storage unit (213), and a control unit (214).

[0036] The wireless communication unit (211) performs functions for transmitting and receiving signals via a wireless channel. For example, the wireless communication unit (211) performs a conversion function between baseband signals and bit streams according to the physical layer specifications of the system. For example, when transmitting data, the wireless communication unit (211) generates complex symbols by encoding and modulating the transmitted bit stream. In addition, when receiving data, the wireless communication unit (211) restores the received bit stream by demodulating and decoding the baseband signal.

[0037] In addition, the wireless communication unit (211) upconverts a baseband signal into an RF (radio frequency) band signal and transmits it through an antenna, and downconverts an RF band signal received through the antenna into a baseband signal. To this end, the wireless communication unit (211) may include a transmission filter, a reception filter, an amplifier, a mixer, an oscillator, a digital to analog convertor (DAC), an analog to digital convertor (ADC), etc. In addition, the wireless communication unit (211) may include a plurality of transmission and reception paths. Furthermore, the wireless communication unit (211) may include at least one antenna array composed of a plurality of antenna elements.

[0038] In terms of hardware, the wireless communication unit (211) may be composed of a digital unit and an analog unit, and the analog unit may be composed of a plurality of sub-units depending on operating power, operating frequency, etc. The digital unit may be implemented with at least one processor (e.g., a digital signal processor (DSP)).

[0039] The wireless communication unit (211) transmits and receives signals as described above. Accordingly, all or part of the wireless communication unit (211) may be referred to as a "transmitter," a "receiver," or a "transceiver." Furthermore, in the following description, transmission and reception performed via a wireless channel are used to mean that the wireless communication unit (211) performs the processing described above.

[0040] The backhaul communication unit (212) provides an interface for performing communication with other nodes within the network. That is, the backhaul communication unit (212) converts a bit string transmitted from the base station (110) to another node, such as another access node, another base station, an upper node, a core network, etc., into a physical signal, and converts a physical signal received from another node into a bit string.

[0041] The storage unit (213) stores data such as basic programs, application programs, and setting information for the operation of the base station (110). The storage unit (213) may be composed of volatile memory, nonvolatile memory, or a combination of volatile and nonvolatile memory. In addition, the storage unit (213) provides stored data upon request from the control unit (214).

[0042] The control unit (214) controls the overall operations of the base station (110). For example, the control unit (214) transmits and receives signals through the wireless communication unit (211) or the backhaul communication unit (212). In addition, the control unit (214) records and reads data in the storage unit (213). In addition, the control unit (214) can perform functions of a protocol stack required by the communication standard. According to another implementation example, the protocol stack can be included in the wireless communication unit (211). For this purpose, the control unit (214) can include at least one processor. According to various embodiments, the control unit (214) can control to perform synchronization using a wireless communication network. For example, the control unit (214) can control the base station (110) to perform operations according to various embodiments described below.

[0043] For example, the at least one processor of the control unit (214) may include various processing circuits and / or multiple processors. For example, the term "processor" as used herein, including in the claims, may include various processing circuits including at least one processor, one or more of which may be configured to individually and / or collectively perform the various functions described below in a distributed manner. As used herein, when "processor," "at least one processor," and "one or more processors" are described as being configured to perform various functions, these terms encompass, for example, and without limitation, situations where one processor performs some of the recited functions and other processor(s) perform other parts of the recited functions, and also situations where one processor may perform all of the recited functions. Additionally, the at least one processor may include a combination of processors that perform the various functions enumerated / disclosed, for example, in a distributed manner. At least one processor is capable of executing program instructions to accomplish or perform various functions.

[0044] Figure 2b illustrates an example of the functional configuration of a terminal in a communication system.

[0045] The configuration illustrated in Fig. 2b can be understood as the configuration of a terminal (120). Terms such as '... unit', '... unit', etc. used hereinafter mean a unit that processes at least one function or operation, and this can be implemented by hardware, software, or a combination of hardware and software.

[0046] Referring to FIG. 2b, the terminal (120) may include a communication unit (221), a storage unit (222), and a control unit (223).

[0047] The communication unit (221) performs functions for transmitting and receiving signals via a wireless channel. For example, the communication unit (221) performs a conversion function between a baseband signal and a bit stream according to the physical layer specifications of the system. For example, when transmitting data, the communication unit (221) generates complex symbols by encoding and modulating a transmission bit stream. In addition, when receiving data, the communication unit (221) restores a reception bit stream by demodulating and decoding a baseband signal. In addition, the communication unit (221) upconverts a baseband signal into an RF band signal and then transmits it through an antenna, and downconverts an RF band signal received through the antenna into a baseband signal. For example, the communication unit (221) may include a transmission filter, a reception filter, an amplifier, a mixer, an oscillator, a DAC, an ADC, etc.

[0048] In addition, the communication unit (221) may include a plurality of transmission and reception paths. Furthermore, the communication unit (221) may include at least one antenna array composed of a plurality of antenna elements. In terms of hardware, the communication unit (221) may be composed of digital circuits and analog circuits (e.g., radio frequency integrated circuits (RFIC)). Here, the digital circuits and analog circuits may be implemented in a single package. In addition, the communication unit (221) may include a plurality of RF chains. Furthermore, the communication unit (221) may perform beamforming.

[0049] The communication unit (221) transmits and receives signals as described above. Accordingly, all or part of the communication unit (221) may be referred to as a "transmitter," a "receiver," or a "transmitting and receiving unit." Furthermore, in the following description, transmission and reception performed via a wireless channel are used to mean that the communication unit (221) performs the processing described above.

[0050] The storage unit (222) stores data such as basic programs, application programs, and setting information for the operation of the terminal (120). The storage unit (222) may be composed of volatile memory, non-volatile memory, or a combination of volatile and non-volatile memory. In addition, the storage unit (222) provides stored data upon request from the control unit (223).

[0051] The control unit (223) controls the overall operations of the terminal (120). For example, the control unit (223) transmits and receives signals through the communication unit (221). In addition, the control unit (223) records and reads data in the storage unit (222). In addition, the control unit (223) can perform the functions of the protocol stack required by the communication standard. To this end, the control unit (223) may include at least one processor or microprocessor, or may be a part of a processor. In addition, a part of the communication unit (221) and the control unit (223) may be referred to as a CP (communication processor). According to various embodiments, the control unit (223) may control to perform synchronization using a wireless communication network. For example, the control unit (223) may control the terminal to perform operations according to various embodiments described below.

[0052] For example, the at least one processor of the control unit (223) may include various processing circuits and / or multiple processors. For example, the term "processor" as used herein, including in the claims, may include various processing circuits including at least one processor, and one or more of the at least one processors may be configured to perform the various functions described below individually and / or collectively in a distributed manner. As used herein, when "processor," "at least one processor," and "one or more processors" are described as being configured to perform various functions, these terms encompass, for example, and without limitation, situations where one processor performs some of the recited functions and other processor(s) perform other parts of the recited functions, and also situations where one processor can perform all of the recited functions. Additionally, the at least one processor may include a combination of processors that perform the various functions enumerated / disclosed, for example, in a distributed manner. At least one processor is capable of executing program instructions to accomplish or perform various functions.

[0053] Figure 2c illustrates an example of the functional configuration of a core network entity in a communication system.

[0054] The core network entity (130) illustrated in FIG. 2c may be understood as a configuration of a device having at least one function among the AMF (130a), the SMF (130b), the UPF (130c), the PCF (130d), the NRF (130e), the UDM (130f), the NEF (130g), the UDR (130h), or the NWDAF (130i) of FIG. 1. However, the embodiments of the present disclosure are not limited thereto. For example, the core network entity (130) of FIG. 2c may be understood as an example of a functional configuration for an entity other than the above-described examples. The entity may be referred to as a node or a network function (NF). Terms such as '... unit', '... unit', etc. used hereinafter mean a unit that processes at least one function or operation, and this may be implemented by hardware, software, or a combination of hardware and software.

[0055] Referring to FIG. 2c, the core network entity (130) may include a communication unit (231), a storage unit (232), and a control unit (233).

[0056] The communication unit (231) provides an interface for communicating with other devices within the network. That is, the communication unit (231) converts a bit string transmitted from the core network entity (130) to another device into a physical signal, and converts a physical signal received from another device into a bit string. That is, the communication unit (231) can transmit and receive signals. Accordingly, the communication unit (231) may be referred to as a modem, a transmitter, a receiver, or a transceiver. In this case, the communication unit (231) enables the core network entity (130) to communicate with other devices or systems via a backhaul connection (e.g., wired backhaul or wireless backhaul) or via a network.

[0057] The storage unit (232) stores data such as basic programs, application programs, and setting information for the operation of the core network entity (130). The storage unit (232) may be composed of volatile memory, non-volatile memory, or a combination of volatile and non-volatile memory. In addition, the storage unit (232) provides the stored data upon request from the control unit (233).

[0058] The control unit (233) controls the overall operations of the core network entity (130). For example, the control unit (233) transmits and receives signals through the communication unit (231). In addition, the control unit (233) records and reads data in the storage unit (232). For this purpose, the control unit (233) may include at least one processor. According to various embodiments, the control unit (233) may control synchronization using a wireless communication network. For example, the control unit (233) may control the core network entity (130) to perform operations according to various embodiments described below.

[0059] For example, the at least one processor of the control unit (233) may include various processing circuits and / or multiple processors. For example, the term "processor" as used herein, including in the claims, may include various processing circuits including at least one processor, wherein one or more of the at least one processors may be configured to individually and / or collectively perform the various functions described below in a distributed manner. As used herein, when "processor," "at least one processor," and "one or more processors" are described as being configured to perform various functions, these terms encompass, for example, and without limitation, situations where one processor performs some of the recited functions and other processor(s) perform other parts of the recited functions, and also situations where one processor may perform all of the recited functions. Additionally, the at least one processor may include a combination of processors that perform the various functions enumerated / disclosed, for example, in a distributed manner. At least one processor is capable of executing program instructions to accomplish or perform various functions.

[0060] Figure 3 illustrates an example of user plane functions (UPFs) deployed within multiple serving areas.

[0061] FIG. 3 illustrates an example (300) in which a core network entity (130) performs selection of UPFs arranged within a plurality of serving areas (310, 320, 330, 340). In the example (300), the core network entity (130) may include an SMF (e.g., an SMF (130b) or an NRF (130e) of FIG. 1) that performs the selection of the UPFs. However, the example (300) is merely an example for convenience of explanation, and the embodiments of the present disclosure are not limited thereto. For example, when the targets of the selection are a plurality of SMFs, the core network entity (130) may include an AMF (e.g., an AMF (130a) of FIG. 1).

[0062] Referring to example (300), each of the plurality of serving areas (310, 320, 330, 340) may be referred to as a specific area where a service is provided. For example, the specific area may be determined based on location information. For example, the specific area may be referred to as a tracking area (TA). For example, the location information may include a tracking area indicator (TAI), a tracking area code (TAC), or an identifier (ID) of an NF group within the serving area. Hereinafter, the serving area may be referred to as an NF area or an entity area.

[0063] Referring to example (300), each of the plurality of serving areas (310, 320, 330, 340) may include at least one entity. For example, the first serving area (310) may include three UPFs (311, 312, 313). For example, the second serving area (320) may include three UPFs (321, 322, 323). For example, the third serving area (330) may include two UPFs (331, 332). For example, the fourth serving area (340) may include four UPFs (341, 342, 343, 344). The at least one entity within each serving area may be referred to as a group. For example, UPFs (311, 312, 313) may be referred to as a first group, UPFs (321, 322, 323) as a second group, UPFs (331, 332, 333) as a third group, and UPFs (341, 342, 343, 344) as a fourth group. At this time, the group may be referred to as an entity group, an NF group, an entity set, or an NF set. For example, entities within the group may have the same location information. For example, the location information may include a TAI, a TAC, or an ID (identifier) ​​of the group.

[0064] In example (300), the core network entity (130) may select a UPF group (e.g., the first group, the second group, the third group, or the fourth group) based on service-related parameters including a data network name (DNN) and a single network slice selection assistance information (S-NSSAI), a location parameter including a tracking area (TA) and a TA list, and a capacity parameter. At this time, the location parameter may be used to reduce delay time by selecting a UPF close to the location of a user equipment (e.g., the terminal (120) of FIG. 1) to be provided with the service. This is because, as edge computing is introduced and UPFs are miniaturized, they are deployed close to base stations and the capacity of the UPFs is limited. The core network entity (130) may implement edge computing by selecting at least one UPF within a UPF group that is deployed closely using the location information of the terminal.

[0065] When implementing the above edge computing, utilizing miniaturized UPFs may improve service quality, but since the capacity (or available capacity) that can be processed within the UPF group is reduced, it may be difficult to distribute the overload in the serving area where the UPF group is located. Furthermore, since the number of UPFs included in the UPF group for the serving area increases, the business operator (or operator) may have difficulty managing and operating the UPFs.

[0066] Hereinafter, the present disclosure provides a method for detecting an overload occurring in a specific entity within an entity group or in entities of the entire entity group, and alleviating the overload of the entity or the entity group by distributing the load. For example, the device and method according to embodiments of the present disclosure can control the overload of the entities within the entity group by using load information collected (or acquired, received) from the entities within the entity group. For example, the SMF (or NRF) can collect load information of each UPF within a UPF group, and determine (or detect) an overload of each UPF within the UPF group based on the load information. The device and method according to embodiments of the present disclosure can distribute the load (or call) to another entity (or an entity of another entity group) within the entity group based on detecting the overload of the specific entity (or the entity group). In addition, the device and method according to embodiments of the present disclosure can reduce the reduction in service quality by re-allocating the load to the specific entity based on detecting that the overload of the specific entity has been resolved (or ended). In addition, the device and method according to embodiments of the present disclosure can prevent the overload by detecting the overload of the specific entity using an artificial intelligence model before the overload is detected and distributing the load to the other entities in advance. Accordingly, the device and method according to the embodiments of the present disclosure can improve the quality of service and reduce the operating cost of the network by having the core network entity (130) automatically detect overload and perform distribution without the intervention of a business operator (or operator).

[0067] Figure 4 illustrates an example of an operational flow for a method of controlling overload of NF.

[0068] At least a part of the method of FIG. 4 may be performed by a first NF that performs NF selection. For example, the first NF may include the core network entity (130) of FIG. 3. For example, the NF (or target NF) that is a target of the NF selection may be an example of an entity within the entity group of FIG. 3. For example, the NF may include a UPF (e.g., UPF (311) of FIG. 3) within the UPF group of FIG. 3 (e.g., UPFs (311, 312, 313) within the first serving area (310) of FIG. 3). In this case, the first NF may include an SMF or an NRF. For example, at least a part of the method may be controlled by a processor of the first NF (e.g., the control unit (233) of FIG. 2c). In the following examples, each operation may be performed sequentially, but is not necessarily performed sequentially. For example, the order of each operation may be changed, and at least two operations may be performed in parallel.

[0069] In operation (400), the first NF may obtain load information for indicating the load of each NF in the NF group. For example, the first NF may collect (or receive) the load information from each NF in the NF group located within one serving area. The NF group may include a second NF and a third NF. For example, the second NF may represent an NF of the same type as the third NF. For example, when the first NF is an SMF or an NRF, the NF group may include a plurality of UPFs. In this case, the second NF may be a first UPF, and the third NF may be a second UPF different from the first UPF.

[0070] In one embodiment, the load information may include information indicating the load of each NF of the NF group. For example, the load information may include a plurality of parameters for identifying the load. For example, the first NF may collect load information of the second NF and load information of the third NF.

[0071] For example, the load information may include a factor related to a service provided by the second NF (or the third NF), a user plane factor, and a control plane factor. For example, the plurality of parameters may include at least one of a factor related to the service, the user plane factor, or the control plane factor. Hereinafter, for convenience of explanation, the load information of the second NF is described as an example, but the embodiments of the present disclosure are not limited thereto. For example, the first NF may collect load information of each NF of the NF group including the second NF.

[0072] For example, the service-related element may include at least one of the number of user equipment (UEs), the number of protocol data unit (PDU) sessions, or the number of quality of service (QoS) flows associated with the second NF. For example, the number of UEs associated with the second NF may include the maximum number of UEs that the second NF can service or the number of UEs that the second NF is providing service to. For example, the number of PDU sessions and the number of QoS flows may indicate the number of PDU sessions and QoS flows that the second NF provides service to. In addition, for example, the service-related element may include at least one of information indicating a load of a central processing unit (CPU), a memory, or a disk of the second NF.

[0073] For example, the user plane element may include at least one of traffic, packet drop rate, or IP (internet protocol) pool usage. For example, the traffic (or throughput) may include the amount of traffic used per unit time and performance capacity information of the maximum serviceable traffic. The packet drop rate may include the number and size of packets dropped due to failure in data transmission and reception. The packet drop rate may be referred to as a drop packet. For example, the IP pool usage may include the number (or usage) of IPs allocated to a specific terminal within the IP pool.

[0074] For example, the control plane element may include information about transactions per second (TPS) or calls. The TPS may include the number of messages per second. The information about the call may include information about attempts, successes, failures, and failure reasons, depending on the call procedure.

[0075] According to one embodiment, the plurality of parameters included in the load information may be determined based on the second NF. For example, the plurality of parameters may be changed based on the function or role of the second NF.

[0076] In one embodiment, the first NF may periodically obtain the load information from the second NF. For example, the first NF may obtain the load information at intervals of a specified length. For example, the specified length may be 5 minutes. However, the embodiments of the present disclosure are not limited thereto.

[0077] In operation (405), the first NF may determine whether each NF in the NF group is in a first state indicating an overload. For example, the first NF may determine, using the load information, whether the load state of each NF in the NF group is the first state. For example, the first state may be referred to as an overload state, an overload mode, or a load limit state. Additionally, a second state different from the first state may be referred to as a default state, a non-overload state, or a default mode.

[0078] According to one embodiment, the first NF may determine whether the load state of the second NF is the first state based on a comparison between a specific parameter and at least one reference value set for the specific parameter. For example, the specific parameter may be included in the plurality of parameters of the load information received from the second NF. The relationship between the specific parameter and the at least one reference value may be referenced in the table below.

[0079]

[0080] Referring to the above table, the AUDIT_TIME may represent a period (e.g., 60 seconds) for collecting the load information, the MINOR_REDUCE may represent a limitation ratio (e.g., 30%) of a call when the level of the parameter is the first level (e.g., minor), the MAJOR_REDUCE may represent a limitation ratio (e.g., 50%) of a call when the level of the parameter is the second level (e.g., major), and the CRITICAL_REDUCE may represent a limitation ratio (e.g., 70%) of a call when the level of the parameter is the third level (e.g., CRITICAL). The limitation ratio may be referred to as a reduction ratio. The period and the limitation ratio of the call in the above table are merely examples for the convenience of description, and the embodiments of the present disclosure are not limited thereto.

[0081] Referring to the above table, the CP_CPU_LEVEL may represent a parameter representing a load of a CPU for a control plane among the plurality of parameters, the MINOR of the CP_CPU_LEVEL may represent a reference value (e.g., 60%) for determining that the CP_CPU_LEVEL is the first level, the MAJOR of the CP_CPU_LEVEL may represent a reference value (e.g., 75%) for determining that the CP_CPU_LEVEL is the second level, and the CRITICAL of the CP_CPU_LEVEL may represent a reference value (e.g., 75%) for determining that the CP_CPU_LEVEL is the third level. For example, when the CP_CPU_LEVEL is 65%, the CP_CPU_LEVEL may be the first level. Alternatively, when the CP_CPU_LEVEL is 80%, the CP_CPU_LEVEL may be the second level. When the CP_CPU_LEVEL is 95%, the CP_CPU_LEVEL may be the third level. For example, the first level, the second level, and the third level may be used to recognize an overload state (i.e., the first state). For example, when the CP_CPU_LEVEL is 50%, the CP_CPU_LEVEL may be a non-overload state (i.e., the second state).

[0082] Referring to the above table, the UP_CPU_LEVEL may represent a parameter representing a CPU load for a user plane among the plurality of parameters, the SESS_CNT may represent the number of PDU sessions among the plurality of parameters, the THROUGHPUT may represent the traffic (or throughput) among the plurality of parameters, and the NI_IPPOOL may represent the IP pool among the plurality of parameters. Although not shown in the above table, at least one reference value (or reference level) may be set for each of the UP_CPU_LEVEL, the SESS_CNT, the THROUGHPUT, and the NI_IPPOOL.

[0083] The above table illustrates an example in which three reference values ​​are set for each parameter, but the embodiments of the present disclosure are not limited thereto. For example, at least one reference value may be set for each parameter, and a different number of reference values ​​may be set. For example, one reference value may be set for the UP_CPU_LEVEL parameter, and two reference values ​​may be set for the SESS_CNT parameter.

[0084] According to one embodiment, the first NF may recognize a level for each of the plurality of parameters of the load information received from the second NF. For example, the first NF may recognize that the load of the CPU for the control plane of the second NF is at the first level, the traffic of the first NF is at the second level, and the IP pool of the second NF is in the second state (i.e., a state that is not overloaded). According to one embodiment, the first NF may determine that the second NF is in the first state when at least one of the plurality of parameters is at the first level. In the example, even if the IP pool is in the second state, the load of the CPU for the control plane is at the first level, and the traffic is at the second level, so the first NF may determine the second NF to be in the second state. However, the embodiments of the present disclosure are not limited thereto. For example, the first NF may determine that the second NF is in the first state if at least one of the plurality of parameters is equal to or higher than the second level or is at the third level. In the following, for convenience of explanation, it is assumed that the state of the second NF is in the first state if at least one of the plurality of parameters is at the first level.

[0085] According to one embodiment, the first NF may determine whether each NF in the NF group is in the first state based on load information (hereinafter, predicted load information) predicted or expected using the load information. For example, the predicted load information may be generated by an artificial intelligence model using the latest (or most recently received) load information. For example, the artificial intelligence model may be trained through load information acquired for a plurality of time intervals prior to the timing at which the latest load information is acquired. Specific details of the artificial intelligence model trained through load information acquired for the plurality of time intervals are described below in FIG. 6A. Specific details of the predicted load information generated by the artificial intelligence model are described below in FIG. 6B. For example, the artificial intelligence model may be referred to as a load prediction model, a load prediction AI model, a statistical model, or a load data prediction model.

[0086] Referring to the above, the first NF may determine the first state of the second NF based on a result predicted using the load information when the prediction accuracy of the artificial intelligence model is relatively high (or when the predicted load information is relatively similar to the actual load information). For example, when a difference between a first load value generated from a first set of load information before a reference time among the plurality of time intervals and a second load value generated from a second set of load information after the reference time among the plurality of time intervals exceeds a reference difference, the first NF may determine that the prediction accuracy is relatively high.

[0087] For example, the first load value may represent a value identified based on load information predicted for after the reference point in time, based on the first set of load information, which is load information before the reference point in time. For example, the second load value may represent a value identified based on the second set of load information. For example, the first load value or the second load value may represent a summation of values ​​calculated based on scored conversion values ​​for each of the plurality of parameters and a ratio between the plurality of parameters. Each of the conversion values ​​may represent a scaled number so as to enable comparison between the plurality of parameters. For example, each of the conversion values ​​may be a value between 0 and 100. For example, since it is difficult to compare the size of the load of the CPU because the number of PDU sessions and the scaling are different, each of the plurality of parameters may be converted into a conversion value.

[0088] In operation (410), the first NF may determine whether all NFs within the NF group are in the first state. For example, the first NF may determine whether the state of each of all NFs is in the first state based on load information collected (or received) from each of all NFs within the NF group.

[0089] In operation (410), the first NF may perform operation (415) if at least one of all NFs of the NF group located within the serving area is not in the first state (i.e., if the at least one is in the second state). Alternatively, in operation (410), the first NF may perform operation (425) if all NFs of the NF group are in the first state.

[0090] In operation (415), the first NF may change a ratio within the NF group for distributing the plurality of calls. For example, the first NF may change a ratio within the NF group for distributing the plurality of calls in response to determining that at least one of all NFs within the NF group is in the second state. Hereinafter, for convenience of explanation, it is assumed that the second NF among the NFs within the NF group is in the first state, and the third NF among the NFs within the NF group is in the second state. The third NF may be referred to as an NF capable of distributing a call or an available NF.

[0091] For example, the plurality of calls may represent newly introduced calls (or loads) with respect to at least one terminal. For example, for the plurality of calls, the first NF may select an NF and assign to the selected NF. In other words, distributing the plurality of calls may indicate selecting at least one NF for the plurality of calls and assigning to the selected at least one NF. The serving area in which the NF group is located may indicate an area that can provide relatively optimal service to the at least one terminal compared to another serving area. In other words, the location information (e.g., TAI) of the at least one terminal and the location information of the serving area may be the same, or the location information of the serving area may indicate a location closest to the at least one terminal.

[0092] In one embodiment, the first NF can change the ratio from the first ratio to a second ratio that is different from the first ratio. For example, the first ratio can represent a distribution ratio that is set before collecting load information from each NF in the NF group. For example, the first NF can distribute calls to the second NF and the third NF based on the first ratio. For example, the first ratio can be 1:1 for the second NF and the third NF. However, the embodiments of the present disclosure are not limited thereto. The first NF can detect (or recognize) the degree of overload of the second NF in the first state. For example, the degree of overload can be determined using the parameter-specific level of the second NF recognized by the first NF, as described in the above-described table. For example, the degree of overload may represent the maximum level among the levels of each parameter of the second NF. Or, for example, the average level of the levels of each parameter of the second NF may represent. In the following, for convenience of explanation, it is assumed that the degree of overload of the second NF is the first level (or minor) and the call restriction ratio according to the first level is 50% (e.g., the MINOR_REDUCE is 50%). For example, the first NF may detect that the second NF in the first state is the first level. Upon detecting that the second NF is the first level, the first NF may calculate the second ratio in which the ratio for the second NF is reduced by 50%. For example, the second ratio may be 0.5:1 for the second NF and the third NF.

[0093] In operation (420), the first NF may assign the plurality of calls to the NF based on the changed ratio. For example, if the second ratio is used instead of the first ratio, the number of calls to be assigned (or distributed) to the second NF may be reduced. For example, according to the first ratio, first calls among the plurality of calls may be assigned to the second NF, and second calls among the plurality of calls may be assigned to the third NF. Conversely, if the second ratio changed from the first ratio is used, third calls, which are fewer than the first calls among the plurality of calls, may be assigned to the second NF, and fourth calls, which are more than the second calls among the plurality of calls, may be assigned to the third NF. In the above example, an example of distributing the plurality of calls to the second NF and the third NF of the NF group is described, but the embodiments of the present disclosure are not limited thereto. For example, the first NF can distribute the plurality of calls to the NF group including three or more NFs.

[0094] As described above, the first NF can resolve (or alleviate) the overload of the NF in the first state (e.g., the second NF) by changing the allocation ratio for the NFs of the NF group within the serving area. Alternatively, if all of the NFs of the NF group within the serving area are in the first state, the first NF can attempt to distribute to the NFs of the other serving area different from the serving area.

[0095] In operation (425), the first NF may obtain other load information to indicate the load of each NF in another NF group. For example, in response to determining that all NFs in the NF group are in the first state, the first NF may receive the other load information from each NF in the other NF group. However, the embodiments of the present disclosure are not limited thereto. For example, the first NF may determine a plurality of NF groups (e.g., the NF group and the other NF group) for providing a service to the at least one terminal, and may first determine whether an NF group capable of providing an optimal service among the plurality of NF groups is overloaded. At this time, the first NF may also collect load information of each of the plurality of NF groups.

[0096] For example, the other NF group may include NFs within the other serving area. For example, the other serving area may have different location information than the serving area in which the NF group is deployed. The location information may include a TAI, a TAC, or an ID of the NF group. Alternatively, for example, the other serving area may be larger in size than the serving area and may be configured to include the serving area. However, the embodiments of the present disclosure are not limited thereto. Specific details related thereto are described with reference to FIG. 5B.

[0097] For example, specific details about the other load information may be substantially identically applied to the content about the load information. For example, the other load information may include a plurality of parameters for identifying the load of each NF within the other NF group. For example, if the other NF group includes a fourth NF, the first NF may collect (or receive) other load information of the fourth NF from the fourth NF.

[0098] In operation (430), the first NF may assign the plurality of calls to NFs within the other NF group. For example, the first NF may determine, based on the other load information, whether each NF within the other NF group is in the first state. For example, the first NF may determine, using the other load information, the state of the fourth NF within the other NF group as the second state. For example, the first NF may assign at least some of the plurality of calls to the fourth NF. In other words, the first NF may select the fourth NF as the NF that will provide the at least some of the plurality of calls.

[0099] As described above, the first NF may allocate at least a portion of the plurality of calls to the fourth NF of the other NF group, not the NF group. Since the other NF group is not the NF group having the same location information as the location information of the at least one terminal, the quality of the service provided by the fourth NF may be relatively lowered. Accordingly, the first NF may perform call redistribution to the at least one NF when the first state for at least one NF of the NF group is resolved. For example, assume that the at least one NF is the second NF. According to one embodiment, the first NF may detect (or recognize) that the state of the second NF is changed from the first state to the second state after allocating the at least a portion to the fourth NF. For example, the change from the first state to the second state may be determined based on load information collected (or received, acquired) from the second NF. For example, the first NF may search for at least some of the at least some at the timing at which the change is detected. The first NF may determine an NF to provide service for the at least some as the second NF of the NF group from the fourth NF of the other NF group. For example, the first NF may cause to perform a PDU session reestablishment procedure to allocate the at least some to the second NF at the timing at which the change is detected. Alternatively, for example, the first NF may cause to perform the PDU session reestablishment procedure when it detects that the at least one terminal transitions to an idle mode after the timing at which the change is detected.This is because the at least one terminal can perform continuous service (e.g., call) over the at least part of the time. According to one embodiment, the first NF may cause the second NF of the NF group to perform the PDU session re-establishment procedure for a specified number of calls during a specified time period, since an overload of the second NF may occur again when a plurality of PDU session re-establishment procedures are performed. For example, the specified time period and the specified number may be determined based on load information received from the second NF. For example, when the load level of the second NF is low (e.g., when the number of currently assigned calls is relatively small), the specified time period may be shortened and the specified number may be increased. Alternatively, for example, when the second NF is not overloaded but the load level is high (e.g., when the number of currently assigned calls is relatively large), the specified time period may be lengthened and the specified number may be decreased.

[0100] As described above, the device and method according to embodiments of the present disclosure can detect that an overload occurs in a specific NF (e.g., the second NF) within an NF group or in NFs of the entire NF group, and relieve the overload of the NF or the NF group by distributing the load. The device and method according to embodiments of the present disclosure can, based on detecting the overload of the specific NF (or the NF group), distribute the load (or call) to another NF (e.g., the third NF) within the NF group (or an NF of another NF group (e.g., the fourth NF)). In addition, the device and method according to embodiments of the present disclosure can reduce a decrease in service quality by reallocating the load back to the specific NF (e.g., the second NF) based on detecting that the overload of the specific NF (e.g., the second NF) is relieved (or terminated). Furthermore, the devices and methods according to embodiments of the present disclosure can prevent overload by detecting overload in advance and distributing the load before the overload of the specific NF is detected using an artificial intelligence model. Accordingly, the devices and methods according to embodiments of the present disclosure can improve service quality and reduce network operating costs by having the first NF automatically detect overload and perform distribution without the intervention of a business operator (or operator).

[0101] Figure 5a illustrates an example of how to control overload of NFs within one serving area.

[0102] FIG. 5A illustrates an example (500) of a method for distributing a call within an NF group based on determining that at least one NF within the NF group is in the second state in operation (410). For example, the NF group may include NFs located within one serving area (510). In example (500), an example is illustrated where the NF group includes UPFs (511, 512, 513, 514), but embodiments of the present disclosure are not limited thereto.

[0103] Referring to example (500), a serving area (510) may include four UPFs (511, 512, 513, 514). The serving area (510) may represent an area for providing services to a terminal (120). In example (500), the serving area (510) is illustrated as having four UPFs (511, 512, 513, 514) providing services to one terminal (120), but the embodiments of the present disclosure are not limited thereto. For example, the serving area (510) may include three or fewer or five or more UPFs. Alternatively, for example, the serving area (510) may provide services to a plurality of terminals. For example, UPFs (511, 512, 513, 514) within the serving area (510) may have the same location information (e.g., TAI, TAC). At this time, the location information of the terminal (120) may also be the same as the location information of each of the UPFs (511, 512, 513, 514) within the serving area (510). In other words, the location information of the UPFs (511, 512, 513, 514) within the serving area (510) may indicate an area including the location of the terminal (120). Although not shown in the example (500), the entity (or NF) that distributes calls to the UPFs (511, 512, 513, 514) may be an SMF (or NRF). Below, for convenience of explanation, an example of the SMF distributing the call is described. However, the embodiments of the present disclosure are not limited thereto.

[0104] For example, the SMF can collect load information from each of the UPFs (511, 512, 513, 514). For example, the SMF can detect the status of each of the UPFs (511, 512, 513, 514) using the load information collected from each of the UPFs (511, 512, 513, 514). In other words, the SMF can determine whether each of the UPFs (511, 512, 513, 514) is in an overload state (or the first state) using the load information. However, the embodiments of the present disclosure are not limited thereto. For example, the SMF may determine whether each of the UPFs (511, 512, 513, 514) is in an overload state by using the load information (or predicted load information) predicted by the artificial intelligence model and the load information.

[0105] For example, the SMF may determine that the UPF (512) is in the first state and the UPFs (511, 513, 514) are in the second state. In response to determining that the UPF (512) is in the first state, the SMF may change a call distribution ratio between the UPF (512) and the UPFs (511, 513, 514). For example, the SMF may change the first ratio (e.g., UPF (511):UPF (512):UPF (513):UPF (514) = 1:1:1:1) to a second ratio. For example, the second ratio may be determined depending on the degree of overload of the UPF (512). For example, when the degree of overload of UPF (512) is the first level, the second ratio may have 1:0.5:1:1 based on the call restriction ratio according to the first level (e.g., 50%). Or, for example, when the degree of overload of UPF (512) is the second level, the second ratio may have 1:0.3:1:1 based on the call restriction ratio according to the second level (e.g., 70%).

[0106] In the example (500) of FIG. 5a, the SMF can relieve (or alleviate) the overload of the UPF (512), which is the first state, by distributing the call among the UPFs (511, 512, 513, 514) within the serving area (510). In the following FIG. 5b, an example of distribution for UPFs within multiple serving areas is described.

[0107] Figure 5b illustrates an example of a method for controlling overload of NFs within multiple serving areas.

[0108] FIG. 5b illustrates an example (550) of a method for distributing a call within another NF group based on determining that all NFs within the NF group are in the first state in operation (410). For example, the NF group may include NFs located within a first serving area (560). For example, the other NF group may include NFs within a second serving area (570) or NFs within a third serving area (580).

[0109] Referring to example (550), the first serving area (560) may include two UPFs (561, 562). The first serving area (560) may indicate an area with the highest quality of service provided to the terminal (120). For example, the UPFs (561, 562) within the first serving area (560) may have the same location information (e.g., TAI, TAC). In this case, the location information of the terminal (120) may also be the same as the location information of each of the UPFs (561, 562) within the first serving area (560). For example, the location information of the UPFs (561, 562) within the serving area (560) may indicate an area including the location of the terminal (120). Also, referring to example (550), the second serving area (570) may include four UPFs (571, 572, 573, 574). The second serving area (570) may represent an area with a higher quality of service, following the first serving area (560). For example, the UPFs (571, 572, 573, 574) within the second serving area (560) may have the same location information (e.g., TAI, TAC). The location information of the UPFs (571, 572, 573, 574) may be different from the location information of the terminal (120) and each of the UPFs (561, 562) within the first serving area (560). For example, the location information of UPFs (571, 572, 573, 573) within the serving area (570) may indicate an area including the location of the terminal (120). In addition, referring to example (550), the third serving area (580) may include four UPFs (581, 582, 583, 584). The third serving area (580) may indicate an area with the lowest service quality among the first serving area (560), the second serving area (570), and the third serving area (580), which are serving areas capable of providing the service to the terminal (120).For example, UPFs (581, 582, 583, 584) within the third serving area (560) may have the same location information (e.g., TAI, TAC). The location information of the UPFs (581, 582, 583, 584) may be different from the location information of the terminal (120), the UPFs (561, 562) within the first serving area (560), and the UPFs (571, 572, 573, 574) within the second serving area (570). For example, the location information of the UPFs (581, 582, 583, 583) within the serving area (580) may indicate an area including the location of the terminal (120). In example (550), an example is shown including one terminal (120), UPFs (561, 562) of a first serving area (560), UPFs (571, 572, 573, 574) of a second serving area (570), and UPFs (581, 582, 583, 584) of a third serving area (580), but the embodiments of the present disclosure are not limited thereto.

[0110] Although not illustrated in example (550), the entity (or NF) distributing the call within multiple serving areas (560, 570, 580) may be an SMF (or NRF). For convenience of explanation, an example of the SMF distributing the call is described below. However, the embodiments of the present disclosure are not limited thereto.

[0111] For example, the SMF can collect load information from each of the UPFs (561, 562). For example, the SMF can detect the state of each of the UPFs (561, 562) using the load information collected from each of the UPFs (561, 562). In other words, the SMF can determine whether each of the UPFs (561, 562) is in an overload state (or the first state) using the load information. However, the embodiments of the present disclosure are not limited thereto. For example, the SMF can also determine whether each of the UPFs (561, 562) is in an overload state using the load information (or predicted load information) predicted by the artificial intelligence model and the load information.

[0112] For example, the SMF may determine that the UPF (561) and the UPF (562) are in the first state. In response to determining that the UPF (561) and the UPF (562) are in the first state, the SMF may collect load information from the UPFs (571, 572, 573, 574) of the second serving area (570). However, embodiments of the present disclosure are not limited thereto. For example, the SMF may collect load information of the UPFs (571, 572, 573, 574) of the second serving area (570) while collecting load information of the UPFs (561, 562) of the first serving area (560).

[0113] For example, the SMF can determine the status of the UPFs (571, 572, 573, 574) by using the load information of the UPFs (571, 572, 573, 574). For example, the SMF can determine that the UPF (571) among the UPFs (571, 572, 573, 574) is in the second status. The SMF can assign newly introduced calls (or new calls) to the UPF (571).

[0114] Alternatively, for example, the SMF may collect load information from the UPFs (581, 582, 583, 584) of the third serving area (580) in response to determining that all of the UPFs (571, 572, 573, 574) are in the first state. However, embodiments of the present disclosure are not limited thereto. For example, the SMF may collect load information of the UPFs (571, 572, 573, 574) of the second serving area (570) as well as load information of the UPFs (581, 582, 583, 584) of the third serving area (580) while collecting load information of the UPFs (561, 562) of the first serving area (560). The above SMF can determine that among the UPFs (581, 582, 583, 584), the UPF (582) and the UPF (583) are in the second state. The SMF can assign newly introduced calls (or new calls) to the UPF (582) and the UPF (583).

[0115] After assigning the calls to the UPF (571) or UPF (582) and UPF (583), the SMF can collect load information from each of the UPFs (561, 562) of the first serving area (560). For example, the SMF can detect the state of each of the UPFs (561, 562) using the load information collected from each of the UPFs (561, 562). For example, the SMF can detect that the UPF (561) changes from the first state to the second state after assigning the calls. Accordingly, the SMF can perform redistribution for the calls assigned to the UPF (571) or UPF (582) and UPF (583). For example, the SMF can assign the calls to the UPF (561). For example, the SMF may cause the PDU session reestablishment procedure to be performed to assign the calls to the UPF (561) at the timing at which the change is detected. Or, for example, the SMF may cause the PDU session reestablishment procedure to be performed when it detects that the terminal (120) transitions to idle mode after the timing at which the change is detected. According to one embodiment, the SMF may cause the PDU session reestablishment procedure to be performed for a specified number of calls during a specified time. For example, the specified time and the specified number may be determined based on load information received from the UPF (561).

[0116] Figure 6a illustrates an example of a method for learning an artificial intelligence model (AI model) based on load information of multiple time intervals.

[0117] The above load information may represent load information collected by the first NF (e.g., SMF or NRF) of FIG. 4 from each NF (e.g., UPF) within the NF group. The artificial intelligence model, which is the learning target of FIG. 6a, may represent an example of the artificial intelligence model (660) of FIG. 6b, which will be described later.

[0118] FIG. 6A illustrates an example (600) in which the first NF collects load information for a specified period (603). Referring to the example (600), the first NF may collect load information for a time interval from a current point in time (601) to a past point in time (602) for a specified period (603). For example, the specified period (603) may be three months. For example, the time interval may be five minutes. The time interval may represent a cycle for collecting the load information. Referring to the example (600), since the current point in time (601) changes over time and the specified period (603) has a specific length, the past point in time (602) may change. The specified period (603) may also represent the length of data that the first NF stores in relation to NFs within the NF group.

[0119] For example, a designated period (603) may be composed of a set of sequences including a plurality of time intervals. For example, the set of sequences may include sequences (610-1, 610-2, 610-3, ..., 610-n). For example, the temporal length of one sequence may be 6 hours. In other words, one sequence may include 72 time intervals.

[0120] For example, the temporal difference between two adjacent sequences among the sequences (610-1, 610-2, 610-3, ..., 610-n) can be defined as one time interval. For example, the difference (615) between sequences (610-1) and (610-2) can correspond to the length of one time interval (e.g., 5 minutes).

[0121] The above artificial intelligence model can perform learning on a sequence-by-sequence basis. For example, the artificial intelligence model can be trained based on a single sequence (610-n). Referring to FIG. 6A, a specific example (620) for sequence (610-n) is illustrated.

[0122] Referring to example (620), the sequence (610-n) may include a plurality of time intervals (640-1, 640-2, 640-3, ..., 640-n, and 650). For example, based on a first set of load information prior to a reference timing (630) among the plurality of time intervals (640-1, 640-2, 640-3, ..., 640-n, and 650), the artificial intelligence model may predict load information after the reference timing (630). In other words, the artificial intelligence model may identify predicted load information. For example, the reference timing (630) may represent a point 5 hours after the earliest time point within the sequence (610-n). For example, the first set of load information may represent load information corresponding to a first time section (645). For example, the temporal length of the first time section (645) may be 5 hours. For example, the load information of the first set may include load information corresponding to a plurality of time sections (640-1, 640-2, 640-3, ..., 640-n).

[0123] The artificial intelligence model may compare the predicted load information with a second set of load information after a reference time point (630). For example, the second set of load information may represent load information corresponding to a second time period (655). For example, the time length of the second time period (655) may be 1 hour. For example, the second set of load information may include load information corresponding to a plurality of time periods including the load information (650). For example, the plurality of parameters included in the second set of load information may include CPU load, traffic, dropped packets, and the number of PDU sessions. However, the embodiments of the present disclosure are not limited thereto. The artificial intelligence model may learn by comparing the predicted load information and the second set of load information.

[0124] Referring to FIG. 6A, a first portion of the sequences (610-1, 610-2, 610-3, ..., 610-n) within a designated period (603) may be used for training the artificial intelligence model. In addition, a second portion of the sequences (610-1, 610-2, 610-3, ..., 610-n) within the designated period (603), which is different from the first portion, may be used to evaluate the prediction accuracy (or precision) of the trained artificial intelligence model. In other words, through the example (620) of FIG. 6A, training the artificial intelligence model may be performed based on each of the sequences of the first portion. The first portion and the second portion may not overlap each other. For example, a portion of the sequences (610-1, 610-2, 610-3, ..., 610-n) included in the first portion may be different from a portion of the sequences (610-1, 610-2, 610-3, ..., 610-n) included in the second portion. For example, the first portion may include 70% of the sequences (610-1, 610-2, 610-3, ..., 610-n). The second portion may include 30% of the sequences (610-1, 610-2, 610-3, ..., 610-n). However, the embodiments of the present disclosure are not limited thereto, and the ratio between the first portion and the second portion may be changed. Additionally, each of the first part and the second part may include any sequence of sequences (610-1, 610-2, 610-3, ..., 610-n). In other words, the sequences included in the first part or the second part may be randomly identified rather than being identified in a time series manner.

[0125] Although FIG. 6A illustrates an example in which the artificial intelligence model compares load information (or a second set of load information) including the plurality of parameters with predicted load information, the present disclosure is not limited thereto. For example, the artificial intelligence model may be trained by comparing a scaled load value based on the plurality of parameters of the load information with a scaled load value of the predicted load information.

[0126] For example, based on a first set of load information prior to a reference time point (630) among a plurality of time intervals (640-1, 640-2, 640-3, ..., 640-n, and 650), the artificial intelligence model can identify a first load value associated with a time after the reference time point (630). For example, the reference time point (630) may represent a point 5 hours after the earliest time point in the sequence (610-n). For example, the first set of load information may represent load information corresponding to a first time section (645). For example, the temporal length of the first time section (645) may be 5 hours. For example, the first set of load information may include load information corresponding to a plurality of time intervals (640-1, 640-2, 640-3, ..., 640-n). The first load value may represent a value related to a time after a predicted reference time (630) based on the first set of load information. The first load value may represent a scaled value based on a plurality of parameters of the first set of load information and a ratio between the plurality of parameters. The artificial intelligence model may compare the first load value with a second load value identified based on a second set of load information after the reference time (630). For example, the second set of load information may represent load information corresponding to a second time section (655). For example, the temporal length of the second time section (655) may be 1 hour. For example, the second set of load information may include load information corresponding to a plurality of time intervals including the load information (650). The second load value may represent a scaled value based on a plurality of parameters of the load information corresponding to the plurality of time intervals including the load information (650) and a ratio between the plurality of parameters.As described above, the first NF can learn the artificial intelligence model by comparing the first load value and the second load value.

[0127] While FIG. 6A illustrates the first NF comprising a single artificial intelligence model as an example, embodiments of the present disclosure are not limited thereto. For example, the first NF may comprise multiple artificial intelligence models. For example, the first NF may train each of the multiple artificial intelligence models based on load information acquired during a specified period (603).

[0128] In addition, the temporal lengths exemplified in FIG. 6A are merely exemplary for the convenience of explanation, and the embodiments of the present disclosure are not limited thereto. For example, the temporal lengths of a designated period (603) may have different lengths. For example, the temporal lengths of a single sequence may have different lengths. For example, the temporal lengths of a first time section and a second time section within a single sequence may have different lengths. For example, the temporal lengths of a single time section may have different lengths. In addition, the first portion and the second portion within the designated period (603) may be identified at different ratios.

[0129] Figure 6b illustrates an example of a method for obtaining predictive load information using an artificial intelligence model.

[0130] An example (670) of a method for learning an artificial intelligence model (660) for each parameter of load information of FIG. 6b and obtaining the predicted load information through the learned artificial intelligence model (660) is shown.

[0131] Referring to example (670), the artificial intelligence model (660) can be implemented with a recurrent neural network (RNN) (or stacked RNN) that learns using sequential inputs and a fully connected neural network (FCNN) for generating outputs. The RNN can utilize multiple gated recurrent units (GRUs). However, the embodiments of the present disclosure are not limited thereto. For example, the RNN can be implemented through an artificial intelligence model of the RNN series (e.g., long short term memory (LSTM) or gated recurrent units (GRUs)). Alternatively, for example, the artificial intelligence model (660) can be implemented through a prediction model for time series data (e.g., Transformer).

[0132] Referring to example (670), the artificial intelligence model (660) can be trained for each parameter of the load information. For example, the load information can include a plurality of parameters (e.g., traffic, number of PDU sessions, CPU load), and the artificial intelligence model (660) can be trained for each parameter. For example, the artificial intelligence model (660) can be trained through inputs (660-1, 660-2, ..., 660-n-1, 660-n) for a plurality of time intervals. For example, each of the inputs (660-1, 660-2, ..., 660-n-1, 660-n) can include a parameter (or load information) and location information (loc). For example, the location information can indicate the location information of an NF of an NF group that provides the load information. For example, the location information can include an ID of a TA or an NF group. For example, the artificial intelligence model (660) can be trained using an input (660-1) including a first time interval and position information in the first time interval, an input (660-2) including a second time interval following the first time interval and position information in the second time interval, an input (660-n-1) including an n-1th time interval and position information in the n-1th time interval, and an input (660-n) including an n-th time interval and position information in the n-th time interval.

[0133] Referring to example (670), the artificial intelligence model (660) can generate output values ​​(665) using inputs (660-1, 660-2, ..., 660-n-1, 660-n). For example, the output values ​​(665) can be 128 output values ​​(O1, O2, O3, ..., O 128) may be included. However, the embodiments of the present disclosure are not limited thereto. For example, the output values ​​(665) may be used to generate the predicted load information. For example, the predicted load information generated from the output values ​​(665) may be generated for a specific parameter. For example, the first NF may evaluate the accuracy (or prediction accuracy) of the artificial intelligence model (660) by comparing the predicted load information generated using the artificial intelligence model (660) with the actually collected load information. For example, the first NF may perform a prediction on the NF of the NF group using the artificial intelligence model (660) when the accuracy is above a threshold level. For example, the first NF may determine whether the NF is in the first state based on an artificial intelligence model (660) by using location information of the terminal (e.g., terminal (120) of FIGS. 5A and 5B) and load information recently acquired from an NF of the NF group.

[0134] Figures 7a and 7b illustrate examples of graphs representing predicted load information and collected load information over time.

[0135] Each of the above-mentioned predicted load information and the above-mentioned collected load information may include multiple parameters. For example, the multiple parameters may include traffic and the number of PDU sessions. For example, the predicted load information may be generated through the artificial intelligence model (660) of FIG. 6B.

[0136] FIG. 7A illustrates an example of a graph (700) that depicts the number of collected traffic over time and the number of predicted traffic using an artificial intelligence model. The horizontal axis of the graph (700) may represent time (unit: hours), and the vertical axis may represent the number of traffic. The graph (700) may include a first line (710) representing the number of collected traffic over time and a second line (720) representing the number of predicted traffic. Comparing the first line (710) and the second line (720), the error rate between the number of collected traffic in the first line (710) and the number of predicted traffic in the second line (720) may be approximately 1.45%. For example, the error rate may be calculated based on the mean absolute percentage error (MAPE).

[0137] FIG. 7B illustrates an example of a graph (730) illustrating the number of collected PDU sessions over time and the number of predicted PDU sessions using an artificial intelligence model. The horizontal axis of the graph (730) may represent time (unit: hour), and the vertical axis may represent the number of PDU sessions. The graph (730) may include a third line (740) representing the number of collected PDU sessions over time and a fourth line (750) representing the number of predicted PDU sessions. Comparing the third line (740) and the fourth line (750), the error rate between the number of collected PDU sessions of the third line (740) and the number of predicted PDU sessions of the fourth line (750) may be about 1.68%. For example, the error rate may be calculated based on MAPE.

[0138] Referring to FIGS. 7A and 7B , the device and method according to the embodiment of the present disclosure can form a low error rate between the load information collected over a specified period (e.g., 3 months) and the load information predicted from the load information using an artificial intelligence model (660). Accordingly, the first NF (e.g., the first NF of FIG. 4 ) including the artificial intelligence model (660) can relatively accurately predict future load information based on load information collected currently or in the past.

[0139] As described above, the first NF according to embodiments of the present disclosure can automatically detect overload of each NF by collecting load information (or load information of the user plane) of each NF of the NF group in real time. In addition, when detecting overload, the first NF can relieve (or alleviate) the overload of the overloaded (or the first state) NF by distributing the load (or call) to the NF for which the overload is detected. The first NF according to embodiments of the present disclosure can quickly detect and relieve the overload, thereby preventing service failure and improving service quality. In addition, since the first NF according to embodiments of the present disclosure automatically detects and distributes the overload, operator errors can be reduced, and the operating cost of a network including the first NF can be reduced.

[0140] In addition, the first NF according to embodiments of the present disclosure may, when detecting an overload of an entire specific NF group, preferentially distribute a call to an NF in an NF group that is different from but closest to the specific NF group. Accordingly, a deterioration in the quality of a service provided based on the call may be reduced. In addition, when the overload state of at least one NF in the specific NF group is resolved, the first NF may perform a recovery operation to redistribute the call to the at least one NF, thereby improving the quality of the service.

[0141] In addition, the first NF according to embodiments of the present disclosure can predict overload in advance before the overload of a specific NF occurs by using an artificial intelligence model (e.g., the artificial intelligence model (660) of FIG. 6B). For example, if a relatively high level of overload occurs for the specific NF, the overload caused by existing traffic may worsen even if the distribution of new calls is restricted. Therefore, the first NF can predict overload using the artificial intelligence model, thereby restricting and distributing new calls introduced at a faster time point. Accordingly, the first NF can prevent service disruption due to overload.

[0142] Figure 8 illustrates an example of an operational flow for how a first NF (network function) controls overload of a second NF of an NF group.

[0143] At least some of the above methods of FIG. 8 may be performed by a first NF that performs NF selection. For example, the first NF may represent an example of the first NF of FIG. 4. For example, the first NF of FIG. 8 may include the core network entity (130) of FIG. 3. For example, the second NF that is the target of the NF selection may be an example of an entity (e.g., UPF (311)) of the entity group (e.g., UPFs (311, 312, 313) in the first serving area (310)) of FIG. 3, a UPF (512) of the UPF group (e.g., UPFs (511, 512, 513, 514)) in the serving area (510) of FIG. 5a, or a UPF (561 or 562) of the UPF group (e.g., UPFs (561, 562)) in the first serving area (560) of FIG. 5b. At this time, the first NF may include an SMF or an NRF.

[0144] For example, at least a portion of the method of FIG. 8 may be controlled by the processor of the first NF (e.g., the control unit (233) of FIG. 2C). In the following embodiments, each operation may be performed sequentially, but is not necessarily performed sequentially. For example, the order of each operation may be changed, and at least two operations may be performed in parallel.

[0145] In operation (810), the first NF may obtain load information for indicating the load of each NF in the NF group including the second NF and the third NF. For example, the first NF may collect the load information of each NF in the NF group. For example, the NF group may be arranged in one serving area. For example, when the NF group includes UPFs and the second NF is a first UPF among the UPFs, the third NF may be a second UPF. The second NF and the third NF in the NF group may have the same location information.

[0146] In one embodiment, the load information may include information indicating the load of each NF of the NF group. For example, the load information may include a plurality of parameters for identifying the load. For example, the first NF may collect load information of the second NF and load information of the third NF.

[0147] For example, the load information may include a factor related to a service provided by the second NF (or the third NF), a user plane factor, and a control plane factor. For example, the plurality of parameters may include at least one of a factor related to the service, the user plane factor, or the control plane factor. Hereinafter, for convenience of explanation, the load information of the second NF is described as an example, but the embodiments of the present disclosure are not limited thereto. For example, the first NF may collect load information of each NF of the NF group including the second NF.

[0148] For example, the service-related element may include at least one of the number of user equipment (UEs), the number of protocol data unit (PDU) sessions, or the number of quality of service (QoS) flows associated with the second NF. For example, the number of UEs associated with the second NF may include the maximum number of UEs that the second NF can service or the number of UEs that the second NF is providing service to. For example, the number of PDU sessions and the number of QoS flows may indicate the number of PDU sessions and QoS flows that the second NF provides service to. In addition, for example, the service-related element may include at least one of information indicating a load of a central processing unit (CPU), a memory, or a disk of the second NF.

[0149] For example, the user plane element may include at least one of traffic, packet drop rate, or IP (internet protocol) pool usage. For example, the traffic (or throughput) may include the amount of traffic used per unit time and performance capacity information of the maximum serviceable traffic. The packet drop rate may include the number and size of packets dropped due to failure in data transmission and reception. The packet drop rate may be referred to as a drop packet. For example, the IP pool usage may include the number (or usage) of IPs allocated to a specific terminal within the IP pool.

[0150] For example, the control plane element may include information about transactions per second (TPS) or calls. The TPS may include the number of messages per second. The information about the call may include information about attempts, successes, failures, and failure reasons, depending on the call procedure.

[0151] According to one embodiment, the plurality of parameters included in the load information may be determined based on the second NF. For example, the plurality of parameters may be changed based on the function or role of the second NF.

[0152] In one embodiment, the first NF may periodically obtain the load information from the second NF. For example, the first NF may obtain the load information at intervals of a specified length. For example, the specified length may be 5 minutes. However, the embodiments of the present disclosure are not limited thereto.

[0153] In operation (820), the first NF may determine, using the load information, whether the first state indicating overload of each NF in the NF group is a first state. For example, the first NF may determine, using the load information, whether the load state of each NF in the NF group is the first state. For example, the first state may be referred to as an overload state, an overload mode, or a load limit state. Additionally, a second state different from the first state may be referred to as a default state, a non-overload state, or a default mode.

[0154] According to one embodiment, the first NF may determine whether the first state is reached based on a comparison between a specific parameter and at least one reference value set for the specific parameter. For example, the specific parameter may be included in the plurality of parameters of the load information received from the second NF. The relationship between the specific parameter and the at least one reference value may be referenced to Table 1.

[0155] According to one embodiment, the first NF may recognize a level for each of the plurality of parameters of the load information received from the second NF. For example, the first NF may recognize that the load of the CPU for the control plane of the second NF is at a first level, the traffic of the first NF is at a second level, and the IP pool of the second NF is in the second state (i.e., a state that is not overloaded). According to one embodiment, the first NF may determine that the second NF is in the first state when at least one of the plurality of parameters is at the first level. In the example, even if the IP pool is in the second state, the load of the CPU for the control plane is at the first level, and the traffic is at the second level, so the first NF may determine the second NF to be in the second state. However, the embodiments of the present disclosure are not limited thereto. For example, the first NF may determine that the second NF is in the first state if at least one of the plurality of parameters is equal to or greater than the second level or is at the third level.

[0156] According to one embodiment, the first NF may determine whether each NF in the NF group is in the first state based on load information (hereinafter, predicted load information) predicted or expected using the load information. For example, the predicted load information may be generated by an artificial intelligence model (e.g., the artificial intelligence model (660) of FIG. 6B) using the latest load information. For example, the artificial intelligence model may be learned through load information acquired for a plurality of time intervals prior to the timing at which the latest load information is acquired and location information of an NF providing the load information. For specific details on the learned artificial intelligence model, reference may be made to FIGS. 6A and 6B.

[0157] According to one embodiment, the first NF may determine the first state of the second NF based on a result predicted using the load information when the prediction accuracy of the artificial intelligence model is equal to or higher than a threshold level. For example, when a difference between a first load value generated from a first set of load information before a reference time point among the plurality of time intervals and a second load value generated from a second set of load information after the reference time point among the plurality of time intervals exceeds a reference difference, the first NF may determine that the prediction accuracy is relatively high. For example, the first load value may represent a value identified based on load information predicted for after the reference time point based on the first set of load information, which is load information before the reference time point. For example, the second load value may represent a value identified based on the second set of load information.

[0158] In operation (830), the first NF may change a ratio between the second NF and the third NF for distributing the plurality of calls from the first ratio to the second ratio based on determining that the second NF is in the first state and the third NF is in the second state. For example, the first NF may determine that the second NF of the NF group is in the first state and the third NF of the NF group is in the second state. In other words, if at least one NF of the NF group is not overloaded, the first NF may change a ratio within the NF group for distributing the plurality of calls. For example, the third NF may be referred to as an NF capable of distributing calls or an available NF.

[0159] In one embodiment, the first NF can change the ratio from the first ratio to the second ratio that is different from the first ratio. For example, the first ratio can represent a distribution ratio that is set before collecting load information from each NF in the NF group. For example, the first NF can distribute calls to the second NF and the third NF based on the first ratio. For example, the first ratio can be a 1:1 ratio for the second NF and the third NF. However, the embodiments of the present disclosure are not limited thereto.

[0160] According to one embodiment, the first NF can detect (or recognize) the degree of overload of the second NF in the first state. For example, the degree of overload can be determined using the parameter-specific levels of the second NF recognized by the first NF, as described in Table 1 described above. For example, the degree of overload can represent the maximum level among the parameter-specific levels of the second NF. Or, for example, it can represent the average level of the parameter-specific levels of the second NF.

[0161] In operation (840), the first NF may assign, based on the changed second ratio, second calls, which are fewer than first calls according to the first ratio among the plurality of calls, to the second NF, and fourth calls, which are more than third calls according to the first ratio among the plurality of calls, to the third NF.

[0162] For example, according to the first ratio, the first calls among the plurality of calls may be allocated to the second NF, and the second calls among the plurality of calls may be allocated to the third NF. In contrast, when the second ratio changed from the first ratio is used, the third calls, which are fewer than the first calls among the plurality of calls, may be allocated to the second NF, and the fourth calls, which are more than the second calls among the plurality of calls, may be allocated to the third NF. In the above example, an example of distributing the plurality of calls to the second NF and the third NF of the NF group is described, but the embodiments of the present disclosure are not limited thereto. For example, the first NF may distribute the plurality of calls to the NF group including three or more NFs. For example, when the second ratio is used instead of the first ratio, the number of calls to be allocated (or distributed) to the second NF may be reduced.

[0163] As described above, a device of a first network function (NF) may include a memory including instructions. The device may include a transceiver. The device may include at least one processor. The instructions, when individually or collectively executed by the at least one processor, may cause the device to obtain load information indicating a load of each NF in a NF group including a second NF and a third NF. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine, using the load information, whether each NF in the NF group is in a first state indicating an overload of the NF. The instructions, when individually or collectively executed by the at least one processor, may cause the device to change a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio, based on determining that the second NF is in the first state and that the third NF is in a second state different from the first state. The instructions, when individually or collectively executed by the at least one processor, may cause the device to allocate, based on the changed second ratio, a second number of calls from among the plurality of calls to the second NF, which is less than a first number of calls according to the first ratio, and a fourth number of calls from among the plurality of calls to the third NF, which is more than a third number of calls according to the first ratio.

[0164] According to one embodiment, the load information of the second NF may include at least one of a factor related to a service provided by the second NF, a user plane factor, or a control plane factor. The factor related to the service may include information indicating the number of user equipments related to the second NF, the number of protocol data unit (PDU) sessions, or the number of quality of service (QoS) flows, and a load of a central processing unit (CPU), memory, or disk of the second NF.

[0165] In one embodiment, the user plane element may include traffic, packet drop rate, or IP (Internet Protocol) pool usage. The control plane element may include information about transactions per second (TPS) or calls.

[0166] According to one embodiment, NFs within the NF group may have location information. The location information may include a tracking area indicator (TAI).

[0167] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the device to determine whether a first parameter of the load information of the second NF exceeds a first threshold value. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine whether a second parameter of the load information of the second NF exceeds a second threshold value. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine a load state of the second NF as the first state based on determining that the first parameter exceeds the first threshold value or that the second parameter exceeds the second threshold value.

[0168] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the device to determine a limitation ratio for changing the ratio from the first ratio to the second ratio when the first parameter exceeds the first reference value and the second parameter is less than the second reference value. The limitation ratio may be determined based on a level of the first parameter.

[0169] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the device to obtain a first load value based on a first set of load information for the second NF prior to a reference timing among a plurality of time intervals. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine whether a difference between the first load value and a second load value obtained based on a second set of load information after the reference timing among the plurality of time intervals exceeds a reference difference. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine whether the second NF is in the first state using the load information of the second NF if the difference exceeds the reference difference. The instructions, when individually or collectively executed by the at least one processor, may cause the device to: obtain predicted load information of the second NF in a time interval after the timing at which the load information is obtained, if the difference is less than the reference difference. The predicted load information may be obtained using an artificial intelligence model (AI model) based on the load information. The instructions, when individually or collectively executed by the at least one processor, may cause the device to: determine whether the second NF is in the first state using the predicted load information, if the difference is less than the reference difference.

[0170] In one embodiment, the artificial intelligence model may include a recurrent neural network (RNN). The artificial intelligence model may be trained based on a first portion of load information for a specified period of time and location information corresponding to the first portion. The first set of load information and the second set of load information related to the plurality of time intervals may be included in a second portion of the load information for the specified period of time, which is different from the first portion.

[0171] According to one embodiment, the first load value may be predicted using the artificial intelligence model based on the first set of load information. The predicted load information may be predicted based on the load information and location information of the user equipment receiving the service by the second NF.

[0172] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the device to obtain another load information indicative of a load of each NF in another NF group including a fourth NF, based on determining that each of all NFs in the NF group is in the first state. The instructions, when individually or collectively executed by the at least one processor, may cause the device to determine, using the another load information, whether each NF in the another NF group is in the first state. The instructions, when individually or collectively executed by the at least one processor, may cause the device to assign at least some of the plurality of calls to the fourth NF, based on determining that the fourth NF is in the second state. The location information of the NF group may be different from other location information of the other NF group.

[0173] In one embodiment, the location information of the NF group may indicate a first area including the location of a terminal associated with at least a portion of the NF group. The other location information of the other NF group may indicate a second area including the location. The first area may be closer to the location than the second area.

[0174] In one embodiment, the size of the second region may be wider than the size of the first region. The second region may include the first region.

[0175] In one embodiment, the instructions, when individually or collectively executed by the at least one processor, may cause the device to allocate at least some of the plurality of calls to the fourth NF, and then detect that a load state of the second NF changes from the first state to the second state using load information obtained from the second NF. The instructions, when individually or collectively executed by the at least one processor, may cause the device to allocate the at least some to the second NF.

[0176] According to one embodiment, the first NF may include a session management function (SMF) or a network repository function (NRF). The NF group may include a plurality of user plane functions (UPFs). When the second NF is a first UPF, the third NF may be a second UPF different from the first UPF.

[0177] The method performed by the device of the first NF (network function) as described above may include an operation of obtaining load information for indicating a load of each NF in a NF group including a second NF and a third NF. The method may include an operation of determining, using the load information, whether each NF in the NF group is in a first state indicating an overload of the NF. The method may include an operation of changing a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio based on determining that the second NF is in the first state and the third NF is in a second state different from the first state. The method may include, based on the changed second ratio: an operation of allocating second calls, which are reduced from first calls to be distributed to the second NF among the plurality of calls, to the second NF, and an operation of allocating fourth calls, which are increased from third calls to be distributed to the third NF among the plurality of calls, to the third NF.

[0178] In one embodiment, the method may include obtaining another load information for indicating a load of each NF in another NF group including a fourth NF, based on determining that each of all NFs in the NF group is in the first state. The method may include determining, using the other load information, whether each NF in the other NF group is in the first state. The method may include assigning at least some of the plurality of calls to the fourth NF, based on determining that the fourth NF is in the second state. The location information of the NF group may be different from other location information of the other NF group.

[0179] In one embodiment, the location information of the NF group may indicate a first area for the location of the terminal associated with at least a portion of the NF group. The other location information of the other NF group may indicate a second area for the location. The first area may be closer to the location than the second area.

[0180] In one embodiment, the size of the second region may be wider than the size of the first region. The second region may include the first region.

[0181] In one embodiment, the method may include an operation of detecting that the second NF changes from the first state to the second state using load information obtained from the second NF after allocating at least some of the plurality of calls to the fourth NF. The method may include an operation of allocating at least some of the calls to the second NF.

[0182] A non-transitory computer-readable storage medium as described above may store one or more programs comprising instructions that, when individually or collectively executed by at least one processor of a device of a first network function (NF) including a transceiver, cause the device to obtain load information indicating a load of each NF in a group of NFs including a second NF and a third NF. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the device to determine, using the load information, whether each NF in the group of NFs is in a first state indicating an overload of the NF. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the device to change a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio, based on determining that the second NF is in the first state and the third NF is in a second state different from the first state. The non-transitory computer-readable storage medium may store one or more programs comprising instructions that, when individually or collectively executed by the at least one processor, cause the device to: assign second calls, which are reduced from first calls to be distributed to the second NF among the plurality of calls, to the second NF; and assign fourth calls, which are increased from third calls to be distributed to the third NF among the plurality of calls, to the third NF.

[0183] The methods according to the embodiments described in the claims or specification of the present disclosure may be implemented in the form of hardware, software, or a combination of hardware and software.

[0184] When implemented in software, a computer-readable storage medium storing one or more programs (software modules) may be provided. The one or more programs stored in the computer-readable storage medium are configured to be executed by one or more processors in an electronic device. The one or more programs include instructions that cause the electronic device to execute methods according to embodiments described in the claims or specification of the present disclosure. The one or more programs may be provided as a computer program product. The computer program product may be traded between a seller and a buyer as a commodity. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read only memory (CD-ROM)), or may be distributed online (e.g., downloaded or uploaded) via an application store (e.g., Play Store™) or directly between two user devices (e.g., smart phones). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created in a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0185] These programs (software modules, software) may be stored in random access memory, non-volatile memory including flash memory, read only memory (ROM), electrically erasable programmable read only memory (EEPROM), magnetic disc storage devices, compact disc-ROM (CD-ROM), digital versatile discs (DVDs) or other forms of optical storage devices, magnetic cassettes, or may be stored in memories formed by a combination of some or all of these. In addition, each configuration memory may include multiple copies.

[0186] Additionally, the program may be stored on an attachable storage device that is accessible via a communication network, such as the Internet, an intranet, a local area network (LAN), a wide area network (WAN), a storage area network (SAN), or a combination thereof. Such a storage device may be connected to a device implementing an embodiment of the present disclosure via an external port. Additionally, a separate storage device on the communication network may be connected to a device implementing an embodiment of the present disclosure.

[0187] In the specific embodiments of the present disclosure described above, components included in the disclosure are expressed singularly or plurally, depending on the specific embodiment presented. However, the singular or plural expressions are selected to suit the presented situation for convenience of explanation, and the present disclosure is not limited to singular or plural components. Components expressed in plural may be composed of singular elements, or components expressed in singular may be composed of plural elements.

[0188] According to embodiments, one or more of the components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., modules or programs) may be integrated into a single component. In such a case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to embodiments, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

[0189] Meanwhile, although the detailed description of the present disclosure has described specific embodiments, it is obvious that various modifications are possible within the scope of the present disclosure.

Claims

1. In the device of the first NF (network function), Memory containing instructions; transceiver; and comprising at least one processor; The above instructions, when individually or collectively executed by the at least one processor, cause the device to: Obtain load information to indicate the load of each NF in the NF group including the second NF and the third NF; Using the above load information, determine whether each NF in the NF group is in a first state indicating an overload of the NF; Based on determining that the second NF is the first state and the third NF is a second state different from the first state, changing a ratio between the second NF and the third NF for distributing a plurality of calls from the first ratio to a second ratio different from the first ratio; and Based on the above changed second ratio: Assigning the second numbers, which are less than the first numbers according to the first ratio among the plurality of numbers, to the second NF; and Causing to assign to the third NF more fourth numbers than third numbers according to the first ratio among the above plurality of numbers, device.

2. In claim 1, The load information of the second NF includes at least one of a factor related to a service provided by the second NF, a user plane element, or a control plane element, Elements related to the above services are: The number of terminals (user equipment), the number of PDU (protocol data unit) sessions, or the number of QoS (quality of service) flows associated with the second NF, and Contains information indicating the load of the CPU (central processing unit), memory, or disk of the second NF. device.

3. In claim 2, The above user plane elements include traffic, packet drop rate, or IP (internet protocol) pool usage, and The above control plane element includes information about TPS (transaction per second) or call. device.

4. In claim 1, NFs within the above NF group have location information, and The above location information includes TAI (tracking area indicator). device.

5. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the device to: Determine whether the first parameter of the load information of the second NF exceeds the first reference value; Determining whether the second parameter of the load information of the second NF exceeds the second reference value; and Causing the load state of the second NF to be determined as the first state based on determining that the first parameter exceeds the first reference value or that the second parameter exceeds the second reference value. device.

6. In claim 5, The above instructions, when individually or collectively executed by the at least one processor, cause the device to: If the first parameter exceeds the first reference value and the second parameter is less than the second reference value, a limitation ratio is determined to change the ratio from the first ratio to the second ratio, The above limit ratio is determined according to the level of the first parameter. device.

7. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the device to: Obtaining a first load value based on a first set of load information for the second NF prior to a reference time point (timing) among multiple time intervals; Determine whether a difference between a second load value obtained based on a second set of load information after the reference time point among the plurality of time intervals and the first load value exceeds a reference difference; If the above difference exceeds the above reference difference, the load information of the second NF is used to determine whether the second NF is in the first state; If the above difference is less than the above reference difference: In a time interval after the timing at which the above load information is acquired, the predicted load information of the second NF is acquired, and the predicted load information is acquired using an artificial intelligence model (AI model) based on the load information; and Using the above predicted load information, causing the second NF to determine whether it is in the first state, device.

8. In claim 7, The above artificial intelligence model includes a recurrent neural network (RNN), The above artificial intelligence model is learned based on a first part of the load information for a specified period and location information corresponding to the first part, and The first set of load information and the second set of load information related to the plurality of time intervals are included in a second part of the load information for the specified period that is different from the first part. device.

9. In claim 7, The above first load value is predicted using the artificial intelligence model based on the load information of the first set, and The above predicted load information is predicted based on the load information and the location information of the terminal (user equipment) receiving the service by the second NF. device.

10. In claim 1, The above instructions, when individually or collectively executed by the at least one processor, cause the device to: Based on determining that each of all NFs in the above NF group is in the first state, another load information is obtained to represent the load of each NF in another NF group including the fourth NF; Using the other load information, determining whether each NF in the other NF group is in the first state; and Based on determining that the fourth NF is the second state, causing at least some of the plurality of calls to be assigned to the fourth NF, The location information of the above NF group is different from other location information of the other NF group. device.

11. In claim 10, The location information of the above NF group indicates a first area including the location of a terminal associated with at least a part of the above, The other location information of the other NF group indicates a second area including the location, The above first region is closer to the above location than the above second region, The size of the second region is wider than the size of the first region, and The second region includes the first region, device.

12. In claim 10, The above instructions, when individually or collectively executed by the at least one processor, cause the device to: After allocating at least some of the above plurality of calls to the fourth NF, detecting that the load state of the second NF changes from the first state to the second state by using the load information obtained from the second NF; and causing at least a portion of said above to be allocated to said second NF, device.

13. In claim 1, The above first NF includes a session management function (SMF) or a network repository function (NRF), The above NF group includes a plurality of UPFs (user plane functions), and When the second NF is the first UPF, the third NF is a second UPF different from the first UPF. device.

14. In a method performed by a device of the first NF (network function), The above method: An operation of obtaining load information to indicate the load of each NF in a NF group including a second NF and a third NF; An operation of using the above load information to determine whether each NF in the NF group is in a first state indicating overload of the NF; An operation of changing a ratio between the second NF and the third NF for distributing a plurality of calls from a first ratio to a second ratio different from the first ratio, based on determining that the second NF is the first state and the third NF is a second state different from the first state; and Based on the above changed second ratio: An operation of allocating second numbers, which are less than first numbers according to the first ratio among the plurality of numbers, to the second NF; and An operation of allocating to the third NF more fourth numbers than third numbers according to the first ratio among the plurality of numbers, method.

15. A non-transitory computer-readable storage medium, when individually or collectively executed by at least one processor of a device of a first NF (network function) including a transceiver, causes the device to: Obtain load information to indicate the load of each NF in the NF group including the second NF and the third NF; Using the above load information, determine whether each NF in the NF group is in a first state indicating an overload of the NF; Based on determining that the second NF is the first state and the third NF is a second state different from the first state, changing a ratio between the second NF and the third NF for distributing a plurality of calls from the first ratio to a second ratio different from the first ratio; and Based on the above changed second ratio: Assigning the second numbers, which are less than the first numbers according to the first ratio among the plurality of numbers, to the second NF; and Storing one or more programs including instructions causing the 4th number, which is more than the 3rd number according to the first ratio, to be assigned to the 3rd NF among the plurality of numbers, A non-transitory computer-readable storage medium.

Images