Apparatus and method for introducing data context using machine learning model

Abstract

An apparatus is provided that includes a transceiver and a processor coupled to the transceiver. The processor and transceiver are configured such that the apparatus includes a network state parameter in a message identifying a machine learning model. The machine learning model is used for obtaining analysis information of the wireless communication network. When the wireless communication network is in a particular network state, the machine learning model has been trained using training data acquired from the wireless communication network. The network state parameter indicates a particular network state.

Description

Technical Field

[0001] The subject matter disclosed herein generally relates to using machine learning models to introduce data context, particularly data context indicated by network state parameters. Background Art

[0002] Network analytics and artificial intelligence (AI) / machine learning (ML) are deployed in the 5G core network by introducing the Network Data Analysis Function (NWDAF). Various analysis types, such as "UE mobility", "NF load", etc., differentiated by different analysis IDs, can be supported. This is discussed in TS23.288.

[0003] Each NWDAF can support one or more analysis IDs and can be responsible for implementing: (i) AI / ML inference, called NWDAF AnLF, or (ii) AI / ML training, called NWDAF MTLF, or (iii) both.

[0004] TS23.288 introduces the Analytics Data Repository Function (ADRF), which supports the storage and retrieval of analytics generated by NWDAF and other collected data.

[0005] US10572321B2 discloses a technology for providing and serving listed repository items (such as algorithms, data, models, pipelines, and / or notebooks). In some examples, a network service provider receives a request for listed repository items from a requester, the request indicating at least a category of the repository items, and each column of the repository items, including an indication of the category to which the listed repository items belong and a storage location of the listed repository items, determines a suggestion of at least one listed repository item based on the request, and provides the suggestion of at least one listed repository item to the requester. Summary of the invention

[0006] Disclosed herein are apparatus and processes for introducing context into data using ML models, for example for storage in an ADRF.

[0007] A device is provided, the device comprising a transceiver and a processor coupled to the transceiver. The processor and the transceiver are configured so that the device: includes a network state parameter in a message identifying a machine learning model, wherein: the machine learning model is used to derive analysis information of a wireless communication network; when the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network; and the network state parameter indicates the specific network state.

[0008] A method is also provided, the method comprising: including, by a processor, a network state parameter in a message identifying a machine learning model. The machine learning model is used to derive analysis information of a wireless communication network. When the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network. The network state parameter indicates the specific network state.

[0009] A device is also provided, the device comprising a memory and a processor coupled to the memory. The processor and the memory are configured so that the device: stores a machine learning model in the memory, the machine learning model is used to obtain analysis information of the wireless communication network, when the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network; and stores a network state parameter associated with the machine learning model in the memory, the network state parameter indicating the specific network state.

[0010] A method is also provided, comprising: storing a machine learning model in a memory, the machine learning model being used to derive analysis information of a wireless communication network, the machine learning model having been trained using training data acquired from the wireless communication network when the wireless communication network is in a specific network state; and storing network state parameters associated with the machine learning model in the memory, the network state parameters indicating a specific network state.

[0011] An analysis consumer is also provided, the analysis consumer comprising a transceiver and a processor coupled to the transceiver. The processor and the transceiver are configured to cause the device to: include a network state parameter in a message, and send the message to another device. The message is a request to be provided with analysis information or a subscription to analysis information. The network state parameter indicates a specific network state. The network state parameter can be used for selection of a machine learning model.

[0012] A method performed by an analytics consumer is also provided. The method includes including a network state parameter in a message, and sending the message to another device. The message is a request to be provided with analytics information or a subscription to analytics information. The network state parameter indicates a particular network state. The network state parameter may be used for selection of a machine learning model. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to describe the manner in which the advantages and features of the present disclosure can be obtained, the present disclosure is described by reference to certain devices and methods shown in the accompanying drawings. Each of these drawings depicts only certain aspects of the present disclosure and therefore should not be considered as limiting the scope thereof. For the sake of clarity, the drawings may have been simplified and are not necessarily drawn to scale.

[0014] A method and apparatus for introducing data context using an ML model for storage in an ADRF will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0015] Figure 1 A wireless communication system is depicted;

[0016] Figure 2 depicting a user equipment device;

[0017] Figure 3 Depicts the network nodes;

[0018] Figure 4 is a schematic diagram of a network and illustrates various types of NWDAFs;

[0019] Figure 5 is a diagram depicting the data storage architecture used to analyze and collect data;

[0020] Figure 6 Depicting an example ML model ID combination or structure;

[0021] Figure 7 is a process flow diagram illustrating the process of requesting and receiving network data analysis;

[0022] Figure 8 is a process flow diagram illustrating the process of subscribing / unsubscribing to be notified of network data analytics;

[0023] Fig. 9 is a process flow diagram illustrating a process of requesting and receiving ML model information;

[0024] Fig.10 is a process flow diagram illustrating a process of subscribing / unsubscribing to be notified of ML model information;

[0025] Fig.11 is a process flow diagram showing various processes;

[0026] Fig.12 is a process flow diagram of a method for introducing data context (which is specifically indicated by network state parameters) using a machine learning model;

[0027] Fig.13 is a process flow diagram of a method for storing data context (which is particularly indicated by network state parameters) using a machine learning model; and

[0028] Fig.14 is a process flow diagram of a method for a consumer to request analytics information for a specific analytics ID, the analytics information indicating a specific data context. DETAILED DESCRIPTION

[0029] Those skilled in the art will appreciate that aspects of the present disclosure may be embodied as systems, devices, methods or program products. Therefore, the arrangements described herein may be implemented in full hardware form, full software form (including firmware, resident software, microcode, etc.), or in a combination of software and hardware aspects.

[0030] For example, the disclosed methods and apparatus may be implemented as hardware circuits, including custom very large scale integrated (VLSI) circuits or gate arrays, off-the-shelf semiconductors such as logic chips, transistors, or other discrete components. The disclosed methods and apparatus may also be implemented in programmable hardware devices, such as field programmable gate arrays, programmable array logic, programmable logic devices, etc. As another example, the disclosed methods and apparatus may include one or more physical or logical blocks of executable code, which may be organized, for example, as objects, procedures, or functions.

[0031] In addition, the method and apparatus may take the form of a program product embodied in one or more computer-readable storage devices storing machine-readable code, computer-readable code, and / or program code, hereinafter referred to as code. The storage device may be tangible, non-transitory, and / or non-transmitting. The storage device may not embody a signal. In some arrangements, the storage device employs only a signal for accessing the code.

[0032] Any combination of one or more computer readable media may be used. The computer readable medium may be a computer readable storage medium. The computer readable storage medium may be a storage device storing code. The storage device may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, holographic, micromechanical, or semiconductor system, device or apparatus, or any suitable combination of the above.

[0033] More specific examples of storage devices (a non-exhaustive list) include the following: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the context of this document, a computer-readable storage medium may be any tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0034] Throughout this specification, references to examples or similar language of a particular method or device indicate that a particular feature, structure, or characteristic described in conjunction with the example is included in at least one implementation of the method and device described herein. Therefore, unless otherwise expressly provided, references to examples or similar language features of a particular method or device may, but not necessarily, all refer to the same example, but mean "one or more but not all examples". Unless otherwise expressly provided, the terms "include", "comprise", "have", and their variants mean "including but not limited to". Unless otherwise expressly provided, a list of items listed does not mean that any or all items are mutually exclusive. Unless otherwise expressly provided, the terms "one", "an", and "the" also refer to "one or more".

[0035] As used herein, a list with the conjunction "and / or" includes any single item in the list or a combination of items in the list. For example, a list of A, B, and / or C includes only A, only B, only C, a combination of A and B, a combination of B and C, a combination of A and C, or a combination of A, B, and C. As used herein, a list using the term "one or more of ..." includes any single item in the list or a combination of items in the list. For example, one or more of A, B, and C includes only A, only B, only C, a combination of A and B, a combination of B and C, a combination of A and C, or a combination of A, B, and C. As used herein, a list using the term "one of ..." includes one and only one of any single item in the list. For example, "one of A, B, and C" includes only A, only B, or only C, and does not include a combination of A, B, and C. As used herein, "a member selected from a group consisting of A, B, and C" includes one and only one of A, B, or C, and does not include a combination of A, B, and C. As used herein, “a member selected from the group consisting of A, B and C and combinations thereof” includes only A, only B, only C, a combination of A and B, a combination of B and C, a combination of A and C, or a combination of A, B and C.

[0036] In addition, the features, structures or characteristics described herein may be combined in any suitable manner. In the following description, many specific details are provided, such as examples of programming, software modules, user selections, network transactions, database queries, database structures, hardware modules, hardware circuits, hardware chips, etc., to provide a comprehensive understanding of the present disclosure. However, those skilled in the relevant art will recognize that the disclosed methods and apparatus may be practiced without one or more of the specific details, or may be practiced using other methods, components, materials, etc. In other cases, well-known structures, materials or operations are not shown or described in detail to avoid confusing aspects of the present disclosure.

[0037] The following is a description of various aspects of the disclosed methods and apparatuses with reference to schematic flow charts and / or schematic block diagrams of methods, apparatuses, systems, and program products. It will be understood that each frame of the schematic flow charts and / or schematic block diagrams and the combination of frames in the schematic flow charts and / or schematic block diagrams can be implemented by code. The code can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine, so that instructions executed by a processor of a computer or other programmable data processing device can create components for implementing the functions / actions specified in the schematic flow charts and / or schematic block diagrams.

[0038] The code may also be stored in a storage device that can direct a computer, other programmable data processing apparatus, or other device to work in a specific manner so that the instructions stored in the storage device produce an article of manufacture that includes instructions for implementing the functions / actions specified in the schematic flowchart and / or schematic block diagram.

[0039] The code may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are executed on the computer, other programmable apparatus, or other device, thereby producing a computer-implemented process, so that the code executed on the computer or other programmable apparatus provides a process for implementing the functions / actions specified in the schematic flowchart and / or schematic block diagram.

[0040] The schematic flow charts and / or schematic block diagrams in the figure illustrate the possible architectures, functions and operations of the devices, systems, methods and program products. In this regard, each box in the schematic flow charts and / or schematic block diagrams can represent a module, segment, or portion of a code, which includes one or more executable instructions for implementing the code of (multiple) specified logical functions.

[0041] It should also be noted that in some alternative implementations, the functions shown in the blocks may not appear in the order shown in the figures. For example, two blocks shown in succession may, in fact, be executed substantially simultaneously, or the blocks may sometimes be executed in the reverse order, depending on the functions involved. Other steps and methods may be envisioned that are equivalent in function, logic or effect to one or more blocks or portions thereof of the illustrated figures.

[0042] The description of an element in each figure may refer to the elements of the subsequent figures. The same numerals represent the same elements throughout the drawings.

[0043] The present disclosure relates to an apparatus and method that introduces a network state or network context that describes the training conditions of an ML model, and the ability to request, store, search, and retrieve an ML model when a consumer's needs match a desired network state.

[0044] ML models in machine learning are created by ML algorithms. In other words, ML algorithms specify a process, such as pattern recognition, that considers data (i.e., training data) to create an ML model. These processes in machine learning can generally be defined as follows:

[0045] ML Model: An ML model can be a mathematical representation of a real-world process. To generate an ML model, training data can be provided to the ML algorithm for learning. The ML model can then be used for ML inference or simple reasoning.

[0046] ML Inference: This is the process of providing observations (i.e., data that the ML model operates on) to the ML model and computing outputs. This process is often referred to as “operating the machine learning model” or “putting the machine learning model into production.”

[0047] ML Algorithms: ML algorithms are sets of assumptions made with real-world data before training begins. For example, consider the set of assumptions for a linear regression algorithm which represents the set of functions with the characteristics defined by linear regression. Within this set of functions, the ML model is the selected function that best fits the training data.

[0048] ML Training: This is the process of feeding training data to the ML algorithm to find patterns so that the input parameters correspond to the target. The output of the training process is an ML model, which can be used to provide analytical results. This process is also called "learning".

[0049] Figure 1 An embodiment of a wireless communication system 100 is depicted in which an apparatus and method for introducing a data context (i.e., a network state or network context) describing training conditions for an ML model may be implemented. In one embodiment, the wireless communication system 100 includes a remote unit 102 and a network unit 104. Although Figure 1 A specific number of remote units 102 and network units 104 are depicted in FIG. 1 , but one skilled in the art will recognize that any number of remote units 102 and network units 104 may be included in the wireless communication system 100.

[0050] In one embodiment, the remote unit 102 may include a computing device such as a desktop computer, a laptop computer, a personal digital assistant (PDA), a tablet computer, a smart phone, a smart TV (e.g., a TV connected to the Internet), a set-top box, a game console, a security system (including a security camera), a vehicle computer, a network device (e.g., a router, a switch, a modem), an aircraft, a drone, etc. In some embodiments, the remote unit 102 includes a wearable device such as a smart watch, a fitness band, an optical head-mounted display, etc. In addition, the remote unit 102 may be referred to as a subscriber unit, a mobile device, a mobile station, a user, a terminal, a mobile terminal, a fixed terminal, a subscriber station, a UE, a user terminal, a device, or other terms used in the art. The remote unit 102 may communicate directly with one or more of the network units 104 via UL communication signals. In some embodiments, the remote unit 102 may communicate directly with other remote units 102 via sidelink communications.

[0051] The network elements 104 may be distributed over a geographical area. In some embodiments, the network elements 104 may also be referred to as access points, access terminals, base stations, base stations, node Bs, eNBs, gNBs, home node Bs, relay nodes, devices, core networks, air servers, radio access nodes, APs, NRs, network entity access and mobility management functions (AMFs), unified data repositories (UDRs), UDM / UDRs, policy control functions (PCFs), location management functions (LMFs), radio access networks (RANs), network slice selection functions (NSSFs), operations, maintenance, and management (OAMs), session management functions (SMFs), user plane functions (UPFs), application functions, authentication server functions (AUSFs), security anchor functions (SEAFs), trusted non-3GPP gateway functions (TNGFs), application functions, service enablement architecture layer (SEAL) functions, vertical application enablement servers, edge enablement servers, border configuration servers, mobile edge computing platform functions, mobile edge computing applications, application data analysis enablement servers, SEAL data delivery servers, middleware entities, network slice capability management servers, or any other terminology used in the art. The network elements 104 are typically part of a radio access network that includes one or more controllers communicatively coupled to one or more corresponding network elements 104. The radio access network is typically communicatively coupled to one or more core networks, which may be coupled to other networks, such as the Internet and a public switched telephone network, etc. These and other elements of the radio access network and core network are not illustrated, but are generally known to those of ordinary skill in the art.

[0052] In one implementation, the wireless communication system 100 complies with the New Radio (NR) protocol standardized in 3GPP, wherein the network unit 104 transmits on the downlink (DL) using an orthogonal frequency division multiplexing (OFDM) modulation scheme, and the remote unit 102 transmits on the uplink (UL) using a single carrier frequency division multiple access (SC-FDMA) scheme or an OFDM scheme. More generally, however, the wireless communication system 100 may implement some other open or proprietary communication protocol, such as WiMAX, IEEE 802.11 variants, GSM, GPRS, UMTS, LTE variants, CDMA2000, ZigBee, Sigfox, etc. The present disclosure is not intended to be limited to the implementation of any particular wireless communication system architecture or protocol.

[0053] The network unit 104 can serve multiple remote units 102 within a service area (eg, a cell or cell sector) via wireless communication links. The network unit 104 sends DL communication signals to serve the remote units 102 in the time, frequency, and / or spatial domains.

[0054] Figure 2 A user equipment device 200 is depicted that can be used to implement the methods described herein. The user equipment device 200 is used to implement one or more of the technical solutions described herein. The user equipment device 200 conforms to one or more of the user equipment devices described in the embodiments of this document. In particular, the user equipment device 200 can be used with the above reference Figure 1 The remote units 102 are identical or identical. The user equipment device 200 includes a processor 205 , a memory 210 , an input device 215 , an output device 220 , and a transceiver 225 .

[0055] The input device 215 and the output device 220 may be combined into a single device, such as a touch screen. In some implementations, the user equipment apparatus 200 does not include any input device 215 and / or output device 220. The user equipment apparatus 200 may include one or more of the following: a processor 205, a memory 210, and a transceiver 225, and may not include an input device 215 and / or an output device 220.

[0056] As depicted in the figure, the transceiver 225 includes at least one transmitter 230 and at least one receiver 235. The transceiver 225 can communicate with one or more cells (or wireless coverage areas) supported by one or more base units. The transceiver 225 can be operable on an unlicensed spectrum. In addition, the transceiver 225 can include multiple UE panels supporting one or more beams. In addition, the transceiver 225 can support at least one network interface 240 and / or application interface 245. (Multiple) application interface 245 can support one or more APIs. (Multiple) network interface 240 can support 3GPP reference points, such as Uu, N1, PC5, etc. As understood by those of ordinary skill in the art, other network interfaces 240 can be supported.

[0057] The processor 205 may include any known controller capable of executing computer-readable instructions and / or capable of performing logical operations. For example, the processor 205 may be a microcontroller, a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), an auxiliary processing unit, a field programmable gate array (FPGA), or a similar programmable controller. The processor 205 may execute instructions stored in the memory 210 to perform the methods and routines described herein. The processor 205 is communicatively coupled to the memory 210, the input device 215, the output device 220, and the transceiver 225.

[0058] Processor 205 may control user equipment device 200 to implement the user equipment device behaviors described herein. Processor 205 may include an application processor (also referred to as a main processor) that manages application domain and operating system (OS) functions, and a baseband processor (also referred to as a baseband radio processor) that manages radio functions.

[0059] The memory 210 may be a computer-readable storage medium. The memory 210 may include a volatile computer storage medium. For example, the memory 210 may include a RAM, including a dynamic RAM (DRAM), a synchronous dynamic RAM (SDRAM), and / or a static RAM (SRAM). The memory 210 may include a non-volatile computer storage medium. For example, the memory 210 may include a hard drive, a flash memory, or any other suitable non-volatile computer storage device. The memory 210 may include both a volatile computer storage medium and a non-volatile computer storage medium.

[0060] The memory 210 may store data related to implementing the traffic class field as described herein. The memory 210 may also store program code and related data, such as an operating system or other controller algorithms operating on the device 200.

[0061] Input device 215 may include any known computer input device, including a touchpad, buttons, keyboard, stylus, microphone, etc. Input device 215 may be integrated with output device 220, for example, as a touch screen or similar touch-sensitive display. Input device 215 may include a touch screen so that text can be entered using a virtual keyboard displayed on the touch screen and / or by handwriting on the touch screen. Input device 215 may include two or more different devices, such as a keyboard and a touchpad.

[0062] Output device 220 may be designed to output visual signals, auditory signals, and / or tactile signals. Output device 220 may include an electronically controllable display or display device capable of outputting visual data to a user. For example, output device 220 may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, an organic LED (OLED) display, a projector, or a similar display device capable of outputting images, text, etc. to a user. As another non-limiting example, output device 220 may include a wearable display that is separate from the rest of user device apparatus 200 but communicatively coupled to the rest, such as a smart watch, smart glasses, a head-up display, etc. In addition, output device 220 may be a component of a smart phone, a personal digital assistant, a television, a desktop computer, a notebook (laptop) computer, a personal computer, a vehicle dashboard, etc.

[0063] Output device 220 may include one or more speakers for producing sound. For example, output device 220 may produce an audible alarm or notification (e.g., a beep or buzzer). Output device 220 may include one or more tactile devices for producing vibration, motion, or other tactile feedback. All or part of output device 220 may be integrated with input device 215. For example, input device 215 and output device 220 may form a touch screen or similar touch-sensitive display. Output device 220 may be located near input device 215.

[0064] The transceiver 225 communicates with one or more network functions of the mobile communication network via one or more access networks. The transceiver 225 operates under the control of the processor 205 to send messages, data, and other signals, and also receives messages, data, and other signals. For example, the processor 205 can selectively activate the transceiver 225 (or part thereof) at a specific time to send and receive messages.

[0065] The transceiver 225 includes at least one transmitter 230 and at least one receiver 235. One or more transmitters 230 can be used to provide uplink communication signals to the base station unit of the wireless communication network. Similarly, one or more receivers 235 can be used to receive downlink communication signals from the base station unit. Although only one transmitter 230 and one receiver 235 are shown, the user equipment device 200 can have any suitable number of transmitters 230 and receivers 235. In addition, (multiple) transmitters 230 and (multiple) receivers 235 can be any suitable type of transmitter and receiver. The transceiver 225 may include a first transmitter / receiver pair for communicating with a mobile communication network on a licensed radio spectrum, and a second transmitter / receiver pair for communicating with a mobile communication network on an unlicensed radio spectrum.

[0066] A first transmitter / receiver pair that can be used to communicate with a mobile communication network on a licensed radio spectrum and a second transmitter / receiver pair that can be used to communicate with a mobile communication network on an unlicensed radio spectrum can be combined into a single transceiver unit, such as a single chip that performs functions for use with both the licensed radio spectrum and the unlicensed radio spectrum. The first transmitter / receiver pair and the second transmitter / receiver pair can share one or more hardware components. For example, some transceivers 225, transmitters 230, and receivers 235 can be implemented as physically separate components that access shared hardware resources and / or software resources, such as a network interface 240.

[0067] One or more transmitters 230 and / or one or more receivers 235 may be implemented and / or integrated into a single hardware component, such as a multi-transceiver chip, a system on chip, an application specific integrated circuit (ASIC), or other types of hardware components. One or more transmitters 230 and / or one or more receivers 235 may be implemented and / or integrated into a multi-chip module. Other components such as a network interface 240 or other hardware components / circuits may be integrated into a single chip with any number of transmitters 230 and / or receivers 235. Transmitters 230 and receivers 235 may be logically configured as transceivers 225 using one or more common control signals, or may be configured as modular transmitters 230 and receivers 235 implemented in the same hardware chip or multi-chip module.

[0068] Figure 3Depicts additional details of a network node 300 that can be used to implement the methods described herein. The network node 300 can be an implementation of an entity in a wireless communication network (e.g., one or more of the wireless communication networks described herein). For example, the network node 300 can be the above-mentioned UE 200, or a network function (NF) or application function (AF) or another entity of one or more wireless communication networks of the embodiments described herein. The network node 300 includes: a processor 305, a memory 310, an input device 315, an output device 320, and a transceiver 325.

[0069] The input device 315 and the output device 320 may be combined into a single device, such as a touch screen. In some implementations, the network node 300 does not include any input device 315 and / or output device 320. The network node 300 may include one or more of a processor 305, a memory 310, and a transceiver 325, and may not include an input device 315 and / or an output device 320.

[0070] As depicted, the transceiver 325 includes at least one transmitter 330 and at least one receiver 335. Here, the transceiver 325 communicates with one or more remote units 102. In addition, the transceiver 325 can support at least one network interface 340 and / or an application interface 345. The (multiple) application interface 345 can support one or more APIs. The (multiple) network interface 340 can support 3GPP reference points such as Uu, N1, N2, and N3. As will be appreciated by one of ordinary skill in the art, other network interfaces 340 may be supported.

[0071] The processor 305 may include any known controller capable of executing computer-readable instructions and / or capable of performing logical operations. For example, the processor 305 may be a microcontroller, a microprocessor, a CPU, a GPU, an auxiliary processing unit, an FPGA, or a similar programmable controller. The processor 305 may execute instructions stored in the memory 310 to perform the methods and routines described herein. The processor 305 is communicatively coupled to the memory 310, the input device 315, the output device 320, and the transceiver 325.

[0072] The memory 310 may be a computer-readable storage medium. The memory 310 may include a volatile computer storage medium. For example, the memory 310 may include a RAM, including a dynamic RAM (DRAM), a synchronous dynamic RAM (SDRAM), and / or a static RAM (SRAM). The memory 310 may include a non-volatile computer storage medium. For example, the memory 310 may include a hard drive, a flash memory, or any other suitable non-volatile computer storage device. The memory 310 may include both a volatile computer storage medium and a non-volatile computer storage medium.

[0073] The memory 310 may store data related to establishing a multipath unicast link and / or mobile operation. For example, as described herein, the memory 310 may store parameters, configurations, resource allocations, policies, etc. The memory 310 may also store program code and related data, such as an operating system or other controller algorithms operating on the network node 300.

[0074] Input device 315 may include any known computer input device, including a touchpad, buttons, keyboard, stylus, microphone, etc. Input device 315 may be integrated with output device 320, for example, as a touch screen or similar touch-sensitive display. Input device 315 may include a touch screen so that text can be entered using a virtual keyboard displayed on the touch screen and / or by handwriting on the touch screen. Input device 315 may include two or more different devices, such as a keyboard and a touchpad.

[0075] Output device 320 can be designed to output visual signals, auditory signals, and / or tactile signals. Output device 320 may include an electronically controllable display or display device capable of outputting visual data to a user. For example, output device 320 may include, but is not limited to, an LCD display, an LED display, an OLED display, a projector, or a similar display device capable of outputting images, text, etc. to a user. As another non-limiting example, output device 320 may include a wearable display that is separate from the rest of network node 300 but communicatively coupled to the rest, such as a smart watch, smart glasses, a head-up display, etc. In addition, output device 320 may be a component of a smart phone, a personal digital assistant, a television, a desktop computer, a notebook (laptop) computer, a personal computer, a vehicle dashboard, etc.

[0076] Output device 320 may include one or more speakers for producing sound. For example, output device 320 may produce an audible alarm or notification (e.g., a beep or buzzer). Output device 320 may include one or more tactile devices for producing vibration, motion, or other tactile feedback. All or part of output device 320 may be integrated with input device 315. For example, input device 315 and output device 320 may form a touch screen or similar touch-sensitive display. Output device 320 may be located near input device 315.

[0077] The transceiver 325 includes at least one transmitter 330 and at least one receiver 335. One or more transmitters 330 can be used to communicate with the UE, as described herein. Similarly, one or more receivers 335 can be used to communicate with network functions in the PLMN and / or RAN, as described herein. Although only one transmitter 330 and one receiver 335 are shown, the network node 300 can have any suitable number of transmitters 330 and receivers 335. In addition, the (multiple) transmitters 330 and (multiple) receivers 335 can be any suitable type of transmitter and receiver.

[0078] The following information is helpful for understanding the method and apparatus for data preparation for analyzing data in the 3GPP architecture described below.

[0079] Currently, network analytics and AI / ML are deployed in the 5G core network via NWDAF. Various analysis types can be supported. Various analysis types can be distinguished using different analysis IDs, such as "UE mobility", "NF load", etc. This is discussed in TS23.288. Each NWDAF can support one or more analysis IDs and can have the following roles: (i) AI / ML reasoning, called NWDAF AnLF; or (ii) AI / ML training, called NWDAF MTLF; or (iii) both.

[0080] NWDAF AnLF (or AnLF for short) and NWDAF MTLF (or MTLF for short) represent logical functions that can be deployed as independent functions or in combination. An AnLF that supports reasoning for a specific analysis ID using an AI / ML model subscribes to the corresponding MTLF, which is responsible for training the same AI / ML model for the corresponding analysis ID.

[0081] Figure 4is a schematic diagram of a network 400, and illustrates various NWDAF "flavors" or types (particularly NWDAFAnLF / MTLF 402, NWDAF AnLF 404, and NWDAF MTLF 406), and their corresponding input data and output result consumers. Specifically, the analysis IDs contained in the NWDAFs 402, 404, 406 rely on various data input sources, including data from the 5G Core NF 408, AF 410, 5G Core Repository 412 (e.g., Network Repository Function (NRF), UDM, etc.), and OAM data 414 (e.g., PM / KPI, CM data, alarms, etc.). The analysis ID is contained in the AnLF, and the analysis output results can be provided to the 5G Core NF 416, AF 418, 5G Core Repository 420 (e.g., UDM, UDR ADRF or OAM MnS user or MF 422). The MTLF and AnLF can exchange AI / ML models, such as via serialization, containerization, etc., including relevant model information. Optionally, the DCCF and MFAF 424 may be involved to distribute or collect duplicate data to or from various data sources.

[0082] TS23.288 introduces ADRF, which supports storage and retrieval of analytics generated from NWDAF and other collected data.

[0083] Figure 5 It is a diagram depicting the data storage architecture for analysis and collection data defined in TS23.288. Figure 5 Depicts the ADRF architecture in 5G. Architecture 500 supports the following options:

[0084] -ADRF 502 provides storage and retrieval of data by other 5GC NF 504 (e.g., NWDAF).

[0085] - Based on the NF request or configuration on the DCCF 506, the DCCF 506 may determine the ADRF 502 and interact directly or indirectly with the ADRF 502 to request or store

[0086] Data. Interactions can be:

[0087] - directly, i.e., DCCF 506 requests data to be stored in ADRF 502, or via notification (e.g., when ADRF 502 requests data collection notification via DCCF 506)

[0088] In addition, DCCF 506 retrieves data from ADRF 502;

[0089] - Indirect: DCCF 506 requests the messaging framework 508 to store data in ADRF

[0090] 502. The message passing framework 508 may include communication between protocols defined by 3GPP.

[0091] One or more adapters for performing conversions;

[0092] - The consumer NF 504 may specify in a request to the DCCF 506 that the data provided by the data source needs to be stored in the ADRF 502 .

[0093] - The ADRF 502 stores data received directly from the NF 504, or received in a notification message from the DCCF 506, MFAF 510, or NWDAF.

[0094] - ADRF 502 checks whether the data consumer is authorized to access the ADRF service, and

[0095] The requested data shall be provided using the procedures specified in TS 23.501 clause 7.1.4.

[0096] In TR 23.700-81, ADRF storage and retrieval services are enhanced to support ML models. In other words, ADRF supports:

[0097] - Trained ML model file(s) and ML model file serialization format stored by NF consumer, i.e., NWDAF including MTLF.

[0098] - retrieval of trained ML model files(s) by NF Consumer,

[0099] That is, NWDAF including AnLF.

[0100] ADRF should be able to store historical ML models and use new ML model storage, provisioning, and update services to deliver historical ML models with a wider range of ML model filtering information.

[0101] By storing ML models in ADRF, more efficient ML model sharing can be enabled without overloading the MTLF that can handle multiple ML models and ML model consumers. Then, ML model consumers can subscribe and request ML models from the corresponding ADRF instead of the corresponding MTLF, thereby reducing their workload.

[0102] The trained ML model profile stored by the NWDAF containing the MTLF and retrieved by the NWDAF containing the AnLF may include one of the following parameters:

[0103] - Identification of the location relevant to the ML model, such as NWDAF ID, ADRFID, etc.

[0104] -The ML model ID associated with each historical ML model version stored.

[0105] - Analysis ID and / or Model Framework (i.e., Analysis Type, which may also include (e.g.,

[0106] Where not explicitly mentioned) corresponds to the standardized input data source).

[0107] -ML model implementation details, such as model platform, model type, algorithm, compilation language, etc.

[0108] -ML model serialization format, which handles interoperability, i.e., whether the retrieved model can be used in a specified platform or vendor.

[0109] -ML model usage information, which may include, but is not limited to, spatial validity, model validity period, model accuracy, model spatial validity, model file, S-NSSAI, region of interest, target object (e.g., (multiple) UE, (multiple) NF), model target period, and other model content information, such as that specified in TS23.288 Section 6.2A.2.

[0110] - The notification endpoint that is expected to receive ML models.

[0111] In some traditional applications, consumers can use application programming interfaces (APIs) to query to find and select shared content to build pipelines and / or cause the use of execution resources and storage to execute or train selected models or algorithms. The query can include one or more of the following: categories and / or subcategories, data information (i.e., data format availability to the consumer), available resources (e.g., processor type), timing information (e.g., expected latency), an indication of the pipeline that the model / algorithm / data will be used for, expected accuracy, content type (algorithm, model, and / or data), etc. The response to the query can include the relative usage of the data, storage location, and some test cases (i.e., inputs and outputs) to be used during verification and testing.

[0112] Current proposals focus on the ability to store and retrieve machine learning models taking into account various filtering information. However, none of these proposals describe and store the conditions and state of the network environment when training ML models at the MLTF. Although analysis IDs pointing to specific data sources that can be used in the training process can be included in the filtering information, the network topology and the use of these data sources are not deterministic and may vary depending on the network state.

[0113] Some prior art proposals in 3GPP (e.g., technical solution #62 in 3GPP TR 23.700-81 v1.1.0) disclose a method in which an analysis consumer includes a use case context identifier. The NWDAF can use this use case context identifier to select a specific ML model based on the user type, specific situation, or requested application. It is assumed that the NWDAF is pre-configured to select a specific ML model based on the requested use case. This process does not allow the NWDAF to determine the network conditions, i.e., network status (such as energy saving, failure, etc.) when such analysis is requested, nor does it allow the NWDAF to select an ML model based on network conditions (such as consumed data sources) when the ML model is trained.

[0114] This is important because the same ML model can be trained in the same geographic area (i.e., same spatial validity) at different times while the network state is different, for example due to failures at certain base stations. Energy conservation or special events (e.g., sports games) may also result in changes in network state. ML models trained in different network states (e.g., using different network topologies and / or load distributions) may produce different trained ML models. Therefore, it is not enough to capture only ML model implementations, interoperability, and usage.

[0115] This disclosure introduces the concept of network state associated with trained ML models, and introduces a method for storing, searching, and / or retrieving ML models when consumers request them for the purpose of inference or further training.

[0116] The present disclosure may assume that more than one ML model is available for selection for a particular analysis ID. Each ML model may carry or be associated with a unique identifier (i.e., the ML model ID proposed in TR 23.700-81 Section 6.61.2). The ML model ID may also be associated with the corresponding ML algorithm, for example, in a hierarchical naming relationship. The same geographic region and data source under different network states may be considered for training ML models, thereby generating different trained ML models that may be used for inference. Therefore, different ML model IDs preferably additionally consider and reflect different versions associated with the corresponding network state for training purposes. The relationship between the ML algorithm, network state version, or ID and the ML model ID is as shown in FIG. Figure 6 as shown in .

[0117] Figure 6 An example ML model ID 600 that may be implemented in embodiments of the present invention is depicted. The ML model ID 600 includes: an ML algorithm ID 602 ​​and a network state version or ID 604.

[0118] For example, a unique ML model ID can be introduced for each different ML algorithm, for example, for peak traffic and non-peak traffic conditions, if these conditions affect the ML model training and result in different trained ML models. The same applies to different network congestion conditions, network failures, and network configurations, or simply when different parts and types of data sources are used to train the ML model.

[0119] From the consumer's perspective, the selection of ML models is preferably adapted to the generation of analytics that reflect the encountered network conditions, environment, and usage scenarios, e.g., for a specific UE, or application, or network configuration, fault, or failure. Traditionally, NWDAF AnLF allows consumers to select accuracy level, time schedule, region of interest, and target object, but these attributes cannot determine the network condition semantics. In other words, assuming that NWDAF AnLF can accommodate multiple ML models, the current attributes are insufficient to allow analytics consumers to select relevant ML models (i.e., ML models trained under the required network conditions).

[0120] In addition, the current ML model information attributes listed in TS23.288 clause 6.2A.2 are often insufficient for the NWDAF to select relevant ML models for perturbation or further training. Explicit indications from analytics consumers and / or ML model consumers on network conditions, environment, and usage can help the NWDAF (AnLF and MTLF) select the most relevant ML models.

[0121] The network state may reflect network conditions, environment, and usage, which may take the form of an identifier or version (e.g., string or scalar) from a preconfigured list (known to analysis consumers, such as AF, NF, and OAM) and is valid throughout the PLMN. The analysis consumer (e.g., AF, NF, or OAM) may indicate the network state identifier or version to the NWDAF AnLF. Optionally, if the AnLF supports multiple ML models trained by different algorithms for the same analysis ID, and the analysis consumer knows this (i.e., knows the specific algorithm ID), the analysis consumer may indicate the desired ML algorithm ID. The network state identifier or version reflection may be a combination of the following:

[0122] - Network conditions or network information related to network objects (e.g., NFs or regions of interest)

[0123] Network configuration;

[0124] - User policy associated with the UE or session / flow (optional if specified in clause 6.1.3 TS

[0125] The target of the analysis report specified in 23.288 indicates a specific UE or a group of UEs).

[0126] When AnLF and MTLF are included in different NWDAF instances, AnLF can select the desired ML model ID from the available ML model IDs using the network status identifier or version. Alternatively, if MTLF supports multiple ML models trained by different algorithms for the same analysis ID, AnLF may need to indicate the desired ML algorithm ID.

[0127] The present disclosure enhances the existing process by introducing new parameters related to the network state version or identifier (and optionally the ML algorithm ID). For example, this can be included in the following items:

[0128] i) Nnwdaf_AnalyticsInfo service,

[0129] ii) Nnwdaf_AnalyticsSubscription service,

[0130] iii) Nnwdaf_MLModelInfo service, and / or

[0131] iv) Nnwdaf_MLModelProvision service.

[0132] In some embodiments, an analytics consumer (e.g., a NWDAF service consumer (e.g., NF / OAM)) can add parameters on the network state version or identifier (and optionally the algorithm ID) to the request and use the Nnwdaf_AnalyticsInfo service to get analytics information from the NWDAF. Figure 7 as shown in .

[0133] Figure 7 is a process flow diagram illustrating an embodiment of a process 700 for requesting and receiving network data analysis.

[0134] Process 700 involves NWDAF service consumer 702 and NWDAF 704 .

[0135] NWDAF service consumer 702 and / or NWDAF 704 may be the same as or consistent with any network entity, function, or node described herein. For example, NWDAF service consumer 702 and NWDAF 704 may be the same as Figure 3 The network node 300 shown is the same as that described in more detail above.

[0136] At step 706, the NWDAF service consumer 702 (e.g., NF / OAM) requests the analysis information by calling the Nnwdaf_AnalyticsInfo_Request service operation. The parameters that the NWDAF service consumer 702 can provide are listed in Section 6.1.3 of TS23.288 and include, but are not limited to, analysis ID, filter information, target object, time schedule, etc. Network state parameters (e.g., network state version or identifier) ​​and optional ML algorithm ID are included as (multiple) additional parameters. The NWDAF service consumer 702 sends a Nnwdaf_AnalyticsInfo_Request message containing the network state parameters to the NWDAF 704.

[0137] At step 708, the NWDAF 704 responds to the NWDAF service consumer 702 with analytical information.

[0138] In some embodiments, an analytics consumer (i.e., NWDAF service consumer (e.g., NF / OAM)) may subscribe / unsubscribe at the NWDAF using the Nnwdaf_AnalyticsSubscription service to indicate the network state parameters (e.g., network state version or identifier) ​​and optional ML algorithm ID to be notified on analytics information. The Nnwdaf_AnalyticsSubscription service may also be used to modify existing analytics subscription(s).

[0139] Figure 8 is a process flow diagram illustrating an embodiment of a process 800 for subscribing / unsubscribing to be notified of network data analytics.

[0140] Process 800 involves NWDAF service consumer 802 and NWDAF 804 .

[0141] NWDAF service consumer 802 and NWDAF 804 may be the same as or consistent with any network entity, function, or node described herein. For example, NWDAF service consumer 802 and / or NWDAF 804 may be the same as Figure 3 The network node 300 shown is the same as that described in more detail above.

[0142] At step 806, the NWDAF service consumer 802 (e.g., NF / OAM) subscribes to or unsubscribes from the analytics information by calling the Nnwdaf_AnalyticsSubscription_Subscribe / Nnwdaf_AnalyticsSubscription_Unsubscribe service operations. The parameters that the NWDAF service consumer may provide are listed in clause 6.1.3 TS 23.288. Network state parameters (e.g., network state version or identifier) ​​and optionally the algorithm ID are included as additional parameters.

[0143] At step 808, if the NWDAF service consumer 802 subscribes to the analysis information, the NWDAF 804 notifies the NWDAF service consumer 802 of the analysis information by calling the Nnwdaf_AnalyticsSubscription_Notify service operation based on the request from the NWDAF service consumer 802 (e.g., the analysis report parameters including the network state version of the identifier (and optionally the algorithm ID)). If the NWDAF 804 provides a termination request, the consumer 802 cancels its subscription to the analysis information by calling the Nnwdaf_AnalyticsSubscription_Unsubscribe service operation.

[0144] In some embodiments, the NWDAF AnLF (i.e., the NWDAF service consumer in this embodiment (according to clause 6.2A.3 TS 23.288)) may add network state parameters (e.g., network state version or identifier), and optionally algorithm ID, when requesting and obtaining ML model information from the NWDAF MTLF using the Nnwdaf_MLModelInfo service. Fig. 9 as shown in .

[0145] Fig. 9 is a process flow diagram illustrating an embodiment of a process 900 for requesting and receiving ML model information.

[0146] Process 900 involves a NWDAF service consumer 902 (ie, NWDAF AnLF) and a NWDAF MTLF 904.

[0147] The NWDAF service consumer 902 (ie, NWDAF AnLF) and the NWDAF MTLF 904 may be the same as or consistent with any network entity, function, or node described herein. For example, the NWDAF service consumer 902 (ie, NWDAF AnLF) and / or the NWDAF MTLF 904 may be the same as or consistent with any network entity, function, or node described herein. Figure 3The network node 300 shown is the same as that described in more detail above.

[0148] At step 906, the NWDAF service consumer 902 (i.e., NWDAF AnLF) requests the ML model(s) associated with the analysis ID(s) by calling the Nnwdaf_MLModelInfo_Request service operation. The parameters that the NWDAF service consumer 902 may provide are listed in clause 6.2A.2 of TS 23.288, and may include network state parameters (e.g., network state version or identifier), and an optional ML algorithm ID.

[0149] When a request for ML model information for analysis is received, the NWDAF MTLF 904 may:

[0150] - Determine whether an existing trained ML model can be used for the request; or

[0151] -Determine whether the request needs to trigger further training of an existing trained ML model.

[0152] If the NWDAF 904 including the MTLF determines that further training is needed, the NWDAF 904 may initiate data collection to generate an ML model.

[0153] At step 908, the NWDAF MTLF 904 responds to the NWDAF service consumer 902 with ML model information (including the file address(es) of the trained ML model) by calling the Nnwdaf_MLModelInfo_Request response service operation. The content of the ML model information that the NWDAF 904 including the MTLF may provide may be specified in clause 6.2A.2 TS 23.288 and also includes network state parameters (e.g., network state version or identifier) ​​and an optional ML algorithm ID.

[0154] In some embodiments, the NWDAF AnLF (i.e., NWDAF service consumer (according to TS23.288 Section 6.2A.1)) uses the Nnwdaf_MLModelProvision service to add network state parameters (e.g., network state version or identifier) ​​and optional ML algorithm ID when subscribing / unsubscribing to the NWDAF containing MTLF, so as to be notified when ML model information for relevant analysis is available.

[0155] Fig.10 is a process flow diagram illustrating an embodiment of a process 1000 of subscribing / unsubscribing to be notified of ML model information.

[0156] Process 1000 involves a NWDAF service consumer 1002 (ie, NWDAF AnLF) and a NWDAF MTLF 1004 .

[0157] The NWDAF service consumer 1002 (ie, NWDAF AnLF) and the NWDAF MTLF 1004 may be the same as or consistent with any network entity, function, or node described herein. For example, the NWDAF service consumer 1002 (ie, NWDAF AnLF) and / or the NWDAF MTLF 1004 may be the same as or consistent with any network entity, function, or node described herein. Figure 3 The network node 300 shown is the same as that described in more detail above.

[0158] At step 1006, the NWDAF service consumer 1002 (i.e., NWDAF AnLF) subscribes, modifies, or unsubscribes to the trained ML model(s) associated with the analysis ID(s) by calling the Nnwdaf_MLModelProvision_Subscribe / Nnwdaf_MLModelProvision_Unsubscribe service operations. The parameters that the NWDAF service consumer may provide are listed in Section 6.2A.2 TS 23.288 and also include network state parameters (e.g., network state version or identifier) ​​and an optional ML algorithm ID.

[0159] When a subscription to a trained ML model associated with an analysis ID is received, the NWDAF 1004 containing the MTLF may:

[0160] - Determine whether an existing trained ML model is available for subscription; or

[0161] -Determine if a subscription needs to trigger further training of an existing trained ML model.

[0162] If the NWDAF 1004 including the MTLF determines that further training is needed, the NWDAF 1004 may initiate data collection to generate an ML model.

[0163] At 1008, the NWDAF MTLF 1004 notifies the NWDAF service consumer 1002 (i.e., NWDAF AnLF) of the trained ML model information (file address (set) containing the trained ML model) by calling the Nnwdaf_MLModelProvision_Notify service operation. The content of the trained ML model information that the NWDAF 1004 containing MTLF can provide can be specified in Section 6.2A.2 of TS23.288, and also includes network state parameters (e.g., network state version or identifier) ​​and an optional ML algorithm ID.

[0164] For example, when the NWDAF MTLF 1004 determines at step 1006 that the previously provisioned trained ML model needs to be retrained (which may be determined when the network status changes), the NWDAF MTLF 1004 calls the Nnwdaf_MLModelProvision_Notify service operation to notify the available retrained ML model.

[0165] When step 1006 is performed for a subscription modification or a network status change (eg, including a subscription correlation ID), the NWDAF MTLF 1004 may provide a new trained ML model or a retrained ML model.

[0166] In some embodiments, the network status parameter may identify, for example, one or more of the following network attributes (i.e., the network status associated with the network version or identifier may be described by, for example, one or more of the following attributes):

[0167] - Network type, which may include network technology and purpose of use, such as:

[0168] - Private networks (e.g., 5G factories) or public networks in rural or urban environments;

[0169] -The radio access technology (RAT) being used, e.g. 5G, 4G, satellite, etc.

[0170] - Network faults, errors and / or failures within or associated with the network. Network status parameters may identify the objects involved and the associated faults, errors or failures, including security events. Network status parameters may identify:

[0171] - Network objects,

[0172] - the type of fault, error or malfunction,

[0173] - the duration of the fault, error or malfunction,

[0174] -Back up network objects or functions that are in use.

[0175] - The energy-saving state of the network. The network state parameters may identify the objects involved and the associated energy modes. The network state parameters may identify:

[0176] - network objects in energy-saving state,

[0177] - the duration that the network object is in the energy-saving state,

[0178] - Energy saving state, i.e., energy mode associated with a specific network device, such as power off or interrupted transmission, etc.

[0179] - The network maintenance status of the network. The network status parameter may identify network objects that are out of operation and involved in the upgrade. The network status parameter may identify:

[0180] - the network object stops operating,

[0181] - Network objects with different maintenance characteristics, e.g. software versions,

[0182] - Duration of time that the network object stopped operating.

[0183] - Network congestion level, which may include the mean, standard deviation, or congestion distribution. The network congestion level may be expressed as:

[0184] - quantitative values, such as high, medium, low,

[0185] -Average network load as a percentage of network capacity.

[0186] - Network configuration. This can indicate network conditions, for example in terms of:

[0187] -Connectivity settings, such as identification of network slices, RATs, reflectors / relays, etc.,

[0188] - Slice type, such as eMBB, mIoT, URLLC, V2X, etc.,

[0189] -DNN, which stores certain provided services,

[0190] - Quality of service, e.g. in terms of SLA, which may include throughput, latency, etc., - Policy and control, e.g. data offloading, route optimization, use of CoMP, etc.,

[0191] -Network ownership or network object ownership, such as edge cloud platforms.

[0192] -The data sources used for ML model training. This can include identifying:

[0193] - the identity of the data source from which the data was obtained,

[0194] -The state or condition of the data source object, such as being in a power-saving state (power off), fault, load condition, congestion level, etc.

[0195] - Duration of time to obtain data.

[0196] - Data statistics, which can specify or take into account the distribution and correlation (i.e., dependencies) of data in terms of:

[0197] - data ranges, such as minimum and maximum values,

[0198] - the time interval between consecutive data acquired,

[0199] - duration of acquisition data,

[0200] - the amount of data or data acquired,

[0201] -The frequency with which data occur or recur, such as a data value or state.

[0202] - Historical data, which may be located in the DCCF / MFAF or ADRF. Historical data may be specified in TS23.288 clauses 5A.4 and 5A.5.

[0203] - Tracking data, which can be represented as a "pointer" or indicator, which is used to identify network tracking data in a corresponding tracking database related to the network status, such as PM, MDT data, or KPI. This "pointer" to the tracking data

[0204] The "needle" or indicator can be described by one or a combination of the following:

[0205] - a time instance, i.e., an initial or starting time instance associated with a desired PM, MDT data, or KPI,

[0206] - Duration associated with desired PM, MDT data, or KPI,

[0207] -Geographic areas or network objects associated with the desired PM, MDT data, or KPIs,

[0208] - Tracking measurements, such as PM, MDT data or KPIs, each consumer potentially requests a different set of tracking to meet its needs.

[0209] It should be noted that the above attributes are not necessarily determined by the NWDAF MTLF. Data-related attributes (i.e., (multiple) data sources, data statistics, and historical data) can be determined by the NWDAF MTLF. Other attributes can be obtained based on interactions with OAM and / or other 5G NFs (e.g., PCF), and can, for example, take into account usage related to user type, specific situations (i.e., network conditions), or applications (i.e., analysis IDs).

[0210] For example, analysis IDs related to areas of interest and / or NFs may require network status information from OAM related to, for example, power saving, faults, etc. Other analysis IDs related to specific UEs, sessions, QoE, or applications may require additional information from PFC related to policy and control.

[0211] Service processes related to ML model storage, registration, search, and / or provision may include, but are not limited to, the following:

[0212] - ML model storage towards ADRF, which may include network state version and identifier, optionally including algorithm ID.

[0213] - ML model registration in DCCF / MFAF or NRF, which may include the network state version and identifier, and may optionally include the ML algorithm ID.

[0214] -ML model search in NRF, which can include network state version and identifier,

[0215] Optionally includes the ML algorithm ID.

[0216] ML model provisioning towards an ML model consumer, which may be a NWDAF AnLF or a NWDAF MTLF, which may include a network state version and identifier, and may optionally include an ML algorithm ID.

[0217] For some embodiments:

[0218] - The network status can be determined by NWDAF MTLF, for example, by one of the following:

[0219] or more to get:

[0220] - Consider the data sources used in the training process. The data sources can provide "fresh" or new data directly. As an alternative or in addition to new data, historical data can be provided via DCCF / MFAF, ADRF, or other data tracking tools.

[0221] - Consider including in the data source additionally the network type associated with the respective data source.

[0222] The consideration further includes the status of the data source, ie the condition of the network object of the corresponding data source.

[0223] - Additional considerations include data statistics that may be calculated by the MTLF or acquired by another data entity.

[0224] -NWDAF MTLF can obtain network status by interacting with OAM.

[0225] The NWDAF MTLF may do one or more of the following:

[0226] - Understand why some of the listed data sources are unavailable or not operating as expected. "Inference" tags can be introduced that can indicate or identify, for example, failures, malfunctions, safety issues, maintenance, energy-saving status, etc.

[0227] - Get status related to a data source, such as <name, availability, start_time, time_duration>.

[0228] - Get a pointer to the corresponding trace data (e.g., data entry number in the trace database), which can enable ML model consumers to interact with the trace database or any other similar tool that keeps network traces and retrieve network conditions when ML model training is executed.

[0229] - Get information related to the region of interest, network entity or object (e.g., MnS, slice, etc.)

[0230] The relative level of network congestion (eg, average level).

[0231] -NWDAF MTLF can obtain network status from 5G NF (e.g., PCF).

[0232] The NWDAF MTLF may do one or more of the following:

[0233] - Get the SLA associated with a specific service.

[0234] - Obtain configuration policies (e.g., route optimization, data offloading, connectivity, etc.) related to the provisioning of services or slices in the form of, for example:<policy_name,network_object_list,start_time,time_duration> .

[0235] - Get ownership information, such as <name, network object>; the name can be an identifier for a vertical, vendor, application provider, or a generic third party.

[0236] Alternatively, NWDAF may instruct OAM and / or other 5G NFs to provide information about the network status directly to ADRF by sharing a model ID or another generic identifier associated with the ML model that requires network status information update in ADRF. This may be provided along with a schedule, region of interest, network objects, etc. In some embodiments, DCCF / MFAF and / or a dedicated NF in ADRF (which may be responsible for coordinating the network status with the ML model storage in ADRF) or a logical NF may help obtain the network status while taking into account the model ID or another generic identifier associated with the ML model, as well as the schedule, region of interest, network objects, etc.

[0237] Embodiments described herein may enhance TR 23.700-81 Technical Solution #43, ML model storage in ADRF, and ML model provisioning of ADRF by introducing network state parameters (e.g., network state version and identifier), and may optionally include an algorithm ID (e.g., Figure 6 As shown and / or as referenced below Fig.11 as described above) to complement existing services.

[0238] Fig.11 11 is a process flow diagram illustrating a process 1100 of ML model storage, registration, and / or provisioning and a process of subscribing / unsubscribing to be notified of ML model information according to an embodiment of the present invention.

[0239] Process 1100 involves ML model consumer 1102 (eg, NWDAF AnLF), NWDAF MTLF 1104 , DCCF / MFAF 1106 , NRF 1108 , and ADRF 1110 .

[0240] ML model consumer 1102, NWDAF MTLF 1104, DCCF / MFAF 1106, NRF 1108, and / or ADRF 1110 may communicate with Figure 3 The network node 300 shown is the same as that described in more detail above.

[0241] ML model creation and storage process Fig.11 1112 and is indicated by reference numeral 1112. The ML model creation and storage process 1112 includes steps 1114 to 1120.

[0242] At step 1114 , the NWDAF MTLF 1104 provides or determines training to the ML model(s).

[0243] At step 1116, the NWDAF MTLF 1104 requests to store the ML model to the ADRF 1110 by calling the Nadrf_MLModel_StorageRequest service operation. The service operation contains the trained ML model(s) and / or ML model(s) information, as well as network state parameters (e.g., network state version and identifier), and optionally includes the ML algorithm ID). Alternatively, the NWDAF MTLF 1104 stores the trained ML model to the ADRF 1110 via the DCCF 1106.

[0244] At step 1118, ADRF 1110 stores the trained ML model(s) and / or ML model(s) information, including network state parameters and optional ML algorithm IDs. ADRF 1110 may determine whether the same trained ML model has already been stored. If the trained ML model has already been stored, ADRF 1110 may decide to update it, including the associated network state parameters and optional ML algorithm IDs. In some embodiments, if the new ML model is the same as the previous ML model, ADRF 1110 may only update the relevant network state information.

[0245] At step 1120 , the ADRF 1110 sends a response to the NWDAF MTLF 1104 to indicate that the ML model is stored or updated using a Nadrf_MLModel_StorageRequestResponse. This can be done directly or via the DCCF 1106.

[0246] Therefore, a ML model creation and storage process is provided.

[0247] The ML model registration process is Fig.11 1122. The first ML model registration process includes steps 1124 to 1128. The second ML model registration process includes steps 1130 to 1134.

[0248] In the first ML model registration process, the NWDAF MTLF 1104 and / or the ADRF 1110 registers the ML model with the DCCF / MFAF 1106. In the second ML model registration process, the NWDAF MTLF 1104 and / or the ADRF 1110 registers the ML model with the NRF 1108. In either case, the ML model may be distributed to the corresponding NWDAF AnLF or NWDAFMTLF that requested it.

[0249] In the first ML model registration process, at step 1124, the NWDAF MTLF 1104 and / or the ADRF 1110 may request registration of the ML model profile (including any appropriate attributes), including network state parameters (e.g., network state version and identifier), and optionally the ML algorithm ID, with the DCCF 1106 by calling Ndccf_MLModel_Registration.

[0250] At step 1126 , the DCCF / MFAF 1106 registers the ML model and corresponding network state parameters (eg, network state version and / or identifier), including an optional ML algorithm ID.

[0251] At step 1128 , the DCCF / MFAF 1106 responds to the NWDAF MTLF 1104 and / or ADRF 1110 with a Ndccf_MLModel_RegistrationResponse.

[0252] In the second ML model registration process, at step 1130 , the NWDAF MTLF 1104 and / or ADRF 1110 registers its ML model profile, including network state parameters (e.g., network state version and / or identifier), and optionally the ML algorithm ID, with the NRF 1108 by calling Nnrf_MLModel_Registration.

[0253] At step 1132 , the NRF 1108 registers the ML model and corresponding network state parameters (eg, network state version and identifier), optionally including the ML algorithm ID.

[0254] At step 1134 , the NRF 1108 responds to the NWDAFMTLF 1104 and / or the ADRF 1110 with a Nnrf_MLModel_RegistrationResponse.

[0255] Therefore, an ML model registration process is provided.

[0256] The ML model can be provisioned to consumers, such as NWDAF AnLF (e.g., for inference) or NWDAF MTLF (e.g., for further training). The ML model provisioning process is described in Fig.11 1136 and is indicated by reference numeral 1136. The ML model provisioning process 1136 includes steps 1138 to 1144.

[0257] At step 1138, the ML model consumer 1102 optionally issues a search request to the NRF 1108 to find the location of the desired ML model by calling Nnrf_MLModel_SearchRequest (e.g., as defined in TS 23.288). This may include the desired network state parameters (e.g., network state version and identifier), and may optionally include the ML algorithm ID.

[0258] At step 1140, the NRF 1108 responds with a Nnrf_MLModel_SearchRequestResponse, which contains the ML model and network state parameters (e.g., network state version and identifier), and may optionally include the ML algorithm ID if the ML model is available. Otherwise (i.e., if the ML model is not available), the NRF 1108 provides an error message indicating that the desired ML model does not exist / is not available.

[0259] At step 1142, the ML model consumer 1102 (e.g., NWDAF AnLF or NWDAF MTLF) may request or subscribe to the DCCF 1106 or ADRF 1110 for (multiple) trained ML model(s) associated with (multiple) analysis ID(s) and network state parameters (e.g., network state version and / or identifier) ​​(optionally including ML algorithm ID) by calling Nadrf_MLModel_Request / Subscribe.

[0260] At step 1144, the ADRF 1110 or DCCF 1106 notifies the ML model consumer 1102 of the trained ML model information, including network state parameters, and optionally the ML algorithm ID. This can be done by responding with Nadrf_MLModel_Request / Subscribe_Response. The response may contain the file address(es) of the trained ML model.

[0261] Therefore, an ML model provisioning process 1136 is provided.

[0262] In an alternative embodiment, if the analysis service consumer includes a use case context parameter, the NWDAF determines the type of ML model algorithm required to support such a use case. The NWDAF can then determine the appropriate ML model algorithm based on the current network state configuration or version. In some embodiments, when the NWDAF determines the network state (e.g., as described in the above text), the NWDAF can interface with the ADRF to obtain an ML model that supports a specific algorithm trained based on the determined network state.

[0263] In one aspect, an apparatus is provided, the apparatus comprising a transceiver and a processor coupled to the transceiver. The processor and the transceiver are configured such that the apparatus: includes a network state parameter in a message identifying a machine learning model. The machine learning model is used to derive (e.g., by inference) analysis information of a wireless communication network; when the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network. The network state parameter indicates a specific network state.

[0264] The processor and transceiver may also be configured to cause the apparatus to: send the message to another apparatus (eg, to a network function on another apparatus).

[0265] The message may be associated with the storage, registration, search, or provision of a machine learning model or an analysis derived therefrom. For example, the message may be a request to store, register, search, provision, or provide a machine learning model and / or an analysis derived therefrom. Also, for example, the message may be a storage request response, a registration request response, a search result associated therewith, or a message to provide a machine learning model and / or an analysis derived therefrom. Also, for example, the message may be a subscription message, such as a subscription request or response for network analysis information and / or ML model information (e.g., as described in more detail above).

[0266] The message may include an identifier of the machine learning model (e.g., a model ID). The identifier may include an identifier of a machine learning algorithm used to define the machine learning model (e.g., an algorithm ID), which is combined (e.g., concatenated) with the network state parameter. Optionally, the message may also include an analysis ID for identifying an analysis (e.g., the requested analysis).

[0267] The network status parameter may be selected from a list of predefined network status parameters. For example, the processor may be configured to select (or refrain from selecting) a network status parameter from a predefined list.

[0268] The network state parameters may specify, include, define, or indicate one or more of the following: one or more data sources from which the training data is obtained; and / or one or more data statistics related to the training data.

[0269] The network state parameter may specify, include, define, or indicate one or more of the following: the type of the wireless communication network (e.g., when in a particular network state); one or more network faults, errors, or malfunctions in the wireless communication network (e.g., when in a particular network state); a power saving state of the wireless communication network (e.g., when in a particular network state); a network maintenance state of the wireless communication network (e.g., when in a particular network state); a network congestion level of the wireless communication network (e.g., when in a particular network state); a network configuration (e.g., configuration policies and / or controls) of the wireless communication network (e.g., when in a particular network state); and / or historical data associated with the wireless communication network (e.g., when in a particular network state). The historical data may be specified in TS23.288 Sections 5A.4 and 5A.5.

[0270] The network status parameter may specify, include, define, or indicate one or more of the following: tracing data or an identifier thereof (e.g., a pointer); and / or a tracing tool or an identifier thereof (e.g., a pointer). The tracing tool may be used to retrieve the network status or network context. The tracing tool may be as defined in TS 32.421 or TS 32.422, such as an MDT server (TCE).

[0271] The device can be selected from: network data analysis function NWDAF; NWDAF model training logic function MTLF; NWDAF analysis logic function AnLF; data collection coordination function DCCF; messaging framework adapter function MFAF; network repository function NRF; analysis data repository function ADRF; and machine learning model consumer.

[0272] The processor and transceiver may also be configured to cause the device to interact with another device (e.g., a network function on another device in a wireless communication network). The processor and transceiver may also be configured to cause the device to: determine a network state parameter based on the interaction and the information provided. The other device interacting with it may be selected from: an operation, administration and maintenance OAM entity; a 5G network function NF; and a policy control function PCF.

[0273] A method executable by an apparatus in a wireless communication network is also provided. Fig.1212 is a process flow diagram illustrating certain steps of the method 1200. The method 1200 includes: including 1202, by a processor, a network state parameter in a message identifying a machine learning model. The machine learning model is used to derive (e.g., by inference) analysis information of a wireless communication network. When the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network. The network state parameter indicates a specific network state. Optionally, the method may also include: sending 1204, by a transceiver, the message, for example, for use by another device.

[0274] A device is also provided, the device comprising a memory and a processor coupled to the memory. The processor and the memory are configured so that the device: stores a machine learning model in the memory, the machine learning model is used to derive (e.g., by inference) analysis information of a wireless communication network, when the wireless communication network is in a specific network state, the machine learning model has been trained using training data obtained from the wireless communication network; and stores a network state parameter associated with the machine learning model in the memory, the network state parameter indicating the specific network state.

[0275] A computer-implemented method executable by a device in a wireless communication network is also provided. Fig.13 is a process flow diagram illustrating certain steps of the method 1300. The method 1300 includes: storing 1302 a machine learning model in a memory, the machine learning model being used to derive analysis information of a wireless communication network, the machine learning model having been trained using training data acquired from the wireless communication network when the wireless communication network is in a specific network state; and storing 1304 a network state parameter associated with the machine learning model in the memory, the network state parameter indicating the specific network state.

[0276] An analysis consumer is also provided, comprising a transceiver and a processor coupled to the transceiver. The processor and the transceiver are configured such that the device: includes a network state parameter in a message. The message is a request to be provided with analysis information or a subscription to analysis information (i.e., a subscription for receiving or being notified of analysis information). The network state parameter indicates a specific network state. The network state parameter can be used for selection of a machine learning model, for example at an NWDAF AnLF. When the NWDAF MTLF receives a retraining request for a specific different network state from the NWDAF AnLF, the network state parameter can also be used to avoid selection of a data source for retraining the machine learning model, for example at the NWDAF MTLF. The processor and the transceiver can also be configured such that the device: sends the message to another device (e.g., to the NWDAF-AnLF).

[0277] A method performed by an analytics consumer is also provided. Fig.14 14 is a process flow diagram illustrating certain steps of the method 1400. The method 1400 includes: including 1402 a network state parameter in a message, wherein: the message is a request to be provided with analysis information or a subscription to analysis information; the network state parameter indicates a specific network state and can be used for selection of a machine learning model ML model (e.g., at or through the NWDAF AnLF); and sending 1404 the message to another device (such as the NWDAF AnLF on another device).

[0278] In some embodiments, methods 1200 , 1300 , and 1400 may be performed by a processor that executes program code, such as a microcontroller, a microprocessor, a CPU, a GPU, an auxiliary processing unit, an FPGA, or the like.

[0279] Traditional approaches tend to focus on the ability to store and retrieve ML models, while considering filtering information related to implementation, interoperability, and usage, including requests and subscriptions. Traditional approaches do not consider conditions and network states when training ML models. ML models can be trained in different network states, for example, due to base station failures, energy conservation, or special events such as sports games. Different network states (e.g., different network topologies with different mobility and load distribution) may produce different trained ML models. These network conditions are currently invisible to ML model consumers.

[0280] The apparatus and method described above introduce network state parameters / identifiers that describe the training conditions of the ML model, for example, from a preconfigured list. Such network state parameters specify, define, or reflect network faults, errors, malfunctions, energy saving states, congestion conditions, maintenance, network configurations, active data sources, and / or data characteristics.

[0281] When a consumer requests an ML model for the purpose of inference or further training under a specific network state, the above-mentioned apparatus and method tend to provide capabilities for requesting information, subscribing, storing, searching, and retrieving the ML model.

[0282] Traditionally, there is no description or indication of the network conditions under which the NWDAF MTLF trains the ML model. Even if an analysis ID pointing to a specific data source that can be used for the training process and the corresponding geographic area and time is indicated, the network conditions may change (for example, due to failures, energy saving, or special events), thereby changing the network topology, mobility, and traffic distribution, which often significantly affects the results of ML model training. If these network conditions are not taken into account in the ML model storage, its use may be problematic (for example, ML models trained under different conditions and / or using different data sources and / or different data source statistics may be selected. This may result in significant performance drift, depending on the use case). The embodiments described herein address this problem.

[0283] Embodiments described herein include network status parameters when requesting and subscribing to analytics and ML model information from corresponding NWDAFs.

[0284] Embodiments described herein include network state parameters when the NWDAF MTLF stores and / or retrieves ML models from the ADRF.

[0285] Embodiments described herein provide for registration, search, and provisioning of ML models including relevant network states in DCCF / MFAF and / or NRF.

[0286] Provided is an apparatus and method that introduces a network state version or identifier to describe the training conditions or network environment associated with a trained ML model for storage, registration, search, and supply to ML model consumers. The network state version and identifier can be combined with the algorithm ID to describe the model ID. The network state or network context may include data sources and data statistics. The network state or network context may further include, but is not limited to, network type, network failure, energy saving state, network maintenance, network congestion level, and / or network configuration policy and control. Pointers to tracking tools that can help ML model consumers retrieve network state or network context may be included. The network state can be obtained in the ML model training entity by interacting with the 5G core network or OAM, or by the model training entity.

[0287] It should be noted that the above methods and apparatus illustrate rather than limit the present invention, and that those skilled in the art will be able to design many alternative arrangements without departing from the scope of the appended claims. The word "comprising" does not exclude the presence of other elements or steps than those listed in the claims, "a" or "an" does not exclude a plurality, and a single processor or other unit may fulfil the functions of several units described in the claims. Any reference signs in the claims should not be construed as limiting their scope.

[0288] Furthermore, although examples have been given in the context of specific communication standards, these examples are not intended to limit the communication standards to which the disclosed methods and apparatus may be applied. For example, although specific examples have been given in the context of 3GPP, the principles disclosed herein may also be applied to another wireless communication system, in fact any communication system that uses routing rules.

[0289] The method may also be embodied in a set of instructions stored on a computer-readable medium, which when loaded into a computer processor, a digital signal processor (DSP), etc., causes the processor to perform the above-described method.

[0290] The described methods and apparatus may be practiced in other specific forms. The described methods and apparatus are considered in all respects to be illustrative only and not restrictive. Therefore, the scope of the present invention is indicated by the appended claims rather than by the foregoing description. All changes within the meaning and equivalent range of the claims should be included within their scope.

[0291] The following abbreviations are relevant to the areas covered by this document:

[0292] 3GPP: Third Generation Partnership Project

[0293] 5G: The fifth generation of mobile communications

[0294] AI / ML: Artificial Intelligence / Machine Learning

[0295] ADRF: Analytical Data Repository Function

[0296] AF: Application Function

[0297] AnLF: Analytical Logic Function

[0298] API: Application Protocol Interface

[0299] DCCF: Data Collection Coordination Function

[0300] DNN: Data Network Name

[0301] eMBB: enhanced mobile broadband

[0302] KPI: Key Performance Indicator

[0303] MDT: Minimization of driving tests

[0304] MF: Management Function

[0305] MFAF: Messaging Framework Adapter Function

[0306] mIoT: Massive Internet of Things

[0307] MnS: Management Services

[0308] MTLF: Model Training Logical Function

[0309] NF: Network Function

[0310] NRF: Network Repository Function

[0311] NWDAF: Network Data Analysis Function

[0312] OAM: Operations, Administration, and Maintenance

[0313] PM: Performance Measurement

[0314] UDM: User Data Manager

[0315] UDR: User Data Repository

[0316] UE: User Equipment

[0317] URLLC: Ultra-Reliable Low-Latency Communications

[0318] V2X: Vehicle to Everything

Claims

1. A device comprising: Transceiver; as well as a processor coupled to the transceiver, the processor and the transceiver being configured to cause the apparatus to: A network state parameter is included in the message identifying the machine learning model, where: The machine learning model is used to obtain analysis information of the wireless communication network; When the wireless communication network is in a specific network state, the machine learning model has been trained using training data acquired from the wireless communication network; and The network status parameter indicates the specific network status. 2 . The apparatus according to claim 1 , wherein the processor and the transceiver are further configured to cause the apparatus to: send the message to another apparatus.

3. An apparatus according to any of the preceding claims, wherein the message is associated with the storage, registration, search, or provision of the machine learning model or an analysis derived from the machine learning model.

4. The apparatus according to any one of the preceding claims, wherein the message comprises: The identifier of the machine learning model.

5. The apparatus of claim 4, wherein the identifier of the machine learning model comprises: an identifier of a machine learning algorithm used to define the machine learning model in combination with the network state parameter; as well as Optional analysis ID identifying the analysis.

6. The apparatus according to any of the preceding claims, wherein the network status parameter is selected from a list of predefined network status parameters.

7. The apparatus according to any one of the preceding claims, wherein the network status parameter indicates one or more of the following: One or more data sources from which the training data is obtained; and / or One or more data statistics associated with the training data.

8. The apparatus according to any one of the preceding claims, wherein the network status parameter indicates one or more of the following: The type of the wireless communication network; one or more network failures in the wireless communication network; an energy-saving state of the wireless communication network; a network maintenance status of the wireless communication network; a network congestion level of the wireless communication network; A network configuration of the wireless communication network; and / or Historical data associated with the wireless communication network.

9. The apparatus according to any one of the preceding claims, wherein the network status parameter indicates one or more of the following: Tracking data or its identifiers; and / or A tracing tool or an identifier thereof, the tracing tool being used to retrieve a network state or a network context.

10. The device according to any of the preceding claims, wherein the device is a device selected from the group consisting of: Network data analysis function NWDAF; NWDAF model training logic function MTLF; NWDAF analysis logic function AnLF; Data Collection Coordination Function DCCF; Messaging Framework Adapter Function MFAF; Network repository function NRF; Analytical Data Repository Function (ADRF); and Machine learning model consumers.

11. The apparatus of any preceding claim, wherein the processor and the transceiver are further configured to cause the apparatus to: interact with another apparatus and determine the network status parameter based on the interaction.

12. The apparatus of claim 11, wherein the other apparatus is a apparatus selected from the group consisting of: Operate, manage and maintain OAM entities; 5G network functions NF; and Policy Control Function PCF.

13. A method comprising: The processor includes, in a message identifying the machine learning model, a network state parameter; wherein: The machine learning model is used to obtain analysis information of the wireless communication network; When the wireless communication network is in a specific network state, the machine learning model has been trained using training data acquired from the wireless communication network; and The network status parameter indicates the specific network status.

14. An apparatus comprising: Memory; as well as a processor coupled to the memory, the processor and the memory being configured to cause the apparatus to: storing a machine learning model in the memory, the machine learning model being used to derive analysis information of a wireless communication network, the machine learning model having been trained using training data acquired from the wireless communication network when the wireless communication network is in a specific network state; as well as A network state parameter associated with the machine learning model is stored in the memory, and the network state parameter indicates the specific network state.

15. A method comprising: storing a machine learning model in a memory, the machine learning model being used to derive analysis information of a wireless communication network, the machine learning model having been trained using training data acquired from the wireless communication network when the wireless communication network is in a specific network state; as well as A network state parameter associated with the machine learning model is stored in the memory, and the network state parameter indicates the specific network state.

16. An analysis of consumers, comprising: Transceiver; as well as a processor coupled to the transceiver, the processor and the transceiver being configured to cause the apparatus to: including network status parameters in the message; as well as sending the message to another device; wherein: The message is a request to be provided with analytical information or a subscription to said analytical information; The network status parameter indicates a specific network status; and The network status parameters can be used for selecting a machine learning model.

17. A method performed by an analytics consumer, the method comprising: including network status parameters in the message; as well as sending the message to another device; wherein: The message is a request to be provided with analytical information or a subscription to said analytical information; The network status parameter indicates a specific network status; and The network status parameters can be used for selecting a machine learning model.

Images