Method for controlling ai / ML model information in wireless communication system, and apparatus therefor

Abstract

The present disclosure relates to a technology for controlling AI / ML model information in a wireless communication system, and a method by which a terminal manages AI / ML model information and an apparatus may be provided, the method comprising the steps of: receiving, from a base station, configuration information for training an AI / ML model and associated identification information associated with the configuration information; training the AI / ML model on the basis of the configuration information and the associated identification information; reporting, to the base station, AI / ML model identification information corresponding to the associated identification information; and receiving, from the base station, indication information for indicating the AI / ML model.

Description

Method and apparatus for controlling AI / ML model information in a wireless communication system

[0001] The present disclosure relates to a method and apparatus for controlling AI / ML model information in a wireless communication system.

[0002] The International Telecommunication Union (ITU) is conducting standardization of next-generation 5th generation (5G) and 6th generation (6G) mobile communications through the International Mobile Telecommunication (IMT)-2020 and IMT-2030 programs.

[0003] The next-generation 6th generation (6G) mobile communication system is defined as an evolutionary system that further enhances the performance of 5th generation (5G) technology, and at the same time, as a next-generation communication platform that encompasses new services and industrial applications. Based on the ultra-high transmission rates, ultra-low latency, and massive terminal connectivity achieved by 5G, 6G aims for ultra-broadband, ultra-precision, ultra-intelligence, and ultra-convergence.

[0004] To achieve this, various technologies are being discussed in 6G, including the utilization of FR3 and terahertz (THz) bands, large-scale antennas and new numerology designs, AI / ML native network optimization, global coverage based on non-terrestrial networks (NTN), dynamic channel control by intelligent reflective surfaces (RIS), and integrated sensing and communication (ISAC). In addition, technologies to enhance network energy efficiency are also being treated as an important research axis in 6G. Therefore, to achieve the goals of ultra-high speed, ultra-low latency, intelligence, global coverage, and convergence services that the next-generation 6th generation (6G) mobile communication system aims for, technical challenges related to AI-Native RAN supporting AI / ML-based network optimization, non-terrestrial networks (NTN) ensuring global connectivity, ISAC converging communication, location, and sensing, intelligent reflective surfaces (RIS) enabling dynamic channel control, and securing energy efficiency for sustainable network operation need to be resolved.

[0005] In particular, when using AI / ML models to predict channel estimation results or estimate optimal beams, sharing information regarding the use of AI / ML models between the terminal and the base station is necessary. To achieve this, mutual awareness and management of AI / ML model information between the terminal and the base station are required for the training, inference, activation, and management of the models.

[0006] Therefore, there is an urgent need for the proposal and research of various technologies capable of meeting the technical requirements for this purpose.

[0007] The present embodiments may provide a method and apparatus for managing AI / ML model information in a wireless communication system.

[0008] The present embodiments may provide a method and apparatus for assigning AI / ML model identification information in a wireless communication system.

[0009] In order to solve the aforementioned problem, in one aspect, the present disclosure may provide a method for a terminal to manage AI / ML model information, comprising the steps of: receiving configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station; performing training of the AI / ML model based on the configuration information and linked identification information; reporting AI / ML model identification information corresponding to the linked identification information to the base station; and receiving instruction information for instructing the AI / ML model from the base station.

[0010] In another aspect, the present disclosure may provide a method for a base station to control the management of AI / ML model information of a terminal, comprising the steps of: transmitting configuration information for training an AI / ML model and linked identification information linked to the configuration information to the terminal; receiving the trained AI / ML model identification information at the terminal; and transmitting instruction information for instructing the AI / ML model to the terminal.

[0011] In another aspect, the present disclosure may provide a terminal device for managing AI / ML model information, comprising a receiving unit that receives configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station, a control unit that performs training of the AI / ML model based on the configuration information and linked identification information, and a transmitting unit that reports AI / ML model identification information corresponding to the linked identification information to the base station, wherein the receiving unit further receives instruction information for instructing the AI / ML model from the base station.

[0012] According to the present disclosure, communication can be performed in a wireless communication system by efficiently using an AI / ML model.

[0013] In addition, according to the present disclosure, a technology for assigning AI / ML model identification information in a wireless communication system can be provided.

[0014] The present disclosure provides the effect of enabling evolution into intelligent and converged networks, going beyond the performance goals of ultra-high speed, ultra-low latency, and massive connectivity aimed at by next-generation 6th generation (6G) mobile communication systems. Specifically, network intelligence and efficient resource management are realized through AI / ML-native autonomous optimization, and new service quality can be provided through ultra-broadband communication utilizing FR3 and terahertz (THz) bands. In addition, energy efficiency enhancement technology enables the implementation of sustainable communication infrastructure, and non-terrestrial networks (NTN) can overcome regional limitations by securing global coverage. Furthermore, intelligent reflective surfaces (RIS) can optimize channel quality in dynamic propagation environments, and integrated sensing and communication (ISAC) can create a new application ecosystem by integrating communication, location, and sensing functions.

[0015] FIG. 1 is a drawing illustrating the structure of a wireless communication system to which the present embodiment can be applied.

[0016] FIG. 2 is a diagram illustrating the logical layer structure between a terminal (UE) and a base station (gNB) in a wireless communication system to which the present embodiment can be applied.

[0017] FIG. 3 is a diagram illustrating an exemplary NG-RAN structure to which the present embodiment can be applied.

[0018] FIG. 4 is a diagram illustrating the configuration of a terminal to which the present embodiment can be applied.

[0019] FIG. 5 is a diagram illustrating an AI / ML operation workflow according to one embodiment of the present disclosure.

[0020] FIG. 6 is a diagram illustrating the data collection and reporting procedure of an AI / ML model to which embodiments of the present disclosure can be applied.

[0021] FIG. 7 is a diagram illustrating the classification according to the form in which an AI / ML model to which embodiments of the present disclosure can be applied is configured.

[0022] FIG. 8 is a diagram illustrating a terminal-network collaboration level using an AI / ML model to which embodiments of the present disclosure can be applied.

[0023] FIG. 9 is a diagram illustrating terminal operation according to one embodiment.

[0024] FIG. 10 is a diagram illustrating the operation of a base station according to one embodiment.

[0025] FIG. 11 is a diagram illustrating the operation of transmitting AI / ML model identification information through DCI according to one embodiment.

[0026] FIG. 12 is a diagram illustrating the operation of transmitting AI / ML model identification information through an SIB according to one embodiment.

[0027] FIG. 13 is a diagram illustrating the operation of transmitting AI / ML model identification information through an MIB according to one embodiment.

[0028] FIG. 14 is a diagram illustrating the operation of transmitting AI / ML model identification information through an RRC message according to one embodiment.

[0029] FIG. 15 is a drawing for explaining a terminal configuration according to one embodiment.

[0030] FIG. 16 is a diagram illustrating a base station configuration according to one embodiment.

[0031] Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the embodiments, if it is determined that a detailed description of related known components or functions may obscure the essence of the technical concept, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless otherwise specified.

[0032] Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms.

[0033] In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another.

[0034] In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used.

[0035] Meanwhile, where numerical values ​​or corresponding information regarding a component (e.g., levels, etc.) are mentioned, even without separate explicit notation, the numerical values ​​or corresponding information may be interpreted as including a range of error that may occur due to various factors (e.g., process factors, internal or external shocks, noise, etc.).

[0036] A wireless communication system in this specification refers to a system for providing various communication services, such as voice and data packets, using wireless resources, and may include a terminal, a base station, or a core network.

[0037] Meanwhile, the term "terminal" in this specification is a comprehensive concept meaning a device including a wireless communication module that communicates with a base station in a wireless communication system, and should be interpreted as a concept that includes not only User Equipment (UE) in WCDMA, LTE, NR, HSPA and IMT-2020 / IMT-2030 (5G, New Radio, and 6G), and wired / wireless communication sensing systems that interact with specific communication systems, but also Mobile Station (MS), User Terminal (UT), Subscriber Station (SS), wireless device, and sensing module / device in GSM. In this specification, the term "terminal" is a comprehensive concept meaning a device including a wireless communication module that communicates with a base station in a wireless communication system. It should be interpreted as a concept that includes not only User Equipment (UE) in WCDMA, LTE, NR, HSPA, and IMT-2020 / IMT-2030 (5G, New Radio, and 6G), and wired / wireless communication sensing systems that interact with specific communication systems, but also Mobile Station (MS), User Terminal (UT), Subscriber Station (SS), wireless device, sensing module / device, etc. in GSM. Furthermore, depending on the usage type, the terminal may be a user portable device such as a smartphone, and in a V2X communication system, it may refer to a device including the vehicle itself or a wireless communication module within the vehicle. Furthermore, in the case of a Machine Type Communication (MTC) system, the concept can be interpreted to include not only MTC terminals, M2M terminals, URLLC terminals, and V2X terminals equipped with communication modules to perform machine type communication, but also sensor network terminals, IoT devices, wearable devices, etc.As such, a “terminal” may refer to another node that transmits and receives data / signals with a specific communication node.

[0038] In this specification, the term "base station" is a comprehensive concept referring to a network device that performs wireless communication with a terminal in a wireless communication system. It should be interpreted to include eNBs (evolved Node Bs) in 3GPP LTE systems, gNBs (gNode Bs) in NR (New Radio) systems, and next-generation base stations in IMT-2020 / IMT-2030 (5G and 6G) ​​systems, as well as Node Bs in WCDMA systems, BTSs (Base Transceiver Stations) in GSM systems, APs (Access Points) in Wi-Fi systems, and RSUs (Road Side Units) in V2X communication systems. Furthermore, base stations can be implemented in various forms, including not only a centralized architecture but also distributed Small Cell units, Repeaters, Relay Nodes, Satellite Gateways, and Intelligent Reflective Surface (RIS) controllers, and all such variations should be interpreted as being included in the term "base station" in this specification. As such, "base station" may refer to another communication node that transmits and receives data / signals with another node in a specific communication system. In this specification, the term "Cell" should be interpreted as a comprehensive concept referring to a logical / physical area in a wireless communication system where a terminal receives wireless services from a base station. In LTE systems, the coverage provided by the base station (eNB) may include not only physical cells, but also multiple cell concepts (e.g., SCell, PCell, BWP-based virtual cell, etc.) based on frequency resources, beams, slot structures, etc., in 5G NR systems. Furthermore, as beam-based communication structures are emphasized along with FR2 / FR3 and terahertz (THz) band expansion, beam-based coverage may be included as the basic unit of a cell.Furthermore, in a Non-Terrestrial Network (NTN) environment, as the beam footprint of an orbiting satellite changes over time, there is a characteristic where the boundaries of cells are not fixed but change dynamically, and this can be defined and included as a cell. Additionally, it may include cell expansion based on Intelligent Reflective Surfaces (RIS), dynamic cell structures considering Integrated Sensing and Communication (ISAC), and cells based on virtualized network slices. This may include a structure in which a terminal connects to multiple cells simultaneously or selectively according to physical / logical resource units. Therefore, the present specification should be interpreted as a concept that includes not only base station-based coverage but also beam-unit logical cells, NTN-based dynamic cells, and new types of cells formed based on RIS or virtualization.

[0039] The present disclosure describes an NR system as an example, but it will be obvious that the present disclosure can be applied to and used in the relevant functions of a 6G mobile communication network, and that the limitations of the present disclosure are not restricted by the said NR system. This specifies that while NR (New Radio) is used as a specific embodiment, the technical concept of the present disclosure can be modified or expanded and applied to future systems such as 6G and the newly defined Beyond_G, and is characterized by not being limited to a specific system (NR). Accordingly, the technical scope of the invention is not limited to an NR system and is characterized by being implementable in various wireless systems including 6G.

[0040] The 6G communication to which the present disclosure applies is a next-generation wireless communication technology following 5G, and standardization is currently underway with the goal of commercialization around 2030. Beyond simple speed improvements, 6G aims for artificial intelligence, hyper-spatial scalability, and reality-digital convergence, and is being prepared as a core infrastructure for the future digital society. This 6G communication aims to support the next-generation digital society by organically combining core technologies such as THz communication capable of ultra-broadband transmission, AI-Native networks integrating AI / ML across all network layers, RIS (Reconfigurable Intelligent Surface) technology that actively reflects / manipulates radio waves, NTN (Non-Terrestrial Network) providing 3D coverage such as satellites, Joint Communication & Sensing (JCAS) integrating communication and sensing functions, ultra-precise location estimation, and Quantum Key Distribution (QKD) for security.

[0041] The structure of these 6G communication networks is being discussed based on the following directions and characteristics. First, as an End-to-End Intelligent Autonomous Network (AI-Native Architecture) structure, AI / ML is fundamentally embedded across all network layers to perform data-driven policy optimization, autonomous network operation, anomaly detection, and Quality of Service (QoS / QoE) assurance; it is expected that AI functions will operate in a distributed manner across the RAN, Core, and service management. Second, by adopting a 3D scalability architecture, it supports a Non-Terrestrial Network (NTN) that integrates the terrestrial network (RAN), satellite, and High Altitude Platform (HAPS); to this end, an integrated cell structure and IAB-based backhaul / fronthaul design are planned to be supported. Third, the Service-Based Architecture (SBA) introduced in 5G will be advanced to strengthen the modularization of Network Functions (NF), and flexible API-based service creation will be enabled through NEF, NWDAF, PCF, etc. Fourth, through a distributed intelligence structure, we plan to move away from a centralized architecture and perform real-time AI-based decisions at edge nodes such as RAN, MEC, and UE, thereby supporting local decision-making in smart transportation, factories, and medical sites. Fifth, by introducing a structure that integrates communication and sensing (Joint Communication & Sensing), we will enable the RAN to perform environmental sensing functions beyond simple communication capabilities; to this end, structures for processing sensing features and asynchronous data collection will also be designed. Finally, through a digital twin-based virtualization structure, we will replicate the physical network state in real time and enable predictive autonomous operation and network simulation, thereby supporting the intelligence and efficiency of the entire network.The next-generation wireless system to which this disclosure applies includes support for both frequency bands below 6 GHz (FR1, Frequency Range 1) and frequency bands above 6 GHz (FR2, Frequency Range 2), and specifically defines support for FR3 (Frequency Range 3) between FR1 and FR2, and proposes various support technologies related to end-to-end interoperability and data connectivity for the FR3 band (7.125-24.25 GHz).

[0042] For example, NR has defined various operation scenarios by adding considerations for satellites, automobiles, and new verticals, and in terms of service, it supports eMBB (Enhanced Mobile Broadband) scenarios, mMTC (Massive Machine Communication) scenarios that require low data rates and asynchronous connections while having high terminal density and being deployed over a wide range, and URLLC (Ultra Reliability and Low Latency) scenarios that require high responsiveness and reliability and can support high-speed mobility.

[0043] To satisfy these scenarios, NR introduces a wireless communication system equipped with new waveform and frame structure technologies, low latency technologies, mmWave support technologies, and forward compatibility technologies. In particular, NR systems present various technical changes in terms of flexibility to provide forward compatibility.

[0044] Although the following description focuses on NR, it can also be applied to next-generation wireless communication systems (e.g., 6G, etc.) as described above. Additionally, while the aforementioned terms may be used as different terms in next-generation wireless communication systems, the device providing the same function may be the device described in this specification.

[0045] FIG. 1 is a drawing illustrating the structure of a wireless communication system to which the present embodiment can be applied.

[0046] Referring to FIG. 1, a wireless communication system may be composed of a core network unit (120), a base station (110), and terminals (100, 101). In addition, the wireless communication system may include additional nodes such as a Radio Unit (RU). From the perspective of terminal expansion, the wireless communication system may further include vehicles (102, 103), aerial mobile bodies (104), etc. From the perspective of base station expansion (110), the wireless communication system may further include satellites (111), etc.

[0047] The base station (110) can provide user plane and control plane protocol termination to the terminal (100). For example, the base station (110) may be named gNB (next generation-Node B) and / or eNB (evolved-Node B), and may be referred to by other terms in next-generation wireless communication systems. For example, the terminal (100) may be fixed or mobile and may be referred to by other terms such as MS (Mobile Station), UT (User Terminal), SS (Subscriber Station), MT (Mobile Terminal), or Wireless Device. For example, the base station (110) may be a fixed station communicating with the terminal (100) and may be referred to by other terms such as BTS (Base Transceiver System) or Access Point.

[0048] Base stations (110) can be connected to each other via protocol interfaces. For example, base stations (110) can be connected to each other via Xn interfaces. Additionally, base stations (110) can be connected to core network entities (120) via protocol interfaces. For example, base stations (110) can be connected to core network entities (120) via NG interfaces. More specifically, base stations (110) can be connected to an access and mobility management function (AMF) (120) via an NG-C interface and to a user plane function (UPF) (120) via an NG-U interface. The AMF is responsible for control planes such as terminal access and mobility control functions, and the UPF is responsible for control functions for user data.

[0049] The satellite (111) can provide communication services to a terminal through a ground antenna or gateway as a core network entity. The satellite (111) may provide communication services while orbiting the Earth or may provide communication services while fixed in orbit on the Earth. The satellite (111) may be configured to transparently transmit messages provided by a ground gateway or base station (110). Alternatively, the satellite (111) may be equipped with some or all of the functions of a base station (110) to perform the function of a base station for the core network entity (120) and the terminal.

[0050] Meanwhile, the terminal (100) can perform direct communication with other terminals (101) without passing through the base station (110) through terminal-to-terminal communication technology. For example, direct communication between terminals can be described as D2D or V2X technology, but is not limited to such terms. For example, vehicles (102, 103) can use terminal-to-terminal communication technology as vehicle-to-vehicle communication technology, and communication can also be performed between the vehicle (103) and the terminal (100). Of course, communication can also be performed between the vehicle (102) and the base station (110). Similarly, a communication link can be established between the aerial mobile body (104) and the base station (110) to perform communication, and communication can also be performed between the aerial mobile body (104) and the vehicle (102, 103) or the terminal (100, 101).

[0051] In NR, CP-OFDM waveforms using a cyclic prefix are used for downlink transmission, while CP-OFDM or DFT-s-OFDM are used for uplink transmission. OFDM technology is easy to combine with MIMO (Multiple Input Multiple Output) and has the advantage of enabling the use of low-complexity receivers along with high frequency efficiency. For next-generation wireless communication systems, various waveforms and modulation technologies may be applied.

[0052] FIG. 2 is a diagram illustrating the logical layer structure between a terminal (UE) and a base station (gNB) in a wireless communication system to which the present embodiment can be applied.

[0053] Referring to FIG. 2, data transmission between a terminal and a base station in a wireless communication system is performed through a multilayer protocol structure. Generally, NR and NG-RAN structures are configured to include at least one of a Service Data Adaptation Protocol (SDAP), a Packet Data Collection Protocol (PDCP), Radio Link Control (RLC), Media Access Control (MAC), and a Physical Layer (PHY).

[0054] In the User Plane, the SDAP layer maps QoS flows defined in the 5G core (5GC) to the Data Radio Bearer (DRB), and the PDCP layer performs encryption, integrity protection, header compression, packet reordering, and deduplication functions. The RLC layer provides splitting and combining, retransmission, and reordering functions, while the MAC layer handles radio resource scheduling, multiplexing, and HARQ. The PHY layer performs modulation, encoding, channel estimation, and beamforming.

[0055] In the control plane, the RRC layer is responsible for wireless bearer setup, measurement and handover procedures, and security parameter exchange, and works in conjunction with the upper layer, NAS, to perform terminal authentication and session management.

[0056] The functions and operations of each layer described above are exemplary, and the functions of a specific layer may be separated and performed by other layers. For instance, the splitting and combining functions of the RLC layer may be performed by PDCP, and the retransmission and reordering functions may be absorbed into similar functions in the MAC or PHY layers, thereby eliminating the RLC layer.

[0057] The embodiments in this specification may be applied to the aforementioned logical layer structure and may also be applied to various logical layer structures that are modified in the future.

[0058] FIG. 3 is a diagram illustrating an exemplary NG-RAN structure to which the present embodiment can be applied.

[0059] Referring to FIG. 3, in the NG-RAN structure, base station (110) functions can be separated into a central unit (CU) and a distributed unit (DU). The CU is responsible for the control plane (RRC / PDCP-C) and the user plane (SDAP / PDCP-U), while the DU is responsible for the RLC / MAC / PHY layer. The interface between the CU and the DU is defined as F1-C and F1-U, and within the CU, an E1 interface is defined between the control plane and the user plane. This allows for support of both distributed and centralized architectures.

[0060] Additionally, the connection between base stations is performed by the CU through the Xn interface, and the base station (110) is connected to the core network entity (120) through the NG interface. Meanwhile, in a 6th generation (6G) wireless communication system, the above logic layer structure can be extended for AI / ML-based network optimization, FR3 and THz band expansion, non-terrestrial network (NTN) support, RIS (Reconfigurable Intelligent Surfaces)-based dynamic channel control, and ISAC (Integrated Sensing and Communication)-based integrated service provision. For example, the RRC and MAC layers may include new control signals for low-power mode and energy efficiency, and the SDAP / PDCP layer may simultaneously perform QoS processing and sensing quality assurance. Additionally, in an NTN environment, RLC / MAC procedures may be added to assist time / frequency synchronization, and new RRC information elements for RIS control may be defined.

[0061] As explained, the communication system to which the present disclosure applies may evolve in a form in which each layer interacts and expands as follows, while inheriting the 5G-Advanced infrastructure for interoperability with 6G or next-generation communication systems. Alternatively, the communication system to which the present disclosure applies may include a separate RAN and Core network structure distinct from the conventional wireless communication structure. That is, each embodiment described in the present disclosure applies to a communication system and may also be applied to various past or future communication systems, such as 4G, 5G, and 6G.

[0062] For example, AI-Native RAN linkage operations can be applied. AI-Native RAN can be connected to RRC, CU / DU orchestration, and the core's PCF. Models and policies are deployed in the DU's real-time (MAC / PHY) control, the CU-CP's near-real-time (RRM / handover / measurement policy), and the core / PCF's minute-to-hour slice / QoS governance, respectively. Higher-level policies are propagated from the PCF through the AMF / SMF path, and model-derived parameters can be linked to terminals and base stations through RRC IE extensions (measurement / handover / DRX / cell reselection, etc.). In addition, metrics such as PDCP retransmission / discard delays, RLC buffer occupancy / retransmissions, MAC / PHY beam quality based on CQI / CSI / BLER / HARQ / PHR / SRS / CSI-RS, and RRC / NGAP / XnAP measurement events / HO failures / delays / jitter can be collected in a schema that clearly defines UE / Cell / Beam scopes and sampling cycles, enabling inference and policy updates. Furthermore, AI learning and inference in the RAN can be expanded in various ways.

[0063] As another example, the frequency band used in a communication system can be extended from FR1 and FR2 to FR3. FR3 can be configured in a very high frequency band at THz. In this case, the complexity of beam / resource configuration at THz can explode. For instance, beam tracking in MAC / PHY and policies in RRC / SDAP must be closely coupled. Specifically, in the ultra-high frequency environment of THz, link stability can change rapidly due to path loss, shielding, and Doppler. To mitigate this, AI can use terminal behavior / environment features to preempt predictive beams and reduce sweeping cycles / SSB sets to alleviate overhead. Resource decisions, such as slot / mini-slot, DMRS / CSI-RS density, and auxiliary carrier usage, can be synchronized with SDAP QoS mapping to allow latency-sensitive traffic and bandwidth-intensive traffic to adopt different beam / carrier strategies.

[0064] As another example, since NTN and terrestrial network duality entails orbit / visibility / Doppler / path delay variations, measurement / handover procedures on the RRC side and timing / synchronization on the DU side must be complemented. For instance, it may be required to predict and correct timing advances and frequency offsets based on satellite orbit / ephemeris / terminal movement models, and to utilize visibility predictions between the terrestrial network and NTN or between NTNs to perform preparatory handovers.

[0065] As another example, from an energy efficiency perspective, green RAN improves energy efficiency by maintaining traffic and SLAs while implementing DU sleep / wake, MIMO chain off, and DRX optimization. For instance, hierarchical coverage strategies can be employed, such as turning off small cells / auxiliary carriers and maintaining wide-area cell anchors during nighttime or low-load periods. To achieve this, RRC low-power profiles by service class, power state transition commands / telemetry (KPM) from CU to DU, and OAM collection of energy KPIs (cell power, RF chain uptime, cooling alarms) can be defined. To address increased handover failures or delays caused by excessive power saving, relevant policies and procedures regarding minimum resources and maximum wake-up delays per SLA can also be established.

[0066] As another example, ISAC integration utilizes PHY signals such as PRS / RS for sensing and treats sensing quality as part of QoS. By including “Sensing-Accuracy / Latency / Update-Rate” auxiliary metrics in SDAP / 5QI to incorporate sensing into the objective function of policy optimization, RRC can adapt PRS / CSI-RS patterns / densities / power according to service characteristics (e.g., industrial location / sensing).

[0067] FIG. 4 is a diagram illustrating the configuration of a terminal to which the present embodiment can be applied.

[0068] Referring to FIG. 4, the terminal (100) may be composed of various elements, components, units, and / or modules. For example, the terminal (100) may include a communication unit (300), a processor (310), and a memory (320). The communication unit (300) may include a communication circuit and transceiver(s). For example, the communication circuit may include circuit configurations such as a signal oscillator for transmitting or receiving a signal through an antenna. For example, the transceiver(s) may be composed of one or more and may include one or more antennas. The processor (310) is electrically connected to the communication unit (300) and the memory (320) and controls the overall operation of the terminal (100). For example, the processor (310) may control the electrical / mechanical operation of the terminal based on a program / code / command / information stored in the memory (320). Additionally, the processor (310) can transmit information stored in memory (320) to an external (e.g., another communication device or base station) via a wireless / wired interface through the communication unit (300), or store information received from an external (e.g., another communication device or base station) via a wireless / wired interface through the communication unit (300) in memory (320).

[0069] Additionally, the processor (310) can perform AI / ML-related data processing by loading program code into memory (320). Alternatively, the processor (310) may perform data analysis and processing operations using signals and / or data received through the communication unit (300).

[0070] In addition to this, the terminal (100) may include various additional elements. For example, the additional elements may be configured in various ways depending on the type of terminal. For example, the additional elements may include at least one of a power unit / battery, an input / output unit (I / O unit), a driving unit, and a computing unit. Although not limited thereto, the terminal (100) may be implemented in the form of a drone, robot, vehicle, XR device, portable device, home appliance, IoT device, digital broadcasting terminal, hologram device, public safety device, MTC device, medical device, fintech device (or financial device), security device, climate / environment device, AI server / device, sensor device, network node, etc. Depending on the use—e.g., service—the terminal (100) may be movable or used in a fixed location.

[0071] Meanwhile, various elements, components, units / parts, and / or modules within the terminal (100) may be entirely interconnected via a wired interface, or at least some of them may be wirelessly connected via a communication unit (300). For example, within the terminal (100), the processor (310) and the communication unit (300) may be wired, and the processor (310) and the memory (320) and / or additional elements may be wirelessly connected via the communication unit (300). Additionally, each element, component, unit / part, and / or module within the terminal (100) may include one or more additional elements. For example, the processor (310) may be composed of a set of a communication control processor, an application processor, an Electronic Control Unit (ECU), a graphics processing processor, a memory control processor, etc. As another example, memory (320) may be composed of RAM (Random Access Memory), DRAM (Dynamic RAM), ROM (Read Only Memory), flash memory, volatile memory, non-volatile memory and / or a combination thereof.

[0072] Wireless communication systems are in the process of adopting artificial intelligence technology to provide more effective and faster communication services. For example, advancements in AI / ML (Artificial intelligence / machine learning) technology are leading to the intelligentization and sophistication of node(s) and terminal(s) constituting wireless communication networks.

[0073] In particular, due to the intelligence of networks and base stations, it is expected that various network / base station decision parameter values ​​can be rapidly optimized, derived, and applied according to various environmental parameters. Environmental parameters may include at least one of the distribution / location of base stations, the distribution / location / material of buildings / furniture, the location / movement direction / speed of terminals, and climate information. However, the parameters described above are merely examples, and the environmental parameters may include other environmental parameters related to the network / base station decision parameters in addition to the parameters listed. Network / base station decision parameter values ​​may include at least one of the transmit / receive power of each base station, transmit power of each terminal, precoder / beam of the base station / terminal, time / frequency resource allocation for each terminal, and duplex method of each base station. However, the parameters described above are merely examples, and the network / base station decision parameter values ​​may include other parameters determined by the network / base station in addition to the parameters listed.

[0074] In a narrow sense, AI / ML can easily be referred to as artificial intelligence based on deep learning, but conceptually, it can be distinguished as follows.

[0075] - Artificial Intelligence: This refers to all automation where machines can take over tasks that humans would otherwise have to perform.

[0076] - Machine Learning: Machines learn patterns for decision-making from data on their own, without explicitly programming rules.

[0077] - Deep Learning: A model based on artificial neural networks that enables machines to perform everything from feature extraction to judgment from unstructured data in a single step. The algorithm relies on multi-layered networks composed of interconnected nodes for feature extraction and transformation, inspired by biological nervous systems, or neural networks. Common deep learning network architectures include Deep Neural Networks (DNN), Recurrent Neural Networks (RNN), and Convolutional Neural Networks (CNN).

[0078] As mentioned above, artificial intelligence (AI) is the broadest concept of AI / ML, and deep learning is the narrowest concept of AI / ML. Machine learning (ML) can be interpreted as a concept that is narrower than artificial intelligence but broader than deep learning.

[0079] In this specification, AI / ML or artificial intelligence is a comprehensive concept that includes the aforementioned deep learning and does not limit specific models or specific learning / inference methods.

[0080] FIG. 5 is a diagram illustrating an AI / ML operation workflow according to one embodiment of the present disclosure.

[0081] Referring to Fig. 5, the AI / ML operations workflow highlights the key stages of the ML model lifecycle. The ML model lifecycle includes Training (510), Testing (520), Emulation (530), Deployment (540), and Inference (550).

[0082] ML model training (Training, 510): Includes the initial training and retraining of an ML model (or family of models). It also includes validation to evaluate the performance of the model when performed on training and validation data. If the validation results fall short of expectations (e.g., variance is outside the acceptable range), the model must be retrained.

[0083] ML Model Testing (Testing, 520): Test the validated ML model to evaluate the performance of the ML model trained on test data. If the test results meet expectations, you can proceed to the next step. If the test results do not meet expectations, the model must be retrained.

[0084] AI / ML Inference Emulation: Before applying to a target network or system, the model is run in an emulation environment to evaluate inference performance. If the emulation results fall short of expectations (e.g., failure to meet target inference performance, negative impact on the performance of existing features), the model must be retrained. AI / ML inference emulation is optional and can be omitted from the ML model lifecycle.

[0085] ML Model Deployment: Includes a model loading procedure that makes the trained model available for use in the target AI / ML inference function. Deployment may not be necessary in cases such as when the training function and the inference function are co-located in the same location.

[0086] AI / ML Inference: The AI / ML inference function performs inference using a trained model. It can also trigger retraining or model updates based on performance monitoring and evaluation.

[0087] Depending on the system implementation method and the configuration of AI / ML functions, both the AI / ML inference emulation and ML model deployment steps may be omitted.

[0088] Meanwhile, models can degrade over time or due to changes in the environment. Therefore, a retraining process may be performed based on feedback regarding the artificial intelligence model's performance.

[0089] In this way, when an artificial intelligence model is applied in a communication system, training, storage, inference, and management operations are performed based on a framework, enabling stable prediction and management of the model.

[0090] FIG. 6 is a diagram illustrating the data collection and reporting procedure of an AI / ML model to which embodiments of the present disclosure can be applied.

[0091] Referring to FIG. 6, 610 and 620 can each be nodes in a communication system. For example, 610 can be a base station and 620 can be a terminal. Conversely, 610 can be a terminal and 620 can be a base station. In addition, 610 can be a first base station and 620 can be a second base station. That is, a data request procedure between base stations can be performed. Similarly, 610 can be a first terminal and 620 can be a second terminal.

[0092] The exchange of Data Collection Request, Data Collection Failure, and Data Collection Update messages between the first node (610) and the second node (620) is described as an example. The first node (610) can send a message requesting data collection to the second node (620) for training, inference, management, and updating of an artificial intelligence model (S601). The message requesting data collection may vary depending on the subject of each node. For example, the message requesting data collection may be transmitted through the Xn interface, NG interface, or Uu interface. The message requesting data collection may be an upper-layer message or a lower-layer message. S601 is a procedure in which the first node (610) requests specific operational data, performance indicators, or measurement data from the second node (620). At this time, the data may be used for integration with the Self-Organizing Network (SON), Network Data Analytics Function (NWDAF), or OAM. Through this, it is possible to link with OAM / MDT (Minimization of Drive Test) data collection or E2E performance analysis.

[0093] The second node (620) determines whether the requested data can be collected. For example, if the 620 does not possess the requested data or if collection is difficult, the second node (620) sends a data collection failure message to the first node (610) (S602). For example, the data collection failure (S502) procedure may return a failure response when the requested data does not exist or when the node cannot provide it due to a lack of authority or resources. This corresponds to an error handling procedure and can be transmitted as a failure cause IE (Information Element) specified in the standardization. For example, it may include unsupported parameter, resource not available, etc. As described above, the data collection failure message may also be configured with various interfaces and message formats depending on the subject of each node. The data collection failure message may include information on the cause of the data collection failure. Alternatively, information reported through updates may include predicted resource status information, terminal performance feedback information, measured terminal path, energy cost, etc.

[0094] If the second node (620) succeeds in collecting the requested data, it may send a data collection update message to the first node (610) (S603). For example, the data collection update (S503) procedure may subsequently send an update message to the first node (610) when the situation improves or when some data becomes available. This includes sending via asynchronous reporting or periodic / conditional update methods. Examples may include RAN performance counters, QoE measurements, QoS Flow utilization, location / traffic patterns, etc. The data collection update message may include collected data information regarding the requested data. In the case of previously transmitted data, it may be included in the data collection update message by indicating an update.

[0095] Through this process, data collection and transmission between each node can be performed.

[0096] As explained above, the artificial intelligence model can be configured in the terminal and / or base station.

[0097] FIG. 7 is a diagram illustrating a classification according to the configuration of an AI / ML model to which embodiments of the present disclosure can be applied. FIG. 7 performs wireless resource management and beam management optimization using an artificial intelligence / machine learning (AI / ML) model.

[0098] First, (a) in the case of a network-sided AI / ML model, an AI / ML model located at a base station (710) performs resource optimization operations such as scheduling, beam selection, and interference management based on channel state information, measurements, performance indicators, etc. collected from a terminal (700).

[0099] Next, (b) in the case of a terminal-side AI / ML model (UE-sided AI / ML Model), the AI / ML model located at the terminal (700) performs beam prediction, performance estimation, and movement pattern analysis on its own and reports the results to the base station (710), thereby improving the accuracy of network control.

[0100] Finally, (c) in the case of a two-sided AI / ML model, the base station (710) and the terminal (700) each perform AI / ML inference, and the utilization of network resources is maximized by mutually exchanging inference results, performance indicators, and confidence level information.

[0101] Accordingly, the present embodiments perform wireless resource optimization operations in a network-only, terminal-only, or network-terminal collaborative structure depending on the deployment location of the AI / ML model, thereby providing the effect of enabling the implementation of ultra-high-speed, ultra-low-latency, and intelligent services for next-generation wireless communication systems.

[0102] Referring again to FIG. 6, the artificial intelligence model may be configured in only one of the terminal (700) or the base station (710). This can be described as a one-side AI / ML model. Alternatively, the artificial intelligence model may be configured in both the terminal (700) and the base station (710). This can be described as a two-side AI / ML model.

[0103] In the form of a one-side AI / ML model, an artificial intelligence model is configured at a base station (710), and a terminal (700) transmits information necessary for model input to the base station (710) via a wireless interface. The base station (710) can generate inference results by providing the received information as input to the model. Through this, the base station (710) can directly utilize the model to perform beam management, terminal mobility management, etc. This is called a NW-side AI / ML model.

[0104] In another form of the one-side AI / ML model, an artificial intelligence model is configured in the terminal (700), and the base station (710) transmits information necessary for model input to the terminal (700) through a wireless interface. For example, a reference signal can be transmitted through the base station (710) beam, or a reference signal for channel measurement can be transmitted. In this case, the base station (710) can transmit the signal that was previously transmitted. The terminal (700) can generate an inference result by providing the received information as input to the model. The terminal (700) transmits the generated inference result to the base station (710). Through this, the terminal (700) performs inference using the model and transmits the inference result to the base station (710) to assist with beam management of the base station (710), mobility management of the terminal, etc. This is called a UE-side AI / ML model.

[0105] In the case of a two-sided AI / ML model, a model may be configured at the terminal (700) and the base station (710), respectively. In this case, the terminal (700) and the base station (710) can input mutually received data into the model to derive an inference result. The terminal (700) can transmit the inference result to the base station (710). The base station (710) can control the operation of the terminal using the inference result received from the terminal (700) and the inference result of the base station (710). The models configured at the terminal (700) and the base station (710) may each be configured with different purposes or functions.

[0106] As such, AI / ML capabilities vary depending on specific use cases and sub-use cases, and may focus on the interaction between the network and the terminal.

[0107] FIG. 8 is a diagram illustrating a terminal-network collaboration level using an AI / ML model to which embodiments of the present disclosure can be applied.

[0108] Referring to FIG. 8, the collaboration levels between the base station (810) and the terminal (800) can be defined as follows.

[0109] 1. Level X (or Collaboration Level 0) is characterized by a method in which the terminal (800) and the base station (810) each independently operate AI / ML models, and the network-side (NW-side) AI model and the terminal-side (UE-side) AI model perform only self-optimization without standardized signaling. For example, the base station (810) uses its own beam management model, and the terminal (800) uses its own power optimization model. That is, the base station (810) and the terminal (800) each independently operate their own AI / ML models, and both sides operate through AI / ML logic implemented within their own or the device without standardized signaling or model sharing. For example, beam management, scheduling, and power control of the base station can be performed using their own AI models, and operations such as power consumption optimization based on the terminal's movement pattern, antenna selection, and local beam prediction can be performed independently. In this Level X, unpredictable operations may occur between them due to the opacity of the AI ​​operations.

[0110] 2. Level Y (Collaboration Level 1) allows the terminal (UE, 800) and the base station (810) to perform the step of exchanging signals with each other to assist in the inference and operation of an AI model. Specifically, the terminal (800) reports the inference results or performance indicators (e.g., Top-K beam prediction accuracy, confidence level, etc.) of the AI / ML model it has performed itself to the base station (810). After receiving the reported information, the base station (810) performs network optimization decisions, such as resource management, beam management, and scheduling, based on this information. In particular, as part of the AI / ML model performance monitoring and auxiliary signaling procedures, signaling definitions such as RRC information elements (RRC IE) or MAC control elements (MAC CE) are required. Therefore, compared to the case where the terminal (800) and the base station (810) operate the AI ​​model independently according to the present disclosure, this provides a technical advantage of being able to efficiently manage signaling overhead while improving mutual collaboration performance.

[0111] 3. Level Z (Collaboration Level 2) allows the base station (810) and the terminal (UE, 800) to directly exchange the AI ​​model itself or learned parameters. Specifically, the base station (810) can download the AI ​​model learned on the network side to the terminal (800) to enhance the inference function at the terminal, while conversely, the terminal (800) can upload parameters learned independently or model update information to the base station (810) to enhance the central management function of the network. To this end, procedures for distributing and updating the AI ​​model can be performed, and include defining dedicated signaling procedures for model transmission, security / integrity guarantees, and transmission efficiency optimization techniques. For example, RRC protocol extensions, new MAC CE definitions, or interoperability structures with NWDAF / AF can be considered. Accordingly, according to the present disclosure, the base station (810) and the terminal (800) raise the level of cooperation to the highest level (full collaboration), providing the advantage that the network and the terminal can perform resource management and service optimization based on the same AI model. This can be applied adaptively at the same time, taking into account issues such as increased signaling overhead and standardization complexity.

[0112] More specifically, Collaboration Level 0 involves the terminal operating through its own AI / ML model or performing only RRC / RRM procedures without AI functions, and includes operations where the base station does not separately configure or intervene with the AI ​​model for the terminal. In other words, it is characterized by the absence of AI / ML-related interaction between the base station and the terminal. Meanwhile, Collaboration Level 1 is characterized by the fact that while the model itself is not transmitted, configuration / signaling information to support AI / ML functions is exchanged between the base station and the terminal. For example, the terminal receives configuration information (e.g., reference signal configuration, evaluation conditions, reporting cycle, etc.) for using the AI / ML model from the base station. Considering this, the terminal reports only the results of AI inference or measurement results, while the model itself exists independently within the terminal. That is, the terminal can report the results of local AI model execution used for CSI enhancement, beam management, mobility enhancement, etc., to the base station. In addition, Collaboration Level 2 enables the base station to transmit the AI / ML model itself or some parameter / configuration information to the terminal, and the terminal to perform AI functions based on the model (e.g., inference module, weight, etc.) received from the base station, or to feed back the model results learned by the terminal to the base station, or to support bidirectional model updates. This can be referred to as collaborative inference and federated learning-like operations. To this end, the terminal and the base station can transmit and receive model-related signaling, such as the configuration, parameters, and training cycle of the AI ​​model. The technology described herein may be applied to at least one of the three collaboration levels mentioned above.

[0113] Meanwhile, when a model is configured on a terminal, various methods may be applied to deliver the model to the terminal.

[0114] For example, the model can be trained on a terminal, a base station, or a separate device. The trained model can be stored on a server via model transfer / delivery. The model stored on the server can be delivered to the terminal as needed and configured on the terminal.

[0115] As another example, the model may be trained on a terminal or a separate device. The trained model is transmitted to a base station through a model delivery procedure and stored at the base station. The base station can then transmit the model to the terminal as needed for configuration on the terminal.

[0116] As another example, a model can be trained at a base station. The trained model is stored at the base station. The base station storing the model may be the same as or different from the base station that performed the training. The base station storing the model can then transmit the model to a terminal as needed to configure the model on the terminal.

[0117] Meanwhile, AI / ML models can be applied to various usage scenarios.

[0118] For example, it can be used for the purpose of saving network energy. For instance, to balance performance and energy savings, it is necessary to optimize cell activation or deactivation decisions based on predicted traffic loads by utilizing AI / ML technology. To this end, internal base station information such as current and predicted resource status, UE path prediction, currently predicted UE traffic, and predicted resource status of adjacent RAN nodes can be used as inputs to the model. Additionally, terminal-related information such as UE location data, UE measurement reports, and cell-level and beam-level UE measurement reports can be used as inputs. Furthermore, neighbor network information such as currently predicted energy efficiency, currently predicted resource status, and current energy status can be used as inputs. The model's output can infer energy saving strategies (recommended cell activation / deactivation), handover strategies, recommended candidate cells to take over traffic, predicted energy efficiency, and predicted energy status information. Feedback considers the resource status of adjacent RAN nodes, energy efficiency, UE performance affected by energy saving measures, and system KPIs.

[0119] As another example, AI / ML technology may be used for load balancing purposes. For instance, this functionality can be applied to enhance the quality of user experience and improve system capacity by improving load balancing performance through AI / ML model-based solutions and predicted loads, based on various metrics, feedback collected from UEs and network nodes, and historical data. To this end, the model's inputs may include base station information such as UE mobility path predictions, currently predicted energy efficiency, and currently predicted resource status, as well as terminal information such as UE location data, UE movement history data, and cell-level and beam-level UE measurement reports. Additionally, neighbor network information may include terminal performance measurement data from adjacent cells to which traffic has been offloaded. The model's outputs may infer target cell selection for load balancing, predicted local resource status information, a resource status model of predicted adjacent RAN nodes, and predicted UE information selected for handover to the target RAN node. Feedback information may include UE performance data of the target NG RAN, updates to the target NG RAN's resource status information, and system KPIs.

[0120] As another example, AI / ML technology may be used from a mobility perspective. For instance, based on AI / ML, functions such as (1) reducing the probability of unintended mobility events (early / late handover, incorrect cell handover), (2) predicting UE location mobility performance, and (3) optimizing traffic steering can be performed. To this end, the input to the model may include base station information such as UE mobility path prediction, currently predicted energy efficiency, and currently predicted resource status, as well as terminal measurement information such as UE location information, UE movement history information, and wireless measurements of adjacent cells to the UE service cell. Additionally, neighbor base station information may include UE history information, the location of the handovered UE, QoS parameter performance information, and information on past UE handovers that were successful or failed in the currently predicted resource status. The output of the model may include UE path prediction, conditional handover-related prediction, UE traffic prediction, and the validity period of the model output.

[0121] In addition, AI / ML capabilities can be used to analyze metrics related to network and UE performance to perform optimal resource management and mobility decisions for network slicing, thereby ensuring the Quality of Service (QoS) of each slice and improving network efficiency. Alternatively, AI / ML capabilities can provide network performance improvements and maintain consistency in the UE experience by proactively detecting and responding to Coverage and Capacity Optimization issues.

[0122] In addition to these scenarios and usability, the operations described in this specification can be applied to various past, present, and future scenarios.

[0123] Below, we will explain several representative scenarios and their applications.

[0124] AI / ML models can be used to enhance CSI feedback. Based on channel state feedback information, CSI feedback enables base stations to accurately identify the user's wireless channel status, allowing for optimal wireless resource allocation and transmission. However, a disadvantage exists in that the feedback load required for accurate channel state information feedback increases significantly as the number of antennas and subbands increases. To address this, technology for delivering accurate channel state information at an appropriate feedback load is required.

[0125] CSI feedback is a critical procedure for adaptive communication technologies such as adaptive modulation, coding, and beamforming. Base stations utilize CSI feedback received from terminals to determine detailed downlink transmission methods, thereby improving overall network efficiency and system performance. However, periodically transmitting CSI feedback can cause system overhead and waste wireless resources, and pose a problem where it becomes difficult to manage as the number of users increases.

[0126] By utilizing AI / ML models to further enhance CSI feedback, it may be possible to provide benefits such as reduced overhead and improved accuracy.

[0127] As mentioned above, in order for the terminal and the base station to perform channel estimation, optimal beam prediction, and mobility operation determination using AI / ML models, management of model information between the terminal and the base station may be required.

[0128] For example, in order to perform retraining on an AI / ML model, train an AI / ML model using a specific dataset, or enable or disable a specific AI / ML model, it must be possible to specify a specific AI / ML model. In addition, information regarding the specified AI / ML model must be able to be recognized and shared identically between the terminal and the base station.

[0129] To this end, it is necessary to manage identification information for AI / ML models.

[0130] For example, model identification of an AI / ML model is initiated by the terminal, and the base station (network) can support the remaining steps of model identification. A model ID may be assigned during the model identification step.

[0131] As another example, model identification of an AI / ML model is initiated by the base station, and the terminal can respond to the remaining steps of model identification. Similarly, a model ID can be assigned during the model identification step.

[0132] Meanwhile, as mentioned above, the AI / ML model can be configured as a two-side model configured on both the UE-side and NW-side or on both the terminal and network sides as needed.

[0133] For example, in the case where AI / ML models are configured on both the terminal and network sides (two-side model), a data set for training can be transmitted from the network side to the terminal side. The terminal can perform training on the terminal part of the two-side model based on the received data set. The terminal can report information about the trained terminal part based on the data set to the base station.

[0134]

[0135] Trained AI / ML models can be used for inference operations.

[0136] For UE-side AL / ML model inference in BM-Case2, it supports reporting the inference results N (N>=1, FFS on N) of future time instances as a single report. In this case, the inference result information for a single time instance is equivalent to a single report for BM-Case 1.

[0137] Therefore, when a base station uses a UE-side model in a beam management model or CSI-related inference process, it must instruct the UE to proceed with inference through the UE-side model, and may send the instruction through at least one of DCI, SIB, RRC, and MIB.

[0138] In addition, it may be necessary to consider cases where the number of terminal-side models may be limited depending on UE capability.

[0139] For example, when a base station selects one of multiple terminal models (N>1), information for the selection can be transmitted by including it in the aforementioned instructions.

[0140] As another example, in the case where the base station only issues an instruction to use the UE-side model and the terminal directly selects one of the multiple models (N=1), only information regarding usage may be included in the aforementioned instruction.

[0141] The following describes various embodiments of operations for identifying and controlling model information in relation to lifecycle management operations. Additionally, various embodiments of operations for transmitting model information between a terminal and a base station are also described. Each embodiment may be applied by combining any combination, or each may be applied and operated individually.

[0142] The AI / ML model identification information described below may be replaced by various terms such as model information, model ID, or AI / ML ID. Additionally, associated identification information refers to identification information used by a network to identify data-related conditions or configuration information for training an AI / ML model within a cell. For example, associated identification information refers to information used to distinguish and identify conditions or configurations necessary for model training, such as training dataset information, training parameters, training iterations, and model structure. Associated identification information can be applied to network-side models, terminal-side models, two-side models, etc.

[0143] FIG. 9 is a diagram illustrating terminal operation according to one embodiment.

[0144] Referring to FIG. 9, a method for a terminal to manage AI / ML model information may include the step of receiving configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station (S900).

[0145] For example, a terminal may receive information necessary for training an AI / ML model from a base station. For instance, the configuration information may include at least one of model parameters for training the AI / ML model, training data collection information, and model entity information. In addition, the configuration information may include condition information set by the base station for training the AI / ML model. For instance, the condition information may be set in various ways, such as the training conditions of the AI / ML model, the number of training iterations, target accuracy information, and the model structure (CNN, neural network, etc.).

[0146] In addition, configuration information can be mapped to linked identification information for differentiation. That is, linked identification information can be set for each configuration information to enable distinct identification of each configuration information.

[0147] Meanwhile, the terminal may receive additional identification information of the AI / ML model to be trained from the base station. For example, in the case of a two-side model, AI / ML cooperative training between the base station and the terminal is required, and they may need to have identical training and characteristics. To this end, the base station may transmit identification information of the AI / ML model to be trained, configuration information, and / or linkage identification information for the training of the terminal part of the two-side model. The terminal can perform training of the designated AI / ML model using configuration information (condition information) identified by the linkage identification information. In this case, the base station can control the training operation of the terminal to ensure that LCM operations, such as training or retraining for the target AI / ML model, are performed.

[0148] The AI / ML model identification information directed by the base station to the terminal may be assigned as a globally unique identifier. Alternatively, the AI / ML model identification information may consist of a combination of identifiers for identifying the model structure and identifiers for distinguishing functions. For example, model identification information may be identified by distinguishing a model performing a classification function of a CNN structure using each identifier.

[0149] Alternatively, AI / ML model identification information may be assigned by the terminal in correspondence with linked identification information.

[0150] A method for a terminal to manage AI / ML model information may include a step of performing training of an AI / ML model based on configuration information and linked identification information (S910).

[0151] For example, the terminal can train an AI / ML model using configuration information. Alternatively, the terminal can train an identified AI / ML model based on AI / ML model identification information instructed by the base station. For training, the terminal can perform training by reflecting training condition information identified by the configuration information and / or linked identification information.

[0152] For example, the terminal can perform data collection for a procedure or target indicated by configuration information and / or linked identification information, and process the collected data according to the conditions of the configuration information to proceed with model training.

[0153] A method for a terminal to manage AI / ML model information may include a step of reporting AI / ML model identification information corresponding to linked identification information to a base station (S920).

[0154] When training is completed, the terminal may report AI / ML model identification information to the base station. In this case, the terminal may also report associated identification information to indicate under which configuration information or conditions the AI / ML model was trained. Alternatively, the terminal may transmit model identification information of the trained AI / ML model to the base station in correspondence with the associated identification information.

[0155] For example, AI / ML model identification information can be assigned by a terminal using global unique identification information.

[0156] As another example, AI / ML model identification information can be assigned by the terminal in correspondence with linked identification information.

[0157] As another example, AI / ML model identification information may be assigned based on at least one of model structure classification, model function classification, training sequence information, and associated identification information. For example, AI / ML model identification information may be assigned as a combination of at least two identifiers among an identifier based on model function classification, an identifier based on model structure classification, and an identifier based on training sequence information. For example, an identifier based on a pre-set function classification according to the function performed by the model may be assigned as the MSB, and a pre-set identifier based on the model structure classification may be assigned as the LSB, thereby assigning AI / ML model identification information through a combination of MBS+LSB. Similarly, a combination of an identifier based on model function classification and an identifier based on model training sequence information at the terminal may be used. Alternatively, a combination of an identifier based on model structure classification and an identifier based on model training sequence may be used. Or, model identification information may be assigned as a combination of three identifiers.

[0158] Meanwhile, AI / ML model identification information may be assigned as the linked identification information itself. That is, the AI / ML model identification information may be reused from the linked identification information associated with the configuration information received from the base station. In this case, since the base station can know in advance that the relevant AI / ML model identification information is assigned as linked identification information, it may receive only whether the training has ended from the terminal.

[0159] A method for a terminal to manage AI / ML model information may include the step of receiving instruction information for instructing an AI / ML model from a base station (S930).

[0160] Once the AI / ML model is fully trained, the terminal can store the model and use it for inference. For example, the terminal can perform inference operations using the AI / ML model instructed by the base station. Alternatively, the terminal may perform the instructed operations by checking the activation status or retraining status of a specific AI / ML model according to the instructions of the base station.

[0161] To this end, the terminal may receive instruction information from the base station. Accordingly, the instruction information may include AI / ML model identification information and / or information regarding instruction actions for instructing an action.

[0162] A terminal can train and store one or more AI / ML models on the terminal. When only one AI / ML model is stored on the terminal, the base station transmits instruction information to the terminal, such as inference instructions, activation instructions, and retraining instructions for the AI / ML model, and the terminal can receive this and perform the relevant operations.

[0163] In contrast, when two or more AI / ML models are stored in the terminal, the base station must clearly specify which model to instruct the operation for. To this end, the terminal may receive instruction information including N bits of AI / ML model identification information and an instruction field that instructs the operation.

[0164] For example, instruction information can be received via Downlink Control Information (DCI) containing AI / ML model identification information. As another example, instruction information may be received via a Master Information Block (MIB) or System Information Block (SIB) containing AI / ML model identification information. As yet another example, instruction information may be received via upper-layer signaling containing AI / ML model identification information. The upper-layer signaling may be a channel state information configuration message. That is, to estimate channel state information using an AI / ML model, the base station transmits channel state information configuration information to the terminal. The channel state information configuration information includes AI / ML model identification information, which can instruct the terminal on which model to use to perform inference in channel state information estimation.

[0165] Similarly, when a terminal-side model or a two-side model is configured in relation to LCM operation, the base station may instruct the terminal to retrain a specific AI / ML model. In this case, the base station may instruct the terminal to identify the AI / ML model subject to retraining and transmit linked identification information to instruct the terminal on which configuration information the AI / ML model should perform retraining. If the AI / ML model configured in the terminal has been trained using linked identification information other than that instructed by the base station, the terminal may check whether the linked identification information instructed by the base station differs from the linked identification information corresponding to the instructed AI / ML model, and if they differ, may retrain the AI / ML model under conditions based on the linked identification information instructed.

[0166] Through the above operations, the terminal and the base station can assign identifiers to AI / ML models and control the operation of specific AI / ML models according to the instructions of the base station. Below, these operations are explained once again from the perspective of the base station.

[0167] FIG. 10 is a diagram illustrating the operation of a base station according to one embodiment.

[0168] Referring to FIG. 10, a method for a base station to control the management of AI / ML model information of a terminal may include the step of transmitting configuration information for training the AI / ML model and linked identification information linked to the configuration information to the terminal (S1000).

[0169] For example, a base station can transmit information necessary for training an AI / ML model to a terminal. For instance, the configuration information may include at least one of model parameters for training the AI / ML model, training data collection information, and model entity information. In addition, the configuration information may include condition information set by the base station for training the AI / ML model. For example, the condition information may be set in various ways, such as the training conditions of the AI / ML model, the number of training iterations, target accuracy information, and the model structure (CNN, neural network, etc.).

[0170] In addition, configuration information can be mapped to linked identification information for differentiation. That is, linked identification information can be set for each configuration information to enable distinct identification of each configuration information.

[0171] Meanwhile, the base station may further transmit identification information of the AI / ML model to be trained to the terminal. For example, in the case of a two-side model, AI / ML cooperative training between the base station and the terminal is required, and they may need to have identical training and characteristics. To this end, the base station may transmit identification information of the AI / ML model to be trained, configuration information, and / or linkage identification information for the training of the terminal part of the two-side model. The terminal can perform training of the designated AI / ML model using configuration information (condition information) identified by the linkage identification information. In this case, the base station can control the training operation of the terminal to ensure that LCM operations, such as training or retraining for the target AI / ML model, are performed.

[0172] The AI / ML model identification information instructed by the base station to the terminal may be assigned as a globally unique identifier. Alternatively, the AI / ML model identification information may consist of a combination of identifiers for identifying the model structure and identifiers for distinguishing functions. For example, model identification information may be identified by distinguishing a model performing a classification function of a CNN structure using each identifier. Alternatively, the AI / ML model identification information may be assigned by the terminal in correspondence with linked identification information.

[0173] The terminal can perform training of an AI / ML model based on received configuration information and linked identification information.

[0174] A method for a base station to control the management of AI / ML model information of a terminal may include the step of receiving AI / ML model identification information trained at the terminal (S1010).

[0175] When training is completed, the terminal may report AI / ML model identification information to the base station. In this case, the terminal may also report associated identification information to indicate under which configuration information or conditions the AI / ML model was trained. Alternatively, the terminal may transmit model identification information of the trained AI / ML model to the base station in correspondence with the associated identification information.

[0176] For example, AI / ML model identification information can be assigned by a terminal using global unique identification information.

[0177] As another example, AI / ML model identification information can be assigned by the terminal in correspondence with linked identification information.

[0178] As another example, AI / ML model identification information may be assigned based on at least one of model structure classification, model function classification, training sequence information, and associated identification information. For example, AI / ML model identification information may be assigned as a combination of at least two identifiers among an identifier based on model function classification, an identifier based on model structure classification, and an identifier based on training sequence information. For example, an identifier based on a pre-set function classification according to the function performed by the model may be assigned as the MSB, and a pre-set identifier based on the model structure classification may be assigned as the LSB, thereby assigning AI / ML model identification information through a combination of MBS+LSB. Similarly, a combination of an identifier based on model function classification and an identifier based on model training sequence information at the terminal may be used. Alternatively, a combination of an identifier based on model structure classification and an identifier based on model training sequence may be used. Or, model identification information may be assigned as a combination of three identifiers.

[0179] Meanwhile, AI / ML model identification information may be assigned as the linked identification information itself. That is, the AI / ML model identification information may be reused from the linked identification information associated with the configuration information received from the base station. In this case, since the base station can know in advance that the relevant AI / ML model identification information is assigned as linked identification information, it may receive only whether the training has ended from the terminal.

[0180] A method for a base station to control the management of AI / ML model information of a terminal may include the step of transmitting instruction information to instruct the terminal to an AI / ML model (S1020).

[0181] Once the AI / ML model is fully trained, the terminal can store the model and use it for inference. For example, the terminal can perform inference operations using the AI / ML model instructed by the base station. Alternatively, the terminal may perform the instructed operations by checking the activation status or retraining status of a specific AI / ML model according to the instructions of the base station.

[0182] To this end, the base station may transmit instruction information to the terminal. Accordingly, the instruction information may include AI / ML model identification information and / or information regarding instruction actions for instructing an action.

[0183] A terminal can train and store one or more AI / ML models on the terminal. When only one AI / ML model is stored on the terminal, the base station transmits instruction information to the terminal, such as inference instructions, activation instructions, and retraining instructions for the AI / ML model, and the terminal can receive this and perform the relevant operations.

[0184] In contrast, when two or more AI / ML models are stored in the terminal, the base station must clearly specify which model to operate on. To this end, the base station may transmit instruction information including N bits of AI / ML model identification information and an instruction field that instructs the operation.

[0185] For example, instruction information can be transmitted via Downlink Control Information (DCI) containing AI / ML model identification information. As another example, instruction information may be transmitted via a Master Information Block (MIB) or System Information Block (SIB) containing AI / ML model identification information. As yet another example, instruction information may be transmitted via upper-layer signaling containing AI / ML model identification information. The upper-layer signaling may be a channel state information configuration message. That is, to estimate channel state information using an AI / ML model, the base station transmits channel state information configuration information to the terminal. The channel state information configuration information includes AI / ML model identification information, which can instruct the terminal on which model to use to perform inference in channel state information estimation.

[0186] Similarly, when a terminal-side model or a two-side model is configured in relation to LCM operation, the base station may instruct the terminal to retrain a specific AI / ML model. In this case, the base station may instruct the terminal to identify the AI / ML model subject to retraining and transmit linked identification information to instruct the terminal on which configuration information the AI / ML model should perform retraining. If the AI / ML model configured in the terminal has been trained using linked identification information other than that instructed by the base station, the terminal may check whether the linked identification information instructed by the base station differs from the linked identification information corresponding to the instructed AI / ML model, and if they differ, may retrain the AI / ML model under conditions based on the linked identification information instructed.

[0187] Through the above operations, the terminal and the base station can assign an identifier to the AI / ML model and control the operation of a specific AI / ML model according to the instructions of the base station.

[0188] Below, various embodiments that can be performed by the aforementioned terminal and base station are distinguished and described. Each embodiment may be combined with others in any combination to form separate embodiments.

[0189] As described above, one or more AI / ML models may be configured on the terminal. Each AI / ML model can be used through various selection processes, such as model transfer / delivery, model activation / deactivation, and model update.

[0190] To achieve this, model identification information must be defined. Additionally, AI / ML models can be classified into various types.

[0191] For example, standardized model identification information may be used. The terminal reports the model identification information of the trained AI / ML model to the base station. Based on the received model identification information, the base station performs a model selection operation to select the model used for inference / activation / deactivation / retraining and can transmit it to the terminal based on the model identification information. To this end, model identification information allocation rules or global unique model identification information, etc., must be shared between the terminal and the base station.

[0192] As another example, model identification information based on terminal capability can be used. The terminal assigns model identification information for the trained model and reports the model identification information, associated identification information related to training conditions, and inference results using the model to the base station. Based on the received information, the base station proceeds with model selection to choose the model used for inference, activation, deactivation, or retraining, and can transmit this to the terminal based on the model identification information.

[0193] As another example, consider a case where model identification information is not used. The terminal reports the trained model configuration, associated identification information, and inference results to the base station. The base station performs model selection based on the received information, selects the model used for inference / activation / deactivation / retraining, and directs it to the terminal. In this case, the associated identification information, information about the model configuration, and acks for information received from the terminal may be used for model identification.

[0194] In the operation where the base station instructs the terminal with model identification information, various message types such as DCI, MIB, SIB, and RRC messages may be used, and the instruction may be given in different forms depending on the number of models stored in the terminal. This will be explained in more detail later with reference to the drawings.

[0195] Model identification information according to an example of the present invention serves as a key element for consistently managing the life cycle of AI / ML models, such as creation, distribution, activation, updating, and disposal, and plays a role in securing the uniqueness of each model and ensuring interoperability. For example, a network or terminal must select one of a plurality of AI / ML models to perform a specific function; in this case, the model identification information can clearly distinguish the type, version, provider, storage location, etc. of the model. Such model identification information may include the following information. For example, it may be defined as at least one combination of a model identifier (Model ID), model type, model function, model version, model provider information, model storage location, model status, and model reference information, and may be defined in the form of a predetermined index. Furthermore, the above AI / ML Model ID is a Global Identifier used to uniquely identify the AI / ML model itself used by the network or terminal, and can be defined in a representative form by considering a combination of the data parameter set, architecture, and function type of the model. Additionally, the model identification information can be updated whenever the structure or parameters of the AI / ML model are changed or whenever a new version is registered, and through this, it can be signaled through the model management procedure between the terminal (UE) and the base station (gNB).

[0196] Such AI / ML model identification information may be assigned by the base station, or assigned by the terminal and reported to the base station. Alternatively, the linked identification information itself may be used as model identification information. Alternatively, the model identification information may be determined according to a pre-set rule. In this case, since the terminal and the base station share the pre-set rule in advance, the reporting procedure for the assignment of separate model identification information may be omitted or simplified.

[0197] As mentioned above, model identification information (model ID) can be used in various fields, such as indicating activation status, designating and reporting models for inference, and management including model retraining. This is described and explained as part of Life Cyclic Management (LCM).

[0198] In the case of a NW-sided model, since LCM is performed in the NW, there is no need to send a separate model ID to the terminal. For UE-sided models and UE-part of two-sided models, the cases for performing model identification are divided into four types (cases where model identification is performed in the UE or NW, cases where identification is initiated in the UE and terminated in the NW, and cases where identification is initiated in the NW and terminated in the UE), and a method and procedure for assigning model IDs are presented.

[0199] According to one embodiment, model identification can be performed at the terminal without over-the-air signaling.

[0200] For example, the terminal can report to the network along with the model ID after the model training is completed. The network can configure LCM information based on the received model ID. The network can transmit the configured LCM information to the terminal. The terminal can perform LCM according to the received LCM information.

[0201] In this case, model identification information may be assigned by the terminal. For example, the terminal may assign a Model ID defined according to the model's Function. Alternatively, the terminal may assign an ID based on the order in which the model was trained. Or, the terminal may assign an ID by adding additional information (associated ID, model configuration, etc.) to the ID defined according to the function after model training. Alternatively, the terminal may define an ID according to the function after model training and transmit additional information (associated ID, model configuration, etc.) along with it. The above-mentioned model association information identifier (Associated ID) includes an identifier introduced to identify additional information (associated data entity) linked to a specific Model ID. It is characterized by including information to track relationships when an AI / ML model has relationships with training datasets, validation results, inference results, federated models, etc., rather than simply being a set of codes or parameters; therefore, the model association information identifier (Associated ID) may further include additional information identifying other AI / ML resources (artifacts) associated with the above-mentioned Model ID.

[0202] As described above, Model Identification Information for identifying an AI / ML model according to one embodiment of the present invention includes at least one Model ID and one or more Associated IDs linked to the Model ID. The Model ID includes the unique structure, function, version, and provider information of the AI / ML model, enabling the model to be identified as a single entity. Additionally, the Associated ID functions as an auxiliary identifier for identifying datasets, inference results, validation history, or other models associated with the model, and an AI / ML model management server or terminal can track the model's training history and interdependencies using the Associated ID. Furthermore, the Associated ID according to the present invention supports a model management system in consistently managing each model's training history, inference results, validation status, and association relationships by systematically identifying various associated information generated, updated, and stored throughout the entire lifecycle of the AI / ML model.

[0203] According to another embodiment, the operation of proceeding with model identification in the NW without over-the-air signaling (the UE reporting training results to the NW after the model training ends) can be performed.

[0204] For example, the terminal can report the end of training to the NW after the model training is completed. The NW can configure LCM information after assigning a model ID. The NW can transmit the configured LCM information to the UE. The terminal can perform LCM based on the received LCM information.

[0205] In this case, model identification information can be defined according to the AI / ML model ID assignment defined by the model function. Alternatively, the NW can assign an AI / ML model ID defined by the entity. Alternatively, the NW can assign an ID based on the order in which the model was trained. Alternatively, the NW can assign an ID by adding additional information (entity, associated ID, model configuration, etc.) to the ID defined by the function after model training. Alternatively, the NW can define an ID based on the function after model training and transmit additional information (entity, associated ID, model configuration, etc.) along with it.

[0206] According to another embodiment, an operation to proceed with model identification in the NW without over-the-air signaling (where the NW assigns information about model training and model ID information) can be performed.

[0207] Before the terminal performs model training, when the network sends configuration information for model training (e.g., Data collection related configurations, associated ID, entity, etc.) to the terminal, the model ID may also be transmitted.

[0208] For example, the NW can transmit model configuration information (entity, data configuration, etc.) and the model ID to the terminal. The terminal can perform model training. The terminal can report the completion of model training to the NW. The NW can transmit LCM information to the terminal. The terminal can perform LCM.

[0209] The Network Workforce (NW) can assign a Model ID defined according to the Function. The Network Workforce (NW) can assign a Model ID defined according to the Model entity. The Network Workforce (NW) can assign an Associated ID. Here, the Associated ID is information used to identify the conditions (configuration information) set by the NW side in relation to the data for training an AI / ML model within a cell. The Associated ID is assigned as a single identifier within the cell. It may be assigned separately in other cells, or the Associated ID may be assigned based on the same base station. Alternatively, the Network Workforce (NW) can assign IDs according to the order in which model training instructions were issued to the terminal. The Network Workforce (NW) can use the ID defined according to the Function, the Associated ID, model configuration, monitoring results, entity information, etc., together for identification.

[0210] According to another embodiment, an LCM operation procedure through over-the-air signaling (initiation of identification in the UE, proceeding to the remaining steps in the NW) can be performed.

[0211] For example, the terminal can assign an ID to the model itself and transmit it to the NW. The NW can reassign the ID by appending additional model ID information to the ID received from the terminal. The NW can transmit the reassigned ID and LCM information to the terminal. The terminal can perform an LCM operation based on the reassigned ID.

[0212] When assigning a Model ID independently, the terminal may consider at least one of the following information. For example, the terminal may assign an ID by considering at least one of the following information: Function, Model entity information, Data set, associated ID, Inference results, AI / ML model capacity, and AI / ML model training order.

[0213] NW may reassign the remaining model ID by considering at least one of the following information. For example, NW may reassign the model ID by considering at least one of the Function, Model entity information, Performance monitoring results, AI / ML model capacity, and AI / ML model report order.

[0214] According to another embodiment, an LCM operation procedure through over-the-air signaling (initiation of identification in NW, proceeding to the remaining steps in UE) can be performed.

[0215] The NW can transmit information for model training to the UE along with a model ID. The terminal can create and train a model based on the received model training information (e.g., configuration information and / or associated identification information). The terminal can independently reconfigure and assign an ID to the created model. The terminal can report the reassigned model ID to the NW. The NW can transmit LCM information to the UE based on the reassigned model ID. The terminal can perform an LCM operation.

[0216] A method for the terminal to independently reassign a Model ID may consider at least one of the following information. For example, the terminal may assign a model ID by considering at least one of the following information: Function, Model entity information, Data set, associated ID, Inference result, Performance metric, Performance results, AI / ML model capacity, and AI / ML model training order.

[0217] The method for assigning a Model ID in NW may consider at least one of the following information. For example, NW may assign a model ID by considering at least one of the following information: Function, Model entity information, Data set, associated ID, AI / ML model capacity, and AI / ML model report order.

[0218] As described above, it can be primarily used to identify a UE-sided model via a model ID for LCM in the network, and the model ID can also be used when reporting model performance to the network even if the UE performs LCM directly. For example, this applies when transmitting performance monitoring results of a UE-sided model to the network.

[0219] Meanwhile, below, we intend to explain the method by which a base station instructs an AI / ML model to a terminal, categorized according to each instruction information format.

[0220] As mentioned above, the number of UE models may be limited depending on the UE capability. A base station may obtain AI / ML model information configured in a terminal by receiving capability information or model training result information from the terminal. In this case, when the base station transmits instruction information instructing the use of an LCM or a model configured in the terminal, the base station may directly select one model among the terminal's multiple models. Alternatively, the base station may only instruct whether to use a terminal-side model, allowing the terminal to directly select one of the multiple models, or there may be cases where only one model is configured in the terminal.

[0221] If multiple models are configured in the terminal and the base station selects one of them, the instruction information field can be configured to indicate one of K models using N bits as shown in Table 1.

[0222] K (Number of UE-side models) N (bits) 11224384125

[0223] The bit format above is an example, and in the case of model instructions via SIB, MIB, RRC messages, etc., it may be composed of bits different from the above.

[0224] FIG. 11 is a diagram illustrating the operation of transmitting AI / ML model identification information through DCI according to one embodiment.

[0225] Referring to FIG. 11, for example, a base station can instruct a terminal (1100) to use a UE-side model through a DCI format.

[0226] The base station (1110) can transmit a command to use a UE-side model to the terminal (1100) by placing it in an N-bit DCI (S1101). In this case, the base station (1110) may instruct the terminal (1100) to use only the model, rather than specifying which model to use. The terminal (1100) selects a suitable model among the UE-side models and obtains an inference result through the measured beam RSRP and the selected model (S1102). The terminal (1100) transmits the inference result to the base station (S1103). At this time, the terminal (1100) may report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0227] As another example, the base station can instruct the terminal to use the UE-side model via the DCI format.

[0228] The base station (1110) can transmit a command to use a specific UE-side model to the terminal (1100) by placing it in an N-bit DCI (S1101). In this case, the terminal (1100) can select the indicated model based on the model identification information indicated by the base station (1110) and obtain the measured beam RSRP and the inference result through the model (S1102). The terminal (1100) transmits the inference result to the base station (S1103). At this time, the terminal (1100) can report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0229] FIG. 12 is a diagram illustrating the operation of transmitting AI / ML model identification information through an SIB according to one embodiment.

[0230] Referring to FIG. 12, the base station (1110) can transmit a command to use a UE-side model to the terminal (1100) by placing it in an N-bit SIB (S1201). In this case, the base station (1110) may instruct the terminal (1100) to use only the model, rather than specifying which model to use. The terminal (1100) selects a suitable model among the UE-side models and obtains an inference result through the measured beam RSRP and the selected model (S1202). The terminal (1100) transmits the inference result to the base station (S1203). At this time, the terminal (1100) may report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0231] As another example, the base station can instruct the terminal to use the UE-side model via the SIB.

[0232] The base station (1110) can transmit a command to use a specific UE-side model to the terminal (1100) by placing it in an N-bit SIB (S1201). In this case, the terminal (1100) can select the indicated model based on the model identification information indicated by the base station (1110) and obtain the measured beam RSRP and the inference result through the model (S1202). The terminal (1100) transmits the inference result to the base station (S1203). At this time, the terminal (1100) can report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0233] To this end, an additional SIB format may be configured. This SIB format may be used to transmit AI / ML model identification information, as well as to broadcast linked identification information, configuration information, etc.

[0234]

[0235] FIG. 13 is a diagram illustrating the operation of transmitting AI / ML model identification information through an MIB according to one embodiment.

[0236] Referring to FIG. 13, the base station (1110) can transmit a command to use a UE-side model to the terminal (1100) by placing it in an N-bit MIB (S1301). In this case, the base station (1110) may instruct the terminal (1100) to use only the model, rather than specifying which model to use. The terminal (1100) selects a suitable model among the UE-side models and obtains an inference result through the measured beam RSRP and the selected model (S1302). The terminal (1100) transmits the inference result to the base station (S1303). At this time, the terminal (1100) may report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0237] As another example, the base station can instruct the terminal to use the UE-side model via the MIB.

[0238] The base station (1110) can transmit a command to use a specific UE-side model to the terminal (1100) by placing it in an N-bit MIB (S1301). In this case, the terminal (1100) can select the indicated model based on the model identification information indicated by the base station (1110) and obtain the measured beam RSRP and the inference result through the model (S1302). The terminal (1100) transmits the inference result to the base station (S1303). At this time, the terminal (1100) can report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0239]

[0240] FIG. 14 is a diagram illustrating the operation of transmitting AI / ML model identification information through an RRC message according to one embodiment.

[0241] Referring to FIG. 14, the base station (1110) can transmit a command to use a UE-side model to the terminal (1100) in an RRC message (S1401). In this case, the base station (1110) may instruct the terminal (1100) to use only the model, rather than specifying which model to use. The terminal (1100) selects a suitable model among the UE-side models and obtains an inference result through the measured beam RSRP and the selected model (S1402). The terminal (1100) transmits the inference result to the base station (S1403). At this time, the terminal (1100) may report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0242] As another example, the base station can instruct the terminal to the UE-side model via an RRC message.

[0243] The base station (1110) can transmit a command to use a specific UE-side model to the terminal (1100) by including it in an RRC message (S1401). In this case, the terminal (1100) can select the indicated model based on the model identification information indicated by the base station (1110) and obtain the measured beam RSRP and the inference result through the model (S1402). The terminal (1100) transmits the inference result to the base station (S1403). At this time, the terminal (1100) can report the AI / ML model identification information used together with the inference result. Alternatively, the terminal (1100) may report only the inference result.

[0244] Here, the RRC message may be a configuration information message for CSI measurement configuration. Alternatively, the RRC message may be an RRC reconstruction message. If inference from an AI / ML model is used during CSI measurement, the base station may include model identification information of the AI / ML model instructing the CSI measurement configuration information.

[0245]

[0246] The terminal and base station configurations capable of performing the operations of each of the aforementioned embodiments are briefly described once again.

[0247] FIG. 15 is a drawing for explaining a terminal configuration according to one embodiment.

[0248] Referring to FIG. 15, a terminal (1500) managing AI / ML model information includes a receiver (1530) that receives configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station, a control unit (1510) that performs training of the AI / ML model based on the configuration information and linked identification information, and a transmitter (1520) that reports AI / ML model identification information corresponding to the linked identification information to the base station. The receiver (1530) may further receive instruction information for instructing the AI / ML model from the base station.

[0249] The receiver (1530) can receive information necessary for training an AI / ML model from a base station. For example, the configuration information may include at least one of model parameters for training the AI / ML model, training data collection information, and model entity information. In addition, the configuration information may include condition information set by the base station for training the AI / ML model. For example, the condition information may be set in various ways, such as the training conditions of the AI / ML model, the number of trainings, target accuracy information, and the structure of the model (CNN, neural network, etc.).

[0250] In addition, configuration information can be mapped to linked identification information for differentiation. That is, linked identification information can be set for each configuration information to enable distinct identification of each configuration information.

[0251] Meanwhile, the receiving unit (1530) may receive additional identification information of the AI / ML model to be trained from the base station. For example, in the case of a two-side model, AI / ML cooperative training between the base station and the terminal is required, and they may have the same training and characteristics. To this end, the base station may transmit identification information of the AI / ML model to be trained, configuration information, and / or linkage identification information for the training of the terminal part of the two-side model. The control unit (1510) may perform training of the instructed AI / ML model using configuration information (condition information) identified by the linkage identification information. In this case, the base station may control the training operation of the terminal to control LCM operations, such as training and retraining for the target AI / ML model, to be performed.

[0252] The AI / ML model identification information instructed by the base station to the terminal may be assigned as a globally unique identifier. Alternatively, the AI / ML model identification information may consist of a combination of identifiers for identifying the model structure and identifiers for distinguishing functions. For example, model identification information may be identified by distinguishing a model performing a classification function of a CNN structure using each identifier. Alternatively, the AI / ML model identification information may be assigned by the terminal in correspondence with linked identification information.

[0253] Additionally, the control unit (1510) can train an AI / ML model using configuration information. Alternatively, the control unit (1510) can train an identified AI / ML model based on AI / ML model identification information instructed by a base station. For training, the control unit (1510) can perform training by reflecting training condition information identified by configuration information and / or linked identification information.

[0254] For example, the control unit (1510) can perform data collection for a procedure or target indicated by configuration information and / or linked identification information, and process the collected data according to the conditions of the configuration information to proceed with model training.

[0255] When training is completed, the transmitter (1520) may report AI / ML model identification information to the base station. In this case, the transmitter (1520) may also report linked identification information to indicate which configuration information or conditions the AI / ML model was trained according to. Alternatively, the transmitter (1520) may transmit model identification information of the trained AI / ML model to the base station in correspondence with the linked identification information.

[0256] For example, AI / ML model identification information can be assigned by a terminal using global unique identification information.

[0257] As another example, AI / ML model identification information can be assigned by the terminal in correspondence with linked identification information.

[0258] As another example, AI / ML model identification information may be assigned based on at least one of model structure classification, model function classification, training sequence information, and associated identification information. For example, AI / ML model identification information may be assigned as a combination of at least two identifiers among an identifier based on model function classification, an identifier based on model structure classification, and an identifier based on training sequence information. For example, an identifier based on a pre-set function classification according to the function performed by the model may be assigned as the MSB, and a pre-set identifier based on the model structure classification may be assigned as the LSB, thereby assigning AI / ML model identification information through a combination of MBS+LSB. Similarly, a combination of an identifier based on model function classification and an identifier based on model training sequence information at the terminal may be used. Alternatively, a combination of an identifier based on model structure classification and an identifier based on model training sequence may be used. Or, model identification information may be assigned as a combination of three identifiers.

[0259] Meanwhile, AI / ML model identification information may be assigned as the linked identification information itself. That is, the AI / ML model identification information may be reused from the linked identification information associated with the configuration information received from the base station. In this case, since the base station can know in advance that the relevant AI / ML model identification information is assigned as linked identification information, it may receive only whether the training has ended from the terminal.

[0260] When the training of the AI / ML model is completed, the control unit (1510) can store the AI / ML model and use it for inference. For example, the control unit (1510) can perform an inference operation using the AI / ML model instructed by the base station. Alternatively, the control unit (1510) may perform an instructed operation by checking whether a specific AI / ML model is activated, whether it is retrained, etc., according to the instructions of the base station.

[0261] To this end, the receiver (1530) can receive instruction information from the base station. Accordingly, the instruction information may include AI / ML model identification information and / or information about instruction actions for instructing actions.

[0262] The terminal can train and store one or more AI / ML models in the terminal. When only one AI / ML model is stored in the terminal, the base station transmits instruction information to the terminal, such as an inference instruction, an activation instruction, or a retraining instruction for the AI / ML model, and the control unit (1510) can receive this and perform the relevant operation.

[0263] In contrast, when two or more AI / ML models are stored in the terminal, the base station must clearly specify which model to instruct the operation for. To this end, the terminal may receive instruction information including N bits of AI / ML model identification information and an instruction field that instructs the operation.

[0264] For example, instruction information can be received via Downlink Control Information (DCI) containing AI / ML model identification information. As another example, instruction information may be received via a Master Information Block (MIB) or System Information Block (SIB) containing AI / ML model identification information. As yet another example, instruction information may be received via upper-layer signaling containing AI / ML model identification information. The upper-layer signaling may be a channel state information configuration message. That is, to estimate channel state information using an AI / ML model, the base station transmits channel state information configuration information to the terminal. The channel state information configuration information includes AI / ML model identification information, which can instruct the terminal on which model to use to perform inference in channel state information estimation.

[0265] Similarly, when a terminal-side model or a two-side model is configured in relation to LCM operation, the base station may instruct the terminal to retrain a specific AI / ML model. In this case, the base station may instruct the terminal to identify the AI / ML model subject to retraining and transmit linked identification information to instruct the terminal on which configuration information the AI / ML model should perform retraining. If the AI / ML model configured in the terminal has been trained using linked identification information other than that instructed by the base station, the terminal may check whether the linked identification information instructed by the base station differs from the linked identification information corresponding to the instructed AI / ML model, and if they differ, may retrain the AI / ML model under conditions based on the linked identification information instructed.

[0266] In addition to this, the control unit (1510) controls the overall operation of the terminal (1500) according to the management operation of AI / ML model information necessary to perform the aforementioned invention.

[0267] The transmitting unit (1520) and the receiving unit (1530) are used to transmit and receive signals, messages, and data necessary to perform the aforementioned invention with a base station.

[0268] FIG. 16 is a diagram illustrating a base station configuration according to one embodiment.

[0269] Referring to FIG. 16, a base station (1600) that controls the management of AI / ML model information of a terminal includes a transmitter (1620) that transmits configuration information for training an AI / ML model and linked identification information linked to the configuration information to the terminal, and a receiver (1630) that receives the trained AI / ML model identification information from the terminal. The transmitter (1620) may further transmit instruction information for instructing the AI / ML model to the terminal.

[0270] For example, the transmitting unit (1620) can transmit information necessary for training an AI / ML model to a terminal. For example, the configuration information may include at least one of model parameters for training the AI / ML model, training data collection information, and model entity information. In addition, the configuration information may include condition information set by a base station for training the AI / ML model. For example, the condition information may be set in various ways, such as the training conditions of the AI / ML model, the number of trainings, target accuracy information, and the structure of the model (CNN, neural network, etc.).

[0271] In addition, configuration information can be mapped to linked identification information for differentiation. That is, linked identification information can be set for each configuration information to enable distinct identification of each configuration information.

[0272] Meanwhile, the transmitter (1620) may further transmit identification information of the AI / ML model to be trained to the terminal. For example, in the case of a two-side model, AI / ML cooperative training between the base station and the terminal is required, and they may have the same training and characteristics. To this end, the transmitter (1620) may transmit identification information of the AI / ML model to be trained, configuration information, and / or linkage identification information for the training of the terminal part of the two-side model. The terminal may perform training of the designated AI / ML model using configuration information (condition information) identified by the linkage identification information. In this case, the control unit (1610) may control the training operation of the terminal to control LCM operations, such as training and retraining for the target AI / ML model, to be performed.

[0273] The AI / ML model identification information instructed by the base station to the terminal may be assigned as a globally unique identifier. Alternatively, the AI / ML model identification information may consist of a combination of identifiers for identifying the model structure and identifiers for distinguishing functions. For example, model identification information may be identified by distinguishing a model performing a classification function of a CNN structure using each identifier. Alternatively, the AI / ML model identification information may be assigned by the terminal in correspondence with linked identification information.

[0274] The terminal can perform training of an AI / ML model based on received configuration information and linked identification information.

[0275] When training is completed, the terminal may report AI / ML model identification information to the base station. In this case, the terminal may also report associated identification information to indicate under which configuration information or conditions the AI / ML model was trained. Alternatively, the terminal may transmit model identification information of the trained AI / ML model to the base station in correspondence with the associated identification information.

[0276] For example, AI / ML model identification information can be assigned by a terminal using global unique identification information.

[0277] As another example, AI / ML model identification information can be assigned by the terminal in correspondence with linked identification information.

[0278] As another example, AI / ML model identification information may be assigned based on at least one of model structure classification, model function classification, training sequence information, and associated identification information. For example, AI / ML model identification information may be assigned as a combination of at least two identifiers among an identifier based on model function classification, an identifier based on model structure classification, and an identifier based on training sequence information. For example, an identifier based on a pre-set function classification according to the function performed by the model may be assigned as the MSB, and a pre-set identifier based on the model structure classification may be assigned as the LSB, thereby assigning AI / ML model identification information through a combination of MBS+LSB. Similarly, a combination of an identifier based on model function classification and an identifier based on model training sequence information at the terminal may be used. Alternatively, a combination of an identifier based on model structure classification and an identifier based on model training sequence may be used. Or, model identification information may be assigned as a combination of three identifiers.

[0279] Meanwhile, AI / ML model identification information may be assigned as the linked identification information itself. That is, the AI / ML model identification information may be reused from the linked identification information associated with the configuration information received from the base station. In this case, since the base station can know in advance that the relevant AI / ML model identification information is assigned as linked identification information, it may receive only whether the training has ended from the terminal.

[0280] Once the AI / ML model is fully trained, the terminal can store the model and use it for inference. For example, the terminal can perform inference operations using the AI / ML model instructed by the base station. Alternatively, the terminal may perform the instructed operations by checking the activation status or retraining status of a specific AI / ML model according to the instructions of the base station.

[0281] To this end, the transmitting unit (1620) can transmit instruction information to a terminal. Accordingly, the instruction information may include AI / ML model identification information and / or information about instruction actions for instructing actions.

[0282] A terminal can train and store one or more AI / ML models on the terminal. When only one AI / ML model is stored on the terminal, the base station transmits instruction information to the terminal, such as inference instructions, activation instructions, and retraining instructions for the AI / ML model, and the terminal can receive this and perform the relevant operations.

[0283] In contrast, when two or more AI / ML models are stored in the terminal, the base station must clearly specify which model to operate on. To this end, the base station may transmit instruction information including N bits of AI / ML model identification information and an instruction field that instructs the operation.

[0284] For example, instruction information can be transmitted via Downlink Control Information (DCI) containing AI / ML model identification information. As another example, instruction information may be transmitted via a Master Information Block (MIB) or System Information Block (SIB) containing AI / ML model identification information. As yet another example, instruction information may be transmitted via upper-layer signaling containing AI / ML model identification information. The upper-layer signaling may be a channel state information configuration message. That is, to estimate channel state information using an AI / ML model, the base station transmits channel state information configuration information to the terminal. The channel state information configuration information includes AI / ML model identification information, which can instruct the terminal on which model to use to perform inference in channel state information estimation.

[0285] Similarly, when a terminal-side model or a two-side model is configured in relation to LCM operation, the base station may instruct the terminal to retrain a specific AI / ML model. In this case, the base station may instruct the terminal to identify the AI / ML model subject to retraining and transmit linked identification information to instruct the terminal on which configuration information the AI / ML model should perform retraining. If the AI / ML model configured in the terminal has been trained using linked identification information other than that instructed by the base station, the terminal may check whether the linked identification information instructed by the base station differs from the linked identification information corresponding to the instructed AI / ML model, and if they differ, may retrain the AI / ML model under conditions based on the linked identification information instructed.

[0286] In addition to this, the control unit (1610) controls the overall operation of the base station (1600) according to the management operation of AI / ML model information necessary to perform the aforementioned invention.

[0287] The transmitting unit (1620) and the receiving unit (1630) are used to transmit and receive signals, messages, and data necessary to perform the aforementioned invention with the terminal.

[0288] Meanwhile, the above-described disclosure may be implemented by a control unit, a transmitter, and a receiver of a terminal and a base station. The control unit, the transmitter, and the receiver may each be composed of a memory and a processor, and may include an antenna as needed.

[0289] The aforementioned embodiments may be supported by standard documents disclosed in at least one of the wireless access systems IEEE 802, 3GPP, and 3GPP2. That is, steps, configurations, and parts in the embodiments that are not described to clearly reveal the technical concept may be supported by the aforementioned standard documents. Furthermore, all terms disclosed in this specification may be explained by the standard documents disclosed above.

[0290] The embodiments described above may be implemented through various means. For example, the embodiments may be implemented by hardware, firmware, software, or a combination thereof.

[0291] In the case of implementation by hardware, the method according to the embodiments may be implemented by one or more ASICs (Application Specific Integrated Circuits), DSPs (Digital Signal Processors), DSPDs (Digital Signal Processing Devices), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), processors, controllers, microcontrollers, or microprocessors.

[0292] In the case of implementation by firmware or software, the method according to the embodiments may be implemented in the form of a device, procedure, or function that performs the functions or operations described above. Software code may be stored in a memory unit and executed by a processor. The memory unit may be located inside or outside the processor and may exchange data with the processor by various means already known.

[0293] Additionally, terms such as "system," "processor," "controller," "component," "module," "interface," "model," or "unit" described above may generally refer to computer-related entities, hardware, combinations of hardware and software, software, or running software. For example, the aforementioned components may be, but are not limited to, processes driven by a processor, processors, controllers, control processors, objects, execution threads, programs, and / or computers. For example, both the application running on the controller or processor and the controller or processor may be components. One or more components may reside within a process and / or execution thread, and the components may be located on a single device (e.g., a system, a computing device, etc.) or distributed across two or more devices.

[0294] The foregoing description is merely an illustrative explanation of the technical concept of the present disclosure, and those skilled in the art to which the present disclosure pertains may make various modifications and variations within the scope of the essential characteristics of the technical concept. Furthermore, since these embodiments are intended to explain, not limit, the scope of the technical concept is not limited by these embodiments. The scope of protection of the present disclosure shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present disclosure.

Claims

1. Regarding the method by which a terminal manages AI / ML model information, A step of receiving configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station; A step of performing training of the AI / ML model based on the above configuration information and the above linked identification information; A step of reporting the AI / ML model identification information corresponding to the above-mentioned linked identification information to the base station; and A method comprising the step of receiving instruction information for instructing the AI / ML model from the base station.

2. In Paragraph 1, The above configuration information is, A method comprising at least one of model parameters for training the above AI / ML model, training data collection information, and model entity information, and distinguished and identified by the above-mentioned linked identification information.

3. In Paragraph 1, The step of receiving the above configuration information and linked identification information is, A method for receiving additional identification information of the above-mentioned AI / ML model to be trained.

4. In Paragraph 3, The above AI / ML model identification information is, A method assigned by the above base station and assigned as a global unique identifier.

5. In Paragraph 1, The above AI / ML model identification information is, A method assigned by the terminal in correspondence with the above-mentioned linked identification information.

6. In Paragraph 5, The above AI / ML model identification information is, A method assigned based on at least one of model structure classification, model function classification, training sequence information and the above-mentioned linked identification information.

7. In Paragraph 6, The above AI / ML model identification information is, A method assigned by combining at least two identifiers among an identifier according to the above model function classification, an identifier according to the above model structure classification, and an identifier according to the above training sequence information.

8. In Paragraph 6, The above AI / ML model identification information is, A method assigned using the above-mentioned linked identification information.

9. In Paragraph 1, The above instruction information is, A method received through Downlink Control Information (DCI) including the above AI / ML model identification information.

10. In Paragraph 1, The above instruction information is, A method of receiving through a Master Information Block (MIB) or System Information Block (SIB) containing the above AI / ML model identification information.

11. In Paragraph 1, The above instruction information is, A method received through upper-layer signaling including the above AI / ML model identification information.

12. In Paragraph 11, The above-mentioned upper-level signaling is, Method for channel status information configuration information messages.

13. A method for a base station to control the management of AI / ML model information of a terminal, A step of transmitting configuration information for training an AI / ML model and linked identification information linked to the configuration information to a terminal; A step of receiving the AI / ML model identification information trained at the terminal; and A method comprising the step of transmitting instruction information to the terminal to instruct the AI / ML model.

14. In Paragraph 13, The above configuration information is, A method comprising at least one of model parameters for training the above AI / ML model, training data collection information, and model entity information, and distinguished and identified by the above-mentioned linked identification information.

15. In Paragraph 13, The above AI / ML model identification information is, A method assigned based on at least one of model structure classification, model function classification, training sequence information and the above-mentioned linked identification information.

16. In Paragraph 13, The above instruction information is, A method for providing information for instructing at least one of an inference instruction, an activation instruction, and a retraining instruction for the above AI / ML model.

17. In Paragraph 13, The above instruction information is, A method of transmitting through any one of the Downlink Control Information (DCI), Master Information Block (MIB), System Information Block (SIB), and Channel State Information Configuration Information messages containing the above AI / ML model identification information.

18. In a terminal managing AI / ML model information, A receiver that receives configuration information for training an AI / ML model and linked identification information linked to the configuration information from a base station; A control unit that performs training of the AI / ML model based on the above configuration information and the above linkage identification information; and It includes a transmitter that reports the AI / ML model identification information corresponding to the above-mentioned linked identification information to the base station, The above receiver is a terminal that further receives instruction information for instructing the AI / ML model from the base station.

19. In Paragraph 18, The above configuration information is, A terminal comprising at least one of model parameters for training the above AI / ML model, training data collection information, and model entity information, and distinguished and identified by the above-mentioned linkage identification information.

20. In Paragraph 18, The above AI / ML model identification information is, A terminal assigned based on at least one of model structure classification, model function classification, training sequence information and the above-mentioned linked identification information.

Images