Method and apparatus for management of models in a wireless communication system

Abstract

A method for a wireless communication system according to one embodiment of the present specification, performed by a first device, includes receiving first information related to groups and receiving second information related to management of the groups.

Description

[Technical Field]

[0001] The present specification relates to a method and apparatus for management of models in a wireless communication system. [Background technology]

[0002] Mobile communication systems were developed to provide voice services while ensuring user activity. However, the scope of mobile communication systems has expanded beyond voice to include data services, and currently, explosive traffic growth is causing resource shortages and users are demanding faster services, so more advanced mobile communication systems are required.

[0003] The next generation of mobile communication systems must be able to accommodate explosive data traffic, dramatically increase the transmission rate per user, accommodate a significantly increased number of connected devices, achieve extremely low end-to-end latency, and be energy efficient. To achieve this, various technologies are being researched, including dual connectivity, massive multiple input / multiple output (MIMO), in-band full duplex, non-orthogonal multiple access (NOMA), super wideband support, and device networking.

[0004] Communication functions may be realized based on the model(s) / functionality(s). In connection with this, operations for life cycle management (LCM) may be performed. As an example, multiple models / functions may be realized based on parameters (and structures) optimized / obtained for each operating environment / scenario. As an example, multiple functions may be supported by one device. Summary of the Invention [Problem to be solved by the invention]

[0005] As mentioned above, when multiple models / functionality are implemented / used for a communication function, LCM-related signaling (e.g., model / functionality activation / deactivation / update / switching / fallback, etc.) must be implemented for various factors related to the management of the model / functionality (e.g., the environment related to the communication function, the scenario related to a specific communication function, etc.).

[0006] In such a case, the overhead for LCM-related signaling may increase excessively. For example, LCM-related signaling may be frequently performed whenever the above-mentioned model / function-based LCM-related elements / characteristics are changed. Furthermore, if such LCM-related signaling is performed for each of multiple models / functions, the signaling overhead may further increase.

[0007] The purpose of this document is to propose a method for solving the aforementioned problems.

[0008] The technical problems to be achieved in this specification are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those having ordinary skill in the art to which the present invention pertains from the following description. [Means for solving the problem]

[0009] A method performed by a first device in a wireless communication system according to one embodiment of the present specification includes receiving (comprising; configuring; establishing; configuring; including; containing; having) first information related to groups and receiving second information related to management of the groups.

[0010] The group may relate to i) one or more models, and / or ii) one or more functionalities.

[0011] Based on the second information, an action associated with the one or more models and / or the one or more features may be directed.

[0012] The action may include at least one of i) activation, ii) deactivation, iii) selection, iv) switching, v) update, and / or vi) fallback.

[0013] The action may be directed based on an ID or unit associated with the group.

[0014] Based on the second information, at least one setting associated with the group may be reset.

[0015] The at least one configuration may include at least one of: i) a reference signal (RS) set associated with a measurement; ii) a prediction window associated with channel state information (CSI) and / or a beam; and / or iii) a granularity associated with reporting the CSI and / or the beam.

[0016] Each of the groups can be associated with a common property.

[0017] The common characteristic may be based on at least one of i) zone, ii) configuration, iii) data traffic, iv) mobility, and / or v) interference.

[0018] In another embodiment of the present specification, a first device operating in a wireless communication system includes one or more transceivers, one or more processors, and one or more memories connected to the one or more processors and configured to store instructions.

[0019] The instructions, when executed by the one or more processors, configure the one or more processors to perform all of the steps of any one of the methods.

[0020] According to still another embodiment of the present disclosure, an apparatus includes one or more memories and one or more processors operatively coupled to the one or more memories.

[0021] The one or more memories are characterized in that they store instructions that, when executed by the one or more processors, cause the one or more processors to perform all steps of any one of the methods.

[0022] According to yet another embodiment of the present disclosure, one or more non-transitory computer-readable media store instructions executable by one or more processors to configure the one or more processors to perform all of the steps of any one of the methods.

[0023] According to yet another embodiment of the present specification, a method performed by a second device in a wireless communication system includes transmitting first information related to groups and transmitting second information related to management of the groups.

[0024] A second device operating in a wireless communication system according to another embodiment of the present specification includes one or more transceivers, one or more processors, and one or more memories connected to the one or more processors and configured to store instructions.

[0025] The instructions, when executed by the one or more processors, configure the one or more processors to perform all of the steps of the method. [Effects of the Invention]

[0026] According to an embodiment of the present specification, model / function management (i.e., LCM) is performed based on predefined groups. Therefore, overhead can be reduced compared to when LCM-related signaling is performed for each model / function. Multiple models / functions can be effectively managed while minimizing signaling overhead.

[0027] Furthermore, the signaling / operations for managing multiple models / functions can be simplified, thereby reducing the complexity required to implement a device that supports the management of multiple models / functions.

[0028] The effects obtained in this specification are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those having ordinary skill in the art to which the present invention pertains from the following description. [Brief explanation of the drawings]

[0029] [Figure 1] Presents a functional framework for AI / ML models. [Figure 2] 1 illustrates a signaling procedure according to an embodiment of the present disclosure. [Figure 3] 1 is a flowchart illustrating a method executed by a first device according to an embodiment of the present specification. [Figure 4] 10 is a flowchart illustrating a method performed by a second device according to another embodiment of the present specification. [Figure 5] 1 is a diagram illustrating the configuration of a first device and a second device according to an embodiment of the present specification. DETAILED DESCRIPTION OF THE INVENTION

[0030] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings. The detailed description disclosed below, together with the accompanying drawings, is intended to describe exemplary embodiments of the present invention and is not intended to represent the only embodiments in which the present invention can be practiced. The following detailed description includes specific details to provide a thorough understanding of the present invention. However, those skilled in the art will recognize that the present invention can be practiced without such specific details.

[0031] In some cases, well-known structures and devices may be omitted or shown in block diagram form, focusing on the core functions of each structure and device, in order to avoid obscuring the concepts of the present invention.

[0032] Hereinafter, downlink (DL) refers to communication from a base station to a terminal, and uplink (UL) refers to communication from the terminal to the base station. In the downlink, the transmitter may be part of the base station, and the receiver may be part of the terminal. In the uplink, the transmitter may be part of the terminal, and the receiver may be part of the base station. The base station may also be expressed as a first communication device, and the terminal may also be expressed as a second communication device. A base station (BS) may also be replaced with terms such as a fixed station, NodeB, evolved-NodeB (eNB), Next Generation NodeB (gNB), base transceiver system (BTS), access point (AP), network (5G network), AI system, road side unit (RSU), vehicle, robot, unmanned aerial vehicle (UAV), augmented reality (AR) device, and virtual reality (VR) device. Furthermore, a terminal may be fixed or mobile, and may be replaced with terms such as UE (User Equipment), MS (Mobile Station), UT (user terminal), MSS (Mobile Subscriber Station), SS (Subscriber Station), AMS (Advance Mobile Station), WT (Wireless terminal), MTC (Machine-Type Communication) device, M2M (Machine-to-Machine) device, D2D (Device-to-Device) device, vehicle, robot, AI module, drone (Unmanned Aerial Vehicle, UAV), AR (Augmented Reality) device, VR (Virtual Reality) device, etc.

[0033] AIML related explanation

[0034] Advances in AI / ML (artificial intelligence / machine learning) technology are making the nodes and terminals that make up wireless communication networks more intelligent and sophisticated.

[0035] In particular, the intelligence of networks / base stations is expected to enable the rapid optimization, derivation, and application of various network / base station decision parameter values ​​according to various environmental parameters.

[0036] The environmental parameters may include at least one of the distribution / location of base stations, the distribution / location / material of buildings / furniture, etc., the location / movement direction / speed of the terminal, and weather information. However, the above parameters are merely examples, and the environmental parameters may further include other environmental parameters associated with the network / base station determination parameters in addition to the listed parameters.

[0037] The network / base station determined parameter values ​​may include at least one of the transmit / receive power of each base station, the transmit power of each terminal, the precoder / beam of the base station / terminal, the time / frequency resource allocation for each terminal, and the duplex mode of each base station, although the above parameters are only examples, and the network / base station determined parameter values ​​may further include other parameters determined by the network / base station in addition to the listed parameters.

[0038] In line with this trend, many standardization organizations (e.g., 3GPP (registered trademark: the same below), O-RAN) are considering the introduction of AI / ML, and active studies on this are also underway.

[0039] Although AI / ML can easily be referred to as deep learning-based artificial intelligence in a narrow sense, conceptually it can be divided as follows:

[0040] - Artificial Intelligence: This refers to any automation in which machines can replace the work that humans should do.

[0041] - Machine Learning: Machines learn patterns for decision-making from data without being explicitly programmed with rules.

[0042] - Deep Learning: An artificial neural network-based model that performs feature extraction and decision-making from unstructured data in one go. The algorithm relies on a multi-layer network of interconnected nodes for feature extraction and transformation inspired by the biological nervous system, i.e., the neural network. Common deep learning network architectures include deep neural networks (DNNs), recurrent neural networks (RNNs), and convolutional neural networks (CNNs).

[0043] 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 narrower concept than artificial intelligence, but broader than deep learning.

[0044] AI / ML typologies based on various criteria

[0045] - Offline vs. Online

[0046] Offline Learning

[0047] - Offline learning strictly follows the sequential steps of database collection, learning, and prediction. In other words, collection and learning are performed offline, and the completed program can be installed in the field and used for prediction work. In most situations, this offline learning method is used.

[0048] Online Learning

[0049] - Recently, data that can be used for learning is continuously generated via the Internet. This method of incrementally improving performance through intensive additional learning using additional data is called online learning.

[0050] Classification by AI / ML Framework Concept

[0051] -Centralized Learning

[0052] In centralized learning, training data collected from multiple different nodes is reported to a centralized node, and all data resources, storage, learning (e.g., supervised, unsupervised, reinforcement learning), etc. are executed on a single centralized node.

[0053] - Federated Learning

[0054] Federated learning is where a collective model is built on data spread across distributed data owners. Instead of bringing the data to the model, the AI / ML model is brought to the data source, allowing local nodes / individual devices to collect data and train their own copies of the model, without the need to report source data back to a central node.

[0055] In federated learning, the parameters / weights of an AI / ML model are returned to a centralized node to support general model training. The advantages of federated learning include increased computational speed and superior information security. This means that there is no need to upload personal data to a central server, preventing the leakage and misuse of personal information.

[0056] - Distributed Learning

[0057] Distributed learning describes the concept of machine learning processes being scaled and distributed across a cluster of nodes: training models are split and shared across multiple nodes working simultaneously to speed up model training.

[0058] Classification by learning method

[0059] - Supervised Learning

[0060] Supervised learning is a machine learning task whose goal is to learn a mapping function from input to output given a labeled dataset. The input data is called training data, and has known labels or outcomes. Examples of supervised learning include:

[0061] 1) Regression: Linear Regression, Logistic Regression

[0062] 2) Instance‐based Algorithms: k‐Nearest Neighbor(KNN)

[0063] 3)Decision Tree Algorithms:CART

[0064] 4) Support Vector Machines (SVM)

[0065] 5) Bayesian Algorithms: Naive Bayes

[0066] 6) Ensemble Algorithms: Extreme Gradient Boosting, Bagging: Random Forest

[0067] Supervised learning can be further grouped into regression and classification problems, where classification is about predicting a label and regression is about predicting a quantity.

[0068] - Unsupervised Learning

[0069] Unsupervised learning is a machine learning task that aims to learn functions that explain hidden structure from unlabeled data. The input data is unlabeled and there is no known outcome. Some examples of unsupervised learning include K-means clustering, principal component analysis (PCA), nonlinear independent component analysis (ICA), and LSTM.

[0070] - Reinforcement Learning

[0071] In reinforcement learning (RL), an agent interacts with the environment based on a trial-and-error process with the aim of optimizing a long-term goal through goal-directed learning. Below are some example RL algorithms:

[0072] 1) Q-learning

[0073] 2) Multi-armed bandit learning

[0074] 3) Deep Q Network

[0075] 4)State-Action-Reward-State-Action (SARSA)

[0076] 5)Temporal Difference Learning

[0077] 6)Actor‐critic reinforcement learning

[0078] 7) Deep deterministic policy gradient

[0079] 8) Monte‐Carlo tree search

[0080] Reinforcement learning can be further grouped into model-based reinforcement learning and model-free reinforcement learning.

[0081] Model-based reinforcement learning: RL algorithms that use predictive models, different dynamic states of the environment, and models that lead to compensation to obtain transition probabilities between states.

[0082] Model-free reinforcement learning: A value- or policy-based RL algorithm that achieves maximum future rewards in a multi-agent environment / state, is computationally less complex, and does not require an accurate representation of the environment.

[0083] RL algorithms can also be classified as value-based RL vs. policy-based RL, policy-based RL vs. policy-free RL, etc.

[0084] Representative model of deep learning

[0085] 1. FFNN (Feed‐Forward Neural Network)

[0086] An FFNN consists of an input layer, a hidden layer, and an output layer.

[0087] 2. RNN (Recurrent Neural Network)

[0088] RNN is a type of artificial neural network in which hidden nodes are connected by directional edges to form a cyclic structure (directed cycle). It is a model suitable for processing sequential data such as voice and text.

[0089] 3. CNN(Convolution Neural Network)

[0090] CNNs are used for two purposes: to reduce model complexity and to extract good features by applying convolution operations, which are commonly used in the fields of video processing and image processing.

[0091] - Kernel or filter: a unit / structure that applies weights to inputs of a certain range / unit

[0092] - Stride: the range of movement of the kernel within the input

[0093] - Feature map: the result of applying the kernel to the input

[0094] - Padding: Value added to adjust the size of the feature map.

[0095] - Pooling: An operation to downsample a feature map to reduce its size (e.g., max pooling, average pooling).

[0096] 4. Auto encoder

[0097] An autoencoder is a neural network that receives a feature vector x and outputs the same or similar vector x'. In an autoencoder, the input and output nodes have the same features.

[0098] Figure 1 shows the functional framework of the AI / ML model.

[0099] The definitions of each term and the operations by function in the framework shown in FIG. 1 are based on Table 1 below.

[0100] [Table 1] JPEG2026507536000003.jpg197144JPEG2026507536000004.jpg196145JPEG2026507536000005.jpg84143

[0101] JPEG2026507536000006.jpg74143

[0102] Data Set

[0103] Data sets used in AI / ML are divided into training data, validation data, and test data, which are defined as follows:

[0104] - Training data

[0105] Data set for training the model

[0106] - Validation data

[0107] A data set for validating a model that has already been trained

[0108] Validation data is a data set that is typically used to prevent overfitting of the training data set.

[0109] Validation data is a dataset used to select the best model from various models trained during the training process. Therefore, validation data can be considered a dataset related to training.

[0110] - Test data

[0111] Dataset for final evaluation. Test data is unrelated to training.

[0112] For the data set, a full training set containing a certain percentage of the aforementioned data may be used.

[0113] As an example, a training set containing training data and validation data in a ratio of 8:2 or 7:3 may be used.

[0114] As an example, a training set containing training data, validation data, and test data in a ratio of 6:2:2 may be used.

[0115] Collaboration level

[0116] Depending on whether or not the AI ​​ / ML function is capable between the base station and the terminal, the cooperation level can be defined as shown in Table 4 below.

[0117] [Table 2]

[0118] The cooperation levels in Table 2 are merely examples and may be modified and utilized differently depending on the implementation. For example, a cooperation level that combines two or more of the cooperation levels illustrated may be defined / utilized. For example, a cooperation level that excludes one or more of the cooperation levels illustrated may be utilized.

[0119] In this document, a method is proposed to reduce the signaling overhead for model / feature management.

[0120] The embodiments described below are not limited to application to management of AI / ML models. As an example, the embodiments described below may be applied to management of models that implement communication functions. As an example, the embodiments described below may be applied to management of models related to communication functions. Hereinafter, "AI / ML model" may be interpreted / substituted as "model," "model related to a (specific) communication function," "model for a (specific) communication function," "model related to a (specific) communication procedure," or "model for a (specific) communication procedure."

[0121] In this document, ' / ' means 'and', 'or', or 'and / or' depending on the context. In this specification, 'terminal' and 'UE' can be used interchangeably with the same / similar meaning, and 'base station', 'network', and 'TRP' can also be used interchangeably with the same / similar meaning from the perspective of the air interface.

[0122] Recently, there have been active attempts to apply AI / ML (artificial intelligence / machine learning) technology to wireless communication networks. The 3GPP Rel-18 AI / ML study item initiated a study on applying AI / ML technology to the air interface between terminals and networks. The study considers beam management (BM), CSI acquisition, and positioning as key use cases for incorporating AI / ML into the air interface, and various related terms are defined and discussed as shown in Table 3 below. The study also discusses methods for managing AI / ML models. For example, AI / ML model management can refer to life cycle management (LCM). LCM can include at least one of model activation / deactivation, model switching / selection, model monitoring, model update / (fine)tuning, and / or fallback.

[0123] [Table 3] JPEG2026507536000009.jpg56141

[0124] JPEG2026507536000010.jpg139143JPEG2026507536000011.jpg121142

[0125] JPEG2026507536000012.jpg206144JPEG2026507536000013.jpg78142

[0126] JPEG2026507536000014.jpg94143

[0127] For the UE-part / side model, two approaches / methods for model management are discussed, as shown in the last agreement in Table 3. Here, the UE-part model refers to the UE part of the two-sided model, and the UE-side model refers to a one-sided model in which AI is implemented only in the UE.

[0128] The functionality-based approach focuses on the functions or capabilities supported by the application of AI / ML models. The standardization target of the functionality-based approach is signaling related to activation / deactivation / switching / fallback for functionality.

[0129] The target of standardization for the model-based method is signaling related to activation / deactivation / switching / fallback for AI / ML models.

[0130] The biggest differences between the two methods mentioned above are:

[0131] It may be assumed that a terminal implements multiple AI / ML models for the same functionality according to various environments / scenarios in which the functionality may be applied, related base station settings, etc.

[0132] Model-based LCM directly manages which models to activate / deactivate / select / switch. On the other hand, functionality-based LCM activates / deactivates / monitors the corresponding models from the perspective of their functionality / performance. Therefore, functionality-based LCM may not require information about how many models the base station (NW) and / or terminal have implemented.

[0133] The subject of monitoring the model / functionality of the UE-part / side model may be the UE or / and the NW. The subject that determines / performs activation / deactivation / selection / switching / update of the model / functionality may be the UE or / and the NW.

[0134] An example of the UE and NW performing the LCM process together is as follows: When the UE recommends / reports a model selection / update decision to the NW, the NW can perform final confirmation / decision on the decision.

[0135] As mentioned above, once the NW is involved in the LCM process, signaling between the relevant UE and the NW is inevitable.

[0136] Table 4 below illustrates LCM signaling between the UE and the NW.

[0137] [Table 4]

[0138] This LCM signaling must be performed whenever the environment / scenario changes, especially when multiple AI / ML models are implemented by optimizing / acquiring parameters (and structure) for each operating environment / scenario. Furthermore, AI / ML models for multiple functionalities can be implemented in one device / terminal (e.g., N1 models for CSI prediction, N2 models for CSI compression, N3 models for beam prediction, N4 models for positioning), which further increases the LCM signaling overhead. A method for solving the above problem will be described in detail below.

[0139] Method 1

[0140] A plurality of models / functionalities can be grouped, and model / functionality LCM signaling can be defined / executed based on the corresponding group (or the unit / ID corresponding to the corresponding group).

[0141] The grouping may be performed as follows: AI / ML models for different functionalities or models trained for the same or similar operating environment / scenario may be grouped together. That is, one or more groups may be determined based on multiple models / functions.

[0142] Here, the operating environment / scenarios may be classified / categorized in various aspects such as indoor / outdoor, urban / suburban, zone / site / configuration (ID), base station / TRP placement / density, terminal placement / density, inter- / intra-cell interference, NW data traffic, terminal movement speed, terminal form factor, etc. According to this embodiment, LCM signaling may be performed as follows.

[0143] As an example, in the case of model-based LCM, model group (ID)-based LCM signaling may be defined / implemented.

[0144] For example, in the case of a functionality-based LCM, individual models may not appear standard. That is, individual models may not be explicitly defined. In this case, LCM signaling may be defined / executed based on a specific unit (corresponding to a model group). The model group may be defined as a certain type of unit or may be a unit distinguished by any ID (e.g., index, indicator).

[0145] For example, zone / setting / scenario / environment IDs can be defined. Model group-based LCM can be performed by defining / executing LCM signaling based on the corresponding IDs. The method can also be applied to model-based LCM. Specifically, the method can be applied as follows: Multiple models can belong to the same zone / setting / scenario / environment ID (e.g., assigning / giving corresponding IDs to information for each model), and LCM signaling can be defined / executed based on the corresponding IDs. Specific application examples are shown in Table 5 below.

[0146] [Table 5]

[0147] When group-based LCM is performed according to the proposed method, the signaling / RS overhead can be further reduced by specifying that the associated RS, channel, reporting settings, etc. are changed by the corresponding LCM signaling.

[0148] In other words, based on the LCM signaling, at least one of the RS, channel, and / or reporting settings associated with the corresponding LCM signaling may be changed / updated. As an example, the change / update of at least one of the RS, channel, and / or reporting settings associated with the LCM signaling may be defined to be performed by the subject executing the LCM (without any additional signaling). As an example, the subject executing the LCM may operate under the assumption that at least one of the RS, channel, and / or reporting settings associated with the corresponding LCM signaling has been changed / updated when the LCM signaling is executed. As an example, the LCM signaling may be utilized as an instruction to change / update not only the LCM but also at least one of the associated RS, channel, and / or reporting settings.

[0149] Below, specific examples of changes / updates of related RSs, channels, reporting settings, etc. by LCM signaling will be described.

[0150] For example, measurement CSI-RS / SSB resource sets / groups related to CSI compression, CSI prediction, beam prediction, positioning, etc. may be changed collectively in accordance with the model group change related instruction. This operation can reduce RS reset signaling in response to changes in the operating environment (e.g., measurement sets are changed based on movement from cell / TRP A to cell / TRP B).

[0151] For example, CSI / beam report setting values ​​(e.g., prediction window, reporting granularity, etc.) related to CSI compression, CSI prediction, beam prediction, etc. may be varied collectively according to the model group-based LCM signaling. Through this operation, it is possible to reduce report reset signaling due to changes in the operating environment.

[0152] In this case, information regarding which group and which RS set / group, report setting value to change / activate / deactivate can be set / defined in advance (e.g., setting the relevant information through an RRC message, defining the relevant information when realizing the terminal / network (base station)).

[0153] Alternatively, it may be specified to apply specific specified values ​​(e.g., values ​​corresponding to fallback operation). For example, when fallback operation is instructed via LCM signaling, it may be defined that specific parameters of report settings related to CSI prediction and beam prediction are set to specified default values ​​(e.g., resetting values ​​related to the prediction window to default values ​​and executing non-predicted CSI / beam report).

[0154] Method 1-1

[0155] (When Method 1 is applied) Based on LCM signaling, RS / channel / report related settings can be reset to pre-configured or promised / specified values.

[0156] Method 1-1 can also be applied when Method 1 is not applied, ie, to single model / functionality-based LCM.

[0157] The proposed methods (e.g., Method 1, Method 1-1) can be extended to model group-based transfer / delivery methods in addition to LCM signaling. For example, model transfer / delivery can be performed for the model group according to changes in the operating environment / scenario.

[0158] In an implementation aspect, the operation of a base station / terminal according to the above-described embodiments (e.g., operation based on at least one of Method 1, Methods 1-1) may be processed by the apparatus of Figure 5 (e.g., processors 110, 210 of Figure 5) described below.

[0159] In addition, the operation of the base station / terminal according to the above-mentioned embodiments (e.g., operation based on at least one of Method 1 and Method 1-1) may be stored in a memory (e.g., 140, 240 in FIG. 5) in the form of instructions / programs (e.g., instructions, executable code) for driving at least one processor (e.g., 110, 210 in FIG. 5).

[0160] In the following we take a closer look at the signalling procedure according to the above mentioned embodiment.

[0161] FIG. 2 illustrates a signaling procedure according to an embodiment of the present disclosure.

[0162] FIG. 2 shows an example of signaling between a user equipment (UE) and a network (NW) based on the proposed method. Here, the UE and network are merely examples, and various devices may be used instead. FIG. 2 is provided for convenience of explanation only and does not limit the scope of this specification. Furthermore, some steps shown in FIG. 2 may be omitted depending on the situation and / or settings.

[0163] This can correspond to any entity such as a base station, BS (Base Station), Node B, or TRP belonging to the NW in Figure 2, and the UE can also be replaced with an entity / server responsible for UE-related AI / ML operations instead of the UE.

[0164] 2 illustrates a model / functionality switching (or selection) procedure, which is one embodiment of the proposed LCM signaling, and other procedures may be performed for other LCM procedures (e.g., model / functionality selection / activation / deactivation / update / fallback / etc.) In addition to the procedure illustrated in FIG. 2, the model / functionality switching procedure may also be performed through other procedures.

[0165] The UE may perform a procedure (S205) of reporting terminal information regarding the AI / ML model and / or its associated features / functionality to the base station. Such information may include terminal capability values ​​(UE capabilities).

[0166] The base station may perform configuration for related functions / parameters and / or AI / ML models based on the terminal capability values ​​reported in S205 (S210).

[0167] Thereafter, a model / functionality switching procedure between the base station and the terminal can be performed. The detailed process of the model / functionality switching procedure may vary, and FIG. 2 shows one embodiment of the procedure. In order for the network to determine model switching, the UE may transmit monitoring information regarding model performance to the base station (S215). Alternatively, if the base station determines model switching without based on the terminal's report information, this procedure may be omitted.

[0168] The network can issue a command for model / functionality switching to the terminal (S220). In this procedure, the proposed methods (e.g., Method 1, Method 1-1) of this specification can be applied. For example, a switching command based on a model group ID or a certain type of zone / setting ID can be used. Also, RS / channel / report-related settings can be changed based on such a command.

[0169] Upon receiving the command (S220), the terminal may perform switching and related operations for the model / functionality and transmit related acknowledgement / confirmation messages to the network (S225).

[0170] As mentioned above, the above-described NW / UE signaling and operations may be realized by the following devices (FIG. 5). For example, the NW may be the first wireless device and the UE may be the second wireless device, or vice versa.

[0171] For example, the above-mentioned NW / UE signaling and operations may be processed by one or more processors 110, 210 of FIG. 5, and the above-mentioned NW / UE signaling and operations may be stored in memory 140, 240 in the form of instructions / programs (e.g., instructions, executable code) for driving at least one processor 110, 210 of FIG. 5.

[0172] Hereinafter, the above-mentioned embodiment will be described in detail with reference to FIGS. 3 and 4 in terms of the operation of the first and second devices. The methods described below are distinguished only for the sake of convenience, and it is understood that some components of one method may be replaced with some components of another method or may be combined with each other. Hereinafter, the first device and the second device may be entities that perform signaling (and / or model transfer / delivery) related to model / function management. For example, the first device may be a terminal, a base station, or a first terminal, and the second device may be a base station, a terminal, or a second terminal.

[0173] FIG. 3 is a flowchart illustrating a method performed by a first device according to an embodiment of the present specification.

[0174] Referring to FIG. 3, a method performed by a first device according to an embodiment of the present specification includes a first information receiving step S310 and a second information receiving step S320.

[0175] In S310, the first device receives first information about groups from the second device. For example, the groups may be set based on the first information. For example, a plurality of models / functions may be classified / determined into the groups based on the first information.

[0176] According to one embodiment, each of the groups may be associated with a common property. This embodiment may be based on Method 1.

[0177] As an example, the common characteristic may be based on at least one of i) zone, ii) configuration, iii) data traffic, iv) mobility, and / or v) interference.

[0178] At S320, the first device receives second information related to the management of the group from the second device.

[0179] According to one embodiment, the group may be associated with i) one or more models and / or ii) one or more functionalities.

[0180] For example, an action related to the one or more models and / or one or more functions may be instructed based on the second information.

[0181] By way of example, the action may include at least one of: i) activation, ii) deactivation, iii) selection, iv) switching, v) update, and / or vi) fallback.

[0182] As an example, the action may be directed based on an ID or unit associated with the group.

[0183] According to an embodiment, at least one setting related to the group may be reset based on the second information. This embodiment may be based on Method 1 and / or Method 1-1.

[0184] As an example, the at least one setting may be related to at least one of CSI compression, CSI prediction, beam prediction, and / or positioning.

[0185] As an example, the at least one configuration may include at least one of: i) a Reference Signal (RS) set associated with a measurement; ii) a prediction window associated with Channel State Information (CSI) and / or a beam; and / or iii) a granularity associated with reporting the CSI and / or beam.

[0186] The operations according to S310 and S320 described above may be implemented by the apparatus of Figure 5. For example, the first device 100 / 200 may control one or more transceivers 130 / 230 and / or one or more memories 140 / 240 to perform the operations according to S310 and S320.

[0187] The above-described embodiment will now be described in detail with reference to the operation of the second device.

[0188] S410 to S420 described below correspond to S310 to S320 described in Fig. 3. In consideration of the correspondence, duplicated explanations will be omitted. That is, specific explanations of the operations of the second device described below can be replaced with the explanations / embodiments of Fig. 3 corresponding to the operations.

[0189] FIG. 4 is a flowchart illustrating a method performed by a second device according to another embodiment of the present disclosure.

[0190] Referring to FIG. 4, a method performed by a second device according to another embodiment of the present specification includes a first information transmitting step S410 and a second information transmitting step S420.

[0191] At S410, the second device transmits first information related to groups to the first device.

[0192] At S420, the second device transmits second information related to the management of the group to the first device.

[0193] The operations according to S410 to S420 described above may be performed by the apparatus of Figure 5. For example, second device 100 / 200 may control one or more transceivers 130 / 230 and / or one or more memories 140 / 240 to perform the operations according to S410 to S420.

[0194] An apparatus to which the embodiments of the present specification can be applied (an apparatus that implements the methods / operations according to the embodiments of the present specification) will be described below with reference to FIG.

[0195] FIG. 5 is a diagram showing the configurations of the first device and the second device according to the embodiment of the present specification.

[0196] The first device 100 may include a processor 110 , an antenna unit 120 , a transceiver 130 , and a memory 140 .

[0197] The processor 110 performs baseband-related signal processing and may include an upper layer processing unit 111 and a physical layer processing unit 115. The upper layer processing unit 111 may process operations of the MAC layer, the RRC layer, or higher layers. The physical layer processing unit 115 may process operations of the PHY layer. For example, when the first device 100 is a base station device in base station-terminal communication, the physical layer processing unit 115 may perform uplink reception signal processing, downlink transmission signal processing, etc. For example, when the first device 100 is a first terminal device in terminal-terminal communication, the physical layer processing unit 115 may perform downlink reception signal processing, uplink transmission signal processing, sidelink transmission signal processing, etc. In addition to performing baseband-related signal processing, the processor 110 may also control the overall operation of the first device 100.

[0198] The antenna unit 120 may include one or more physical antennas, and when multiple antennas are included, MIMO transmission and reception may be supported. The transceiver 130 may include an RF (Radio Frequency) transmitter and an RF receiver. The memory 140 may store information processed by the processor 110, as well as software, an operating system, applications, etc. related to the operation of the first device 100, and may also include components such as buffers.

[0199] The processor 110 of the first device 100 can be configured to implement the operation of a base station in base station-terminal communication (or the operation of a first terminal device in terminal-terminal communication) in the embodiments described in this disclosure.

[0200] The second device 200 may include a processor 210 , an antenna unit 220 , a transceiver 230 , and a memory 240 .

[0201] The processor 210 performs baseband-related signal processing and may include an upper layer processing unit 211 and a physical layer processing unit 215. The upper layer processing unit 211 may process operations of the MAC layer, the RRC layer, or higher layers. The physical layer processing unit 215 may process operations of the PHY layer. For example, when the second device 200 is a terminal device in base station-terminal communication, the physical layer processing unit 215 may perform downlink reception signal processing, uplink transmission signal processing, etc. For example, when the second device 200 is a second terminal device in terminal-terminal communication, the physical layer processing unit 215 may perform downlink reception signal processing, uplink transmission signal processing, sidelink reception signal processing, etc. In addition to performing baseband-related signal processing, the processor 210 may also control the overall operation of the second device 200.

[0202] The antenna unit 220 may include one or more physical antennas, and if multiple antennas are included, MIMO transmission and reception may be supported. The transceiver 230 may include an RF transmitter and an RF receiver. The memory 240 may store information processed by the processor 210, as well as software, an operating system, applications, etc. related to the operation of the second device 200, and may also include components such as buffers.

[0203] The processor 210 of the second device 200 may be configured to implement the operation of a terminal in base station-terminal communication (or the operation of a second terminal device in terminal-terminal communication) in the embodiments described in this disclosure.

[0204] In the operation of the first device 100 and the second device 200, the matters described in the examples of the present disclosure regarding the base station and terminal in base station-terminal communication (or the first terminal and second terminal in terminal-terminal communication) can be equally applied, and duplicate explanations will be omitted.

[0205] Here, the wireless communication technology implemented by the devices 100 and 200 of the present disclosure may include not only LTE, NR, and 6G, but also Narrowband Internet of Things (NB-IoT) for low-power communication. For example, the NB-IoT technology is an example of a Low Power Wide Area Network (LPWAN) technology, and can be implemented by standards such as LTE Cat NB1 and / or LTE Cat NB2, and is not limited to the aforementioned names.

[0206] Additionally or alternatively, the wireless communication technology implemented in the devices 100 and 200 of the present disclosure may perform communication based on LTE-M technology. For example, LTE-M technology is an example of LPWAN technology and is referred to by various names such as enhanced machine type communication (eMTC). For example, LTE-M technology may be implemented by at least one of various standards such as 1) LTE CAT 0, 2) LTE Cat M1, 3) LTE Cat M2, 4) LTE non-BL (non-Bandwidth Limited), 5) LTE-MTC, 6) LTE Machine Type Communication, and / or 7) LTE M, and is not limited to the above names.

[0207] Additionally or alternatively, the wireless communication technology implemented in the devices 100 and 200 of the present disclosure may include at least one of ZigBee, Bluetooth, and a Low Power Wide Area Network (LPWAN), which consider low-power communication, but is not limited to the aforementioned names. For example, ZigBee technology can create personal area networks (PANs) related to small / low-power digital communication based on various standards such as IEEE 802.15.4, and may be called by various names.

[0208] [Claims at the time of international application] [Claim 1] 1. A method performed by a first device in a wireless communication system, comprising: receiving first information related to groups; receiving second information related to management of the group. [Claim 2] The group is i) one or more models, and / or ii) a method according to claim 1, characterized in that it is related to one or more functionalities. [Claim 3] 3. The method of claim 2, wherein an action related to the one or more models and / or the one or more functions is indicated based on the second information. [Claim 4] The operation is i) activation; ii) deactivation; iii) selection, iv) switching, v) Updates, and / or 4. The method of claim 3, further comprising at least one of: vi) a fallback. [Claim 5] 4. The method of claim 3, wherein the action is directed based on an ID or unit associated with the group. [Claim 6] 2. The method of claim 1, wherein at least one setting associated with the group is reset based on the second information. [Claim 7] The at least one setting may include: i) a set of Reference Signals (RS) associated with the measurement; ii) Channel State Information (CSI) and / or a prediction window associated with the beam, and / or 10. The method of claim 6, further comprising at least one of: iii) CSI and / or beam reporting related granularity. [Claim 8] 2. The method of claim 1, wherein each of the groups is associated with a common property. [Claim 9] The common features are: i) zone, ii) configuration, iii) data traffic; iv) mobility, and / or 9. The method of claim 8, wherein the method is based on at least one of: v) interference. [Claim 10] A first device operating in a wireless communication system, comprising: one or more transceivers; one or more processors; one or more memories coupled to the one or more processors and configured to store instructions; A first device, characterized in that the instructions, when executed by the one or more processors, configure the one or more processors to perform all steps of the method according to any one of claims 1 to 9. [Claim 11] 1. An apparatus comprising: one or more memories; one or more processors operatively connected to the one or more memories; An apparatus, characterized in that the one or more memories store instructions that, when executed by the one or more processors, cause the one or more processors to perform all steps of the method described in any one of claims 1 to 9. [Claim 12] one or more non-transitory computer-readable media storing instructions, One or more non-transitory computer-readable media, characterized in that the instructions executable by one or more processors configure the one or more processors to perform all steps of the method of any one of claims 1 to 9. [Claim 13] 1. A method in a wireless communication system performed by a second device, comprising: transmitting first information related to groups; transmitting second information related to management of the group. [Claim 14] a second device operating in a wireless communication system, one or more transceivers; one or more processors; one or more memories coupled to the one or more processors and configured to store instructions; 14. The second device, wherein the instructions, when executed by the one or more processors, configure the one or more processors to perform all of the steps of the method of claim 13.

Claims

1. 1. A method performed by a first device in a wireless communication system, comprising: receiving first information related to groups; receiving second information related to the management of the group.

2. The group is i) one or more models, and / or 2. The method of claim 1, wherein ii) the method is associated with one or more functionalities.

3. 3. The method of claim 2, wherein an action associated with the one or more models and / or the one or more functions is directed based on the second information.

4. The operation is i) activation; ii) deactivation; iii) selection; iv) switching; v) update, and / or vi) a fallback.

5. 4. The method of claim 3, wherein the action is directed based on an ID or unit associated with the group.

6. 2. The method of claim 1, wherein at least one setting associated with the group is reset based on the second information.

7. The at least one setting may include: i) a set of Reference Signals (RS) associated with the measurement; ii) Channel State Information (CSI) and / or a prediction window associated with a beam, and / or 10. The method of claim 6, further comprising at least one of: iii) CSI and / or beam reporting related granularity.

8. 2. The method of claim 1, wherein each of the groups is associated with a common property.

9. The common features are: i) zone; ii) configuration; iii) data traffic; iv) mobility, and / or 9. The method of claim 8, wherein the method is based on at least one of: v) interference.

10. 1. A first device operating in a wireless communication system, comprising: one or more transceivers; one or more processors; one or more memories coupled to the one or more processors and configured to store instructions; The first device, characterized in that the instructions, when executed by the one or more processors, configure the one or more processors to perform all steps of the method of any one of claims 1 to 9.

11. 1. An apparatus comprising: one or more memories; one or more processors operatively connected to the one or more memories; The one or more memories store instructions that, when executed by the one or more processors, configure the one or more processors to perform all steps of the method of any one of claims 1 to 9.

12. one or more non-transitory computer-readable media storing instructions, One or more non-transitory computer-readable media, characterized in that the instructions executable by one or more processors configure the one or more processors to perform all the steps of the method of any one of claims 1 to 9.

13. 1. A method performed by a second device in a wireless communication system, comprising: transmitting first information related to groups; transmitting second information related to the management of the group.

14. a second device operating in a wireless communication system, one or more transceivers; one or more processors; one or more memories coupled to the one or more processors and configured to store instructions; 14. The second device, wherein the instructions, upon being executed by the one or more processors, configure the one or more processors to perform all of the steps of the method of claim 13.

Images