Systems and methods for UE-side AI / ML model configuration for beam management

Abstract

A system and method performed by a UE in a wireless communication system includes: receiving a message corresponding to a data collection request of the UE from a base station; transmitting a data collection request having at least one of the preferred configuration or the time interval to the base station; receiving, from the base station, a response for enabling UE data collection in response to the data collection request, the response comprising a resource configuration and a condition identifier; and performing data collection through the measurement resource configuration.

Description

[0001] Cross Reference to Related Applications

[0002] This application claims priority to U.S. Provisional Application No. 63 / 649,008, filed May 17, 2024, the disclosure of which is incorporated by reference herein in its entirety, as if fully set forth herein. TECHNICAL FIELD

[0003] The present disclosure generally relates to wireless communication systems employing artificial intelligence (AI) and machine learning (ML). More specifically, the subject matter disclosed herein relates to improvements in procedures and signaling of network (NW) side additional conditions in user equipment (UE) side AI / ML models for beam management use cases. BACKGROUND

[0004] The Third Generation Partnership Project (3GPP) Radio Access NW (RAN) has introduced AI / ML based features for the New Radio (NR) air interface. Specifically, in Release 19, 3GPP supports both UE side AI / ML models and NW side AI / ML models, where the UE side models are trained by the UE vendor using data measured from a specific NW configuration. During inference, these models are applied to cells whose configuration is aligned with the configuration used during training. Differences between the training and inference conditions, such as differences in the mapping between beam codebooks and beam indices or subsets of beams, can degrade model performance. SUMMARY

[0005] To address this issue, previous approaches implicitly provide the NW side additional conditions through an associated identity (ID). However, these approaches do not clearly define the procedures for ID allocation, validation, and data acquisition for training specific to a cell or a base station (gNB).

[0006] To overcome these issues, the embodiments described herein provide improved signaling and procedures to manage UE data collection based on associated IDs provided by the NW. Specifically, detailed procedures for allocating IDs, validating NW side conditions, and determining applicable AI / ML functions based on these IDs are disclosed. Embodiments also propose handling data collection during handover, thereby ensuring model consistency and performance.

[0007] The above approach improves previous solutions by clearly defining the conditions that trigger data collection requests, mechanisms for updating and validating NW side conditions, and procedures for managing AI / ML model training during handover.

[0008] According to an embodiment, a method performed by a UE in a wireless communication system includes receiving, from a base station, a message corresponding to a data collection request of the UE; transmitting, to the base station, the data collection request with at least one of a preference configuration or a time interval; receiving, from the base station, a response for enabling UE data collection in response to the data collection request, the response including a resource configuration and a condition identifier; and performing data collection by measuring the resource configuration.

[0009] According to another embodiment, a method performed by a base station in a wireless communication system includes transmitting, to a UE, a message corresponding to a data collection request of the UE; receiving, from the UE, the data collection request with at least one of a preference configuration or a time interval; transmitting, to the UE, a response for enabling UE data collection in response to the data collection request, the response including a resource configuration and a condition identifier; and enabling data collection at the UE by measuring the resource configuration.

[0010] According to another embodiment, a non-transitory computer-readable medium storing instructions is provided. The instructions, when executed by one or more processors of a UE in a wireless communication system, cause the UE to receive, from a base station, a message corresponding to a data collection request of the UE; transmit, to the base station, the data collection request with at least one of a preference configuration or a time interval; receive, from the base station, a response for enabling UE data collection in response to the data collection request, the response including a resource configuration and a condition identifier; and perform data collection by measuring the resource configuration. BRIEF DESCRIPTION OF DRAWINGS

[0011] In the following description, aspects of the subject matter disclosed herein will be described in reference to exemplary embodiments shown in the drawings, wherein:

[0012] Figure 1 A transmitting device or a receiving device in a communication system according to an embodiment is illustrated;

[0013] Figure 2 An exemplary NW environment depicting UE-side AI / ML model training with associated NW-side conditions according to an embodiment is illustrated;

[0014] Figure 3 Signaling exchange between a UE and a gNB for data collection according to an embodiment is illustrated;

[0015] Figure 4 Signaling exchange between a UE and a gNB for configuring and / or activating applicable functions according to an embodiment is illustrated;

[0016] Figure 5 Signaling exchange for an exemplary gNB handover scenario according to an embodiment is illustrated;

[0017] Figure 6Ais a flowchart illustrating a method for beam management model training performed by a UE according to an embodiment;

[0018] Figure 6B is a flowchart illustrating a method for supporting beam management model training at a UE performed by a base station according to an embodiment;

[0019] Figure 7 is a block diagram of an electronic device in a NW environment according to an embodiment; and

[0020] Figure 8 is a system including a UE and a base station gNB in communication with each other according to an embodiment. DETAILED DESCRIPTION

[0021] In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the disclosure. However, it will be understood by those skilled in the art that the disclosed aspects can be practiced without these specific details. In other instances, well-known methods, procedures, components and circuits have not been described in detail so as not to obscure the subject matter disclosed herein.

[0022] References throughout this specification to “one embodiment” or “an embodiment” mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. The appearances of the phrase “in one embodiment” or “in an embodiment” or “according to one embodiment” (or other similar phrases) throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In this respect, the term “exemplary” as used herein means “serving as an example, instance, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Moreover, depending on the context, singular terms can include their corresponding plural forms and plural terms can include their corresponding singular forms. Similarly, terms with hyphenated

[0023] Furthermore, depending on the context, singular terms can include their corresponding plural forms, and plural terms can include their corresponding singular forms. It will be further understood that various drawings shown and discussed herein are merely illustrative and are not drawn to scale. For example, the dimensions of some of the elements can be exaggerated relative to other elements for clarity. Further, if considered appropriate, reference numerals can be repeated among the figures to indicate corresponding and / or analogous elements.

[0024] The terminology used herein is for the purpose of describing some example embodiments only and is not intended to be limiting of the claimed subject matter. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025] It will be understood that when an element or layer is referred to as being "on" or "connected to" or "coupled to" another element or layer, it can be directly on, connected or coupled to the other element or layer, or intervening elements or layers can be present. In contrast, when an element is referred to as being "directly on," "directly connected to," or "directly coupled to" another element or layer, there are no intervening elements or layers present. Like reference numerals refer to like elements throughout. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0026] As used herein, the terms "first," "second," and the like, are used as labels for nouns that they follow so that one element is distinguished from another, and are not necessarily used in their ordinal sense. Also, the use of "adapted to" or "configured to," or "arranged to" or similar terminology connects to the sufficient structure of one or more elements, as is used in the description, and does not impose constructional or architectural requirements to the sufficient structure. Moreover, the use of "based on," "configured by" or "arranged by" is not limiting, and is used to indicate one or more elements.

[0027] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this subject matter belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0028] A "module" as used herein refers to any combination of software, firmware, and / or hardware configured to provide the functionality described herein with respect to the module. For example, software can be embodied in software packages, code, and / or instructions sets or instruction sets, and the term "hardware" as used herein in any embodiment described herein can include, for example, assemblies, hardwired circuits, programmable circuits, state machine circuits, and / or firmware storing instructions for execution by programmable circuits, individually or in any combination. Modules can be collectively or individually embodied as circuits that form part of a larger system, such as, but not limited to, integrated circuits (ICs), systems on chips (SoCs), assemblies, and the like.

[0029] A "data collection request" as used herein refers to a message or command sent by a UE to a base station indicating that the UE is looking for measurement resources for AI / ML model training. Some examples of "data collection requests" are radio resource control (RRC) messages generated when certain pre-defined conditions are met.

[0030] A "resource configuration" as used herein refers to a set of parameters or instructions provided by a base station that enable a UE to collect specific data used in AI / ML model training. Some examples of "resource configurations" are channel state information reference signal (CSI-RS) allocations, time intervals for measurements, or beam measurement specifications that a UE follows when collecting training data.

[0031] A "condition ID" (also referred to as "associated ID") as used herein refers to an identifier conveyed by the network to represent one or more additional conditions under which an AI / ML model should be trained or applied. Some examples of "condition IDs" are beam patterns or codebooks associated with a particular cell, cell identifiers that differentiate different coverage areas, or timing constraints for data collection.

[0032] A "NW-side additional condition" as used herein refers to an aspect that is assumed for AI / ML model training and inference but is not directly part of the UE's inherent capabilities. Some examples of "NW-side additional conditions" are specific beam configurations, codebooks, or timestamp constraints that the network requires the UE to consider during measurements and model training.

[0033] Figure 1 A transmitting device or a receiving device in a communication system according to an embodiment is shown.

[0034] Reference Figure 1, the device 100 can be used as a UE, such as a client device, or as a base station (gNB). The device 100 includes a controller module 101 (e.g., one or more processors), a storage module 102, and an antenna module 103 for performing the AI / ML related processes and signaling described herein.

[0035] The controller module 101 performs the primary processing tasks and manages device operations. It can include processors dedicated to specific tasks, such as digital signal processing (DSP), for handling signal conditioning, demodulation, synchronization, equalization, and other complex signal processing functions. These functions can be used to train and infer AI / ML models related to beam management. The DSP can employ advanced computational techniques, such as fast Fourier transforms (FFT), inverse FFT (IFFT), and digital filtering, to ensure the reliability and accuracy of signals for AI / ML training and inference processing.

[0036] In addition, the controller module 101 can include an application processor (AP) that executes software applications, including web browsing, media playback, and other interactive applications. Such processing capabilities support advanced functions that can be enhanced using AI / ML insights for user experience and NW efficiency.

[0037] The storage module 102 can include temporary or non-temporary memory for storing executable instructions, AI / ML models, training data sets, and various data required for NW condition verification processes. The stored instructions can include instructions necessary to perform the processes described herein, such as requesting and verifying associated IDs, managing measurement resources, and processing data collection during handover. The storage module 102 can also contain a communication protocol stack, including layers such as physical (PHY), medium access control (MAC), radio link control (RLC), packet data convergence protocol (PDCP), and radio resource control (RRC).

[0038] The antenna module 103 can include one or more antennas responsible for transmitting and receiving wireless signals between the device 100 and other NW components, such as UEs or gNBs. The antenna module 103 can receive wireless signals from a base station, convert these signals into electrical signals suitable for processing, and transmit data from the device to other nodes within the NW.

[0039] Embodiments disclosed herein provide a signaling and procedural framework for facilitating efficient data collection for training AI / ML models within a wireless communication system. Some embodiments can use an associated ID that the UE uses to request data collection when certain pre-defined conditions are met. Furthermore, to maintain the integrity and relevance of these NW-side additional conditions, various validation mechanisms are proposed, including the implementation of a refresh timer, a validity label, and a timestamp or token. These mechanisms enable the UE to validate the validity of the NW-side conditions and manage the associated ID.

[0040] Various embodiments provide different signaling methods between the base station (gNB) and the UE, allowing the UE to request initiation or termination of data collection configuration based on the identified conditions. Procedures are also provided by which the gNB communicates validation-related information to the UE, which can allow the UE to reset or release the associated ID when the validation conditions have changed. Furthermore, the UE can be equipped with a specialized procedure to determine the applicable AI / ML model or function by interpreting the received gNB configuration along with the associated ID and validity indicator.

[0041] Furthermore, some solutions enhance the robustness of AI / ML operations during handover between cells or base stations. Specifically, the source gNB can communicate the current resource configuration and data collection requirements to the target gNB during handover. The target gNB can maintain the existing resource configuration or propose a new resource configuration, allowing the UE to continue or reset its ongoing model training after receiving the handover command message.

[0042] The present disclosure is related to AI / ML models located at the UE side, where these models can be trained directly within the device or on a server associated with the UE vendor. Beam management includes two main sub-cases: spatial domain downlink beam prediction, where the prediction of one set of beams (set A) is based on measurements from another set of beams (set B); and temporal downlink beam prediction, where the prediction of set A beams is derived from historical measurements of set B beams.

[0043] Depending on the specific deployment strategy and the object of operation, beam configurations and patterns can vary significantly between different cells. Therefore, AI / ML models should be trained on data collected under the same or closely related NW configuration as the one used during inference to ensure consistent and accurate performance.

[0044] Figure 2 An exemplary NW environment depicting UE-side AI / ML model training with associated NW-side conditions is shown in accordance with an embodiment.

[0045] Reference Figure 2The model 201 is trained with data specific to the cell 201A, which utilizes a different beam configuration. Conversely, the model 202 is trained with data from the cell 202B, which represents a different beam configuration. This difference in the models is necessary because optimal AI / ML model performance relies on consistency between the training conditions (beam codebook, indices, and mapping of Set A and Set B) and the subsequent inference conditions within each respective cell.

[0046] To ensure the relevance and applicability of AI / ML models trained under varying conditions, one or more embodiments disclosed herein define NW-side additional conditions that supplement the UE capabilities. The NW can provide such additional conditions via an associated condition ID supported by the base station (gNB), enabling the UE to accurately identify and utilize the applicable AI / ML functionality associated with each specific condition ID. Synchronization between the UE and gNB regarding these condition IDs ensures that both parties share a common understanding of the associated conditions.

[0047] Furthermore, a comprehensive performance monitoring procedure for improving the lifecycle management (LCM) of AI / ML models can be used. This includes two measurement approaches: the first (Type 1) involves the direct reporting of actual prediction results, such as Layer 1 Reference Signal Received Power (L1-RSRP) and beam identifiers, whose data size varies depending on the measurement range. The second (Type 2) involves performance metrics or event-driven management decisions, where the UE calculates specific metrics and communicates these metrics to the gNB, which can include a relatively small amount of data and support a dynamic decision-making process regarding model selection, activation, or deactivation.

[0048] The embodiments described herein address the challenge of maintaining consistency between AI / ML model training and inference by utilizing NW-side additional conditions represented by condition IDs. These condition IDs can be globally unique or specific to a single cell or base station (gNB). In some systems, there can be uncertainty regarding which NW entity allocates these condition IDs and how these IDs are precisely linked to the corresponding NW-side additional conditions. Furthermore, the manner in which the UE obtains the necessary dataset for model training tailored to a specific cell or gNB should be clearly defined.

[0049] To address these issues, detailed procedures and signaling mechanisms are provided. Specifically, a structured data collection procedure is proposed based on the use of condition IDs. When the UE requests data collection for training purposes, the gNB responds by providing a resource configuration along with a condition ID. The UE can retain these condition IDs linked to the trained model even after releasing the RRC connection with a specific cell.

[0050] Further, to address the change of the NW-side additional conditions, three verification methods are proposed. These methods include the use of a refresh timer, validity label, and timestamp or token. Other verification methods can also be included. Such verification mechanisms ensure the ongoing accuracy and applicability of the trained model by allowing the UE to dynamically confirm whether the NW-side conditions remain valid or need to be updated.

[0051] Further, to assist the UE in identifying applicable functions based on its AI / ML capabilities, the gNB can provide the relevant condition IDs. The UE determines the appropriate function and applicability-related information upon receiving these IDs. A value label or timestamp can additionally be employed to maintain the validity of the NW-side additional conditions.

[0052] Further, procedures and signaling mechanisms designed to support data collection continuity during handover events are also detailed. In these scenarios, the source gNB communicates the current resource configuration and associated data collection request information to the target gNB. The target gNB can choose to maintain the original resource configuration or assign a different resource configuration. The UE adjusts its model training process accordingly upon receiving the handover command, seamlessly continuing or initiating model reset, thereby ensuring consistent and reliable performance across handovers.

[0053] The embodiments described herein provide mechanisms to ensure consistency between AI / ML model training and inference by enabling the NW to provide additional conditions. These NW-side additional conditions are communicated via condition IDs, which can be globally unique, cell-specific, or specific to a particular base station (gNB). Based on these condition IDs, the UE can effectively determine the applicable AI / ML model trained under the corresponding NW configuration.

[0054] Figure 3 Signaling exchange between a UE and a gNB for data collection is shown in accordance with an embodiment.

[0055] Reference Figure 3 In step 301, the UE initiates the procedure by sending a data collection request to the gNB. This initiation can be allowed or facilitated by NW signaling such as system information block (SIB) or dedicated RRC signaling. Within this request, the UE can specify a preferred time interval or configuration, which can be indicated by the number of measurement occasions. Multiple scenarios can trigger this data collection request, including lack of a trained model for the current cell, difference between the provided condition ID and the existing trained model, updated condition ID due to change of NW configuration, performance-based trigger (e.g., performance monitoring values such as accuracy, link quality, beam prediction reliability of the AI / ML model), exceeding the valid training interval threshold of the model (e.g., threshold time after model training expires), or requirement during handover procedure (e.g., handover event).

[0056] In response to the UE’s request, in step 302, the gNB provides resource configuration for performing measurements along with an associated condition ID (condition ID_1). The configuration provided by the gNB can include CSI-RS for measuring set B beams and can also include CSI-RS that provides ground truth data for set A beams.

[0057] Subsequently, in step 303, the UE performs model training using the data sets acquired from the resources configured from the gNB. During ongoing data collection and training, the NW configuration can change, prompting the gNB to send an updated data collection response in step 304 that includes new resource configuration and a new or modified condition ID (condition ID_2). Upon receiving the updated configuration, the UE continues AI / ML model training in step 305 with the newly provided measurement data sets.

[0058] Once the model training is complete, the UE retains the condition ID linked to the trained model and the associated value labels even after the data collection session and RRC connection with the cell have ended. These identifiers are stored along with the specific cell identity (physical cell ID or global cell ID) to facilitate future applicability verification.

[0059] Upon completion of the required training, the UE initiates termination of the data collection procedure in step 306 by sending a request to the gNB to stop data collection. In response, the gNB acknowledges the request and releases the previously allocated measurement resources by sending a corresponding data collection response in step 307.

[0060] Recognizing that the UE retains the condition ID even after transitioning to idle or inactive mode (e.g., such that the UE has released the dedicated configuration provided in connected mode) or performing handover and cell reselection, various embodiments also address the validity of the additional conditions on the NW side. To avoid prolonged and unnecessary storage of outdated models, several methods are proposed to ensure that the UE’s stored model and associated condition ID remain relevant and aligned with the current NW configuration.

[0061] According to an embodiment, a method involves the use of a refresh timer that enables the NW to indicate to the UE when to discard the trained model associated with a particular condition ID. Such a refresh timer can be condition ID specific, cell specific, or PLMN specific. In the case of a condition ID specific timer, the UE starts this timer upon completion of the model training or immediately after requesting the release of the previously allocated resources for data collection. Alternatively, a cell specific timer can be conveyed via a SIB or by dedicated RRC signaling, prompting the UE to start the timer upon reception of this information. Furthermore, a PLMN specific timer can be delivered upon the UE establishing a connection with the NW or can be pre-determined within the system specification, thus starting whenever the UE connects to the corresponding NW or when the timer value is available.

[0062] According to another embodiment, the use of value tags can be used. In this case, along with the condition ID, the gNB provides a value tag, which can be condition ID specific, cell specific, or PLMN specific. These value tags sent via a SIB or dedicated RRC signaling allow the UE to detect changes in the additional conditions on the NW side. If the recently received value tag is different from the UE's stored value tag, the UE can determine that the associated additional conditions on the NW side have changed and thus initiate a new data collection procedure.

[0063] According to another embodiment, the use of timestamps or tokens is also introduced as a mechanism to keep the stored condition IDs valid. If the corresponding additional conditions on the NW side have been updated since the model training, the gNB can configure the UE to discard the condition ID. To facilitate this, the UE can record the actual timestamp of the moment of model training or the token provided by the gNB during the initial resource allocation configuration for data collection. This recorded timestamp or token can be used as a reference, enabling the UE and the NW to verify the continued applicability of the trained model with respect to the current NW configuration.

[0064] According to an embodiment, a method is provided to determine the applicable functionality based on the condition ID. In this case, upon establishing the RRC connection, the base station (gNB) can indicate the supported additional conditions on the NW side by sending a list of condition IDs to the UE. This initial communication step enables the UE to identify the functionality related to model inference that aligns with the provided condition IDs.

[0065] Figure 4 A signaling exchange between a UE and a gNB for configuring and / or activating the applicable functionality according to an embodiment is shown.

[0066] Reference Figure 4The procedure starts with the gNB sending a list of condition IDs to the UE in step 401. This step can happen separately or jointly with the actual configuration of the AI / ML functions. Upon receiving these condition IDs, the UE evaluates and identifies the applicable functions corresponding to the model trained under the indicated NW-side additional conditions in step 402. In this evaluation, the UE can compare the newly received value labels with the previously stored value labels if the value labels are sent with the condition IDs through SIB or dedicated RRC signaling. In addition, the UE can alternatively report the applicability related information such as the candidate beams for set A or set B, the boundary conditions or other related context parameters instead of identifying the functions.

[0067] Subsequently, the gNB evaluates the functions or the applicability related information indicated by the UE to configure the applicable functions in step 403. After configuring these functions, the gNB activates one or more functions for actual use in model inference in step 404. Once activated, the UE performs model inference based on the activated functions using the AI / ML model aligned with the configured condition IDs and value labels in step 405. Then, the UE reports the inference results back to the gNB in step 406, enabling the NW to evaluate and further optimize the effectiveness of the activated AI / ML functions.

[0068] According to embodiments, a detailed procedure is provided to manage data collection and AI / ML model training continuity during handover scenarios. When a UE participating in model training initiates handover to another cell, the UE can pause or continue data collection based on the similarity or difference of the NW-side additional conditions between the source cell and the target cell.

[0069] Figure 5 Signaling exchange for an exemplary gNB handover scenario according to embodiments is shown.

[0070] Reference Figure 5 In step 501, the UE sends a data collection request to the source gNB. In step 502, the source gNB responds by providing a resource configuration along with the associated condition ID (condition ID_1). The UE then performs model training using the data collected according to this provided resource configuration in step 503.

[0071] If handover becomes necessary during ongoing training, the continuity of data collection can depend on the alignment of the NW-side additional conditions between the source and target cell. If the configuration in the target cell matches (or is within a predefined range) the configuration of the source cell, the UE can seamlessly continue model training based on the resource configuration provided by the target cell. To facilitate this, the source gNB should include the current data collection resource configuration and request information initially provided by the UE within the handover request message sent to the target gNB in step 504. Then, in step 505, the target gNB assesses this information and communicates its decision to the source gNB via the handover response message, which indicates the same or different resource configuration and possibly a new condition ID (condition ID_2).

[0072] Due to the cell-specific nature of the condition ID, the target gNB can issue a different condition ID and associated value tag even if the NW-side additional conditions remain consistent. Thus, if data collection spans multiple cells, the UE can associate multiple condition IDs and corresponding value tags with a single trained model and store.

[0073] If the NW-side additional conditions differ between the source and target cell, the source gNB or the target gNB can indicate whether the existing data collection resources should be released and new resources allocated. After handover, the UE can initiate training of a new AI / ML model with data collected in the target cell, thereby reflecting the changed conditions. If the source gNB does not provide the target gNB with the UE’s original data collection request information, the UE can itself directly initiate a new data collection request to the target gNB after completing handover.

[0074] In step 506, the source gNB sends a handover message (reconfiguration with synchronization information) to the UE detailing the final resource configuration and related condition ID of the target cell. After successful handover, in step 507, the UE continues model training with the newly provided resources or resets the model training process based on the extent of the configuration change. If the target cell configuration closely aligns with the source cell configuration, the UE can choose to continue training the existing model, associating both the source cell condition ID and the target cell condition ID. Alternatively, a substantial difference in resource configuration can prompt the UE to reset model training.

[0075] Furthermore, some embodiments can involve incomplete signaling or procedures. For example, data collection can be initiated by the NW without an explicit UE request. Furthermore, the NW can explicitly release the associated condition ID without sending corresponding validity information such as a timer, tag, or timestamp.

[0076] Figure 6A is a flowchart illustrating a method for beam management model training performed by a UE according to an embodiment.

[0077] Referring to Figure 6A In step 601A, the UE receives a message from a base station (gNB) corresponding to a data collection request for the UE (e.g., a message enabling the UE to initiate a data collection request). The message can be provided via a SIB or by dedicated RRC signaling and can serve as a trigger for the UE to start its decision process on whether data collection is needed. Upon receiving the message, the UE can determine that data collection is appropriate based on conditions such as lack of a trained model, expiration of a training interval, mismatch between stored value labels and received value labels, or pending handover.

[0078] In step 602A, the UE sends a data collection request to the base station, which can include a preferred configuration for measurements or a desired time interval. The request can provide context to the NW for allocating appropriate measurement resources tailored to the UE’s configuration or state.

[0079] In response to the request, in step 603A, the UE receives a response message to enable UE data collection including a resource configuration and a condition identifier. The resource configuration can define specific measurement parameters such as CSI-RS or other reference signals, while the condition identifier can identify a set of NW-side additional conditions related to the AI / ML model.

[0080] In step 604A, the UE performs data collection by the measurement resource configuration. The data collected under these parameters can be stored and used to train or update an AI / ML model supporting beam prediction or other processing tasks.

[0081] Figure 6B is a flowchart illustrating a method for supporting beam management model training at a UE performed by a base station according to an embodiment.

[0082] Referring to Figure 6B In step 601B, the base station sends a message to the UE corresponding to a data collection request for the UE (e.g., a message enabling the UE to initiate a data collection request). The message can be provided via a SIB or by dedicated RRC signaling and can define the context in which the UE can initiate data collection. The base station can broadcast or unicast the message as part of a larger configuration strategy for managing AI / ML-based functionality.

[0083] In step 602B, the base station receives a data collection request from the UE, which can include a preferred configuration for the measurements or a desired time interval. The request can provide context about the UE’s configuration or current operating state, such as a lack of a previously trained model, a training interval expiration, or other relevant performance-based conditions.

[0084] In step 603B, the base station sends a response message to enable UE data collection including a resource configuration and a condition identifier. The resource configuration can specify one or more measurement parameters that the UE should use to collect data, such as CSI-RS or other reference signals. The condition identifier can reference a set of NW-side additional conditions that apply to the AI / ML model and can ensure consistency between training and inference.

[0085] In step 604B, the base station enables the UE to perform data collection by allowing the UE to measure the resource configuration (providing information needed for the UE to measure the specified resources). The UE can then use the collected data to train or update the AI / ML model. The base station can subsequently receive a release message from the UE indicating that the training is complete, at which point it can deallocate the measurement resources or update the condition identifier’s validity accordingly.

[0086] Figure 7 is a block diagram of an electronic device in a NW environment according to an embodiment.

[0087] Figure 7 is a block diagram of an electronic device in a NW environment 700 according to an embodiment.

[0088] Reference Figure 7The electronic device 701 in the NW environment 700 can communicate with the electronic device 702 via a first NW 798 (e.g., a short-range wireless communication NW) or the electronic device 704 or the server 708 via a second NW 799 (e.g., a long-range wireless communication NW). The electronic device 701 can communicate with the electronic device 704 via the server 708. The electronic device 701 can include a processor 720, a memory 730, an input device 750, a sound output device 755, a display device 760, an audio module 770, a sensor module 776, an interface 777, a haptic module 779, a camera module 780, a power management module 788, a battery 789, a communication module 790, a subscriber identification module (SIM) 796, or an antenna module 797. In one embodiment, at least one (e.g., the display device 760 or the camera module 780) of the components can be omitted from the electronic device 701, or one or more other components can be added in the electronic device 701. Some of the components can be implemented as single integrated circuit (IC) or multiple ICs. The sensor module 776 (e.g., a fingerprint sensor, an iris sensor, or an illuminance sensor) can be embedded in the display device 760 (e.g., a display).

[0089] The processor 720 can execute software (e.g., a program 740) to control at least one other component (e.g., a hardware or software component) of the electronic device 701 coupled with the processor 720 and can perform various data processing or computation. The program 740 can be stored in the memory 730 or transmitted from or through the electronic device 701. The program 740 can include a kernel 741, middleware 742, an application programming interface (API) 743, or an application 744.

[0090] As at least a part of the data processing or computation, the processor 720 can load a command or data received from another component (e.g., the sensor module 776 or the communication module 790) to the volatile memory 732, process the command or the data stored in the volatile memory 732, and store new data generated by the processing in the non-volatile memory 734. The processor 720 can include a main processor 721 (e.g., a central processing unit (CPU) or an AP) and an auxiliary processor 723 (e.g., a graphics processing unit (GPU), an image signal processor (ISP), a sensor hub processor, or a communication processor (CP)) that is operable independently from, or in conjunction with, the main processor 721. Additionally or alternatively, the auxiliary processor 723 can be adapted to consume less power than the main processor 721 or to perform a specific function. The auxiliary processor 723 can be implemented as separate from or as part of the main processor 721.

[0091] In particular embodiments, the processor 720 is configured to implement processes for requesting or receiving data collection configurations associated with one or more condition identifiers, as described herein. For example, upon detecting a specified condition for AI / ML model training, the processor 720 can orchestrate signaling exchanges via the communication module 790 to transmit a data collection request to a base station, receive a resource configuration, and manage data measurement tasks performed by the sensor module 776 or other components. Further, the memory 730 can store a validity indicator for a condition identifier, allowing the processor 720 to determine whether an associated AI / ML model remains valid or needs to be discarded in response to changes in the NW-side additional conditions.

[0092] Accordingly, the processor 720 and the memory 730 can thereby support storage and execution of AI / ML model training data, model parameters, and condition identifiers. In particular, the main processor 721 or the auxiliary processor 723 can process beam-related measurements collected by the communication module 790 to train or update an AI / ML model. The communication module 790, together with the antenna module 797, can forward a release message to the NW upon completion of training.

[0093] The auxiliary processor 723 can replace the main processor 721 while the main processor 721 is in an inactive (e.g., sleep) state, or control at least some of the functions or states related to at least one component (e.g., the display device 760, the sensor module 776, or the communication module 790) among the components of the electronic device 701, together with the main processor 721 while the main processor 721 is in an active state (e.g., executing an application). The auxiliary processor 723 (e.g., an image signal processor or a communication processor) can be implemented as a part of another component (e.g., the camera module 780 or the communication module 790) functionally related to the auxiliary processor 723.

[0094] The memory 730 can store various data used by at least one component (e.g., the processor 720 or the sensor module 776) of the electronic device 701. The various data can include, for example, software (e.g., the program 740) and input data or output data about a command related thereto. The memory 730 can include the volatile memory 732 or the non-volatile memory 734. The non-volatile memory 734 can include the internal memory 736 and / or the external memory 738.

[0095] The program 740 can be stored in the memory 730 as software, and can include, for example, an operating system (OS) 742, middleware 744, or an application 746.

[0096] The input device 750 can receive a command or data to be used by another component (e.g., the processor 720) of the electronic device 701, from the outside (e.g., a user) of the electronic device 701. The input device 750 can include, for example, a microphone, a mouse, or a keyboard.

[0097] The sound output device 755 can output sound signals to the outside of the electronic device 701. The sound output device 755 can include, for example, a speaker or a receiver. The speaker can be used for general purposes, such as playing multimedia or recording, and the receiver can be used for receiving an incoming call. The receiver can be implemented as separate from, or as part of, the speaker.

[0098] The display device 760 can visually provide information to the outside (e.g., a user) of the electronic device 701. The display device 760 can include, for example, a display, a hologram device, or a projector and a control circuit for controlling a corresponding one of the display, the hologram device, and the projector. The display device 760 can include a touch circuit adapted to detect a touch or a sensor circuit (e.g., a pressure sensor) adapted to measure the intensity of force incurred by the touch.

[0099] The audio module 770 can convert a sound into an electrical signal and vice versa. The audio module 770 can obtain sound data, via a speaker, a microphone, or a communication circuit, or output sound data, via the speaker or a receiver.

[0100] The sensor module 776 can detect an operational state (e.g., power or temperature) of the electronic device 701 or an environmental state (e.g., a state of a user) external to the electronic device 701, and then generate an electrical signal or data value corresponding to the detected state. The sensor module 776 can include, for example, a gesture sensor, a gyro sensor, an atmospheric pressure sensor, a magnetic sensor, an acceleration sensor, a grip sensor, a proximity sensor, a color sensor, an infrared (IR) sensor, a biometric sensor, a temperature sensor, a humidity sensor, or an illuminance sensor.

[0101] The interface 777 can support one or more designated protocols for the electronic device 701 to be coupled with the external electronic device 702 directly (e.g., wiredly) or wirelessly. The interface 777 can include, for example, a high definition multimedia interface (HDMI), a universal serial bus (USB) interface, a secure digital (SD) card interface, or an audio interface.

[0102] The connection terminal 778 can include a connector to which the electronic device 701 can be physically connected with the external electronic device 702. The connection terminal 778 can include, for example, a HDMI connector, a USB connector, a SD card connector, or an audio connector (e.g., a headphone connector).

[0103] The haptic module 779 can convert electrical signal into a mechanical stimulus (e.g., vibration or movement) or electrical stimulus that can be recognized by a user via tactile feeling or kinesthetic sense. The haptic module 779 can include, for example, a motor, a piezoelectric element, or an electrical stimulus.

[0104] The camera module 780 can capture still images or moving images. The camera module 780 can include one or more lenses, image sensors, image signal processors, or flashes. The power management module 788 can manage power supplied to the electronic device 701. The power management module 788 can be implemented as at least a part of, for example, a power management integrated circuit (PMIC).

[0105] The battery 789 can supply power to at least one component of the electronic device 701. The battery 789 can include, for example, a primary cell which is not rechargeable, a secondary cell which is rechargeable, or a fuel cell.

[0106] The communication module 790 can support establishing a direct (e.g., wired) communication channel or a wireless communication channel between the electronic device 701 and an external electronic device (e.g., the electronic device 702, the electronic device 704, or a server 708) and performing communication via the established communication channel. The communication module 790 can include one or more communication processors that are operable independently from the processor 720 (e.g., an AP) and supports a direct (e.g., wired) communication or a wireless communication. The communication module 790 can include a wireless communication module 792 (e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module) or a wired communication module 794 (e.g., a local area NW (LAN) communication module or a power line communication (PLC) module). A corresponding one of these communication modules can communicate with the external electronic device via a first NW 798 (e.g., a short-range communication NW, such as BLUETOOTH, wireless-fidelity (Wi-Fi) direct, or infrared data association (IrDA)) or a second NW 799 (e.g., a long-range communication NW, such as a cellular NW, the Internet, or a computer NW (e.g., LAN or wide area NW (WAN)). These various types of communication modules can be implemented as a single component (e.g., a single IC) or can be implemented as separate components (e.g., separate ICs) from each other. The wireless communication module 792 can identify and authenticate the electronic device 701 in a communication NW, such as the first NW 798 or the second NW 799, using subscriber information (e.g., an international mobile subscriber identity (IMSI)) stored in the subscriber identification module 796. TM

[0107] ​The antenna module 797 can transmit or receive a signal or power to or from an external electronic device (e.g., an external electronic device) of the electronic device 701. The antenna module 797 can include one or more antennas, and can, for example, select at least one antenna appropriate for a communication scheme used in a communication NW, such as the first NW 798 or the second NW 799, from among the one or more antennas, by the communication module 790 (e.g., the wireless communication module 792). Then, the signal or power can be transmitted or received between the communication module 790 and the external electronic device via the selected at least one antenna.

[0108] Commands or data can be transmitted or received between the electronic device 701 and an external electronic device 704 via the server 708 coupled with the second NW 799. Each of the electronic devices 702 and 704 can be a device of a same type as or different from the electronic device 701. All or some of the operations to be performed at the electronic device 701 can be performed at one or more of the external electronic devices 702, 704, or 708. For example, if the electronic device 701 is to automatically perform a function or a service, or to perform a function or a service in response to a request from a user or another device, the electronic device 701 can request one or more external electronic devices to perform at least part of the function or service, instead of, or in addition to, performing the function or service. The one or more external electronic devices receiving the request can perform at least part of the requested function or service, or an additional function or an additional service related to the request, and transmit a result of the performance to the electronic device 701. The electronic device 701 can provide the result with or without further processing of the result, as at least part of a reply to the request. To that end, for example, cloud computing, distributed computing, or client-server computing technology can be used.

[0109] Figure 8 is a system including a UE and a base station gNB in communication with each other according to an embodiment.

[0110] Reference Figure 8 The UE can include a radio 815 and processing circuitry (or means for processing) 820, which can perform various methods disclosed herein, e.g., with reference to the methods illustrated in FIG. 6. For example, the processing circuitry 820 can receive a transmission from a NW node (gNB) 810 via the radio 815, and the processing circuitry 820 can transmit a signal to the gNB 810 via the radio 815. Figure 3

[0111] ​Embodiments of the subject matter and operations described in this specification can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a computer storage medium for execution by, or to control the operation of, data processing apparatus. Additionally or alternatively, the program instructions can be encoded on an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus. The computer storage medium can be, or include, a computer-readable storage device, a computer-readable storage substrate, a random or serial access memory array or device, or a combination of one or more of them. Moreover, while a computer storage medium is not a propagated signal, a computer storage medium can be a source or destination of computer program instructions encoded in an artificially generated propagated signal. The computer storage medium can also be, or include, one or more separate physical components or media (e.g., multiple CDs, disks, or other storage devices) within or across one or more devices (e.g., within a single computer or across multiple computers). Furthermore, the operations described in this specification can be implemented as operations performed by a data processing apparatus on data stored on one or more computer-readable storage devices or received from other sources.

[0112] While this specification contains many specific implementation details, these should not be construed as limitations on the scope of any claims, but rather as descriptions of features specific to particular embodiments. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features can be described above as acting in particular combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination and the claimed combination can be directed to a subcombination or variation of a subcombination.

[0113] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring such order, nor that all illustrated operations be performed, to implement and / or benefit from the embodiments. Rather, the order of the operations can be modified in some implementations. In certain circumstances, multitasking and parallel processing can be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated in a single software product or packaged into multiple software products.

[0114] Thus, particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, acts recited in the claims can be performed in a different order and still achieve desirable results. Additionally, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0115] As will be recognized by those skilled in the art, the innovative concepts described in this document can be modified and varied widely. Therefore, the scope of the claimed subject matter is not to be limited to any of the specific exemplary teachings discussed above, but is instead being defined only by the following claims — the legal scope of which can be broadened by the court.

Claims

1. A method performed by a user equipment (UE) in a wireless communication system, the method comprising: Receive a message from the base station corresponding to the data collection request of the UE; Send the data collection request to the base station, which has at least one of a preference configuration or a time interval; The base station receives a response for enabling UE data collection in response to the data collection request, the response including resource configuration and condition identifier; as well as Data collection is performed by measuring the resource configuration.

2. The method according to claim 1, wherein, The data collection request is sent to the base station based on at least one of the following: The lack of trained models The received value label does not match the previously stored value label. Performance monitoring values, The threshold time after model training expires, or Switching events.

3. The method according to claim 1, further comprising: The condition identifier is stored, and in response to the condition identifier being configured, validity information associated with the condition identifier is stored when the UE is in idle mode, in inactive mode, performing handover, or performing cell reselection.

4. The method according to claim 1, wherein, The resource configuration includes at least one reference signal for collecting beam measurements or validity information associated with the condition identifier.

5. The method of claim 1, further comprising sending a message to the base station to request the cessation of data collection, and releasing the resource configuration upon completion of model training.

6. The method of claim 1, further comprising discarding or refreshing the trained model associated with the at least one condition identifier when a determination is made that the network-side additional condition is invalid.

7. The method according to claim 6, wherein, The determination is based on at least one of a refresh timer, a value tag, or a timestamp.

8. The method according to claim 1, further comprising: Perform a handover from the base station to another base station; as well as Receive an indication from the other base station as to whether the resource configuration has been changed.

9. The method according to claim 8, wherein, When the resource configuration remains unchanged, multiple condition identifiers are associated with the trained model.

10. A method performed by a base station in a wireless communication system, the method comprising: Send a message corresponding to the data collection request of the user equipment (UE); The UE receives the data collection request having at least one of a preference configuration or a time interval; In response to the data collection request, a response for enabling UE data collection is sent to the UE, the response including resource configuration and condition identifier; as well as Enable data collection at the UE by measuring the resource configuration.

11. The method according to claim 10, wherein, The data collection request is received from the UE based on at least one of the following: The lack of trained models The received value label does not match the previously stored value label. Performance monitoring values, The threshold time after model training expires, or Switching events.

12. The method of claim 10, further comprising: In response to the condition identifier being configured for the UE, validity information associated with the condition identifier is provided when the UE is in idle mode, in inactive mode, performing handover, or performing cell reselection.

13. The method according to claim 10, wherein, The resource configuration includes at least one reference signal for collecting beam measurements or validity information associated with the condition identifier at the UE.

14. The method of claim 10, further comprising: The system receives a message from the UE requesting the cessation of data collection and releases the resource configuration upon completion of model training performed by the UE.

15. The method of claim 10, further comprising: When a determination is made that the network-side additional conditions are invalid, the system indicates whether the UE should discard or refresh the trained model associated with the at least one condition identifier.

16. The method according to claim 15, wherein, The determination is based on at least one of a refresh timer, a value tag, or a timestamp.

17. The method of claim 10, further comprising: Perform a handover from the base station to another base station; as well as Send an indication to the UE as to whether the resource configuration has been changed.

18. The method according to claim 17, wherein, When the resource configuration remains unchanged, multiple condition identifiers are associated with the trained model.

19. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a user equipment (UE) in a wireless communication system, cause the UE to: Receive a message from the base station corresponding to the data collection request of the UE; Send the data collection request to the base station, which has at least one of a preference configuration or a time interval; The base station receives a response for enabling UE data collection in response to the data collection request, the response including resource configuration and condition identifier; as well as Data collection is performed by measuring the resource configuration.

20. The non-transitory computer-readable medium according to claim 19, wherein, The data collection request is sent to the base station based on at least one of the following: The lack of trained models The received value label does not match the previously stored value label. Performance monitoring values, The threshold time after model training expires, or Switching events.

Images