Advanced gNB-UE model monitoring method

Abstract

The present disclosure describes a method of configuring AI / ML-based model monitoring and signaling activation / deactivation of model monitoring in a wireless mobile communication system that includes a base station (e.g., gNB) and a mobile station (e.g., UE). When applying the AI / ML model to a radio access network, the model performance is monitored (such as reasoning) such that any potential drift occurrences can be detected to ensure the desired communication quality. Depending on the applied model characteristics and channel environment, the model monitors a monitoring cycle configured to support changes associated with different modes of drift occurrences in the model life cycle.

Description

Technical Field

[0001] This disclosure relates to AI (artificial intelligence) / ML (machine learning) based model monitoring, wherein techniques are proposed for configuring and signaling specific information to avoid model performance degradation. Background Art

[0002] In December 2021, the 3GPP (Third Generation Partnership Project) approved its Release-18 technology package, and one of the selected research items was AI / ML (artificial intelligence / machine learning), as described in the relevant document (RP-213599) discussed at 3GPP TSG (Technical Specification Group) RAN (Radio Access Network) meeting #94e. The AI / ML research item is officially titled "Study on AI / ML for NR Air Interface," and RAN Working Group 1 (WG1) and WG2 are currently actively developing specifications. The goal of this research item is to define a common AI / ML framework and identify areas where AI / ML-based technologies can benefit from using use cases. From a radio access network perspective, one of the key areas of scope is to identify various levels of collaboration between base stations (BS / gNB) and user equipment (UE).

[0003] According to 3GPP, the main goal of this research project is to study the AI / ML framework of the air interface for target use cases by considering performance, complexity, and potential regulatory impact. In particular, one of the key work areas will be the AI / ML model, terminology, and description used to determine the common and specific characteristics of the framework. For the selected target use cases, two issues will be included, such as performance benefit evaluation and potential regulatory impact assessment. Regarding the AI / ML framework, various aspects are considered for research, and one of the key projects is about the lifecycle management of AI / ML models, which includes multiple necessary stages such as model training, model deployment, model inference, model monitoring, and model updates.

[0004] One of the key agreements in the 3GPP TSG RAN WG1 #110bis meeting was to research multiple metrics and methods for AI / ML model monitoring in lifecycle management, targeting each use case. AI / ML model monitoring was identified as a key issue because drift often occurs during AI / ML model operation, as model performance degrades over time. Once drift is detected through model monitoring, countermeasures are needed to compensate for the degradation or maintain the required KPIs.

[0005] On the other hand, in 3GPP, the terminology of the work list contains a set of high-level descriptions about AI / ML model training, inference, verification, testing, UE-side model, network-side model, single-sided model, dual-sided model, etc. The UE-side model and network-side model indicate that the AI / ML model operates on the UE side and the network side, respectively. In a similar context, the single-sided model and dual-sided model indicate that the AI / ML model is located on one side and on both sides, respectively. As the definition of terms is still under discussion and subject to further revisions, not all signaling aspects to support the above items have been specified at this time. Any potential impact on the standard using new or enhanced mechanisms to support AI / ML models using the above work list items is one of the key areas of investigation in the AI / ML research project.

[0006] The present invention solves the problem of how to establish an efficient model monitoring mechanism when operating AI / ML models in radio access networks.

[0007] Currently, there is no mechanism that allows the UE to shorten the model monitoring period without relevant specifications. Therefore, the UE will continue to monitor the model operation, which increases power consumption.

[0008] WO 2021 / 044192 discloses a method and system for detecting and / or predicting data drift, for example in a distributed cloud.

[0009] US 2021 / 0256310 describes a machine learning platform that deploys a machine learning model and monitors the performance of the machine learning model after deployment.

[0010] US 2019 / 0147371 describes a device that identifies training data and scoring data for a model and removes bias from the training data to generate unbiased training data.

[0011] WO 2022 / 182272 provides a network analysis method and apparatus that can identify unexpected behavior of components in a communication network without having to measure performance at the individual component level.

[0012] WO 2022 / 008037 relates to the ability (inability) of a UE to execute and / or train an ML model, and network-initiated triggering of execution and / or training of an ML model in view of the ability (inability) of the UE to execute and / or train an ML model.

[0013] US 2020 / 0151619 describes the impact of concept drift when choosing machine learning training methods to address the identified effects. Summary of the Invention

[0014] This solution is provided by activating model monitoring configured by the gNB, with the UE performing model monitoring based on the configured monitoring period. The UE only monitors the model based on the configured period, thus reducing device power consumption. AI / ML model monitoring is a critical issue because drift often occurs during AI / ML model operation due to degradation in model performance over time.

[0015] According to a first aspect, the present disclosure relates to a method for activating model monitoring configured by a gNB, wherein a UE performs model monitoring based on a configured monitoring period, wherein the gNB sends model monitoring configuration information to the UE through RRC signaling, and the gNB reconfigures the model monitoring period of the UE through an RRC message based on an AI / ML model operating between the gNB and the UE.

[0016] In some embodiments of the method according to the first aspect, the method is characterized in that, for a UE group having the same AI / ML model, the configured model monitoring is applied to multiple UEs in the group.

[0017] In some embodiments of the method according to the first aspect, the method is characterized in that the gNB enables / disables model monitoring of the UE.

[0018] In some embodiments of the method according to the first aspect, the method is characterized in that an indication message for enabling / disabling pattern monitoring is sent for (multiple) UEs via PDCCH / MAC CE (UE specific) and system information (to all UEs).

[0019] In some embodiments of the method according to the first aspect, the method is characterized in that, when the gNB determines model retraining or model switching during the model monitoring operation, an indication message is sent to the UE to stop or temporarily disable model monitoring while reconfiguring the model monitoring for updating.

[0020] In some embodiments of the method according to the first aspect, the method is characterized in that the gNB can configure different monitoring configurations according to the RRC state, such that:

[0021] In the RRC_CONNECTED state, the gNB can configure monitoring information via dedicated RRC messages.

[0022] In the RRC_INACTIVE state, the gNB can configure monitoring information through system information messages.

[0023] In the RRC_IDLE state, the gNB can configure monitoring information through the RRC Release message.

[0024] In some embodiments of the method according to the first aspect, the method is characterized in that both the one-sided model and the two-sided model can apply the model monitoring signaling and the associated configuration information.

[0025] In some embodiments of the method according to the first aspect, the method is characterized in that, in one of the model deployment scenarios, the UE determines to activate or deactivate model monitoring based on assistance information from the gNB.

[0026] According to a second aspect, the present disclosure relates to a wireless device comprising at least one memory and at least one processor configured to perform the method according to any one of the embodiments of the first aspect.

[0027] According to a third aspect, the present disclosure relates to a user equipment (UE), which comprises a wireless device according to any one of the embodiments of the present disclosure.

[0028] According to a fourth aspect, the present disclosure relates to a base station BS, comprising at least one memory and at least one processor, wherein the at least one processor is configured to execute the method according to any one of the embodiments of the first aspect.

[0029] According to a fifth aspect, the present disclosure relates to a wireless communication system, which includes at least one base station as described in any one of the embodiments of the present disclosure and at least one user equipment as described in any one of the embodiments of the present disclosure.

[0030] According to a sixth aspect, the present disclosure relates to a computer program product comprising instructions that, when executed by at least one processor, configure the at least one processor to perform the method according to the first aspect, and configure the at least one processor to perform the method for exchanging data according to any one of the embodiments of the present disclosure. The computer program product may use any programming language and may be in the form of source code, object code, or any intermediate code between source code and object code, such as a partially compiled form, or any other desired form.

[0031] According to a sixth aspect, the present disclosure relates to a computer-readable storage medium comprising instructions which, when executed by at least one processor, configure the at least one processor to perform a method according to any one of the embodiments of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A flow chart showing gNB behavior for model monitoring configuration and associated downlink signaling.

[0033] Figure 2 A flow chart showing UE behavior for model monitoring operation and reception of a model monitoring configuration is shown.

[0034] Figure 3 The signaling flow of gNB-UE communication for model monitoring operation is shown. DETAILED DESCRIPTION

[0035] The detailed description set forth below with reference to the accompanying drawings is intended as a description of various configurations and is not intended to represent the only configuration in which the concepts described herein may be practiced. The detailed description includes specific details to provide a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts may be practiced without these specific details. In particular, although terminology from 3GPP 5G NR may be used in this disclosure to illustrate the embodiments herein, this should not be considered as limiting the scope of the invention.

[0036] Some embodiments contemplated herein will now be described more fully with reference to the accompanying drawings. However, other embodiments are also within the scope of the subject matter disclosed herein, and the disclosed subject matter should not be construed as being limited to the embodiments set forth herein; rather, these embodiments are provided by way of example to convey the scope of the subject matter to those skilled in the art.

[0037] Generally, all terms used herein should be interpreted according to their ordinary meaning in the relevant technical field, unless a different meaning is clearly given and / or a different meaning is implied from the context of its use. Unless otherwise expressly stated, all references to one / a kind / this element, device, part, mode, step, etc. should be openly interpreted as referring to at least one instance of an element, device, part, mode, step, etc. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless a step is explicitly described as being after or before another step and / or it is implied that a step must be after or before another step. Where appropriate, any feature of any embodiment disclosed herein may be applicable to any other embodiment. Similarly, any advantage of any embodiment may be applicable to any other embodiment, and vice versa. Based on the following description, other purposes, features and advantages of the attached embodiments will become apparent.

[0038] In some embodiments, the more general term "network node" may be used, which may correspond to any type of radio network node or any network node that communicates with a UE (directly or via another node) and / or communicates with another network node. Examples of network nodes are NodeB, MeNB, ENB, a network node belonging to an MCG or SCG, a base station (BS), a multi-standard radio (MSR) radio node (such as an MSR BS, eNodeB, gNodeB), a network controller, a radio network controller (RNC), a base station controller (BSC), a relay, a donor node controlled relay, a base transceiver station (BTS), an access point (AP), a transmission point, a transmission node, an RRU, an RRH, a node in a distributed antenna system (DAS), a core network node (such as a mobile switching center (MSC), a mobility management entity (MME), etc.), operations and maintenance (O&M), an operations support system (OSS), a self-optimizing network (SON), a positioning node (such as an evolved serving mobile positioning center (E-SMLC)), minimization of drive tests (MDT), test equipment (physical node or software), etc.

[0039] In some embodiments, the non-limiting term user equipment (UE) or wireless device may be used and may refer to any type of wireless device that communicates with a network node and / or another UE in a cellular or mobile communication system. Examples of UEs are target devices, device-to-device (D2D) UEs, machine-type UEs or UEs capable of machine-to-machine (M2M) communication, PDAs, PADs, tablet computers, mobile terminals, smartphones, laptop embedded devices (LEEs), laptop mounted equipment (LMEs), USB dongles, M1 category UEs, M2 category UEs, ProSe UEs, V2V UEs, V2X UEs, and the like.

[0040] Furthermore, terms such as base station / gNodeB and UE should be considered non-restrictive and, in particular, do not imply a hierarchical relationship between the two. In general, a "gNodeB" can be considered device 1 and a "UE" can be considered device 2, with the two devices communicating with each other over a radio channel. In the following, a transmitter or receiver can be either a gNodeB (gNB) or a UE.

[0041] As will be appreciated by those skilled in the art, aspects of the embodiments may be embodied as a system, apparatus, method, or program product. Thus, the embodiments may take the form of entirely hardware embodiments, entirely software embodiments (including firmware, resident software, microcode, etc.), or embodiments combining software and hardware aspects.

[0042] For example, the disclosed embodiments may be implemented as hardware circuits comprising custom very large scale integrated ("VLSI") circuits or gate arrays, off-the-shelf semiconductors (e.g., logic chips, transistors, or other discrete components). The disclosed embodiments may also be implemented in programmable hardware devices such as field programmable gate arrays, programmable array logic, programmable logic devices, and the like. As another example, the disclosed embodiments may comprise one or more physical or logical blocks of executable code, which blocks may be organized, for example, as objects, procedures, or functions.

[0043] Furthermore, embodiments may take the form of a program product embodied in one or more computer-readable storage devices storing machine-readable code, computer-readable code, and / or program code (hereinafter referred to as code). The storage device may be tangible, non-transitory, and / or non-transmissive. The storage device may not embody signals. In certain embodiments, the storage device utilizes only signals to access the code.

[0044] Any combination of one or more computer-readable media may be utilized. The computer-readable medium may be a computer-readable storage medium. The computer-readable storage medium may be a storage device that stores code. The storage device may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, holographic, micromechanical, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing.

[0045] More specific examples of storage devices (a non-exhaustive list) would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, random access memory ("RAM"), read-only memory ("ROM"), erasable programmable read-only memory ("EPROM" or flash memory), a portable compact disk read-only memory ("CD-ROM"), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer-readable storage medium may be any tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0046] The code for performing the operations of the embodiment can be any number of lines and can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Python, Ruby, Java, Smalltalk, C++, and conventional procedural programming languages ​​such as the "C" programming language, and / or machine languages ​​such as assembly language. The code can be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network ("LAN"), a wireless LAN ("WLAN"), or a wide area network ("WAN"), or can be connected to an external computer (e.g., via the Internet using an Internet Service Provider ("ISP")).

[0047] Furthermore, the described features, structures, or characteristics of the embodiments may be combined in any suitable manner. In the following description, numerous specific details (e.g., examples of programming, software modules, user selections, network transactions, database queries, database structures, hardware modules, hardware circuits, hardware chips, etc.) are provided to provide a thorough understanding of the embodiments. However, those skilled in the relevant art will recognize that the embodiments may be practiced without one or more of the specific details, or with other methods, components, materials, etc. In other instances, well-known structures, materials, or operations are not shown or described in detail to avoid obscuring aspects of the embodiments. Reference throughout this specification to "one embodiment," "an embodiment," or similar language means that at least one embodiment includes the specific features, structures, or characteristics described in connection with that embodiment. Therefore, unless expressly stated otherwise, the phrases "one embodiment," "an embodiment," and similar language throughout this specification may, but do not necessarily, refer to the same embodiment, but rather to "one or more, but not all, embodiments." Unless expressly stated otherwise, the terms "including," "comprising," "having," and variations thereof mean "including, but not limited to." The enumerated listing of items does not imply that any or all of the items are mutually exclusive unless expressly specified otherwise.The terms "a" and "an" and "the" also mean "one or more" unless expressly specified otherwise.

[0048] Various aspects of the embodiments are described below with reference to schematic flow charts and / or schematic block diagrams of methods, apparatuses, systems, and program products according to the embodiments. It should be understood that each block of the schematic flow charts and / or schematic block diagrams, as well as combinations of blocks in the schematic flow charts and / or schematic block diagrams, can be implemented by code. The code can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to generate a machine, such that instructions executed by the processor of the computer or other programmable data processing device create means for implementing the functions / actions specified in the flow charts and / or block diagrams.

[0049] The code may also be stored in a storage device that can direct a computer, other programmable data processing apparatus, or other device to operate in a specific manner so that the instructions stored in the storage device produce an article of manufacture including instructions for implementing the functions / actions specified in the flowcharts and / or block diagrams.

[0050] The code may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus, or other device to produce a computer-implemented process, such that the code executed on the computer or other programmable apparatus provides a process for implementing the functions / actions specified in the flowcharts and / or block diagrams.

[0051] The flowcharts and / or blocks in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of devices, systems, methods, and program products according to various embodiments. In this regard, each block in the flowcharts and / or block diagrams may represent a module, segment, or portion of code, which includes one or more executable instructions for implementing the specified logical function(s).

[0052] It should also be noted that in some alternative implementations, the functions indicated in the blocks may not occur in the order indicated in the figures. For example, two blocks shown in succession may actually be executed substantially in parallel, or the blocks may sometimes be executed in the reverse order depending on the functions involved. Other steps and methods may be conceived that are equivalent in function, logic, or effect to one or more blocks or portions thereof in the illustrated figures.

[0053] Although various arrow types and line types may be used in the flowcharts and / or block diagrams, it should be understood that they do not limit the scope of the corresponding embodiments. In fact, some arrows or other connectors may be used only to indicate the logical flow of the depicted embodiments. For example, arrows can indicate waiting or monitoring periods of unspecified duration between the enumerated steps of the depicted embodiments. It should also be noted that each block of the block diagrams and / or flowcharts and the combination of blocks in the block diagrams and / or flowcharts can be implemented by a dedicated hardware-based system or a combination of dedicated hardware and code that performs the specified function or action.

[0054] The description of an element in each figure may refer to an element in a subsequent figure. In all figures, the same reference numerals refer to the same elements, including alternative embodiments of the same elements.

[0055] The detailed description set forth below with reference to the accompanying drawings is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein may be practiced. The detailed description includes specific details to provide a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts may be practiced without these specific details. For example, although 3GPP terminology from, for example, 5G NR may be used in this disclosure to illustrate the embodiments herein, this should not be viewed as limiting the scope of this disclosure.

[0056] Generally, all terms used herein are to be interpreted according to their ordinary meaning in the relevant technical field, unless a different meaning is explicitly assigned and / or a different meaning is implied from the context of their use. Unless expressly stated otherwise, all references to an element, device, component, method, step, etc. are to be interpreted as referring to at least one instance of that element, device, component, method, step, etc. Furthermore, the order of steps in any method disclosed herein, and particularly in the accompanying drawings, is for illustrative purposes only and is not intended to limit the present disclosure. The present disclosure may apply to the same steps performed in a different order, and / or to all or part of the steps performed in parallel or in combination, unless a step is explicitly described as preceding or following another step and / or a step is implied to necessarily precede or follow another step. Furthermore, in the accompanying drawings, steps indicated by dashed lines are considered optional with respect to the embodiment depicted in that figure. Where appropriate, any feature of any embodiment disclosed herein may apply to any other embodiment. Similarly, any advantage of any embodiment may apply to any other embodiment, and vice versa. Other objects, features, and advantages of the accompanying embodiments will become apparent from the following description.

[0057] This disclosure relates to a wireless communication system, which may be, for example, a 5G NR wireless communication system. More specifically, it refers to the radio access network (RAN) of the wireless communication system, which is used to exchange data with user equipment (UEs) via radio signals. For example, the RAN may send data (downlink DL), such as data received from a core network (CN), to the UE. The RAN may also receive data (uplink UL) from the UE, which may be forwarded to the CN.

[0058] In the illustrated example, the RAN includes one base station (BS). Of course, the RAN may include more than one BS to increase the coverage of the wireless communication system. Depending on the implemented wireless communication standard(s), each of these BSs may be referred to as a NB, eNodeB (or eNB), gNodeB (or, in the case of a 5G NR wireless communication system, a gNB), access point, etc.

[0059] The UE is located within the coverage of the BS. For example, the coverage of the BS corresponds to an area in which the UE can decode the PDCCH transmitted by the BS.

[0060] An example of a wireless device suitable for implementing any of the methods discussed in this disclosure performed at a UE corresponds to an apparatus that provides a wireless connection to a wireless communication system's radio access network (RAN) and can be used to exchange data with the RAN. Such a wireless device can be included in a UE. For example, a UE can be a cellular phone, a wireless modem, a wireless communication device, a handheld device, a laptop computer, etc. A UE can also be an Internet of Things (IoT) device such as a wireless camera, a smart sensor, a smart meter, smart glasses, a vehicle (manned or unmanned), a Global Positioning System device, etc., or any other device that can run an application that requires exchanging data with a remote recipient via a wireless device.

[0061] The wireless device includes one or more processors and one or more memories. The one or more processors may include, for example, a central processing unit (CPU), a digital signal processor (DSP), a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), etc. The one or more memories may include any type of computer-readable volatile and non-volatile memory (magnetic hard disk, solid-state drive, optical disk, electronic memory, etc.). The one or more memories may store a computer program product in the form of a set of program code instructions, which are executed by the one or more processors to implement all or part of the steps of the method for exchanging data performed on the UE side according to any of the embodiments disclosed herein.

[0062] The wireless device may also include a main radio (MR) unit. The MR unit corresponds to the wireless device's main wireless communication unit and is used to exchange data with a base station of the RAN using radio signals. The MR unit can implement one or more wireless communication protocols and can be, for example, a 3G, 4G, 5G, NR, WiFi, WiMax, or other transceiver. In a preferred embodiment, the MR unit corresponds to a 5G NR wireless communication unit.

[0063] The following explanation provides a detailed description of the mechanisms for AI / ML-based model monitoring operations in radio access networks. AI / ML-based technologies are currently being applied to many different applications, and 3GPP has also begun technical research to apply them to multiple use cases based on the observed potential benefits.

[0064] The AI / ML lifecycle can be divided into several stages, such as data collection / preprocessing, model training, model testing / validation, model deployment / update, model monitoring, etc., each of which is equally important for achieving the target performance with any specific (multiple) models.

[0065] A challenging issue when applying AI / ML models to any use case or application is managing their lifecycle. This is primarily due to data / model drift, which can occur during model deployment and inference, leading to degradation in AI / ML model performance. Essentially, after model deployment, dataset statistics change, and the model's ability to reason with unseen data as input is also affected. Similarly, the statistical properties of the dataset and the relationship between the input and output of a trained model can change with drift.

[0066] To address this issue, AI / ML models require model monitoring after deployment, as model performance can drift and become unsustainable. Updated feedback is then needed to retrain / update the model or select an alternative model. Therefore, it is crucial to address AI / ML data / model drift by tracking model performance (e.g., predictability, accuracy, etc.).

[0067] When deploying wireless communication networks that support AI / ML models, it is important to consider how to handle drift in AI / ML model monitoring during wireless device operations (e.g., model training, inference, updates, etc.). In parallel, another aspect is how to handle the increase in power consumption used to perform model monitoring, which necessitates close detection of degradation in model performance.

[0068] Figure 1A flowchart illustrates gNB behavior when model monitoring is configured and associated downlink signaling is generated for a UE. Activation of model monitoring is configured by the gNB, with the UE performing model monitoring based on the configured monitoring period. Specifically, the gNB configures model monitoring (e.g., a UE-specific model monitoring period) and may use model drift pattern learning as a reference before configuring model monitoring, as drift patterns can vary depending on model characteristics and the application / environment. Model monitoring may then be adapted to the configuration, and a model drift pattern learning phase may be helpful. After model monitoring information is configured, it is sent to the UE, and this transmission may occur using radio resource control (RRC) signaling. Furthermore, when a group of UEs uses the same AI / ML model, the configured model monitoring can be applied to multiple UEs in the group based on configured metrics (such as the same UE priority or other QoS parameters). For model monitoring operation, the gNB can configure different monitoring configurations based on the RRC state. For example, in the RRC_CONNECTED state, the gNB can configure monitoring information via a dedicated RRC message. In the RRC_INACTIVE state, the gNB can configure monitoring information via a system information message. In the RRC_IDLE state, the gNB can configure monitoring information through the RRC Release message.

[0069] Figure 2 This flowchart illustrates UE behavior when receiving configuration information for model monitoring and model operation is monitored. Upon receiving the associated information from the gNB, the UE is signaled to activate the configured model monitoring, and then a UE-specific model monitoring period is applied, thereby reducing power consumption due to monitoring operations. The configured model monitoring information helps the UE reduce device battery power consumption by adjusting the model monitoring period based on different model performance behaviors. The specific model monitoring period used for an application is implementation-specific, and the configuration information may include a set of parameters related to the application and adjustment of the model monitoring period and its characteristics. Alternatively, depending on the model operation scenario, the UE may determine whether to activate or deactivate model monitoring based on assistance information from the gNB. By default, the gNB provides signaling information regarding the activation / disablement of model monitoring via configuration information, and the UE may also overrule the activation or deactivation of model monitoring via the configuration information for model monitoring.

[0070] Figure 3The signaling flow for gNB-UE communication when model monitoring is applied is shown. Model drift pattern learning can be performed between the gNB and UE, allowing model performance to be learned from drift behavior. Different models may have different drift patterns depending on the application and environment. Model drift pattern information facilitates model monitoring for any drift detection or performance degradation events. The gNB configures model monitoring (e.g., a UE-specific model monitoring period), and the configured model monitoring information can be sent via RRC signaling. For example, the gNB can generate a set of model monitoring patterns as index values ​​(e.g., monitoring periods) to command UE behavior based on the configured model monitoring operation. Using model monitoring update reports from the UE, the gNB can reconfigure model monitoring information by adjusting the monitoring period.

[0071] Depending on the state of model operation, the gNB also enables / disables model monitoring for the UE and sends an indication message for this to (multiple) UEs via PDCCH / MAC CE (UE-specific) and system information (to all UEs). Configurable criteria for enabling / disabling model monitoring are signaled by the network. For example, when the gNB determines that model retraining or model switching is required during model monitoring operation, it sends an indication message to the UE to stop or temporarily disable model monitoring while reconfiguring model monitoring for an update.

[0072] In other aspects, both one-sided and two-sided models may apply model monitoring signaling and associated configuration information, where the one-sided model is defined as a network-side model or a UE-side model, and the two-sided model is defined as a paired model with joint model operation between the gNB and the UE.

[0073] This application aims to provide basic mechanisms for AI / ML-enabled interaction and data information flow in radio access network collaboration.

[0074] Based on the proposed application, the gNB-UE behavior for supporting AI / ML operations for wireless communications can be greatly improved in potential scenarios such as two-sided AI / ML models.

Claims

1. A method for monitoring activation patterns configured by a gNB, wherein: The UE performs model monitoring based on the configured monitoring period, wherein: The gNB sends the model monitoring configuration information to the UE via RRC signaling. The gNB reconfigures the UE’s model monitoring period via RRC messages based on the AI / ML model operating between the gNB and the UE.

2. The method according to claim 1, wherein For a UE group with the same AI / ML model, the configured model monitoring is applied to multiple UEs in the group.

3. A method according to any preceding claim, wherein: The gNB enables / disables the UE's model monitoring.

4. A method according to any preceding claim, wherein: An indication message for enabling / disabling pattern monitoring is sent to (multiple) UEs via PDCCH / MAC CE (UE-specific) and system information (to all UEs).

5. A method according to any preceding claim, wherein: When the gNB determines model retraining or model switching during model monitoring operation, it sends an indication message to the UE to stop or temporarily disable model monitoring while reconfiguring model monitoring for updating.

6. A method according to any preceding claim, wherein: The gNB can configure different monitoring configurations based on the RRC state, such that: In the RRC_CONNECTED state, the gNB can configure monitoring information through dedicated RRC messages. In the RRC_INACTIVE state, the gNB can configure monitoring information through system information messages. In the RRC_IDLE state, the gNB can configure monitoring information through the RRC Release message.

7. A method according to any preceding claim, wherein: Both one-sided and two-sided models can apply model monitoring signaling and associated configuration information.

8. A method according to any preceding claim, wherein: In one of the model deployment scenarios, the UE determines whether to activate or deactivate model monitoring based on assistance information from the gNB.

9. A wireless device comprising at least one memory and at least one processor configured to perform the method according to any one of the preceding claims.

10. A user equipment (UE), comprising the wireless device according to claim 9.

11. A base station BS, comprising at least one memory and at least one processor, wherein the at least one processor is configured to execute the method according to any one of claims 1 to 9.

12. A wireless communication system comprising at least one base station according to claim 11 and at least one user equipment according to claim 10.

13. A computer program product comprising instructions which, when executed by at least one processor, configure the at least one processor to perform the method according to any one of claims 1 to 9.

14. A computer-readable storage medium comprising instructions that, when executed by at least one processor, configure the at least one processor to perform the method according to any one of claims 1 to 8.

Images