Functionality-based implicative ML inference parameter

Abstract

A method and related apparatus for wireless communication at a user equipment (UE) are provided. In the method, a UE receives, from a network entity, a first indication indicating a target set of machine learning (ML) functions of a plurality of defined ML functions, and communicates with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions. The method provides a new signaling framework dedicated to ML model switching and enables the UE to switch parameters of the ML function with reduced overall signaling overhead. The method improves the efficiency of wireless communication.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of and priority to PCT International Application No. PCT / CN2023 / 078054, filed on February 24, 2023, entitled “FUNCTIONALITY BASED IMPLICIT MLINFERENCE PARAMETER-GROUP SWITCH FOR BEAM PREDICTION”, the entire text of which is expressly incorporated herein by reference. Technical Field

[0003] The present disclosure relates generally to communication systems, and more particularly to functionality-based implicit machine learning (ML) inference parameter set switching in wireless communications. Background Art

[0004] Wireless communication systems are widely deployed to provide a variety of telecommunication services, such as telephony, video, data, messaging, and broadcasts. Typical wireless communication systems employ multiple-access technologies capable of supporting communication with multiple users by sharing available system resources. Examples of such multiple-access technologies include code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal frequency division multiple access (OFDMA), single-carrier frequency division multiple access (SC-FDMA), and time division synchronous code division multiple access (TD-SCDMA).

[0005] These multiple access technologies have been adopted in various telecommunication standards to provide a common protocol that enables different wireless devices to communicate at the city, national, regional, and even global levels. An example telecommunication standard is 5G New Radio (NR). 5G NR is part of the continued evolution of mobile broadband, promulgated by the 3rd Generation Partnership Project (3GPP) to meet new requirements associated with latency, reliability, security, scalability (e.g., for the Internet of Things (IoT)), and other requirements. 5G NR includes services associated with enhanced mobile broadband (eMBB), massive machine-type communications (mMTC), and ultra-reliable low-latency communications (URLLC). Certain aspects of 5G NR may be based on the 4G Long Term Evolution (LTE) standard. Further improvements to 5G NR technology are needed. In addition, these improvements may also be applicable to other multiple access technologies and telecommunication standards that adopt these technologies. Summary of the Invention

[0006] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of these aspects. This summary is not an extensive overview of all contemplated aspects. This summary does not identify key or critical elements of all aspects, nor does it delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that will be presented later.

[0007] In one aspect of the present disclosure, a method, computer-readable medium, and apparatus for wireless communications at a user equipment (UE) are provided. The apparatus may include a memory; and at least one processor coupled to the memory. Based at least in part on information stored in the memory, the at least one processor may be configured to: receive a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions; and communicate with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions.

[0008] In one aspect of the present disclosure, a method, computer-readable medium, and apparatus for wireless communication at a network entity are provided. The apparatus may include a memory and at least one processor coupled to the memory. Based at least in part on information stored in the memory, the at least one processor may be configured to: provide a first indication to a UE indicating a target set of ML functions from a plurality of defined ML functions; and communicate with the UE based on one or more operating parameters associated with the target set of ML functions.

[0009] To accomplish the foregoing and related objectives, one or more aspects may include the features hereinafter fully described and particularly pointed out in the claims. The following description and accompanying drawings set forth in detail some illustrative features of one or more aspects. However, these features are indicative of only some of the various ways in which the principles of the various aspects may be employed. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 is a diagram illustrating an example of a wireless communication system and an access network.

[0011] Figure 2A is a diagram illustrating an example of a first frame according to various aspects of the present disclosure.

[0012] Figure 2B is a diagram illustrating an example of downlink (DL) channels within a subframe according to various aspects of the present disclosure.

[0013] Figure 2C is a diagram illustrating an example of a second frame according to various aspects of the present disclosure.

[0014] Figure 2D is a diagram illustrating an example of uplink (UL) channels within a subframe according to various aspects of the present disclosure.

[0015] Figure 3 is a diagram illustrating an example of a base station and a user equipment (UE) in an access network.

[0016] Figure 4 is a diagram illustrating an example of an AI / ML algorithm in wireless communication.

[0017] Figure 5 is a diagram illustrating example AI / ML functionality in accordance with various aspects of the present disclosure.

[0018] Figure 6 is a call flow diagram illustrating a method of wireless communication according to various aspects of the present disclosure.

[0019] Figure 7 is a flowchart illustrating a method of wireless communication at a UE according to various aspects of the present disclosure.

[0020] Figure 8 is a flowchart illustrating a method of wireless communication at a UE according to various aspects of the present disclosure.

[0021] Figure 9 is a flow chart illustrating a method of wireless communication at a network entity according to various aspects of the present disclosure.

[0022] Figure 10 is a flow chart illustrating a method of wireless communication at a network entity according to various aspects of the present disclosure.

[0023] Figure 11 are diagrams illustrating examples of hardware implementations for example apparatuses and / or network entities.

[0024] Figure 12 is a diagram illustrating an example of a hardware implementation for an example network entity. DETAILED DESCRIPTION

[0025] When switching artificial intelligence / machine learning (AI / ML) models, operations such as time domain (TD) beam prediction using the AI / ML models may incur significant signaling overhead. For example, the network may transmit a large amount of information to the user equipment (UE) to initiate and facilitate model switching, resulting in increased latency and power consumption, especially when handling periodic, semi-persistent, or aperiodic channel state information (CSI) reports. Additionally, considering the different AI / ML models that can be trained locally at the UE, each AI / ML model is associated with a certain combination of operating parameters, the complexity and overhead of model switching become more significant. The example aspects presented herein address these issues by providing an efficient mechanism that minimizes signaling overhead while seamlessly enabling AI / ML model switching in wireless communications.

[0026] Various aspects generally relate to wireless communication systems. Some aspects more specifically relate to functionality-based implicit ML inference parameter set switching for ML models used in wireless communications. In some examples, a set of functionalities may include a defined set of parameters for an artificial intelligence (AI) or machine learning (ML) model. For example, the various functionalities may include beam prediction, channel state feedback, or positioning. The defined set of functionalities may include various levels of sub-functionalities, each with a corresponding set of parameters for the ML model. In some aspects, various ML parameters for different functionalities may be defined in wireless standards. A network may indicate a target set of functionalities to a UE. The UE may then switch between different parameters for the ML model based on the functionality at the UE, e.g., switching between various parameter sets without requiring additional signaling from the network. For example, the UE may receive a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions. The UE may then communicate with the network entity based on one or more operating parameters associated with the indicated target set of ML functions from the plurality of defined ML functions. In some aspects, the UE may configure one or more operating parameters associated with the target set of ML functions from the plurality of defined operating parameters based on the first indication. In some aspects, the plurality of defined operating parameters may include one or more of: a first set of parameters related to UE ML interference behavior of the plurality of ML functions; a second set of parameters related to expected ML inputs and outputs of the plurality of ML functions; or a third set of parameters related to assistance information.

[0027] Certain aspects of the subject matter described in this disclosure can be implemented to achieve one or more of the following potential advantages. In some examples, by receiving a first indication indicating a target set of ML functions; and communicating with a network entity based on one or more operating parameters associated with the target set of ML functions, the described techniques can be used to provide a new signaling framework dedicated to artificial intelligence or ML (AI / ML) model switching, and enable a UE to switch parameters of the ML functions with reduced overall signaling overhead.

[0028] The detailed description set forth below in conjunction with the accompanying drawings is a description of various configurations and does not 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, these concepts may be practiced without these specific details. In some cases, well-known structures and components are shown in block diagram form to avoid obscuring such concepts.

[0029] Several aspects of telecommunications systems are presented with reference to various apparatus and methods. These apparatus and methods are described in the following detailed description and illustrated in the accompanying drawings by various blocks, components, circuits, processes, algorithms, etc. (collectively referred to as "elements"). These elements can be implemented using electronic hardware, computer software, or any combination thereof. Whether these elements are implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system.

[0030] As an example, an element, or any part of an element, or any combination of elements, can be implemented as a "processing system" that includes one or more processors. When multiple processors are implemented, the multiple processors can perform functions individually or in combination. Examples of processors include microprocessors, microcontrollers, graphics processing units (GPUs), central processing units (CPUs), application processors, digital signal processors (DSPs), reduced instruction set computing (RISC) processors, systems on a chip (SoCs), baseband processors, field programmable gate arrays (FPGAs), programmable logic devices (PLDs), state machines, gating logic, discrete hardware circuits, and other suitable hardware configured to perform the various functionalities described throughout this disclosure. One or more processors in a processing system can execute software. Whether referred to as software, firmware, middleware, microcode, hardware description language, or other terms, software should be broadly interpreted to mean instructions, instruction sets, codes, code segments, program codes, programs, subroutines, software components, applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, processes, functions, or any combination thereof.

[0031] Thus, in one or more example aspects, implementations, and / or use cases, the functionality described may be implemented in hardware, software, or any combination thereof. If implemented in software, the functionality may be stored or encoded as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media. A storage medium can be any available medium that can be accessed by a computer. By way of example, such computer-readable media may include random access memory (RAM), read-only memory (ROM), electrically erasable programmable ROM (EEPROM), optical disk storage, magnetic disk storage, other magnetic storage devices, combinations of these types of computer-readable media, or any other medium that can be used to store computer-executable code in the form of instructions or data structures that can be accessed by a computer.

[0032] While aspects, implementations, and / or use cases are described herein through the lens of a few examples, additional or different aspects, implementations, and / or use cases may arise in many different arrangements and scenarios. The aspects, implementations, and / or use cases described herein may be implemented across many different platform types, devices, systems, shapes, sizes, and packaging arrangements. For example, the aspects, implementations, and / or use cases may arise via integrated chip implementations and other non-module component-based devices (e.g., end-user devices, vehicles, communications equipment, computing devices, industrial equipment, retail / purchase equipment, medical devices, artificial intelligence (AI)-enabled devices, etc.). While some examples may or may not be specifically targeted at use cases or applications, the examples described may have broad applicability. The aspects, implementations, and / or use cases may range from chip-level or modular components to non-modular, non-chip-level implementations, and further to aggregated, distributed, or original equipment manufacturer (OEM) devices or systems incorporating one or more of the techniques described herein. In some practical settings, devices incorporating the described aspects and features may also include additional components and features for implementing and practicing the claimed and described aspects. For example, the transmission and reception of wireless signals necessarily involve multiple components for both analog and digital purposes (e.g., hardware components including antennas, RF chains, power amplifiers, modulators, buffers, processors, interleavers, adders / summers, etc.). The techniques described herein can be practiced in a wide variety of devices of various sizes, shapes, and configurations, including chip-level components, systems, distributed arrangements, aggregated or disaggregated components, end-user devices, and the like.

[0033] The deployment of a communication system (such as a 5G NR system) can be arranged in a variety of ways using various components or parts. In a 5G NR system or network, a network node, a network entity, a mobility element of the network, a radio access network (RAN) node, a core network node, a network element, or network equipment (such as a base station (BS)), or one or more units (or one or more components) performing base station functionality can be implemented in a converged or disaggregated architecture. For example, a base station (such as a node B (NB), an evolved NB (eNB), an NR base station, a 5G NB, an access point (AP), a transmit / receive point (TRP), or a cell) can be implemented as a converged base station (also known as a standalone base station or a monolithic base station) or a disaggregated base station.

[0034] A converged base station may be configured to utilize a radio protocol stack that is physically or logically integrated within a single RAN node. A disaggregated base station may be configured to utilize a protocol stack that is physically or logically distributed across two or more units, such as one or more central or centralized units (CUs), one or more distributed units (DUs), or one or more radio units (RUs). In some aspects, a CU may be implemented within a RAN node, and one or more DUs may be co-located with the CU, or alternatively, may be geographically or virtually distributed across one or more other RAN nodes. A DU may be implemented to communicate with one or more RUs. Each of the CU, DU, and RU may be implemented as a virtual unit, namely a virtual central unit (VCU), a virtual distributed unit (VDU), or a virtual radio unit (VRU).

[0035] Base station operation or network design can take into account the aggregated nature of base station functionality. For example, a disaggregated base station can be used in an integrated access backhaul (IAB) network, an open radio access network (O-RAN (a network configuration such as that promoted by the O-RAN Alliance)), or a virtualized radio access network (vRAN, also known as a cloud radio access network (C-RAN)). Disaggregation can include distributing functionality across two or more units at various physical locations, as well as virtually distributing the functionality of at least one unit, which can enable flexibility in network design. The various units of a disaggregated base station or disaggregated RAN architecture can be configured for wired or wireless communication with at least one other unit.

[0036] Figure 1FIG100 is a diagram illustrating an example of a wireless communication system and access network. The illustrated wireless communication system includes a disaggregated base station architecture. The disaggregated base station architecture may include one or more CUs 110, which may communicate directly with a core network 120 via a backhaul link or indirectly with the core network 120 through one or more disaggregated base station units, such as a near real-time (near-RT) RAN intelligent controller (RIC) 125 via an E2 link, a non-real-time (non-RT) RIC 115 associated with a service management and orchestration (SMO) framework 105, or both. The CUs 110 may communicate with one or more DUs 130 via corresponding midhaul links, such as the F1 interface. The DUs 130 may communicate with one or more RUs 140 via corresponding fronthaul links. The RUs 140 may communicate with corresponding UEs 104 via one or more radio frequency (RF) access links. In some implementations, a UE 104 may be served simultaneously by multiple RUs 140.

[0037] Each of the units (i.e., CU 110, DU 130, RU 140, as well as near-RT RIC 125, non-RT RIC 115, and SMO framework 105) may include or be coupled to one or more interfaces configured to receive or transmit signals, data, or information (collectively, signals) via a wired or wireless transmission medium. Each of the units, or an associated processor or controller that provides instructions to the communication interfaces of these units, may be configured to communicate with one or more of the other units via the transmission medium. For example, these units may include a wired interface configured to receive signals or transmit signals to one or more of the other units via the wired transmission medium. Additionally, these units may include a wireless interface that may include a receiver, transmitter, or transceiver (such as an RF transceiver) configured to receive signals or transmit signals to one or more of the other units via the wireless transmission medium.

[0038] In some aspects, the CU 110 may host one or more higher-layer control functions. Such control functions may include radio resource control (RRC), packet data convergence protocol (PDCP), service data adaptation protocol (SDAP), and the like. Each control function may be implemented using an interface configured to communicate signals with other control functions hosted by the CU 110. The CU 110 may be configured to handle user plane functionality (i.e., central unit-user plane (CU-UP)), control plane functionality (i.e., central unit-control plane (CU-CP)), or a combination thereof. In some implementations, the CU 110 may be logically split into one or more CU-UP units and one or more CU-CP units. When implemented in an O-RAN configuration, the CU-UP units may communicate bidirectionally with the CU-CP units via an interface, such as the E1 interface. As needed, the CU 110 may be implemented to communicate with the DU 130 for network control and signaling.

[0039] The DU 130 may correspond to a logical unit that includes one or more base station functions for controlling the operation of one or more RUs 140. In some aspects, the DU 130 may host one or more of a radio link control (RLC) layer, a medium access control (MAC) layer, and one or more higher physical (PHY) layers (such as modules for forward error correction (FEC) encoding and decoding, scrambling, modulation and demodulation, etc.), at least in part according to a functional split (such as those defined by 3GPP). In some aspects, the DU 130 may also host one or more lower PHY layers. Each layer (or module) may be implemented using an interface configured to communicate signals with other layers (and modules) hosted by the DU 130 or with control functions hosted by the CU 110.

[0040] Lower layer functionality may be implemented by one or more RUs 140. In some deployments, a RU 140 controlled by a DU 130 may correspond to a logical node that hosts RF processing functionality or low-PHY layer functionality (such as performing fast Fourier transforms (FFTs), inverse FFTs (iFFTs), digital beamforming, physical random access channel (PRACH) extraction and filtering), or both, based at least in part on a functional split (such as a lower layer functional split). In this architecture, the RU 140 may be implemented to handle over-the-air (OTA) communications with one or more UEs 104. In some implementations, both real-time and non-real-time aspects of control and user plane communications with the RU 140 may be controlled by the corresponding DU 130. In some scenarios, this configuration may enable the implementation of the DU 130 and CU 110 in a cloud-based RAN architecture, such as a vRAN architecture.

[0041] The SMO framework 105 can be configured to support RAN deployment and provisioning of both non-virtualized and virtualized network elements. For non-virtualized network elements, the SMO framework 105 can be configured to support the deployment of dedicated physical resources for RAN coverage requirements, which can be managed via an operations and maintenance interface (such as the O1 interface). For virtualized network elements, the SMO framework 105 can be configured to interact with a cloud computing platform (such as Open Cloud (O-Cloud) 190) to perform network element lifecycle management (such as instantiating virtualized network elements) via cloud computing platform interfaces (such as the O2 interface). Such virtualized network elements may include, but are not limited to, the CU 110, DU 130, RU 140, and near-RT RIC 125. In some implementations, the SMO framework 105 can communicate with hardware aspects of the 4G RAN (such as the Open eNB (O-eNB) 111) via the O1 interface. Additionally, in some implementations, the SMO framework 105 can communicate directly with one or more RUs 140 via the O1 interface. The SMO framework 105 may also include a non-RT RIC 115 configured to support the functionality of the SMO framework 105 .

[0042] The non-RT RIC 115 may be configured to include logic that enables non-real-time control and optimization of RAN elements and resources, artificial intelligence (AI) / machine learning (ML) (AI / ML) workflows including model training and updating, or policy-based guidance of applications / features in the near-RT RIC 125. The non-RT RIC 115 may be coupled to or in communication with the near-RT RIC 125 (e.g., via an A1 interface). The near-RT RIC 125 may be configured to include logic that enables near-real-time control and optimization of RAN elements and resources through data collection and actions via an interface (e.g., via an E2 interface) that connects one or more CUs 110, one or more DUs 130, or both, and the O-eNB with the near-RT RIC 125.

[0043] In some implementations, to generate AI / ML models to be deployed in the near-RT RIC 125, the non-RT RIC 115 may receive parameters or external enrichment information from an external server. This information may be utilized by the near-RT RIC 125 and may be received from non-network data sources or from network functions at the SMO framework 105 or the non-RT RIC 115. In some examples, the non-RT RIC 115 or the near-RT RIC 125 may be configured to tune RAN behavior or performance. For example, the non-RT RIC 115 may monitor long-term trends and patterns in performance and employ AI / ML models to execute corrective actions through the SMO framework 105 (such as via reconfiguration of O1) or by creating RAN management policies (such as A1 policies).

[0044] At least one of the CU 110, DU 130, and RU 140 may be referred to as a base station 102. Thus, base station 102 may include one or more of CU 110, DU 130, and RU 140 (each component is indicated by a dashed line to indicate that each component may or may not be included in base station 102). Base station 102 provides a UE 104 with access to core network 120. Base station 102 may include a macro cell (a high-power cellular base station) and / or a small cell (a low-power cellular base station). Small cells include femto cells, pico cells, and micro cells. A network that includes both small cells and macro cells may be referred to as a heterogeneous network. A heterogeneous network may also include a home evolved Node B (eNB) (HeNB), which may provide services to a restricted group known as a closed subscriber group (CSG). The communication link between RU 140 and UE 104 may include uplink (UL) (also known as reverse link) transmissions from UE 104 to RU 140 and / or downlink (DL) (also known as forward link) transmissions from RU 140 to UE 104. The communication link may utilize multiple-input, multiple-output (MIMO) antenna technology, including spatial multiplexing, beamforming, and / or transmit diversity. The communication link may be over one or more carriers. Base station 102 / UE 104 may utilize spectrum with a bandwidth of up to Y MHz (e.g., 5 MHz, 10 MHz, 15 MHz, 20 MHz, 100 MHz, 400 MHz, etc.) for each carrier allocated in a carrier aggregation for transmission in each direction, totaling up to Yx MHz (x component carriers). These carriers may or may not be adjacent to each other. Carrier allocation may be asymmetric with respect to DL and UL (e.g., more or fewer carriers may be allocated for DL ​​compared to UL). Component carriers may include a primary component carrier and one or more secondary component carriers. The primary component carrier may be referred to as a primary cell (PCell) and the secondary component carrier may be referred to as a secondary cell (SCell).

[0045] Certain UEs 104 may communicate with each other using device-to-device (D2D) communication links 158. D2D communication links 158 may use DL / UL wireless wide area network (WWAN) spectrum. D2D communication links 158 may use one or more sidelink channels, such as a physical sidelink broadcast channel (PSBCH), a physical sidelink discovery channel (PSDCH), a physical sidelink shared channel (PSSCH), and a physical sidelink control channel (PSCCH). D2D communication may be performed using various wireless D2D communication systems, such as, for example, Bluetooth™ (Bluetooth is a trademark of the Bluetooth Special Interest Group (SIG)), Wi-Fi™ (Wi-Fi is a trademark of the Wi-Fi Alliance) based on the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standard, LTE, or NR.

[0046] The wireless communication system may also include a Wi-Fi AP 150 that communicates with a UE 104 (also referred to as a Wi-Fi station (STA)) via a communication link 154, for example, in the 5 GHz unlicensed spectrum. When communicating in the unlicensed spectrum, the UE 104 / AP 150 may perform a clear channel assessment (CCA) to determine whether the channel is available before communicating.

[0047] The electromagnetic spectrum is typically subdivided into various categories, bands, channels, and so on, based on frequency / wavelength. In 5G NR, two initial operating bands have been identified as frequency ranges designated FR1 (410 MHz-7.125 GHz) and FR2 (24.25 GHz-52.6 GHz). Although a portion of FR1 extends beyond 6 GHz, FR1 is often (interchangeably) referred to as the "sub-6 GHz" band in various documents and articles. A similar naming issue sometimes arises with FR2, which is often (interchangeably) referred to as the "millimeter wave" band in documents and articles, despite being distinct from the extremely high frequency (EHF) band (30 GHz-300 GHz), which is designated as a "millimeter wave" band by the International Telecommunication Union (ITU).

[0048] Frequencies between FR1 and FR2 are often referred to as mid-band frequencies. Recent 5G NR research has identified an operating band for these mid-band frequencies as frequency range designation FR3 (7.125 GHz-24.25 GHz). Frequency bands falling within FR3 may inherit FR1 characteristics and / or FR2 characteristics, effectively extending the features of FR1 and / or FR2 to mid-band frequencies. Furthermore, higher frequency bands are currently being explored to extend 5G NR operation beyond 52.6 GHz. For example, three higher operating bands have been identified as frequency range designations FR2-2 (52.6 GHz-71 GHz), FR4 (71 GHz-114.25 GHz), and FR5 (114.25 GHz-300 GHz). Each of these higher frequency bands falls within the EHF band.

[0049] With the above in mind, unless otherwise specified, if the term "sub-6 GHz" or the like is used herein, it may broadly refer to frequencies that may be less than 6 GHz, may be within FR1, or may include mid-band frequencies. Furthermore, unless otherwise specified, if the term "millimeter wave" or the like is used herein, it may broadly refer to frequencies that may include mid-band frequencies, may be within FR2, FR4, FR2-2, and / or FR5, or may be within the EHF band.

[0050] Base station 102 and UE 104 may each include multiple antennas (such as antenna elements, antenna panels, and / or antenna arrays) to facilitate beamforming. Base station 102 may transmit beamformed signals 182 to UE 104 in one or more transmit directions. UE 104 may receive beamformed signals from base station 102 in one or more receive directions. UE 104 may also transmit beamformed signals 184 to base station 102 in one or more transmit directions. Base station 102 may receive beamformed signals from UE 104 in one or more receive directions. Base station 102 / UE 104 may perform beam training to determine the optimal receive and transmit directions for each of base station 102 / UE 104. The transmit and receive directions of base station 102 may or may not be the same. The transmit and receive directions of UE 104 may or may not be the same.

[0051] The base station 102 may include and / or be referred to as a gNB, Node B, eNB, access point, base transceiver station, radio base station, radio transceiver, transceiver function, basic service set (BSS), extended service set (ESS), TRP, network node, network entity, network equipment, or some other suitable terminology. The base station 102 may be implemented as an integrated access and backhaul (IAB) node, a relay node, a sidelink node, a converged (monolithic) base station having a baseband unit (BBU) (including a CU and a DU) and a RU, or as a disaggregated base station including one or more of a CU, a DU, and / or a RU. A collection of base stations that may include disaggregated base stations and / or converged base stations may be referred to as a next generation (NG) RAN (NG-RAN).

[0052] The core network 120 may include an access and mobility management function (AMF) 161, a session management function (SMF) 162, a user plane function (UPF) 163, a unified data management (UDM) 164, one or more location servers 168, and other functional entities. The AMF 161 is a control node that handles signaling between the UE 104 and the core network 120. The AMF 161 supports registration management, connection management, mobility management, and other functions. The SMF 162 supports session management and other functions. The UPF 163 supports packet routing, packet forwarding, and other functions. The UDM 164 supports the generation of authentication and key agreement (AKA) credentials, user identity handling, access authorization, and subscription management. The one or more location servers 168 are exemplified as including a gateway mobile location center (GMLC) 165 and a location management function (LMF) 166. However, in general, the one or more location servers 168 may include one or more location / positioning servers, which may include one or more of the GMLC 165, LMF 166, Position Determination Entity (PDE), Serving Mobile Location Center (SMLC), Mobile Positioning Center (MPC), etc. The GMLC 165 and LMF 166 support UE location services. The GMLC 165 provides an interface for clients / applications (e.g., emergency services) to access UE positioning information. The LMF 166 receives measurements and assistance information from the NG-RAN and UE 104 via the AMF 161 to calculate the position of the UE 104. The NG-RAN may utilize one or more positioning methods to determine the position of the UE 104. Positioning the UE 104 may involve signal measurements, position estimates, and optional velocity calculations based on these measurements. Signal measurements may be performed by the UE 104 and / or the base station 102 serving the UE 104. The measured signals may be based on one or more of a satellite positioning system (SPS) 170 (e.g., one or more of a global navigation satellite system (GNSS), a global positioning system (GPS), a non-terrestrial network (NTN), or other satellite positioning / positioning systems), LTE signals, wireless local area network (WLAN) signals, Bluetooth signals, a terrestrial beacon system (TBS), sensor-based information (e.g., an atmospheric pressure sensor, a motion sensor), NR enhanced cell ID (NR E-CID) methods, NR signals (e.g., multi-round trip time (multi-RTT), DL angle of departure (DL-AoD), DL time difference of arrival (DL-TDOA), UL time difference of arrival (UL-TDOA), and UL angle of arrival (UL-AoA) positioning), and / or other systems / signals / sensors.

[0053] Examples of UE 104 include a cellular phone, a smartphone, a Session Initiation Protocol (SIP) phone, a laptop, a personal digital assistant (PDA), a satellite radio, a global positioning system, a multimedia device, a video device, a digital audio player (e.g., an MP3 player), a camera, a game console, a tablet, a smart device, a wearable device, a vehicle, an electric meter, a gas pump, a large or small kitchen appliance, a healthcare device, an implant, a sensor / actuator, a display, or any other similarly functional device. Some of UE 104 may be referred to as IoT devices (e.g., a parking meter, a gas pump, a toaster, a vehicle, a heart rate monitor, etc.). UE 104 may also be referred to as a station, a mobile station, a subscriber station, a mobile unit, a subscriber unit, a wireless unit, a remote unit, a mobile device, a wireless device, a wireless communication device, a remote device, a mobile subscriber station, an access terminal, a mobile terminal, a wireless terminal, a remote terminal, a handset, a user agent, a mobile client, a client, or some other suitable terminology. In some scenarios, the term UE may also apply to one or more companion devices, such as in a device constellation arrangement. One or more of these devices may access the network collectively and / or individually.

[0054] Reference again Figure 1 In certain aspects, the UE 104 may include an AI / ML function switching component 198. The AI / ML function switching component 198 may be configured to: receive a first indication indicating a target set of ML functions from a plurality of defined ML functions from a network entity; and communicate with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions. In certain aspects, the base station 102 may include an AI / ML function switching component 199. The AI / ML function switching component 199 may be configured to: provide a first indication indicating a target set of ML functions from a plurality of defined ML functions to the UE; and communicate with the UE based on one or more operating parameters associated with the target set of ML functions. Although the following description may focus on 5G NR, the concepts described herein may be applicable to other similar areas, such as LTE, LTE-A, CDMA, GSM, and other wireless technologies.

[0055] Figure 2A FIG200 is a diagram illustrating an example of a first subframe within a 5G NR frame structure. Figure 2B FIG230 is a diagram illustrating an example of DL channels within a 5G NR subframe. Figure 2C FIG250 is a diagram illustrating an example of a second subframe within a 5G NR frame structure. Figure 2DFIG280 is a diagram illustrating an example of UL channels within a 5G NR subframe. The 5G NR frame structure may be frequency division duplex (FDD), where for a particular set of subcarriers (carrier system bandwidth), subframes within that subcarrier set are dedicated to either DL or UL, or time division duplex (TDD), where for a particular set of subcarriers (carrier system bandwidth), subframes within that subcarrier set are dedicated to both DL and UL. Figure 2A 、 Figure 2C In the example provided, the 5G NR frame structure is assumed to be TDD, with subframe 4 configured with slot format 28 (mostly DL), where D stands for DL, U stands for UL, and F stands for flexible use between DL / UL, and subframe 3 configured with slot format 1 (all UL). While subframes 3 and 4 are shown with slot formats 1 and 28, respectively, any particular subframe can be configured with any of the various available slot formats 0-61. Slot formats 0 and 1 are all DL and all UL, respectively. The other slot formats 2-61 include a mix of DL, UL, and flexible symbols. The UE is configured with the slot format via a received slot format indicator (SFI), either dynamically via DL control information (DCI) or semi-statically / statically via radio resource control (RRC) signaling. Note that the following description also applies to the 5G NR frame structure as TDD.

[0056] Figures 2A to 2D This example illustrates a frame structure, and aspects of this disclosure are applicable to other wireless communication technologies that may have different frame structures and / or different channels. A frame (10 ms) can be divided into 10 equally sized subframes (1 ms). Each subframe may include one or more slots. A subframe may also include mini-slots, which may include 7, 4, or 2 symbols. Each slot may include 14 or 12 symbols, depending on whether the cyclic prefix (CP) is normal or extended. For a normal CP, each slot may include 14 symbols, and for an extended CP, each slot may include 12 symbols. Downlink symbols may be CP-orthogonal frequency division multiplexing (OFDM) symbols. Uplink symbols may be CP-OFDM symbols (for high-throughput scenarios) or discrete Fourier transform (DFT)-spread OFDM (DFT-s-OFDM) symbols (for power-limited scenarios; limited to single-stream transmission). The number of slots within a subframe depends on the CP and the parameter set. The parameter set defines the subcarrier spacing (SCS) (see Table 1). Symbol length / duration can be scaled with 1 / SCS.

[0057]

[0058] Table 1: Parameter set, SCS and CP

[0059] For normal CP (14 symbols / slot), different parameter sets µ 0 to 4 allow 1, 2, 4, 8, and 16 slots per subframe, respectively. For extended CP, parameter set 2 allows 4 slots per subframe. Therefore, for normal CP and parameter set µ, there are 14 symbols / slot and 2 µ time slots / subframe. The subcarrier spacing can be equal to ,in For parameter sets 0 to 4. Therefore, the subcarrier spacing for parameter set µ=0 is 15 kHz, and the subcarrier spacing for parameter set µ=4 is 240 kHz. The symbol length / duration is inversely related to the subcarrier spacing. Figures 2A to 2D An example is provided for a normal CP with 14 symbols per slot and a parameter set µ=2 with 4 slots per subframe. The slot duration is 0.25ms, the subcarrier spacing is 60kHz, and the symbol duration is approximately 16.67µs. Within a frame set, there may be one or more different bandwidth parts (BWPs) frequency-division multiplexed (see Figure 2B ). Each BWP may have a specific parameter set and CP (normal or extended).

[0060] A resource grid can be used to represent the frame structure. Each slot consists of a resource block (RB) (also known as a physical RB (PRB)) that extends over 12 consecutive subcarriers. The resource grid is divided into multiple resource elements (REs). The number of bits carried by each RE depends on the modulation scheme.

[0061] like Figure 2A As illustrated, some of the REs carry reference (pilot) signals (RS) for the UE. The RSs may include a demodulation RS (DM-RS) (indicated as R for one specific configuration, but other DM-RS configurations are possible) and a channel state information reference signal (CSI-RS) for channel estimation at the UE. The RSs may also include a beamforming RS (BRS), a beam refinement RS (BRRS), and a phase tracking RS (PT-RS).

[0062] Figure 2BExamples of various downlink channels within a subframe of a frame are illustrated. The physical downlink control channel (PDCCH) carries DCI within one or more control channel elements (CCEs) (e.g., 1, 2, 4, 8, or 16 CCEs). Each CCE consists of six resource element groups (REGs), with each REG comprising 12 contiguous REs within an OFDM symbol of a RB. The PDCCH within a BWP is referred to as a control resource set (CORESET). During PDCCH monitoring opportunities within a CORESET, the UE is configured to monitor PDCCH search spaces (e.g., common search space, UE-specific search space) for PDCCH candidates with different DCI formats and aggregation levels. Additional BWPs may be located at higher and / or lower frequencies across the channel bandwidth. The primary synchronization signal (PSS) may be within symbol 2 of specific subframes of a frame. The PSS is used by UE 104 to determine subframe / symbol timing and physical layer identification. The secondary synchronization signal (SSS) may be within symbol 4 of specific subframes of a frame. The SSS is used by the UE to determine the physical layer cell identity group number and radio frame timing. Based on the physical layer identity and the physical layer cell identity group number, the UE can determine the physical cell identifier (PCI). Based on the PCI, the UE can determine the location of the DM-RS. The physical broadcast channel (PBCH), which carries the master information block (MIB), can be logically grouped with the PSS and SSS to form a synchronization signal (SS) / PBCH block (also known as an SS block (SSB)). The MIB provides the system frame number (SFN) and the number of RBs in the system bandwidth. The physical downlink shared channel (PDSCH) carries user data, broadcast system information not sent via the PBCH (such as the system information block (SIB)), and paging messages.

[0063] like Figure 2C As illustrated, some of the REs carry DM-RSs (indicated as R for a specific configuration, but other DM-RS configurations are possible) for channel estimation at the base station. The UE can transmit DM-RSs for the physical uplink control channel (PUCCH) and DM-RSs for the physical uplink shared channel (PUSCH). The PUSCH DM-RS can be transmitted in the first or first two symbols of the PUSCH. The PUCCH DM-RS can be transmitted in different configurations depending on whether a short or long PUCCH is transmitted and the specific PUCCH format used. The UE can transmit a sounding reference signal (SRS). The SRS can be transmitted in the last symbol of the subframe. The SRS can have a comb structure, and the UE can transmit the SRS on one of the teeth of the comb. The SRS can be used by the base station for channel quality estimation to achieve frequency-dependent scheduling of the UL.

[0064] Figure 2DExamples of various UL channels within a subframe of a frame are illustrated. The PUCCH may be located at the position indicated in one configuration. The PUCCH carries uplink control information (UCI), such as a scheduling request, a channel quality indicator (CQI), a precoding matrix indicator (PMI), a rank indicator (RI), and hybrid automatic repeat request (HARQ) acknowledgement (ACK) (HARQ-ACK) feedback (i.e., one or more HARQ ACK bits indicating one or more ACKs and / or negative ACKs (NACKs)). The PUSCH carries data and may additionally be used to carry a buffer status report (BSR), a power headroom report (PHR), and / or UCI.

[0065] Figure 3 Figure 3 is a block diagram of a base station 310 communicating with a UE 350 in an access network. In the DL, Internet Protocol (IP) packets may be provided to a controller / processor 375. The controller / processor 375 implements Layer 3 and Layer 2 functionality. Layer 3 includes the Radio Resource Control (RRC) layer, and Layer 2 includes the Service Data Adaptation Protocol (SDAP) layer, the Packet Data Convergence Protocol (PDCP) layer, the Radio Link Control (RLC) layer, and the Medium Access Control (MAC) layer. The controller / processor 375 provides RRC layer functionality associated with broadcasting of system information (e.g., MIB, SIB), RRC connection control (e.g., RRC connection paging, RRC connection establishment, RRC connection modification, and RRC connection release), inter-radio access technology (RAT) mobility, and measurement configuration for UE measurement reporting; PDCP layer functionality associated with header compression / decompression, security (ciphering, deciphering, integrity protection, integrity verification), and handover support functions; RLC layer functionality associated with delivery of upper layer packet data units (PDUs), error correction through ARQ, concatenation, segmentation and reassembly of RLC service data units (SDUs), resegmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto transport blocks (TBs), demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction through HARQ, priority handling, and logical channel prioritization.

[0066] The transmit (TX) processor 316 and receive (RX) processor 370 implement Layer 1 functionality associated with various signal processing functions. Layer 1, which includes the physical (PHY) layer, may include error detection on transport channels, forward error correction (FEC) coding / decoding of transport channels, interleaving, rate matching, mapping onto physical channels, modulation / demodulation of physical channels, and MIMO antenna processing. The TX processor 316 handles the mapping onto signal constellations based on various modulation schemes (e.g., binary phase-shift keying (BPSK), quadrature phase-shift keying (QPSK), M-order phase-shift keying (M-PSK), and M-order quadrature amplitude modulation (M-QAM)). The coded and modulated symbols are then separated into parallel streams. Each stream is then mapped to an OFDM subcarrier, multiplexed with a reference signal (e.g., a pilot) in the time and / or frequency domain, and then combined using an inverse fast Fourier transform (IFFT) to produce a physical channel carrying a time-domain OFDM symbol stream. The OFDM stream is spatially precoded to generate multiple spatial streams. Channel estimates from a channel estimator 374 may be used to determine the coding and modulation schemes, as well as for spatial processing. The channel estimates may be derived from a reference signal and / or channel condition feedback transmitted by the UE 350. Each spatial stream may then be provided to a different antenna 320 via a separate transmitter 318Tx. Each transmitter 318Tx may modulate a radio frequency (RF) carrier using a corresponding spatial stream for transmission.

[0067] At the UE 350, each receiver 354Rx receives a signal via its corresponding antenna 352. Each receiver 354Rx recovers the information modulated onto the RF carrier and provides the information to a receive (RX) processor 356. The TX processor 368 and the RX processor 356 implement Layer 1 functionality associated with various signal processing functions. The RX processor 356 performs spatial processing on the information to recover any spatial streams destined for the UE 350. If multiple spatial streams are destined for the UE 350, they may be combined by the RX processor 356 into a single OFDM symbol stream. The RX processor 356 then converts the OFDM symbol stream from the time domain to the frequency domain using a fast Fourier transform (FFT). The frequency-domain signal includes a separate OFDM symbol stream for each subcarrier of the OFDM signal. The symbols on each subcarrier, as well as the reference signal, are recovered and demodulated by determining the most likely signal constellation point transmitted by the base station 310. These soft decisions may be based on channel estimates calculated by the channel estimator 358. The soft decisions are then decoded and deinterleaved to recover the data and control signals originally sent on the physical channel by base station 310. The data and control signals are then provided to a controller / processor 359, which implements layer 3 and layer 2 functionality.

[0068] The controller / processor 359 may be associated with at least one memory 360 that stores program codes and data. The at least one memory 360 may be referred to as a computer-readable medium. In the UL, the controller / processor 359 provides demultiplexing between transport and logical channels, packet reassembly, decryption, header decompression, and control signal processing to recover IP packets. The controller / processor 359 is also responsible for error detection using ACK and / or NACK protocols to support HARQ operations.

[0069] Similar to the functionality described in conjunction with DL transmissions performed by the base station 310, the controller / processor 359 provides RRC layer functionality associated with system information (e.g., MIB, SIB) acquisition, RRC connection, and measurement reporting; PDCP layer functionality associated with header compression / decompression and security (encryption, decryption, integrity protection, integrity verification); RLC layer functionality associated with delivery of upper layer PDUs, error correction through ARQ, concatenation, segmentation, and reassembly of RLC SDUs, resegmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto TBs, demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction through HARQ, priority handling, and logical channel prioritization.

[0070] Channel estimates derived by the channel estimator 358 based on a reference signal or feedback transmitted by the base station 310 may be used by the TX processor 368 to select the appropriate coding and modulation schemes and to facilitate spatial processing. The spatial streams generated by the TX processor 368 may be provided to different antennas 352 via separate transmitters 354Tx. Each transmitter 354Tx may modulate an RF carrier with a corresponding spatial stream for transmission.

[0071] UL transmissions are processed at the base station 310 in a manner similar to that described in conjunction with the receiver functionality at the UE 350. Each receiver 318Rx receives a signal through its corresponding antenna 320. Each receiver 318Rx recovers information modulated onto an RF carrier and provides the information to the RX processor 370.

[0072] The controller / processor 375 may be associated with at least one memory 376 that stores program codes and data. The at least one memory 376 may be referred to as a computer-readable medium. In the UL, the controller / processor 375 provides demultiplexing between transport and logical channels, packet reassembly, decryption, header decompression, and control signal processing to recover IP packets. The controller / processor 375 is also responsible for error detection using ACK and / or NACK protocols to support HARQ operations.

[0073] At least one of the TX processor 368, the RX processor 356, and the controller / processor 359 may be configured to perform operations related to Figure 1 The AI / ML functional switching component 198 combines various aspects.

[0074] At least one of the TX processor 316, the RX processor 370, and the controller / processor 375 may be configured to perform operations related to Figure 1 The AI / ML function switching component 199 combines various aspects.

[0075] If combined Figure 1As described in 182 and 184 of , the UE and the network may communicate using one or more directional beams. The UE and the network may perform various aspects of beam management to select beams for transmission and reception. In some aspects, beam management may be performed using a Tracking Reference Signal (TRS), for example, for UEs in RRC Inactive or RRC Idle states. For initial access, the UE may, for example, use a combination of SSBs and a wide beam sweeping procedure to identify a beam for initial access. For contention-based random access (CBRA), the UE may use a random access opportunity (RO) and a preamble corresponding to the selected SSB / beam. In the RRC Connected state, the UE and / or the network may perform various aspects of beam management, including, for example, the P1, P2, and P3 procedures using SSBs or CSI-RS measurements; the U1, U2, and U3 procedures using SRS transmission and measurement; and Layer 1 Reference Signal Received Power (L1-RSRP) reporting. The network may configure one or more TCI state configurations for the UE and may indicate the TCI state for the UE from a set of TCI state configurations. In some aspects, the UE may provide Layer 1 signal-to-interference-plus-noise ratio (L1-SINR) reporting, which can reduce overhead and latency and enable CC group beam updates or faster UL beam updates. In some aspects, the UE may communicate with the network using unified TCI states, L1 / L2-centric mobility (also referred to as L1 / L2 triggered mobility (LTM)), dynamic TCI updates, and / or uplink multi-panel selection, with maximum permitted exposure (MPE) migration. Beam management may be used in specific scenarios, such as high speeds (e.g., high-speed trains (HST)), single-frequency networks (SNFs), and multiple transmit / receive points (mTRPs). Based on measurements, the UE may identify beam failure detection (BFD) and perform beam failure recovery (BFD). In some aspects, BFD or BFR may be used for the primary cell (PCell) or primary / secondary cell (PSCell). BFD may be based on the BFD reference signal (BFD-RS) and the PDCCH block error rate (BLER). BFR may be based on contention-free random access (CFRA). For SCells, BFD and BFR may include a link recovery request via a Scheduling Request (SR) or Medium Access Control-Control Element (MAC-CE)-based BFR for SCells. If BFR is unsuccessful, the UE may identify a radio link failure.

[0076] Some wireless communications may include the use of AI or ML at the network and / or UE. In various examples, AI / ML may be used for beam management at the UE and / or network, including for performing beam prediction in the time and / or spatial domains. As another example, AI / ML models may be used for channel state feedback. As another example, AI / ML models may be used for positioning. The use of AI / ML models may reduce latency or overhead and improve the accuracy of beam selection. Models may be provided that support various levels of network and UE collaboration and various use cases. The use of AI / ML models may include various aspects such as model training, model deployment, model inference, model monitoring, and model updates.

[0077] Figure 4 4 is an example of an AI / ML algorithm 400 for a method of wireless communication, and illustrates various aspects of model training, model inference, model feedback, and model updates. The AI / ML algorithm 400 may include various functions, including data collection 402, model training function 404, model inference function 406, and participants 408.

[0078] Data collection 402 may be a function that provides input data to model training function 404 and model inference function 406. The data collection 402 function may include any form of data preparation, and it may not be specific to a particular implementation of an AI / ML algorithm (e.g., data pre-processing and cleaning, formatting, and transformation).

[0079] Examples of input data may include, but are not limited to, measurements from entities including UEs or network nodes (such as RSRP measurements or other TCI candidate information, channel measurements, positioning measurements), feedback from participants 408 (e.g., which may be UEs or network nodes), and output from another AI / ML model. Data collection 402 may include training data and inference data, where training data refers to data to be transmitted as input to AI / ML model training function 404 and inference data refers to data to be transmitted as input to AI / ML model inference function 406.

[0080] The model training function 404 may be a function that performs ML model training, validation, and testing, and may generate model performance metrics as part of the model testing process. The model training function 404 may also be responsible for data preparation (e.g., data preprocessing and cleaning, formatting, and transformation) based on the training data delivered or received from the data collection 402 function. The model training function 404 may deploy or update the trained, validated, and tested AI / ML models to the model inference function 406, and receive model performance feedback from the model inference function 406. As described above, there may be various functionalities to be performed by the AI / ML model for wireless communication.

[0081] Model inference function 406 may be a function that provides AI / ML model inference output (e.g., predictions or decisions). Model inference function 406 may also perform data preparation (e.g., data preprocessing and cleaning, formatting, and transformation) based on the inferred data delivered from data collection 402 function. The output of model inference function 406 may include the inferred output of the AI / ML model generated by model inference function 406. The details of the inferred output may be use case specific. For example, the output may include beam predictions for beam management. The predictions may be network-specific or UE-specific. The outputs may include, for example, positioning information for the UE. The outputs may include channel state feedback. The outputs may include various outputs for different functionalities. In some aspects, the participant may be a base station or a component of a core network. In other aspects, the participant may be a UE communicating with a wireless network.

[0082] Model performance feedback may refer to information derived from model inference functionality 406 that may be suitable for improving the AI / ML model trained in model training functionality 404. Feedback from participants 408 or other network entities (via data collection 402 functionality) may be implemented for model inference functionality 406 to create model performance feedback.

[0083] Participant 408 may be a function that receives output from model inference function 406 and triggers or executes corresponding actions. Participants may trigger actions for other network entities or their own network entities. Participant 408 may also provide feedback information to model training function 404 or model inference function 406 regarding training or inference data or performance feedback. This feedback may be sent back to data collection 402.

[0084] The network can use machine learning algorithms, deep learning algorithms, neural networks, reinforcement learning, regression, boosting or advanced signal processing methods for various aspects of wireless communication, including various functionalities such as beam management, CSF or positioning.

[0085] In some aspects described herein, the network can train one or more neural networks to learn the dependence of the quality of the measurement on various parameters. In addition, examples of machine learning models or neural networks that can be included in the network entity include artificial neural networks (ANNs); decision tree learning; convolutional neural networks (CNNs); deep learning architectures, in which the outputs of neurons in a first layer become the inputs of neurons in a second layer, etc.; support vector machines (SVMs), for example, which include a separating hyperplane (e.g., a decision boundary) for classifying data; regression analysis; Bayesian networks; genetic algorithms; deep convolutional networks (DCNs), which are configured with additional pooling and normalization layers; and deep belief networks (DBNs).

[0086] A machine learning model, such as an artificial neural network (ANN), may include a set of interconnected artificial neurons (e.g., a neuron model) and may be a computing device or may represent a method to be executed by a computing device. The connections of the neuron model may be modeled as weights. Machine learning models can be trained using datasets to provide predictive models, adaptive control, and other applications. The model can adapt based on external or internal information processed by the machine learning model. Machine learning can provide nonlinear statistical data models or decision making, and can model complex relationships between input data and output information.

[0087] A machine learning model may include multiple layers and / or operations, which can be formed by cascading one or more of the operations mentioned above. Examples of these operations include: extraction of various features from the data, convolution operations, fully connected operations that can be activated or deactivated, compression, decompression, quantization, flattening, and the like. As used herein, a "layer" of a machine learning model may refer to an operation performed on the input data. For example, a convolutional layer, a fully connected layer, and the like may refer to associated operations performed on the data input to the layer. A convolution AxB operation converts multiple input features A into multiple output features B. A "kernel size" may refer to the number of adjacent coefficients combined in a dimension. As used herein, "weights" may refer to one or more coefficients used in operations within each layer to combine rows and / or columns of the input data. For example, a fully connected layer operation may have an output y determined at least in part based on the sum of the product of the input matrix x and weights A (which may be a matrix) and bias values ​​B (which may be a matrix). The term "weights" may be used herein to refer generally to both weights and bias values. Weights and biases are examples of parameters of a trained machine learning model. Different layers of a machine learning model can be trained independently.

[0088] Machine learning models can include various connectivity patterns, such as any feedforward network, hierarchy, recursive architecture, feedback connections, etc. The connections between the layers of the neural network can be fully connected or locally connected. In a fully connected neural network, a neuron in a first layer can communicate its output to every neuron in a second layer, and each neuron in the second layer can receive input from every neuron in the first layer. In a locally connected network, a neuron in a first layer can be connected to a limited number of neurons in a second layer. In some aspects, a convolutional network can be locally connected and configured with shared connection strengths associated with the inputs of each neuron in the second layer. The locally connected layers of a network can be configured so that each neuron in the layer has the same or similar connectivity pattern, but with different connection strengths.

[0089] A machine learning model or neural network can be trained. For example, a machine learning model can be trained based on supervised learning. During training, the machine learning model is presented with inputs that the model uses to calculate and produce an output. The actual output can be compared to the target output, and the difference can be used to adjust the machine learning model's parameters (such as weights and biases) to provide an output closer to the target output. Before training, the output may be incorrect or less accurate, and the error, or difference, between the actual and target outputs can be calculated. The weights of the machine learning model can then be adjusted to align the output more closely with the target. To adjust the weights, the learning algorithm can calculate a gradient vector for the weights. The gradient indicates the amount by which the error will increase or decrease if the weights are slightly adjusted. At the top layer, the gradient may directly correspond to the value of the weights connecting the activated neurons in the penultimate layer to the neurons in the output layer. In lower layers, the gradient may depend on the value of the weights and the error gradient calculated in higher layers. The weights can then be adjusted to reduce the error or move the output closer to the target. This method of adjusting weights is referred to as backpropagation through the neural network. This process can continue until the achievable error rate stops decreasing, or until the error rate reaches a target level.

[0090] These machine learning models can involve computational complexity and a large number of processors for training machine learning models. The output of one node is connected as an input to another node. The connection between nodes can be called an edge, and weights can be applied to the connection / edge to adjust the output from one node, which is used as input to another node. Nodes can apply thresholds to determine whether or when to provide an output to a connected node. The output of each node can be calculated as a nonlinear function of the sum of the inputs to the node. A neural network can include any number of nodes and any type of connection between the nodes. A neural network can include one or more hidden nodes. Nodes can be grouped into layers, and different layers of a neural network can perform different types of transformations on the input. A signal can travel through multiple layers of a neural network from an input at the first layer to an output at the last layer of the neural network, and can traverse these layers multiple times.

[0091] The signaling overhead of AI / ML model switching can be significant. The network may transmit a large amount of information to the UE in order to enable the UE to switch to and start using a different AI / ML model. Using AI / ML model switching for time division (TD) beam prediction as an example, the UE may communicate with the UE via, for example, a message about the AI / ML model used for beam The UE may also report its predicted L1-RSRP with respect to the CSI report of the future time (where N and X are integers). beam Continuous historical measurement opportunities and ms basic beam management (BM) period (where M, Y and P are integers), based on which the future L1-RSRP can be predicted. Different AI / ML models may have been trained locally at the UE (e.g., in combination with Figure 4 described), and each model can be used with The base station can also determine the mode to which the UE is to switch based on different observations (e.g., out-of-loop performance or AI / ML performance monitoring observations).

[0092] To establish such AI / ML model switching, the base station may send a lot of signaling to the UE. For example, for periodic channel state information (P-CSI) reporting, the UE may be configured with a large number of CSI reporting settings associated with different operating modes, which may result in a large amount of configuration overhead. For semi-persistent CSI (SP-CSI) reporting, the UE may receive an indication (which is based on a combination of multiple pre-configured CSI reporting settings) and may result in a large amount of signaling overhead. For aperiodic CSI (AP-CSI) reporting, the UE is configured with a number of AP CSI triggering states to reflect (It may be limited due to the following reasons: the different combinations of activatable AP CSI triggering states may be limited due to DCI overhead limitations).

[0093] Example aspects herein provide methods and apparatus for implicit ML inference parameter set switching based on functionality. In some aspects, a new signaling framework dedicated to AI / ML model switching is introduced. Based on this new signaling framework, parameters of AI / ML models, such as those discussed above, can be implicitly switched based on reduced overall signaling overhead with low specification changes. The proposed methods may be described in the context of beam prediction, but the methods are not limited thereto and are generally applicable to other use cases.

[0094] In some aspects, in a method for functionality-based implicit ML inference parameter set switching, a set of functionalities may be defined for AI / ML-based UE operations. In some aspects, this set of functionalities may be defined in a wireless standard or otherwise known to the UE and network without requiring network signaling to the UE. In some aspects, the defined functionalities and their corresponding AI / ML model parameters may be referred to as "predefined" because they are defined and known to the UE and network, rather than being signaled to the UE by the network. In some aspects, the defined functionalities may include multiple primary functionalities, such as beam prediction, CSF, and / or positioning. Each of these primary functionalities may also include multiple functionalities for the next level, which may be referred to as "sub-functionalities." For example, the primary functionality of beam prediction may also include sub-functionalities for time domain (TD) beam prediction, spatial domain (SD) beam prediction, and frequency domain (FD) beam prediction. Each of these sub-functionalities may also include one or more lower-level functionalities. For example, a sub-functionality may include multiple functionalities at the next level, which may be referred to as "sub-sub-functionalities." The operating parameters of these functionalities may include parameters defining specific UE behavior, expected AI / ML input / output, and expected assistance information or reference signals from the base station. The operating parameters may also be defined in wireless standards. As used herein, the term "standard" may refer to a set of technical specifications or guidelines that have been agreed upon within the industry for wireless communications.

[0095] In some aspects, a UE may receive an indication from a base station requesting the UE to switch to one or more of the defined functionalities (e.g., one or more of the primary functionality or sub-functionality) such that operating parameters associated with the requested functionality (e.g., the requested primary functionality or sub-functionality) may be switched implicitly without further signaling.

[0096] In some aspects, defined functionality (e.g., functionality defined in a wireless standard) may include one or more of: TD / SD / FD beam prediction, AI / ML-based CSI feedback, and / or AI / ML-based positioning.

[0097] Figure 5 is a diagram 500 illustrating example AI / ML functionality according to various aspects of the present disclosure. Figure 5As shown, AI / ML functionality 502 may include several main functionalities 510. Main functionalities 510 may include, for example, beam prediction functionality 512, CSF functionality 514, and positioning functionality 516. Each of main functionalities 510 may also include lower-level functionality, which may be referred to as sub-functionality 520. For example, beam prediction functionality 512 may include sub-functionality for TD beam prediction 522, SD beam prediction 524, and FD beam prediction 526. Each of sub-functionality 520 may also include lower-level functionality, which may be referred to as, for example, sub-subfunctionality 530. Upon receiving a command to switch to a specific functionality, for example, from a base station, the UE may select a corresponding parameter set for performing the specific functionality without further signaling. For example, a parameter set may include parameters associated with certain CSI report settings for reporting prediction results for the specific functionality.

[0098] Functionality handover commands to the UE can be implemented in various ways. In some aspects, within defined AI / ML functionality (e.g., functionality defined in a wireless standard), the UE can be configured with a subset of the defined functionality via RRC. The defined subset of functionality can be based on the capabilities supported by the UE. For example, the UE can report its support for an AI / ML model as a UE capability provided to the network during initial access. The base station can configure the UE for one or more of the AI / ML functionalities supported by the UE. For example, the base station can configure the UE for a subset of functionality from the functionality that the UE can support based on the capabilities reported by the UE. For example, the base station can signal the defined functionality or sub-functionality to the UE in RRC signaling. The RRC signaling can indicate the ML of the functionality with reference to a defined AI / ML model parameter for the indicated functionality or sub-functionality. In some aspects, a maximum number of sub-functionalities that the base station can configure for a certain primary functionality can be defined, and the base station can determine which of the maximum number of sub-functionalities to be configured by RRC for the UE.

[0099] In some aspects, the network may indicate or activate one or more defined AI / ML functionalities (e.g., defined in a wireless standard) previously indicated in RRC signaling to the UE via one or more MAC-CEs, thereby activating one or more subsets of these RRC-configured, defined functionalities. In one example, candidate AI / ML functionalities may be first indexed according to defined rules (e.g., rules defined in the wireless standard or defined rules otherwise known to the UE), and the UE may be instructed via a single MAC-CE, which may include the functional IDs of the functionalities to be activated, to activate one or more subsets of these functionalities. In another example, the UE may be instructed to activate one or more subsets of these functionalities via multiple MAC-CEs. Each MAC-CE format of the multiple MAC-CEs may be associated with a dedicated primary functionality or a sub-functionality within the primary functionality. In some examples, the MAC-CE may identify the primary functionality or sub-functionality to which the MAC-CE is associated, and the remaining payload may further indicate the candidate functional IDs to be activated.

[0100] In some aspects, the network may further instruct the UE to switch to or apply specific AI / ML functionality within a set configured via RRC and / or a subset activated via MAC-CE. For example, among defined AI / ML functionality (e.g., functionality defined in a wireless standard), RRC-configured functionality, and / or MAC-CE-activated functionality, the UE may be further instructed to switch to one or more subsets of these functionalities via one or more MAC-CEs or one or more DCIs. In one example, candidate AI / ML functionalities may first be indexed, for example, based on a defined rule or configuration, and the UE may receive a single MAC-CE or DCI that may include the functional ID of the functionality to be switched to. In another example, the UE may be instructed to switch to one or more subsets of these functionalities via multiple MAC-CEs or DCIs. Each of the multiple MAC-CE (or DCI) formats may be associated with a dedicated primary functionality or a sub-functionality within the primary functionality. In some examples, the MAC-CE (or DCI) may identify the primary functionality or sub-functionality to which the MAC-CE (or DCI) is associated, and the remaining payload may then further indicate candidate functionality IDs of the functionality to be switched to.

[0101] In some aspects, one or more of the signaling frameworks described above (e.g., RRC configuration for a defined set of functionalities with associated AI / ML models, MAC-CE activation for a subset of the RRC-configured set, and / or MAC-CE or DCI indication for applying or switching to one or more of the configured and / or activated functionalities and corresponding AI / ML model parameters) may be combined to signal a UE to switch to specific AI / ML functionality. For example, a UE may switch to specific, defined AI / ML functionality based on MAC-CE or DCI without requiring RRC configuration or MAC-CE activation. A UE may switch to specific, defined AI / ML functionality based on MAC-CE activation and DCI indication without requiring RRC configuration of an initial set. A UE may switch to specific, defined AI / ML functionality based on RRC configuration and MAC-CE or DCI indication. A UE may switch to specific, defined AI / ML functionality based on an RRC-configured set, MAC-CE activation for a subset, and a DCI indication of an activated subset from the RRC-configured set of defined AI / ML functionalities.

[0102] The operating parameters associated with the AI / ML functionality may be defined, for example, in a wireless standard. In other examples, the AI / ML functionality may include channel characteristic prediction, beam prediction, positioning prediction, etc. The operating parameters may include at least one of the following: parameters associated with UE AI / ML inference behavior, parameters associated with expected AI / ML input or output, and parameters associated with expected assistance information from the network or RS.

[0103] Parameters associated with UE AI / ML inference behavior may include prediction parameters indicating prediction types associated with multiple defined ML functions. For example, the prediction type may indicate whether beam prediction is performed only in the spatial domain, in a combination of the spatial and time domains, only in the frequency domain, in a combination of the frequency and time domains, in a combination of the spatial and frequency domains, or in a combination of the spatial, frequency, and time domains.

[0104] Parameters associated with UE AI / ML interference behavior may also include feedback parameters indicating feedback for channel characteristic predictions associated with multiple defined ML functions. For example, the feedback parameters may indicate whether the UE should feedback the predicted L1-RSRP, the predicted L1-SINR, the channel quality indicator (CQI) of the channel between the UE and the network entity, the rank indicator (RI), or the top K predicted target resources among candidate predicted target resources based on their predicted L1-RSRP / L1-SINR.

[0105] Parameters associated with UE AI / ML interference behavior may also include an accuracy parameter that indicates the expected prediction accuracy or confidence level associated with a plurality of defined ML functions. For example, the accuracy parameter may indicate whether the UE is to provide feedback regarding the confidence level associated with its reported prediction results. The accuracy parameter may also indicate the expected prediction accuracy or confidence level to be met by the UE, and if the expected prediction accuracy or confidence level cannot be met during inference, indicate whether the UE is to report such issues.

[0106] Parameters associated with expected AI / ML inputs or outputs may include a first resource indicator for a first measurement resource for an input associated with a plurality of defined ML functions, a second resource indicator for a predicted target resource for an output associated with the plurality of defined ML functions, and a characteristic parameter indicating a relationship between the first measurement resource and the predicted target resource. The first resource indicator may include the number of measurement resources to be used for the AI / ML input, and whether the number of measurement resources will vary across different measurement occasions, and if so, the pattern of variation of the measurement resources. The second resource indicator may indicate the number of predicted target resources or beams to be derived from the AI / ML output, and whether such output will vary across different predicted target occasions, and if so, the pattern of variation of the predicted target. The characteristic parameter may indicate whether the measurement resource is a subset of the predicted target resource, or whether the measurement resource does not overlap with the predicted target resource in terms of beam pointing direction and / or beam width.

[0107] Parameters associated with expected assistance information from the network or the RS may include a signaling indicator for an indication request associated with a plurality of defined ML functions, a target indicator for transmitting a target result associated with the plurality of defined ML functions, and a measurement indicator for measuring a target result associated with the plurality of defined ML functions.

[0108] In some aspects, the signaling indicator may indicate whether the UE can expect further signaling from the base station regarding transmit beam information for the measurement resources and / or prediction targets, and if so, whether the signaling is based on explicit beam width and pointing direction, based on transmit panel structure and precoding codebook, or based on implicit indication (e.g., based solely on beam neighbor information for the measurement resources and / or prediction targets, or based on an association of a linear combination of the measurement resources and prediction targets). The target indicator may indicate whether the UE can expect the target prediction resource or beam to be transmitted (although less frequently than the measurement resource used to identify the prediction result), or not at all. If the UE can expect the target prediction resource / beam to be transmitted, the target indicator may also indicate the frequency with which the UE can expect such transmission. The measurement indicator may indicate whether the UE is expected to measure the actually transmitted prediction target (for performance monitoring) via SSB, via CSI-RS, or via DMRS.

[0109] In some aspects, operational parameters associated with AI / ML functionality may include parameters specific to the particular AI / ML functionality. For example, when the ML functionality includes TD beam prediction, parameters associated with UE AI / ML inference behavior may also include one or more of: a history parameter indicating an association between historical measurement resources and a set of spatial transmit filters associated with the target ML functionality; and an accuracy adjustment parameter indicating an adjustment to the accuracy or confidence level for different future prediction occasions for the ML functionality. In some aspects, the history parameter may indicate whether the UE expects measurement resources for historical measurements to be associated with the same set of spatial transmit filters for the same measurement resources across different measurement occasions. The accuracy adjustment parameter may indicate different accuracy or confidence levels for different future prediction occasions (e.g., a lower accuracy or confidence level may be associated with another prediction occasion).

[0110] For example, when the ML functionality includes TD beam prediction, the parameters associated with the expected AI / ML inputs and outputs may further include one or more of: an interval parameter indicating the interval between adjacent measurement opportunities and between adjacent prediction opportunities; and a future prediction parameter indicating the number of future prediction opportunities. In some aspects, the interval parameter may indicate at least one of the interval between adjacent measurement opportunities and the interval between adjacent prediction opportunities. The future prediction parameter may indicate the number of future prediction opportunities that the UE predicts based on a specific number of measurement opportunities.

[0111] In some aspects, when the ML function includes FD beam prediction, the parameters associated with the UE AI / ML inference behavior may further include one or more of the following: a frequency range (FR) or CC that conveys the second measurement resource and the prediction target; and an accuracy adjustment parameter that indicates an adjustment to the accuracy or confidence level for different FRs or CCs of the ML function. In some examples, the accuracy adjustment parameter may indicate different accuracy or confidence levels for different FR / CC distances (e.g., a larger difference between FRs / CCs corresponds to a lower accuracy or confidence level).

[0112] In some aspects, when the ML function includes FD beam prediction, the parameters associated with the expected AI / ML input and output may further include a TD association parameter indicating a first association of a prediction target with a TD timing of the ML function. In some examples, the TD association parameter may indicate whether the prediction target is with respect to the same TD timing associated with the measurement resource (e.g., for pure FD prediction) or with respect to a future TD timing (e.g., for both FD and TD prediction).

[0113] In some aspects, when the ML functionality includes FD beam prediction, the parameters associated with the expected base station assistance information or RS may further include: a co-location indicator indicating a second association of a third measurement resource with an antenna; and a location parameter indicating location information of the antenna of the network entity. In some examples, the co-location indicator may indicate whether the measurement resource and prediction target are associated with a co-located antenna or a non-positioned antenna. If the measurement resource and prediction target are associated with a non-positioned antenna, the location parameter may indicate specific location / distance information for such antenna.

[0114] In some aspects, for certain primary functionalities or sub-functionalities for UE-side channel characteristic prediction (including TD / SD / FD beam prediction), the UE may be configured with a single CSI reporting setting. The CSI reporting setting may additionally include an ID of the corresponding primary functionality or sub-functionality. In some examples, coordinated multipoint reception (CMR) associated with the CSI reporting setting may be used as input to identify the predicted channel characteristics. The parameter reportQuantity associated with the CSI reporting setting may include the predicted channel characteristics. In some aspects, upon receiving an AI / ML functionality switch command associated with the primary functionality or sub-functionality ID under consideration, the parameters associated with the CSI reporting setting may be adaptively switched without further signaling from, for example, a base station.

[0115] In one example, the CSI reporting settings may be associated with a set of sub-functionalities for TD beam prediction. beam consecutive historical measurement opportunities, taking into account the basic period in adjacent prediction / measurement opportunities = ms, reports about a specific reporting instance via the associated CSI report relative to the predicted future L1-RSRP beam For example, refer to Figure 5 For TD beam prediction 522, the UE may report about 8 future opportunities for 4 historical measurements (at 532), about 8 future opportunities for 8 historical measurements (at 534), or about 8 future opportunities for 16 historical measurements (at 536) via associated CSI reporting relative to the predicted future L1-RSRP.

[0116] The set ID of the functionality can be configured in the CSI reporting settings. The combination of defines a set of sub-functionalities in the wireless standard. For example, CMRs may be configured in order to have more than actually expected, and the previous maximum value may be expected based on the network's indication of the sub-functionality. indivual.

[0117] For a set of sub-functionalities, the UE may be expected to be further instructed with auxiliary information regarding beam pointing direction neighboring information related to CMR and predicted target resources, without expecting further explicit information. Therefore, the set of sub-functionalities may be further defined in the wireless standard with this type of auxiliary information.

[0118] For a set of sub-functionalities, whether the UE can expect to receive the actual RS associated with the prediction target can be changed as a parameter of the sub-functionality. Therefore, the sub-functionality can be further defined in the wireless standard by being controllable by whether the actual RS associated with the prediction target is to be received. This can be further based on the fact that the wireless standard can define that the CSI reporting setting will have pre-configured potential CSI-RS resources associated with the prediction target resources to convey the potential actual transmission of such prediction target. However, if the network's indication of the sub-functionality implies that there is no such actual RS transmission, then the reception of such CSI-RS resources may not be expected.

[0119] Alternatively, the base station may use a dedicated MAC-CE. The first portion of the MAC-CE payload may contain a set ID for the sub-functionality, and the second portion of the MAC-CE payload may contain the aforementioned parameters or a sub-functionality ID from the set. The UE can adaptively change parameters associated with the CSI reporting configuration without receiving a dedicated command to change the CSI reporting configuration.

[0120] Figure 6 6 is a call flow diagram illustrating a method of wireless communication according to various aspects of the present disclosure. Although various aspects are described with respect to base station 604, these aspects may be performed by base stations in an aggregation and / or by one or more components of base station 604 (e.g., such as CU 110, DU 130, and / or RU 140).

[0121] like Figure 6 As shown, UE 602 may send a UE capability indication to base station 604. The UE capability indication may indicate a supported set of ML functions from defined ML functions. For example, referring to Figure 5 , the supported set of ML functions may be TD beam prediction 522 , SD beam prediction 524 and FD beam prediction 526 .

[0122] At 608, UE 602 may receive RRC signaling from base station 604 indicating a first subset of ML functions from a plurality of defined ML functions. Figure 5, a first subset of ML functions may be TD beam prediction 522 .

[0123] At 610, UE 602 may receive RRC signaling from base station 604 indicating an RRC-configured subset of ML functions from a plurality of defined ML functions. Figure 5 , the RRC configured subset of ML functionality may be SD beam prediction 524.

[0124] At 612, UE 602 may receive one or more MAC-CEs activating a subset of the defined ML functions from base station 604. For example, referring to

[0125] At 614, the UE 602 may receive a first indication from the base station 604 indicating a target set of ML functions from among the plurality of defined ML functions. Figure 5 , the first indication may indicate sub-sub-functionalities 532 and 534 as a target set of ML functions.

[0126] At 616, the UE 602 may configure one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters based on the first indication. Figure 5 If the target set of ML functions includes sub-sub-functionalities 532 and 534, the UE may configure one or more operating parameters based on the sub-sub-functionalities 532 and 534. For example, the one or more operating parameters may include parameters of a CSI report setting for reporting prediction results.

[0127] At 618, UE 602 may communicate with base station 604 based on one or more operating parameters associated with a target set of ML functions indicated from a plurality of defined ML functions. Figure 5 , based on the one or more operating parameters configured at 616 , the UE may communicate with base station 604 .

[0128] At 620, UE 602 may adjust parameters of the CSI reporting settings based on the first indication. Figure 5 If the first indication indicates that the target set of ML functions includes sub-sub-functionalities 532 and 534, the UE may adjust parameters of the CSI report setting based on the first indication.

[0129] At 622, UE 602 may send a CSI report to base station 604. For example, referring to Figure 5 If the first indication indicates that the target set of ML functions includes sub-sub-functionalities 532 and 534 , the UE may send a CSI report to the base station, and the CSI report settings of the CSI report may be based on the sub-sub-functionalities 532 and 534 .

[0130] Figure 7 700 is a flowchart illustrating a method for wireless communication at a UE according to various aspects of the present disclosure. The method may be performed by a UE. The UE may be UE 104, 350, 602, or Figure 11 The method provides a new signaling framework dedicated to AI / ML model switching and enables the UE to switch the parameters of the ML function with reduced total signaling overhead. Therefore, the method improves the efficiency of wireless communication.

[0131] like Figure 7 As shown, at 702, the UE may receive a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions. The network entity may be Figure 1 A base station or a component of a base station in an access network, or a core network component (e.g., base station 102, 310, 604; or Figure 11 Network entity 1102 in a hardware specific implementation). Figure 5 and Figure 6 Various aspects of the steps in conjunction with flowchart 700 are illustrated. For example, referring to Figure 6 , the UE 602 may receive a first indication at 614 from a network entity (base station 604) indicating a target set of ML functions from a plurality of defined ML functions. Figure 5 , the target set of ML functions may include sub-sub-functionalities 532 , 534 , and 536 . In some aspects, 702 may be performed by AI / ML function switch component 198 .

[0132] At 704, the UE may communicate with a network entity based on one or more operating parameters associated with a target set of ML functions indicated from a plurality of defined ML functions. Figure 6 , UE 602 may communicate with a network entity (base station 604) based on one or more operating parameters associated with a target set of ML functions indicated from a plurality of defined ML functions at 618. In some aspects, 704 may be performed by AI / ML function switching component 198.

[0133] Figure 8 800 is a flowchart illustrating a method for wireless communication at a UE according to various aspects of the present disclosure. The method may be performed by a UE. The UE may be UE 104, 350, 602, or Figure 11 The method provides a new signaling framework dedicated to AI / ML model switching and enables the UE to switch the parameters of the ML function with reduced total signaling overhead. Therefore, the method improves the efficiency of wireless communication.

[0134] like Figure 8As shown, at 810, the UE may receive a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions. The network entity may be Figure 1 A base station or a component of a base station in an access network, or a core network component (e.g., base station 102, 310, 604; or Figure 11 Network entity 1102 in a hardware specific implementation). Figure 5 and Figure 6 Various aspects of the steps in conjunction with flowchart 800 are illustrated. For example, referring to Figure 6 , the UE 602 may receive a first indication at 614 from a network entity (base station 604) indicating a target set of ML functions from a plurality of defined ML functions. Figure 5 , the target set of ML functions may include sub-sub-functionalities 532 , 534 , and 536 . In some aspects, 810 may be performed by AI / ML function switch component 198 .

[0135] At 814, the UE may communicate with a network entity based on one or more operating parameters associated with a target set of ML functions indicated from a plurality of defined ML functions. Figure 6 , UE 602 may communicate with a network entity (base station 604) based on one or more operating parameters associated with a target set of ML functions indicated from a plurality of defined ML functions at 618. In some aspects, 814 may be performed by AI / ML function switching component 198.

[0136] At 812, the UE may configure one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters based on the first indication. Figure 6 , the UE 602 may configure one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters based on the first indication at 616. In some aspects, 812 may be performed by the AI / ML function switching component 198.

[0137] In some aspects, the plurality of defined ML functions may include one or more of: a beam prediction function; a CSF function; or a positioning function. Figure 5 , the plurality of defined ML functions may include one or more of: a beam prediction function 512 ; a CSF function 514 ; or a positioning function 516 .

[0138] In some aspects, to receive the first indication, the UE may be configured to receive, via one or more MAC-CEs or DCIs, a first indication indicating a target set of ML functions. Figure 6When the UE 602 receives the first indication at 614, the UE may receive the first indication via one or more MAC-CEs or DCIs.

[0139] In some aspects, the ML functions in the plurality of defined ML functions may be indexed and each ML function may be associated with a functionality ID according to defined rules. Figure 5 , ML functions (e.g., TD beam prediction 522, SD beam prediction 524, and FD beam prediction 526) among multiple defined ML functions can be indexed, and according to defined rules, each ML function (e.g., TD beam prediction 522, SD beam prediction 524, and FD beam prediction 526) can be associated with a functional ID.

[0140] In some aspects, a single MAC-CE or a single DCI may include a functional ID corresponding to an ML function in a target set of ML functions. Figure 6 When the UE 602 receives the first indication indicating the target set of ML functions through the MAC-CE or DCI at 614, the UE 602 may receive a single MAC-CE or a single DCI including a functional ID corresponding to the ML function in the target set of ML functions.

[0141] In some aspects, the one or more MAC-CEs or DCIs include multiple MAC-CEs or multiple DCIs. Each of the multiple MAC-CEs or multiple DCIs may have a format associated with a corresponding ML functionality, and each of the multiple MAC-CEs or multiple DCIs may include a functional ID of at least one ML function in a target set of ML functions associated with the corresponding ML functionality. For example, referring to Figure 6 When the UE 602 receives the first indication indicating the target set of ML functions through a MAC-CE or a DCI at 614, the UE 602 may receive multiple MAC-CEs or multiple DCIs. Each of the multiple MAC-CEs or the multiple DCIs may have a format associated with a corresponding ML functionality, and each of the multiple MAC-CEs or the multiple DCIs may include a functional ID of at least one ML function in the target set of ML functions associated with the corresponding ML functionality.

[0142] At 804, the UE may receive RRC signaling from a network entity indicating a first subset of ML functions from a plurality of defined ML functions. The target set of ML functions may be within the first subset of ML functions indicated in the RRC signaling. For example, referring to Figure 6At 608, UE 602 may receive RRC signaling from a network entity (base station 604) indicating a first subset of ML functions from a plurality of defined ML functions. The target set of ML functions (at 614) may be within the first subset of ML functions indicated in the RRC signaling. In some aspects, 804 may be performed by AI / ML function switching component 198.

[0143] At 802, the UE may send a UE capability indication to a network entity indicating a supported set of ML functions from a defined ML function. The first subset of ML functions may be within the supported set of ML functions. For example, referring to Figure 6 , UE 602 may send a UE capability indication indicating a supported set of ML functions from a defined set of ML functions to a network entity (base station 604) at 606. The first subset of ML functions (at 614) may be within the supported set of ML functions. In some aspects, 802 may be performed by AI / ML function switching component 198.

[0144] At 808, the UE may receive one or more MAC-CEs from the network entity that activate a subset of the defined ML functions. The target set of ML functions may be within the subset of ML functions activated by the one or more MAC-CEs. For example, referring to Figure 6 At 612, UE 602 may receive one or more MAC-CEs from a network entity (base station 604) activating a subset of the defined ML functions. The target set of ML functions (at 614) may be within the subset of ML functions activated (at 612) by the one or more MAC-CEs. In some aspects, 808 may be performed by AI / ML function switching component 198.

[0145] In some aspects, the one or more MAC-CEs may include a single MAC-CE that includes a functionality ID corresponding to an ML function in a subset of ML functions. Figure 6 At 612, the one or more MAC-CEs may include a single MAC-CE including a functionality ID corresponding to an ML function in the subset of ML functions.

[0146] In some aspects, the one or more MAC-CEs may include a plurality of MAC-CEs. Each MAC-CE in the plurality of MAC-CEs may have a MAC-CE format associated with a corresponding ML functionality, and each MAC-CE in the plurality of MAC-CEs may include an identifier for at least one ML function in a target set of ML functions associated with the corresponding ML functionality. For example, referring to Figure 6The one or more MAC-CEs may include a plurality of MAC-CEs at 612. Each MAC-CE in the plurality of MAC-CEs may have a MAC-CE format associated with a corresponding ML functionality, and each MAC-CE in the plurality of MAC-CEs may include (at 612) an identifier of at least one ML function in a target set of ML functions associated with the corresponding ML functionality.

[0147] At 806, the UE may receive RRC signaling from the network entity indicating an RRC-configured subset of ML functions from a plurality of defined ML functions. The subset of ML functions may be within the RRC-configured subset of ML functions. For example, referring to Figure 6 At 610, UE 602 may receive RRC signaling from a network entity (base station 604) indicating an RRC-configured subset of ML functions from a plurality of defined ML functions. The subset of ML functions (at 612) may be within the RRC-configured subset of ML functions. In some aspects, 806 may be performed by AI / ML function switching component 198.

[0148] In some aspects, the plurality of defined operating parameters may include one or more of: a first set of parameters related to UE ML interference behavior of the plurality of ML functions; a second set of parameters related to expected ML inputs and outputs of the plurality of ML functions; and a third set of parameters related to assistance information. Figure 6 When the UE 602 configures one or more operating parameters from a plurality of defined operating parameters at 616, the plurality of defined operating parameters may include one or more of: a first set of parameters related to UE ML interference behavior of a plurality of ML functions; a second set of parameters related to expected ML inputs and outputs of the plurality of ML functions; and a third set of parameters related to assistance information.

[0149] In some aspects, a first set of parameters may include one or more of: a prediction parameter indicating a prediction type associated with a plurality of defined ML functions; a feedback parameter indicating feedback for channel characteristic predictions associated with a plurality of defined ML functions; and an accuracy parameter indicating an expected prediction accuracy or confidence level associated with a plurality of defined ML functions. A second set of parameters may include one or more of: a first resource indicator for a first measurement resource for an input associated with a plurality of defined ML functions; a second resource indicator for a predicted target resource for an output associated with a plurality of defined ML functions; and a characteristic parameter indicating a relationship between the first measurement resource and the predicted target resource. A third set of parameters may include one or more of: a signaling indicator for an indication request associated with a plurality of defined ML functions; a target indicator for sending a target result associated with a plurality of defined ML functions; and a measurement indicator for measuring a target result associated with a plurality of defined ML functions. For example, with reference to Figure 6 When the UE 602 configures one or more operating parameters from the plurality of defined operating parameters at 616, a first set of parameters (included in the plurality of defined operating parameters) may include one or more of the following: a prediction parameter indicating a prediction type associated with the plurality of defined ML functions; a feedback parameter indicating feedback for channel characteristic predictions associated with the plurality of defined ML functions; and an accuracy parameter indicating an expected prediction accuracy or confidence level associated with the plurality of defined ML functions. A second set of parameters (included in the plurality of defined operating parameters) may include one or more of the following: a first resource indicator for a first measurement resource of an input associated with the plurality of defined ML functions; a second resource indicator for a predicted target resource of an output associated with the plurality of defined ML functions; and a characteristic parameter indicating a relationship between the first measurement resource and the predicted target resource. The third set of parameters (which are included in the plurality of defined operating parameters) may include one or more of: a signaling indicator for an indication request associated with the plurality of defined ML functions; a target indicator for sending a target result associated with the plurality of defined ML functions; and a measurement indicator for measuring a target result associated with the plurality of defined ML functions.

[0150] In some aspects, the prediction type may include one or more of: SD beam prediction; TD beam prediction; and FD beam prediction. Figure 5 , the prediction type may include one or more of the following: SD beam prediction 524 ; TD beam prediction 522 ; and FD beam prediction 526 .

[0151] In some aspects, the channel characteristic prediction may include one or more of the following: predicted L1-RSRP; predicted L1-SINR; CQI of the channel between the UE and the network entity; RI; or predicted top K prediction target resources, the predicted top K prediction target resources being associated with the predicted L1-RSRP and the predicted L1-SINR. For example, referring to Figure 6 At 616, the channel characteristic prediction (which is included in the first set of parameters) may include one or more of the following: a predicted L1-RSRP; a predicted L1-SINR; a CQI of a channel between the UE and the network entity; an RI; or predicted top K prediction target resources, the predicted top K prediction target resources being associated with the predicted L1-RSRP and the predicted L1-SINR.

[0152] In some aspects, the ML function may include TD beam prediction, and the first set of parameters may further include one or more of the following: a history parameter indicating an association of historical measurement resources and a set of spatial transmit filters associated with the target ML function; and an accuracy adjustment parameter indicating an adjustment to the accuracy or confidence level of different future prediction opportunities for the ML function. The second set of parameters may further include one or more of the following: an interval parameter indicating an interval between adjacent measurement opportunities and between adjacent prediction opportunities; and a future prediction parameter indicating a number of future prediction opportunities. For example, with reference to Figure 6 When the multiple defined ML functions include TD beam prediction, the first set of parameters (which the UE 602 may configure at 616) may further include one or more of the following: a history parameter indicating an association between historical measurement resources and a set of spatial transmit filters associated with the target ML function; and an accuracy adjustment parameter indicating an adjustment to the accuracy or confidence level of different future prediction opportunities for the ML function. The second set of parameters may further include one or more of the following: an interval parameter indicating an interval between adjacent measurement opportunities and between adjacent prediction opportunities; and a future prediction parameter indicating the number of future prediction opportunities.

[0153] In some aspects, the ML function may include FD beam prediction, and the first set of parameters may also include one or more of the following: a FR or CC that conveys a second measurement resource and a prediction target; an accuracy adjustment parameter that indicates an adjustment to the accuracy or confidence level for different FRs or CCs for the ML function. The second set of parameters may also include: a TD association parameter that indicates a first association of the prediction target with a TD opportunity of the ML function. The third set of parameters may also include one or more of the following: a co-location indicator that indicates a second association of a third measurement resource with an antenna; and a location parameter that indicates location information of an antenna of a network entity. For example, with reference to Figure 6 When the multiple defined ML functions include FD beam prediction, the first set of parameters (which the UE 602 may configure at 616) may further include one or more of the following: a frame rate or a center frequency (CC) that conveys the second measurement resource and the prediction target; and an accuracy adjustment parameter that indicates an adjustment to the accuracy or confidence level for different frame rates or CCs for the ML function. The second set of parameters may further include a time delay (TD) association parameter that indicates a first association of the prediction target with a TD opportunity for the ML function. The third set of parameters may further include one or more of the following: a co-location indicator that indicates a second association of the third measurement resource with an antenna; and a location parameter that indicates location information of the antenna of the network entity.

[0154] At 818, the UE may send a CSI report to the network entity. Parameters of the CSI report settings for the CSI report may include: a functional ID of a primary function associated with a target set of ML functions; and prediction information associated with the target set of ML functions. For example, referring to Figure 6 UE 602 may transmit a CSI report to a network entity (base station 604) at 622. Parameters of a CSI report setting for the CSI report may include: a functional ID of a primary function associated with a target set of ML functions; and prediction information associated with the target set of ML functions. In some aspects, 818 may be performed by AI / ML function switching component 198.

[0155] In some aspects, the prediction information may include: a predicted channel characteristic of a channel between the UE and a network entity; a CMR, wherein the CMR may be used as an input to identify the predicted channel characteristic; and a set of prediction resources, the set of prediction resources identifying a prediction target for the channel characteristic. Figure 6At 622, the prediction information (one of the parameters of the CSI report setting of the CSI report) may include: a predicted channel characteristic of a channel between the UE and the network entity; a CMR, wherein the CMR may be used as an input to identify the predicted channel characteristic; and a prediction resource set, which identifies a prediction target of the channel characteristic.

[0156] At 816, the UE may adjust parameters of the CSI reporting configuration based on the first indication. These parameters may be adjusted without signaling from the network entity. For example, referring to Figure 6 , UE 602 may adjust parameters of the CSI reporting settings based on the first indication at 620. These parameters may be adjusted without signaling from the network entity (base station 604). In some aspects, 816 may be performed by AI / ML functional switching component 198.

[0157] Figure 9 900 is a flowchart illustrating a method for wireless communication at a network entity according to various aspects of the present disclosure. The method may be performed by a network entity. The network entity may be Figure 1 A base station or a component of a base station in an access network, or a core network component (e.g., base station 102, 310, 604; or Figure 11 The method provides a new signaling framework dedicated to AI / ML model switching and enables the UE to switch the parameters of the ML function with reduced total signaling overhead. Therefore, the method improves the efficiency of wireless communication.

[0158] like Figure 9 As shown, at 902, the network entity may provide a first indication to the UE indicating a target set of ML functions from a plurality of defined ML functions. The UE may be UE 104, 350, 602, or Figure 11 The device 1104 is implemented in hardware. Figure 5 and Figure 6 Various aspects of the steps in conjunction with flowchart 900 are illustrated. For example, referring to Figure 6 , the network entity (base station 604) may provide a first indication indicating a target set of ML functions from among a plurality of defined ML functions to the UE 602 at 614. In some aspects, 902 may be performed by the AI / ML function switching component 199.

[0159] At 904, the network entity may communicate with the UE based on one or more operating parameters associated with the target set of ML functions. Figure 6At 618 , the network entity (base station 604 ) can communicate with UE 602 based on one or more operating parameters associated with the target set of ML functions. In some aspects, 904 can be performed by AI / ML function switching component 199 .

[0160] Figure 10 1000 is a flowchart illustrating a method for wireless communication at a network entity according to various aspects of the present disclosure. The method may be performed by a network entity. The network entity may be Figure 1 A base station or a component of a base station in an access network, or a core network component (e.g., base station 102, 310, 604; or Figure 11 The method provides a new signaling framework dedicated to AI / ML model switching and enables the UE to switch the parameters of the ML function with reduced total signaling overhead. Therefore, the method improves the efficiency of wireless communication.

[0161] like Figure 10 As shown, at 1002, a network entity may provide a first indication to a UE indicating a target set of ML functions from a plurality of defined ML functions. The UE may be UE 104, 350, 602, or Figure 11 The device 1104 is implemented in hardware. Figure 5 and Figure 6 Various aspects of the steps in conjunction with flowchart 1000 are illustrated. For example, referring to Figure 6 , the network entity (base station 604) may provide a first indication indicating a target set of ML functions from among a plurality of defined ML functions to the UE 602 at 614. In some aspects, 1002 may be performed by the AI / ML function switching component 199.

[0162] At 1004, the network entity may communicate with the UE based on one or more operating parameters associated with the target set of ML functions. Figure 6 At 618 , the network entity (base station 604 ) can communicate with UE 602 based on one or more operating parameters associated with the target set of ML functions. In some aspects, 1004 can be performed by AI / ML function switching component 199 .

[0163] In some aspects, at 1006, the first indication may indicate to the UE to configure one or more operating parameters associated with the target set of ML functions. Figure 6 , the first indication (at 614 ) may indicate to the UE 602 to configure, at 616 , one or more operating parameters associated with the target set of ML functions.

[0164] In some aspects, at 1008, the plurality of defined ML functions may include one or more of: a beam prediction function; a CSF function; or a positioning function. Figure 5 , the plurality of defined ML functions may include one or more of: a beam prediction function 512 ; a CSF function 514 ; or a positioning function 516 .

[0165] In some aspects, to provide the first indication, the network entity may be configured to provide, via one or more MAC-CEs or DCIs, a first indication indicating a target set of ML functions. Figure 6 , when the network entity (base station 604) provides (at 614) the first indication, the network entity (base station 604) may provide the first indication via one or more MAC-CEs or DCIs.

[0166] In some aspects, the ML functions in the plurality of defined ML functions may be indexed and each ML function may be associated with a functionality ID according to defined rules. Figure 5 , ML functions (e.g., TD beam prediction 522, SD beam prediction 524, and FD beam prediction 526) among multiple defined ML functions can be indexed, and according to defined rules, each ML function (e.g., TD beam prediction 522, SD beam prediction 524, and FD beam prediction 526) can be associated with a functional ID.

[0167] In some aspects, a single MAC-CE or a single DCI may include a functional ID corresponding to an ML function in a target set of ML functions. Figure 6 , when the network entity (base station 604) provides the first indication (at 614), the network entity (base station 604) may provide the first indication via a single MAC-CE or a single DCI, and the single MAC-CE or the single DCI may include a functional ID corresponding to the ML function in the target set of ML functions.

[0168] Figure 11Diagram 1100 illustrates an example of a hardware implementation for an apparatus 1104. Apparatus 1104 may be a UE, a component of a UE, or may implement UE functionality. In some aspects, apparatus 1104 may include at least one cellular baseband processor (or processing circuit) 1124 (also referred to as a modem) coupled to one or more transceivers 1122 (e.g., a cellular RF transceiver). Cellular baseband processor (or processing circuit) 1124 may include at least one on-chip memory (or memory circuit) 1124′. In some aspects, apparatus 1104 may also include one or more subscriber identity module (SIM) cards 1120, and at least one application processor (or processing circuit) 1106 coupled to a secure digital (SD) card 1108 and a screen 1110. Application processor (or processing circuit) 1106 may include on-chip memory (or memory circuit) 1106′. In some aspects, device 1104 may also include a Bluetooth module 1112, a WLAN module 1114, an SPS module 1116 (e.g., a GNSS module), one or more sensor modules 1118 (e.g., a barometric pressure sensor / altimeter; a motion sensor such as an inertial measurement unit (IMU), a gyroscope, and / or an accelerometer; light detection and ranging (LIDAR), radio-aided detection and ranging (RADAR), sound navigation and ranging (SONAR), a magnetometer, audio, and / or other technologies for positioning), an additional memory module 1126, a power source 1130, and / or a camera 1132. The Bluetooth module 1112, the WLAN module 1114, and the SPS module 1116 may include an on-chip transceiver (TRX) (or, in some cases, only a receiver (RX)). The Bluetooth module 1112, the WLAN module 1114, and the SPS module 1116 may include their own dedicated antennas and / or utilize an antenna 1180 for communication. Cellular baseband processor (or processing circuitry) 1124 communicates with UE 104 and / or RUs associated with network entity 1102 via transceiver 1122 via one or more antennas 1180. Cellular baseband processor (or processing circuitry) 1124 and application processor (or processing circuitry) 1106 may each include computer-readable media / memory (or memory circuitry) 1124', 1106', respectively. Additional memory module 1126 may also be considered a computer-readable medium / memory (or memory circuitry). Each computer-readable medium / memory (or memory circuitry) 1124', 1106', 1126 may be non-transitory. Cellular baseband processor (or processing circuitry) 1124 and application processor (or processing circuitry) 1106 are each responsible for general processing, including the execution of software stored on the computer-readable media / memory (or memory circuitry).The software, when executed by cellular baseband processor (or processing circuitry) 1124 / application processor (or processing circuitry) 1106, causes cellular baseband processor (or processing circuitry) 1124 / application processor (or processing circuitry) 1106 to perform the various functions described above. Cellular baseband processor (or processing circuitry) 1124 and application processor (or processing circuitry) 1106 are configured to perform the various functions described above based, at least in part, on information stored in memory. That is, cellular baseband processor (or processing circuitry) 1124 and application processor (or processing circuitry) 1106 can be configured to perform a first subset of the various functions described above without information stored in memory, and can be configured to perform a second subset of the various functions described above based on information stored in memory. Computer-readable media / memory (or memory circuitry) can also be used to store data manipulated by cellular baseband processor (or processing circuitry) 1124 / application processor (or processing circuitry) 1106 when executing the software. The cellular baseband processor (or processing circuit) 1124 / application processor (or processing circuit) 1106 may be a component of the UE 350 and may include at least one memory 360 and / or at least one of the TX processor 368, the RX processor 356, and the controller / processor 359. In one configuration, the device 1104 may be at least one processor chip (modem and / or applications) and include only the cellular baseband processor (or processing circuit) 1124 and / or the application processor (or processing circuit) 1106, while in another configuration, the device 1104 may be the entire UE (e.g., see. Figure 3 UE 350 ) and includes additional modules of device 1104.

[0169] As discussed above, component 198 may be configured to: receive a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions; and communicate with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions. Component 198 may be further configured to perform a combination Figure 7 or Figure 8 The flowchart described and / or Figure 6Component 198 may be within the cellular baseband processor (or processing circuitry) 1124, the application processor (or processing circuitry) 1106, or both. Component 198 may be one or more hardware components specifically configured to perform the recited processes / algorithms, implemented by one or more processors configured to perform the recited processes / algorithms, stored on a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors may individually or in combination perform the recited processes / algorithms. As shown, device 1104 may include a variety of components configured for various functions. In one configuration, the apparatus 1104, in particular the cellular baseband processor (or processing circuit) 1124 and / or the application processor (or processing circuit) 1106, may include: means for receiving a first indication indicating a target set of ML functions from a plurality of defined ML functions from a network entity; and means for communicating with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions. The apparatus 1104 may also include means for performing a combination Figure 7 and Figure 8 Aspects of the flowcharts described and / or by Figure 6 602 in the UE 602. A component may be a component 198 of the apparatus 1104 configured to perform the functions recited by the component. As described above, the apparatus 1104 may include the TX processor 368, the RX processor 356, and the controller / processor 359. Thus, in one configuration, the component may be the TX processor 368, the RX processor 356, and / or the controller / processor 359 configured to perform the functions recited by the component.

[0170] Figure 12Diagram 1200 illustrates an example hardware implementation for a network entity 1202. Network entity 1202 may be a base station (BS), a component of a BS, or may implement BS functionality. Network entity 1202 may include at least one of a CU 1210, a DU 1230, or a RU 1240. For example, depending on the layer functionality handled by component 199, network entity 1202 may include a CU 1210; both the CU 1210 and the DU 1230; each of the CU 1210, the DU 1230, and the RU 1240; the DU 1230; both the DU 1230 and the RU 1240; or the RU 1240. CU 1210 may include at least one CU processor (or processing circuit) 1212. CU processor (or processing circuit) 1212 may include on-chip memory (or memory circuit) 1212′. In some aspects, the CU 1210 may also include additional memory modules 1214 and a communication interface 1218. The CU 1210 communicates with the DU 1230 via a midhaul link (such as an F1 interface). The DU 1230 may include at least one DU processor (or processing circuit) 1232. The DU processor (or processing circuit) 1232 may include on-chip memory (or memory circuit) 1232′. In some aspects, the DU 1230 may also include additional memory modules 1234 and a communication interface 1238. The DU 1230 communicates with the RU 1240 via a fronthaul link. The RU 1240 may include at least one RU processor (or processing circuit) 1242. The RU processor (or processing circuit) 1242 may include on-chip memory (or memory circuit) 1242′. In some aspects, the RU 1240 may also include additional memory modules 1244, one or more transceivers 1246, an antenna 1280, and a communication interface 1248. RU 1240 communicates with UE 104. On-chip memory (or memory circuits) 1212', 1232', 1242' and additional memory modules 1214, 1234, 1244 can each be considered a computer-readable medium / memory (or memory circuit). Each computer-readable medium / memory (or memory circuit) can be non-transitory. Each processor (or processing circuit) 1212, 1232, 1242 is responsible for general processing, including the execution of software stored on the computer-readable medium / memory (or memory circuit). When executed by the corresponding processor, this software causes the processor to perform the various functions described above. The computer-readable medium / memory (or memory circuit) can also be used to store data manipulated by the processor when executing the software.

[0171] As discussed above, component 199 may be configured to: provide a first indication to the UE indicating a target set of ML functions from a plurality of defined ML functions; and communicate with the UE based on one or more operating parameters associated with the target set of ML functions. Component 199 may be further configured to perform a combination Figure 9 or Figure 10 The flowcharts described in and / or by Figure 6 Any of the aspects performed by the base station 604 in . Component 199 may be within one or more processors of one or more of the CU 1210, DU 1230, and RU 1240. Component 199 may be one or more hardware components specifically configured to perform the stated process / algorithm, implemented by one or more processors configured to perform the stated process / algorithm, stored in a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors may perform the stated process / algorithm individually or in combination. The network entity 1202 may include multiple components configured for various functions. In one configuration, the network entity 1202 may include: a component for providing a first indication to the UE indicating a target set of ML functions from a plurality of defined ML functions; and a component for communicating with the UE based on one or more operating parameters associated with the target set of ML functions. The network entity 1202 may also include a component for performing a combination of Figure 9 and Figure 10 Aspects described in the flowcharts and / or by Figure 6 The network entity 1202 may include a TX processor 316, an RX processor 370, and a controller / processor 375. Thus, in one configuration, the component may be the TX processor 316, the RX processor 370, and / or the controller / processor 375 configured to perform the functions recited by the component.

[0172] The present disclosure provides a method for wireless communication at a UE. The method may include: receiving a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions; and communicating with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions. The method provides a new signaling framework dedicated to AI / ML model switching and enables the UE to switch parameters of ML functions with reduced overall signaling overhead. Thus, the method improves the efficiency of wireless communication.

[0173] It should be understood that the specific order or hierarchy of blocks in the disclosed process / flowchart is merely illustrative of an exemplary method. It should be understood that the specific order or hierarchy of blocks in the process / flowchart may be rearranged based on design preferences. In addition, some blocks may be combined or omitted. The accompanying method claims provide elements of the various blocks in a sample order, but are not limited to the specific order or hierarchy provided.

[0174] The foregoing description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects. Therefore, the claims are not limited to the various aspects described herein, but should be given the full scope consistent with the language claims. Unless otherwise specified, references to elements in the singular form do not mean "one and only one", but "one or more". Terms such as "if", "when" and "while" do not imply a direct temporal relationship or reaction. That is, these phrases, such as "when...", do not mean immediate action in response to the occurrence of an action or during the occurrence of an action, but simply imply that if the conditions are met, the action will occur, but there is no need for a specific or immediate time limit for the occurrence of the action. The word "exemplary" is used herein to mean "used as an example, instance, or illustration". Any aspect described as "exemplary" herein is not necessarily interpreted as being preferred or having advantages over other aspects. Unless otherwise specified, the term "some" refers to one or more. Combinations such as “at least one of A, B, or C,” “one or more of A, B, or C,” “at least one of A, B, and C,” “one or more of A, B, and C,” and “A, B, C, or any combination thereof” include any combination of A, B, and / or C, which may include multiple As, multiple Bs, or multiple Cs. Specifically, combinations such as “at least one of A, B, or C,” “one or more of A, B, or C,” “at least one of A, B, and C,” “one or more of A, B, and C,” and “A, B, C, or any combination thereof” may be only A, only B, only C, A and B, A and C, B and C, or A, B, and C, where any such combination may include one or more members of A, B, or C. A set should be interpreted as a set of elements, where the number of elements is one or more. Thus, for a set of X, X will include one or more elements. When at least one processor is configured to perform a set of functions, the at least one processor is configured to perform the set of functions individually or in any combination. Thus, each of the at least one processor can be configured to perform a specific subset of the set of functions, where the subset is the complete set, a suitable subset of the set, or an empty subset of the set. If a first device receives data from or sends data to a second device, the data may be received / sent directly between the first and second devices, or indirectly between the first and second devices via a set of devices. A device configured to "output" data (such as a transmission, signal, or message) may, for example, send the data using a transceiver, or may transmit the data to the device sending the data.A device configured to "obtain" data (such as a transmission, signal, or message) may, for example, receive the data using a transceiver, or may obtain the data from a device that receives the data. Information stored in a memory includes instructions and / or data. All structural and functional equivalents to the elements of the various aspects described throughout this disclosure that are or later become known to those of ordinary skill in the art are expressly incorporated herein by reference and are covered by the claims. In addition, nothing disclosed herein is intended to be dedicated to the public, regardless of whether such disclosure is expressly recited in the claims. Words such as "module," "mechanism," "element," and "device" are not intended to be substituted for the word "component." Therefore, no claim element will be construed as part-plus-function unless the element is expressly recited using the phrase "component for..."

[0175] As used herein, the phrase "based on" should not be interpreted as referring to a closed set of information, one or more conditions, one or more factors, etc. In other words, the phrase "based on A" (where "A" can be information, a condition, a factor, etc.) should be interpreted as "based at least on A" unless specifically stated differently.

[0176] The following aspects are merely illustrative and may be combined with other aspects or teachings described herein without limitation.

[0177] Aspect 1 is a method of wireless communication at a UE. The method may include: receiving a first indication from a network entity indicating a target set of ML functions from a plurality of defined ML functions; and communicating with the network entity based on one or more operating parameters associated with the target set of ML functions indicated from the plurality of defined ML functions.

[0178] Aspect 2 is a method according to aspect 1, wherein the method may further include: configuring the one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters based on the first indication.

[0179] Aspect 3 is a method according to any one of aspects 1 to 2, wherein the plurality of defined ML functions may include one or more of the following: a beam prediction function; a CSF function; or a positioning function.

[0180] Aspect 4 is a method according to any one of aspects 1 to 3, wherein receiving the first indication may include: receiving the first indication indicating the target set of ML functions via one or more MAC-CEs or DCIs.

[0181] Aspect 5 is the method according to aspect 4, wherein the ML functions in the plurality of defined ML functions can be indexed, and each ML function can be associated with a functionality ID according to a defined rule.

[0182] Aspect 6 is the method according to aspect 5, wherein a single MAC-CE or a single DCI may include the functional ID corresponding to the ML function in the target set of ML functions.

[0183] Aspect 7 is a method according to aspect 5, wherein the one or more MAC-CEs or DCIs may include multiple MAC-CEs or multiple DCIs, each of the multiple MAC-CEs or the multiple DCIs having a format associated with a corresponding ML functionality. Each of the multiple MAC-CEs or the multiple DCIs may include the functional ID of at least one ML function in the target set of the ML functions associated with the corresponding ML functionality.

[0184] Aspect 8 is a method according to any one of aspects 1 to 7, wherein the method may further include: before receiving the first indication: receiving RRC signaling from the network entity indicating a first subset of ML functions from the plurality of defined ML functions. The target set of ML functions may be within the first subset of the ML functions indicated in the RRC signaling.

[0185] Aspect 9 is a method according to aspect 8, wherein the method may further include: sending a UE capability indication to the network entity indicating a supported set of ML functions from a defined ML function. The first subset of the ML functions may be within the supported set of the ML functions.

[0186] Aspect 10 is a method according to any one of aspects 1 to 9, wherein the method may further include: before receiving the first indication, receiving one or more MAC-CEs from the network entity activating a subset of the defined ML functions. The target set of ML functions may be within the subset of ML functions activated by the one or more MAC-CEs.

[0187] Aspect 11 is the method according to aspect 10, wherein the one or more MAC-CEs may include a single MAC-CE, the single MAC-CE including a functional ID corresponding to the ML function in the subset of the ML functions.

[0188] Aspect 12 is a method according to aspect 10, wherein the one or more MAC-CEs may include a plurality of MAC-CEs. Each MAC-CE in the plurality of MAC-CEs may have a MAC-CE format associated with a corresponding ML functionality, and each MAC-CE in the plurality of MAC-CEs may include an identifier for at least one ML function in the target set of ML functions associated with the corresponding ML functionality.

[0189] Aspect 13 is a method according to aspect 10, wherein the method may further include: before receiving the first indication: receiving RRC signaling from the network entity indicating an RRC-configured subset of ML functions among the plurality of defined ML functions. The subset of ML functions may be within the RRC-configured subset of ML functions.

[0190] Aspect 14 is a method according to aspect 3, wherein the plurality of defined operating parameters may include one or more of the following: a first set of parameters related to UE ML interference behavior of a plurality of ML functions; a second set of parameters related to expected ML inputs and outputs of the plurality of ML functions; and a third set of parameters related to assistance information.

[0191] Aspect 15 is a method according to aspect 14, wherein the first set of parameters includes one or more of the following: a prediction parameter indicating a prediction type associated with the plurality of defined ML functions; a feedback parameter indicating feedback for channel characteristic predictions associated with the plurality of defined ML functions; and an accuracy parameter indicating an expected prediction accuracy or confidence level associated with the plurality of defined ML functions. The second set of parameters may include one or more of the following: a first indicator of a first measurement resource for input associated with the plurality of defined ML functions; a second resource indicator for a predicted target resource for output associated with the plurality of defined ML functions; and a characteristic parameter indicating a relationship between the first measurement resource and the predicted target resource. The third set of parameters may include one or more of the following: a signaling indicator for an indication request associated with the plurality of defined ML functions; a target indicator for sending a target result associated with the plurality of defined ML functions; and a measurement indicator for measuring the target result associated with the plurality of defined ML functions.

[0192] Aspect 16 is a method according to aspect 15, wherein the prediction type may include one or more of the following: spatial domain beam prediction; time domain beam prediction; and frequency domain beam prediction.

[0193] Aspect 17 is a method according to aspect 15, wherein the channel characteristic prediction may include one or more of the following: predicted L1-RSRP; predicted L1-SINR; CQI of the channel between the UE and the network entity; RI; or predicted top K predicted target resources, the predicted top K predicted target resources being associated with the predicted L1-RSRP and the predicted L1-SINR.

[0194] Aspect 18 is a method according to aspect 15, wherein the ML function may include time-domain beam prediction. The first set of parameters may also include one or more of the following: a history parameter indicating an association between historical measurement resources and a set of spatial transmit filters associated with the ML function; and an accuracy adjustment parameter indicating an adjustment of the accuracy or confidence level of different future prediction opportunities for the ML function. The second set of parameters may also include one or more of the following: an interval parameter indicating an interval between adjacent measurement opportunities and between adjacent prediction opportunities; and a future prediction parameter indicating a number of future prediction opportunities.

[0195] Aspect 19 is a method according to any one of Aspect 15, wherein the ML function may include frequency domain beam prediction, and the first set of parameters may also include one or more of the following: a FR or CC, the FR or the CC conveying a second measurement resource and a prediction target; an accuracy adjustment parameter, the accuracy adjustment parameter indicating an adjustment of the accuracy or confidence level for different FRs or CCs for the ML function. The second set of parameters may also include: a TD association parameter, the TD association parameter indicating a first association of the prediction target with a TD opportunity of the ML function. The third set of parameters may also include one or more of the following: a co-location indicator, the co-location indicator indicating a second association of a third measurement resource with an antenna; and a location parameter, the location parameter indicating location information of the antenna of the network entity.

[0196] Aspect 20 is a method according to any one of Aspects 1 to 19, wherein the method may further include: sending a CSI report to the network entity, wherein the parameters of the CSI report setting of the CSI report may include: the functional ID of the main function associated with the target set of ML functions; and prediction information associated with the target set of ML functions.

[0197] Aspect 21 is a method according to aspect 20, wherein the prediction information may include: predicted channel characteristics of the channel between the UE and the network entity; CMR, wherein the CMR is used as input to identify the predicted channel characteristics; and a prediction resource set, wherein the prediction resource set identifies a prediction target of the channel characteristics.

[0198] Aspect 22 is a method according to aspect 20, wherein the method may further include: adjusting the parameter of the CSI report setting based on the first indication. The parameter may be adjusted without signaling from the network entity.

[0199] Aspect 23 is an apparatus for wireless communication at a UE, the apparatus comprising: a processing system, the processing system comprising a processor circuit and a memory circuit, the memory circuit storing code and coupled to the processor circuit, the processing system being configured to cause the UE to perform the method according to one or more of aspects 1 to 22.

[0200] Aspect 24 is an apparatus for wireless communication at a UE, the apparatus comprising: at least one memory; and at least one processor, the at least one processor being coupled to the at least one memory and, based at least in part on information stored in the at least one memory, the at least one processor being configured, alone or in any combination, to perform a method according to any one of Aspects 1 to 22.

[0201] Aspect 25 is an apparatus for wireless communication at a UE, the apparatus comprising means for implementing the method according to any one of aspects 1 to 22.

[0202] Aspect 26 is an apparatus according to any one of aspects 23 to 25, further comprising a transceiver configured to receive or transmit in association with the method according to any one of aspects 1 to 22.

[0203] Aspect 27 is a computer-readable medium (e.g., a non-transitory computer-readable medium) storing computer-executable code at a UE, wherein the code, when executed by at least one processor, causes the at least one processor to implement a method according to any one of aspects 1 to 22, alone or in any combination.

[0204] Aspect 28 is a method of wireless communication at a network entity. The method may include: providing a first indication to a UE indicating a target set of ML functions from a plurality of defined ML functions; and communicating with the UE based on one or more operating parameters associated with the target set of ML functions.

[0205] Aspect 29 is a method according to aspect 28, wherein the first indication may indicate to the UE to configure the one or more operating parameters associated with the target set of ML functions.

[0206] Aspect 30 is a method according to any one of aspects 28 to 29, wherein the plurality of defined ML functions may include one or more of: a beam prediction function; a CSF function; or a positioning function.

[0207] Aspect 31 is a method according to any one of aspects 28 to 30, wherein providing the first indication may include providing the first indication indicating the target set of ML functions via one or more MAC-CEs or DCIs.

[0208] Aspect 32 is the method according to aspect 31, wherein the ML functions in the plurality of defined ML functions can be indexed, and each ML function can be associated with a functionality ID according to a defined rule.

[0209] Aspect 33 is a method according to aspect 32, wherein a single MAC-CE or a single DCI may include the functional ID corresponding to the ML function in the target set of ML functions.

[0210] Aspect 34 is an apparatus for wireless communication at a network entity, the apparatus comprising: a processing system comprising a processor circuit and a memory circuit, the memory circuit storing code and coupled to the processor circuit, the processing system configured to cause the network entity to perform the method according to one or more of aspects 28 to 33.

[0211] Aspect 35 is an apparatus for wireless communication at a network entity, the apparatus comprising: at least one memory; and at least one processor, the at least one processor being coupled to the at least one memory and, based at least in part on information stored in the at least one memory, the at least one processor being configured, alone or in any combination, to perform a method according to any one of aspects 28 to 33.

[0212] Aspect 36 is an apparatus for wireless communication, the apparatus comprising means for implementing the method according to any one of aspects 28 to 33.

[0213] Aspect 37 is an apparatus according to any one of aspects 34 to 36, further comprising at least a transceiver configured to receive or transmit in association with the method according to any one of aspects 28 to 33.

[0214] Aspect 38 is a computer-readable medium (e.g., a non-transitory computer-readable medium) storing computer-executable code at a network entity, wherein the code, when executed by at least one processor, causes the at least one processor to implement a method according to any one of aspects 28 to 33, alone or in any combination.

Claims

1. An apparatus for wireless communication at a user equipment (UE), the apparatus comprising: at least one memory; and at least one processor coupled to the at least one memory and, based at least in part on information stored in the at least one memory, configured, alone or in any combination, to cause the UE to: receiving a first indication from a network entity indicating a target set of machine learning (ML) functions among a plurality of defined ML functions; and Communicating with the network entity based on one or more operational parameters associated with the target set of ML functions indicated from the plurality of defined ML functions.

2. The apparatus according to claim 1, further comprising: a transceiver coupled to the at least one processor, wherein, to receive the first indication, the at least one processor is configured, alone or in any combination, to cause the UE to receive the first indication via the transceiver, and wherein the at least one processor is further configured, alone or in any combination, to cause the UE to: configuring, based on the first indication, the one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters, wherein the plurality of defined ML functions comprises one or more of the following: Beam prediction function; Channel State Feedback (CSF) functionality; or Positioning function.

3. The device according to claim 2, wherein To receive the first indication, the at least one processor is configured, alone or in any combination, to cause the UE to: The first indication indicating the target set of ML functions is received via one or more Medium Access Control-Control Elements (MAC-CEs) or Downlink Control Information (DCIs). 4 . The apparatus of claim 3 , wherein the ML functions of the plurality of defined ML functions are indexed, and each ML function is associated with a functionality identifier (ID) according to a defined rule. 5 . The apparatus of claim 4 , wherein a single MAC-CE or a single DCI includes a functional ID corresponding to the ML function in the target set of ML functions.

6. The apparatus of claim 4 , wherein the one or more MAC-CEs or DCIs comprise a plurality of MAC-CEs or a plurality of DCIs, each of the plurality of MAC-CEs or the plurality of DCIs having a format associated with a corresponding ML functionality, and each of the plurality of MAC-CEs or the plurality of DCIs comprises the functionality ID of at least one ML function in the target set of ML functions associated with the corresponding ML functionality.

7. The apparatus of claim 1 , wherein prior to being configured to receive the first indication, the at least one processor is further configured, alone or in any combination, to cause the UE to: Radio resource control (RRC) signaling is received from the network entity indicating a first subset of ML functions among the plurality of defined ML functions, and wherein the target set of ML functions is within the first subset of ML functions indicated in the RRC signaling.

8. The apparatus of claim 7, wherein the at least one processor, alone or in any combination, is further configured to cause the UE to: A UE capability indication is sent to the network entity indicating a supported set of ML functions from the plurality of defined ML functions, wherein the first subset of the ML functions is within the supported set of ML functions.

9. The apparatus of claim 1 , wherein prior to being configured to receive the first indication, the at least one processor is further configured, alone or in any combination, to cause the UE to: One or more medium access control-control elements (MAC-CEs) activating a subset of ML functions among the plurality of defined ML functions are received from the network entity, wherein the target set of ML functions is within the subset of ML functions activated by the one or more MAC-CEs.

10. The apparatus of claim 9, wherein the one or more MAC-CEs include a single MAC-CE including a functionality identifier (ID) corresponding to the ML function in the subset of ML functions.

11. The apparatus of claim 9 , wherein the one or more MAC-CEs comprise a plurality of MAC-CEs, each of the plurality of MAC-CEs having a MAC-CE format associated with a corresponding ML functionality, and each of the plurality of MAC-CEs comprising an identifier for at least one ML function in the target set of ML functions associated with the corresponding ML functionality.

12. The apparatus of claim 9, wherein prior to being configured to receive the first indication, the at least one processor is further configured, alone or in any combination, to cause the UE to: Radio Resource Control (RRC) signaling is received from the network entity indicating a RRC-configured subset of ML functions among the plurality of defined ML functions, wherein the subset of ML functions activated by the one or more MAC-CEs is within the RRC-configured subset of ML functions.

13. The apparatus of claim 2, wherein the plurality of defined operating parameters comprises one or more of: a first set of parameters related to UE ML interference behavior of the plurality of defined ML functions; a second set of parameters related to expected ML inputs and outputs of the plurality of defined ML functions; and A third set of parameters related to auxiliary information.

14. The apparatus of claim 13, wherein the first set of parameters comprises one or more of: a prediction parameter indicating a prediction type associated with the plurality of defined ML functions, wherein the prediction type comprises one or more of: spatial domain beam prediction, time domain beam prediction, and frequency domain beam prediction; feedback parameters indicating feedback for channel characteristic predictions associated with the plurality of defined ML functions; or an accuracy parameter indicating an expected prediction accuracy or confidence level associated with the plurality of defined ML functions; and The second set of parameters includes one or more of the following: a first resource indicator of a first measurement resource for inputs associated with the plurality of defined ML functions; a second resource indicator of a target resource for a prediction of outputs associated with the plurality of defined ML functions; and a characteristic parameter indicating a relationship between the first measured resource and the predicted target resource; and The third set of parameters includes one or more of the following: a signaling indicator for indicating a request associated with the plurality of defined ML functions; a target indicator for transmitting target results associated with the plurality of defined ML functions; and Measurement indicators for measuring the target outcomes associated with the plurality of defined ML functions.

15. The apparatus according to claim 14, wherein the channel characteristic prediction comprises one or more of the following: Predicted layer 1 reference signal received power (L1-RSRP); The predicted layer 1 signal to interference plus noise ratio (L1-SINR); a channel quality indicator (CQI) of a channel between the UE and the network entity; Rank Indicator (RI); or The predicted top K prediction target resources are associated with the predicted L1-RSRP and the predicted L1-SINR.

16. The apparatus of claim 14, wherein the target ML function comprises the time-domain beam prediction, and the first set of parameters further comprises one or more of: a history parameter indicating an association of a historical measurement resource and a spatial transmit filter set associated with the target ML function; or an accuracy adjustment parameter indicating an adjustment to the accuracy or confidence level of different future prediction opportunities for the target ML function; and The second set of parameters also includes one or more of the following: an interval parameter indicating an interval between adjacent measurement opportunities and between adjacent prediction opportunities; or A future prediction parameter indicates a number of future prediction opportunities.

17. The apparatus of claim 14, wherein the target ML function comprises the frequency-domain beam prediction, and the first set of parameters further comprises one or more of: a frequency range (FR) or a component carrier (CC), the frequency range (FR) or the component carrier (CC) conveying a second measurement resource and a prediction target; an accuracy adjustment parameter indicating an adjustment to the accuracy or confidence level of different FRs or CCs for the target ML function; The second set of parameters further comprises a TD association parameter indicating a first association of a prediction target with a TD opportunity of the target ML function; or The third set of parameters further includes one or more of the following: a co-location indicator indicating a second association of a third measurement resource with the antenna; and A location parameter indicating location information of the antenna of the network entity.

18. The apparatus of claim 2, wherein the at least one processor, alone or in any combination, is further configured to cause the UE to: Sending a channel state information (CSI) report to the network entity, wherein parameters of a CSI report setting of the CSI report include: the functional ID of the primary function associated with said target set of ML functions; and Prediction information associated with the target set of ML functions, wherein the prediction information includes: predicted channel characteristics of a channel between the UE and the network entity; Coordinated Multipoint Reception (CMR), wherein the CMR is used as input to identify the predicted channel characteristics; and A prediction resource set identifies a prediction target of the predicted channel characteristic.

19. The apparatus of claim 18, wherein the at least one processor, alone or in any combination, is further configured to cause the UE to: The parameter of the CSI reporting setting is adjusted based on the first indication, wherein the parameter is adjusted without signaling from the network entity.

20. An apparatus for wireless communication at a network entity, the apparatus comprising: at least one memory; and at least one processor coupled to the at least one memory and, based at least in part on information stored in the at least one memory, configured, alone or in any combination, to cause the network entity to: providing a first indication to a user equipment (UE) indicating a target set of machine learning (ML) functions among a plurality of defined ML functions; and Communicating with the UE based on one or more operating parameters associated with the target set of ML functions.

21. The apparatus according to claim 20, further comprising: a transceiver coupled to the at least one processor, wherein, to provide the first indication, the at least one processor is configured, alone or in any combination, to cause the network entity to provide the first indication via the transceiver, and wherein the first indication indicates to the UE to configure the one or more operating parameters associated with the target set of ML functions.

22. The apparatus of claim 21 , wherein the plurality of defined ML functions comprises one or more of: Beam prediction function; Channel State Feedback (CSF) functionality; or Positioning function.

23. The device according to claim 22, wherein To provide the first indication, the at least one processor is configured, alone or in any combination, to cause the network entity to: The first indication indicating the target set of ML functions is provided via one or more Medium Access Control-Control Elements (MAC-CEs) or Downlink Control Information (DCIs). 24 . The apparatus of claim 23 , wherein the ML functions of the plurality of defined ML functions are indexed, and each ML function is associated with a functionality identifier (ID) according to a defined rule.

25. The apparatus of claim 24, wherein a single MAC-CE or a single DCI includes a functionality ID corresponding to the ML function in the target set of ML functions.

26. A method of wireless communication at a user equipment (UE), the method comprising: receiving, from a network entity, a first indication indicating a target set of machine learning (ML) functions among a plurality of defined ML functions; as well as Communicating with the network entity based on one or more operational parameters associated with the target set of ML functions indicated from the plurality of defined ML functions.

27. The method according to claim 26, further comprising: configuring, based on the first indication, the one or more operating parameters associated with the target set of ML functions from a plurality of defined operating parameters, wherein the plurality of defined ML functions comprises one or more of the following: Beam prediction function; Channel State Feedback (CSF) functionality; or Positioning function.

28. The method of claim 27, wherein receiving the first indication comprises: The first indication indicating the target set of ML functions is received via one or more Medium Access Control-Control Elements (MAC-CEs) or Downlink Control Information (DCIs).

29. The method of claim 28, wherein the ML functions of the plurality of defined ML functions are indexed and each ML function is associated with a functionality identifier (ID) according to a defined rule.

30. A method of performing wireless communication at a network entity, the method comprising: providing a first indication to a user equipment (UE) indicating a target set of machine learning (ML) functions among a plurality of defined ML functions; as well as Communicating with the UE based on one or more operating parameters associated with the target set of ML functions.

Images