WO2025123239A1

WO2025123239A1 - Beam-mapping pattern consistency for machine learning model training and inference

Info

Publication number: WO2025123239A1
Application number: PCT/CN2023/138369
Authority: WO
Inventors: Qiaoyu Li; Tao Luo; Mahmoud Taherzadeh Boroujeni; Hamed Pezeshki
Original assignee: Qualcomm Inc
Current assignee: Qualcomm Inc
Priority date: 2023-12-13
Filing date: 2023-12-13
Publication date: 2025-06-19
Anticipated expiration: 2026-06-13

Abstract

Methods, systems, and devices for wireless communication are described. A user equipment (UE) may receive an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE. The UE may transmit a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

Description

BEAM-MAPPING PATTERN CONSISTENCY FOR MACHINE LEARNING MODEL TRAINING AND INFERENCE

TECHNICAL FIELD

The following relates to wireless communication, including techniques that support beam-mapping pattern consistency for machine learning model training and inference.

DESCRIPTION OF THE RELATED TECHNOLOGY

Wireless communications systems are widely deployed to provide various types of communication content such as voice, video, packet data, messaging, broadcast, and so on. These systems may be capable of supporting communication with multiple users by sharing the available system resources (e.g., time, frequency, and power) . Examples of such multiple-access systems include fourth generation (4G) systems such as Long Term Evolution (LTE) systems, LTE-Advanced (LTE-A) systems, or LTE-A Pro systems, and fifth generation (5G) systems which may be referred to as New Radio (NR) systems. These systems may employ technologies such as code division multiple access (CDMA) , time division multiple access (TDMA) , frequency division multiple access (FDMA) , orthogonal FDMA (OFDMA) , or discrete Fourier transform spread orthogonal frequency division multiplexing (DFT-S-OFDM) . A wireless multiple-access communications system may include one or more base stations, each supporting wireless communication for communication devices, which may be known as user equipment (UE) .

A UE may support artificial intelligence (AI) or machine learning (ML) -based beam prediction (e.g., spatial beam prediction) , referred to generally as ML beam prediction. For example, a UE may support an AI or ML model (or functionality) that enables ML beam prediction. The term ML model may be an umbrella term that encompasses both an ML model and an AI model as well as an ML functionality and an AI functionality.

Before a UE uses an ML model for inference, the ML model may be trained. During training, the ML model may use measurements of a first set of beams (referred to as training set B beams) to predict measurements for a second set of beams (referred to as training set A beams) . The ML model may adapt itself using feedback on the predicted measurements for the training set A beams, thereby evolving and improving future beam measurement predictions. During inference, the ML model may use measurements of a third set of beams (referred to as inference Set B beams) to predict measurements for a fourth set of beams (referred to as inference set A beams) . In general, set B beams may refer to beams for which measured channel characteristics are input into the ML model as features (e.g., for training, for inference) and set A beams may refer to beams for which associated channel characteristics are predicted by the ML model.

During training, the set B beam measurements input into the ML model may be for set B beams that are selected in accordance with a training beam-mapping pattern, where a beam-mapping pattern maps set B beams to set A beams. During inference, the set B beam measurements input into the ML model may be for set B beams that are selected in accordance with an inference beam-mapping pattern. The performance of an ML model may be a function of the beam-mapping patterns used for training and inference. Techniques for selecting beam-mapping patterns that improve ML model performance may be desired.

SUMMARY

The systems, methods and devices of this disclosure each have several innovative aspects, no single one of which is solely responsible for the desirable attributes disclosed herein.

One innovative aspect of the subject matter described in this disclosure can be implemented in a method for wireless communication at a user equipment (UE) . The method includes receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

Another innovative aspect of the subject matter described in this disclosure can be implemented in a method for wireless communication at a network entity. The method includes outputting, to a UE, an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows an example of a wireless communications system that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 2 shows an example of a wireless communications system that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 3 shows an example of beam mapping patterns that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 4 shows an example of beam-mapping patterns that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 5 shows an example of prediction cycles that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 6 shows an example of beam-mapping patterns that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 7 shows an example of beam-mapping patterns that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 8 shows an example of probabilities associated with beam-mapping patterns in accordance with one or more aspects of the present disclosure.

Figure 9 shows an example of a process flow that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 10 shows an example of a process flow that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 11 shows an example of a process flow that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figures 12 and 13 show devices that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 14 shows a communications manager that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 15 shows a diagram of a system including a device that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figures 16 and 17 show devices that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 18 shows a communications manager that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figure 19 shows a diagram of a system including a device that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

Figures 20 and 21 show flowcharts illustrating methods that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure.

DETAILED DESCRIPTION

The following description is directed to some particular examples for the purposes of describing innovative aspects of this disclosure. However, a person having ordinary skill in the art will readily recognize that the teachings herein can be applied in a multitude of different ways. Some or all of the described examples may be implemented in any device, system or network that is capable of transmitting and receiving radio frequency (RF) signals according to one or more of the Long Term Evolution (LTE) , 3G, 4G or 5G (New Radio (NR) ) standards promulgated by the 3rd Generation Partnership Project (3GPP) , the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standards, the IEEE 802.15 standards, the standards as defined by the Bluetooth Special Interest Group (SIG) , among others. The described examples can be implemented in any device, system or network that is capable of transmitting and receiving RF signals according to one or more of the following technologies or techniques: code division multiple access (CDMA) , time division multiple access (TDMA) , orthogonal frequency division multiplexing (OFDM) , frequency division multiple access (FDMA) , orthogonal FDMA (OFDMA) , single-carrier FDMA (SC-FDMA) , spatial division multiple access (SDMA) , rate-splitting multiple access (RSMA) , multi-user shared access (MUSA) , single-user (SU) multiple-input multiple-output (MIMO) and multi-user (MU) -MIMO (MU-MIMO) . The described examples also can be implemented using other wireless communication protocols or RF signals suitable for use in one or more of a wireless personal area network (WPAN) , a wireless local area network (WLAN) , a wireless wide area network (WWAN) , a wireless metropolitan area network (WMAN) , or an internet of things (IOT) network.

A UE with an ML model may train the ML model before using the ML model for inference. For example, during training the ML model may operate on measurements of training set B beams to predict measurements for training set A beams. The ML model may adapt itself using feedback on the predicted measurements for the training set A beams, thereby evolving and improving future beam measurement predictions. During inference, the ML model may use measurements of inference Set B beams to predict measurements for inference set A beams. Unless otherwise indicated, set B beams may refer to beams for which measured channel characteristics are input into the ML model as features (e.g., for training, for inference) and set A beams may refer to beams for which associated channel characteristics are predicted by the ML model.

Various aspects generally relate to machine learning (ML) -based beam prediction, and more specifically to signaling techniques for selecting beam-mapping patterns for use during ML model training and inference. In a scenario where the training set B beams are a subset of the set A beams (referred to as narrow-to-narrow beam prediction) , the performance of an ML model may suffer if the inference beam-mapping pattern is inconsistent with the training beam-mapping pattern (e.g., if the training set B beams are inconsistent with the inference set B beams) . But the network entity responsible for activating and deactivating the ML model may not have sufficient information to ensure consistency between the training beam-mapping pattern and the inference beam-mapping pattern (collectively referred to as a beam-mapping pattern pair) .

According to the described techniques, a user equipment (UE) and a network entity may exchange signaling that ensures consistency between the training beam-mapping pattern used for an ML model and the inference beam-mapping pattern used for the ML model. For instance, the UE may exchange signaling with the network entity so that the network entity is able to determine whether a candidate inference beam-mapping pattern is consistent with the training beam-mapping pattern used for the ML model. If the candidate beam-mapping pattern is consistent with the training beam-mapping pattern, the network entity may signal the UE to activate the ML model (and in some examples may indicate the beam-mapping pattern to use for inference) . If the candidate beam-mapping pattern is inconsistent with the training beam-mapping pattern, the network entity may signal the UE to deactivate the ML model.

Particular aspects of the subject matter described in this disclosure can be implemented to realize one or more of the following potential advantages. In some examples, by ensuring an activated ML model has consistency between beam-mapping patterns as described, the performance of an ML model may be improved, which may allow for improved beam prediction by the UE. Improved beam prediction in turn may enable improved beam management by the network entity, which may increase resource-use efficiency and communication reliability, which may result in greater spectral efficiency in more deployment scenarios, higher data rates, lower latency, and greater capacity, among other benefits.

Aspects of the disclosure are initially described in the context of wireless communications systems. Additional aspects of the disclosure are described with reference to beam-mapping patterns, prediction cycles, and process flows. Aspects of the disclosure are further illustrated by and described with reference to apparatus diagrams, system diagrams, and flowcharts that relate to beam-mapping pattern consistency for machine learning model training and inference.

Figure 1 shows an example of a wireless communications system 100 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The wireless communications system 100 may include one or more network entities 105, one or more UEs 115, and a core network 130. In some examples, the wireless communications system 100 may be a Long Term Evolution (LTE) network, an LTE-Advanced (LTE-A) network, an LTE-A Pro network, a New Radio (NR) network, or a network operating in accordance with other systems and radio technologies, including future systems and radio technologies not explicitly mentioned herein.

The network entities 105 may be dispersed throughout a geographic area to form the wireless communications system 100 and may include devices in different forms or having different capabilities. In various examples, a network entity 105 may be referred to as a network element, a mobility element, a radio access network (RAN) node, or network equipment, among other nomenclature. In some examples, network entities 105 and Ues 115 may wirelessly communicate via one or more communication links 125 (e.g., a radio frequency (RF) access link) . For example, a network entity 105 may support a coverage area 110 (e.g., a geographic coverage area) over which the Ues 115 and the network entity 105 may establish one or more communication links 125. The coverage area 110 may be an example of a geographic area over which a network entity 105 and a UE 115 may support the communication of signals according to one or more radio access technologies (RATs) .

The Ues 115 may be dispersed throughout a coverage area 110 of the wireless communications system 100, and each UE 115 may be stationary, or mobile, or both at different times. The Ues 115 may be devices in different forms or having different capabilities. Some example Ues 115 are illustrated in Figure 1. The Ues 115 may be capable of supporting communications with various types of devices, such as other Ues 115 or network entities 105, as shown in Figure 1.

As described herein, a node of the wireless communications system 100, which may be referred to as a network node, or a wireless node, may be a network entity 105 (e.g., any network entity described herein) , a UE 115 (e.g., any UE described herein) , a network controller, an apparatus, a device, a computing system, one or more components, or another suitable processing entity configured to perform any of the techniques described herein. For example, a node may be a UE 115. As another example, a node may be a network entity 105. As another example, a first node may be configured to communicate with a second node or a third node. In one aspect of this example, the first node may be a UE 115, the second node may be a network entity 105, and the third node may be a UE 115. In another aspect of this example, the first node may be a UE 115, the second node may be a network entity 105, and the third node may be a network entity 105. In yet other aspects of this example, the first, second, and third nodes may be different relative to these examples. Similarly, reference to a UE 115, network entity 105, apparatus, device, computing system, or the like may include disclosure of the UE 115, network entity 105, apparatus, device, computing system, or the like being a node. For example, disclosure that a UE 115 is configured to receive information from a network entity 105 also discloses that a first node is configured to receive information from a second node.

In some examples, network entities 105 may communicate with the core network 130, or with one another, or both. For example, network entities 105 may communicate with the core network 130 via one or more backhaul communication links 120 (e.g., in accordance with an S1, N2, N3, or other interface protocol) . In some examples, network entities 105 may communicate with one another via a backhaul communication link 120 (e.g., in accordance with an X2, Xn, or other interface protocol) either directly (e.g., directly between network entities 105) or indirectly (e.g., via a core network 130) . In some examples, network entities 105 may communicate with one another via a midhaul communication link 162 (e.g., in accordance with a midhaul interface protocol) or a fronthaul communication link 168 (e.g., in accordance with a fronthaul interface protocol) , or any combination thereof. The backhaul communication links 120, midhaul communication links 162, or fronthaul communication links 168 may be or include one or more wired links (e.g., an electrical link, an optical fiber link) , one or more wireless links (e.g., a radio link, a wireless optical link) , among other examples or various combinations thereof. A UE 115 may communicate with the core network 130 via a communication link 155.

One or more of the network entities 105 described herein may include or may be referred to as a base station 140 (e.g., a base transceiver station, a radio base station, an NR base station, an access point, a radio transceiver, a NodeB, an eNodeB (eNB) , a next-generation NodeB or a giga-NodeB (either of which may be referred to as a gNB) , a 5G NB, a next-generation eNB (ng-eNB) , a Home NodeB, a Home eNodeB, or other suitable terminology) . In some examples, a network entity 105 (e.g., a base station 140) may be implemented in an aggregated (e.g., monolithic, standalone) base station architecture, which may be configured to utilize a protocol stack that is physically or logically integrated within a single network entity 105 (e.g., a single RAN node, such as a base station 140) .

In some examples, a network entity 105 may be implemented in a disaggregated architecture (e.g., a disaggregated base station architecture, a disaggregated RAN architecture) , which may be configured to utilize a protocol stack that is physically or logically distributed among two or more network entities 105, such as an integrated access backhaul (IAB) network, an open RAN (O-RAN) (e.g., a network configuration sponsored by the O-RAN Alliance) , or a virtualized RAN (vRAN) (e.g., a cloud RAN (C-RAN) ) . For example, a network entity 105 may include one or more of a central unit (CU) 160, a distributed unit (DU) 165, a radio unit (RU) 170, a RAN Intelligent Controller (RIC) 175 (e.g., a Near-Real Time RIC (Near-RT RIC) , a Non-Real Time RIC (Non-RT RIC) ) , a Service Management and Orchestration (SMO) 180 system, or any combination thereof. An RU 170 may also be referred to as a radio head, a smart radio head, a remote radio head (RRH) , a remote radio unit (RRU) , or a transmission reception point (TRP) . One or more components of the network entities 105 in a disaggregated RAN architecture may be co-located, or one or more components of the network entities 105 may be located in distributed locations (e.g., separate physical locations) . In some examples, one or more network entities 105 of a disaggregated RAN architecture may be implemented as virtual units (e.g., a virtual CU (VCU) , a virtual DU (VDU) , a virtual RU (VRU) ) .

The split of functionality between a CU 160, a DU 165, and an RU 170 is flexible and may support different functionalities depending on which functions (e.g., network layer functions, protocol layer functions, baseband functions, RF functions, and any combinations thereof) are performed at a CU 160, a DU 165, or an RU 170. For example, a functional split of a protocol stack may be employed between a CU 160 and a DU 165 such that the CU 160 may support one or more layers of the protocol stack and the DU 165 may support one or more different layers of the protocol stack. In some examples, the CU 160 may host upper protocol layer (e.g., layer 3 (L3) , layer 2 (L2) ) functionality and signaling (e.g., Radio Resource Control (RRC) , service data adaption protocol (SDAP) , Packet Data Convergence Protocol (PDCP) ) . The CU 160 may be connected to one or more Dus 165 or Rus 170, and the one or more Dus 165 or Rus 170 may host lower protocol layers, such as layer 1 (L1) (e.g., physical (PHY) layer) or L2 (e.g., radio link control (RLC) layer, medium access control (MAC) layer) functionality and signaling, and may each be at least partially controlled by the CU 160. Additionally, or alternatively, a functional split of the protocol stack may be employed between a DU 165 and an RU 170 such that the DU 165 may support one or more layers of the protocol stack and the RU 170 may support one or more different layers of the protocol stack. The DU 165 may support one or multiple different cells (e.g., via one or more Rus 170) . In some cases, a functional split between a CU 160 and a DU 165, or between a DU 165 and an RU 170 may be within a protocol layer (e.g., some functions for a protocol layer may be performed by one of a CU 160, a DU 165, or an RU 170, while other functions of the protocol layer are performed by a different one of the CU 160, the DU 165, or the RU 170) . A CU 160 may be functionally split further into CU control plane (CU-CP) and CU user plane (CU-UP) functions. A CU 160 may be connected to one or more Dus 165 via a midhaul communication link 162 (e.g., F1, F1-c, F1-u) , and a DU 165 may be connected to one or more Rus 170 via a fronthaul communication link 168 (e.g., open fronthaul (FH) interface) . In some examples, a midhaul communication link 162 or a fronthaul communication link 168 may be implemented in accordance with an interface (e.g., a channel) between layers of a protocol stack supported by respective network entities 105 that are in communication via such communication links.

In wireless communications systems (e.g., wireless communications system 100) , infrastructure and spectral resources for radio access may support wireless backhaul link capabilities to supplement wired backhaul connections, providing an IAB network architecture (e.g., to a core network 130) . In some cases, in an IAB network, one or more network entities 105 (e.g., IAB nodes 104) may be partially controlled by each other. One or more IAB nodes 104 may be referred to as a donor entity or an IAB donor. One or more Dus 165 or one or more Rus 170 may be partially controlled by one or more Cus 160 associated with a donor network entity 105 (e.g., a donor base station 140) . The one or more donor network entities 105 (e.g., IAB donors) may be in communication with one or more additional network entities 105 (e.g., IAB nodes 104) via supported access and backhaul links (e.g., backhaul communication links 120) . IAB nodes 104 may include an IAB mobile termination (IAB-MT) controlled (e.g., scheduled) by Dus 165 of a coupled IAB donor. An IAB-MT may include an independent set of antennas for relay of communications with Ues 115, or may share the same antennas (e.g., of an RU 170) of an IAB node 104 used for access via the DU 165 of the IAB node 104 (e.g., referred to as virtual IAB-MT (vIAB-MT) ) . In some examples, the IAB nodes 104 may include Dus 165 that support communication links with additional entities (e.g., IAB nodes 104, Ues 115) within the relay chain or configuration of the access network (e.g., downstream) . In such cases, one or more components of the disaggregated RAN architecture (e.g., one or more IAB nodes 104 or components of IAB nodes 104) may be configured to operate according to the techniques described herein.

In the case of the techniques described herein applied in the context of a disaggregated RAN architecture, one or more components of the disaggregated RAN architecture may be configured to support beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, some operations described as being performed by a UE 115 or a network entity 105 (e.g., a base station 140) may additionally, or alternatively, be performed by one or more components of the disaggregated RAN architecture (e.g., IAB nodes 104, Dus 165, Cus 160, Rus 170, RIC 175, SMO 180) .

A UE 115 may include or may be referred to as a mobile device, a wireless device, a remote device, a handheld device, or a subscriber device, or some other suitable terminology, where the “device” may also be referred to as a unit, a station, a terminal, or a client, among other examples. A UE 115 may also include or may be referred to as a personal electronic device such as a cellular phone, a personal digital assistant (PDA) , a tablet computer, a laptop computer, or a personal computer. In some examples, a UE 115 may include or be referred to as a wireless local loop (WLL) station, an Internet of Things (IoT) device, an Internet of Everything (IoE) device, or a machine type communications (MTC) device, among other examples, which may be implemented in various objects such as appliances, or vehicles, meters, among other examples.

The Ues 115 described herein may be able to communicate with various types of devices, such as other Ues 115 that may sometimes act as relays as well as the network entities 105 and the network equipment including macro eNBs or gNBs, small cell eNBs or gNBs, or relay base stations, among other examples, as shown in Figure 1.

The Ues 115 and the network entities 105 may wirelessly communicate with one another via one or more communication links 125 (e.g., an access link) using resources associated with one or more carriers. The term “carrier” may refer to a set of RF spectrum resources having a defined physical layer structure for supporting the communication links 125. For example, a carrier used for a communication link 125 may include a portion of a RF spectrum band (e.g., a bandwidth part (BWP) ) that is operated according to one or more physical layer channels for a given radio access technology (e.g., LTE, LTE-A, LTE-A Pro, NR) . Each physical layer channel may carry acquisition signaling (e.g., synchronization signals, system information) , control signaling that coordinates operation for the carrier, user data, or other signaling. The wireless communications system 100 may support communication with a UE 115 using carrier aggregation or multi-carrier operation. A UE 115 may be configured with multiple downlink component carriers and one or more uplink component carriers according to a carrier aggregation configuration. Carrier aggregation may be used with both frequency division duplexing (FDD) and time division duplexing (TDD) component carriers. Communication between a network entity 105 and other devices may refer to communication between the devices and any portion (e.g., entity, sub-entity) of a network entity 105. For example, the terms “transmitting, ” “receiving, ” or “communicating, ” when referring to a network entity 105, may refer to any portion of a network entity 105 (e.g., a base station 140, a CU 160, a DU 165, a RU 170) of a RAN communicating with another device (e.g., directly or via one or more other network entities 105) .

Signal waveforms transmitted via a carrier may be made up of multiple subcarriers (e.g., using multi-carrier modulation (MCM) techniques such as orthogonal frequency division multiplexing (OFDM) or discrete Fourier transform spread OFDM (DFT-S-OFDM) ) . In a system employing MCM techniques, a resource element may refer to resources of one symbol period (e.g., a duration of one modulation symbol) and one subcarrier, in which case the symbol period and subcarrier spacing may be inversely related. The quantity of bits carried by each resource element may depend on the modulation scheme (e.g., the order of the modulation scheme, the coding rate of the modulation scheme, or both) , such that a relatively higher quantity of resource elements (e.g., in a transmission duration) and a relatively higher order of a modulation scheme may correspond to a relatively higher rate of communication. A wireless communications resource may refer to a combination of an RF spectrum resource, a time resource, and a spatial resource (e.g., a spatial layer, a beam) , and the use of multiple spatial resources may increase the data rate or data integrity for communications with a UE 115.

The time intervals for the network entities 105 or the Ues 115 may be expressed in multiples of a basic time unit which may, for example, refer to a sampling period of T_s=1/ (Δf_max·N_f) seconds, for which Δf_max may represent a supported subcarrier spacing, and N_f may represent a supported discrete Fourier transform (DFT) size. Time intervals of a communications resource may be organized according to radio frames each having a specified duration (e.g., 10 milliseconds (ms) ) . Each radio frame may be identified by a system frame number (SFN) (e.g., ranging from 0 to 1023) .

Each frame may include multiple consecutively-numbered subframes or slots, and each subframe or slot may have the same duration. In some examples, a frame may be divided (e.g., in the time domain) into subframes, and each subframe may be further divided into a quantity of slots. Alternatively, each frame may include a variable quantity of slots, and the quantity of slots may depend on subcarrier spacing. Each slot may include a quantity of symbol periods (e.g., depending on the length of the cyclic prefix prepended to each symbol period) . In some wireless communications systems 100, a slot may further be divided into multiple mini-slots associated with one or more symbols. Excluding the cyclic prefix, each symbol period may be associated with one or more (e.g., N_f) sampling periods. The duration of a symbol period may depend on the subcarrier spacing or frequency band of operation.

A subframe, a slot, a mini-slot, or a symbol may be the smallest scheduling unit (e.g., in the time domain) of the wireless communications system 100 and may be referred to as a transmission time interval (TTI) . In some examples, the TTI duration (e.g., a quantity of symbol periods in a TTI) may be variable. Additionally, or alternatively, the smallest scheduling unit of the wireless communications system 100 may be dynamically selected (e.g., in bursts of shortened TTIs (sTTIs) ) .

Physical channels may be multiplexed for communication using a carrier according to various techniques. A physical control channel and a physical data channel may be multiplexed for signaling via a downlink carrier, for example, using one or more of time division multiplexing (TDM) techniques, frequency division multiplexing (FDM) techniques, or hybrid TDM-FDM techniques. A control region (e.g., a control resource set (CORESET) ) for a physical control channel may be defined by a set of symbol periods and may extend across the system bandwidth or a subset of the system bandwidth of the carrier. One or more control regions (e.g., CORESETs) may be configured for a set of the Ues 115. For example, one or more of the Ues 115 may monitor or search control regions for control information according to one or more search space sets, and each search space set may include one or multiple control channel candidates in one or more aggregation levels arranged in a cascaded manner. An aggregation level for a control channel candidate may refer to an amount of control channel resources (e.g., control channel elements (CCEs) ) associated with encoded information for a control information format having a given payload size. Search space sets may include common search space sets configured for sending control information to multiple Ues 115 and UE-specific search space sets for sending control information to a specific UE 115.

In some examples, a network entity 105 (e.g., a base station 140, an RU 170) may be movable and therefore provide communication coverage for a moving coverage area 110. In some examples, different coverage areas 110 associated with different technologies may overlap, but the different coverage areas 110 may be supported by the same network entity 105. In some other examples, the overlapping coverage areas 110 associated with different technologies may be supported by different network entities 105. The wireless communications system 100 may include, for example, a heterogeneous network in which different types of the network entities 105 provide coverage for various coverage areas 110 using the same or different radio access technologies.

The wireless communications system 100 may be configured to support ultra-reliable communications or low-latency communications, or various combinations thereof. For example, the wireless communications system 100 may be configured to support ultra-reliable low-latency communications (URLLC) . The Ues 115 may be designed to support ultra-reliable, low-latency, or critical functions. Ultra-reliable communications may include private communication or group communication and may be supported by one or more services such as push-to-talk, video, or data. Support for ultra-reliable, low-latency functions may include prioritization of services, and such services may be used for public safety or general commercial applications. The terms ultra-reliable, low-latency, and ultra-reliable low-latency may be used interchangeably herein.

In some examples, a UE 115 may be configured to support communicating directly with other Ues 115 via a device-to-device (D2D) communication link 135 (e.g., in accordance with a peer-to-peer (P2P) , D2D, or sidelink protocol) . In some examples, one or more Ues 115 of a group that are performing D2D communications may be within the coverage area 110 of a network entity 105 (e.g., a base station 140, an RU 170) , which may support aspects of such D2D communications being configured by (e.g., scheduled by) the network entity 105. In some examples, one or more Ues 115 of such a group may be outside the coverage area 110 of a network entity 105 or may be otherwise unable to or not configured to receive transmissions from a network entity 105. In some examples, groups of the Ues 115 communicating via D2D communications may support a one-to-many (1: M) system in which each UE 115 transmits to each of the other Ues 115 in the group. In some examples, a network entity 105 may facilitate the scheduling of resources for D2D communications. In some other examples, D2D communications may be carried out between the Ues 115 without an involvement of a network entity 105.

The core network 130 may provide user authentication, access authorization, tracking, Internet Protocol (IP) connectivity, and other access, routing, or mobility functions. The core network 130 may be an evolved packet core (EPC) or 5G core (5GC) , which may include at least one control plane entity that manages access and mobility (e.g., a mobility management entity (MME) , an access and mobility management function (AMF) ) and at least one user plane entity that routes packets or interconnects to external networks (e.g., a serving gateway (S-GW) , a Packet Data Network (PDN) gateway (P-GW) , or a user plane function (UPF) ) . The control plane entity may manage non-access stratum (NAS) functions such as mobility, authentication, and bearer management for the Ues 115 served by the network entities 105 (e.g., base stations 140) associated with the core network 130. User IP packets may be transferred through the user plane entity, which may provide IP address allocation as well as other functions. The user plane entity may be connected to IP services 150 for one or more network operators. The IP services 150 may include access to the Internet, Intranet (s) , an IP Multimedia Subsystem (IMS) , or a Packet-Switched Streaming Service.

The wireless communications system 100 may operate using one or more frequency bands, which may be in the range of 300 megahertz (MHz) to 300 gigahertz (GHz) . Generally, the region from 300 MHz to 3 GHz is known as the ultra-high frequency (UHF) region or decimeter band because the wavelengths range from approximately one decimeter to one meter in length. UHF waves may be blocked or redirected by buildings and environmental features, which may be referred to as clusters, but the waves may penetrate structures sufficiently for a macro cell to provide service to the Ues 115 located indoors. Communications using UHF waves may be associated with smaller antennas and shorter ranges (e.g., less than 100 kilometers) compared to communications using the smaller frequencies and longer waves of the high frequency (HF) or very high frequency (VHF) portion of the spectrum below 300 MHz.

The wireless communications system 100 may utilize both licensed and unlicensed RF spectrum bands. For example, the wireless communications system 100 may employ License Assisted Access (LAA) , LTE-Unlicensed (LTE-U) radio access technology, or NR technology using an unlicensed band such as the 5 GHz industrial, scientific, and medical (ISM) band. While operating using unlicensed RF spectrum bands, devices such as the network entities 105 and the Ues 115 may employ carrier sensing for collision detection and avoidance. In some examples, operations using unlicensed bands may be based on a carrier aggregation configuration in conjunction with component carriers operating using a licensed band (e.g., LAA) . Operations using unlicensed spectrum may include downlink transmissions, uplink transmissions, P2P transmissions, or D2D transmissions, among other examples.

A network entity 105 (e.g., a base station 140, an RU 170) or a UE 115 may be equipped with multiple antennas, which may be used to employ techniques such as transmit diversity, receive diversity, multiple-input multiple-output (MIMO) communications, or beamforming. The antennas of a network entity 105 or a UE 115 may be located within one or more antenna arrays or antenna panels, which may support MIMO operations or transmit or receive beamforming. For example, one or more base station antennas or antenna arrays may be co-located at an antenna assembly, such as an antenna tower. In some examples, antennas or antenna arrays associated with a network entity 105 may be located at diverse geographic locations. A network entity 105 may include an antenna array with a set of rows and columns of antenna ports that the network entity 105 may use to support beamforming of communications with a UE 115. Likewise, a UE 115 may include one or more antenna arrays that may support various MIMO or beamforming operations. Additionally, or alternatively, an antenna panel may support RF beamforming for a signal transmitted via an antenna port.

Beamforming, which may also be referred to as spatial filtering, directional transmission, or directional reception, is a signal processing technique that may be used at a transmitting device or a receiving device (e.g., a network entity 105, a UE 115) to shape or steer an antenna beam (e.g., a transmit beam, a receive beam) along a spatial path between the transmitting device and the receiving device. Beamforming may be achieved by combining the signals communicated via antenna elements of an antenna array such that some signals propagating along particular orientations with respect to an antenna array experience constructive interference while others experience destructive interference. The adjustment of signals communicated via the antenna elements may include a transmitting device or a receiving device applying amplitude offsets, phase offsets, or both to signals carried via the antenna elements associated with the device. The adjustments associated with each of the antenna elements may be defined by a beamforming weight set associated with a particular orientation (e.g., with respect to the antenna array of the transmitting device or receiving device, or with respect to some other orientation) .

A network entity 105 or a UE 115 may use beam sweeping techniques as part of beamforming operations. For example, a network entity 105 (e.g., a base station 140, an RU 170) may use multiple antennas or antenna arrays (e.g., antenna panels) to conduct beamforming operations for directional communications with a UE 115. Some signals (e.g., synchronization signals, reference signals, beam selection signals, or other control signals) may be transmitted by a network entity 105 multiple times along different directions. For example, the network entity 105 may transmit a signal according to different beamforming weight sets associated with different directions of transmission. Transmissions along different beam directions may be used to identify (e.g., by a transmitting device, such as a network entity 105, or by a receiving device, such as a UE 115) a beam direction for later transmission or reception by the network entity 105.

Some signals, such as data signals associated with a particular receiving device, may be transmitted by transmitting device (e.g., a transmitting network entity 105, a transmitting UE 115) along a single beam direction (e.g., a direction associated with the receiving device, such as a receiving network entity 105 or a receiving UE 115) . In some examples, the beam direction associated with transmissions along a single beam direction may be determined based on a signal that was transmitted along one or more beam directions. For example, a UE 115 may receive one or more of the signals transmitted by the network entity 105 along different directions and may report to the network entity 105 an indication of the signal that the UE 115 received with a highest signal quality or an otherwise acceptable signal quality.

In some examples, transmissions by a device (e.g., by a network entity 105 or a UE 115) may be performed using multiple beam directions, and the device may use a combination of digital precoding or beamforming to generate a combined beam for transmission (e.g., from a network entity 105 to a UE 115) . The UE 115 may report feedback that indicates precoding weights for one or more beam directions, and the feedback may correspond to a configured set of beams across a system bandwidth or one or more sub-bands. The network entity 105 may transmit a reference signal (e.g., a cell-specific reference signal (CRS) , a channel state information reference signal (CSI-RS) ) , which may be precoded or unprecoded. The UE 115 may provide feedback for beam selection, which may be a precoding matrix indicator (PMI) or codebook-based feedback (e.g., a multi-panel type codebook, a linear combination type codebook, a port selection type codebook) . Although these techniques are described with reference to signals transmitted along one or more directions by a network entity 105 (e.g., a base station 140, an RU 170) , a UE 115 may employ similar techniques for transmitting signals multiple times along different directions (e.g., for identifying a beam direction for subsequent transmission or reception by the UE 115) or for transmitting a signal along a single direction (e.g., for transmitting data to a receiving device) .

A receiving device (e.g., a UE 115) may perform reception operations in accordance with multiple receive configurations (e.g., directional listening) when receiving various signals from a transmitting device (e.g., a network entity 105) , such as synchronization signals, reference signals, beam selection signals, or other control signals. For example, a receiving device may perform reception in accordance with multiple receive directions by receiving via different antenna subarrays, by processing received signals according to different antenna subarrays, by receiving according to different receive beamforming weight sets (e.g., different directional listening weight sets) applied to signals received at multiple antenna elements of an antenna array, or by processing received signals according to different receive beamforming weight sets applied to signals received at multiple antenna elements of an antenna array, any of which may be referred to as “listening” according to different receive configurations or receive directions. In some examples, a receiving device may use a single receive configuration to receive along a single beam direction (e.g., when receiving a data signal) . The single receive configuration may be aligned along a beam direction determined based on listening according to different receive configuration directions (e.g., a beam direction determined to have a highest signal strength, highest signal-to-noise ratio (SNR) , or otherwise acceptable signal quality based on listening according to multiple beam directions) .

In some examples, a UE 115 may support ML-based beam prediction in which the UE 115 predicts one or more channel characteristics for a set of downlink beams, which the UE 115 may report to a network entity 105 in a measurement report. The network entity 105 may use the measurement report for beam management (e.g., initial access, beam management in RRC connected mode, beam failure recovery (BFR) , radio link failure recovery response) .

To enable narrow-to-narrow machine learning (ML) -based beam prediction, in which set B beams are a subset of set A beams, the ML model of the UE 115 may undergo training in which the ML model operates on measured channel characteristics (referred to as training set B channel characteristics) of measurement signals transmitted using training set B beams and uses those measured channel characteristics to predict channel characteristics of measurement signals transmitted using training set A beams (referred to as training set A channel characteristics) . During use of the trained ML model (referred to as inference) to predict channel characteristics of measurement signals transmitted using inference set A beams (referred to as inference set A channel characteristics) , the ML model may operate on measured channel characteristics of measurement signals transmitted using inference set B beams (referred to as inference set B channel characteristics) . The performance of the ML model may suffer in the scenario in which the training set B beams are inconsistent with the inference set B beams.

According to the described techniques, a UE 115 and a network entity 105 may exchange signaling to ensure that, for a given ML model, the training set B beams are consistent with the inference set B beams.

In a first example, described more fully with reference to Figure 9, beam-mapping pattern consistency may be enabled via beam-mapping pattern negotiation between a UE 115 and a network entity 105. For instance, a UE 115 with an ML model may exchange signaling with a network entity 105 so that the network entity can determine whether a candidate inference beam-mapping pattern is consistent with the training beam-mapping pattern used for the ML model. If the candidate beam-mapping pattern is consistent with the training beam-mapping pattern, the network entity 105 may signal the UE 115 to activate the ML model (and in some examples may indicate the beam-mapping pattern to use for inference) . If the candidate beam-mapping pattern is inconsistent with the training beam-mapping pattern, the network entity may signal the UE 115 to deactivate the ML model.

In a second example, described more fully with reference to Figure 10, beam-mapping pattern consistency may be enabled by a network entity 105 providing an indication of the training beam-mapping pattern to the UE 115 (or by the UE 115 reporting the training beam-mapping pattern to the network entity 105) so that the training beam-mapping pattern is known to the network entity 105. Accordingly, when the network entity 105 activates the ML model for inference, the network entity 105 may instruct the UE 115 to use an inference beam-mapping pattern that is consistent with the training beam-mapping pattern. In some examples, a second network entity that manages ML model information for the wireless system may assist the network entity 105 with selecting a consistent inference beam-mapping pattern.

In a third example, described more fully with reference to Figure 11, beam-mapping pattern consistency may be enabled via ML model negotiation between a UE 115 and a network entity 105. For example, by communicating with the UE 115, the network entity 105 may determine a set of ML models that both A) are supported by both the UE 115 and the network entity 105 and B) have preconfigured beam-mapping pattern pairs that are consistent. So, the network entity 105 may ensure beam-mapping pattern consistency by selectively activating an ML model with an inference beam-mapping pattern that satisfies those two conditions.

Although described separately, aspects of the different examples disclosed herein may be combined and implemented together.

In some examples, a beam-mapping pattern may be described relative to a prediction cycle (described in more detail with reference to Figure 5) and may be fixed across prediction cycles or may vary across prediction cycles. As noted, a beam-mapping pattern may map set B beams to set A beams (described in more detail with reference to Figures 3 and 4) .

The wireless communications system 100 may support different definitions for beam-mapping pattern consistency. In some examples, a beam-mapping pattern pair may be regarded as consistent if the inference beam-mapping pattern is the same as the training beam-mapping pattern, a condition referred to as Consistency Condition 1 (described in more detail with reference to Figure 6) .

In some examples, a beam-mapping pattern pair may be regarded as consistent if a training interval factor D is the same as an inference interval factor D, a condition referred to as Consistency Condition 2-1 (described in more detail herein and with reference to Figure 7) . In some examples, a beam-mapping pattern pair may be regarded as consistent if there is strict consistency between the quantity of training set B beams and the quantity of inference set B beams (such that the quantity of training set B beams is the same as the quantity of inference set B beams) , a condition referred to as Consistency Condition 2-2 (1) (described in more detail herein and with reference to Figure 7) . In some examples, a beam-mapping pattern pair may be regarded as consistent if there is statistical consistency between the quantity of training set B beams and the quantity of inference set B beams, a condition referred to as Consistency Condition 2-2 (2) (described in more detail herein and with reference to Figure 7) .

In some examples, a first beam-mapping pattern may be consistent with a second beam-mapping pattern if, for each set A beam associated with the second beam-mapping pattern, the probability of that set A beam being selected as (e.g., mapped to) a set B beam for inference is within a threshold margin of the probability of the corresponding set A beam (e.g., a set A beam with the same identifier) associated with the first beam-mapping pattern being selected as a set B beam for training, a condition referred to as Consistency Condition 3 (described in more detail with reference to Figure 8) .

Figure 2 shows an example of a wireless communication system 200 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The wireless communications system 200 may implement or may be implemented by aspects of the wireless communications system 100. For example, the wireless communications system 200 may include a UE 115-a, which may be an example of a UE 115. The wireless communications system 200 may include a network entity 105-a, which may be an example of a network entity 105. The UE 115-a may communicate with the network entity 105-a to ensure beam-mapping pattern consistency across training and inference of an ML model. A beam-mapping pattern may also be referred to as a sub-sampling pattern or other suitable terminology.

The UE 115-a may communicate with the network entity 105-a using a communication link 125-a. The communication link 125-a may be an example of an NR or LTE link between the UE 115-a and the network entity 105-a. The communication link 125-a may include a bi-directional link that enables both uplink and downlink communications. For example, the UE 115-a may transmit uplink signals (e.g., uplink control signals, uplink data signals) , to the network entity 105-a using the communication link 125-a and the network entity 105-a may transmit downlink signals (e.g., downlink control signals, downlink data signals, downlink measurement signals) to the UE 115-b using the communication link 125-a. In some examples, the network entity 105-a may perform beamforming procedures to transmit downlink signals 210 to the UE 115-a via one or more beams 220.

The UE 115-a may perform beam prediction using an ML model (e.g., ML model #X, where #X refers to the ML model identifier) to predict future measurements of channel characteristics associated with the beams 220. A training procedure and inference procedure for the ML model is described with reference to the UE 115-a and the network entity 105-a, but in some examples the ML model may be trained by one or more other devices and uploaded to the UE 115-a after training.

To facilitate training of the ML model, the network entity 105-a (or other training device) may use set A beams (which may include one or more of the beams 220) to transmit measurement signals 210 (e.g., synchronization signal blocks (SSBs) , channel state information (CSI) reference signals (CSI-RSs) ) in measurement resources (e.g., SSB resources, CSI-RS resources, virtual resources) . For example, for each sample occasion 205, the network entity 105-a may transmit CSI-RS in corresponding measurement resources (labeled CSI-RS resource indicator (CRI) ) using 32 Set A beams. Other types of measurement signals and quantities of Set A beams are contemplated and within the scope of the present disclosure.

During training, the UE 115-a (or other training device) may measure the measurement signals in the measurement resources to determine channel characteristics (e.g., RSRP, layer 1 RSRP (L1-RSRP) , signal-to-interference-and-noise ratio (SINR) , the N measurement resources with the highest L1-RSRP or SINR) associated with the set A beams. For example, the UE 115-a may measure the CSI-RS L1-RSRP of each CRI (for a total of 32 CSI-RS L1-RSRP) . However, the UE 115-a may only input (e.g., as features) into the ML model the channel characteristics associated with the training set B beams, which are a subset of the set A beams as defined by the training beam-mapping pattern. The ML model may operate on the measured channel characteristics associated with the training set B beams to predict channel characteristics (e.g., output features) associated with some or all of the set A beams, which the ML model may check against the actual channel characteristics measured for the set A beams (which may serve as ground truth labels) .

During inference for the set A beams, the UE 115-a may measure a subset of the measurement signals in the measurement resources (e.g., the CRI associated with the inference set B beams) to determine channel characteristics associated with the inference set B beams. The UE 115-a may input into the ML model the channel characteristics for the inference set B beams, which are a subset of the set A beams as defined by the inference beam-mapping pattern. The ML model may operate on the channel characteristics of the inference set B beams to predict channel characteristics of the set A beams or a remainder of the set A beams (e.g., the set A beams not mapped to set B beams) . The UE 115-a may report the measured channel characteristics, the predicted channel characteristics, or both, in one or more measurement reports 215, which may be transmitted on a prediction cycle-basis, measurement occasion-basis, or other basis. In some examples, a measurement report 215 may be included in an uplink control information (UCI) message or a MAC-CE message. The network entity 105-a may use the reported channel characteristics for beam management.

To improve performance of the ML model, the UE 115-a and the network entity 105-a may work together, a process referred to generally as negotiation, to ensure that the training beam-mapping pattern is consistent with the inference beam-mapping pattern (e.g., so that the training set B beams are consistent with the inference set B beams) . For example, the UE 115-a and the network entity 105-a may implement aspects of the process flows described with reference to Figures 9–11. Before the UE 115-a and the network entity 105-a engage in negotiation, the UE 115-a may indicate to the network entity 105-a (e.g., via a UE capability report) that the UE 115-a is capable of negotiation. In response to the UE capability report, the network entity 105-a may use a control message to enable or disable negotiation at the UE 115-a. The network entity 105-a may transmit the control message dynamically (e.g., as a DCI message) or semi-statically (e.g., as an RRC message or MAC-CE message) .

For performance purposes, the same set A beams may be used for training of the AI model and for inference performed by the AI model. A same set of beams refers to beams that have resource quantity consistency, beam-shape consistency, order consistency, or a combination thereof.

Beams may have resource quantity consistency (also referred to as number consistency) if the same quantity of measurement resources (e.g., CSI-RS resources) are configured as set A beams (e.g., across training and inference) . Beams may have beam-shape consistency if (with respect to two corresponding measurement resources of the measurement resources associated with set A beams for training and inference) the difference between the relative pointing directions of the beams is below a threshold (also referred to as a predefined tolerance) , if the difference between the respective beamwidths is below a threshold, or both. Beams may have order consistency if the UE 115-a uses L1-RSRP with respect to measurement resource indicator n (e.g., SSB resource indicator (SSBRI) = n, CRI = n, virtual resource indicator (VRI) = n) for Set-A beams as a label with respect to the nth output feature during training, and if the UE 115-a uses the value from the nth output feature to derive prediction results with respect to measurement resource index n for set A beams during inference.

Figure 3 shows an example of a beam-mapping patterns 300 that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The beam-mapping patterns 300 may be examples of beam-mapping patterns used (e.g., for ML model training, for ML model inference) by a UE 115. The beam-mapping patterns 300 may map set B beams to set A beams such that the set B beams are a subset of the set A beams. Although described with reference to 32 set A beams (denoted by circles of any shading) and 8 set B beams (denoted by black circles) , a beam-mapping pattern may include any quantity of set A beams and any quantity of set B beams.

Beam-mapping pattern 300-a may be a fixed beam-mapping pattern in which the set B beams across prediction cycles are the same. For example, the set B beams in prediction cycles n, n+1, n+2, and n+3 may be the same (e.g., have the same set B beam identifiers) . Beam-mapping pattern 300-b may be a varying beam-mapping pattern in which the set B beams across prediction cycles are different. For example, the set B beams for each of prediction cycles n, n+1, n+2, and n+#may be different (e.g., have different set B beam identifiers) . In some examples, a beam-mapping pattern may repeat as illustrated by beam-mapping pattern 300-c, in which the beam-mapping pattern repeats every four prediction cycles (collectively referred to as a measurement occasion) . Although shown encompassing four prediction cycles, a measurement occasion may include any quantity of prediction cycles, including one prediction cycle.

Within a measurement resource set configured (e.g., via RRC) or indicated (e.g., via DCI, via MAC control element (MAC-CE) ) by a network entity, a UE may identify a first quantity of measurement resources (e.g., SSB resources, CSI-RS resources, virtual resources) as set A beams associated with an ML model (e.g., ML model #X) . The kth set A beam, where k is a positive integer and refers to the set A beam identifier or index, may be defined as the beam associated with the kth entry within the measurement resource set. In the scenario in which SSBs/CSI-RSs are considered, the periodicity of the SSBs/CSI-RSs resources may be the same or substantially the same. In the scenario in which virtual resources are considered, the measurement signals (e.g., SSBs/CSI-RSs) may not be transmitted.

In a given prediction cycle, the UE may identify a second quantity of measurement resources as set B beams associated with the ML model (e.g., ML model #X) . In some examples, the UE may identify the second quantity of measurement resources as set B beams associated with the ML model based on a measurement resource configuration configured or indicated by the network entity 105. The jth set B beam, where j is a positive integer and refers to the set B beam identifier or index, may be defined as the beam associated with the jth entry within the measurement resource set.

A UE may identify set B beams at different prediction cycles based on the same measurement resource set or based on different measurement resource sets (e.g., as shown in Figure 4) . In the scenario in which the UE identifies set B beams at different prediction cycles based on different measurement resource sets, the quantity of measurement resources within the different involved measurement resource sets may be the same or different. As illustrated, set B beams may be based on the same temporal periodicity but different temporal offsets. As noted, the UE may use channel characteristic measurements (e.g., RSRP, L1-RSRP, SINR) of the set B beams to predict channel characteristics associated with the set A beams.

Figure 4 shows an example of beam-mapping patterns 400 that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The beam-mapping patterns 400 may be examples of beam-mapping patterns used (e.g., for ML model training, for ML model inference) by a UE 115. The beam-mapping patterns 400 may map set B beams to set A beams and may be examples of time-varying beam-mapping patterns in which set B beams vary across prediction cycles. Although described with reference to 32 set A beams and 8 set B beams (denoted by black circles) , a beam-mapping pattern may include any quantity of set A beams and any quantity of set B beams. A time-varying beam-mapping pattern may include sub-patterns denoted P1, P2, P3, and P4.

In some examples, a beam-mapping pattern (which maps set B beams to Set A beams) may be defined such that, for each set B beam identifier, there is one (e.g., a single) set A beam identifier with the considered set B beams (such that different involved set B beams are associated with different set A beam identifiers) . In some examples, a beam-mapping pattern may be further defined such that the beam shapes associated with the linked (e.g., mapped) set B beam identifier and the set A beam identifier are identical, where beam shape may be considered as referring to the absolute pointing direction, the angular-specific beamforming gains, or both, of a corresponding beam.

A network entity 105 may indicate a beam-mapping pattern using RRC signaling, MAC-CE signaling, or DCI signaling. In some examples, a network entity 105 may use RRC-based beam-mapping pattern signaling in which each resource associated with a set B beam identifier is further RRC-configured with the linked set A beam identifier. In some examples, a set A beam identifier may be updated via MAC-CE signaling or via DCI signaling (e.g., for measurement resources with respect to set B beams) .

As noted, a beam-mapping pattern may be a time-varying beam-mapping pattern in that the beam-mapping pattern may vary (e.g., deterministically, randomly) across different measurement occasions or prediction cycles. The beam-mapping pattern variation may be signaled (e.g., DCI signaled, RRC signaled, MAC-CE signaled) to a UE 115 by a network entity 105.

In a first example of a time-varying beam-mapping pattern, illustrated by beam-mapping pattern 400-a, a single measurement resource set (e.g., a single CSI-RS resource set with 32 resources) may be configured as set B beams. In such an example, the set B beams in different prediction cycles may be configured via different measurement resources in the measurement resource set. The temporal offsets (e.g., between set B resources) may be short for intra-prediction cycle set B beams and may be longer for inter-cycle set B beams. In some examples, the beam-mapping pattern 400-a may be indicated for each respective measurement resource.

In a second example of a time-varying beam-mapping pattern, illustrated by beam-mapping pattern 400-b, a single measurement resource set (e.g., a single CSI-RS resource set with 8 resources) may be configured as set B beams. In such an example, the set B beams in different prediction cycles may reuse the same resources within the measurement resource set, and the prediction cycle-specific beam-mapping patterns may be indicated separately. For instance, L prediction cycles, in which L is a positive integer, may be periodically repeated and the beam-mapping pattern (e.g., the sub-patterns P1 through P4) for each of the repeated prediction cycles may configured or indicated by the network entity.

In a third example of a time-varying beam-mapping pattern, illustrated by beam-mapping pattern 400-c, multiple measurement resource sets (e.g., multiple CSI-RS resource sets with 8 resources each) may be configured as set B beams. In such an example, the set B beams in the same prediction cycle may be configured in the same measurement resource set, and the set B beams with respect to different prediction cycles may be associated with different measurement resource set. In some examples a similar configuration may be implemented for L periodically repeated prediction cycles.

So, a beam-mapping pattern may vary across different prediction cycles.

Figure 5 shows an example of prediction cycles 500 that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. A first option (Option 1) for defining adjacent prediction cycles 500 is illustrated on the left and a second option (Option 2) for defining adjacent prediction cycles 500 is illustrated on the right. The predictions cycles 500 may include four prediction cycles labeled PC#1 through PC#4 and in some examples may have corresponding prediction cycle identifiers. Each prediction cycle may have an associated measurement report, denote report 1 through report 4, that includes channel characteristic information (e.g., measured channel characteristics, predicted channel characteristics) for that prediction cycle. The reports may be transmitted during reporting occasions (e.g., in UCI messages, in MAC-CE message) . The channel characteristic information may be generated by an ML model and may be based on set B beams.

In Option 1, adjacent prediction cycles may include overlapping durations (e.g., adjacent prediction cycles may be overlapping in the time domain) . For example, a UE that implements Option 1 may start monitoring (e.g., measuring) set B beams for the Kth prediction cycle, in which K is a positive integer, before the UE transmits the report for the (K-1) th prediction cycle. In Option 2, adjacent prediction cycles may include non-overlapping durations (e.g., adjacent prediction cycles may be non-overlapping in the time domain) . For example, a UE that implements Option 2 may start monitoring (e.g., measuring) set B beams for the Kth prediction cycle after the UE has transmitted the report for the (K-1) th prediction cycle.

The first prediction cycle (e.g., PC#1) may start when the UE receives the first symbol among all set B beams for the first prediction cycle. In some examples, the first prediction cycle may end at the last symbol of the first report (e.g., report 1) , which may include predicted channel characteristic information (e.g., L1-RSRP, layer 1 SINR (LI-SINR) , the N measurement resources, in which N is a positive integer, with the highest L1-RSRP or L1-SINR) for the set A beams. The predicted channel characteristics included in the first report may be based on the set B beams received at least a duration 505 before the first symbol of the first report. The duration 505 may be X symbols or X slots in which X is a positive integer. In some examples, the duration 505 may be based on a capability of the UE.

The Kth prediction cycle may start at the first symbol among all set B beams used to derive the channel characteristic information (with respect to the Kth report) . In some examples, the Kth prediction cycle may end at the last symbol of the Kth report (for the Kth prediction cycle) , which may include predicted channel characteristic information (e.g., L1-RSRP, LI-SINR) , the N measurement resources with the highest L1-RSRP or L1-SINR) for the set A beams. The predicted channel characteristics included in the Kth report may be based on the set B beams received at least a duration 505 before the first symbol of the Kth report, where the duration 505 may be X symbols or X slots. As noted, the duration 505 may be based on a capability of the UE.

For a given prediction cycle, a UE and network entity may coordinate so that the set B beams used for training and the set B beams used for inference are consistent (e.g., satisfy a consistency condition as described with reference to Figures 9–11) . In some examples, a consistency condition may also be referred to as a consistency rule or other suitable terminology.

Figure 6 shows an example of beam-mapping patterns 600 that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The beam-mapping patterns 600 may be examples of beam-mapping patterns used for ML-based beam prediction. The beam-mapping patterns 600 may be examples of training beam- mapping patterns (which map training set B beams to set A beams) or inference beam-mapping patterns (which map inference set B beams to set A beams) in a narrow-to-narrow beam scenario (in which set B beams are a subset of the set A beams) .

A device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam-mapping patterns satisfy a consistency condition, such as Consistency Condition 1.

According to Consistency Condition 1, a first beam-mapping pattern may be consistent with a second beam-mapping pattern if, for each prediction cycle, the first beam-mapping pattern is the same (e.g., temporally, spatially) as the second beam-mapping pattern. Put another way, starting from the first prediction cycle, the set A beams selected as (e.g., mapped to) set B beams for each prediction cycle should be strictly identical (e.g., the same) across training and inference. For instance, in a given prediction cycle (e.g., PC#1) a first beam-mapping pattern may be consistent with a second beam-mapping pattern if the identifiers of the set B beams in the first beam-mapping pattern (e.g., set B beam IDs 1, 5, 9, 13, 17, 21, 25, and 29) match the identifiers of the set B beams in the second beam-mapping pattern.

So, based on Consistency Condition 1, the device may determine that beam-mapping pattern 600-a is consistent with beam-mapping pattern 600-b and may determine that beam-mapping pattern 600-c is inconsistent with both beam-mapping pattern 600-a and beam-mapping pattern 600-b.

Figure 7 shows an example of beam-mapping patterns 700 that support beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The beam-mapping patterns 700 may be examples of beam-mapping patterns used for ML-based beam prediction. The beam-mapping patterns 700 may be examples of training beam-mapping patterns (which map training set B beams to set A beams) or inference beam-mapping patterns (which map inference set B beams to set A beams) in a narrow-to-narrow beam scenario (in which set B beams are a subset of the set A beams) .

A device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam-mapping patterns satisfy a consistency condition, such as Consistency Condition 2-1, Consistency Condition 2-2 (1) , or Consistency Condition 2-2 (2) .

In some examples, a beam-mapping pattern may be a function of or associated with an interval factor D (also referred to as a sub-sampling factor) that is a positive integer and that represents an interval between the set A beam identifiers mapped to set B beams. For example, the beam-mapping pattern may define the set B beams as the set A beams with identifiers {1+d, 1+D+d, 1+2D+d, ... } , where the offset d is a positive integer that represents the offset for the first (e.g., numerically) set A beam identifier mapped to a set B beam. So, the identifiers mapped to set B beams may be shifted across prediction cycles according to the offset d even if the interval factor is the same for those prediction cycles.

In Figure 7, beam-mapping pattern 700-a and beam mapping pattern 700-b may each be associated with an interval factor D = 4 whereas beam-mapping pattern 700-c may be associated with an interval factor D = 4.

The offset d (where d ∈ {0, 1, 2, …, D-1} ) may be fixed or varied (e.g., deterministically, randomly) across prediction cycles. In some examples, as illustrated by beam-mapping pattern 700-c, the offset d may be uniformly varied across prediction cycles (e.g., d may be uniformly varied across prediction cycles among all candidates in {0, 1, 2, …, D-1) . In some examples, the offset d may vary across prediction cycles (e.g., d may vary across prediction cycles among all candidates {0, 1, 2, …, D-1} based on particular distributions, where the distributions are a part of the interval factor D) . In some examples, the interval factor D may be represented by the variation pattern of the offset d across prediction cycles.

A device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam-mapping patterns satisfy a consistency condition, such as Consistency Condition 2-1.

According to Consistency Condition 2-1, a first beam-mapping pattern may be consistent with a second beam-mapping pattern if, for each prediction cycle, the interval factor D associated with the first beam-mapping pattern is equal to the interval factor D associated with the second beam-mapping pattern. For instance, the set B beams of each prediction cycle may be based on selection of set A beams with a fixed interval factor that is consistent across training and inference (however, the quantity of set B beams across different prediction cycles may be different if the total quantity of set A beams cannot be divided by the interval factor) . So, according to Consistency Condition 2-1, beam-mapping pattern 700-a (with interval factor D = 4) may be consistent with beam-mapping pattern 700-b (with interval factor D =4) . Further, beam-mapping pattern 700-c (with interval factor D =2) may be inconsistent with beam mapping pattern 700-a and beam-mapping pattern 700-b.

Alternatively, a device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam-mapping patterns satisfy a different consistency condition, such as Consistency Condition 2-2 (1) .

According to Consistency Condition 2-2 (1) , a first beam-mapping pattern (e.g., a training beam-mapping pattern) may be consistent with a second beam-mapping pattern (e.g., an inference beam-mapping pattern) if, for each prediction cycle across the predictions cycles, the quantity of set B beams associated with first beam-mapping patten is the same as the quantity of set B beams associated with the second beam-mapping pattern (e.g., if the quantity of set B beams is absolutely fixed across different prediction cycles and also consistent across training and inference) . Put another way, the quantity of set B beams of each prediction cycle may be fixed across different prediction cycles and consistent across training and inference. So, according to Consistency Condition 2-2 (1) , beam-mapping pattern 700-a (with eight set B beams for each prediction cycle) may be consistent with beam-mapping pattern 700-b (with eight set B beams for each prediction cycle) , even though the set B beam identifiers in a given prediction cycle are different. However, beam-mapping pattern 700-c (with sixteen set B beams) may be inconsistent with beam mapping pattern 700-a and beam-mapping pattern 700-b.

Alternatively, a device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam- mapping patterns satisfy a different consistency condition, such as Consistency Condition 2-2 (2) .

According to Consistency Condition 2-2 (2) , a first beam-mapping pattern (e.g., a training beam-mapping pattern) may be consistent with a second beam-mapping pattern (e.g., an inference beam-mapping pattern) if, for each prediction cycle across the predictions cycles, the quantity of set B beams associated with first beam-mapping patten and the quantity of set B beams associated with the second beam-mapping pattern are statistically fixed (e.g., such that the mean or variance of the quantity of set B beams across prediction cycles and across training/inference is consistent) .

As an example of Consistency Condition 2-2 (2) , beam-mapping pattern 700-a may be consistent with beam-mapping pattern 700-b if, for each prediction cycle across the prediction cycles, the statistical mean (e.g., average) of the quantity of set B beams associated with beam-mapping pattern 700-a is within a threshold margin of the statistical mean of the quantity of set B beams associated with beam-mapping pattern 700-b. As another example of Consistency Condition 2-2 (2) , beam-mapping pattern 700-a may be consistent with beam-mapping pattern 700-b if, for each prediction cycle across the prediction cycles, the statistical variance of the quantity of set B beams associated with beam-mapping pattern 700-a is within a threshold margin of the statistical variance of the quantity of set B beams associated with beam-mapping pattern 700-b. A threshold may also be referred to as a tolerance or other suitable terminology.

So, according to Consistency Condition 2-2 (2) , a first beam-mapping pattern may be consistent with a second beam-mapping pattern if, for each prediction cycle across the prediction cycles, a first statistical value associated with the quantity of training set B beams is within a threshold margin of a second statistical value associated with the quantity of inference set B beams.

A device may use a consistency condition such as Consistency Condition 2-1, Consistency Condition 2-2 (1) , or Consistency Condition 2-2 (2) to determine the consistency between a pair of beam-mapping patterns for training and inference.

Figure 8 shows an example of probabilities 800 associated with beam-mapping patterns in accordance with one or more aspects of the present disclosure. The beam-mapping patterns may be examples of beam-mapping patterns used for ML-based beam prediction. The beam-mapping patterns 700 may be examples of training beam-mapping patterns (which map training set B beams to set A beams) or inference beam-mapping patterns (which map inference set B beams to set A beams) in a narrow-to-narrow beam scenario (in which set B beams are a subset of the set A beams) .

The set A beams associated with a beam-mapping pattern may have a probability p (Train) of being selected as (e.g., mapped to) set B beams for training. For example, each set A beam ID associated with the beam-mapping pattern 1 may have a respective probability p (Train) of being mapping to a set B beam for training. In some examples, the probabilities p (Train) may be the same (e.g., 3.15%) for each set A beam associated with the beam-mapping pattern 1.

The set A beams associated with a beam-mapping pattern may have a probability p (Inf) of being selected as (e.g., mapped to) set B beams for inference. For example, each set A beam ID associated with the beam-mapping pattern 2 may have a respective probability p (Inf) of being mapping to a set B beam for inference. Similarly, each set A beam ID associated with the beam-mapping pattern 3 may have a respective probability p (Inf) of being mapping to a set B beam for inference.

A device (e.g., a network entity, a UE) may determine whether a beam-mapping pattern pair (e.g., comprising a training beam-mapping pattern and an inference beam-mapping pattern) is consistent by determining whether the two beam-mapping patterns satisfy a consistency condition, such as Consistency Condition 3.

According to Consistency Condition 3, a first beam-mapping pattern may be consistent with a second beam-mapping pattern if, for each set A beam associated with the second beam-mapping pattern (and for one or more prediction cycles, or for each prediction cycle) , the probability p (Inf) of that set A beam is within a threshold margin of the probability p (Train) of the corresponding set A beam associated with the first beam-mapping pattern. Put another way, if the probability of the nth set A beam being selected as a set B beam among all set A beams (across different prediction cycles during training and inference) is {Ptrain (n) , Pinference (n) } , then the Consistency Condition 3 is satisfied if |Ptrain (n) –Pinference (n) | < Tld (n) , where Tdl (n) refers to the threshold margin for the nth set A beam, and n is a positive integer. The threshold margin Tdl may be fixed across different values of n (e.g., fixed across different set A beam IDs) or may vary across different values of n (e.g., vary across different set A beam IDs) . So, each set A beam ID may respectively have an associated threshold margin Tdl, which may be the same or different across set A beam IDs.

According to Consistency Condition 3, in the illustrated example, if the threshold margin Tdl for each set A beam ID is . 025%, the device may determine that the beam-mapping pattern 1 is consistent with the beam-mapping pattern 2 (e.g., because each set A beam ID associated with beam-mapping pattern 2 has p (Inf) within .025%of p (Train) of the corresponding set A beam ID associated with beam mapping pattern 1) and inconsistent with the beam-mapping pattern 3. Put another way, the device may determine that the beam-mapping pattern 1 is consistent with the beam mapping pattern 2 because the difference between A) the likelihood of the mth set A beam being mapped to a set B beam across prediction cycles during training, and B) the likelihood of the mth set A beam being mapped to a set B beam across prediction cycles during inference is under the threshold margin for the mth set A beam, in which m is a positive integer.

So, a device may use a consistency condition such as Consistency Condition 3 to determine the consistency between a pair of beam-mapping patterns.

Figure 9 shows an example of a process flow 900 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The process flow 900 may be implemented by a UE 115-b and a network entity 105-b. The UE 115-b may be configured with an ML model #X. The UE 115-b and the network entity 105-b may exchange information about supported beam-mapping patterns to enable activation or deactivation of the ML model, a process that may be referred to as beam negotiation.

Before implementing aspects of the process flow 900, the UE 115-b may indicate to the network entity 105-b (e.g., via a UE capability report) that the UE 115-b is capable of beam negotiation. In response to the UE capability report, the network entity 105-b may use a control message to enable or disable beam negotiation at the UE 115-b. The network entity 105-b may transmit the control message dynamically (e.g., as a DCI message) or semi-statically (e.g., as an RRC message or MAC-CE message) .

At 905, the UE 115-b, the network entity 105-b, or both, may train the ML model. Alternatively, (e.g., if the ML model is already trained) , the network entity 105-b (or some other device) may indicate the ML model to the UE 115-b. Regardless of how training occurs, the ML model may be trained with a first beam-mapping pattern referred to as the training beam-mapping pattern.

At 910, the network entity 105-b may select a consistency condition to use for determining the consistency between the training beam-mapping pattern and potential inference beam-mapping patterns. In some examples, the consistency condition may be selected based on a recommendation from the UE 115-b. In some examples, the consistency condition may be configured at the network entity 105-b. At 915, the network entity 105-b may transmit an indication of the consistency condition to the UE 115-b. The indication of the consistency condition may be included in an RRC message, a DCI message, or a MAC-CE message.

At 920, the UE 115-b and the network entity 105-b may engage in beam-mapping pattern negotiation. The negotiation may be network entity-initiated or UE-initiated.

As an example of network entity-initiated negotiation, the network entity 105-b may indicate one or more candidate inference beam-mapping patterns to the UE 115-b. In response, the UE 115-b may indicate whether the UE 115-b supports the candidate inference beam-mapping patterns (e.g., the UE 115-b may indicate which of the candidate inference beam-mapping patterns the UE 115-b supports) .

In some examples, the network entity 105-b may indicate a candidate inference beam-mapping pattern (which may be a fixed beam-mapping pattern, such as beam-mapping pattern 300-a, or a varying beam-mapping pattern such as beam-mapping pattern 300-b) by indicating the specific beam-mapping pattern. Such indication may be compatible with any consistency condition. In some examples the network entity 105-b may indicate a candidate beam-mapping pattern on a per-prediction cycle basis.

In some examples (e.g., if Consistency Condition 2-1 is selected) , the network entity 105-b may indicate a candidate inference beam-mapping pattern by indicating the interval factor D associated with the candidate inference beam-mapping pattern (e.g., for one or more prediction cycles) . In some examples, (e.g., if Consistency Condition 2-2 (1) is selected) , the network entity 105-b may indicate a candidate inference beam-mapping pattern by indicating the quantity of training set B beams associated with the candidate inference beam-mapping pattern (e.g., for each prediction cycle) . In some examples (e.g., if Consistency Condition 2-2 (2) is selected) , the network entity 105-b may indicate a candidate inference beam-mapping pattern by indicating the mean or the variance of the training set B beams associated with the candidate inference beam-mapping pattern (e.g., for each prediction cycle) .

In some examples (e.g., if Consistency Condition 3 is selected) , the network entity 105-b may indicate a candidate inference beam-mapping pattern by indicating the probabilities of each respective set A beam (associated with the candidate inference beam-mapping pattern) being selected as (e.g., mapped to) a set B beam. For instance, the network entity 105-b may indicate the probability p (Inf) for each set A beam ID associated with the candidate inference beam-mapping pattern. In some examples, the network entity 105-b may indicate a common threshold margin Tld for the set A beams or network entity 105-b may indicate a respective threshold margin Tld for each set A beam. In some examples, the UE 115-b may indicate the probabilities of each respective set A beam being selected as (e.g., mapped to) a set A beam ID associated with a beam-mapping pattern supported by the UE 115-b. For instance, the UE 115-b may indicate the probability p (Inf) for each set A beam ID associated with the supported beam-mapping pattern. Alternatively, the UE 115-b may indicate the difference between the probabilities indicated by the network entity 105-b and the probabilities associated with a beam-mapping pattern supported by the UE 115-b. In some examples, multiple allowable combinations of probabilities and/or tolerances may be indicated.

As an example of UE-initiated negotiation, the UE 115-b may indicate a set of one or more beam-mapping patterns supported by the UE 115-b. In some examples, the UE 115-b may indicate a supported beam-mapping pattern (which may be a fixed beam-mapping pattern, such as beam-mapping pattern 300-a, or a varying beam-mapping pattern such as beam-mapping pattern 300-b) by indicating the specific beam-mapping pattern. Such indication may be compatible with any consistency condition. In some examples the UE 115-b may indicate a candidate beam-mapping pattern on a per-prediction cycle basis.

In some examples (e.g., if Consistency Condition 2-1 is selected) , the UE 115-b may indicate a supported beam-mapping pattern by indicating the interval factor D associated with the beam-mapping pattern (e.g., for one or more prediction cycles) . In some examples, (e.g., if Consistency Condition 2-2 (1) is selected) , the UE 115-b may indicate a supported beam-mapping pattern by indicating the quantity of training set B beams associated with the beam-mapping pattern (e.g., for each prediction cycle) . In some examples (e.g., if Consistency Condition 2-2 (2) is selected) , the UE 115-b may indicate a supported beam-mapping pattern by indicating the mean or the variance of the training set B beams associated with the beam-mapping pattern (e.g., for each prediction cycle) .

In some examples (e.g., if Consistency Condition 3 is selected) , the UE 115-b may indicate a supported beam-mapping pattern by indicating the probabilities of each respective set A beam (associated with the beam-mapping pattern) being selected as (e.g., mapped to) a set B beam. For instance, the UE 115-b may indicate the probability p(Inf) for each set A beam ID associated with the supported beam-mapping pattern. Alternatively, the UE 115-b may indicate the difference between the probabilities indicated by the network entity 105-b and the probabilities associated with the beam-mapping pattern supported by the UE 115-b. In some examples, multiple allowable combinations of probabilities and/or tolerances may be indicated.

In either example (e.g., network entity-initiated negotiation or UE-initiated negotiation) , the network entity 105-b may indicate the set of one or more candidate beam-mapping patterns using an RRC message, a MAC-CE message, or a DCI message. In either example (e.g., network entity-initiated negotiation or UE-initiated negotiation) , the UE 115-b may indicate the set of one or more supported beam-mapping patterns using a UE capability report or using a dynamic update (e.g., using a MAC-CE message, using a UCI message) . In some examples, the UE 115-b may further indicate the identifier of the ML model associated with the indicated beam-mapping patterns. In some examples (e.g., in network entity-initiated negotiation) , the set of supported beam-mapping patterns indicated by the UE 115-b may be a subset of the candidate beam-mapping patterns indicated by the network entity 105-b.

At 925, the network entity 105-b may determine whether a second beam-mapping pattern is consistent with the training beam-mapping pattern. The second beam-mapping pattern may be one of the candidate inference beam-mapping patterns indicated by the network entity 105-b, may be one of the supported beam-mapping patterns indicated by the UE 115-b, or both. If the second beam-mapping pattern is inconsistent with the training beam-mapping pattern (and no other beam-mapping pattern is consistent with the training beam-mapping pattern) , the network entity 105-b may transmit, at 935, activation information that deactivates the ML model (assuming the ML model is already activated) .

If the second beam-mapping pattern is consistent with the training beam-mapping pattern, the network entity 105-b may, at 930, select the second beam-mapping pattern as the inference beam-mapping pattern for the ML model. Accordingly, at 935, the network entity 105-b may transmit activation information that activates the ML model. The activation information transmitted at 935 (whether activating or deactivating the ML model) may indicate the identifier of the ML model. The activation information may be included in an RRC message, a MAC-CE message, or a DCI message.

At 940, if the ML model is activated at 935, the network entity 105-b may indicate (e.g., in an RRC message, in a MAC-CE message, in a DCI message) the second beam-mapping pattern to the UE 115-b for use during inference of the ML model. The network entity 105-b may also (in the same message or a different message) indicate the identifier of the ML model associated with the second beam-mapping pattern.

Accordingly, at 945, the UE 115-b may use the ML model to perform ML-based beam prediction (e.g., to predict channel characteristic measurements) using the second beam-mapping pattern as the inference beam-mapping pattern. At 950, the UE 115-b may transmit a measurement report (e.g., in a UCI message, in a MAC-CE message) that indicates one or more predicted channel characteristic measurements for the set A beams associated with the second beam-mapping pattern.

So, the UE 115-b and the network entity 105 may exchange information about supported beam-mapping patterns to enable activation or deactivation of the ML model.

Figure 10 shows an example of a process flow 1000 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The process flow 1000 may be implemented by a UE 115-c, a first network entity 105-c, and a second network entity 105-d. As another example of beam negotiation, the UE 115-c, the first network entity 105-c, and potentially the second network entity 105-d, may exchange information about supported beam-mapping patterns to enable use of an inference beam-mapping pattern that is consistent with a training beam-mapping pattern. In some examples, the process flow 1000 may be used for an online ML model (e.g., an ML model that undergoes online training, that is used for online inference, or both) .

Before implementing aspects of the process flow 1000, the UE 115-c may indicate to the network entity 105-c (e.g., via a UE capability report) that the UE 115-c is capable of beam negotiation. In response to the UE capability report, the network entity 105-c may use a control message to enable or disable beam negotiation at the UE 115-c. The network entity 105-c may transmit the control message dynamically (e.g., as a DCI message) or semi-statically (e.g., as an RRC message or MAC-CE message) .

The second network entity 105-d may be responsible for managing ML models in a wireless communication system that includes the UE 115-c and the first network entity 105-c. For example, the second network entity105-d may track the identifiers of ML models supported by, or in use, by the devices in the wireless communications system. In some examples, the second network entity 105-d may also track the beam-mapping patterns associated with the identifiers of the MLs models. For instance, for a given ML model identifier the second network entity 105-d may receive an indication of the beam-mapping pattern used to train the associated ML model. Alternatively (e.g., if the ML models are standardized) , for a given ML model identifier the second network entity 105-d may be configured with the beam-mapping pattern used for training the associated ML model. The second network entity 105-d may communicate with the device in the wireless communications system to ensure that consistent beam-mapping patterns are used for ML models employed by the devices. So, ML model identifiers may be identified with assistance from the second network entity 105-d (which is a separate network entity from the first network entity 105-c) . For example, upon initiating a data collection procedure (e.g., by the first network entity 105-c) , an ML model identifier may be agreed upon by the first network entity 105-c, the second network entity 105-d, and the UE 115-c.

At 1005, the first network entity 105-c may transmit (e.g., in an RRC message, in a MAC-CE message, in a DCI message) training information to the UE 115-c. The training information may include the identifier of a trained ML model, an indication of a first beam-mapping pattern associated with training the ML model, or both. In some examples, the first network entity 105-c may receive some or all of the training information from the second network entity 105-d (e.g., as part of ML model management signaling 1015, which may occur at any point or multiple points in the process flow 1000) .

At 1010, the UE 115-c, the first network entity 105-c, or both, may train the ML model (e.g., based on the training beam-mapping pattern indicated at 1005) . Alternatively, (e.g., if the ML model is already trained) , the first network entity 105-c may indicate the ML model to the UE 115-c.

At 1020, the first network entity 105-c may select a second beam-mapping pattern for use by the ML model during inference. The second beam-mapping pattern may be selected based on the second beam-mapping pattern satisfying a consistency condition relative to the first beam-mapping pattern associated with training the ML model. The first network entity 105-c may select the second beam-mapping pattern based on information received from the second network entity 105-d (e.g., as part of ML model management signaling 1015) .

At 1025, the first network entity 105-c may transmit (e.g., in an RRC message, in a MAC-CE message, in a DCI message) activation information that activates the ML model. The activation information may include the identifier of the ML model to be used for inference, an indication of the second beam-mapping pattern associated with the ML model, or both. So, the same ML model identifier may be identified during both data collection for training and during inference (e.g., ML-based beam prediction) , which may ensure consistency between the training and inference beam-mapping patterns associated with that ML model identifier. In some examples, the beam-mapping patterns may be configured or indicated to the UE 115-c during data collection for UE-side model training.

At 1030, the UE 115-c may use the ML model associated with the ML model identifier to perform ML-based beam prediction (e.g., to predict channel characteristic measurements) using the second beam-mapping pattern as the inference beam-mapping pattern. At 1035, the UE 115-c may transmit (e.g., in a UCI message, in a MAC-CE message) a measurement report that indicates one or more predicted channel characteristic measurements for the set A beams associated with the second beam-mapping pattern.

So, the UE 115-c, the first network entity 105-c, and potentially the second network entity 105-d, may exchange information about supported beam-mapping patterns to enable use of an inference beam-mapping pattern that is consistent with a training beam-mapping pattern.

Figure 11 shows an example of a process flow 1100 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The process flow 1100 may be implemented by a UE 115-d and a network entity 105-e. In a process that may be referred to as model negotiation, the UE 115-d and network entity 105-e may exchange information about supported ML models, which may be associated with respective ML model identifiers, to enable use of an inference beam-mapping pattern that is consistent with a training beam-mapping pattern. In some examples, the process flow 1100 may be used for an offline ML model (e.g., an ML model that undergoes offline training) .

Before implementing aspects of the process flow 1100, the UE 115-d may indicate to the network entity 105-e (e.g., via a UE capability report) that the UE 115-d is capable of model negotiation. In response to the UE capability report, the network entity 105-e may use a control message to enable or disable model negotiation at the UE 115-d. The network entity 105-e may transmit the control message dynamically (e.g., as a DCI message) or semi-statically (e.g., as an RRC message or MAC-CE message) .

The UE 115-d may support one or more ML models each of which is associated with a consistent training beam-mapping pattern and inference beam-mapping pattern. For example, the UE 115-d may support ML model #X (e.g., the ML model associated with ML model identifier X) which may be associated with a training beam-mapping pattern A and inference beam-mapping pattern A, which are consistent with each other. Similarly, the UE 115-d may support ML model #Y (e.g., the ML model associated with ML model identifier Y) which may be associated with a training beam-mapping pattern B and inference beam-mapping pattern B, which are consistent with each other.

The network entity 105-e may also support one or more ML models each of which is associated with a consistent training beam-mapping pattern and inference beam-mapping pattern. In some examples, the consistent beam-mapping patterns associated with an ML model may be pre-agreed upon by a standards body or by the manufacturers of the UE 115-d and the network entity 105-e. In some examples, the ML models associated with consistent beam-mapping patterns may also be associated with one or more performance metrics (e.g., beam prediction accuracy, channel characteristic prediction accuracy) .

At 1105, the network entity 105-e may transmit (e.g., in an RRC message, in a DCI message, in a MAC-CE message) a first set of one or more ML model identifiers associated with one or more ML models that are supported by the network entity 105-e (referred to as network entity-supported ML models) . At 1110, the UE 115-d may transmit (e.g., in a UCI message, in a MAC-CE message) a second set of one or more ML model identifiers associated with one or more ML models that are supported by the UE 115-d (referred to as UE-supported ML models) . The second set of ML model identifiers may be a subset of the first set of ML model identifiers.

At 1115, the network entity 105-e may determine the ML model identifiers that are common to (e.g., included in) both the first set of ML model identifiers and the second set of ML model identifiers. At 1120, the network entity 105-e may determine which of the ML model identifiers determined at 1115 (e.g., which of the ML model identifiers associated with ML models supported by both the UE 115-d and the network entity 105-e) are associated with consistent beam-mapping patterns.

At 1125, the network entity 105-e may select an ML model identifier for activation. The network entity 105-e may select the ML model identifier for activation based on the ML model identifier A) being included in both the first set of ML model identifiers and the second set of ML model identifiers and B) being associated with a consistent pair of training and inference beam-mapping patterns. In some examples, the network entity 105-e may select the ML model identifier for activation based on the performance metrics (e.g., beam prediction accuracy, channel characteristic prediction accuracy) associated with the ML model identifier.

At 1130, the network entity 105-e may transmit (e.g., in an RRC message, in a DCI message, in a MAC-CE message) activation information to the UE 115-d. The activation information may include the ML model identifier selected at 1125. However, in some examples, the activation information may exclude an indication of beam-mapping patterns associated with the ML model identifier (e.g., because the associated beam-mapping patterns are already associated with the ML model identifier) .

At 1135, the UE 115-d may use the ML model associated with the ML model identifier to perform ML-based beam prediction (e.g., to predict channel characteristic measurements) using the inference beam-mapping pattern associated with the ML model identifier. At 1140, the UE 115-d may transmit a report (e.g., in a UCI message, in a MAC-CE message) that indicates one or more predicted channel characteristic measurements for the set A beams associated with the inference beam-mapping pattern.

So, the UE 115-d and network entity 105-e may exchange information about supported ML model identifiers to enable use of an inference beam-mapping pattern that is consistent with a training beam-mapping pattern. As long as the same ML model identifier is identified during training (e.g., offline training) and inference (e.g., online inference) , the UE 115-d and the network entity 105-e may avoid signaling that indicates their respective supported beam-mapping patterns. In some examples, the ML model identifiers with consistent beam-mapping pairs may be standardized.

Figure 12 shows a device 1205 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1205 may be an example of aspects of a UE 115 as described herein. The device 1205 may include a receiver 1210, a transmitter 1215, and a communications manager 1220. The device 1205, or one or more components of the device 1205 (e.g., the receiver 1210, the transmitter 1215, and the communications manager 1220) , may include at least one processor, which may be coupled with at least one memory, to, individually or collectively, support or enable the described techniques. Each of these components may be in communication with one another (e.g., via one or more buses) .

The receiver 1210 may provide a means for receiving information such as packets, user data, control information, or any combination thereof associated with various information channels (e.g., control channels, data channels, information channels related to beam-mapping pattern consistency for machine learning model training and inference) . Information may be passed on to other components of the device 1205. The receiver 1210 may utilize a single antenna or a set of multiple antennas.

The transmitter 1215 may provide a means for transmitting signals generated by other components of the device 1205. For example, the transmitter 1215 may transmit information such as packets, user data, control information, or any combination thereof associated with various information channels (e.g., control channels, data channels, information channels related to beam-mapping pattern consistency for machine learning model training and inference) . In some examples, the transmitter 1215 may be co-located with a receiver 1210 in a transceiver module. The transmitter 1215 may utilize a single antenna or a set of multiple antennas.

The communications manager 1220, the receiver 1210, the transmitter 1215, or various combinations thereof or various components thereof may be examples of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1220, the receiver 1210, the transmitter 1215, or various combinations or components thereof may be capable of performing one or more of the functions described herein.

In some examples, the communications manager 1220 may be configured to perform various operations (e.g., receiving, obtaining, monitoring, outputting, transmitting) using or otherwise in cooperation with the receiver 1210, the transmitter 1215, or both. For example, the communications manager 1220 may receive information from the receiver 1210, send information to the transmitter 1215, or be integrated in combination with the receiver 1210, the transmitter 1215, or both to obtain information, output information, or perform various other operations as described herein.

The communications manager 1220 may support wireless communications in accordance with examples as disclosed herein. For example, the communications manager 1220 is capable of, configured to, or operable to support a means for receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The communications manager 1220 is capable of, configured to, or operable to support a means for transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

By including or configuring the communications manager 1220 in accordance with examples as described herein, the device 1205 (e.g., at least one processor controlling or otherwise coupled with the receiver 1210, the transmitter 1215, the communications manager 1220, or a combination thereof) may support techniques for improved beam management (e.g., due to improved ML-based beam prediction) , which in turn may enable more efficient utilization of communication resources.

Figure 13 shows a device 1305 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1305 may be an example of aspects of a device 1205 or a UE 115 as described herein. The device 1305 may include a receiver 1310, a transmitter 1315, and a communications manager 1320. The device 1305, or one of more components of the device 1305 (e.g., the receiver 1310, the transmitter 1315, and the communications manager 1320) , may include at least one processor, which may be coupled with at least one memory, to support the described techniques. Each of these components may be in communication with one another (e.g., via one or more buses) .

The receiver 1310 may provide a means for receiving information such as packets, user data, control information, or any combination thereof associated with various information channels (e.g., control channels, data channels, information channels related to beam-mapping pattern consistency for machine learning model training and inference) . Information may be passed on to other components of the device 1305. The receiver 1310 may utilize a single antenna or a set of multiple antennas.

The transmitter 1315 may provide a means for transmitting signals generated by other components of the device 1305. For example, the transmitter 1315 may transmit information such as packets, user data, control information, or any combination thereof associated with various information channels (e.g., control channels, data channels, information channels related to beam-mapping pattern consistency for machine learning model training and inference) . In some examples, the transmitter 1315 may be co-located with a receiver 1310 in a transceiver module. The transmitter 1315 may utilize a single antenna or a set of multiple antennas.

The device 1305, or various components thereof, may be an example of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1320 may include an activation component 1325 a report component 1330, or any combination thereof. In some examples, the communications manager 1320, or various components thereof, may be configured to perform various operations (e.g., receiving, obtaining, monitoring, outputting, transmitting) using or otherwise in cooperation with the receiver 1310, the transmitter 1315, or both. For example, the communications manager 1320 may receive information from the receiver 1310, send information to the transmitter 1315, or be integrated in combination with the receiver 1310, the transmitter 1315, or both to obtain information, output information, or perform various other operations as described herein.

The communications manager 1320 may support wireless communications in accordance with examples as disclosed herein. The activation component 1325 is capable of, configured to, or operable to support a means for receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam- mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The report component 1330 is capable of, configured to, or operable to support a means for transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

Figure 14 shows a communications manager 1420 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The communications manager 1420, or various components thereof, may be an example of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1420 may include an activation component 1425, a report component 1430, a pattern component 1435, an identifier component 1440, a consistency component 1445, or any combination thereof. Each of these components, or components or subcomponents thereof (e.g., one or more processors, one or more memories) , may communicate, directly or indirectly, with one another (e.g., via one or more buses) .

The communications manager 1420 may support wireless communications in accordance with examples as disclosed herein. The activation component 1425 is capable of, configured to, or operable to support a means for receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The report component 1430 is capable of, configured to, or operable to support a means for transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam- mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

In some examples, the first beam-mapping pattern includes a beam-mapping pattern associated with inference by the machine learning model, and the second beam-mapping pattern includes a beam-mapping pattern associated with training the machine learning model.

In some examples, the pattern component 1435 is capable of, configured to, or operable to support a means for transmitting an indication of a set of beam-mapping patterns supported by the UE, where the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns. In some examples, a UE capability report, a UCI message, or a MAC-CE message.

In some examples, the pattern component 1435 is capable of, configured to, or operable to support a means for receiving a set of candidate beam-mapping patterns supported by a network entity, where the indication of the set of beam-mapping patterns is transmitted in accordance with receiving the set of candidate beam-mapping patterns and includes a subset of the set of candidate beam-mapping patterns.

In some examples, the pattern component 1435 is capable of, configured to, or operable to support a means for receiving an indication that the second beam-mapping pattern is associated with training the machine learning model, where the indication of the first beam-mapping pattern is received after receiving the indication that the second beam-mapping pattern is associated with training the machine learning model.

In some examples, the identifier component 1440 is capable of, configured to, or operable to support a means for receiving a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, where the indication to activate the machine learning model includes the machine learning model identifier.

In some examples, the indication of the first beam-mapping pattern includes a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, and the identifier component 1440 is capable of, configured to, or operable to support a means for receiving a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by a network entity. In some examples, the indication of the first beam-mapping pattern includes a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, and the identifier component 1440 is capable of, configured to, or operable to support a means for transmitting a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, where the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.

In some examples, the consistency component 1445 is capable of, configured to, or operable to support a means for receiving an indication of the consistency condition, where the indication to activate the machine learning model is received after receiving the indication of the consistency condition.

In some examples, the consistency component 1445 is capable of, configured to, or operable to support a means for transmitting an indication of the consistency condition, where the indication to activate the machine learning model is received after transmitting the indication of the consistency condition.

In some examples, the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first beam-mapping pattern is the same as the second beam-mapping pattern.

In some examples, the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams. In some examples, each beam in the second set of beams has a first respective probability of being mapped to the first set of beams and has a second respective probability of being mapped to the third set of beams. In some examples, the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that, for each beam in the second set of beams, a difference between the first respective probability and the second respective probability is below a threshold for that beam.

In some examples, the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams. In some examples, the first beam- mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, where the first interval factor specifies an interval between identifiers associated with the first set of beams, and where the second interval factor specifies an interval between identifiers associated with the third set of beams.

In some examples, the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams. In some examples, the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first set of beams has a same quantity of beams as the third set of beams.

In some examples, the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams. In some examples, the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a difference between a first statistical value associated with a quantity of the first set of beams and a second statistical value associated with a quantity of the third set of beams is below a threshold. In some examples, an RRC message, a DCI message, or a MAC-CE message.

Figure 15 shows a diagram of a system including a device 1505 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1505 may be an example of or include the components of a device 1205, a device 1305, or a UE 115 as described herein. The device 1505 may communicate (e.g., wirelessly) with one or more network entities 105, one or more UEs 115, or any combination thereof. The device 1505 may include components for bi-directional voice and data communications including components for transmitting and receiving communications, such as a communications manager 1520, an input/output (I/O) controller 1510, a transceiver 1515, an antenna 1525, at least one memory 1530, code 1535, and at least one processor 1540. These components may be in electronic communication or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more buses (e.g., a bus 1545) .

The I/O controller 1510 may manage input and output signals for the device 1505. The I/O controller 1510 may also manage peripherals not integrated into the device 1505. In some examples, the I/O controller 1510 may represent a physical connection or port to an external peripheral. In some examples, the I/O controller 1510 may utilize an operating system such as or another known operating system. Additionally or alternatively, the I/O controller 1510 may represent or interact with a modem, a keyboard, a mouse, a touchscreen, or a similar device. In some examples, the I/O controller 1510 may be implemented as part of one or more processors, such as the at least one processor 1540. In some examples, a user may interact with the device 1505 via the I/O controller 1510 or via hardware components controlled by the I/O controller 1510.

In some examples, the device 1505 may include a single antenna 1525. However, in some other cases, the device 1505 may have more than one antenna 1525, which may be capable of concurrently transmitting or receiving multiple wireless transmissions. The transceiver 1515 may communicate bi-directionally, via the one or more antennas 1525, wired, or wireless links as described herein. For example, the transceiver 1515 may represent a wireless transceiver and may communicate bi-directionally with another wireless transceiver. The transceiver 1515 may also include a modem to modulate the packets, to provide the modulated packets to one or more antennas 1525 for transmission, and to demodulate packets received from the one or more antennas 1525. The transceiver 1515, or the transceiver 1515 and one or more antennas 1525, may be an example of a transmitter 1215, a transmitter 1315, a receiver 1210, a receiver 1310, or any combination thereof or component thereof, as described herein.

The at least one memory 1530 may include random access memory (RAM) and read-only memory (ROM) . The at least one memory 1530 may store computer-readable, computer-executable code 1535 including instructions that, when executed by the at least one processor 1540, cause the device 1505 to perform various functions described herein. The code 1535 may be stored in a non-transitory computer-readable medium such as system memory or another type of memory. In some cases, the code 1535 may not be directly executable by the at least one processor 1540 but may cause a computer (e.g., when compiled and executed) to perform functions described herein. In some cases, the at least one memory 1530 may contain, among other things, a basic I/O system (BIOS) which may control basic hardware or software operation such as the interaction with peripheral components or devices.

The at least one processor 1540 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, a CPU, a microcontroller, an ASIC, an FPGA, a programmable logic device, a discrete gate or transistor logic component, a discrete hardware component, or any combination thereof) . In some cases, the at least one processor 1540 may be configured to operate a memory array using a memory controller. In some other cases, a memory controller may be integrated into the at least one processor 1540. The at least one processor 1540 may be configured to execute computer-readable instructions stored in a memory (e.g., the at least one memory 1530) to cause the device 1505 to perform various functions (e.g., functions or tasks supporting beam-mapping pattern consistency for machine learning model training and inference) . For example, the device 1505 or a component of the device 1505 may include at least one processor 1540 and at least one memory 1530 coupled with or to the at least one processor 1540, the at least one processor 1540 and at least one memory 1530 configured to perform various functions described herein. In some examples, the at least one processor 1540 may include multiple processors and the at least one memory 1530 may include multiple memories. One or more of the multiple processors may be coupled with one or more of the multiple memories, which may, individually or collectively, be configured to perform various functions herein. In some examples, the at least one processor 1540 may be a component of a processing system, which may refer to a system (such as a series) of machines, circuitry (including, for example, one or both of processor circuitry (which may include the at least one processor 1540) and memory circuitry (which may include the at least one memory 1530) ) , or components, that receives or obtains inputs and processes the inputs to produce, generate, or obtain a set of outputs. The processing system may be configured to perform one or more of the functions described herein. As such, the at least one processor 1540 or a processing system including the at least one processor 1540 may be configured to, configurable to, or operable to cause the device 1505 to perform one or more of the functions described herein. Further, as described herein, being “configured to, ” being “configurable to, ” and being “operable to” may be used interchangeably and may be associated with a capability, when executing code stored in the at least one memory 1530 or otherwise, to perform one or more of the functions described herein.

The communications manager 1520 may support wireless communications in accordance with examples as disclosed herein. For example, the communications manager 1520 is capable of, configured to, or operable to support a means for receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The communications manager 1520 is capable of, configured to, or operable to support a means for transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

By including or configuring the communications manager 1520 in accordance with examples as described herein, the device 1505 may support techniques for improved beam management (e.g., due to improved ML-based beam prediction) , which in turn may enable more efficient utilization of communication resources.

In some examples, the communications manager 1520 may be configured to perform various operations (e.g., receiving, monitoring, transmitting) using or otherwise in cooperation with the transceiver 1515, the one or more antennas 1525, or any combination thereof. Although the communications manager 1520 is illustrated as a separate component, in some examples, one or more functions described with reference to the communications manager 1520 may be supported by or performed by the at least one processor 1540, the at least one memory 1530, the code 1535, or any combination thereof. For example, the code 1535 may include instructions executable by the at least one processor 1540 to cause the device 1505 to perform various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein, or the at least one processor 1540 and the at least one memory 1530 may be otherwise configured to, individually or collectively, perform or support such operations.

Figure 16 shows a device 1605 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1605 may be an example of aspects of a network entity 105 as described herein. The device 1605 may include a receiver 1610, a transmitter 1615, and a communications manager 1620. The device 1605, or one or more components of the device 1605 (e.g., the receiver 1610, the transmitter 1615, and the communications manager 1620) , may include at least one processor, which may be coupled with at least one memory, to, individually or collectively, support or enable the described techniques. Each of these components may be in communication with one another (e.g., via one or more buses) .

The receiver 1610 may provide a means for obtaining (e.g., receiving, determining, identifying) information such as user data, control information, or any combination thereof (e.g., I/Q samples, symbols, packets, protocol data units, service data units) associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) . Information may be passed on to other components of the device 1605. In some examples, the receiver 1610 may support obtaining information by receiving signals via one or more antennas. Additionally, or alternatively, the receiver 1610 may support obtaining information by receiving signals via one or more wired (e.g., electrical, fiber optic) interfaces, wireless interfaces, or any combination thereof.

The transmitter 1615 may provide a means for outputting (e.g., transmitting, providing, conveying, sending) information generated by other components of the device 1605. For example, the transmitter 1615 may output information such as user data, control information, or any combination thereof (e.g., I/Q samples, symbols, packets, protocol data units, service data units) associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) . In some examples, the transmitter 1615 may support outputting information by transmitting signals via one or more antennas. Additionally, or alternatively, the transmitter 1615 may support outputting information by transmitting signals via one or more wired (e.g., electrical, fiber optic) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 1615 and the receiver 1610 may be co-located in a transceiver, which may include or be coupled with a modem.

The communications manager 1620, the receiver 1610, the transmitter 1615, or various combinations thereof or various components thereof may be examples of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1620, the receiver 1610, the transmitter 1615, or various combinations or components thereof may be capable of performing one or more of the functions described herein.

In some examples, the communications manager 1620 may be configured to perform various operations (e.g., receiving, obtaining, monitoring, outputting, transmitting) using or otherwise in cooperation with the receiver 1610, the transmitter 1615, or both. For example, the communications manager 1620 may receive information from the receiver 1610, send information to the transmitter 1615, or be integrated in combination with the receiver 1610, the transmitter 1615, or both to obtain information, output information, or perform various other operations as described herein.

The communications manager 1620 may support wireless communications in accordance with examples as disclosed herein. For example, the communications manager 1620 is capable of, configured to, or operable to support a means for outputting, to a UE, an indication to activating a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The communications manager 1620 is capable of, configured to, or operable to support a means for obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

By including or configuring the communications manager 1620 in accordance with examples as described herein, the device 1605 (e.g., at least one processor controlling or otherwise coupled with the receiver 1610, the transmitter 1615, the communications manager 1620, or a combination thereof) may support techniques for improved beam management (e.g., due to improved ML-based beam prediction) , which in turn may enable more efficient utilization of communication resources.

Figure 17 shows a device 1705 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1705 may be an example of aspects of a device 1605 or a network entity 105 as described herein. The device 1705 may include a receiver 1710, a transmitter 1715, and a communications manager 1720. The device 1705, or one of more components of the device 1705 (e.g., the receiver 1710, the transmitter 1715, and the communications manager 1720) , may include at least one processor, which may be coupled with at least one memory, to support the described techniques. Each of these components may be in communication with one another (e.g., via one or more buses) .

The receiver 1710 may provide a means for obtaining (e.g., receiving, determining, identifying) information such as user data, control information, or any combination thereof (e.g., I/Q samples, symbols, packets, protocol data units, service data units) associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) . Information may be passed on to other components of the device 1705. In some examples, the receiver 1710 may support obtaining information by receiving signals via one or more antennas. Additionally, or alternatively, the receiver 1710 may support obtaining information by receiving signals via one or more wired (e.g., electrical, fiber optic) interfaces, wireless interfaces, or any combination thereof.

The transmitter 1715 may provide a means for outputting (e.g., transmitting, providing, conveying, sending) information generated by other components of the device 1705. For example, the transmitter 1715 may output information such as user data, control information, or any combination thereof (e.g., I/Q samples, symbols, packets, protocol data units, service data units) associated with various channels (e.g., control channels, data channels, information channels, channels associated with a protocol stack) . In some examples, the transmitter 1715 may support outputting information by transmitting signals via one or more antennas. Additionally, or alternatively, the transmitter 1715 may support outputting information by transmitting signals via one or more wired (e.g., electrical, fiber optic) interfaces, wireless interfaces, or any combination thereof. In some examples, the transmitter 1715 and the receiver 1710 may be co-located in a transceiver, which may include or be coupled with a modem.

The device 1705, or various components thereof, may be an example of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1720 may include an activation component 1725 a report component 1730, or any combination thereof. In some examples, the communications manager 1720, or various components thereof, may be configured to perform various operations (e.g., receiving, obtaining, monitoring, outputting, transmitting) using or otherwise in cooperation with the receiver 1710, the transmitter 1715, or both. For example, the communications manager 1720 may receive information from the receiver 1710, send information to the transmitter 1715, or be integrated in combination with the receiver 1710, the transmitter 1715, or both to obtain information, output information, or perform various other operations as described herein.

The communications manager 1720 may support wireless communications in accordance with examples as disclosed herein. The activation component 1725 is capable of, configured to, or operable to support a means for outputting, to a UE, an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The report component 1730 is capable of, configured to, or operable to support a means for obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

Figure 18 shows a communications manager 1820 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The communications manager 1820, or various components thereof, may be an example of means for performing various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein. For example, the communications manager 1820 may include an activation component 1825, a report component 1830, a pattern component 1835, an identifier component 1840, a consistency component 1845, or any combination thereof. Each of these components, or components or subcomponents thereof (e.g., one or more processors, one or more memories) , may communicate, directly or indirectly, with one another (e.g., via one or more buses) which may include communications within a protocol layer of a protocol stack, communications associated with a logical channel of a protocol stack (e.g., between protocol layers of a protocol stack, within a device, component, or virtualized component associated with a network entity 105, between devices, components, or virtualized components associated with a network entity 105) , or any combination thereof.

The communications manager 1820 may support wireless communications in accordance with examples as disclosed herein. The activation component 1825 is capable of, configured to, or operable to support a means for outputting, to a UE, an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The report component 1830 is capable of, configured to, or operable to support a means for obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

In some examples, the first beam-mapping pattern includes a beam-mapping pattern associated with inference by the machine learning model, and the second beam- mapping pattern includes a beam-mapping pattern associated with training the machine learning model.

In some examples, the pattern component 1835 is capable of, configured to, or operable to support a means for obtaining an indication of a set of beam-mapping patterns supported by the UE, where the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns.

In some examples, a UE capability report, a UCI message, or a MAC-CE message.

In some examples, the pattern component 1835 is capable of, configured to, or operable to support a means for outputting a set of candidate beam-mapping patterns supported by the network entity, where the indication of the set of beam-mapping patterns is obtained in accordance with outputting the set of candidate beam-mapping patterns and includes a subset of the set of candidate beam-mapping patterns.

In some examples, the pattern component 1835 is capable of, configured to, or operable to support a means for outputting an indication that the second beam-mapping pattern is associated with training the machine learning model, where the indication of the first beam-mapping pattern is outputted after outputting the indication that the second beam-mapping pattern is associated with training the machine learning model.

In some examples, the identifier component 1840 is capable of, configured to, or operable to support a means for outputting a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, where the indication to activate the machine learning model includes the machine learning model identifier.

In some examples, the indication of the first beam-mapping pattern includes a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, and the identifier component 1840 is capable of, configured to, or operable to support a means for outputting a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by the network entity. In some examples, the indication of the first beam-mapping pattern includes a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, and the identifier component 1840 is capable of, configured to, or operable to support a means for obtaining a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, where the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.

In some examples, the consistency component 1845 is capable of, configured to, or operable to support a means for outputting an indication of the consistency condition, where the indication to activate the machine learning model is outputted after outputting the indication of the consistency condition.

In some examples, the consistency component 1845 is capable of, configured to, or operable to support a means for obtaining an indication of the consistency condition, where the indication to activate the machine learning model is outputted after obtaining the indication of the consistency condition.

In some examples, the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams. In some examples, the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, where the first interval factor specifies an interval between identifiers associated with the first set of beams, and where the second interval factor specifies an interval between identifiers associated with the third set of beams.

Figure 19 shows a diagram of a system including a device 1905 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with one or more aspects of the present disclosure. The device 1905 may be an example of or include the components of a device 1605, a device 1705, or a network entity 105 as described herein. The device 1905 may communicate with one or more network entities 105, one or more UEs 115, or any combination thereof, which may include communications over one or more wired interfaces, over one or more wireless interfaces, or any combination thereof. The device 1905 may include components that support outputting and obtaining communications, such as a communications manager 1920, a transceiver 1910, an antenna 1915, at least one memory 1925, code 1930, and at least one processor 1935. These components may be in electronic communication or otherwise coupled (e.g., operatively, communicatively, functionally, electronically, electrically) via one or more buses (e.g., a bus 1940) .

The transceiver 1910 may support bi-directional communications via wired links, wireless links, or both as described herein. In some examples, the transceiver 1910 may include a wired transceiver and may communicate bi-directionally with another wired transceiver. Additionally, or alternatively, in some examples, the transceiver 1910 may include a wireless transceiver and may communicate bi-directionally with another wireless transceiver. In some examples, the device 1905 may include one or more antennas 1915, which may be capable of transmitting or receiving wireless transmissions (e.g., concurrently) . The transceiver 1910 may also include a modem to modulate signals, to provide the modulated signals for transmission (e.g., by one or more antennas 1915, by a wired transmitter) , to receive modulated signals (e.g., from one or more antennas 1915, from a wired receiver) , and to demodulate signals. In some implementations, the transceiver 1910 may include one or more interfaces, such as one or more interfaces coupled with the one or more antennas 1915 that are configured to support various receiving or obtaining operations, or one or more interfaces coupled with the one or more antennas 1915 that are configured to support various transmitting or outputting operations, or a combination thereof. In some implementations, the transceiver 1910 may include or be configured for coupling with one or more processors or one or more memory components that are operable to perform or support operations based on received or obtained information or signals, or to generate information or other signals for transmission or other outputting, or any combination thereof. In some implementations, the transceiver 1910, or the transceiver 1910 and the one or more antennas 1915, or the transceiver 1910 and the one or more antennas 1915 and one or more processors or one or more memory components (e.g., the at least one processor 1935, the at least one memory 1925, or both) , may be included in a chip or chip assembly that is installed in the device 1905. In some examples, the transceiver 1910 may be operable to support communications via one or more communications links (e.g., a communication link 125, a backhaul communication link 120, a midhaul communication link 162, a fronthaul communication link 168) .

The at least one memory 1925 may include RAM, ROM, or any combination thereof. The at least one memory 1925 may store computer-readable, computer-executable code 1930 including instructions that, when executed by one or more of the at least one processor 1935, cause the device 1905 to perform various functions described herein. The code 1930 may be stored in a non-transitory computer-readable medium such as system memory or another type of memory. In some cases, the code 1930 may not be directly executable by a processor of the at least one processor 1935 but may cause a computer (e.g., when compiled and executed) to perform functions described herein. In some examples, the at least one memory 1925 may contain, among other things, a BIOS which may control basic hardware or software operation such as the interaction with peripheral components or devices. In some examples, the at least one processor 1935 may include multiple processors and the at least one memory 1925 may include multiple memories. One or more of the multiple processors may be coupled with one or more of the multiple memories which may, individually or collectively, be configured to perform various functions herein (for example, as part of a processing system) .

The at least one processor 1935 may include an intelligent hardware device (e.g., a general-purpose processor, a DSP, an ASIC, a CPU, an FPGA, a microcontroller, a programmable logic device, discrete gate or transistor logic, a discrete hardware component, or any combination thereof) . In some examples, the at least one processor 1935 may be configured to operate a memory array using a memory controller. In some other cases, a memory controller may be integrated into one or more of the at least one processor 1935. The at least one processor 1935 may be configured to execute computer-readable instructions stored in a memory (e.g., one or more of the at least one memory 1925) to cause the device 1905 to perform various functions (e.g., functions or tasks supporting beam-mapping pattern consistency for machine learning model training and inference) . For example, the device 1905 or a component of the device 1905 may include at least one processor 1935 and at least one memory 1925 coupled with one or more of the at least one processor 1935, the at least one processor 1935 and the at least one memory 1925 configured to perform various functions described herein. The at least one processor 1935 may be an example of a cloud-computing platform (e.g., one or more physical nodes and supporting software such as operating systems, virtual machines, or container instances) that may host the functions (e.g., by executing code 1930) to perform the functions of the device 1905. The at least one processor 1935 may be any one or more suitable processors capable of executing scripts or instructions of one or more software programs stored in the device 1905 (such as within one or more of the at least one memory 1925) . In some examples, the at least one processor 1935 may include multiple processors and the at least one memory 1925 may include multiple memories. One or more of the multiple processors may be coupled with one or more of the multiple memories, which may, individually or collectively, be configured to perform various functions herein. In some examples, the at least one processor 1935 may be a component of a processing system, which may refer to a system (such as a series) of machines, circuitry (including, for example, one or both of processor circuitry (which may include the at least one processor 1935) and memory circuitry (which may include the at least one memory 1925) ) , or components, that receives or obtains inputs and processes the inputs to produce, generate, or obtain a set of outputs. The processing system may be configured to perform one or more of the functions described herein. As such, the at least one processor 1935 or a processing system including the at least one processor 1935 may be configured to, configurable to, or operable to cause the device 1905 to perform one or more of the functions described herein. Further, as described herein, being “configured to, ” being “configurable to, ” and being “operable to” may be used interchangeably and may be associated with a capability, when executing code stored in the at least one memory 1925 or otherwise, to perform one or more of the functions described herein.

In some examples, a bus 1940 may support communications of (e.g., within) a protocol layer of a protocol stack. In some examples, a bus 1940 may support communications associated with a logical channel of a protocol stack (e.g., between protocol layers of a protocol stack) , which may include communications performed within a component of the device 1905, or between different components of the device 1905 that may be co-located or located in different locations (e.g., where the device 1905 may refer to a system in which one or more of the communications manager 1920, the transceiver 1910, the at least one memory 1925, the code 1930, and the at least one processor 1935 may be located in one of the different components or divided between different components) .

In some examples, the communications manager 1920 may manage aspects of communications with a core network 130 (e.g., via one or more wired or wireless backhaul links) . For example, the communications manager 1920 may manage the transfer of data communications for client devices, such as one or more UEs 115. In some examples, the communications manager 1920 may manage communications with other network entities 105, and may include a controller or scheduler for controlling communications with UEs 115 in cooperation with other network entities 105. In some examples, the communications manager 1920 may support an X2 interface within an LTE/LTE-A wireless communications network technology to provide communication between network entities 105.

The communications manager 1920 may support wireless communications in accordance with examples as disclosed herein. For example, the communications manager 1920 is capable of, configured to, or operable to support a means for outputting, to a UE, an indication to activating a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The communications manager 1920 is capable of, configured to, or operable to support a means for obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

By including or configuring the communications manager 1920 in accordance with examples as described herein, the device 1905 may support techniques for improved beam management (e.g., due to improved ML-based beam prediction) , which in turn may enable more efficient utilization of communication resources.

In some examples, the communications manager 1920 may be configured to perform various operations (e.g., receiving, obtaining, monitoring, outputting, transmitting) using or otherwise in cooperation with the transceiver 1910, the one or more antennas 1915 (e.g., where applicable) , or any combination thereof. Although the communications manager 1920 is illustrated as a separate component, in some examples, one or more functions described with reference to the communications manager 1920 may be supported by or performed by the transceiver 1910, one or more of the at least one processor 1935, one or more of the at least one memory 1925, the code 1930, or any combination thereof (for example, by a processing system including at least a portion of the at least one processor 1935, the at least one memory 1925, the code 1930, or any combination thereof) . For example, the code 1930 may include instructions executable by one or more of the at least one processor 1935 to cause the device 1905 to perform various aspects of beam-mapping pattern consistency for machine learning model training and inference as described herein, or the at least one processor 1935 and the at least one memory 1925 may be otherwise configured to, individually or collectively, perform or support such operations.

Figure 20 shows a flowchart illustrating a method 2000 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with aspects of the present disclosure. The operations of the method 2000 may be implemented by a UE or its components as described herein. For example, the operations of the method 2000 may be performed by a UE 115 as described with reference to FIGs. 1–15. In some examples, a UE may execute a set of instructions to control the functional elements of the UE to perform the described functions. Additionally, or alternatively, the UE may perform aspects of the described functions using special-purpose hardware.

At 2005, the method may include receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The operations of block 2005 may be performed in accordance with examples as disclosed herein. In some examples, aspects of the operations of 2005 may be performed by an activation component 1425 as described with reference to FIG. 14.

At 2010, the method may include transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams. The operations of block 2010 may be performed in accordance with examples as disclosed herein. In some examples, aspects of the operations of 2010 may be performed by a report component 1430 as described with reference to FIG. 14.

Figure 21 shows a flowchart illustrating a method 2100 that supports beam-mapping pattern consistency for machine learning model training and inference in accordance with aspects of the present disclosure. The operations of the method 2100 may be implemented by a network entity or its components as described herein. For example, the operations of the method 2100 may be performed by a network entity as described with reference to FIGs. 1–11 and 16–19. In some examples, a network entity may execute a set of instructions to control the functional elements of the network entity to perform the described functions. Additionally, or alternatively, the network entity may perform aspects of the described functions using special-purpose hardware.

At 2105, the method may include outputting, to a UE, an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams. The operations of block 2105 may be performed in accordance with examples as disclosed herein. In some examples, aspects of the operations of 2105 may be performed by an activation component 1825 as described with reference to FIG. 18.

At 2110, the method may include obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern. The operations of block 2110 may be performed in accordance with examples as disclosed herein. In some examples, aspects of the operations of 2110 may be performed by a report component 1830 as described with reference to FIG. 18.

The following provides an overview of aspects of the present disclosure:

Aspect 1: A method for wireless communications by a UE, comprising: receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.

Aspect 2: The method of aspect 1, wherein the first beam-mapping pattern comprises a beam-mapping pattern associated with inference by the machine learning model, and the second beam-mapping pattern comprises a beam-mapping pattern associated with training the machine learning model.

Aspect 3: The method of any of aspects 1 through 2, further comprising: transmitting an indication of a set of beam-mapping patterns supported by the UE, wherein the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns.

Aspect 4: The method of aspect 3, wherein the indication of the first beam-mapping pattern is included in one or more of a UE capability report, an uplink control information (UCI) message, or a medium access control (MAC) control element (MAC-CE) message.

Aspect 5: The method of any of aspects 3 through 4, further comprising: receiving a set of candidate beam-mapping patterns supported by a network entity, wherein the indication of the set of beam-mapping patterns is transmitted in accordance with receiving the set of candidate beam-mapping patterns and comprises a subset of the set of candidate beam-mapping patterns.

Aspect 6: The method of any of aspects 1 through 5, further comprising: receiving an indication that the second beam-mapping pattern is associated with training the machine learning model, wherein the indication of the first beam-mapping pattern is received after receiving the indication that the second beam-mapping pattern is associated with training the machine learning model.

Aspect 7: The method of aspect 6, further comprising: receiving a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, wherein the indication to activate the machine learning model comprises the machine learning model identifier.

Aspect 8: The method of any of aspects 1 through 7, wherein the indication of the first beam-mapping pattern comprises a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, the method further comprising: receiving a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by a network entity; and transmitting a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, wherein the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.

Aspect 9: The method of any of aspects 1 through 8, further comprising: receiving an indication of the consistency condition, wherein the indication to activate the machine learning model is received after receiving the indication of the consistency condition.

Aspect 10: The method of any of aspects 1 through 9, further comprising: transmitting an indication of the consistency condition, wherein the indication to activate the machine learning model is received after transmitting the indication of the consistency condition.

Aspect 11: The method of any of aspects 1 through 10, wherein the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first beam-mapping pattern is the same as the second beam-mapping pattern.

Aspect 12: The method of any of aspects 1 through 10, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, each beam in the second set of beams has a first respective probability of being mapped to the first set of beams and has a second respective probability of being mapped to the third set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that, for each beam in the second set of beams, a difference between the first respective probability and the second respective probability is below a threshold for that beam.

Aspect 13: The method of any of aspects 1 through 10, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, wherein the first interval factor specifies an interval between identifiers associated with the first set of beams, and wherein the second interval factor specifies an interval between identifiers associated with the third set of beams.

Aspect 14: The method of any of aspects 1 through 10, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first set of beams has a same quantity of beams as the third set of beams.

Aspect 15: The method of any of aspects 1 through 10, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a difference between a first statistical value associated with a quantity of the first set of beams and a second statistical value associated with a quantity of the third set of beams is below a threshold.

Aspect 16: The method of any of aspects 1 through 15, wherein the indication of the first beam-mapping pattern is included in one or more of an RRC message, a DCI message, or a medium access control (MAC) control element (MAC-CE) message.

Aspect 17: A method for wireless communications by a network entity, comprising: outputting, to a UE, an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.

Aspect 18: The method of aspect 17, wherein the first beam-mapping pattern comprises a beam-mapping pattern associated with inference by the machine learning model, and the second beam-mapping pattern comprises a beam-mapping pattern associated with training the machine learning model.

Aspect 19: The method of any of aspects 17 through 18, further comprising: obtaining an indication of a set of beam-mapping patterns supported by the UE, wherein the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns.

Aspect 20: The method of aspect 19, wherein the indication of the first beam-mapping pattern is included in one or more of a UE capability report, an uplink control information (UCI) message, or a medium access control (MAC) control element (MAC-CE) message.

Aspect 21: The method of any of aspects 19 through 20, further comprising: outputting a set of candidate beam-mapping patterns supported by the network entity, wherein the indication of the set of beam-mapping patterns is obtained in accordance with outputting the set of candidate beam-mapping patterns and comprises a subset of the set of candidate beam-mapping patterns.

Aspect 22: The method of any of aspects 17 through 21, further comprising: outputting an indication that the second beam-mapping pattern is associated with training the machine learning model, wherein the indication of the first beam-mapping pattern is outputted after outputting the indication that the second beam-mapping pattern is associated with training the machine learning model.

Aspect 23: The method of aspect 22, further comprising: outputting a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, wherein the indication to activate the machine learning model comprises the machine learning model identifier.

Aspect 24: The method of any of aspects 17 through 23, wherein the indication of the first beam-mapping pattern comprises a machine learning model identifier that is associated with the machine learning model and that is associated with the first beam-mapping pattern, the method further comprising: outputting a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by the network entity; and obtaining a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, wherein the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.

Aspect 25: The method of any of aspects 17 through 24, further comprising: outputting an indication of the consistency condition, wherein the indication to activate the machine learning model is outputted after outputting the indication of the consistency condition.

Aspect 26: The method of any of aspects 17 through 25, further comprising: obtaining an indication of the consistency condition, wherein the indication to activate the machine learning model is outputted after obtaining the indication of the consistency condition.

Aspect 27: The method of any of aspects 17 through 26, wherein the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first beam-mapping pattern is the same as the second beam-mapping pattern.

Aspect 28: The method of any of aspects 17 through 26, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, each beam in the second set of beams has a first respective probability of being mapped to the first set of beams and has a second respective probability of being mapped to the third set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that, for each beam in the second set of beams, a difference between the first respective probability and the second respective probability is below a threshold for that beam.

Aspect 29: The method of any of aspects 17 through 26, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, wherein the first interval factor specifies an interval between identifiers associated with the first set of beams, and wherein the second interval factor specifies an interval between identifiers associated with the third set of beams.

Aspect 30: The method of any of aspects 17 through 26, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first set of beams has a same quantity of beams as the third set of beams.

Aspect 31: The method of any of aspects 17 through 26, wherein the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a difference between a first statistical value associated with a quantity of the first set of beams and a second statistical value associated with a quantity of the third set of beams is below a threshold.

Aspect 32: The method of any of aspects 17 through 31, wherein the indication of the first beam-mapping pattern is included in one or more of an RRC message, a DCI message, or a medium access control (MAC) control element (MAC-CE) message.

Aspect 33: A UE for wireless communications, comprising one or more memories storing processor-executable code, and one or more processors coupled with the one or more memories and individually or collectively operable to execute the code to cause the UE to perform a method of any of aspects 1 through 16.

Aspect 34: A UE for wireless communications, comprising at least one means for performing a method of any of aspects 1 through 16.

Aspect 35: A non-transitory computer-readable medium storing code for wireless communications, the code comprising instructions executable by one or more processors to perform a method of any of aspects 1 through 16.

Aspect 36: A network entity for wireless communications, comprising one or more memories storing processor-executable code, and one or more processors coupled with the one or more memories and individually or collectively operable to execute the code to cause the network entity to perform a method of any of aspects 17 through 32.

Aspect 37: A network entity for wireless communications, comprising at least one means for performing a method of any of aspects 17 through 32.

Aspect 38: A non-transitory computer-readable medium storing code for wireless communications, the code comprising instructions executable by one or more processors to perform a method of any of aspects 17 through 32.

It should be noted that the methods described herein describe possible implementations, and that the operations and the steps may be rearranged or otherwise modified and that other implementations are possible. Further, aspects from two or more of the methods may be combined.

Although aspects of an LTE, LTE-A, LTE-A Pro, or NR system may be described for purposes of example, and LTE, LTE-A, LTE-A Pro, or NR terminology may be used in much of the description, the techniques described herein are applicable beyond LTE, LTE-A, LTE-A Pro, or NR networks. For example, the described techniques may be applicable to various other wireless communications systems such as Ultra Mobile Broadband (UMB) , Institute of Electrical and Electronics Engineers (IEEE) 802.11 (Wi-Fi) , IEEE 802.16 (WiMAX) , IEEE 802.20, Flash-OFDM, as well as other systems and radio technologies not explicitly mentioned herein.

Information and signals described herein may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that may be referenced throughout the description may be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

The various illustrative blocks and components described in connection with the disclosure herein may be implemented or performed using a general-purpose processor, a DSP, an ASIC, a CPU, an FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor but, in the alternative, the processor may be any processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices (e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration) . Any functions or operations described herein as being capable of being performed by a processor may be performed by multiple processors that, individually or collectively, are capable of performing the described functions or operations.

The functions described herein may be implemented using hardware, software executed by a processor, firmware, or any combination thereof. If implemented using software executed by a processor, the functions may be stored as or transmitted using one or more instructions or code of a computer-readable medium. Other examples and implementations are within the scope of the disclosure and appended claims. For example, due to the nature of software, functions described herein may be implemented using software executed by a processor, hardware, firmware, hardwiring, or combinations of any of these. Features implementing functions may also be physically located at various positions, including being distributed such that portions of functions are implemented at different physical locations.

Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of a computer program from one location to another. A non-transitory storage medium may be any available medium that may be accessed by a general-purpose or special-purpose computer. By way of example, and not limitation, non-transitory computer-readable media may include RAM, ROM, electrically erasable programmable ROM (EEPROM) , flash memory, compact disk (CD) ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that may be used to carry or store desired program code means in the form of instructions or data structures and that may be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) , or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of computer-readable medium. Disk and disc, as used herein, include CD, laser disc, optical disc, digital versatile disc (DVD) , floppy disk and Blu-ray disc. Disks may reproduce data magnetically, and discs may reproduce data optically using lasers. Combinations of the above are also included within the scope of computer-readable media. Any functions or operations described herein as being capable of being performed by a memory may be performed by multiple memories that, individually or collectively, are capable of performing the described functions or operations.

As used herein, including in the claims, “or” as used in a list of items (e.g., a list of items prefaced by a phrase such as “at least one of” or “one or more of” ) indicates an inclusive list such that, for example, a list of at least one of A, B, or C means A or B or C or AB or AC or BC or ABC (i.e., A and B and C) . Also, as used herein, the phrase “based on” shall not be construed as a reference to a closed set of conditions. For example, an example step that is described as “based on condition A” may be based on both a condition A and a condition B without departing from the scope of the present disclosure. In other words, as used herein, the phrase “based on” shall be construed in the same manner as the phrase “based at least in part on. ”

As used herein, including in the claims, the article “a” before a noun is open-ended and understood to refer to “at least one” of those nouns or “one or more” of those nouns. So, the terms “a, ” “at least one, ” “one or more, ” “at least one of one or more” may be interchangeable. For example, if a claim recites “a component” that performs one or more functions, each of the individual functions may be performed by a single component or by any combination of multiple components. So, the term “a component” having characteristics or performing functions may refer to “at least one of one or more components” having a particular characteristic or performing a particular function. Subsequent reference to a component introduced with the article “a” using the terms “the” or “said” may refer to any or all of the one or more components. For example, a component introduced with the article “a” may be understood to mean “one or more components, ” and referring to “the component” subsequently in the claims may be understood to be equivalent to referring to “at least one of the one or more components. ” Similarly, subsequent reference to a component introduced as “one or more components” using the terms “the” or “said” may refer to any or all of the one or more components. For example, referring to “the one or more components” subsequently in the claims may be understood to be equivalent to referring to “at least one of the one or more components. ”

The term “determine” or “determining” encompasses a variety of actions and, therefore, “determining” can include calculating, computing, processing, deriving, investigating, looking up (such as via looking up in a table, a database or another data structure) , ascertaining and the like. Also, “determining” can include receiving (e.g., receiving information) , accessing (e.g., accessing data stored in memory) and the like. Also, “determining” can include resolving, obtaining, selecting, choosing, establishing, and other such similar actions.

In the appended figures, similar components or features may have the same reference label. Further, various components of the same type may be distinguished by following the reference label by a dash and a second label that distinguishes among the similar components. If just the first reference label is used in the specification, the description is applicable to any one of the similar components having the same first reference label irrespective of the second reference label, or other subsequent reference label.

The description set forth herein, in connection with the appended drawings, describes example configurations and does not represent all the examples that may be implemented or that are within the scope of the claims. The term “example” used herein means “serving as an example, instance, or illustration, ” and not “preferred” or “advantageous over other examples. ” The detailed description includes specific details for the purpose of providing an understanding of the described techniques. These techniques, however, may be practiced without these specific details. In some instances, known structures and devices are shown in block diagram form in order to avoid obscuring the concepts of the described examples.

The description herein is provided to enable a person having ordinary skill in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to a person having ordinary skill in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. So, the disclosure is not limited to the examples and designs described herein but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.

Claims

A user equipment (UE) , comprising:

a processing system that includes processor circuitry and memory circuitry that stores code, the processing system configured to cause the UE to:

receive an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and

transmit a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.
The UE of claim 1, wherein the first beam-mapping pattern comprises a beam-mapping pattern associated with inference by the machine learning model, and wherein the second beam-mapping pattern comprises a beam-mapping pattern associated with training the machine learning model.
The UE of claim 1, wherein the processing system is further configured to cause the UE to:

transmit an indication of a set of beam-mapping patterns supported by the UE, wherein the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns.
The UE of claim 3, wherein the indication of the first beam-mapping pattern is included in one or more of a UE capability report, an uplink control information (UCI) message, or a medium access control (MAC) control element (MAC-CE) message.
The UE of claim 3, wherein the processing system is further configured to cause the UE to:

receive a set of candidate beam-mapping patterns supported by a network entity, wherein the indication of the set of beam-mapping patterns is transmitted in accordance with receiving the set of candidate beam-mapping patterns and comprises a subset of the set of candidate beam-mapping patterns.
The UE of claim 1, wherein the processing system is further configured to cause the UE to:

receive an indication that the second beam-mapping pattern is associated with training the machine learning model, wherein the indication of the first beam-mapping pattern is received after receiving the indication that the second beam-mapping pattern is associated with training the machine learning model.
The UE of claim 6, wherein the processing system is further configured to cause the UE to:

receive a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, wherein the indication to activate the machine learning model comprises the machine learning model identifier.
The UE of claim 1, wherein the indication of the first beam-mapping pattern comprises a machine learning model identifier that is associated with the machine learning model, and wherein the processing system is further configured to cause the UE to:

receive a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by a network entity; and

transmit a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, wherein the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.
The UE of claim 1, wherein the processing system is further configured to cause the UE to:

receive an indication of the consistency condition, wherein the indication to activate the machine learning model is received after receiving the indication of the consistency condition.
The UE of claim 1, wherein the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first beam-mapping pattern is the same as the second beam-mapping pattern.
The UE of claim 1, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams,

each beam in the second set of beams has a first respective probability of being mapped to the first set of beams and has a second respective probability of being mapped to the third set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that, for each beam in the second set of beams, a difference between the first respective probability and the second respective probability is below a threshold for that beam.
The UE of claim 1, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, wherein the first interval factor specifies an interval between identifiers associated with the first set of beams, and wherein the second interval factor specifies an interval between identifiers associated with the third set of beams.
The UE of claim 1, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first set of beams has a same quantity of beams as the third set of beams.
The UE of claim 1, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a difference between a first statistical value associated with a quantity of the first set of beams and a second statistical value associated with a quantity of the third set of beams is below a threshold.
The UE of claim 1, wherein the indication of the first beam-mapping pattern is included in one or more of a radio resource control (RRC) message, a downlink control information (DCI) message, or a medium access control (MAC) control element (MAC-CE) message.
A network entity, comprising:

a processing system that includes processor circuitry and memory circuitry that stores code, the processing system configured to cause the network entity to:

output, to a user equipment (UE) , an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and

obtain a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.
The network entity of claim 16, wherein the first beam-mapping pattern comprises a beam-mapping pattern associated with inference by the machine learning model, and wherein the second beam-mapping pattern comprises a beam-mapping pattern associated with training the machine learning model.
The network entity of claim 16, wherein the processing system is further configured to cause the network entity to:

obtain an indication of a set of beam-mapping patterns supported by the UE, wherein the indication to activate the machine learning model is in accordance with the indication of the set of beam-mapping patterns.
The network entity of claim 18, wherein the processing system is further configured to cause the network entity to:

output a set of candidate beam-mapping patterns supported by the network entity, wherein the indication of the set of beam-mapping patterns is obtained in accordance with outputting the set of candidate beam-mapping patterns and comprises a subset of the set of candidate beam-mapping patterns.
The network entity of claim 16, wherein the processing system is further configured to cause the network entity to:

output an indication that the second beam-mapping pattern is associated with training the machine learning model, wherein the indication of the first beam-mapping pattern is outputted after outputting the indication that the second beam-mapping pattern is associated with training the machine learning model.
The network entity of claim 20, wherein the processing system is further configured to cause the network entity to:

output a machine learning model identifier that is associated with training the machine learning model and that is associated with the first beam-mapping pattern, wherein the indication to activate the machine learning model comprises the machine learning model identifier.
The network entity of claim 16, wherein the indication of the first beam-mapping pattern comprises a machine learning model identifier that is associated with the machine learning model, and wherein the processing system is further configured to cause the network entity to:

output a first set of machine learning model identifiers associated with one or more candidate machine learning models supported by the network entity; and

obtain a second set of machine learning model identifiers associated with one or more machine learning models supported by the UE, wherein the machine learning model identifier is included in the first set of machine learning model identifiers and the second set of machine learning model identifiers.
The network entity of claim 16, wherein the processing system is further configured to cause the network entity to:

output an indication of the consistency condition, wherein the indication to activate the machine learning model is outputted after outputting the indication of the consistency condition.
The network entity of claim 16, wherein the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first beam-mapping pattern is the same as the second beam-mapping pattern.
The network entity of claim 16, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams,

each beam in the second set of beams has a first respective probability of being mapped to the first set of beams and has a second respective probability of being mapped to the third set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that, for each beam in the second set of beams, a difference between the first respective probability and the second respective probability is below a threshold for that beam.
The network entity of claim 16, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a first interval factor associated with the first beam-mapping pattern is equal to a second interval factor associated with the second beam-mapping pattern, wherein the first interval factor specifies an interval between identifiers associated with the first set of beams, and wherein the second interval factor specifies an interval between identifiers associated with the third set of beams.
The network entity of claim 16, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that the first set of beams has a same quantity of beams as the third set of beams.
The network entity of claim 16, wherein:

the second beam-mapping pattern maps a third set of beams for measurement to the second set of beams, and

the first beam-mapping pattern satisfies the consistency condition relative to the second beam-mapping pattern in that a difference between a first statistical value associated with a quantity of the first set of beams and a second statistical value associated with a quantity of the third set of beams is below a threshold.
A method for wireless communications by a user equipment (UE) , comprising:

receiving an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and

transmitting a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern and in accordance with one or more actual measurements of the channel characteristic for one or more of the first set of beams.
A method for wireless communications by a network entity, comprising:

outputting, to a user equipment (UE) , an indication to activate a machine learning model for beam prediction and an indication of a first beam-mapping pattern that satisfies a consistency condition relative to a second beam-mapping pattern supported by the UE, the first beam-mapping pattern mapping a first set of beams for measurement by the UE to a second set of beams for measurement prediction by the machine learning model, the first set of beams being a subset of the second set of beams; and

obtaining a report indicating one or more predicted measurements of a channel characteristic for one or more of the second set of beams, the one or more predicted measurements being in accordance with the first beam-mapping pattern.