AP CSI-RS based set-b beam sub-sample pattern consistency
By ensuring spatial and temporal consistency in sub-sample patterns for AI/ML models using aperiodic CSI-RS resources, the prediction accuracy and robustness of beam management in wireless communication systems are enhanced, facilitating efficient resource allocation and flexible deployment.
Patent Information
- Application Number
- PCT/CN2024/078254
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-02-23
- Publication Date
- 2025-08-28
AI Technical Summary
Maintaining consistency in sub-sample patterns for beam predictions using AI/ML models is challenging when aperiodic CSI-RS resources are used in wireless communication systems, affecting prediction accuracy and robustness.
Implementing methods and apparatus that ensure spatial and temporal domain consistency for AI/ML models by measuring aperiodic reference signals on a first set of resources and predicting channel characteristics for a second set of resources using specific mapping patterns, ensuring consistency conditions are met during training and inference.
Improves prediction accuracy and robustness of AI/ML models, particularly with aperiodic CSI-RS resources, leading to more efficient resource allocation and flexible deployment across various network environments.
Smart Images

Figure CN2024078254_28082025_PF_FP_ABST
Abstract
Description
AP CSI-RS BASED SET-B BEAM SUB-SAMPLE PATTERN CONSISTENCYTECHNICAL FIELD
[0001] The present disclosure relates generally to communication systems, and more particularly, to wireless communication including beam management.
[0002] INTRODUCTION
[0003] Wireless communication systems are widely deployed to provide various telecommunication services such as telephony, video, data, messaging, and broadcasts. Typical wireless communication systems may employ multiple-access technologies capable of supporting communication with multiple users by sharing available system resources. Examples of such multiple-access technologies include code division multiple access (CDMA) systems, time division multiple access (TDMA) systems, frequency division multiple access (FDMA) systems, orthogonal frequency division multiple access (OFDMA) systems, single-carrier frequency division multiple access (SC-FDMA) systems, and time division synchronous code division multiple access (TD-SCDMA) systems.
[0004] These multiple access technologies have been adopted in various telecommunication standards to provide a common protocol that enables different wireless devices to communicate on a municipal, national, regional, and even global level. An example telecommunication standard is 5G New Radio (NR) . 5G NR is part of a continuous mobile broadband evolution promulgated by Third Generation Partnership Project (3GPP) to meet new requirements associated with latency, reliability, security, scalability (e.g., with Internet of Things (IoT) ) , and other requirements. 5G NR includes services associated with enhanced mobile broadband (eMBB) , massive machine type communications (mMTC) , and ultra-reliable low latency communications (URLLC) . Some aspects of 5G NR may be based on the 4G Long Term Evolution (LTE) standard. There exists a need for further improvements in 5G NR technology. These improvements may also be applicable to other multi-access technologies and the telecommunication standards that employ these technologies.
[0005] BRIEF SUMMARY
[0006] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects. This summary neither identifies key or critical elements of all aspects nor delineates the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.
[0007] In an aspect of the disclosure, a method, a computer-readable medium, and an apparatus are provided for wireless communication at a user equipment (UE) . The apparatus may include at least one memory and at least one processor coupled to the at least one memory. Based at least in part on information stored in the at least one memory, the at least one processor, individually or in any combination, may be configured to measure an aperiodic reference signal received on a first set of resources; predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and report, to a network entity, the one or more channel characteristics predicted for the second set of resources.
[0008] In an aspect of the disclosure, a method, a computer-readable medium, and an apparatus are provided for wireless communication at a network entity. The apparatus may include at least one memory and at least one processor coupled to the at least one memory. Based at least in part on information stored in the at least one memory, the at least one processor, individually or in any combination, may be configured to provide an aperiodic reference signal on a first set of resources; indicate for a UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and receive, from the UE, the one or more channel characteristics predicted for the second set of resources.
[0009] To the accomplishment of the foregoing and related ends, the one or more aspects may include the features hereinafter fully described and particularly pointed out in the claims. The following description and the drawings set forth in detail certain illustrative features of the one or more aspects. These features are indicative, however, of but a few of the various ways in which the principles of various aspects may be employed.BRIEF DESCRIPTION OF THE DRAWINGS
[0010] FIG. 1 is a diagram illustrating an example of a wireless communication system and an access network.
[0011] FIG. 2A is a diagram illustrating an example of a first frame, in accordance with various aspects of the present disclosure.
[0012] FIG. 2B is a diagram illustrating an example of downlink (DL) channels within a subframe, in accordance with various aspects of the present disclosure.
[0013] FIG. 2C is a diagram illustrating an example of a second frame, in accordance with various aspects of the present disclosure.
[0014] FIG. 2D is a diagram illustrating an example of uplink (UL) channels within a subframe, in accordance with various aspects of the present disclosure.
[0015] FIG. 3 is a diagram illustrating an example of a base station and user equipment (UE) in an access network.
[0016] FIG. 4 is an illustrative block diagram of an example machine learning (ML) model represented by an artificial neural network (ANN) , in accordance with various aspects of the present disclosure.
[0017] FIG. 5 is a diagram illustrating an example of an artificial intelligence (AI) and machine learning (ML) (AI / ML) algorithm of a method of wireless communication.
[0018] FIG. 6 is a diagram illustrating an example of an AI / ML model training based on a sub-sample pattern in accordance with various aspects of the present disclosure.
[0019] FIG. 7 is a diagram illustrating an example of an AI / ML model inference based on a sub-sample pattern in accordance with various aspects of the present disclosure.
[0020] FIG. 8 is a diagram illustrating an example of an AI / ML model training in accordance with various aspects of the present disclosure.
[0021] FIG. 9 is a diagram illustrating an example of an AI / ML model inference in accordance with various aspects of the present disclosure.
[0022] FIG. 10A is a diagram illustrating an example of AI / ML model training and inference when the periodicities may not be guaranteed for aperiodic channel state information –reference signal (CSI-RS) resources in accordance with various aspects of the present disclosure.
[0023] FIG. 10B is a diagram illustrating an example of AI / ML model training and inference where equivalent periodicities are achievable by the aperiodic CSI-RS resources in accordance with various aspects of the present disclosure.
[0024] FIG. 11 is a call flow diagram illustrating a method of wireless communication in accordance with various aspects of the present disclosure.
[0025] FIG. 12 is a flowchart illustrating methods of wireless communication at a UE in accordance with various aspects of the present disclosure.
[0026] FIG. 13 is a flowchart illustrating methods of wireless communication at a UE in accordance with various aspects of the present disclosure.
[0027] FIG. 14 is a flowchart illustrating methods of wireless communication at a network entity in accordance with various aspects of the present disclosure.
[0028] FIG. 15 is a flowchart illustrating methods of wireless communication at a network entity in accordance with various aspects of the present disclosure.
[0029] FIG. 16 is a diagram illustrating an example of a hardware implementation for an example apparatus and / or network entity.
[0030] FIG. 17 is a diagram illustrating an example of a hardware implementation for an example network entity.
[0031] FIG. 18 is an illustrative block diagram of an example ML architecture of first wireless device in communication with second wireless device, in accordance with various aspects of the present disclosure.DETAILED DESCRIPTION
[0032] Artificial intelligence (AI) and machine learning (ML) (AI / ML) models may be used for wireless communication. For example, based on measurement results on the first set of beams (e.g., which may be referred to as Set-B beams) , an AI / ML model may be used for spatial or temporal domain beam predictions about a second set of beams (e.g., which may be referred to as Set-A beams) . In some aspects, the first set of beams (e.g., Set-B beams) may be a subset of the second set of beams (e.g., Set-A beams) , and the first set of beams (e.g., Set-B beams) may be obtained by sub-sampling the second set of beams (e.g., Set-A beams) . Temporal varied sub-sample patterns may be used to improve the prediction accuracy and robustness in making the beam predictions for the Set-A beams. It can be challenging to maintain consistency between the sub-sample patterns used between the training and inference of the AI / ML models, e.g., when aperiodic channel state information -reference signal (CSI-RS) resources are used. Hence, there is a need for methods and apparatus that ensure the spatial and temporal domain consistency for an AI / ML model when aperiodic CSI-RS resources are scheduled for the training and inference of an AI / ML model.
[0033] Various aspects relate generally to wireless communication. Some aspects more specifically relate to beam management or beam prediction for wireless communication that improves sub-sample pattern consistency for aperiodic reference signals. In some examples, a UE measures an aperiodic reference signal received on a first set of resources (e.g., resources associated with Set-B beams) and predicts one or more channel characteristics associated with a second set of resources (e.g., resources associated with Set-A beams) using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern. The first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. The UE further reports, to a network entity, the one or more channel characteristics predicted for the second set of resources. In some examples, the prediction of the one or more channel characteristics is based on an AI / ML model, and the initial training may be the training of the AI / ML model.
[0034] 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 spatial and temporal consistency in Set-B resources sub-sample patterns between the training and inference of an AI / ML model, the described techniques can be used to improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization. In some examples, by ensuring spatial and temporal consistency under different scenarios, such as for aperiodic CSI-RS resources, the described techniques ensure the performance of the AI / ML model under a variety of network environments, thereby improving the flexibility of the AI / ML model deployment.
[0035] The detailed description set forth below in connection with the drawings describes various configurations and does not represent the only configurations in which the concepts described herein may be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of various concepts. However, these concepts may be practiced without these specific details. In some instances, well known structures and components are shown in block diagram form in order to avoid obscuring such concepts.
[0036] Several aspects of telecommunication systems are presented with reference to various apparatus and methods. These apparatus and methods are described in the following detailed description and illustrated in the accompanying drawings by various blocks, components, circuits, processes, algorithms, etc. (collectively referred to as “elements” ) . These elements may be implemented using electronic hardware, computer software, or any combination thereof. Whether such elements are implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system.
[0037] By way of example, an element, or any portion of an element, or any combination of elements may be implemented as a “processing system” that includes one or more processors. When multiple processors are implemented, the multiple processors may perform the functions individually or in combination. Examples of processors include microprocessors, microcontrollers, graphics processing units (GPUs) , central processing units (CPUs) , application processors, digital signal processors (DSPs) , reduced instruction set computing (RISC) processors, systems on a chip (SoC) , baseband processors, field programmable gate arrays (FPGAs) , programmable logic devices (PLDs) , state machines, gated logic, discrete hardware circuits, and other suitable hardware configured to perform the various functionality described throughout this disclosure. One or more processors in the processing system may execute software. Software, whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise, shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software components, applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, or any combination thereof.
[0038] Accordingly, in one or more example aspects, implementations, and / or use cases, the functions described may be implemented in hardware, software, or any combination thereof. If implemented in software, the functions may be stored on or encoded as one or more instructions or code on a computer-readable medium. Computer-readable media includes computer storage media. Storage media may be any available media that can be accessed by a computer. By way of example, such computer-readable media can include a random-access memory (RAM) , a read-only memory (ROM) , an electrically erasable programmable ROM (EEPROM) , optical disk storage, magnetic disk storage, other magnetic storage devices, combinations of the types of computer-readable media, or any other medium that can be used to store computer executable code in the form of instructions or data structures that can be accessed by a computer.
[0039] While aspects, implementations, and / or use cases are described in this application by illustration to some examples, additional or different aspects, implementations and / or use cases may come about in many different arrangements and scenarios. Aspects, implementations, and / or use cases described herein may be implemented across many differing platform types, devices, systems, shapes, sizes, and packaging arrangements. For example, aspects, implementations, and / or use cases may come about via integrated chip implementations and other non-module-component based devices (e.g., end-user devices, vehicles, communication devices, computing devices, industrial equipment, retail / purchasing devices, medical devices, artificial intelligence (AI) -enabled devices, etc. ) . While some examples may or may not be specifically directed to use cases or applications, a wide assortment of applicability of described examples may occur. Aspects, implementations, and / or use cases may range a spectrum from chip-level or modular components to non-modular, non-chip-level implementations and further to aggregate, distributed, or original equipment manufacturer (OEM) devices or systems incorporating one or more techniques herein. In some practical settings, devices incorporating described aspects and features may also include additional components and features for implementation and practice of claimed and described aspect. For example, transmission and reception of wireless signals necessarily includes a number of components for analog and digital purposes (e.g., hardware components including antenna, RF-chains, power amplifiers, modulators, buffer, processor (s) , interleaver, adders / summers, etc. ) . Techniques described herein may be practiced in a wide variety of devices, chip-level components, systems, distributed arrangements, aggregated or disaggregated components, end-user devices, etc. of varying sizes, shapes, and constitution.
[0040] Deployment of communication systems, such as 5G NR systems, may be arranged in multiple manners with various components or constituent parts. In a 5G NR system, or network, a network node, a network entity, a mobility element of a network, a radio access network (RAN) node, a core network node, a network element, or a network equipment, such as a base station (BS) , or one or more units (or one or more components) performing base station functionality, may be implemented in an aggregated or disaggregated architecture. For example, a BS (such as a Node B (NB) , evolved NB (eNB) , NR BS, 5G NB, access point, a transmission reception point (TRP) , or a cell, etc. ) may be implemented as an aggregated base station (also known as a standalone BS or a monolithic BS) or a disaggregated base station.
[0041] An aggregated base station may be configured to utilize a radio protocol stack that is physically or logically integrated within a single RAN node. A disaggregated base station may be configured to utilize a protocol stack that is physically or logically distributed among two or more units (such as one or more central or centralized units (CUs) , one or more distributed units (DUs) , or one or more radio units (RUs) ) . In some aspects, a CU may be implemented within a RAN node, and one or more DUs may be co-located with the CU, or alternatively, may be geographically or virtually distributed throughout one or multiple other RAN nodes. The DUs may be implemented to communicate with one or more RUs. Each of the CU, DU and RU can be implemented as virtual units, i.e., a virtual central unit (VCU) , a virtual distributed unit (VDU) , or a virtual radio unit (VRU) .
[0042] Base station operation or network design may consider aggregation characteristics of base station functionality. For example, disaggregated base stations may be utilized in an integrated access backhaul (IAB) network, an open radio access network (O-RAN (such as the network configuration sponsored by the O-RAN Alliance) ) , or a virtualized radio access network (vRAN, also known as a cloud radio access network (C-RAN) ) . Disaggregation may include distributing functionality across two or more units at various physical locations, as well as distributing functionality for at least one unit virtually, which can enable flexibility in network design. The various units of the disaggregated base station, or disaggregated RAN architecture, can be configured for wired or wireless communication with at least one other unit.
[0043] FIG. 1 is a diagram 100 illustrating an example of a wireless communications system and an access network. The illustrated wireless communications system includes a disaggregated base station architecture. The disaggregated base station architecture may include one or more CUs 110 that can communicate directly with a core network 120 via a backhaul link, or indirectly with the core network 120 through one or more disaggregated base station units (such as a Near-Real Time (Near-RT) RAN Intelligent Controller (RIC) 125 via an E2 link, or a Non-Real Time (Non-RT) RIC 115 associated with a Service Management and Orchestration (SMO) Framework 105, or both) . A CU 110 may communicate with one or more DUs 130 via respective midhaul links, such as an F1 interface. The DUs 130 may communicate with one or more RUs 140 via respective fronthaul links. The RUs 140 may communicate with respective UEs 104 via one or more radio frequency (RF) access links. In some implementations, the UE 104 may be simultaneously served by multiple RUs 140.
[0044] Each of the units, i.e., the CUs 110, the DUs 130, the RUs 140, as well as the Near-RT RICs 125, the Non-RT RICs 115, and the SMO Framework 105, may include one or more interfaces or be coupled to one or more interfaces configured to receive or to transmit signals, data, or information (collectively, signals) via a wired or wireless transmission medium. Each of the units, or an associated processor or controller providing instructions to the communication interfaces of the units, can be configured to communicate with one or more of the other units via the transmission medium. For example, the units can include a wired interface configured to receive or to transmit signals over a wired transmission medium to one or more of the other units. Additionally, the units can include a wireless interface, which may include a receiver, a transmitter, or a transceiver (such as an RF transceiver) , configured to receive or to transmit signals, or both, over a wireless transmission medium to one or more of the other units.
[0045] In some aspects, the CU 110 may host one or more higher layer control functions. Such control functions can include radio resource control (RRC) , packet data convergence protocol (PDCP) , service data adaptation protocol (SDAP) , or the like. Each control function can be implemented with an interface configured to communicate signals with other control functions hosted by the CU 110. The CU 110 may be configured to handle user plane functionality (i.e., Central Unit –User Plane (CU-UP) ) , control plane functionality (i.e., Central Unit –Control Plane (CU-CP) ) , or a combination thereof. In some implementations, the CU 110 can be logically split into one or more CU-UP units and one or more CU-CP units. The CU-UP unit can communicate bidirectionally with the CU-CP unit via an interface, such as an E1 interface when implemented in an O-RAN configuration. The CU 110 can be implemented to communicate with the DU 130, as necessary, for network control and signaling.
[0046] The DU 130 may correspond to a logical unit that includes one or more base station functions to control the operation of one or more RUs 140. In some aspects, the DU 130 may host one or more of a radio link control (RLC) layer, a medium access control (MAC) layer, and one or more high physical (PHY) layers (such as modules for forward error correction (FEC) encoding and decoding, scrambling, modulation, demodulation, or the like) depending, at least in part, on a functional split, such as those defined by 3GPP. In some aspects, the DU 130 may further host one or more low PHY layers. Each layer (or module) can be implemented with an interface configured to communicate signals with other layers (and modules) hosted by the DU 130, or with the control functions hosted by the CU 110.
[0047] Lower-layer functionality can be implemented by one or more RUs 140. In some deployments, an RU 140, controlled by a DU 130, may correspond to a logical node that hosts RF processing functions, or low-PHY layer functions (such as performing fast Fourier transform (FFT) , inverse FFT (iFFT) , digital beamforming, physical random access channel (PRACH) extraction and filtering, or the like) , or both, based at least in part on the functional split, such as a lower layer functional split. In such an architecture, the RU (s) 140 can be implemented to handle over the air (OTA) communication with one or more UEs 104. In some implementations, real-time and non-real-time aspects of control and user plane communication with the RU (s) 140 can be controlled by the corresponding DU 130. In some scenarios, this configuration can enable the DU (s) 130 and the CU 110 to be implemented in a cloud-based RAN architecture, such as a vRAN architecture.
[0048] The SMO Framework 105 may be configured to support RAN deployment and provisioning of non-virtualized and virtualized network elements. For non-virtualized network elements, the SMO Framework 105 may be configured to support the deployment of dedicated physical resources for RAN coverage requirements that may be managed via an operations and maintenance interface (such as an O1 interface) . For virtualized network elements, the SMO Framework 105 may be configured to interact with a cloud computing platform (such as an open cloud (O-Cloud) 190) to perform network element life cycle management (such as to instantiate virtualized network elements) via a cloud computing platform interface (such as an O2 interface) . Such virtualized network elements can include, but are not limited to, CUs 110, DUs 130, RUs 140 and Near-RT RICs 125. In some implementations, the SMO Framework 105 can communicate with a hardware aspect of a 4G RAN, such as an open eNB (O-eNB) 111, via an O1 interface. Additionally, in some implementations, the SMO Framework 105 can communicate directly with one or more RUs 140 via an O1 interface. The SMO Framework 105 also may include a Non-RT RIC 115 configured to support functionality of the SMO Framework 105.
[0049] The Non-RT RIC 115 may be configured to include a logical function that enables non-real-time control and optimization of RAN elements and resources, artificial intelligence (AI) / machine learning (ML) (AI / ML) workflows including model training and updates, or policy-based guidance of applications / features in the Near-RT RIC 125. The Non-RT RIC 115 may be coupled to or communicate with (such as via an A1 interface) the Near-RT RIC 125. The Near-RT RIC 125 may be configured to include a logical function that enables near-real-time control and optimization of RAN elements and resources via data collection and actions over an interface (such as via an E2 interface) connecting one or more CUs 110, one or more DUs 130, or both, as well as an O-eNB, with the Near-RT RIC 125.
[0050] In some implementations, to generate AI / ML models to be deployed in the Near-RT RIC 125, the Non-RT RIC 115 may receive parameters or external enrichment information from external servers. Such information may be utilized by the Near-RT RIC 125 and may be received at the SMO Framework 105 or the Non-RT RIC 115 from non-network data sources or from network functions. In some examples, the Non-RT RIC 115 or the Near-RT RIC 125 may be configured to tune RAN behavior or performance. For example, the Non-RT RIC 115 may monitor long-term trends and patterns for performance and employ AI / ML models to perform corrective actions through the SMO Framework 105 (such as reconfiguration via O1) or via creation of RAN management policies (such as A1 policies) .
[0051] At least one of the CU 110, the DU 130, and the RU 140 may be referred to as a base station 102. Accordingly, a base station 102 may include one or more of the CU 110, the DU 130, and the RU 140 (each component indicated with dotted lines to signify that each component may or may not be included in the base station 102) . The base station 102 provides an access point to the core network 120 for a UE 104. The base station 102 may include macrocells (high power cellular base station) and / or small cells (low power cellular base station) . The small cells include femtocells, picocells, and microcells. A network that includes both small cell and macrocells may be known as a heterogeneous network. A heterogeneous network may also include Home Evolved Node Bs (eNBs) (HeNBs) , which may provide service to a restricted group known as a closed subscriber group (CSG) . The communication links between the RUs 140 and the UEs 104 may include uplink (UL) (also referred to as reverse link) transmissions from a UE 104 to an RU 140 and / or downlink (DL) (also referred to as forward link) transmissions from an RU 140 to a UE 104. The communication links may use multiple-input and multiple-output (MIMO) antenna technology, including spatial multiplexing, beamforming, and / or transmit diversity. The communication links may be through one or more carriers. The base station 102 / UEs 104 may use spectrum up to Y MHz (e.g., 5, 10, 15, 20, 100, 400, etc. MHz) bandwidth per carrier allocated in a carrier aggregation of up to a total of Yx MHz (x component carriers) used for transmission in each direction. The carriers may or may not be adjacent to each other. Allocation of carriers may be asymmetric with respect to DL and UL (e.g., more or fewer carriers may be allocated for DL than for UL) . The component carriers may include a primary component carrier and one or more secondary component carriers. A primary component carrier may be referred to as a primary cell (PCell) and a secondary component carrier may be referred to as a secondary cell (SCell) .
[0052] Certain UEs 104 may communicate with each other using device-to-device (D2D) communication link 158. The D2D communication link 158 may use the DL / UL wireless wide area network (WWAN) spectrum. The D2D communication link 158 may use one or more sidelink channels, such as a physical sidelink broadcast channel (PSBCH) , a physical sidelink discovery channel (PSDCH) , a physical sidelink shared channel (PSSCH) , and a physical sidelink control channel (PSCCH) . D2D communication may be through a variety of wireless D2D communications systems, such as for example, BluetoothTM (Bluetooth is a trademark of the Bluetooth Special Interest Group (SIG) ) , Wi-FiTM (Wi-Fi is a trademark of the Wi-Fi Alliance) based on the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standard, LTE, or NR.
[0053] The wireless communications system may further include a Wi-Fi access point (AP) 150 in communication with UEs 104 (also referred to as Wi-Fi stations (STAs) ) via communication link 154, e.g., in a 5 GHz unlicensed frequency spectrum or the like. When communicating in an unlicensed frequency spectrum, the UEs 104 / AP 150 may perform a clear channel assessment (CCA) prior to communicating in order to determine whether the channel is available.
[0054] The electromagnetic spectrum is often subdivided, based on frequency / wavelength, into various classes, bands, channels, etc. In 5G NR, two initial operating bands have been identified as frequency range designations FR1 (410 MHz –7.125 GHz) and FR2 (24.25 GHz –52.6 GHz) . Although a portion of FR1 is greater than 6 GHz, FR1 is often referred to (interchangeably) as a “sub-6 GHz” band in various documents and articles. A similar nomenclature issue sometimes occurs with regard to FR2, which is often referred to (interchangeably) as a “millimeter wave” band in documents and articles, despite being different from the extremely high frequency (EHF) band (30 GHz –300 GHz) which is identified by the International Telecommunications Union (ITU) as a “millimeter wave” band.
[0055] The frequencies between FR1 and FR2 are often referred to as mid-band frequencies. Recent 5G NR studies have identified an operating band for these mid-band frequencies as frequency range designation FR3 (7.125 GHz –24.25 GHz) . Frequency bands falling within FR3 may inherit FR1 characteristics and / or FR2 characteristics, and thus may effectively extend features of FR1 and / or FR2 into mid-band frequencies. In addition, higher frequency bands are currently being explored to extend 5G NR operation beyond 52.6 GHz. For example, three higher operating bands have been identified as frequency range designations FR2-2 (52.6 GHz –71 GHz) , FR4 (71 GHz –114.25 GHz) , and FR5 (114.25 GHz –300 GHz) . Each of these higher frequency bands falls within the EHF band.
[0056] With the above aspects in mind, unless specifically stated otherwise, the term “sub-6 GHz” or the like if used herein may broadly represent frequencies that may be less than 6 GHz, may be within FR1, or may include mid-band frequencies. Further, unless specifically stated otherwise, the term “millimeter wave” or the like if used herein may broadly represent frequencies that may include mid-band frequencies, may be within FR2, FR4, FR2-2, and / or FR5, or may be within the EHF band.
[0057] The base station 102 and the UE 104 may each include a plurality of antennas, such as antenna elements, antenna panels, and / or antenna arrays to facilitate beamforming. The base station 102 may transmit a beamformed signal 182 to the UE 104 in one or more transmit directions. The UE 104 may receive the beamformed signal from the base station 102 in one or more receive directions. The UE 104 may also transmit a beamformed signal 184 to the base station 102 in one or more transmit directions. The base station 102 may receive the beamformed signal from the UE 104 in one or more receive directions. The base station 102 / UE 104 may perform beam training to determine the best receive and transmit directions for each of the base station 102 / UE 104. The transmit and receive directions for the base station 102 may or may not be the same. The transmit and receive directions for the UE 104 may or may not be the same.
[0058] The base station 102 may include and / or be referred to as a gNB, Node B, eNB, an access point, a base transceiver station, a radio base station, a radio transceiver, a transceiver function, a basic service set (BSS) , an extended service set (ESS) , a TRP, network node, network entity, network equipment, or some other suitable terminology. The base station 102 can be implemented as an integrated access and backhaul (IAB) node, a relay node, a sidelink node, an aggregated (monolithic) base station with a baseband unit (BBU) (including a CU and a DU) and an RU, or as a disaggregated base station including one or more of a CU, a DU, and / or an RU. The set of base stations, which may include disaggregated base stations and / or aggregated base stations, may be referred to as next generation (NG) RAN (NG-RAN) .
[0059] The core network 120 may include an Access and Mobility Management Function (AMF) 161, a Session Management Function (SMF) 162, a User Plane Function (UPF) 163, a Unified Data Management (UDM) 164, one or more location servers 168, and other functional entities. The AMF 161 is the control node that processes the signaling between the UEs 104 and the core network 120. The AMF 161 supports registration management, connection management, mobility management, and other functions. The SMF 162 supports session management and other functions. The UPF 163 supports packet routing, packet forwarding, and other functions. The UDM 164 supports the generation of authentication and key agreement (AKA) credentials, user identification handling, access authorization, and subscription management. The one or more location servers 168 are illustrated as including a Gateway Mobile Location Center (GMLC) 165 and a Location Management Function (LMF) 166. However, generally, the one or more location servers 168 may include one or more location / positioning servers, which may include one or more of the GMLC 165, the LMF 166, a position determination entity (PDE) , a serving mobile location center (SMLC) , a mobile positioning center (MPC) , or the like. The GMLC 165 and the LMF 166 support UE location services. The GMLC 165 provides an interface for clients / applications (e.g., emergency services) for accessing UE positioning information. The LMF 166 receives measurements and assistance information from the NG-RAN and the UE 104 via the AMF 161 to compute the position of the UE 104. The NG-RAN may utilize one or more positioning methods in order to determine the position of the UE 104. Positioning the UE 104 may involve signal measurements, a position estimate, and an optional velocity computation based on the measurements. The signal measurements may be made by the UE 104 and / or the base station 102 serving the UE 104. The signals measured may be based on one or more of a satellite positioning system (SPS) 170 (e.g., one or more of a Global Navigation Satellite System (GNSS) , global position system (GPS) , non-terrestrial network (NTN) , or other satellite position / location system) , LTE signals, wireless local area network (WLAN) signals, Bluetooth signals, a terrestrial beacon system (TBS) , sensor-based information (e.g., barometric pressure sensor, motion sensor) , NR enhanced cell ID (NR E-CID) methods, NR signals (e.g., multi-round trip time (Multi-RTT) , DL angle-of-departure (DL-AoD) , DL time difference of arrival (DL-TDOA) , UL time difference of arrival (UL-TDOA) , and UL angle-of-arrival (UL-AoA) positioning) , and / or other systems / signals / sensors.
[0060] Examples of UEs 104 include a cellular phone, a smart phone, a session initiation protocol (SIP) phone, a laptop, a personal digital assistant (PDA) , a satellite radio, a global positioning system, a multimedia device, a video device, a digital audio player (e.g., MP3 player) , a camera, a game console, a tablet, a smart device, a wearable device, a vehicle, an electric meter, a gas pump, a large or small kitchen appliance, a healthcare device, an implant, a sensor / actuator, a display, or any other similar functioning device. Some of the UEs 104 may be referred to as IoT devices (e.g., parking meter, gas pump, toaster, vehicles, heart monitor, etc. ) . The UE 104 may also be referred to as a station, a mobile station, a subscriber station, a mobile unit, a subscriber unit, a wireless unit, a remote unit, a mobile device, a wireless device, a wireless communications device, a remote device, a mobile subscriber station, an access terminal, a mobile terminal, a wireless terminal, a remote terminal, a handset, a user agent, a mobile client, a client, or some other suitable terminology. In some scenarios, the term UE may also apply to one or more companion devices such as in a device constellation arrangement. One or more of these devices may collectively access the network and / or individually access the network.
[0061] Referring again to FIG. 1, in certain aspects, the UE 104 may include the resource consistency component 198. The resource consistency component 198 may be configured to measure an aperiodic reference signal received on a first set of resources; predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and report, to a network entity, the one or more channel characteristics predicted for the second set of resources. In certain aspects, the base station 102 may include the resource consistency component 199. The resource consistency component 199 may be configured to provide an aperiodic reference signal on a first set of resources; indicate for a UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and receive, from the UE, the one or more channel characteristics predicted for the second set of resources. Although the following description may be focused on 5G NR, the concepts described herein may be applicable to other similar areas, such as LTE, LTE-A, CDMA, GSM, and other wireless technologies.
[0062] FIG. 2A is a diagram 200 illustrating an example of a first subframe within a 5G NR frame structure. FIG. 2B is a diagram 230 illustrating an example of DL channels within a 5G NR subframe. FIG. 2C is a diagram 250 illustrating an example of a second subframe within a 5G NR frame structure. FIG. 2D is a diagram 280 illustrating an example of UL channels within a 5G NR subframe. The 5G NR frame structure may be frequency division duplexed (FDD) in which for a particular set of subcarriers (carrier system bandwidth) , subframes within the set of subcarriers are dedicated for either DL or UL, or may be time division duplexed (TDD) in which for a particular set of subcarriers (carrier system bandwidth) , subframes within the set of subcarriers are dedicated for both DL and UL. In the examples provided by FIGs. 2A, 2C, the 5G NR frame structure is assumed to be TDD, with subframe 4 being configured with slot format 28 (with mostly DL) , where D is DL, U is UL, and F is flexible for use between DL / UL, and subframe 3 being configured with slot format 1 (with all UL) . While subframes 3, 4 are shown with slot formats 1, 28, respectively, any particular subframe may be configured with any of the various available slot formats 0-61. Slot formats 0, 1 are all DL, UL, respectively. Other slot formats 2-61 include a mix of DL, UL, and flexible symbols. UEs are configured with the slot format (dynamically through DL control information (DCI) , or semi-statically / statically through radio resource control (RRC) signaling) through a received slot format indicator (SFI) . Note that the description infra applies also to a 5G NR frame structure that is TDD.
[0063] FIGs. 2A-2D illustrate a frame structure, and the aspects of the present disclosure may be applicable to other wireless communication technologies, which may have a different frame structure and / or different channels. A frame (10 ms) may be divided into 10 equally sized subframes (1 ms) . Each subframe may include one or more time slots. Subframes may also include mini-slots, which may include 7, 4, or 2 symbols. Each slot may include 14 or 12 symbols, depending on whether the cyclic prefix (CP) is normal or extended. For normal CP, each slot may include 14 symbols, and for extended CP, each slot may include 12 symbols. The symbols on DL may be CP orthogonal frequency division multiplexing (OFDM) (CP-OFDM) symbols. The symbols on UL may be CP-OFDM symbols (for high throughput scenarios) or discrete Fourier transform (DFT) spread OFDM (DFT-s-OFDM) symbols (for power limited scenarios; limited to a single stream transmission) . The number of slots within a subframe is based on the CP and the numerology. The numerology defines the subcarrier spacing (SCS) (see Table 1) . The symbol length / duration may scale with 1 / SCS.
[0064] Table 1: Numerology, SCS, and CP
[0065] For normal CP (14 symbols / slot) , different numerologies μ 0 to 4 allow for 1, 2, 4, 8, and 16 slots, respectively, per subframe. For extended CP, the numerology 2 allows for 4 slots per subframe. Accordingly, for normal CP and numerology μ, there are 14 symbols / slot and 2μ slots / subframe. The subcarrier spacing may be equal to 2μ* 15 kHz, where μ is the numerology 0 to 4. As such, the numerology μ=0 has a subcarrier spacing of 15 kHz and the numerology μ=4 has a subcarrier spacing of 240 kHz. The symbol length / duration is inversely related to the subcarrier spacing. FIGs. 2A-2D provide an example of normal CP with 14 symbols per slot and numerology μ=2 with 4 slots per subframe. The slot duration is 0.25 ms, the subcarrier spacing is 60 kHz, and the symbol duration is approximately 16.67 μs. Within a set of frames, there may be one or more different bandwidth parts (BWPs) (see FIG. 2B) that are frequency division multiplexed. Each BWP may have a particular numerology and CP (normal or extended) .
[0066] A resource grid may be used to represent the frame structure. Each time slot includes a resource block (RB) (also referred to as physical RBs (PRBs) ) that extends 12 consecutive subcarriers. The resource grid is divided into multiple resource elements (REs) . The number of bits carried by each RE depends on the modulation scheme.
[0067] As illustrated in FIG. 2A, some of the REs carry reference (pilot) signals (RS) for the UE. The RS may include demodulation RS (DM-RS) (indicated as R for one particular configuration, but other DM-RS configurations are possible) and channel state information reference signals (CSI-RS) for channel estimation at the UE. The RS may also include beam measurement RS (BRS) , beam refinement RS (BRRS) , and phase tracking RS (PT-RS) .
[0068] FIG. 2B illustrates an example of various DL channels within a subframe of a frame. The physical downlink control channel (PDCCH) carries DCI within one or more control channel elements (CCEs) (e.g., 1, 2, 4, 8, or 16 CCEs) , each CCE including six RE groups (REGs) , each REG including 12 consecutive REs in an OFDM symbol of an RB. A PDCCH within one BWP may be referred to as a control resource set (CORESET) . A UE is configured to monitor PDCCH candidates in a PDCCH search space (e.g., common search space, UE-specific search space) during PDCCH monitoring occasions on the CORESET, where the PDCCH candidates have different DCI formats and different aggregation levels. Additional BWPs may be located at greater and / or lower frequencies across the channel bandwidth. A primary synchronization signal (PSS) may be within symbol 2 of particular subframes of a frame. The PSS is used by a UE 104 to determine subframe / symbol timing and a physical layer identity. A secondary synchronization signal (SSS) may be within symbol 4 of particular subframes of a frame. The SSS is used by a UE to determine a physical layer cell identity group number and radio frame timing. Based on the physical layer identity and the physical layer cell identity group number, the UE can determine a physical cell identifier (PCI) . Based on the PCI, the UE can determine the locations of the DM-RS. The physical broadcast channel (PBCH) , which carries a master information block (MIB) , may be logically grouped with the PSS and SSS to form a synchronization signal (SS) / PBCH block (also referred to as SS block (SSB) ) . The MIB provides a number of RBs in the system bandwidth and a system frame number (SFN) . The physical downlink shared channel (PDSCH) carries user data, broadcast system information not transmitted through the PBCH such as system information blocks (SIBs) , and paging messages.
[0069] As illustrated in FIG. 2C, some of the REs carry DM-RS (indicated as R for one particular configuration, but other DM-RS configurations are possible) for channel estimation at the base station. The UE may transmit DM-RS for the physical uplink control channel (PUCCH) and DM-RS for the physical uplink shared channel (PUSCH) . The PUSCH DM-RS may be transmitted in the first one or two symbols of the PUSCH. The PUCCH DM-RS may be transmitted in different configurations depending on whether short or long PUCCHs are transmitted and depending on the particular PUCCH format used. The UE may transmit sounding reference signals (SRS) . The SRS may be transmitted in the last symbol of a subframe. The SRS may have a comb structure, and a UE may transmit SRS on one of the combs. The SRS may be used by a base station for channel quality estimation to enable frequency-dependent scheduling on the UL.
[0070] FIG. 2D illustrates an example of various UL channels within a subframe of a frame. The PUCCH may be located as indicated in one configuration. The PUCCH carries uplink control information (UCI) , such as scheduling requests, a channel quality indicator (CQI) , a precoding matrix indicator (PMI) , a rank indicator (RI) , and hybrid automatic repeat request (HARQ) acknowledgment (ACK) (HARQ-ACK) feedback (i.e., one or more HARQ ACK bits indicating one or more ACK and / or negative ACK (NACK) ) . The PUSCH carries data, and may additionally be used to carry a buffer status report (BSR) , a power headroom report (PHR) , and / or UCI.
[0071] FIG. 3 is a block diagram of a base station 310 in communication with a UE 350 in an access network. In the DL, Internet protocol (IP) packets may be provided to a controller / processor 375. The controller / processor 375 implements layer 3 and layer 2 functionality. Layer 3 includes a radio resource control (RRC) layer, and layer 2 includes a service data adaptation protocol (SDAP) layer, a packet data convergence protocol (PDCP) layer, a radio link control (RLC) layer, and a medium access control (MAC) layer. The controller / processor 375 provides RRC layer functionality associated with broadcasting of system information (e.g., MIB, SIBs) , RRC connection control (e.g., RRC connection paging, RRC connection establishment, RRC connection modification, and RRC connection release) , inter radio access technology (RAT) mobility, and measurement configuration for UE measurement reporting; PDCP layer functionality associated with header compression / decompression, security (ciphering, deciphering, integrity protection, integrity verification) , and handover support functions; RLC layer functionality associated with the transfer of upper layer packet data units (PDUs) , error correction through ARQ, concatenation, segmentation, and reassembly of RLC service data units (SDUs) , re-segmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto transport blocks (TBs) , demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction through HARQ, priority handling, and logical channel prioritization.
[0072] The transmit (TX) processor 316 and the receive (RX) processor 370 implement layer 1 functionality associated with various signal processing functions. Layer 1, which includes a physical (PHY) layer, may include error detection on the transport channels, forward error correction (FEC) coding / decoding of the transport channels, interleaving, rate matching, mapping onto physical channels, modulation / demodulation of physical channels, and MIMO antenna processing. The TX processor 316 handles mapping to signal constellations based on various modulation schemes (e.g., binary phase-shift keying (BPSK) , quadrature phase-shift keying (QPSK) , M-phase-shift keying (M-PSK) , M-quadrature amplitude modulation (M-QAM) ) . The coded and modulated symbols may then be split into parallel streams. Each stream may then be mapped to an OFDM subcarrier, multiplexed with a reference signal (e.g., pilot) in the time and / or frequency domain, and then combined together using an Inverse Fast Fourier Transform (IFFT) to produce a physical channel carrying a time domain OFDM symbol stream. The OFDM stream is spatially precoded to produce multiple spatial streams. Channel estimates from a channel estimator 374 may be used to determine the coding and modulation scheme, as well as for spatial processing. The channel estimate may be derived from a reference signal and / or channel condition feedback transmitted by the UE 350. Each spatial stream may then be provided to a different antenna 320 via a separate transmitter 318Tx. Each transmitter 318Tx may modulate a radio frequency (RF) carrier with a respective spatial stream for transmission.
[0073] At the UE 350, each receiver 354Rx receives a signal through its respective antenna 352. Each receiver 354Rx recovers information modulated onto an RF carrier and provides the information to the receive (RX) processor 356. The TX processor 368 and the RX processor 356 implement layer 1 functionality associated with various signal processing functions. The RX processor 356 may perform spatial processing on the information to recover any spatial streams destined for the UE 350. If multiple spatial streams are destined for the UE 350, they may be combined by the RX processor 356 into a single OFDM symbol stream. The RX processor 356 then converts the OFDM symbol stream from the time-domain to the frequency domain using a Fast Fourier Transform (FFT) . The frequency domain signal includes a separate OFDM symbol stream for each subcarrier of the OFDM signal. The symbols on each subcarrier, and the reference signal, are recovered and demodulated by determining the most likely signal constellation points transmitted by the base station 310. These soft decisions may be based on channel estimates computed by the channel estimator 358. The soft decisions are then decoded and deinterleaved to recover the data and control signals that were originally transmitted by the base station 310 on the physical channel. The data and control signals are then provided to the controller / processor 359, which implements layer 3 and layer 2 functionality.
[0074] The controller / processor 359 can be associated with at least one memory 360 that stores program codes and data. The at least one memory 360 may be referred to as a computer-readable medium. In the UL, the controller / processor 359 provides demultiplexing between transport and logical channels, packet reassembly, deciphering, header decompression, and control signal processing to recover IP packets. The controller / processor 359 is also responsible for error detection using an ACK and / or NACK protocol to support HARQ operations.
[0075] Similar to the functionality described in connection with the DL transmission by the base station 310, the controller / processor 359 provides RRC layer functionality associated with system information (e.g., MIB, SIBs) acquisition, RRC connections, and measurement reporting; PDCP layer functionality associated with header compression / decompression, and security (ciphering, deciphering, integrity protection, integrity verification) ; RLC layer functionality associated with the transfer of upper layer PDUs, error correction through ARQ, concatenation, segmentation, and reassembly of RLC SDUs, re-segmentation of RLC data PDUs, and reordering of RLC data PDUs; and MAC layer functionality associated with mapping between logical channels and transport channels, multiplexing of MAC SDUs onto TBs, demultiplexing of MAC SDUs from TBs, scheduling information reporting, error correction through HARQ, priority handling, and logical channel prioritization.
[0076] Channel estimates derived by a channel estimator 358 from a reference signal or feedback transmitted by the base station 310 may be used by the TX processor 368 to select the appropriate coding and modulation schemes, and to facilitate spatial processing. The spatial streams generated by the TX processor 368 may be provided to different antenna 352 via separate transmitters 354Tx. Each transmitter 354Tx may modulate an RF carrier with a respective spatial stream for transmission.
[0077] The UL transmission is processed at the base station 310 in a manner similar to that described in connection with the receiver function at the UE 350. Each receiver 318Rx receives a signal through its respective antenna 320. Each receiver 318Rx recovers information modulated onto an RF carrier and provides the information to a RX processor 370.
[0078] The controller / processor 375 can be associated with at least one memory 376 that stores program codes and data. The at least one memory 376 may be referred to as a computer-readable medium. In the UL, the controller / processor 375 provides demultiplexing between transport and logical channels, packet reassembly, deciphering, header decompression, control signal processing to recover IP packets. The controller / processor 375 is also responsible for error detection using an ACK and / or NACK protocol to support HARQ operations.
[0079] At least one of the TX processor 368, the RX processor 356, and the controller / processor 359 may be configured to perform aspects in connection with the resource consistency component 198 of FIG. 1.
[0080] At least one of the TX processor 316, the RX processor 370, and the controller / processor 375 may be configured to perform aspects in connection with the resource consistency component 199 of FIG. 1.
[0081] Some aspects and techniques as described herein may be implemented, at least in part, using an artificial intelligence (AI) program, such as a program that includes a machine learning (ML) or artificial neural network (ANN) model. An example ML model may include mathematical representations or define computing capabilities for making inferences from input data based on patterns or relationships identified in the input data. As used herein, the term “inferences” can include one or more decisions, predictions, determinations, or values, which may represent outputs of the ML model. The computing capabilities may be defined in terms of certain parameters of the ML model, such as weights and biases. Weights may indicate relationships between certain input data and certain outputs of the ML model, and biases are offsets that may indicate a starting point for the outputs of the ML model. An example ML model operating on input data may start at an initial output based on the biases and then update its output based on a combination of the input data and the weights.
[0082] In some aspects, an ML model may be configured to provide computing capabilities for wireless communications. Such an ML model may be configured with weights and biases to perform predictions regarding a set of resources (e.g., Set-A beams) based on measurements of another set of resources (e.g., Set-B beams) . Thus, during the operation of a device, the ML model may receive input data (such as measurements associated with the first set of resources (e.g., Set-B beam measurements) and make inferences (such as predictions for Set-A beams) based on the weights and biases. The ML model may be employed to assist in beam management or beam selection using a reduced set of measurements.
[0083] ML models may be deployed in one or more devices (for example, network entities and user equipments (UEs) ) and may be configured to enhance various aspects of a wireless communication system. For example, an ML model may be trained to identify patterns or relationships in data corresponding to a network, a device, an air interface, or the like. An ML model may support operational decisions relating to one or more aspects associated with wireless communications devices, networks, or services. For example, an ML model may be utilized for supporting or improving aspects such as signal coding / decoding, network routing, energy conservation, transceiver circuitry controls, frequency synchronization, timing synchronization, channel state estimation, channel equalization, channel state feedback, modulation, demodulation, device positioning, beamforming, load balancing, operations and management functions, security, etc.
[0084] ML models may be characterized in terms of types of learning that generate specific types of learned models that perform specific types of tasks. For example, different types of machine learning include supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, etc. ML models may be used to perform different tasks, such as classification or regression, where classification refers to determining one or more discrete output values from a set of predefined output values, and regression refers to determining continuous values that are not bounded by predefined output values. Some example ML models configured for performing such tasks include ANNs such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) , transformers, diffusion models, regression analysis models (such as statistical models) , large language models (LLMs) , decision tree learning (such as predictive models) , support vector networks (SVMs) , and probabilistic graphical models (such as a Bayesian network) , etc.
[0085] The description herein illustrates, by way of some examples, how one or more tasks or problems in wireless communications may benefit from the application of one or more ML models for the prediction of one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on a first set of resources based on a first mapping pattern. The first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. To facilitate the discussion, an ML model configured using an ANN is used, but it should be understood that other types of ML models may be used instead of an ANN. Hence, unless expressly recited, subject matter regarding an ML model is not necessarily intended to be limited to an ANN solution. Further, it should be understood that, unless otherwise specifically stated, terms such “AI / ML model, ” “ML model, ” “trained ML mode, ” “ANN, ” “model, ” “algorithm, ” or the like are intended to be interchangeable.
[0086] FIG. 4 is an illustrative block diagram of an example machine learning (ML) model represented by an artificial neural network (ANN) 400. ANN 400 may receive input data 406, which may include one or more bits of data A02, pre-processed data output from pre-processor 404 (optional) , or some combination thereof. Here, data 402 may include training data, verification data, application-related data, or the like, based, for example, on the stage of deployment of ANN 400. Pre-processor 404 may be included within ANN 400 in some other implementations. Pre-processor 404 may, for example, process all or a portion of data 402, which may result in some of data 402 being changed, replaced, deleted, etc. In some implementations, pre-processor 404 may add additional data to data 402. In some implementations, the pre-processor 404 may be an ML model, such as an ANN. As an example, the input may include measurements performed on Set-B beams.
[0087] The ANN 400 includes at least one first layer 408 of artificial neurons 410 to process input data 406 and provide resulting first layer data via connections or “edges” such as edges 412 to at least a portion of at least one second layer 414. Second layer 414 processes data received via edges 412 and provides second layer output data via edges 416 to at least a portion of at least one third layer 418. Third layer 418 processes data received via edges 416 and provides third layer output data via edges 420 to at least a portion of a final layer 422, including one or more neurons to provide output data 424. All or part of output data 424 may be further processed in some manner by (optional) post-processor 426. Thus, in certain examples, ANN 400 may provide output data 428 that is based on output data 424, post-processed data output from post-processor 426, or some combination thereof. As an example, the output may include a set of resource (e.g., beam) predictions for Set-A beams. A base station or UE may then select a beam for use in transmission and / or reception based on the beam predictions for the Set-Abeams output from the AI / ML model.
[0088] Post-processor 426 may be included within ANN 400 in some other implementations. Post-processor 426 may, for example, process all or a portion of output data 424, which may result in output data 428 being different, at least in part, from output data 424, as a result of data being changed, replaced, deleted, etc. In some implementations, post-processor 426 may be configured to add additional data to output data 424. In this example, second layer 414 and third layer 418 represent intermediate or hidden layers that may be arranged in a hierarchical or other like structure. Although not explicitly shown, there may be one or more further intermediate layers between the second layer 414 and the third layer 418. In some implementations, the post-processor 426 may be an ML model, such as an ANN.
[0089] The structure and training of artificial neurons 410 in the various layers may be tailored to the specific requirements of an application. Within a given layer, such as first layer 408, second layer 414, or third layer 418 of ANN 400, some or all of the neurons may be configured to process information provided to the layer and output corresponding transformed information from the layer. For example, transformed information from a layer may represent a weighted sum of the input information associated with or otherwise based on a non-linear activation function or other activation function used to “activate” the artificial neurons of the next layer. Artificial neurons in such a layer may be activated by or be responsive to parameters such as the previously described weights and biases of ANN 400. The weights and biases of ANN 400 may be adjusted during a training process or during operation of ANN 400. The weights of the various artificial neurons may control the strength of connections between layers or artificial neurons, while the biases may control the direction of connections between the layers or artificial neurons. An activation function may select or determine whether an artificial neuron transmits its output to the next layer or not in response to its received data.
[0090] Different activation functions may be used to model different types of non-linear relationships. By introducing non-linearity into an ML model, an activation function allows the configuration for the ML model to change in response to identifying or detecting complex patterns and relationships in the input data A06. Some non-exhaustive example activation functions include a sigmoid based activation function, a hyperbolic tangent (tanh) based activation function, a convolutional activation function, up-sampling, pooling, and a rectified linear unit (ReLU) based activation function.
[0091] Training of an ML model, such as ANN 400, may be conducted using training data. Training data may include one or more datasets that ANN 400 may use to identify patterns or relationships. Training data may represent various types of information, including written, visual, audio, environmental context, operational properties, etc. During training, the parameters (such as the weights and biases) of artificial neurons 410 may be changed, such as to minimize or otherwise reduce a loss function or a cost function. A training process may be repeated multiple times to fine-tune ANN 400 with each iteration.
[0092] Various ANN model structures are available for consideration. For example, in a feedforward ANN structure, each artificial neuron 410 in layer A14 receives information from the previous layer (such as one or more artificial neurons 410 in layer 408) and produces information for the next layer (such as one or more artificial neurons 410 in layer 418) . In a convolutional ANN structure, some layers may be organized into filters that extract features from data, such as the training data or the input data. In a recurrent ANN structure, some layers may have connections that allow for the processing of data across time, such as for processing information having a temporal structure, such as time series data forecasting.
[0093] In an autoencoder ANN structure, compact representations of data may be processed and the model trained to predict or potentially reconstruct original data from a reduced set of features. An autoencoder ANN structure may be useful for tasks related to dimensionality reduction and data compression.
[0094] A generative adversarial ANN structure may include a generator ANN and a discriminator ANN that are trained to compete with each other. Generative-adversarial networks (GANs) are ANN structures that may be useful for tasks relating to generating synthetic data or improving the performance of other models.
[0095] A transformer ANN structure makes use of attention mechanisms that may enable the model to process input sequences in a parallel and efficient manner. An attention mechanism allows the model to focus on different parts of the input sequence at different times. Attention mechanisms may be implemented using a series of layers known as attention layers to compute weighted sums of input features based on a similarity between different elements of the input sequence. A transformer ANN structure may include a series of feedforward ANN layers whose configurations may change in response to identifying non-linear relationships between the input and output sequences, which may also be referred to as a process of “learning” by the ANN layers. The output of a transformer ANN structure may be obtained by applying a linear transformation to the output of a final attention layer. A transformer ANN structure may be of particular use for tasks that involve sequence modeling, or other like processing.
[0096] Another example type of ANN structure is a model with one or more invertible layers. Models of this type may be inverted or “unwrapped” to reveal the input data that was used to generate the output of a layer. Other example types of ANN model structures include fully connected neural networks (FCNNs) and long short-term memory (LSTM) networks.
[0097] ANN 400 or other ML models may be implemented in various types of processing circuits along with memory and applicable instructions therein. For example, general-purpose hardware circuits, such as one or more central processing units (CPUs) , one or more graphics processing units (GPUs) , or suitable combinations thereof, may be employed to implement a model. In some implementations, one or more tensor processing units (TPUs) , neural processing units (NPUs) , or other special-purpose processors, field-programmable gate arrays (FPGAs) , application-specific integrated circuits (ASICs) , or the like may also be employed. In some implementations, the ML model may be implemented by an NPU or a TPU embedded in a system on chip (SoC) along with other components, such as one or more CPUs, GPUs, etc. A SoC includes several components manufactured on a shared semiconductor substrate. The NPU or TPU may be controlled by the one or more CPUs by configuring the ML model implemented by the NPU or TPU with weights and biases, providing certain training data to the ML model to configure the ML model, or providing input data to the ML model to obtain related inferences. The one or more CPUs may also receive the inferences and be configured to perform certain actions based on the inferences produced by the ML model. The actions performed by the one or more CPUs may include sending commands to other components of the SoC or components external to the SoC to perform certain actions. For example, the CPU may send commands to an RF transceiver based on the outputs or inferences obtained from an ML model to cause the RF transceiver to operate on a wireless network in accordance with the ML model.
[0098] In some examples, an ML model may be trained prior to, or at some point following, operation of the ML model, such as ANN 400, on input data. When training the ML model, information in the form of applicable training data may be gathered or otherwise created for use in training an ANN accordingly. For example, training data may be gathered or otherwise created regarding information associated with received / transmitted signal strengths, interference, and resource usage data, as well as any other relevant data that might be useful for training a model to address one or more problems or issues in a communication system. In certain instances, all or part of the training data may originate in a user equipment (UE) or other device in a wireless communication system, or one or more network entities, or aggregated from multiple sources (such as a UE and a network entity / entities, one or more other UEs, the Internet, or the like) . For example, wireless network architectures, such as self-organizing networks (SON) or mobile drive test (MDT) networks, may be adapted to support the collection of data for ML model applications. In another example, training data may be generated or collected online, offline, or both online and offline by a UE, network entity, or other device (s) , and all or part of such training data may be transferred or shared (in real or near-real time) , such as through store and forward functions or the like.
[0099] Offline training may refer to creating and using a static training dataset, such as in a batched manner, whereas online training may refer to the real-time collection and use of training data. For example, an ML model at a network device (such as a UE) may be trained or fine-tuned using online or offline training. For offline training, data collection and training can occur in an offline manner at the network side (such as at a base station or other network entity) or at the UE side. For online training, the training of a UE-side ML model may be performed locally at the UE or by a server device (such as a server hosted by a UE vendor) in a real-time or near-real-time manner based on data provided to the server device from the UE. In certain instances, all or part of the training data may be shared within a wireless communication system or even shared (or obtained from) outside of the wireless communication system.
[0100] Once an ANN has been configured by setting parameters, including weights and biases, from training data, the ANN’s performance may be evaluated. In some scenarios, evaluation / verification tests may use a validation dataset, which may include data not in the training data, to compare the model’s performance to baseline or other benchmark information. The ANN configuration may be further refined, for example, by changing its architecture, retraining it on the data, or using different optimization techniques, etc.
[0101] As part of a training process, parameters affecting the functioning of the artificial neurons and layers may be adjusted. For example, backpropagation techniques may be used to train an ANN by iteratively adjusting weights or biases of certain artificial neurons associated with errors between a predicted output of the model and a desired output that may be known or otherwise deemed acceptable. Backpropagation may include a forward pass, a loss function, a backward pass, and a parameter update that may be performed in training iteration. The process may be repeated for a certain number of iterations for each set of training data until the weights of the artificial neurons / layers are adequately tuned.
[0102] Backpropagation techniques associated with a loss function may measure how well a model is able to predict a desired output for a given input. An optimization algorithm may be used during a training process to adjust weights and biases as needed to reduce or minimize the loss function, which should improve the performance of the model. There are a variety of optimization algorithms that may be used along with backpropagation techniques or other training techniques. Some initial examples include a gradient descent based optimization algorithm and a stochastic gradient descent based optimization algorithm. A stochastic gradient descent technique may be used to adjust weights / biases in order to minimize or otherwise reduce a loss function. A mini-batch gradient descent technique, which is a variant of gradient descent, may involve updating weights / biases using a small batch of training data rather than the entire dataset. A momentum technique may accelerate an optimization process by adding a momentum term to update or otherwise affect certain weights / biases.
[0103] An adaptive learning rate technique may adjust the learning rate of an optimization algorithm associated with one or more characteristics of the training data. A batch normalization technique may be used to normalize inputs to a model in order to stabilize a training process and potentially improve the performance of the model. A “dropout” technique may be used to randomly drop out some of the artificial neurons from a model during a training process, for example, in order to reduce overfitting and potentially improve the generalization of the model. An “early stopping” technique may be used to stop an ongoing training process early, such as when a performance of the model using a validation dataset starts to degrade.
[0104] Another example technique includes data augmentation to generate additional training data by applying transformations to all or part of the training information. A transfer learning technique may be used which involves using a pre-trained model as a starting point for training a new model, which may be useful when training data is limited or when there are multiple tasks that are related to each other. A multi-task learning technique may be used which involves training a model to perform multiple tasks simultaneously to potentially improve the performance of the model on one or more of the tasks. Hyperparameters or the like may be input and applied during a training process in certain instances.
[0105] Another example technique that may be useful with regard to an ANN is a “pruning” technique. A pruning technique, which may be performed during a training process or after a model has been trained, involves the removal of unnecessary or less necessary, or possibly redundant features from a model. In certain instances, a pruning technique may reduce the complexity of a model or improve the efficiency of a model without undermining the intended performance of the model.
[0106] Pruning techniques may be particularly useful in the context of wireless communication, where the available resources (such as power and bandwidth) may be limited. Some example pruning techniques include a weight pruning technique, a neuron pruning technique, a layer pruning technique, a structural pruning technique, and a dynamic pruning technique. Pruning techniques may, for example, reduce the amount of data corresponding to a model that may need to be transmitted or stored. Weight pruning techniques may involve removing some of the weights from a model. Neuron pruning techniques may involve removing some neurons from a model. Layer pruning techniques may involve removing some layers from a model. Structural pruning techniques may involve removing some connections between neurons in a model. Dynamic pruning techniques may involve adapting a pruning strategy of a model associated with one or more characteristics of the data or the environment. For example, in certain wireless communication devices, a dynamic pruning technique may more aggressively prune a model for use in a low-power or low-bandwidth environment and less aggressively prune the model for use in a high-power or high-bandwidth environment. In certain example implementations, pruning techniques may also be applied to training data, for example, to remove outliers. In some implementations, pre-processing techniques directed to all or part of a training dataset may improve model performance or promote faster convergence of a model. For example, training data may be pre-processed to change or remove unnecessary data, extraneous data, incorrect data, or otherwise identifiable data. Such pre-processed training data may, for example, lead to a reduction in potential overfitting or otherwise improve the performance of the trained model.
[0107] One or more of the example training techniques presented above may be employed as part of a training process. Some example training processes that may be used to train an ANN include supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning techniques. With supervised learning, a model is trained on a labeled training dataset, wherein the input data is accompanied by a correct or otherwise acceptable output. With unsupervised learning, a model is trained on an unlabeled training dataset, such that the model will need to learn to identify patterns and relationships in the data without the explicit guidance of a labeled training dataset. With semi-supervised learning, a model is trained using some combination of supervised and unsupervised learning processes, for example, when the amount of labeled data is somewhat limited. With reinforcement learning, a model may learn from interactions with its operation / environment, such as in the form of feedback akin to rewards or penalties. Reinforcement learning may be particularly beneficial when used to improve or attempt to optimize the behavior of a model deployed in a dynamically changing environment, such as a wireless communication network.
[0108] Distributed, shared, or collaborative learning techniques may be used for the training process. For example, techniques such as federated learning may be used to decentralize the training process and rely on multiple devices, network entities, or organizations for training various versions or copies of an ML model without relying on a centralized training mechanism. Federated learning may be particularly useful in scenarios where data is sensitive or subject to privacy constraints, or where it is impractical, inefficient, or expensive to centralize data. In the context of wireless communication, for example, federated learning may be used to improve performance by allowing an ANN to be trained on data collected from a wide range of devices and environments. For example, an ANN may be trained on data collected from a large number of wireless devices in a network, such as distributed wireless communication nodes, smartphones, or internet-of-things (IoT) devices, to improve the network's performance and efficiency. With federated learning, a user equipment (UE) or other device may receive a copy of all or part of a global or shared model and perform local training on the local model using locally available training data. The UE may provide updated information regarding the locally trained model to one or more other devices (such as a network entity or a server) , where the updates from other-like devices (such as other UEs) may be aggregated and used to provide an update to the global or shared model. A federated learning process may be repeated iteratively until all or part of a model obtains a satisfactory level of performance. Federated learning may enable devices to protect the privacy and security of local data, while supporting collaboration regarding training and updating of all or part of a shared model.
[0109] In some implementations, one or more devices or services may support processes relating to an ML model’s usage, maintenance, activation, reporting, or the like. In certain instances, all or part of a dataset or model may be shared across multiple devices to provide or otherwise augment or improve processing. In some examples, signaling mechanisms may be utilized at various nodes of wireless networks to signal the capabilities for performing specific functions related to ML models, support for specific ML models, capabilities for gathering, creating, and transmitting training data, or other ML related capabilities. ML models in wireless communication systems may, for example, be employed to support decisions or improve performance relating to wireless resource allocation or selection, wireless channel condition estimation, interference mitigation, beam management, positioning accuracy, energy savings, or modulation or coding schemes, etc. In some implementations, model deployment may occur jointly or separately at various network levels, such as a UE, a network entity such as a base station, or a disaggregated network entity such as a central unit (CU) , a distributed unit (DU) , a radio unit (RU) , or the like.
[0110] FIG. 5 is an illustrative block diagram of an example ML architecture 500 that may be used for wireless communications in any of the various implementations, processes, environments, networks, or use cases listed above. As illustrated, architecture 500 includes multiple logical entities, such as model training host 502, model inference host 504, data source (s) 506, and agent 508. Model inference host 504 is configured to run an ML model based on inference data 512 provided by data source (s) 506. Model inference host 504 may produce output 514, which may include a prediction or inference, such as a discrete or continuous value based on inference data 512, which may then be provided as input to the agent 508.
[0111] Agent 508 may represent an element or an entity of a wireless communication system including, for example, a radio access network (RAN) , a wireless local area network, a device-to-device (D2D) communications system, etc. As an example, agent 508 may be a user equipment (such as UE 104, referring to FIG. 1, for example) , a base station (such as base station 102, referring to FIG. 1, for example) , or a disaggregated network entity (such as a CU 110, DU 130, or RU 140 in FIG. 1) , an access point, a wireless station, a RAN intelligent controller (RIC) in a cloud-based RAN, among some examples. Additionally, agent 508 may also be a type of agent that depends on the type of tasks performed by model inference host 504, the type of inference data 512 provided to model inference host 504, or the type of output 514 produced by model inference host 504. As an example, the input may be measurements associated with a set of resources (e.g., Set-B beams / resources) , and the output may include a set of predictions for a different set of resources (e.g., Set-A beams / resources) . A base station or UE may then select a beam for use in transmission and / or reception based on the beam predictions for the Set-A beams output from the AI / ML model.
[0112] Agent 508 may perform one or more actions associated with receiving output 514 from model inference host 504, e.g., selection, use, and / or reporting regarding the predictions made for the different set of resources (e.g., Set-A beams / resources) . Agent 508 may indicate the one or more actions performed to at least one subject of action 510. In some cases, agent 508 and the subject of action 510 are the same entity.
[0113] Data can be collected from data sources 506, and may be used as training data 516 for training an ML model, or as inference data 512 for feeding an ML model inference operation. Data sources 506 may collect data from various subject of action 510 entities (such as the UE or the network entity) and provide the collected data to a model training host 502 for ML model training. Aspects presented herein improve spatial and / or temporal parameter consistency between model training and model inference, e.g., for predictions based on prior measurements that are based at least in part on aperiodic reference signals. In some examples, if output 514 provided to agent 508 is inaccurate (or the accuracy is below an accuracy threshold) , model training host 502 may provide feedback to model inference host 504 to modify or retrain the ML model used by model inference host 504, such as via an ML model deployment update.
[0114] Model training host 502 may be deployed at the same or a different entity than that in which model inference host 504 is deployed. For example, in order to offload model training processing, which can impact the performance of model inference host 504, model training host 502 may be deployed at a model server.
[0115] In some aspects, for example, an ML model may be deployed at or on a UE (such as UE 104) for predicting channel characteristic (s) associated with a second set of resources using measurement of an aperiodic reference signal on a first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition, as presented herein. More specifically, a model inference host, such as model inference host 504 in FIG. 5, may be deployed at or on the UE for performing beam predictions for the second set of resources (e.g., Set-A beams / resources) based on measurements performed on the first set of resources (e.g., Set-B beams / resources) .
[0116] Example aspects presented herein provide methods and apparatus to ensure spatial and temporal parameter consistency between training and inference for predictions based on measurements of a set of resources (e.g., Set-B beams / resources) , particularly for those based on measurements of aperiodic CSI-RS. For example, when the first set of beams (e.g., Set-B beams) that is based on aperiodic CSI-RS is used for training AI / ML models, aspects presented herein help to ensure that corresponding reference signal (RS) configuration or parameters are used during the inference of the AI / ML models.
[0117] The AI / ML technologies can be used in wireless communication. Examples of these technologies include AI / ML frameworks for air interfaces corresponding to various target use cases, which may have different performance, complexity, and potential impact on existing wireless communication specifications. For example, AI / ML frameworks for beam management may improve beam prediction both in time and spatial domains, enable overhead and latency reduction, and enhance the accuracy of beam selection. AI / ML may be used for the finalization of representative sub use cases for characterization and baseline performance evaluations of these use cases. The AI / ML approaches for these sub use cases may be diverse enough to support various levels of collaboration between the network (e.g., a base station) and the user equipment (UE) . Additionally, the AI / ML frameworks may include the lifecycle management of AI / ML models, including stages such as model training, model deployment, model inference, model monitoring, and model updating.
[0118] In some examples, the AI / ML based beam management may include spatial-domain downlink (DL) beam prediction for a group of beams (e.g., which may be referred to as Set-A beams) based on measurement results on another group of beams (which may be referred to as Set-B beams) . The resources associated with the Set-A beams and Set-B beams may be CSI-RS resources or SSB resources, among other examples. The measurement of the Set-B beams may include a measurement of the CSI-RS or SSB received on the Set-B beams, for example. In some examples, the Set-B beams may be a subset of the Set-A beams, and the Set-B beams to be used for measurement may be obtained by sub-sampling the Set-A beams via a sub-sample pattern. In some examples, the Set-A beams and the Set-B beams may be based on different beams. For example, the Set-A beams may be a set of narrow beams, while the Set-B beams may be a set of wide beams.
[0119] For spatial-domain beam predictions, when the Set-B beams are a subset of the Set-A beams, temporal varied sub-sample patterns may be used to improve prediction accuracy and robustness. These sub-sample patterns may remain consistent in the spatial domain throughout the model’s training and inference (assuming the resource number or identifier (ID) , beam shape, and relative pointing directions are already consistent for Set-A beams across model training and inference) . Additionally, the temporal periodicity of Set-B beams may also be consistent during model inference and model training.
[0120] For example, in the spatial domain, if a model (e.g., an AI / ML model) is trained based on a sub-sample factor of 4 with equal intervals between Set-A beam IDs with temporal varied choices of starting beam IDs (meaning that the Set-B beams used for the model training are selected from the Set-A beams based on a 1-to-4 ratio) , these patterns may be replicated during the model inference. In the temporal domain, if, during the model training, the Set-A beams were measured with a periodicity (e.g., 20 ms) and the model training was also based on a periodicity (e.g., 20 ms) with temporally varied Set-B beams as inputs, then, during the model inference, the Set-B beams can be measurable with temporal variation based on the same periodicity (e.g., 20 ms) .
[0121] The periodicity relationship between model training and inference described above may apply to synchronization signal blocks (SSBs) or periodic or semi-persistent channel state information -reference signal (CSI-RS) resources. However, for network implementations that involve aperiodic CSI-RS resources rather than periodic or semi-persistent CSI-RS resources, it is difficult to ensure spatial and temporal consistency.
[0122] Example aspects presented herein provide methods and apparatus to ensure spatial and temporal parameter consistency when aperiodic CSI-RS resources are scheduled for model training and inference. For example, example aspects provide a consistency related condition for sub-sample patterns for Set-B beams when aperiodic CSI-RS resources are scheduled without a periodicity (e.g., when the aperiodic CSI-RS resources are scheduled with varying time intervals) . On the other hand, if a UE may expect that the aperiodic CSI-RS resources are scheduled based on an equivalent periodicity (e.g., when the aperiodic CSI-RS resources are scheduled following a constant time interval) , the spatial and temporal domain consistency condition for periodic or semi-persistent CSI-RS resources may be used to define the spatial and temporal parameter consistencies for aperiodic CSI-RS resources.
[0123] FIG. 6 is a diagram 600 illustrating an example of an AI / ML model training based on a sub-sample pattern between Set-A beams (e.g., for prediction) and Set-B beams (for measurement as input for the prediction of the Set-A beams) in accordance with various aspects of the present disclosure. Although different sets of beams are used to illustrate the concept, the concept may also be applied to the measurement of a set of resources to make a prediction of channel characteristics for a different set of resources, and is not limited to beam predictions. In FIG. 6, the AI / ML model 650 may be implemented either on the UE side (e.g., by UE 104) or the network side (e.g., by base station 102) and may be used for beam prediction (e.g., spatial-domain beam prediction) , where the AI / ML model 650 may predict one or more best beams within Set-A beams based on measurement results on Set-B beams. The Set-A beams 602 may be RRC configured by a network entity, such as a base station. In the example in FIG. 6, the Set-A beams 602 may include multiple narrow beams (e.g., 32 narrow beams) , and each Set-A beam may correspond to a resource (e.g., CSI-RS resource or other reference signal) . The Set-A beams (e.g., Set-A beams 602) may collectively correspond to a resource set (e.g., a CSI-RS resource set) . Similarly, each Set-B beam may correspond to a resource (e.g., CSI-RS resource) , and the Set-B beams (e.g., Set-B beams 642) may collectively correspond to another resource set (e.g., a CSI-RS resource set) . In some examples, the Set-B beams (642 or 644) may be a subset of Set-A beams 602 and may be obtained by sub-sampling the Set-A beams 602. The Set-B beams (642 or 644) may be mapped to the Set-A beams 602 through a mapping pattern (also referred to as a sub-sample pattern) .
[0124] In some examples, during the training of the AI / ML model 650, the UE 104 or the base station 102 may obtain the measurement results for the training via over-the-air (OTA) data collection 604. During the OTA data collection 604, the UE 104 or the base station 102 may collect measurement results on the Set-A beams 602 on various measurement occasions, such as measurement occasion 620 (e.g., Sample #1) , measurement occasion 622 (e.g., Sample #2) , measurement occasion 624 (e.g., Sample #K) , and measurement occasion 626 (e.g., Sample #K+1) . In some examples, the measurement occasions may have a periodicity (e.g., 20 ms) in the temporal domain. The measurement results on the Set-A beams 602 may include one or more characteristics associated with the reference signals (e.g., CSI-RS) transmitted over the Set-A beams 602. For example, the measurement results may include layer 1 (L1) reference signal received power (L1-RSRP) for each of the Set-A beams 602, or L1 signal-to-interference-plus-noise ratio (L1-SINR) for each of the Set-A beams 602. For example, as shown in FIG. 6, at measurement occasion 624 (e.g., Sample #K) , the UE 104 or the base station 102 may measure the L1-RSRPs for multiple CSI-RSs (e.g., 32 CSI-RSs) respectively associated with the Set-A beams 602. These CSI-RSs may correspond to a range of CSI-RS resource indicators (CRIs 630) from, for example, 1 to 32.
[0125] The training of the AI / ML model 650 may be performed, in some examples, in an offline manner (e.g., not during active beam prediction) based on the measurement results obtained at OTA data collection 604. For example, as shown in FIG. 6, in the offline model training 606, the L1-RSRPs of sub-sampled CSI-RSs from the Set-A beams 602 may be used as the input features 660 to train the AI / ML model 650, and the output features 670 of the AI / ML model 650 may be compared with the corresponding L1-RSRP with CRI=n (n = 1, …, 32) collected at OTA data collection 604, which may serve at the ground truth of the output features of the AI / ML model 650 for the training purposes. In some examples, the sub-sampled CSI-RSs for the input features 660 may be obtained by applying a fixed sub-sample pattern on Set-A beams 602. For example, using a fixed sub-sample pattern, the Set-B beams 642 (e.g., 642-a) sampled from the Set-A beams 602 at the first measurement occasion (e.g., 620) may be the same (e.g., having the same beam shape, including the same pointing direction and beam width) as the Set-B beams 642 (e.g., 642-b) sampled from the Set-A beams 602 at the second measurement occasion (e.g., 622) . In some examples, the sub-sampled CSI-RSs for the input features 660 may be obtained by applying variable sub-sample patterns on Set-A beams 602. In these cases, the sub-sample pattern information may also be provided to the AI / ML model 650. For example, with variable sub-sample patterns applied on the Set-A beams 602, the Set-B beams 644 (e.g., 644-a1) sampled from the Set-A beams 602 at the first measurement occasion (e.g., 620) may be different (e.g., having different beam shapes, such as different the pointing directions or beam widths) from the Set-B beams 644 (e.g., 644-b1) sampled from the Set-A beams 602 at the second measurement occasion (e.g., 622) .
[0126] FIG. 7 is a diagram 700 illustrating an example of an AI / ML model inference based on a sub-sample pattern between Set-A and Set-B beams in accordance with various aspects of the present disclosure. In FIG. 7, the trained AI / ML model 650 may be used to predict one or more best beams (e.g., based on characteristics such as L1-RSRP or L1-SINR) of the Set-A beams 702 based on measurement results on Set-B beams (e.g., Set-B beams 742 or 744) . In the example in FIG. 7, the Set-A beams 702 may include multiple narrow beams (e.g., 32 narrow beams) , and the Set-B beams (e.g., 742 or 744) may be a subset of Set-A beams 702 and may be obtained by sub sampling the Set-A beams 702. The Set-B beams (e.g., 742 or 744) may be mapped to the Set-A beams 702 through a mapping pattern (or a sub-sample pattern) .
[0127] As shown in FIG. 7, during the model inference of the AI / ML model 650, the UE 104 (or the base station 102) may perform the measurement 704 on the Set-B beams (e.g., 742 or 744) on various prediction cycles, such as prediction cycles 720 (e.g., Sample #1) , prediction cycles 722 (e.g., Sample #2) , prediction cycles 724 (e.g., Sample #K) , and prediction cycles 726 (e.g., Sample #K+1) . In some examples, the prediction cycles may have a periodicity (e.g., 20 ms) in the temporal domain. The measurement results on the Set-B beams (e.g., 742 or 744) may include one or more characteristics associated with the reference signals (e.g., CSI-RS) transmitted over the Set-B beams (e.g., 742 or 744) . For example, the measurement results may include L1-RSRP or L1-SINR for each of the Set-B beams (e.g., 742 or 744) . For example, as shown in FIG. 7, at prediction cycles 724 (e.g., Sample #K) , the UE 104 (or the base station 102) may measure the L1-RSRPs for multiple CSI-RSs (e.g., 8 CSI-RSs) associated with the Set-B beams (e.g., 742 or 744) . These CSI-RSs may correspond to a range of CSI-RS resource indicators (CRIs 730) from, for example 1 to 8.
[0128] In some examples, during the model inference, the Set-B beams (e.g., 742 or 744) may be sampled from the Set-A beams 702 based on different mapping patterns (or sub-sample patterns) . In some examples, the beam shapes of the Set-B beams 742 may be strictly fixed across prediction cycles. For example, the Set-B beams (e.g., 742-a, 742-b, 742-c, and 742-d) sampled from the Set-A beams at different prediction cycles (e.g., 720, 722, 724, 726) may have the same beam shapes. In some examples, the beam shapes of the Set-B beams 744 may vary across prediction cycles. For example, the beam shapes of the Set-B beams 744 (e.g., 744-a1, 744-b1, 744-c1, and 744-d1) may vary at different prediction cycles (e.g., 720, 722, 724, 726) . In these cases, the sub-sample pattern information may also be provided to the AI / ML model 650.
[0129] The measurement results obtained at 704 may be used as the input features 760 to the AI / ML model 650, and the AI / ML model 650 may predict one or more characteristics (e.g., L1-RSRP or L1-SINR) for the Set-A beams 702. For example, as shown in FIG. 7, in the online model inference 706, the L1-RSRPs of the sub-sampled CSI-RSs (e.g., based on Set-B beams 742 or 744) may be used as the input features 760 to the AI / ML model 650, and the output features 770 of the AI / ML model 650 may include the predicted L1-RSRPs on the Set-A beams (e.g., L1-RSRP with CRI=n (n = 1, …, 32) ) . In some examples, the sub-sampled CSI-RSs may be obtained by applying a fixed sub-sample pattern on the Set-A beams 702. For example, by applying a fixed sub-sample pattern on the Set-A beams 702, the Set-B beams 742 (e.g., 742-a, 742-b, 742-c, and 742-d) sampled from the Set-A beams at different prediction cycles (e.g., 720, 722, 724, 726) may have the same beam shapes. In some examples, the beam shapes of the Set-B beams 744 may vary across prediction cycles. For example, the beam shapes of the Set-B beams 744 (e.g., 744-a1, 744-b1, 744-c1, and 744-d1) may vary at different prediction cycles (e.g., 720, 722, 724, 726) .
[0130] In some examples, the Set-A beams may be maintained consistent across model training and model inference. For example, the number of CSI-RS resources configured for Set-A beams may be maintained consistent during model training (e.g., at 606) and model inference (e.g., at 706) . The beam shape, including the pointing direction and beam width of the physical beams in the two sets of CSI-RS resources, may not vary beyond predefined tolerances. Furthermore, the UE may maintain the order consistency across the model training and inference. For example, the UE may label the L1-RSRP in relation to the CRI=n for Set-A beams as corresponding to the nth output feature during model training. Then, in the model inference, the UE may use the value from the nth output feature to derive prediction results in relation to CRI=n for Set-A beams.
[0131] Additionally, the sub-sample patterns used in the model training and inference may be based on a consistency condition. In some examples, the consistency condition may include that both spatial and temporal sub-sample patterns may be identical during training and inference. In some examples, the consistency condition may include maintaining a consistent sub-sample factor, which may include either fixed or variable Set-A beam-ID sampling intervals per prediction cycle across model training and inference. In some examples, the consistency condition may include the likelihood of a Set-A beam with respect to CRI=m being sub-sampled as a Set-B beam during model training is similar to the likelihood of a Set-A beam with respect to CRI=m being sub-sampled as a Set-B beam across prediction cycles during model inference, with any difference falling within a predefined tolerance.
[0132] FIG. 8 is diagram 800 illustrating an example of an AI / ML model training for temporal domain beam prediction in accordance with various aspects of the present disclosure. In FIG. 8, historical Set-B beam measurements (e.g., measured L1-RSRPs of the historical Set-B beams) collected within a time duration 848 (e.g., 200 ms) may be used as the input features 860 to train the AI / ML model 650. The historical Set-B beam measurements used in the model training may have a periodicity (e.g., 40 ms) , which may be equal to greater than the periodicity of available Set-B beam measurements in the collected data (e.g., 20 ms) . During the model training, the output features 870 of the AI / ML model 650 may include the measurements (e.g., L1-RSRPs) of Set-A beams 802 on occasions later than the latest Set-B beam measurement occasion (t0) . For example, the output features 870 may include L1-RSRPs of Set-A beams 802 on occasions 40 ms, 80 ms, 120 ms, and 160 ms later than t0.
[0133] FIG. 9 is diagram 900 illustrating an example of an AI / ML model inference for temporal domain beam prediction in accordance with various aspects of the present disclosure. In FIG. 9, based on historical measurement results (e.g., L1-RSRPs) on Set-B beams 942 within a time duration 948, the trained AI / ML model 650 may, based on the output features 970 of the AI / ML model 650, predict one or more best beams (e.g., based on characteristics such as L1-RSRP or L1-SINR) within Set-A beams 902 on occasion later than the latest Set-B measurement occasion (t0) . During the model inference, the time duration 948 for the historical Set-B beam measurements used as the input features 960 may be equal to or longer than the time duration 848 of the historical Set-B beam measurements used as the input features 840 for model training. For example, if the time duration 848 of the historical Set-B beam measurements used for model training is 200 ms, the time duration 948 for the historical Set-B beam measurements for model inference may be 250 ms. During the model inference, the periodicity of the historical Set-B beam measurements may be equal to or less than the periodicity of the historical Set-B beam measurements used for model training. For example, when the historical Set-B beam measurements used for model training have a periodicity of 40 ms, the historical Set-B beam measurements used for model interference may have a periodicity of 40 ms or less (e.g., 20 ms) . During model inference, the AI / ML model 650 may predict characteristics (e.g., L1-RSRP) of the Set-A beams on various occasions later than the latest Set-B beam measurement occasion (t0) . In some aspects, the time difference between the prediction time for Set-A beams 902 and the latest Set-B beam measurement occasion (t0) during model inference may not exceed the maximum time difference between the occasions of the Set-A beams used as the output features 870 and latest Set-B beam measurement occasion (t0) during model training. For example, if the maximum time difference between the occasions of the Set-A beams used as the output features 870 and latest Set-B beam measurement occasion (t0) during model training is 160 ms, then in model inference (as shown in FIG. 9) , the maximum time difference between the prediction time for Set-A beams 902 and the latest Set-B beam measurement occasion (t0) gap may be no more than 160 ms. For example, the time difference for model inference may be 40 ms or 120 ms.
[0134] In some aspects, the spatial and temporal parameters may be maintained consistent between model training and inference for aperiodic CSI-RS resources in beam prediction (e.g., narrow-to-narrow beam prediction) . In some examples, the UE may be indicated by the network (e.g., a gNB) to predict and report channel characteristics, which may include L1-RSRP or L1-SINR or the K best resources (top-K-resources) in terms of L1-RSRP / SINR, with respect to Set-A beams. In some examples, the Set-A beams may be configured based on a synchronization signal blocks (SSB) resource set, a CSI-RS resource set, or a virtual resource set. The predictions may be based on measurements of Set-B beams, which may be scheduled based on aperiodic CSI-RS resources.
[0135] In some examples, each Set-B beam may be associated with a Set-A beam. In some examples, the association between the Set-B beam and the corresponding Set-A beam may be defined as the Set-B beam and the Set-A beam having the same (e.g., identical) beam shapes (including pointing in the same direction and having the same beam width) . The association of the Set-B beam to the corresponding Set-A beam may be based on a “Set-B beam sub-sample pattern. ” In some examples, the prediction process may be based on an AI / ML model operated at the UE side, and the training of the AI / ML mode may use data collected via aperiodic CSI-RS resources associated with Set-A beams. In these scenarios, one Set-B beam sub-sample pattern may be used during the inference of the AI / ML model, which may be referred to as “the first mapping pattern, ” while another Set-B beam sub-sample pattern may be used during the training of the AI / ML model, which may be referred to as “the second mapping pattern. ”
[0136] In some aspects, the Set-B beam sub-sample pattern used during the model inference (e.g., the first mapping pattern) and the Set-B beam sub-sample pattern used during the model training (e.g., the second mapping pattern) may meet at least one of a spatial domain consistency condition or a temporal domain consistency condition. For example, if the periodicities may not be guaranteed for the aperiodic CSI-RS resources (e.g., when the aperiodic CSI-RS resources are scheduled with varying time intervals) , the Set-B beam sub-sample pattern used during the model inference (e.g., the first mapping pattern) and that used during the model training (e.g., the second mapping pattern) may be based on meeting the spatial domain consistency condition, while the temporal domain consistency condition may not apply between the first and second mapping patterns.
[0137] FIG. 10A is a diagram 1000 illustrating an example of AI / ML model training and inference when the periodicities may not be guaranteed for aperiodic CSI-RS resources, in accordance with various aspects of the present disclosure. In FIG. 10A, the periodicities may not be guaranteed for aperiodic CSI-RS resources. For example, the adjacent occasions of the Set-B beams (e.g., Set-B beams 1010, 1020, 1030, and 1040) for model training or inference may not have a fixed periodicity (e.g., p01<p02<p03) . In these scenarios, the Set-B sub-sample patterns (e.g., the sub-sample pattern mapping Set-B beams 1010 to Set-A beams 1012) used during the model inference and model training may meet the spatial domain consistency condition and the spatial domain consistency condition may not apply.
[0138] On the other hand, in the scenarios where an “equivalent” periodicity may be achieved by the aperiodic CSI-RS resources during model training and inference (e.g., when the aperiodic CSI-RS resources are scheduled following a constant time interval) , the Set-B beam sub-sample pattern used during the model inference (e.g., the first mapping pattern) and that used during the model training (e.g., the second mapping pattern) may meet the spatial domain consistency condition and the temporal domain consistency condition.
[0139] FIG. 10B is a diagram 1050 illustrating an example of AI / ML model training and inference where an equivalent periodicity can be achievable by the aperiodic CSI-RS resources in accordance with various aspects of the present disclosure. In FIG. 10B, the adjacent occasions of the Set-B beams (e.g., Set-B beams 1060, 1070, 1080, and 1090) for model training or inference may have an equivalent periodicity (e.g., p1) . In these scenarios. The Set-B sub-sample patterns (e.g., the sub-sample pattern mapping Set-B beams 1060 to Set-A beams 1062) used during the model inference and model training may meet the spatial domain consistency condition and the spatial domain consistency condition.
[0140] The spatial domain consistency condition for the first and second mapping patterns may include various restrictions or conditions between the mapping patterns used in model training and inference. In some examples, the spatial domain consistency condition may indicate a strict spatial and temporal domain consistency between the first and the second mapping patterns. For example, the spatial domain consistency condition may include the first mapping pattern and the second mapping pattern being based on the same mapping pattern. For example, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The first mapping pattern may be the same as the second mapping pattern. In this scenario, a temporal variable “Set-B beam sub-sample pattern, ” such as the sub-sample patterns that map different Set-B beams (e.g., 744-a1, 744-b1) to Set-A beams 702 at different prediction cycles (e.g., 720, 722) , may not be allowed, as aperiodic CSI-RS resources may be activated at any time. Hence, at any given prediction instance during model training, only a single Set-B beam sub-sample pattern (i.e., the second mapping pattern) is allowed, and similarly, only one Set-B beam sub-sample pattern (e.g., the first mapping pattern) is allowed at any prediction instance during model inference. The patterns during training and inference (e.g., the first and second mapping patterns) may be identical.
[0141] In some examples, the spatial domain consistency condition may indicate a sub-sample factor consistency condition for the first and second mapping patterns. For example, the spatial domain consistency condition may indicate that Set-B beams sub-sampled at a particular prediction instance may be selected from Set-A beams with a fixed resource ID interval, as defined in the resource set configured for Set-A beams. The length of this interval may remain consistent throughout the model training and inference phases. For example, a first interval factor associated with the first mapping pattern may specify a first interval between the first set of resources IDs associated with a set of resources for the model inference, and a second interval factor associated with the second mapping pattern may specify a second interval between the second set of resource IDs associated with a set of resources for the model training. The spatial domain consistency condition may include that the first interval factor is equal to the second interval factor. For example, referring to FIG. 6 and FIG. 7, the first interval factor associated with the first mapping pattern may specify a first interval between the first set of resources IDs associated with the set of resources (e.g., resources associated with Set-B beams 744) for the model inference 706, and a second interval factor associated with the second mapping pattern may specify a second interval between the second set of resource IDs associated with the set of resources (e.g., resources associated with Set-B beams 644) for the model training 606. The first interval may be the interval between Set-B beams 744-a1 and 744-a2, Set-B beams 744-b1 and 744-b2, Set-B beams 744-c1 and 744-c2, and Set-B beams 744-d1 and 744-d2. The second interval may be the interval between Set-B beams 644-a1 and 644-a2, Set-B beams 644-b1 and 644-b2. The spatial domain consistency condition may include that the first interval factor being equal to the second interval factor (e.g., equal to 4) . Furthermore, the selection of the starting resource ID at various prediction instances may follow a specific distribution, such as uniform or non-uniform distribution. The distribution of the starting resource ID, like the interval length, may also be consistent across the model training and inference. In some examples, the distribution of the starting resource ID may be uniformly distributed across the model training and inference.
[0142] In some examples, the spatial domain consistency condition may indicate a statistical consistency condition for the first and second mapping patterns. For example, one beam in the Set-A beams (e.g., the mth Set-A beam) , as defined by the associated resource ID in the resource set of Set-A beams, may have a likelihood (denoted as the first probability) of being sub-sampled as a Set-B beam across different prediction cycles during the model inference. This likelihood may be comparable to the likelihood (donated as the second probability) of the same Set-A beam being sub-sampled as a Set-B beam during the model training. For example, the spatial domain consistency condition may include the difference between the first probability and the second probability being less than a predefined tolerance level (e.g., less than a threshold) . For example, referring to FIG. 6 and FIG. 7, the spatial domain consistency condition may include the difference between the probability of one beam (e.g., 744-a1) in the Set-A beams 702 of being sub-sampled as a Set-B beam during model inference (at 706) and the probability of one beam (e.g., 644-a1) in the Set-A beams 602 of being sub-sampled as a Set-B beam during model training (at 606) is less than a threshold. This spatial domain consistency condition may apply to the scenarios when aperiodic CSI-RS resources are used for both model training and inference.
[0143] In some aspects, the “equivalent” periodicities may be achieved (or guaranteed) by the aperiodic CSI-RS resources during model training and inference (e.g., based on the AI / ML functionality and model life cycle management (LCM) frameworks) . For example, as shown in FIG. 10B, the aperiodic CSI-RS resources may be scheduled with an equivalent periodicity of p1. In these scenarios, the Set-B beam sub-sample patterns used during the model inference (e.g., the first mapping pattern) and the model training (e.g., the second mapping pattern) may meet the spatial domain consistency condition and the temporal domain consistency condition. In some examples, the model inference may be conducted with respect to AI / ML functionality and model LCM. For example, the network may signal to the UE that certain scheduled aperiodic CSI-RS resources may be considered as Set-B beams for predicting certain Set-A beams’ characteristics, such as L1-RSRPs or L1-SINR. Additionally, the associated AI / ML functionality and model ID may also be signaled to the UE.
[0144] When the model training is conducted with respect to AI / ML functionality and model LCM, the network may signal to the UE that certain scheduled aperiodic CSI-RS resources may be measured by the UE for data collection for training a certain model ID associated with a certain AI / ML functionality (e.g., narrow-to-narrow beam prediction) .
[0145] In some examples, during or after data collection for model training, the network may signal the shortest periodicity (e.g., p1 in FIG. 10B) achieved by the aperiodic CSI- RS resources during the data collection for the model training. The network may also signal other periodicities that are integer multiples of the shortest periodicity, which the UE should consider for model training in terms of Set-B beam sub-sampling. During model inference associated with the AI / ML functionality and model ID, the network may guarantee at least one of the periodicities (e.g., p1 in FIG. 10B) signaled when transmitting the aperiodic CSI-RS resources as Set-B beams, and the network may signal the guaranteed periodicities (e.g., p1 in FIG. 10B) . In some examples, the network may signal the periodicities based on the AI / ML functionality and model LCM frameworks.
[0146] In some examples, given that the same AI / ML functionality and associated model ID are used during both model training and inference, the network may not signal the shortest periodicity (or the multiples of the shortest periodicity) for the UE during the data collection for model training (e.g., when interoperability of device testing (IoDT) was considered for data collection) . The network, however, may still guarantee at least one of the shortest periodicity or the integer multiples of the shortest periodicity during the model inference, and the network may also signal the guaranteed periodicity to the UE.
[0147] FIG. 11 is a call flow diagram 1100 illustrating a method of wireless communication in accordance with various aspects of this present disclosure. Various aspects are described in connection with a UE 1102 and a base station 1104. The aspects may be performed by the UE 1102 or the base station 1104 in aggregation and / or by one or more components of a base station 1104 (e.g., a CU 110, a DU 130, and / or an RU 140) .
[0148] As shown in FIG. 11, at 1106, a UE 1102 may receive, from a base station 1104, a first indication to predict, based on the measurement of an aperiodic reference signal on the first set of resources, one or more channel characteristics associated with the second set of resources. In some examples, the prediction of the one or more channel characteristics may be based on an AI / ML model 1150 or 650. The AI / ML model may be based on any of the aspects described in connection with FIGs. 4-10B, for example. The first set of resources may be the resources associated with Set-B beams (e.g., Set-B beams 742) for model inference (e.g., 706) , and the second set of resources may be the resources associated with Set-A beams (e.g., Set-A beams 702) for the prediction.
[0149] At 1108, the UE 1102 may receive, from the base station 1104, a second indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model 1150. The training of the AI / ML model 1150 may be associated with the AI / ML functionality or the model identifier. In some examples, the training of the AI / ML model 1150 may be based on data collected via aperiodic training reference signal resources (e.g., aperiodic CSI-RS resources) associated with the second set of resources (e.g., resources associated with Set-A beams 702) .
[0150] At 1110, the UE 1102 may receive, from the base station 1104, a periodicity indicator. In some example, the periodicity indicator may be indicative of a periodicity of the aperiodic training reference signal resources for training the AI / ML model 1150. In some examples, the periodicity indicator may include one or more of the shortest periodicity for the aperiodic training reference signal resources or an additional periodicity for the aperiodic training reference signal resources. The additional periodicity may be an integer multiple of the shortest periodicity, such as twice, three times, or four times the length of the shortest periodicity. For example, referring to FIG. 10B, the shortest periodicity may be the periodicity p1, and the additional periodicity may include one or more integer multiples of the shortest periodicity (e.g., twice, three time, or four times of p1) .
[0151] At 1111, the base station 1104 may provide (e.g., transmit) an aperiodic reference signal to the UE 1102. For example, the aperiodic reference signal may be an aperiodic CSI-RS, which may be transmitted on the first set of resources (e.g., resources associated with Set-B beams 742) .
[0152] At 1114, the UE 1102 may measure the aperiodic reference signal received on a first set of resources. For example, the first set of resources may be resources associated with Set-B beams (e.g., 742 or 744) for model inference (e.g., 706) using the AI / ML model 1150. For example, referring to FIG. 7, the UE may measure an aperiodic reference signal (e.g., an aperiodic CSI-RS) received on a first set of resources (e.g., resources associated with Set-B beams 744) .
[0153] At 1116, the base station 1104 may indicate for the UE 1102 to predict one or more channel characteristics associated with a second set of resources. For example, referring to FIG. 7, the second set of resources may include the resources associated with Set-A beams 702.
[0154] At 1118, the UE 1102 may predict one or more channel characteristics associated with the second set of resources using measurement of the aperiodic reference signal on the first set of resources. The second set of resources may be the resources associated with Set-A beams (e.g., Set-A beams 702) for model inference (e.g., at 706) . The prediction may be based on the AI / ML model 1150 or 650. The first set of resources (e.g., resources associated with Set-B beams 742) may be mapped to (or sub-sampled from) the second set of resources (e.g., resources associated with Set-A beams 702) via a first mapping pattern. During the training of the AI / ML model 1150, a third set of resources (e.g., resources associated with Set-B beams 642 for model training 606) may be mapped to (or sub-sampled from) the second set of resources (resources associated with Set-A beams 702) via a second mapping pattern. The first mapping pattern and the second mapping pattern may meet a spatial domain consistency condition or a temporal domain consistency condition, or both.
[0155] In some examples, the spatial domain consistency condition may include the first mapping pattern and the second mapping pattern being based on the same mapping pattern. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The spatial domain consistency condition may include the first mapping pattern and the second mapping pattern being based on the same mapping pattern.
[0156] In some examples, the spatial domain consistency condition may include a first interval factor associated with the first mapping pattern being equal to a second interval factor associated with the second mapping pattern. The first interval factor may specify a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor may specify a second interval between a second set of identifiers associated with the third set of resources. The first distribution of a starting identifier of the first set of resources may be consistent with the second distribution of the starting identifier of the third set of resources. For example, referring to FIG. 6 and FIG. 7, the first interval factor associated with the first mapping pattern may specify a first interval between the first set of resources IDs associated with the first set of resources (e.g., resources associated with Set-B beams 744) for the model inference (e.g., at 706) , and a second interval factor associated with the second mapping pattern may specify a second interval between the second set of resource IDs associated with the third set of resources (e.g., resources associated with Set-B beams 644) for the model training (e.g., at 606) . The first interval may be, for example, the interval between Set-B beams 744-a1 and 744-a2, Set-B beams 744-b1 and 744-b2, Set-B beams 744-c1 and 744-c2, and Set-B beams 744-d1 and 744- d2, The second interval may be, for example, the interval between Set-B beams 644-a1 and 644-a2, Set-B beams 644-b1 and 644-b2. The spatial domain consistency condition may include that the first interval factor is equal to the second interval factor (e.g., equal to 4) .
[0157] In some examples, each resource in the second set of resources (e.g., resources associated with Set-A beams 702) may have a first probability of being selected to the first set of resources (e.g., resources associated with Set-B beams 742 or 744 for model inference) and have a second respective probability of being selected to the third set of resources (e.g., resources associated with Set-B beams 642 or 644 for model training) . The spatial domain consistency condition may include the difference between the first respective probability and the second respective probability for each resource in the second set of resources (e.g., resources associated with Set-A beams 702) being less than a threshold. For example, referring to FIG. 6 and FIG. 7, the spatial domain consistency condition may include the difference between the probability of one beam (e.g., 744-a1) in the Set-A beams 702 of being sub-sampled as a Set-B beam during model inference (at 706) and the probability of one beam (e.g., 644-a1) in the Set-A beams 602 of being sub-sampled as a Set-B beam during model training (at 606) being less than a threshold.
[0158] In some examples, the temporal domain consistency condition may include a first periodicity associated with inference aperiodic CSI-RS resources being consistent with a second periodicity associated with the aperiodic training reference signal resources. For example, referring to FIG. 10B, when the aperiodic CSI-RS resources are scheduled with an equivalent periodicity (e.g., p1) , the temporal domain consistency condition may include a first periodicity associated with inference aperiodic CSI-RS resources being consistent with a second periodicity associated with the aperiodic training reference signal resources. For example, both the first periodicity and the second periodicity may equal to p1.
[0159] In some examples, the predicted one or more channel characteristics associated with the second set of resources (e.g., resources associated with Set-A beams 702) may include one or more of: L1-RSRP for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , or a subset of resources (e.g., the top K resources) of the second set of resources (e.g., resources associated with Set-A beams 702) based on the L1-RSRP or the L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) .
[0160] At 1120, the base station 1104 may report, to the base station 1104, the one or more channel characteristics predicted for the second set of resources. For example, referring to FIG. 7, after the model inference 706, the UE may report one or more best beams (e.g., based on the predicted L1-RSRPs of the Set-A beams) from the Set-A beams 702 to the base station 1104.
[0161] As illustrated at 1122, the UE 1102 and the base station 1104 may exchange (e.g., transmit and / or receive) wireless communication based on the channel characteristics predicted at the UE and reported to the base station. For example, the base station 1104 and / or the UE 1102 may use a beam to transmit or receive the wireless communication, the beam selection based on the prediction. In some examples, the base station 1104 may schedule communication with the UE 1102 and / or provide control signaling to the UE 1102 based on the predicted channel characteristic (s) .
[0162] Although not illustrated, in some aspects, the UE 1102 may transmit an indication of support to the base station 1104, indicating support for AI / ML channel characteristic prediction. Based on the UE’s support, or capability, for AI / ML based channel characteristic prediction, the base station 1104 may transmit a configuration to the UE 1102 to perform the prediction and / or to report the prediction. For example, the configuration may include aspects such as those described at 1106 and / or 1108.
[0163] FIG. 12 is a flowchart 1200 illustrating methods of wireless communication at a UE in accordance with various aspects of the present disclosure. The method may be performed by a UE. The UE may be the UE 104, 350, 1102; the apparatus 1604 in the hardware implementation of FIG. 16; or the first wireless device 1802 in FIG. 18. By ensuring spatial and temporal consistency in Set-B beams sub-sample patterns between the training and inference of an AI / ML model, the methods improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization.
[0164] As shown in FIG. 12, at 1202, the UE may measure an aperiodic reference signal received on a first set of resources. FIG. 6, FIG. 7, FIG. 8, FIG. 9, FIG. 10A, FIG. 10B, and FIG. 11 illustrate various aspects of the steps in connection with flowchart 1200. For example, referring to FIG. 11, the UE 1102 may, at 1114, measure an aperiodic reference signal (e.g., an aperiodic CSI-RS) received on a first set of resources. Referring to FIG. 7 and FIG. 9, the first set of resources may be the resources for Set-B beams 742 or 744 for model inference 706. In some aspects, 1202 may be performed by the resource consistency component 198.
[0165] At 1204, the UE may predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern. The first mapping pattern may map the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training may meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. For example, referring to FIG. 11 and FIG. 7, the UE 1102 may, at 1118, predict one or more channel characteristics associated with a second set of resources (e.g., resources associated with Set-A beams 702) using measurement of the aperiodic reference signal on the first set of resources (e.g., resources associated with Set-B beams 742 or 744) based on a first mapping pattern. Referring to FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams (e.g., 742, 744) to Set-A beams 702 during model inference 706. The second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. In some aspects, 1204 may be performed by the resource consistency component 198.
[0166] At 1206, the UE may report, to a network entity, the one or more channel characteristics predicted for the second set of resources. The network entity may be a base station, or a component of a base station, in the access network of FIG. 1 or a core network component (e.g., base station 102, 310, 1104; the network entity 1602 in the hardware implementation of FIG. 16; or the second wireless device 1804 in FIG. 18) . For example, referring to FIG. 11, the UE 1102 may report, at 1120 to a network entity (base station 1104) , the one or more channel characteristics predicted for the second set of resources (e.g., resources associated with Set-A beams 702 or 902) . In some aspects, 1206 may be performed by the resource consistency component 198.
[0167] FIG. 13 is a flowchart 1300 illustrating methods of wireless communication at a UE in accordance with various aspects of the present disclosure. The method may be performed by a UE. The UE may be the UE 104, 350, 1102; the apparatus 1604 in the hardware implementation of FIG. 16; or the first wireless device 1802 in FIG. 18, for example. By ensuring spatial and temporal consistency in Set-B beams sub-sample patterns between the training and inference phases of an AI / ML model, the methods improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization.
[0168] As shown in FIG. 13, at 1308, the UE may measure an aperiodic reference signal received on a first set of resources. FIG. 6, FIG. 7, FIG. 8, FIG. 9, FIG. 10A, FIG. 10B, and FIG. 11 illustrate various aspects of the steps in connection with flowchart 1300. For example, referring to FIG. 11, the UE 1102 may, at 1114, measure an aperiodic reference signal received on a first set of resources. Referring to FIG. 7 and FIG. 9, the first set of resources may be the resources for Set-B beams 742 or 744 for model inference 706. In some aspects, 1308 may be performed by the resource consistency component 198.
[0169] At 1310, the UE may predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern. The first mapping pattern may map the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training may meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. For example, referring to FIG. 11 and FIG. 7, the UE 1102 may, at 1118, predict one or more channel characteristics associated with a second set of resources (e.g., resources associated with Set-A beams 702) using measurement of the aperiodic reference signal on the first set of resources (e.g., resources associated with Set-B beams 742 or 744) based on a first mapping pattern. Referring to FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams (e.g., 742, 744) to Set-A beams 702 during model inference 706. The second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. In some aspects, 1310 may be performed by the resource consistency component 198.
[0170] At 1312, the UE may report, to a network entity, the one or more channel characteristics predicted for the second set of resources. The network entity may be a base station, or a component of a base station, in the access network of FIG. 1 or a core network component (e.g., base station 102, 310, 1104; or the network entity 1602 in the hardware implementation of FIG. 16) . For example, referring to FIG. 11, the UE 1102 may report, at 1120 to a network entity (base station 1104) , the one or more channel characteristics predicted for the second set of resources (e.g., resources associated with Set-A beams 702, 902) . In some aspects, 1312 may be performed by the resource consistency component 198.
[0171] In some aspects, at 1302, the UE may receive, from the network entity, a first indication to predict the one or more channel characteristics associated with the second set of resources based on the measurement of the aperiodic reference signal on the first set of resources. For example, referring to FIG. 7, FIG. 9, and FIG. 11, the UE 1102 may, at 1106, receive from the network entity (base station 1104) a first indication to predict the one or more channel characteristics associated with the second set of resources (e.g., resources associated with Set-A beams 702, 902) based on the measurement of the aperiodic reference signal on the first set of resources (e.g., resources associated with Set-B beams 742, 744, 942) . In some aspects, 1302 may be performed by the resource consistency component 198.
[0172] In some aspects, the prediction of the one or more channel characteristics (at 1310) may be based on an AI / ML model, and the initial training (at 1310) may be a training of the AI / ML model. For example, referring to FIG. 6, FIG. 7, and FIG. 11, the predicting (at 1118) may be based on an AI / ML model (e.g., 1150 or 650) , and the initial training may be the training of the AI / ML model 650 at model training 606.
[0173] In some aspects, at 1304, the UE may receive, from the network entity, a second indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model. The training of the AI / ML model may be associated with the AI / ML functionality or the model identifier. For example, referring to FIG. 11, the UE 1102 may receive, at 1108 from the network entity (base station 1104) , a second indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model 1150. The training of the AI / ML model (e.g., at 606) may be associated with the AI / ML functionality or the model identifier. In some aspects, 1304 may be performed by the resource consistency component 198.
[0174] In some aspects, the training of the AI / ML model may be based on data collected via aperiodic training reference signal resources associated with the second set of resources, and the second mapping pattern may map a third set of resources to the second set of resources for the training of the AI / ML model. For example, referring to FIG. 6, the training of the AI / ML model 650 (at 606) may be based on data collected via aperiodic training reference signal resources (e.g., aperiodic CSI-RS resources) associated with the second set of resources (e.g., resources associated with Set-A beams 602) , and the second mapping pattern may map a third set of resources (e.g., resources associated with Set-B beams 642 or 644) to the second set of resources (e.g., resources associated with Set-A beams 602) for the training of the AI / ML model 650 (at 606) .
[0175] In some aspects, the one or more channel characteristics associated with the second set of resources (at 1310) may include one or more of: L1-RSRP for each resource of the second set of resources, L1-SINR for each resource of the second set of resources, or a subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources. For example, referring to FIG. 11, the one or more channel characteristics associated with the second set of resources (at 1118) may include one or more of: L1-RSRP for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , or a subset of resources of the second set of resources (e.g., resources associated with Set-A beams 702) based on the L1-RSRP or the L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) .
[0176] In some aspects, the one or more channel characteristics may be predicted for the second set of resources (at 1310) based on a configuration of one or more of: an SSB resource set, a CSI-RS resource set, or a virtual resource set. For example, referring to FIG. 7, the one or more channel characteristics may be predicted (at 706) for the second set of resources (e.g., resources associated with Set-A beams 702) based on a configuration of one or more of: an SSB resource set, a CSI-RS resource set, or a virtual resource set.
[0177] In some aspects, at 1320, the first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition.
[0178] In some aspects, at 1322, the spatial domain consistency condition may include the first mapping pattern and the second mapping pattern are based on the same mapping pattern. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The spatial domain consistency condition may include the first mapping pattern and the second mapping pattern are based on the same mapping pattern.
[0179] In some aspects, at 1324, the spatial domain consistency condition may include a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern. The first interval factor may specify a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor may specify a second interval between a second set of identifiers associated with the third set of resources. The first distribution of a starting identifier of the first set of resources may be consistent with the second distribution of the starting identifier of the third set of resources. For example, referring to FIG. 6 and FIG. 7, the first interval factor associated with the first mapping pattern may specify a first interval between the first set of identifiers associated with the first set of resources (e.g., resources associated with Set-B beams 744) for the model inference 706, and a second interval factor associated with the second mapping pattern may specify a second interval between the second set of identifiers associated with the third set of resources (e.g., resources associated with Set-B beams 644) for the model training 606. The first interval may be the interval between Set-B beams 744-a1 and 744-a2, Set-B beams 744-b1 and 744-b2, Set-B beams 744-c1 and 744-c2, and Set-B beams 744-d1 and 744-d2. The second interval may be the interval between Set-B beams 644-a1 and 644-a2, Set-B beams 644-b1 and 644-b2. The spatial domain consistency condition may include that the first interval factor is equal to the second interval factor (e.g., equal to 4) .
[0180] In some aspects, at 1326, each resource in the second set of resources may have a first respective probability of being selected to the first set of resources and a second respective probability of being selected to the third set of resources. The spatial domain consistency condition may include: the difference between the first respective probability and the second respective probability for each resource in the second set of resources is less than a threshold. For example, referring to FIG. 6 and FIG. 7, each resource in the second set of resources (e.g., resources associated with Set-A beams 702) may have a first respective probability of being selected to the first set of resources (e.g., resources associated with Set-B beams 742 or 744) and a second respective probability of being selected to the third set of resources (e.g., resources associated with Set-B beams 642 or 644) . The spatial domain consistency condition may include the difference between the probability of one beam (e.g., 744-a1) in the Set-A beams 702 of being sub-sampled as a Set-B beam during model inference (at 706) and the probability of one beam (e.g., 644-a1) in the Set-A beams 602 of being sub-sampled as a Set-B beam during model training (at 606) is less than a threshold.
[0181] In some aspects, at 1330, the first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition and the temporal domain consistency condition. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition and the temporal domain consistency condition.
[0182] In some aspects, the spatial domain consistency condition may include one of: the first mapping pattern and the second mapping pattern are based on the same mapping pattern (at 1322) , a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern (at 1324) , where the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, or, for each resource in the second set of resources, the difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold (at 1326) . For example, referring to FIG. 6 and FIG. 7, the spatial domain consistency condition may include one of: the first mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706) and the second mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606) are based on the same mapping pattern, or a first interval factor associated with the first mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706) that is equal to a second interval factor associated with the second mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606) , where the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources (e.g., resources associated with Set-B beams 742) , and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources (e.g., resources associated with Set-B beams 642) . The spatial domain consistency condition may also include, for each resource in the second set of resources (e.g., resources associated with Set-A beams 602) , the difference between a first probability of being selected to the first set of resources (e.g., resources associated with Set-B beams 742) and a second probability of being selected to the third set of resources (e.g., resources associated with Set-B beams 642) is less than a threshold.
[0183] In some aspects, at 1332, the temporal domain consistency condition may include a first periodicity associated with inference aperiodic CSI-RS resources is consistent with a second periodicity associated with the aperiodic training reference signal resources. For example, referring to FIG. 8 and FIG. 9, the temporal domain consistency condition may include a first periodicity (e.g., the periodicity of Set-B beams 942) associated with inference aperiodic CSI-RS resources is consistent with a second periodicity (e.g., the periodicity associated with Set-B beams 842) associated with the aperiodic training reference signal resources.
[0184] In some aspects, at 1306, the UE may receive, from the network entity, a periodicity indicator. The periodicity indicator may be indicative of the second periodicity or includes one or more of: the shortest periodicity for the aperiodic training reference signal resources, or an additional periodicity for the aperiodic training reference signal resources, where the additional periodicity is an integer multiple of the shortest periodicity. For example, referring to FIG. 11, the UE 1102 may receive, at 1110 from the network entity (base station 1104) , a periodicity indicator. The periodicity indicator may be indicative of the second periodicity or includes one or more of: the shortest periodicity for the aperiodic training reference signal resources (e.g., the shortest periodicity of 20 ms for Set-B beams 842) , or an additional periodicity for the aperiodic training reference signal resources. The additional periodicity may be an integer multiple of the shortest periodicity (e.g., twice or three times of 20 ms) . In some aspects, 1306 may be performed by the resource consistency component 198.
[0185] In some aspects, the first periodicity may be one of the shortest periodicity or the additional periodicity. For example, referring to FIG. 9, the first periodicity (e.g., the periodicity of the Set-B beams 942) may be one of the shortest periodicity or the additional periodicity (e.g., 20 ms or an integer multiple of 20 ms) .
[0186] FIG. 14 is a flowchart 1400 illustrating methods of wireless communication at a network entity in accordance with various aspects of the present disclosure. The method may be performed by a network entity. The network entity may be a base station, or a component of a base station, in the access network of FIG. 1 or a core network component (e.g., base station 102, 310, 1104; or the network entity 1602 in the hardware implementation of FIG. 16) . By ensuring spatial and temporal consistency in Set-B beams sub-sample patterns between the training and inference phases of an AI / ML model, the methods improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization.
[0187] As shown in FIG. 14, at 1402, the network entity may provide an aperiodic reference signal on a first set of resources. In some examples, the network entity may provide the aperiodic reference signal for a UE. The UE may be the UE 104, 350, 1102, or the apparatus 1604 in the hardware implementation of FIG. 16. FIG. 6, FIG. 7, FIG. 8, FIG. 9, FIG. 10A, FIG. 10B, and FIG. 11 illustrate various aspects of the steps in connection with flowchart 1400. For example, referring to FIG. 11, the network entity (base station 1104) may, at 1112, provide for the UE 1102 an aperiodic reference signal on a first set of resources. Referring to FIG. 7, the first set of resources may be the resources associated with Set-B beams 742 or 744. In some aspects, 1402 may be performed by the resource consistency component 199.
[0188] At 1404, the network entity may indicate for the UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern. The first mapping pattern may map the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training may meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. For example, referring to FIG. 11, the network entity (base station 1104) may, at 1116, indicate for the UE 1102 to predict one or more channel characteristics associated with a second set of resources (e.g., resources associated with Set-A beams 702) . Referring to FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams (e.g., 742, 744) to Set-A beams 702 during model inference 706. Referring to FIG. 6, the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. In some aspects, 1404 may be performed by the resource consistency component 199.
[0189] At 1406, the network entity may receive, from the UE, the one or more channel characteristics predicted for the second set of resources. For example, referring to FIG. 11, the network entity (base station 1104) may receive, at 1120, from the UE 1102, the one or more channel characteristics predicted for the second set of resources (e.g., resources associated with Set-A beams 702) . In some aspects, 1406 may be performed by the resource consistency component 199.
[0190] FIG. 15 is a flowchart 1500 illustrating methods of wireless communication at a network entity in accordance with various aspects of the present disclosure. The method may be performed by a network entity. The network entity may be a base station, or a component of a base station, in the access network of FIG. 1 or a core network component (e.g., base station 102, 310, 1104; or the network entity 1602 in the hardware implementation of FIG. 16) . By ensuring spatial and temporal consistency in Set-B beams sub-sample patterns between the training and inference phases of an AI / ML model, the methods improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization.
[0191] As shown in FIG. 15, at 1506, the network entity may provide an aperiodic reference signal on a first set of resources. In some examples, the network entity may provide the aperiodic reference signal for a UE. The UE may be the UE 104, 350, 1102, or the apparatus 1604 in the hardware implementation of FIG. 16. FIG. 6, FIG. 7, FIG. 8, FIG. 9, FIG. 10A, FIG. 10B, and FIG. 11 illustrate various aspects of the steps in connection with flowchart 1500. For example, referring to FIG. 11, the network entity (base station 1104) may, at 1112, provide for the UE 1102 an aperiodic reference signal on a first set of resources. Referring to FIG. 7, the first set of resources may be the resources associated with Set-B beams 742 or 744. In some aspects, 1506 may be performed by the resource consistency component 199.
[0192] At 1508, the network entity may indicate for the UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern. The first mapping pattern may map the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training may meet one or more of a spatial domain consistency condition or a temporal domain consistency condition. For example, referring to FIG. 11, the network entity (base station 1104) may, at 1116, indicate for the UE 1102 to predict one or more channel characteristics associated with a second set of resources (e.g., resources associated with Set-A beams 702) . Referring to FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams (e.g., 742, 744) to Set-A beams 702 during model inference 706. Referring to FIG. 6, the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. In some aspects, 1508 may be performed by the resource consistency component 199.
[0193] At 1510, the network entity may receive, from the UE, the one or more channel characteristics predicted for the second set of resources. For example, referring to FIG. 11, the network entity (base station 1104) may receive, at 1120, from the UE 1102, the one or more channel characteristics predicted for the second set of resources (e.g., resources associated with Set-A beams 702) . In some aspects, 1510 may be performed by the resource consistency component 199.
[0194] In some aspects, at 1522, the prediction of the one or more channel characteristics (at 1508) may be based on an AI / ML model, and the initial training (at 1508) may be a training of the AI / ML model. For example, referring to FIG. 7, the one or more channel characteristics may be predicted based on an AI / ML model 650 at 706. Referring to FIG. 6, the initial training may be the model training 606 of the AI / ML model 650.
[0195] In some aspects, at 1502, the network entity may provide, for the UE, an indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model. The training of the AI / ML model may be associated with the AI / ML functionality or the model identifier. For example, referring to FIG. 11, the network entity (base station 1104) may, at 1108, provide, for the UE 1102, an indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model. Referring to FIG. 6, the training of the AI / ML model 650 (at 606) may be associated with the AI / ML functionality or the model identifier. In some aspects, 1502 may be performed by the resource consistency component 199.
[0196] In some aspects, the training of the AI / ML model may be based on data collected via aperiodic training reference signal resources associated with the second set of resources, and the second mapping pattern may map a third set of resources to the second set of resources for the training of the AI / ML model. For example, referring to FIG. 6, the training of the AI / ML model 650 (at 606) may be based on data collected via aperiodic training reference signal resources associated with the second set of resources (e.g., resources associated with Set-A beams 602) , and the second mapping pattern may map a third set of resources (e.g., resources associated with Set-B beams 642 or 644) to the second set of resources (e.g., resources associated with Set-A beams 602) for the training of the AI / ML model 650 (at 606) .
[0197] In some aspects, at 1524, the one or more channel characteristics associated with the second set of resources may include one or more of: L1-RSRP for each resource of the second set of resources, L1-SINR for each resource of the second set of resources, or a subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources. For example, referring to FIG. 11, the one or more channel characteristics associated with the second set of resources (at 1118) may include one or more of: L1-RSRP for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) , or a subset of resources of the second set of resources (e.g., resources associated with Set-A beams 702) based on the L1-RSRP or the L1-SINR for each resource of the second set of resources (e.g., resources associated with Set-A beams 702) .
[0198] In some aspects, at 1526, the one or more channel characteristics may be predicted for the second set of resources based on a configuration of one or more of: an SSB resource set, a CSI-RS resource set, or a virtual resource set. For example, referring to FIG. 7, the one or more channel characteristics may be predicted (at 706) for the second set of resources (e.g., resources associated with Set-A beams 702) based on a configuration of one or more of: an SSB resource set, a CSI-RS resource set, or a virtual resource set.
[0199] In some aspects, the first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition.
[0200] In some aspects, the spatial domain consistency condition may include one or more of: the first mapping pattern and the second mapping pattern are based on the same mapping pattern, a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, where the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, wherein a first distribution of a starting identifier of the first set of resources is consistent with a second distribution of the starting identifier of the third set of resources, or, for each resource in the second set of resources, the difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold. For example, referring to FIG. 6 and FIG. 7, the spatial domain consistency condition may include one of: the first mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706) and the second mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606) are based on the same mapping pattern, or a first interval factor associated with the first mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706) that is equal to a second interval factor associated with the second mapping pattern (e.g., the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606) , where the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources (e.g., resources associated with Set-B beams 742) , and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources (e.g., resources associated with Set-B beams 642) . The spatial domain consistency condition may also include, for each resource in the second set of resources (e.g., resources associated with Set-A beams 602) , the difference between a first probability of being selected to the first set of resources (e.g., resources associated with Set-B beams 742) and a second probability of being selected to the third set of resources (e.g., resources associated with Set-B beams 642) is less than a threshold.
[0201] In some aspects, the first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition and the temporal domain consistency condition. For example, referring to FIG. 6 and FIG. 7, the first mapping pattern may be the sub-sample pattern that maps Set-B beams 742 to Set-A beams 702 during model inference 706, and the second mapping pattern may be the sub-sample pattern that maps Set-B beams 642 to Set-A beams 602 during model training 606. The first mapping pattern and the second mapping pattern may meet the spatial domain consistency condition and the temporal domain consistency condition.
[0202] In some aspects, the temporal domain consistency condition may include a first periodicity associated with aperiodic inference reference signal resources is consistent with a second periodicity associated with the aperiodic training reference signal resources. For example, referring to FIG. 8 and FIG. 9, the temporal domain consistency condition may include a first periodicity (e.g., the periodicity of Set-B beams 942) associated with inference aperiodic CSI-RS resources is consistent with a second periodicity (e.g., the periodicity associated with Set-B beams 842) associated with the aperiodic training reference signal resources.
[0203] In some aspects, at 1504, the network entity may transmit, for the UE, a periodicity indicator. The periodicity indicator may be indicative of the second periodicity or may include one or more of: a shortest periodicity for the aperiodic training reference signal resources, or an additional periodicity for the aperiodic training reference signal resources, wherein the additional periodicity is an integer multiple of the shortest periodicity. For example, referring to FIG. 11, the network entity (base station 1104) may transmit, at 1110 for the UE 1102, a periodicity indicator. The periodicity indicator may be indicative of the second periodicity or includes one or more of: the shortest periodicity for the aperiodic training reference signal resources (e.g., the shortest periodicity of 20 ms for Set-B beams 842) , or an additional periodicity for the aperiodic training reference signal resources. The additional periodicity may be an integer multiple of the shortest periodicity (e.g., twice or three times of 20 ms) . In some aspects, 1306 may be performed by the resource consistency component 198. In some aspects, 1504 may be performed by the resource consistency component 199.
[0204] In some aspects, the first periodicity may be one of the shortest periodicity or the additional periodicity. For example, referring to FIG. 9, the first periodicity (e.g., the periodicity of the Set-B beams 942) may be one of the shortest periodicity or the additional periodicity (e.g., 20 ms or an integer multiple of 20 ms) .
[0205] FIG. 16 is a diagram 1600 illustrating an example of a hardware implementation for an apparatus 1604. The apparatus 1604 may be a UE, a component of a UE, or may implement UE functionality. In some aspects, the apparatus 1604 may include at least one cellular baseband processor (or processing circuitry) 1624 (also referred to as a modem) coupled to one or more transceivers 1622 (e.g., cellular RF transceiver) . The cellular baseband processor (s) (or processing circuitry) 1624 may include at least one on-chip memory (or memory circuitry) 1524'. In some aspects, the apparatus 1604 may further include one or more subscriber identity modules (SIM) cards 1620 and at least one application processor (or processing circuitry) 1606 coupled to a secure digital (SD) card 1608 and a screen 1610. The application processor (s) (or processing circuitry) 1606 may include on-chip memory (or memory circuitry) 1506'. In some aspects, the apparatus 1604 may further include a Bluetooth module 1612, a WLAN module 1614, an SPS module 1616 (e.g., GNSS module) , one or more sensor modules 1618 (e.g., barometric pressure sensor / altimeter; motion sensor such as inertial measurement unit (IMU) , gyroscope, and / or accelerometer (s) ; light detection and ranging (LIDAR) , radio assisted detection and ranging (RADAR) , sound navigation and ranging (SONAR) , magnetometer, audio and / or other technologies used for positioning) , additional memory modules 1626, a power supply 1630, and / or a camera 1632. The Bluetooth module 1612, the WLAN module 1614, and the SPS module 1616 may include an on-chip transceiver (TRX) (or in some cases, just a receiver (RX) ) . The Bluetooth module 1612, the WLAN module 1614, and the SPS module 1616 may include their own dedicated antennas and / or utilize the antennas 1680 for communication. The cellular baseband processor (s) (or processing circuitry) 1624 communicates through the transceiver (s) 1622 via one or more antennas 1680 with the UE 104 and / or with an RU associated with a network entity 1602. The cellular baseband processor (s) (or processing circuitry) 1624 and the application processor (s) (or processing circuitry) 1606 may each include a computer-readable medium / memory (or memory circuitry) 1624', 1606', respectively. The additional memory modules 1626 may also be considered a computer-readable medium / memory (or memory circuitry) . Each computer-readable medium / memory (or memory circuitry) 1624', 1606', 1626 may be non-transitory. The cellular baseband processor (s) (or processing circuitry) 1624 and the application processor (s) (or processing circuitry) 1606 are each responsible for general processing, including the execution of software stored on the computer-readable medium / memory (or memory circuitry) . The software, when executed by the cellular baseband processor (s) (or processing circuitry) 1624 / application processor (s) (or processing circuitry) 1606, causes the cellular baseband processor (s) (or processing circuitry) 1624 / application processor (s) (or processing circuitry) 1606 to perform the various functions described supra. The cellular baseband processor (s) (or processing circuitry) 1624 and the application processor (s) (or processing circuitry) 1606 are configured to perform the various functions described supra based at least in part of the information stored in the memory (or memory circuitry) . That is, the cellular baseband processor (s) (or processing circuitry) 1624 and the application processor (s) (or processing circuitry) 1606 may be configured to perform a first subset of the various functions described supra without information stored in the memory and may be configured to perform a second subset of the various functions described supra based on the information stored in the memory. The computer-readable medium / memory (or memory circuitry) may also be used for storing data that is manipulated by the cellular baseband processor (s) (or processing circuitry) 1624 / application processor (s) (or processing circuitry) 1606 when executing software. The cellular baseband processor (s) (or processing circuitry) 1624 / application processor (s) (or processing circuitry) 1606 may be a component of the UE 350 and may include the at least one memory 360 and / or at least one of the TX processor 368, the RX processor 356, and the controller / processor 359. In one configuration, the apparatus 1604 may be at least one processor chip (modem and / or application) and include just the cellular baseband processor (s) (or processing circuitry) 1624 and / or the application processor (s) (or processing circuitry) 1606, and in another configuration, the apparatus 1604 may be the entire UE (e.g., see UE 350 of FIG. 3) and include the additional modules of the apparatus 1604.
[0206] As discussed supra, the component 198 may be configured to measure an aperiodic reference signal received on a first set of resources; predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and report, to a network entity, the one or more channel characteristics predicted for the second set of resources. The component 198 may be further configured to perform any of the aspects described in connection with the flowcharts in FIG. 12 and FIG. 13, and / or performed by the UE 1102 in FIG. 11. The component 198 may be within the cellular baseband processor (s) (or processing circuitry) 1624, the application processor (s) (or processing circuitry) 1606, or both the cellular baseband processor (s) (or processing circuitry) 1624 and the application processor (s) (or processing circuitry) 1606. The component 198 may be one or more hardware components specifically configured to carry out the stated processes / algorithm, implemented by one or more processors configured to perform the stated processes / algorithm, stored within a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors may perform the stated processes / algorithm individually or in combination. As shown, the apparatus 1604 may include a variety of components configured for various functions. In one configuration, the apparatus 1604, and in particular the cellular baseband processor (s) (or processing circuitry) 1624 and / or the application processor (s) (or processing circuitry) 1606, includes means for measuring an aperiodic reference signal received on a first set of resources, means for predicting one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition, and means for reporting, to a network entity, the one or more channel characteristics predicted for the second set of resources. The apparatus 1604 may further include means for performing any of the aspects described in connection with the flowcharts in FIG. 12 and FIG. 13, and / or aspects performed by the UE 1102 in FIG. 11. The means may be the component 198 of the apparatus 1604 configured to perform the functions recited by the means. As described supra, the apparatus 1604 may include the TX processor 368, the RX processor 356, and the controller / processor 359. As such, in one configuration, the means may be the TX processor 368, the RX processor 356, and / or the controller / processor 359 configured to perform the functions recited by the means.
[0207] FIG. 17 is a diagram 1700 illustrating an example of a hardware implementation for a network entity 1702. The network entity 1702 may be a BS, a component of a BS, or may implement BS functionality. The network entity 1702 may include at least one of a CU 1710, a DU 1730, or an RU 1740. For example, depending on the layer functionality handled by the component 199, the network entity 1702 may include the CU 1710; both the CU 1710 and the DU 1730; each of the CU 1710, the DU 1730, and the RU 1740; the DU 1730; both the DU 1730 and the RU 1740; or the RU 1740. The CU 1710 may include at least one CU processor (or processing circuitry) 1712. The CU processor (s) (or processing circuitry) 1712 may include on-chip memory (or memory circuitry) 1712'. In some aspects, the CU 1710 may further include additional memory modules 1714 and a communications interface 1718. The CU 1710 communicates with the DU 1730 through a midhaul link, such as an F1 interface. The DU 1730 may include at least one DU processor (or processing circuitry) 1732. The DU processor (s) (or processing circuitry) 1732 may include on-chip memory (or memory circuitry) 1632'. In some aspects, the DU 1730 may further include additional memory modules 1734 and a communications interface 1738. The DU 1730 communicates with the RU 1740 through a fronthaul link. The RU 1740 may include at least one RU processor (or processing circuitry) 1742. The RU processor (s) (or processing circuitry) 1742 may include on-chip memory (or memory circuitry) 1642'. In some aspects, the RU 1740 may further include additional memory modules 1744, one or more transceivers 1746, antennas 1780, and a communications interface 1748. The RU 1740 communicates with the UE 104. The on-chip memory (or memory circuitry) 1712', 1732', 1742' and the additional memory modules 1714, 1734, 1744 may each be considered a computer-readable medium / memory (or memory circuitry) . Each computer-readable medium / memory (or memory circuitry) may be non-transitory. Each of the processors (or processing circuitry) 1712, 1732, 1742 is responsible for general processing, including the execution of software stored on the computer-readable medium / memory (or memory circuitry) . The software, when executed by the corresponding processor (s) (or processing circuitry) causes the processor (s) (or processing circuitry) to perform the various functions described supra. The computer-readable medium / memory (or memory circuitry) may also be used for storing data that is manipulated by the processor (s) (or processing circuitry) when executing software.
[0208] As discussed supra, the component 199 may be configured to provide an aperiodic reference signal on a first set of resources; indicate for a UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and receive, from the UE, the one or more channel characteristics predicted for the second set of resources. The component 199 may be further configured to perform any of the aspects described in connection with the flowcharts in FIG. 14 and FIG. 15, and / or performed by the base station 1104 in FIG. 11. The component 199 may be within one or more processors (or processing circuitry) of one or more of the CU 1710, DU 1730, and the RU 1740. The component 199 may be one or more hardware components specifically configured to carry out the stated processes / algorithm, implemented by one or more processors configured to perform the stated processes / algorithm, stored within a computer-readable medium for implementation by one or more processors, or some combination thereof. When multiple processors are implemented, the multiple processors may perform the stated processes / algorithm individually or in combination. The network entity 1702 may include a variety of components configured for various functions. In one configuration, the network entity 1702 includes means for providing an aperiodic reference signal on a first set of resources, means for indicating for a UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition, and means for receiving, from the UE, the one or more channel characteristics predicted for the second set of resources. The network entity 1702 may further include means for performing any of the aspects described in connection with the flowcharts in FIG. 14 and FIG. 15, and / or aspects performed by the base station 1104 in FIG. 11. The means may be the component 199 of the network entity 1702 configured to perform the functions recited by the means. As described supra, the network entity 1702 may include the TX processor 316, the RX processor 370, and the controller / processor 375. As such, in one configuration, the means may be the TX processor 316, the RX processor 370, and / or the controller / processor 375 configured to perform the functions recited by the means.
[0209] FIG. 18 is an illustrative block diagram of an example ML architecture of first wireless device 1802 in communication with second wireless device 1804. In some aspects, the first wireless device 1802 may be a UE (e.g., UE 104, 350, 1102, or apparatus 1604) . First wireless device 1802 may be configured to measure an aperiodic reference signal received on a first set of resources; predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and report, to a network entity, the one or more channel characteristics predicted for the second set of resources. The component 198 may be further configured to perform any of the aspects described in connection with the flowcharts in FIG. 12 and FIG. 13, and / or performed by the UE 1102 in FIG. 11. Similarly, the second wireless device may be configured to provide an aperiodic reference signal on a first set of resources; indicate for a UE to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and receive, from the UE, the one or more channel characteristics predicted for the second set of resources. The component 199 may be further configured to perform any of the aspects described in connection with the flowcharts in FIG. 14 and FIG. 15, and / or performed by the base station 1104 in FIG. 11. Note that the example ML architecture of first wireless device 1802 may be applied to second wireless device 1804, and vice versa.
[0210] First wireless device 1802 may be, or may include, a chip, system on chip (SoC) , chipset, package or device that includes one or more processors, processing blocks or processing elements (collectively “processor 1810” ) and one or more memory blocks or elements (collectively “memory 1820” ) . Processor 1810 may be coupled to transceiver 1840, which includes radio frequency (RF) circuitry 1842 coupled to antennas 1846 via interface 1844, for transmitting or receiving signals.
[0211] One or more ML models 1830 (collectively “ML model 1830” ) may be stored in memory 1820 and accessible to processor (s) 1810. In some aspects, an ML mode, and / or updates for an ML model may be received and / or provided to a model server 1850. Individual or groups of ML models 1830 may be associated with respective model identifiers. In some aspects, different ML models 1830, which may optionally be associated with different model identifiers, may have different characteristics. One or more ML models 1830 may be selected based on respective features, characteristics, or applications, as well as characteristics or conditions of first wireless device 1802 (such as a power state, a mobility state, a battery reserve, a temperature, etc. ) . For example, ML models 1830 may have different inference data and output pairings (such as different types of inference data produce different types of output) , different levels of accuracies associated with the predictions, different latencies associated with producing the predictions, different ML model sizes, different coefficients, different parameters, etc.
[0212] Processor 1810 may deploy ML models 1830 to produce respective output data based on input data. As an example, the ML model 1830 may be configured to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition, e.g., as described in connection with any of FIGs. 6-13.
[0213] This disclosure provides a method for wireless communication at a UE. The method may include measuring an aperiodic reference signal received on a first set of resources; predicting one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, where the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and reporting, to a network entity, the one or more channel characteristics predicted for the second set of resources. By ensuring spatial and temporal consistency in Set-B beams sub-sample patterns between the training and inference phases of an AI / ML model, the methods improve the prediction accuracy and robustness of the AI / ML model, particularly when aperiodic CSI-RS resources are used, thereby leading to more efficient resource allocation and utilization.
[0214] It is understood that the specific order or hierarchy of blocks in the processes / flowcharts disclosed is an illustration of example approaches. Based upon design preferences, it is understood that the specific order or hierarchy of blocks in the processes / flowcharts may be rearranged. Further, some blocks may be combined or omitted. The accompanying method claims present elements of the various blocks in a sample order, and are not limited to the specific order or hierarchy presented.
[0215] The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects. Thus, the claims are not limited to the aspects described herein, but are to be accorded the full scope consistent with the language claims. Reference to an element in the singular does not mean “one and only one” unless specifically so stated, but rather “one or more. ” Terms such as “if, ” “when, ” and “while” do not imply an immediate temporal relationship or reaction. That is, these phrases, e.g., “when, ” do not imply an immediate action in response to or during the occurrence of an action, but simply imply that if a condition is met then an action will occur, but without requiring a specific or immediate time constraint for the action to occur. The word “exemplary” is used herein to mean “serving as an example, instance, or illustration. ” Any aspect described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other aspects. Unless specifically stated otherwise, the term “some” refers to one or more. Combinations such as “at least one of A, B, or C, ” “one or more of A, B, or C, ” “at least one of A, B, and C, ” “one or more of A, B, and C, ” and “A, B, C, or any combination thereof” include any combination of A, B, and / or C, and may include multiples of A, multiples of B, or multiples of C. Specifically, combinations such as “at least one of A, B, or C, ” “one or more of A, B, or C, ” “at least one of A, B, and C, ” “one or more of A, B, and C, ” and “A, B, C, or any combination thereof” may be A only, B only, C only, A and B, A and C, B and C, or A and B and C, where any such combinations may contain one or more member or members of A, B, or C. Sets should be interpreted as a set of elements where the elements number one or more. Accordingly, for a set of X, X would include one or more elements. When at least one processor is configured to perform a set of functions, the at least one processor, individually or in any combination, is configured to perform the set of functions. Accordingly, each processor of the at least one processor may be configured to perform a particular subset of the set of functions, where the subset is the full set, a proper subset of the set, or an empty subset of the set. A processor may be referred to as processor circuitry. A memory / memory module may be referred to as memory circuitry. If a first apparatus receives data from or transmits data to a second apparatus, the data may be received / transmitted directly between the first and second apparatuses, or indirectly between the first and second apparatuses through a set of apparatuses. A device configured to “output” data or “provide” data, such as a transmission, signal, or message, may transmit the data, for example with a transceiver, or may send the data to a device that transmits the data. A device configured to “obtain” data, such as a transmission, signal, or message, may receive, for example with a transceiver, or may obtain the data from a device that receives the data. Information stored in a memory includes instructions and / or data. All structural and functional equivalents to the elements of the various aspects described throughout this disclosure that are known or later come to be known to those of ordinary skill in the art are expressly incorporated herein by reference and are encompassed by the claims. Moreover, nothing disclosed herein is dedicated to the public regardless of whether such disclosure is explicitly recited in the claims. The words “module, ” “mechanism, ” “element, ” “device, ” and the like may not be a substitute for the word “means. ” As such, no claim element is to be construed as a means plus function unless the element is expressly recited using the phrase “means for. ”
[0216] As used herein, the phrase “based on” shall not be construed as a reference to a closed set of information, one or more conditions, one or more factors, or the like. In other words, the phrase “based on A” (where “A” may be information, a condition, a factor, or the like) shall be construed as “based at least on A” unless specifically recited differently.
[0217] The following aspects are illustrative only and may be combined with other aspects or teachings described herein, without limitation.
[0218] Aspect 1 is a method of wireless communication at a UE. The method includes measuring an aperiodic reference signal received on a first set of resources; predicting one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and reporting, to a network entity, the one or more channel characteristics predicted for the second set of resources.
[0219] Aspect 2 is the method of aspect 1, where the method further includes receiving, from the network entity, a first indication to predict the one or more channel characteristics associated with the second set of resources based on the measurement of the aperiodic reference signal on the first set of resources.
[0220] Aspect 3 is the method of any of aspects 1 to 2, wherein the prediction of the one or more channel characteristics is based on an artificial intelligence (AI) / machine learning (MI) (AI / ML) model, and the initial training is a training of the AI / ML model.
[0221] Aspect 4 is the method of any of aspects 1 to 3, where the method further includes receiving, from the network entity, a second indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model, wherein the training of the AI / ML model is associated with the AI / ML functionality or the model identifier.
[0222] Aspect 5 is the method of any of aspects 1 to 3, wherein the training of the AI / ML model is based on data collected via aperiodic training reference signal resources associated with the second set of resources, wherein the second mapping pattern maps a third set of resources to the second set of resources for the training of the AI / ML model.
[0223] Aspect 6 is the method of aspect 5, wherein the one or more channel characteristics associated with the second set of resources include one or more of: layer 1 (L1) reference signal received power (L1-RSRP) for each resource of the second set of resources, L1 signal-to-interference-plus-noise ratio (L1-SINR) for each resource of the second set of resources, or a subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources.
[0224] Aspect 7 is the method of aspect 6, wherein the one or more channel characteristics are predicted for the second set of resources based on a configuration of one or more of: a synchronization signal block (SSB) resource set, a CSI-RS resource set, or a virtual resource set.
[0225] Aspect 8 is the method of any of aspects 1 to 5, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition.
[0226] Aspect 9 is the method of aspect 8, wherein the spatial domain consistency condition includes the first mapping pattern and the second mapping pattern are based on a same mapping pattern.
[0227] Aspect 10 is the method of aspect 8, wherein the spatial domain consistency condition includes a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, and wherein a first distribution of a starting identifier of the first set of resources is consistent with a second distribution of the starting identifier of the third set of resources.
[0228] Aspect 11 is the method of aspect 8, wherein each resource in the second set of resources has a first respective probability of being selected to the first set of resources and has a second respective probability of being selected to the third set of resources, and the spatial domain consistency condition includes: a difference between the first respective probability and the second respective probability for each resource in the second set of resources is less than a threshold.
[0229] Aspect 12 is the method of any of aspects 1 to 5, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition and the temporal domain consistency condition.
[0230] Aspect 13 is the method of aspect 12, wherein the spatial domain consistency condition includes one of: the first mapping pattern and the second mapping pattern are based on a same mapping pattern, a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, or, for each resource in the second set of resources, a difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold.
[0231] Aspect 14 is the method of aspect 12, wherein the temporal domain consistency condition includes a first periodicity associated with inference aperiodic CSI-RS resources is consistent with a second periodicity associated with the aperiodic training reference signal resources.
[0232] Aspect 15 is the method of aspect 14, where the method further includes receiving, from the network entity, a periodicity indicator, wherein the periodicity indicator is indicative of the second periodicity or includes one or more of: a shortest periodicity for the aperiodic training reference signal resources, or an additional periodicity for the aperiodic training reference signal resources, wherein the additional periodicity is an integer multiple of the shortest periodicity.
[0233] Aspect 16 is the method of aspect 15, wherein the first periodicity is one of the shortest periodicity or the additional periodicity.
[0234] Aspect 17 is an apparatus for wireless communication at a UE, comprising: a processing system that includes processor circuitry and memory circuitry that stores code and is coupled with the processor circuitry, the processing system configured to cause the UE to perform the method of one or more of aspects 1-16.
[0235] Aspect 18 is an apparatus for wireless communication at a UE, comprising: at least one memory; and at least one processor coupled to the at least one memory and, where the at least one processor, individually or in any combination, is configured to perform the method of any of aspects 1-16.
[0236] Aspect 19 is the apparatus for wireless communication at a UE, comprising means for performing each step in the method of any of aspects 1-16.
[0237] Aspect 20 is an apparatus of any of aspects 17-19, further comprising a transceiver configured to receive or to transmit in association with the method of any of aspects 1-16.
[0238] Aspect 21 is a computer-readable medium (e.g., a non-transitory computer-readable medium) storing computer executable code at a UE, the code when executed by at least one processor causes the at least one processor to, individually or in any combination, perform the method of any of aspects 1-16.
[0239] Aspect 22 is a method of wireless communication at a network entity. The method includes providing an aperiodic reference signal on a first set of resources; indicating for a user equipment (UE) to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; and receiving, from the UE, the one or more channel characteristics predicted for the second set of resources.
[0240] Aspect 23 is the method of aspect 22, wherein the prediction of the one or more channel characteristics is based on an artificial intelligence (AI) / machine learning (MI) (AI / ML) model, and the initial training is a training of the AI / ML model.
[0241] Aspect 24 is the method of aspect 23, where the method further includes providing, for the UE, an indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model, wherein the training of the AI / ML model is associated with the AI / ML functionality or the model identifier.
[0242] Aspect 25 is the method of aspect 24, wherein the training of the AI / ML model is based on data collected via aperiodic training reference signal resources associated with the second set of resources, wherein the second mapping pattern maps a third set of resources to the second set of resources for the training of the AI / ML model.
[0243] Aspect 26 is the method of aspect 25, wherein the one or more channel characteristics associated with the second set of resources include one or more of: layer 1 (L1) reference signal received power (L1-RSRP) for each resource of the second set of resources, L1 signal-to-interference-plus-noise ratio (L1-SINR) for each resource of the second set of resources, or a subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources.
[0244] Aspect 27 is the method of aspect 26, wherein the one or more channel characteristics are predicted for the second set of resources based on a configuration of one or more of: a synchronization signal block (SSB) resource set, a CSI-RS resource set, or a virtual resource set.
[0245] Aspect 28 is the method of any of aspects 22 to 25, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition.
[0246] Aspect 29 is the method of aspect 28, wherein the spatial domain consistency condition includes one or more of: the first mapping pattern and the second mapping pattern are based on a same mapping pattern, a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, wherein a first distribution of a starting identifier of the first set of resources is consistent with a second distribution of the starting identifier of the third set of resources, or, for each resource in the second set of resources, a difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold.
[0247] Aspect 30 is the method of any of aspects 22 to 25, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition and the temporal domain consistency condition.
[0248] Aspect 31 is the method of aspect 30, wherein the temporal domain consistency condition includes a first periodicity associated with aperiodic inference reference signal resources is consistent with a second periodicity associated with the aperiodic training reference signal resources.
[0249] Aspect 32 is the method of aspect 31, where the method further includes transmitting, for the UE, a periodicity indicator, wherein the periodicity indicator is indicative of the second periodicity or includes one or more of: a shortest periodicity for the aperiodic training reference signal resources, or an additional periodicity for the aperiodic training reference signal resources, wherein the additional periodicity is an integer multiple of the shortest periodicity.
[0250] Aspect 33 is the method of aspect 32, wherein the first periodicity is one of the shortest periodicity or the additional periodicity.
[0251] Aspect 34 is an apparatus for wireless communication at a network entity, comprising: a processing system that includes processor circuitry and memory circuitry that stores code and is coupled with the processor circuitry, the processing system configured to cause the network entity to perform the method of one or more of aspects 22-33.
[0252] Aspect 35 is an apparatus for wireless communication at a network entity, comprising: at least one memory; and at least one processor coupled to the at least one memory and, where the at least one processor, individually or in any combination, is configured to perform the method of any of aspects 22-33.
[0253] Aspect 36 is the apparatus for wireless communication at a network entity, comprising means for performing each step in the method of any of aspects 22-33.
[0254] Aspect 37 is an apparatus of any of aspects 34-36, further comprising a transceiver configured to receive or to transmit in association with the method of any of aspects 22-33.
[0255] Aspect 38 is a computer-readable medium (e.g., a non-transitory computer-readable medium) storing computer executable code at a network entity, the code when executed by at least one processor causes the at least one processor to, individually or in any combination, perform the method of any of aspects 22-33.
Claims
1.An apparatus for wireless communication at a user equipment (UE) , comprising:at least one memory; andat least one processor coupled to the at least one memory and, based at least in part on information stored in the at least one memory, the at least one processor, individually or in any combination, is configured to cause the UE to:measure an aperiodic reference signal received on a first set of resources;predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; andreport, to a network entity, the one or more channel characteristics predicted for the second set of resources.2.The apparatus of claim 1, further comprising a transceiver coupled to the at least one processor, wherein to report the one or more channel characteristics predicted for the second set of resources, the at least one processor, individually or in any combination, is configured to cause the UE to report the one or more channel characteristics predicted for the second set of resources via the transceiver, and wherein the at least one processor, individually or in any combination, is further configured to cause the UE to:receive, from the network entity, a first indication to predict the one or more channel characteristics associated with the second set of resources based on the measurement of the aperiodic reference signal on the first set of resources.3.The apparatus of claim 1, wherein a prediction of the one or more channel characteristics is based on an artificial intelligence (AI) / machine learning (MI) (AI / ML) model, and the initial training is a training of the AI / ML model.4.The apparatus of claim 3, wherein the at least one processor, individually or in any combination, is further configured to cause the UE to:receive, from the network entity, a second indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model, wherein the training of the AI / ML model is associated with the AI / ML functionality or the model identifier.5.The apparatus of claim 3, wherein the training of the AI / ML model is based on data collected via aperiodic training reference signal resources associated with the second set of resources, wherein the second mapping pattern maps a third set of resources to the second set of resources for the training of the AI / ML model.6.The apparatus of claim 5, wherein the one or more channel characteristics associated with the second set of resources include one or more of:layer 1 (L1) reference signal received power (L1-RSRP) for each resource of the second set of resources,L1 signal-to-interference-plus-noise ratio (L1-SINR) for each resource of the second set of resources, ora subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources.7.The apparatus of claim 6, wherein the one or more channel characteristics are predicted for the second set of resources based on a configuration of one or more of:a synchronization signal block (SSB) resource set,a CSI-RS resource set, ora virtual resource set.8.The apparatus of claim 5, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition.9.The apparatus of claim 8, wherein the spatial domain consistency condition includes the first mapping pattern and the second mapping pattern are based on a same mapping pattern.10.The apparatus of claim 8, wherein the spatial domain consistency condition includes a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, andwherein a first distribution of a starting identifier of the first set of resources is consistent with a second distribution of the starting identifier of the third set of resources.11.The apparatus of claim 8, wherein each resource in the second set of resources has a first respective probability of being selected to the first set of resources and has a second respective probability of being selected to the third set of resources, and the spatial domain consistency condition includes: a difference between the first respective probability and the second respective probability for each resource in the second set of resources is less than a threshold.12.The apparatus of claim 5, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition and the temporal domain consistency condition.13.The apparatus of claim 12, wherein the spatial domain consistency condition includes one of:the first mapping pattern and the second mapping pattern are based on a same mapping pattern,a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, orfor each resource in the second set of resources, a difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold.14.The apparatus of claim 12, wherein the temporal domain consistency condition includes a first periodicity associated with inference aperiodic CSI-RS resources is consistent with a second periodicity associated with the aperiodic training reference signal resources.15.The apparatus of claim 14, wherein the at least one processor, individually or in any combination, is further configured to cause the UE to:receive, from the network entity, a periodicity indicator, wherein the periodicity indicator is indicative of the second periodicity or includes one or more of:a shortest periodicity for the aperiodic training reference signal resources, oran additional periodicity for the aperiodic training reference signal resources, wherein the additional periodicity is an integer multiple of the shortest periodicity.16.The apparatus of claim 15, wherein the first periodicity is one of the shortest periodicity or the additional periodicity.17.An apparatus for wireless communication at a network entity, comprising:at least one memory; andat least one processor coupled to the at least one memory and, based at least in part on information stored in the at least one memory, the at least one processor, individually or in any combination, is configured to cause the network entity to:provide an aperiodic reference signal on a first set of resources;indicate for a user equipment (UE) to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; andreceive, from the UE, the one or more channel characteristics predicted for the second set of resources.18.The apparatus of claim 17, further comprising a transceiver coupled to the at least one processor, wherein, to receive the one or more channel characteristics predicted for the second set of resources, the at least one processor, individually or in any combination, is configured to cause the network entity to receive the one or more channel characteristics predicted for the second set of resources via the transceiver, and wherein a prediction of the one or more channel characteristics is based on an artificial intelligence (AI) / machine learning (MI) (AI / ML) model, and the initial training is a training of the AI / ML model.19.The apparatus of claim 18, wherein the at least one processor, individually or in any combination, is further configured to cause the network entity to:provide, for the UE, an indication of one or more of an AI / ML functionality and a model identifier for the AI / ML model, wherein the training of the AI / ML model is associated with the AI / ML functionality or the model identifier.20.The apparatus of claim 19, wherein the training of the AI / ML model is based on data collected via aperiodic training reference signal resources associated with the second set of resources, wherein the second mapping pattern maps a third set of resources to the second set of resources for the training of the AI / ML model.21.The apparatus of claim 20, wherein the one or more channel characteristics associated with the second set of resources include one or more of:layer 1 (L1) reference signal received power (L1-RSRP) for each resource of the second set of resources,L1 signal-to-interference-plus-noise ratio (L1-SINR) for each resource of the second set of resources, ora subset of resources of the second set of resources based on the L1-RSRP or the L1-SINR for each resource of the second set of resources.22.The apparatus of claim 21, wherein the one or more channel characteristics are predicted for the second set of resources based on a configuration of one or more of:a synchronization signal block (SSB) resource set,a CSI-RS resource set, ora virtual resource set.23.The apparatus of claim 20, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition.24.The apparatus of claim 23, wherein the spatial domain consistency condition includes one or more of:the first mapping pattern and the second mapping pattern are based on a same mapping pattern,a first interval factor associated with the first mapping pattern that is equal to a second interval factor associated with the second mapping pattern, wherein the first interval factor specifies a first interval between a first set of identifiers associated with the first set of resources, and the second interval factor specifies a second interval between a second set of identifiers associated with the third set of resources, wherein a first distribution of a starting identifier of the first set of resources is consistent with a second distribution of the starting identifier of the third set of resources, orfor each resource in the second set of resources, a difference between a first probability of being selected to the first set of resources and a second probability of being selected to the third set of resources is less than a threshold.25.The apparatus of claim 20, wherein the first mapping pattern and the second mapping pattern meet the spatial domain consistency condition and the temporal domain consistency condition.26.The apparatus of claim 25, wherein the temporal domain consistency condition includes a first periodicity associated with aperiodic inference reference signal resources is consistent with a second periodicity associated with the aperiodic training reference signal resources.27.The apparatus of claim 26, wherein the at least one processor, individually or in any combination, is further configured to cause the network entity to:transmit, for the UE, a periodicity indicator, wherein the periodicity indicator is indicative of the second periodicity or includes one or more of:a shortest periodicity for the aperiodic training reference signal resources, oran additional periodicity for the aperiodic training reference signal resources, wherein the additional periodicity is an integer multiple of the shortest periodicity.28.The apparatus of claim 27, wherein the first periodicity is one of the shortest periodicity or the additional periodicity.29.A method of wireless communication at a user equipment (UE) , comprising:measuring an aperiodic reference signal received on a first set of resources;predicting one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; andreporting, to a network entity, the one or more channel characteristics predicted for the second set of resources.30.A method of wireless communication at a network entity, comprising:providing an aperiodic reference signal on a first set of resources;indicating for a user equipment (UE) to predict one or more channel characteristics associated with a second set of resources using measurement of the aperiodic reference signal on the first set of resources and based on a first mapping pattern, wherein the first mapping pattern maps the first set of resources to the second set of resources, and the first mapping pattern and a second mapping pattern associated with an initial training meet one or more of a spatial domain consistency condition or a temporal domain consistency condition; andreceiving, from the UE, the one or more channel characteristics predicted for the second set of resources.
Citation Information
Patent Citations
Information determination method and device, communication equipment and readable storage medium
CN117294407A
Channel state information (CSI) reference signal (RS) configuration with cross-component carrier CSI prediction algorithm
US20210258991A1
Method and apparatus for configurable measurement resources and reporting
US20230247454A1
Model-based channel state information
US20230291518A1
Methods and apparatus for leveraging transfer learning for channel state information enhancement
WO2023212059A1