Method of network data collection for beam management based on machine learning

Abstract

This disclosure provides systems, apparatus, devices, and methods, including computer programs encoded on a storage medium, by a UE to collect beam measurement data for training, refining, and monitoring machine learning models for beam management on the network side. The signaling framework allows a network entity to control the range of collected data and configure the content and time domain behavior of the report. The UE (102) receives (1104), from a network entity (104), control signaling for configuring one or more sets of channel measurement resources (CMRs) and reports for data collection associated with machine learning based beam management. The UE (102) receives (1110) a downlink reference signal from the network entity (104) via the one or more CMR sets. The UE (102) transmits (1112) the report to the network entity (104) based on a beam measurement corresponding to the downlink reference signal.

Description

TECHNICAL FIELD

[0001] The present disclosure relates generally to wireless communications, and more particularly, to techniques for a user equipment (UE) to collect beam measurement data for training, refining, and monitoring machine learning models on the network side for beam management. BACKGROUND

[0002] The Third Generation Partnership Project (3GPP) specifies a radio interface called Fifth Generation (5G) New Radio (NR) (5G NR). The architecture of a 5G NR wireless communication system includes a 5G core (5GC) network, a 5G radio access network (5G-RAN), user equipment (UE), and the like. In comparison to previous generation cellular communication systems, the 5G NR architecture seeks to provide increased data rates, reduced latency, and / or increased capacity.

[0003] Generally, wireless communication systems are based on multiple access technologies that support communication with multiple UEs, such as orthogonal frequency division multiple access (OFDMA) technologies. Improvements in mobile broadband technologies continue to expand the availability of wireless communication services. For example, for beam management, a UE and a network entity can cooperate to identify and maintain optimal or preferred beams for transmissions in uplink and downlink directions. Beam management can also be used to support beamforming at the network entity and / or the UE. Efficient beam management is crucial as communication systems provide increased capacity in different deployment scenarios. SUMMARY

[0004] 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. It is not intended to identify key or critical elements of all aspects or to delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.

[0005] In beam management, a UE and a network entity can collaborate to identify and maintain optimal or preferred beams for transmission in the uplink and downlink directions. Beam management can also be used to support beamforming at the network entity and / or the UE. For example, in the downlink direction, the network entity and the UE can perform beam management procedures in a hierarchical manner to identify relatively wide beams for initial acquisition and then identify more directional and higher gain beams for physical downlink shared channel (PDSCH) and physical downlink control channel (PDCCH). Beam selection and refinement can be based on downlink reference signals, such as synchronization signal / physical broadcast channel (SS / PBCH) blocks (referred to as SSBs) and channel state information reference signals (CSI-RSs) configured as channel measurement resources (CMRs). The network entity can apply beamforming coefficients to a set of SSBs to generate relatively wide beams for initial acquisition by the UE. The network entity can then apply beam coefficients to a set of CSI-RS resources to generate more directional beams (e.g., narrower beams) for subsequent beam refinement. The UE can measure the downlink reference signals and provide feedback to the network entity in a CSI report to allow for fast and responsive switching between beams. The CSI report can include a SSB block resource indicator (SSBRI) or a CSI-RS resource indicator (CRI) to indicate one or more preferred SSB or CSI-RS beams. The CSI report can also include a layer 1 reference signal received power (L1-RSRP) or a layer 1 signal to interference plus noise ratio (L1-SINR) of the preferred SSB block or CSI-RS beam measured by the UE.

[0006] Machine learning can be used to assist in beam management. For example, in spatial domain beam prediction, a machine learning model can predict one or more best downlink beams based on beam quality measurements of a limited number of beams of downlink reference signals, such as CSI-RS resources configured as CMRs. In temporal domain beam prediction, a machine learning model can predict one or more best downlink beams for a plurality of future time instances based on a limited number of beam quality measurements of downlink reference signals at different time instances in the past. One key step in machine learning is data collection, which is the collection of input and output data used by the machine learning model for model training, model refinement, model monitoring, etc. The UE can collect input and output data including beam quality measurements and indices of the best beams. For example, the collected data can include one or more of L1-RSRP, L1-SINR, SSBRI, or CRI.

[0007] Machine learning models for beam management can reside on the network entity side. Data collection to support machine learning based beam management, such as spatial domain and time domain beam prediction, can require more functionality than that provided by existing UE beam measurement and CSI reporting capabilities. Data collection for machine learning based beam management also introduces other complexities. For example, a UE can perform receive beam sweeping to identify a best UE receive beam to receive a downlink reference signal for data collection. If the downlink reference signal overlaps with other downlink signals in time domain, the UE can face a problem of determining whether and how to receive the downlink reference signal to identify the best beam. In addition, there can be measurement errors when the UE makes beam quality measurements. Measurement errors of data collected by the UE can degrade the performance of machine learning. The UE can consider how to reduce the negative impact of measurement errors when reporting beam quality measurements.

[0008] Aspects of the present disclosure address the above and other deficiencies associated with data collection by a UE to support machine learning based beam prediction performed on the network entity side. In some aspects, a signaling framework is disclosed to allow a network entity to flexibly control the scope of data collected by a UE, such as the number, frequency, timing, etc. of beam quality measurements. In some aspects, when a UE performs receive beam sweeping to receive a downlink reference signal, there can be scheduling restrictions to suppress other downlink signals transmitted by the network entity that overlap in time with the downlink reference signal. In some aspects, if there are no such scheduling restrictions, the UE can determine a receive beam based on a priority rule to apply a receive beam corresponding to a downlink signal with a highest priority. In some aspects, a network entity can configure the content and time domain behavior of a CSI report. In some embodiments, a UE can report an L1-RSRP of a configured CMR and an index of a best beam based on a hypothetical measurement error of the L1-RSRP, L1-SINR, and / or a configured CMR that satisfies one or more criteria.

[0009] According to some aspects, a UE receives, from a network entity, control signaling for configuring one or more channel measurement resource (CMR) sets and reporting of data collection associated with machine learning based beam management. The UE receives, from the network entity, a downlink reference signal via the one or more CMR sets. The UE transmits, to the network entity, a report based on beam measurements corresponding to the downlink reference signal.

[0010] According to some aspects, the network entity transmits control signaling to the UE for configuring one or more sets of Channel Measurement Resources (CMRs) and reports on data collection associated with machine learning-based beam management on the network entity. The network entity transmits downlink reference signals to the UE via one or more sets of CMRs. The network entity receives reports from the UE based on beam measurements corresponding to the downlink reference signals. Attached Figure Description

[0011] Figure 1 An illustration of a wireless communication system according to an embodiment is shown, the wireless communication system including a plurality of user equipments (UEs) and network entities communicating through one or more cells.

[0012] Figure 2 An example of a machine learning model is shown, based on beam quality measurements of a finite number of beams, to predict the optimal set of beams in the spatial domain, according to an embodiment.

[0013] Figure 3 An example of a machine learning model, according to an embodiment, is shown to predict the optimal beam in the time domain based on beam quality measurements from multiple time report instances.

[0014] Figure 4 This is a signaling diagram illustrating the communication between a UE and a network entity according to an embodiment, used by the UE to collect data for supporting machine learning-based beam management on the network side.

[0015] Figure 5 An example of a beam report is shown, according to an embodiment, when the joint input / output beam report used for a machine learning model includes a subset of beam quality measurements from a set of channel measurement resources (CMRs) and a beam index of the best beam of the CMRs.

[0016] Figure 6 An example of a beam report based on beam measurements of the CMR at multiple report instances prior to a reference time is shown, according to an embodiment.

[0017] Figure 7 An example of a beam report based on averaging beam measurements of the CMR for each of a plurality of report instances prior to a reference time is shown, according to an embodiment.

[0018] Figure 8 An example of a beam report is shown according to an embodiment when a network entity configures two CMR sets for beam measurement and the joint input / output beam report for a machine learning model includes a subset of beam quality measurements from the first CMR set and the beam index of the best beam from the second CMR set.

[0019] Figure 9An example of a beam report is shown according to an embodiment when a network entity configures two CMR sets for beam measurement and the separate beam reports for the input and output data of the machine learning model include a subset of beam quality measurements from the first CMR set and a beam index of the best beam from the second CMR set.

[0020] Figure 10 An example of beam measurement of CMR based on a configured measurement window that instructs the UE to measure CMR to generate a beam report is shown according to an embodiment.

[0021] Figure 11 This is a flowchart illustrating a method for receiving CMR at the UE and reporting beam measurements of CMR in a beam report, according to an embodiment, to support wireless communication based on machine learning-based beam management on the network side.

[0022] Figure 12 This is a flowchart of a method for transmitting CMR and receiving beam measurements of CMR at a network entity according to an embodiment to support wireless communication based on machine learning-based beam management on the network side.

[0023] Figure 13 This is a diagram illustrating a hardware implementation of an example UE device according to some embodiments.

[0024] Figure 14 This is a diagram illustrating a hardware implementation of one or more example network entities according to some embodiments. Detailed Implementation

[0025] Figure 1A diagram 100 illustrates a wireless communication system associated with multiple cells 190 according to one embodiment. The wireless communication system includes user equipment (UE) 102 and base station / network entity 104. Some base stations may include an aggregated base station architecture, while others may include a decomposed base station architecture. The aggregated base station architecture utilizes a radio protocol stack physically or logically integrated within a single radio access network (RAN) node. The decomposed base station architecture utilizes a protocol stack physically or logically distributed across two or more units (e.g., radio unit (RU) 106, distributed unit (DU) 108, central unit (CU) 110). For example, CU 110 is implemented within a RAN node, and one or more DU 108 may be located in the same location as CU 110, or alternatively, may be geographically or virtually distributed across one or more other RAN nodes. DU 108 may be implemented to communicate with one or more RU 106. Any of RU 106, DU 108, and CU 110 can be implemented as a virtual unit, such as a virtual radio unit (VRU), a virtual distributed unit (VDU), or a virtual central unit (VCU). Base station / network entity 104 (e.g., an aggregated base station or a decomposed unit of a base station, such as RU 106 or DU 108) can be referred to as a transmit receiver point (TRP).

[0026] The operation and / or network design of base station 104 can be based on the aggregation characteristics of base station functions. For example, a decomposed base station architecture can be utilized in an Integrated Access Backhaul (IAB) network, an Open Radio Access Network (O-RAN) network, or a Virtual Radio Access Network (vRAN) (which may also be referred to as a Cloud Radio Access Network (C-RAN)). Decomposition can include distributing functions among two or more units located in various physical locations, as well as virtually distributing the functions of at least one unit, which allows for flexibility in network design. Various units in a decomposed base station architecture or a decomposed RAN architecture can be configured to communicate with at least one other unit via wired or wireless communication. For example, base stations 104d, 104e and / or RUs 106a, 106b, 106c, 106d can communicate with UEs 102a, 102b, 102c, 102d and / or 102s via one or more radio frequency (RF) access links based on a Uu interface. In the example, multiple RUs 106 and / or base stations 104 can simultaneously serve UE 102, such as through intra-cell and / or inter-cell access links between UE 102 and RUs 106 / base stations 104.

[0027] RU 106, DU 108, and CU 110 may include (or may be coupled to) one or more interfaces configured to transmit or receive information / signals via wired or wireless transmission media. For example, a wired interface may be configured to transmit or receive information / signals via a wired transmission medium—such as a fronthaul link 160 between RU 106d and a baseband unit (BBU) 112 of base station 104d associated with cell 190d. BBU 112 includes DU 108 and CU 110, and may also have a wired interface (e.g., a midhaul link) configured between DU 108 and CU 110 for transmitting or receiving information / signals between DU 108 and CU 110. In a further example, a wireless interface that may include a receiver, transmitter, or transceiver (such as an RF transceiver) may be configured to transmit and / or receive information / signals via a wireless transmission medium, such as information transmitted between RU 106a in cell 190a and base station 104e in cell 190e via inter-cell communication beams 136-138 of RU 106a and base station 104e.

[0028] RU 106 can be configured to implement low-level functions. For example, RU 106 is controlled by DU 108 and can correspond to a logical node that manages RF processing functions or low-level PHY functions such as performing Fast Fourier Transform (FFT), Inverse FFT (iFFT), digital beamforming, Physical Random Access Channel (PRACH) extraction, and filtering. The functionality of RU 106 can be based on functional partitioning, such as low-level functional partitioning.

[0029] RU 106 can transmit or receive over-the-air (OTA) communications with one or more UEs 102. For example, RU 106b of cell 190b communicates with UE 102b of cell 190b via a first communication beamset 132 of RU 106b and a second communication beamset 134b of UE 102b, which may correspond to inter-cell communication beams or, in some examples, inter-cell communication beams. For example, UE 102b of cell 190b can communicate with RU 106a of cell 190a via a third communication beamset 134a of UE 102b and a fourth communication beamset 136 of RU 106a. DU 108 can control the real-time and non-real-time characteristics of control plane and user plane communications of RU 106.

[0030] Any combination or individual reference to RU 106, DU 108, and CU 110 may correspond to base station 104. Therefore, base station 104 may include at least one of RU 106, DU 108, or CU 110. Base station 104 provides UE 102 with access to the core network. Base station 104 may relay communication between UE 102 and the core network (not shown). Base station 104 may be associated with macro cells of high-power cellular base stations and / or small cells of low-power cellular base stations. For example, cell 190e may correspond to a macro cell, while cells 190a-190d may correspond to small cells. Small cells include femtocells, picocells, microcells, etc. A network including at least one macro cell and at least one small cell may be referred to as a "heterogeneous network".

[0031] Transmissions from UE 102 to base station 104 / RU 106 are called uplink (UL) transmissions, while transmissions from base station 104 / RU 106 to UE 102 are called downlink (DL) transmissions. Uplink transmissions can also be called reverse link transmissions, and downlink transmissions can also be called forward link transmissions. For example, RU 106d uses the antenna of base station 104d in cell 190d to transmit downlink / forward link communication to UE 102d via the Uu interface associated with the access link between UE 102d and base station 104d / RU 106d, or receives uplink / reverse link communication from UE 102d.

[0032] The communication link between UE 102 and base station 104 / RU 106 can be based on multiple-input multiple-output (MIMO) antenna technology, including spatial multiplexing, beamforming, and / or transmit diversity. The communication link can be associated with one or more carriers. UE 102 and base station 104 / RU 106 can utilize up to a total of Yx Each carrier allocated in MHz carrier aggregation Y MHz (e.g., 5 MHz, 10 MHz, 15 MHz, 20 MHz, 100 MHz, 400 MHz, 800 MHz, 1600 MHz, 2000 MHz, etc.) spectrum bandwidth, of which x Each component carrier (CC) is used for communication in each of the uplink and downlink directions. The carriers may or may not be adjacent to each other along the spectrum. In the example, uplink and downlink carriers may be allocated asymmetrically, with more or fewer carriers allocated to the uplink or downlink. A component carrier may include a primary component carrier and one or more secondary component carriers. The primary component carrier may be associated with a primary cell (PCell), and the secondary component carriers may be associated with secondary cells (SCells).

[0033] Some UEs 102 (such as UEs 102a and 102s) can perform device-to-device (D2D) communication via a sidelink. For example, the sidelink communication / D2D link utilizes the spectrum of the Wireless Wide Area Network (WWAN) associated with uplink and downlink communication. Such sidelink / D2D communication can be performed by various wireless communication systems, such as Wi-Fi, Bluetooth, LTE, and NR systems.

[0034] UE 102 and base station 104 / RU 106 may each include multiple antennas. These multiple antennas may correspond to antenna elements, antenna panels, and / or antenna arrays that facilitate beamforming operation. For example, RU 106b transmits downlink beamforming signals to UE 102b based on a first communication beamset 132 in one or more transmission directions of RU 106b. UE 102b may receive downlink beamforming signals from RU 106b based on a second communication beamset 134b in one or more reception directions of UE 102b. In a further example, UE 102b may also transmit uplink beamforming signals (e.g., sounding reference signals (SRS)) to RU 106b based on the second communication beamset 134b in one or more transmission directions of UE 102b. RU 106b may receive uplink beamforming signals from UE 102b in one or more reception directions of RU 106b. UE 102b can perform beam training to determine the optimal reception and transmission directions for beamformed signals. The transmission and reception directions of UE 102 and base station 104 / RU 106 may be the same or different.

[0035] In a further example, beamforming signals can be transmitted between a first base station / RU 106a and a second base station 104e. For example, base station 104e of cell 190e can transmit beamforming signals to RU 106a based on communication beam 138 in one or more transmission directions of base station 104e. RU 106a can receive beamforming signals from base station 104e of cell 190e based on RU communication beam 136 in one or more reception directions of RU 106a. In a further example, base station 104e transmits downlink beamforming signals to UE 102e based on communication beam 138 in one or more transmission directions of base station 104e. UE 102e receives downlink beamforming signals from base station 104e based on UE communication beam 130 in one or more reception directions of UE 102e. UE 102e can also transmit uplink beamforming signals to base station 104e based on UE communication beam 130 in one or more transmission directions of UE 102e, so that base station 104e can receive uplink beamforming signals from UE 102e in one or more receiving directions of base station 104e.

[0036] Base station 104 may include and / or be referred to as a network entity. That is, a "network entity" may refer to base station 104 or at least one element of base station 104, such as RU 106, DU 108, and / or CU 110. Base station 104 may also include and / or be referred to as Next Generation Evolved Node B (ng-eNB), Next Generation NB (gNB), Evolved NB (eNB), access point, base transceiver, radio base station, radio transceiver, transceiver function, Basic Service Set (BSS), Extended Service Set (ESS), TRP, network node, network device, or other related terms. Base station 104 or the entity at base station 104 may be implemented as an IAB node, relay node, sidelink node, aggregated (monolithic) base station, or a decomposed base station including one or more RU 106, DU 108, and / or CU 110. Aggregated or decomposed base station sets may be referred to as Next Generation Radio Access Network (NG-RAN). In some examples, UE 102a operates in dual connectivity (DC) with base station 104e and base station / RU 106a. In such a case, base station 104e can be the primary node, while base station / RU 160a can be the secondary node.

[0037] Still referencing Figure 1In some aspects, any of UEs 102 may include a data collection component 140 (also referred to as ML data collection component 140) configured to collect data to support the training, refinement, or monitoring of a machine learning model for downlink beam management on the network side. The ML data collection component 140 may receive control signaling from the base station / network entity 104 for configuring one or more sets of channel measurement resources (CMRs) and for configuring reports for data collection associated with machine learning-based beam management. The ML data collection component 140 may receive downlink reference signals from the base station / network entity 104 via one or more sets of CMRs. The ML data collection component 140 may transmit reports to the base station / network entity 104 based on beam measurements corresponding to the downlink reference signals.

[0038] In some aspects, any of the base stations 104 or the network entity of base station 104 may include a network-side machine learning-based beam management configuration component 150 (also referred to as ML data collection configuration component 150) configured to control UE data collection to support the training, refinement, or monitoring of a machine learning model for downlink beam management on the network side. The ML data collection configuration component 150 may transmit control signaling to any of the UEs 102 for configuring one or more CMR sets and for configuring reports for data collection associated with machine learning-based beam management on base station / network entity 104. The ML data collection configuration component 150 may transmit downlink reference signals to the UEs 102 via one or more CMR sets. The ML data collection configuration component 150 may receive reports from the UE based on beam measurements corresponding to the downlink reference signals.

[0039] therefore, Figure 1 A wireless communication system that can be implemented in conjunction with one or more other figures described herein is described. Furthermore, although the following description may focus on 5G NR, the concepts described herein are applicable to other similar fields, such as 5G-Advanced and future versions, LTE, LTE-advanced (LTE-A), and other wireless technologies such as 6G.

[0040] Machine learning models used to support downlink beam management or prediction can reside on the network side or the UE side. The machine learning model can predict one or more optimal downlink beams in the spatial or temporal domain based on a finite number of beam measurements of a downlink reference signal configured for CMR. Training, refinement, or monitoring of the machine learning model relies on data collection of beam measurements performed by UE 102. For example, UE 102 can collect beam quality measurements (such as L1-RSRP or L1-SINR of the downlink reference signal beam) as input training data applied to the machine learning model. UE 102 can receive the downlink reference signal within a range of azimuth and elevation angles (also referred to as zenith angles). UE 102 can determine one or more optimal beams based on the L1-RSRP or L1-SINR of the downlink reference signal as output training data applied to the machine learning model. The optimal beams can be specified with preferred azimuth and elevation angles. When the machine learning model resides on the network side, UE 102 can report the input and output training data used for the machine learning model as beam reports to network entity 104. Once trained, network entity 104 can use a machine learning model to predict the optimal beam for downlink transmission of data or control signals based on a limited number of beam measurements of the downlink reference signal monitored by UE 102.

[0041] Figure 2 Example 200 illustrates a machine learning model 210 that predicts the optimal beam set 230 in the spatial domain based on beam quality measurements of a finite number of beams, according to one embodiment. Network entity 104 can transmit beams carrying downlink reference signals in the spatial domain to cover ranges from azimuth (AoD) to zenith (ZoD). The downlink reference signal can be a synchronization signal / physical broadcast channel (SS / PBCH) block (referred to as SSB) configured for CMR or a channel state information reference signal (CSI-RS). Figure 2 An array of 4 beams in the ZoD dimension and 8 beams in the AoD dimension for a total of 32 beams in the spatial domain is shown. Instead of measuring all 32 beams, UE 102 can measure the beam quality of four randomly selected beams 220 of the downlink reference signal to reduce the computational load for beam management. Machine learning model 210 can apply the beam quality measurements of the four beams 220 to predict or infer the optimal beam set 230 for downlink transmission. Network entity 104 can use a subset of the optimal beams 230 to transmit the Physical Downlink Shared Channel (PDSCH) and Physical Downlink Control Channel (PDCCH) to UE 102.

[0042] Figure 3Example 300 of a machine learning model 310 is shown, based on beam quality measurements from multiple time reporting instances, to predict the optimal beam or optimal beam set 360 in the time domain according to one embodiment. Network entity 104 may transmit beams carrying downlink reference signals over multiple time instances, such as by periodically transmitting beams over time spans. In one aspect, for each time instance, network entity 104 may transmit multiple beams to cover the range of AoD and ZoD in the spatial domain. UE 102 may perform beam measurements 320, 330, and 340 at three time reporting instances. Each beam measurement for a time reporting instance may include beam quality measurements of multiple beams received at the time reporting instance. Machine learning model 310 may apply beam quality measurements 320, 330, and 340 from the three time reporting instances to predict or infer optimal beams 350 and 360 for downlink transmission at two future time instances.

[0043] This disclosure addresses aspects of the complexities associated with data collection performed by the UE to support machine learning-based beam prediction performed on the network entity side. In some aspects, a signaling framework is disclosed to allow the network entity flexible control over the scope of data collected by the UE, such as the quantity, frequency, and timing of beam quality measurements. In some aspects, when the UE performs a receive beam scan to receive a downlink reference signal, scheduling constraints may exist to suppress the network entity from transmitting other downlink signals that temporally overlap with the downlink reference signal. In some aspects, without such scheduling constraints, the UE can determine the receive beam based on priority rules to apply the receive beam corresponding to the downlink signal with the highest priority. In some aspects, the network entity can configure the content and temporal behavior of CSI reports.

[0044] In some embodiments, the UE may report the L1-RSRP of the downlink reference signal and the index of the optimal beam for the configured CMR based on the L1-RSRP, L1-SINR, and / or the assumed measurement error of the configured CMR that meets one or more criteria. Advantageously, the data collection techniques described herein can support model training, refinement, and monitoring of machine learning models used for beam management. The collected data can improve the performance and prediction accuracy of the machine learning model, thereby allowing network entities to select better beams to improve system performance.

[0045] Figure 4This is a signaling diagram 400 illustrating communication between a UE 102 and a network entity 104 according to one embodiment, in which the UE 102 collects data for supporting network-side machine learning-based beam management. The network entity 104 may correspond to a base station or a unit of a base station (such as RU 106, DU 108, CU 110, etc.).

[0046] UE 102 may transmit (or receive from UE 104) information regarding the UE's capabilities related to supported configurations for data collection used for beam prediction. In one implementation, this capability information may include supported configurations for beam reporting, such as the maximum number of optimal beams to report, the maximum number of time reporting instances for downlink reference signals, the maximum number of beam measurements per time reporting instance, minimum processing for beam reporting, supported time-domain behavior for beam reporting, etc. Network entity 104 may configure data collection based on the UE's capability information.

[0047] Network entity 104 can transmit 404 (or UE 102 can receive 404) control signaling from network entity 104 to configure beam measurements for one or more CMR sets and beam measurement-based reporting. The reports can represent data collection for machine learning-based beam management. In one implementation, network entity 104 can communicate via Radio Resource Control (RRC) signaling (e.g., RRCReconfiguration The network entity can configure at least one CSI report configuration for data collection, wherein it can configure at least one downlink reference signal set (e.g., SSB or CSI-RS) as CMR. The network entity can further configure the number of reports for data collection via RRC signaling. This number of reports can include the content and time-domain behavior of the CSI reports, such as the number, frequency, and timing of beam quality measurements. For periodic reports, the network entity can further configure the periodicity and time slot offset of the reports.

[0048] Network entity 104 may transmit a 406 trigger signal to UE 102 (or UE 102 may receive a 406 trigger signal from network entity 104) to trigger a configured set of downlink reference signals and / or beam reporting for data collection. For example, for semi-persistent or aperiodic reporting, network entity 104 may transmit a Media Access Control (MAC) element (CE) or downlink control information (DCI) (e.g., in a PDCCH) to trigger beam reporting. For semi-persistent or aperiodic downlink reference signals, network entity 104 may transmit a MAC CE or DCI to trigger the downlink reference signal.

[0049] UE 102 can determine the receive beam 408 used to receive the downlink reference signal. In one implementation, UE 102 can perform a receive beam scan to identify the optimal UE receive beam for receiving the downlink reference signal used for data collection. If the downlink reference signal overlaps with other downlink signals in the time domain, UE 102 can determine whether and how to receive the downlink reference signal to identify the optimal beam. In some aspects, when UE 102 performs a receive beam scan to receive the downlink reference signal, there may be scheduling constraints to suppress network entities from transmitting other downlink signals that overlap with the downlink reference signal in time. In some aspects, if there are no such scheduling constraints, UE 102 can determine the receive beam based on priority rules to apply the receive beam corresponding to the downlink signal with the highest priority.

[0050] Network entity 104 may transmit a downlink reference signal 410 (or UE 102 may receive 410) via one or more configured CMR sets for UE 102 to perform beam rate measurements. UE 102 may transmit beam reports 412 (or network entity 104 may receive 412) to network entity 104 based on a configured number of reports and the received downlink reference signal. Beam reports can provide beam quality data corresponding to one or more CMR sets to support machine learning-based beam management.

[0051] The following is a detailed discussion of how UE 102 can be configured to measure the beam of one or more CMR sets to collect data for supporting machine learning-based beam prediction on the network side. In one aspect, network entity 104 can configure UE 102 to collect both input and output data for a machine learning model in a single beam report based on a single CMR set. In another aspect, network entity 104 can configure UE 102 to collect model input and model output data in a single beam report based on multiple separate CMR sets. In yet another aspect, network 104 can configure UE 102 to collect model input and model output data in separate beam reports based on multiple separate CMR sets.

[0052] Figure 5Example 500 of a beam report is shown, according to one embodiment, when the joint input / output beam report for a machine learning model includes a subset of beam quality measurements from a set of channel measurement resources (CMRs) and a beam index of the best beam from the CMRs. Network entity 104 may configure UE 102 to collect beam measurements of both input and output data for the machine learning model based on a single set of CMRs. In one aspect, network entity 104 may use RRC signaling to configure UE 102 to generate beam reports, such as CSI reports.

[0053] A single CMR set may include downlink reference signals transmitted on a spatial array of 4 beams in the ZoD dimension and 8 beams in the AoD dimension of a total of 32 beams (510). UE 102 may perform beam quality measurements on all 32 beams (510), but network entity 104 may configure a CMR reporting subset limit to request UE 102 to report the beam quality of a subset of the CMR beams.

[0054] Figure 5 The CMR report subset limitation indication is shown as a 4-beam subset (520). The network entity can further configure the UE to report N beam indices corresponding to the optimal beam, such as SSB Resource Indicators (SSBRIs) or CSI-RS Resource Indicators (CRIs). In one implementation, the value of N can be predefined, for example, N=1. In another implementation, network entity 104 can configure the value of N via RRC signaling (e.g., RRC parameters in the CSI report configuration). In one implementation, UE 102 can report the maximum value of N via UE capability information. Figure 5 The diagram shows network entity 104 configuring UE 102 to report the beam index (N=1) for the best beam. In one implementation, network entity 104 may further configure UE 102 to report the beam quality corresponding to the N best beams.

[0055] UE 102 can report beam quality 530 for the four beams (520) as indicated by the CMR report subset limitation. UE 102 can evaluate beam quality measurements for all 32 beams to determine the beam with optimal beam quality. The beam report can further include a beam index for the optimal beam 540. The beam quality of the optimal beam 540 may not be part of the beam quality 530 for the four beams (520) requested by the CMR report subset limitation. In this case, UE 102 can report the beam quality of the optimal beam 540.

[0056] Network entity 104 can transmit beams carrying downlink reference signals with CMR over multiple time instances, such as by periodically transmitting beams. In one implementation, for time beam prediction, network entity 104 can configure UE 102 to report the beam quality of N beam indices, N best beams, and / or the beam quality of a subset of CMRs configured in the CMR subset limit for each of X time reporting instances. For example, network entity 104 can configure UE 102 to report the beam quality of a subset of CMRs and the beam indices of N beams for each of X periodic or aperiodic time reporting instances. In one implementation, X can be predefined, for example, X = 4. In another implementation, network entity 104 can configure X via RRC signaling (e.g., RRC parameters in the CSI reporting configuration). In one implementation, UE 102 can report the maximum value of X via UE capability information. In another implementation, UE 102 can report X via UE capability information. The maximum value of N.

[0057] Figure 6 Example 600 of a beam report 660 based on beam measurements of CMR at multiple reporting instances prior to reference time 650, according to an embodiment, is shown. The CMR may be periodic and is represented as CMR instance 1 (610), CMR instance 2 (620), CMR instance 3 (630), and CMR instance 4 (640). Network entity 104 may configure UE 102 to collect data from three time reporting instances of the CMR prior to reference time 650 (e.g., CMR instance 2 (620), CMR instance 3 (630), and CMR instance 4 (640)). These time reporting instances of the CMR to which UE 102 reports beam measurement data may be referred to as reporting instances.

[0058] In one implementation, the reference time 650 may be predefined; for example, the reference time may occur before the latest possible occurrence of the CMR determined from the shortest processing delay of the beam report for data collection, prior to the first symbol of the beam report. In one implementation, the shortest processing delay may be predefined (e.g., Z1 and Z1′ as defined in Section 5.4 of 3GPP TS 38.214). In one implementation, network entity 104 may configure the shortest processing time. In one implementation, UE 102 may report the shortest processing time via UE capability information. In one implementation, network entity 104 may configure the reference time via RRC signaling, MAC CE, or DCI. In one implementation, network entity 104 may configure the reference time as an offset before the first symbol of the beam report for data collection, or an offset after the PDCCH or MAC CE triggers the beam report for data collection. In one implementation, UE 102 may report the reference time in the beam report. In one implementation, UE 102 can report the measurement instance of the reported beam or the measured time slot index in the beam report.

[0059] In one implementation, to improve measurement accuracy, network entity 104 can configure the number of average instances (or the number of measurement instances per time-reporting instance) Y and the number of time-reporting instances X. UE 102 can determine the beam quality of the time-reporting instances based on the measurements of Y instances of CMR, such as by averaging the beam measurements of Y measurement instances. In one implementation, UE 102 can report the maximum value of Y (e.g., the maximum value of measurement instances for each time-reporting instance) and X via UE capability information. The maximum value of Y (e.g., the maximum value of all measurement instances for all time reporting instances), Y N (e.g., the maximum product of the measurement instances for each time-reporting instance and the number of optimal beams to report), X N (e.g., the maximum product of time reporting instances and the number of optimal beams to be reported), and / or X Y N (e.g., the maximum product of the total number of measurement instances for all time reporting instances and the number of the best beams to be reported).

[0060] Figure 7Example 700 is shown, according to an embodiment, of beam reporting 790 based on averaging beam measurements of CMR for each of a plurality of reporting instances prior to reference time 780. CMR may be periodic and is represented as CMR instance 1 (710), CMR instance 2 (720), CMR instance 3 (730), and CMR instance 4 (740), CMR instance 5 (750), CMR instance 6 (760), and CMR instance 7 (770). Network entity 104 may configure UE 102 to collect data from three time reporting instances of CMR prior to reference time 780. The data collected for each time reporting instance may be based on an averaging window of two measurement instances of CMR. For example, the first time reporting instance may be based on the average window 725 of the beam measurements of CMR instance 2 (720) and CMR instance 3 (730); the second time reporting instance may be based on the average window 745 of the beam measurements of CMR instance 4 (740) and CMR instance 5 (750); and the third time reporting instance may be based on the average window 765 of the beam measurements of CMR instance 6 (760) and CMR instance 7 (770).

[0061] In one implementation, network entity 104 can configure the number of measurement instances Y in each average window (e.g., the number of average instances). In another implementation, network entity 104 can configure the total number of measurement instances Z and the number of time reporting instances X. UE 102 can then derive the number of average instances Y = Z / X. In yet another implementation, network entity 104 can configure the total number of measurement instances Z and the number of average instances Y. UE can then derive the number of time reporting instances X = Z / Y.

[0062] In one aspect, network entity 104 can configure UE 102 to collect model input and model output data in a single beam report based on multiple separate CMR sets. The beams of these multiple separate CMR sets can have different beamwidths. For example, the SSB configured in the first CMR set can have a relatively wide beamwidth, and the CSI-RS configured in the second CMR set can have a more directional beam. A machine learning model for beam management can apply beam quality measurements from the first CMR set as input training data and the beam index of the best beam from the second CMR set as output training data, allowing the machine learning model to predict the optimal narrower beam based on the wider beam measurements.

[0063] Figure 8Example 800 of a beam report is illustrated according to an embodiment, where a network entity configures two CMR sets for beam measurements and a joint input / output beam report for a machine learning model includes a subset of beam quality measurements from a first CMR set and a beam index of the best beam from a second CMR set. Network entity 104 may configure UE 102 to collect beam measurements for input and output data used in a machine learning model, based on the first and second CMR sets, respectively.

[0064] The first CMR set may include downlink reference signals transmitted on a spatial array of 4 beams in the ZoD dimension and 8 beams in the AoD dimension, totaling 32 beams. UE 102 may perform beam quality measurements on all 32 beams, but network entity 104 may configure a CMR reporting subset limit to request UE 102 to report the beam quality of a 4-beam subset (810) of the first CMR set.

[0065] The second CMR set may also include downlink reference signals transmitted on a spatial array of 32 beams (820) in total, consisting of 4 beams in the ZoD dimension and 8 beams in the AoD dimension. In one implementation, the first and second CMR sets may have different beam patterns (e.g., the two CMR sets have different azimuth / elevation widths). Network entity 104 may also configure UE 102 to report the N beam indices—SSBRI or CRI—of the best beam in the second CMR set. In one implementation, the value of N may be predefined, for example, N=1. In another implementation, network entity 104 may configure the value of N via RRC signaling (e.g., RRC parameters in the CSI reporting configuration). In one implementation, UE 102 may report the maximum value of N via UE capability information. Figure 8 The diagram shows network entity 104 configuring UE 102 to report the beam index (N=1) of the best beam. In one implementation, network entity 104 may further configure UE 102 to report beam quality corresponding to the N best beams.

[0066] UE 102 can report beam quality 830 for four beams (810) of a first CMR set indicated by CMR report subset restrictions. UE 102 can perform beam quality measurements on all 32 beams (820) of a second CMR set. UE 102 can evaluate the beam quality measurements on all 32 beams (820) of the second CMR set to determine the beam with optimal beam quality. The beam report can further include a beam index 840 of the optimal beam from the second CMR set. In one implementation, UE 102 can report the beam quality 840 of the optimal beam.

[0067] In one implementation, for time beam prediction, network entity 104 can configure UE 102 to report N beam indices, beam quality of N best beams, and / or beam quality of CMR subset constraints for each of X time report instances for beams in a first CMR set and a second CMR set. The number X of time report instances for the two CMR sets can be common or separate. In one implementation, X can be predefined, for example, X = 1 for the first CMR set. In another implementation, network entity 104 can configure X for the second CMR set via RRC signaling (e.g., RRC parameters in the CSI report configuration). In one implementation, UE 102 can report the maximum value of X for the first CMR set and / or the second CMR set via UE capability information. In one implementation, UE 102 can report X via UE capability information. The maximum value of N.

[0068] UE 102 can measure and report beam quality for X time report instances of a first CMR set or a second CMR set prior to a reference time. The reference time 650 can be predefined; for example, it can occur before the latest possible occurrence of a CMR determined by the shortest processing delay of the beam report used for data collection, prior to the first symbol of the beam report. In one implementation, the shortest processing delay can be predefined. In one implementation, network entity 104 can configure the shortest processing time. In one implementation, UE 102 can report the shortest processing time via UE capability information. In one implementation, network entity 104 can configure the reference time via RRC signaling, MAC CE, or DCI. In one implementation, network entity 104 can configure the reference time as an offset before the first symbol of the beam report used for data collection, or an offset after the PDCCH or MAC CE triggers the beam report used for data collection. In one implementation, UE 102 can report the reference time in the beam report. In one implementation, UE 102 can report the measurement instance of the reported beam or the measured time slot index in the beam report.

[0069] In one aspect, network 104 can configure UE 102 to collect model input and model output data based on multiple separate CMR sets in separate beam quality reports. Network entity 104 can configure separate CSI report configurations for collecting beam quality data for model inputs and outputs used for machine learning models via RRC signaling. Network entity 104 can configure the CMR set for each CSI report configuration.

[0070] Figure 9 Example 900 of a beam report is shown, according to an embodiment, when a network entity configures two CMR sets for beam measurement and separate beam reports for input and output data of a machine learning model include a subset of beam quality measurements from a first CMR set and a beam index of the best beam from a second CMR set. Network entity 104 can configure separate CSI report configurations for these two CMR sets.

[0071] In one implementation, for the first CSI report configuration, network entity 104 may configure a CMR report subset limit to request UE 102 to report the beam quality of a 4-beam subset (910) of the first CMR set. The first CMR set may include downlink reference signals transmitted on a spatial array of 4 beams in the ZoD dimension and 8 beams in the AoD dimension, totaling 32 beams.

[0072] For the second CSI reporting configuration, network entity 104 can configure UE 102 to report the N beam indices—SSBRI or CRI—of the best beam in the second CMR set. The second CMR set may also include downlink reference signals transmitted on a spatial array of 4 beams in the ZoD dimension and 8 beams in the AoD dimension, totaling 32 beams (920). In one implementation, the first and second CMR sets may have different beam patterns (e.g., the two CMR sets have different azimuth / elevation widths). In one implementation, the value of N may be predefined, for example, N=1. In another implementation, network entity 104 can configure the value of N via RRC signaling (e.g., RRC parameters in the CSI reporting configuration). In one implementation, UE 102 can report the maximum value of N via UE capability information. Figure 8 The diagram shows network entity 104 configuring UE 102 to report the beam index (N=1) of the best beam. In one implementation, network entity 104 may further configure UE 102 to report beam quality corresponding to the N best beams.

[0073] UE 102 can report beam quality 930 for four beams (910) of a first CMR set indicated by a CMR report subset limitation configured via a first CSI report configuration. UE 102 can perform beam quality measurements on all 32 beams (920) of a second CMR set configured via a second CSI configuration. UE 102 can evaluate the beam quality measurements on all 32 beams (920) of the second CMR set to determine the beam with optimal beam quality. The beam report can further include a beam index of the optimal beam 940 from the second CMR set. In one implementation, UE 102 can report the beam quality of the optimal beam 940.

[0074] In one implementation, for time beam prediction, network entity 104 can configure UE 102 to report beam quality for each of X time reporting instances, including N beam indices, N best beams, and / or CMR subset-limited beam quality, for beams in a first CMR set configured by a first CSI reporting configuration and a second CMR set configured by a second CSI configuration. The UE can measure and report the beam quality of X time reporting instances of the first or second CMR set prior to a reference time. For example, for... Figure 8 As discussed, UE 102 can determine the number X of time report instances and the reference time; for the sake of brevity, its details will not be repeated.

[0075] In one aspect, when UE 102 performs a receive beam scan to receive a downlink reference signal configured by CMR for data collection, scheduling constraints may exist to suppress network entity 104 from transmitting other downlink signals that overlap with the downlink reference signal in the time domain. For example, when the UE receive beam is applicable (e.g., Quasi-Cooperative Positioning (QCL) Type D indicating spatial reception parameters is applicable), network entity 104 may suppress the transmission of other downlink reference signals that overlap with the downlink reference signal for data collection in the time domain in the same component carrier (CC) or different CCs within the same frequency band or combination of frequency bands, because UE 102 needs to perform a beam scan.

[0076] In one implementation, network entity 104 may suppress the transmission of downlink reference signals with QCL-TypeD properties that are different from those of downlink reference signals that overlap with downlink reference signals used for data collection in the same component carrier (CC) or different CCs within a frequency band or combination of frequency bands.

[0077] In one implementation, network entity 104 can configure a measurement window, via RRC signaling, MAC CE, or DCI, to indicate when UE 102 can perform measurements of downlink reference signals for data collection. In another implementation, network entity 104 can configure the periodicity, slot offset, and duration of the measurement window. The aforementioned scheduling restrictions then apply only when the UE needs to measure the downlink reference signals for data collection.

[0078] Figure 10 Example 1000 of beam measurement of CMR according to an embodiment is shown, based on a configured measurement window that instructs the UE to measure CMR to generate a beam report. CMR may be periodic and is represented as CMR instance 1 (1011), CMR instance 2 (1012), CMR instance 3 (1023), CMR instance 4 (1024), CMR instance 5 (1015), CMR instance 6 (1016), CMR instance 7 (1027), and CMR instance 8 (1028).

[0079] Network entity 104 can configure measurement window 1030 to be open for UE 102 to measure CMR instance 1 (1011) and CMR instance 2 (1012); similarly, network entity 104 can configure measurement window 1050 to be open for UE 102 to measure CMR instance 5 (1015) and CMR instance 6 (1016). During the periods when measurement windows 1030 and 1050 are open, network entity 1014 can implement scheduling restrictions 1010 to suppress the transmission of other downlink reference signals.

[0080] In contrast, network entity 104 can configure measurement window 1040 to be closed during CMR instance 3 (1023) and CMR instance 4 (1024); similarly, network entity 104 can configure measurement window 1060 to be closed during CMR instance 7 (1027) and CMR instance 8 (1028). During the periods when measurement windows 1040 and 1060 are closed, network entity 1014 can transmit CMR instances without any scheduling restrictions 1020.

[0081] In one aspect, when a UE receive beam is applicable (e.g., QCL-Type D is applicable), if another downlink reference signal exists in the time domain among the same component carrier (CC) or different CCs within a frequency band or band combination that overlaps with the downlink reference signal used for data collection, the UE 102 can determine the receive beam based on a priority rule. Therefore, the UE 102 can apply the receive beam corresponding to the downlink signal with the highest priority.

[0082] In one implementation, network entity 104 can configure the priority of downlink signals via RRC signaling, MAC CE, or DCI. In another implementation, the priority of downlink signals can be predefined. Priorities can be determined based on channel type, channel configuration (e.g., channel search space type, channel temporal behavior, channel triggering behavior), etc. In one implementation, the priority from high to low can be defined as: PDCCH in the common search space > PDSCH scheduled by the PDCCH in the common search space > PDCCH in the UE-specific search space > PDSCH scheduled by the UE-specific search space > downlink reference signal for data collection > aperiodic CSI-RS > semi-persistent CSI-RS. In other implementations, network entity 104 and UE 102 can determine different priority orders for downlink signals. For example, if there are overlapping downlink signals (e.g., PDSCH) with higher priority, UE 102 may not measure the downlink reference signal to report beam quality measurements for data collection.

[0083] In one implementation, UE 102 may suppress the reception of downlink signals having a QCL type attribute different from the determined QCL-TypeD of the downlink signal with the highest priority. In another implementation, UE 102 may receive all downlink signals based on the determined QCL type of the downlink signal with the highest priority.

[0084] In one aspect, the temporal behavior of beam reports providing collected data may include: aperiodic reporting, semi-persistent reporting, or periodic reporting. Network entity 104 may configure temporal behavior for beam reports. In one implementation, network entity 104 may suppress the configuration of one or more temporal behaviors for reports (e.g., aperiodic reporting). In one implementation, UE 102 may report UE capabilities indicating supported temporal behaviors for beam reports used for data collection.

[0085] In one implementation, UE 102 can transmit beam reports for data collection via at least one PUCCH resource configured by network entity 104 via RRC signaling, MAC CE, or DCI. In another implementation, UE 102 can transmit beam reports for data collection via a PUSCH configured by network entity 104 via RRC signaling, MAC CE, or DCI. In yet another implementation, UE 102 can transmit beam reports for data collection as uplink control information multiplexed on the PUSCH. In yet another implementation, UE 102 can transmit beam reports via MAC CE.

[0086] In one aspect, for the content of the beam report providing the collected data, network entity 104 can configure UE 102 to report the L1-RSRP for each beam. In one implementation, UE 102 can report the absolute L1-RSRP for each beam. If UE 102 is configured to report the L1-RSRP for a configured subset of CMRs or one or more sets of CMRs, UE 102 can report the L1-RSRP based on the order of beam indices within the CMR subset or set. Table 1 shows an example of absolute L1-RSRP reporting for all configured CMRs in a subset or set.

[0087] Table 1: Example of absolute L1-RSRP report for all K configured CMRs in a subset or set.

[0088] In one implementation, UE 102 can report the absolute L1-RSRP of the best beam and the differential L1-RSRP for the remaining beams. If UE 102 is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, UE 102 can report the beam index indicating the beam index with the strongest L1-RSRP and the strongest L1-RSRP, and can report the differential L1-RSRP based on the order of the remaining beam indices within the CMR subset or set. Table 2 shows an example of absolute plus differential L1-RSRP reporting for all configured CMRs in a subset or set.

[0089] Table 2: Example of L1-RSRP report based on absolute difference for all K configured CMRs in a subset or set.

[0090] In one implementation, UE 102 may determine to report L1-RSRPs for one or more CMRs if a specific criterion for the CMR is met. This criterion may include at least one of the following: the measured L1-RSRPs of one or more CMRs are higher than a first threshold; the measured L1-SINRs of one or more CMRs are higher than a second threshold; or the assumed measurement error of one or more CMRs is lower than a third threshold. In one implementation, the first, second, and third thresholds may be predefined. In one implementation, network entity 104 may configure the first, second, and third thresholds via RRC signaling, MAC CE, or DCI. In one implementation, UE 102 may determine the assumed measurement error based on the measured L1-SINR and its receiving algorithm. In one implementation, UE 102 may report an indicator to indicate the number of reported L1-RSRPs higher than the first threshold.

[0091] In one implementation, UE 102 may determine to report the L1-RSRP of all configured CMRs if certain criteria are met. These criteria may include at least one of the following: the measured L1-RSRP of one or more CMRs is higher than a first threshold; the measured L1-SINR of the CMR is higher than a second threshold; or the assumed measurement error of the CMR is lower than a third threshold.

[0092] In one implementation, for beam reports that provide beam quality, network entity 104 can configure UE 102 to report L1-RSRP and L1-SINR for each beam. Network entity 104 can then determine, based on the received L1-SINR, whether to use the reported L1-RSRP for model training, refinement, monitoring, and other purposes.

[0093] In one implementation, UE 102 may report the absolute L1-RSRP and absolute L1-SINR for each beam. If the UE is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, then UE 102 may report the L1-RSRP and associated L1-SINR based on the order of beam indices within the CMR subset or set of CMRs.

[0094] In one implementation, UE 102 can report the absolute L1-RSRP of the best beam and the differential L1-RSRP of the remaining beams. The UE can report the absolute L1-SINR for each beam. If the UE is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, UE 102 can report the beam index indicating the beam index with the strongest L1-RSRP, and the strongest L1-RSRP, and can report the differential L1-RSRP based on the order of the remaining beam indices within the CMR subset or set.

[0095] In one implementation, UE 102 can report the absolute L1-RSRP and L1-SINR of the best beam, as well as the differential L1-RSRP and L1-SINR of the remaining beams. If the UE is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, the UE can report the beam index indicating the beam index with the strongest L1-RSRP, along with the L1-RSRP and associated L1-SINR, and can report the differential L1-RSRP and L1-SINR based on the order of the remaining beam indices within the CMR subset or set.

[0096] In one implementation, for each CMR, UE 102 may report absolute and / or differential L1-RSRP and an indicator indicating whether L1-SINR is above a threshold. In one implementation, the threshold may be predefined, such as 3dB, or configured by network entity 104 via RRC signaling.

[0097] In one implementation, for beam reports that provide beam quality, network entity 104 can configure UE 102 to report L1-RSRP and hypothetical measurement error for each beam. Network entity 104 can then determine, based on the received hypothetical measurement error, whether to use the reported L1-RSRP for model training, refinement, monitoring, and other purposes.

[0098] In one implementation, UE 102 can report the absolute L1-RSRP and absolute assumed measurement error for each beam. If the UE is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, UE 102 can report the L1-RSRP and assumed measurement error based on the order of beam indices within the CMR subset or set of CMRs.

[0099] In one implementation, UE 102 can report the absolute L1-RSRP of the best beam and the differential L1-RSRP of the remaining beams. UE 102 can report the absolute assumed measurement error for each beam. If the UE is configured to report the L1-RSRP of a configured subset of CMRs or one or more sets of CMRs, UE 102 can report the beam index indicating the beam index with the strongest L1-RSRP, along with the L1-RSRP, and can report the differential L1-RSRP based on the order of the remaining beam indices within the CMR subset or set.

[0100] In one implementation, for each CMR, UE 102 may report absolute and / or differential L1-RSRP and an indicator indicating whether the assumed measurement error is below a threshold. In one implementation, the threshold may be predefined, such as 3 dB, or configured by network entity 104 via RRC signaling.

[0101] Figures 11 to 12 The following diagram illustrates the implementation. Figures 2 to 10 One or more aspects of the method. Specifically, Figure 11 The UE 102 is shown to be paired with Figures 2 to 10 The implementation of one or more aspects. Figure 12 104 pairs of network entities are shown. Figures 2 to 10 The implementation of one or more aspects.

[0102] Figure 11 This is a flowchart 1100 of a method at the UE for receiving CMR and reporting beam measurements of CMR in a beam report to support network-side machine learning-based beam management wireless communication, according to an embodiment. (See reference...) Figure 1 , Figure 4 and Figure 13 The method can be executed by UE 102, UE device 1302, etc., which may include memories 1326', 1306', 1316, and may correspond to the entire UE 102 or the entire UE device 1302, or components of UE 102 or UE device 1302 (such as wireless baseband processor 1326 and / or application processor 1306).

[0103] The UE transmits 1102 information to the network entity regarding the UE capabilities supported for beam reporting configurations used for data collection associated with machine learning-based beam management. For example, refer to... Figure 4UE 102 transmits 402 information to network entity 104 regarding UE capabilities for supported configurations of beam reporting for data collection. In one implementation, this capability information may include supported configurations for beam reporting, such as the maximum number of optimal beams to report, the maximum number of time reporting instances for downlink reference signals, the maximum number of beam measurements per time reporting instance, minimum processing for beam reporting, supported time-domain behavior for beam reporting, etc. Network entity 104 can configure data collection based on the UE's capability information.

[0104] The UE receives control signaling 1104 from the network entity for configuring the Channel Measurement Resources (CMR) set and for reporting on data collection associated with machine learning-based beam management. For example, refer to... Figure 4 UE 102 receives control signaling 404 from network entity 104 configuring beam measurements of one or more Channel Measurement Resources (CMR) sets and beam reports based on these beam measurements. In one implementation, the report may represent data collection for machine learning-based beam management. In another implementation, network entity 104 may use RRC signaling (e.g., RRCReconfiguration The network entity can configure at least one CSI report configuration for data collection, wherein it can configure at least one set of downlink reference signals (e.g., SSB or CSI-RS) as CMR. The network entity can further configure the number of reports for data collection via RRC signaling. This number of reports can include the content and time-domain behavior of the CSI reports, such as the number, frequency, and timing of beam quality measurements. For periodic reports, the network entity can further configure the periodicity and time slot offset of the reports.

[0105] The UE receives 1106 from the network entity for receiving the CMR set or for triggering the transmission of a report. For example, refer to Figure 4 UE 102 receives from network entity 104 406 a triggering signaling for receiving one or more CMR sets or for transmitting beam reports. In one implementation, for semi-persistent or aperiodic reports, network entity 104 may transmit MAC CE or DCI to trigger beam reports. For semi-persistent or aperiodic downlink reference signals, network entity 104 may transmit MAC CE or DCI to trigger downlink reference signals.

[0106] The UE receives the 1110 downlink reference signal from the network entity via this CMR set. For example, the reference... Figure 4 UE 102 receives from network entity 104 one or more sets of CMRs to be used for beam measurement. These CMRs may include downlink reference signals such as SSB and / or CSI-RS.

[0107] The UE transmits a report to the network entity based on beam measurements corresponding to the downlink reference signal. For example, the reference... Figure 4 The UE transmits a beam report based on beam measurement 412 to network entity 104 to provide beam quality data corresponding to one or more CMR sets for supporting machine learning-based beam management. In one implementation, the beam report may be based on a configured number of reports and the received downlink reference signal.

[0108] Figure 11 A method is described from the UE side of the wireless communication link, while Figure 12 A method is described from the network side of the wireless communication link.

[0109] Figure 12 This is a flowchart 1200 of a method for transmitting CMR and receiving beam measurements of CMR in a beam report at a network entity according to an embodiment to support wireless communication based on machine learning-based beam management on the network side. (See reference...) Figure 1 , Figure 4 and Figure 14 The method can be performed by one or more network entities 104, which may correspond to a base station or a unit of a base station (such as RU 106, DU 108, CU 110, RU processor 1406, DU processor 1426, CU processor 1446, etc.). One or more network entities 104 may include memories 1406' / 1426' / 1446', which may correspond to the entirety of one or more network entities 104, or components of one or more network entities 104 (such as RU processor 1406, DU processor 1426, or CU processor 1446).

[0110] The network entity receives UE capability information from the UE (1202) regarding supported configurations for beam reporting used for data collection associated with machine learning-based beam management. For example, refer to... Figure 4 Network entity 104 receives from UE 102 402 information about UE capabilities supported for beam reporting for data collection. In one implementation, this capability information may include supported configurations for beam reporting, such as the maximum number of optimal beams to report, the maximum number of time reporting instances for downlink reference signals, the maximum number of beam measurements per time reporting instance, minimum processing for beam reporting, supported time-domain behavior for beam reporting, etc. Network entity 104 can configure data collection based on the UE's capability information.

[0111] The network entity transmits control signaling 1204 to the UE for configuring the Channel Measurement Resources (CMR) set and for reporting on data collection associated with machine learning-based beam management. For example, refer to... Figure 4 Network entity 104 transmits control signaling 404 to UE 102 to configure beam measurements and beam reports based on one or more Channel Measurement Resources (CMR) sets. In one implementation, the report may represent data collection for machine learning-based beam management. In another implementation, network entity 104 may transmit control signaling via RRC (e.g., RRCReconfiguration This allows the network entity to configure at least one CSI report configuration for data collection, whereby the network entity can configure at least one downlink reference signal set (e.g., SSB or CSI-RS) as CMR. The network entity can further configure the number of reports for data collection via RRC signaling. The number of reports can include the content and time-domain behavior of the CSI reports, such as the number, frequency, and timing of beam quality measurements. For periodic reports, the network entity can further configure the periodicity and time slot offset of the reports.

[0112] The network entity transmits 1206 to the UE for receiving the CMR set or for triggering the transmission of the report. For example, refer to Figure 4 Network entity 104 transmits signaling 406 to UE 102 for receiving one or more CMR sets or for transmitting beam reports. In one implementation, for semi-persistent or aperiodic reports, network entity 104 may transmit MAC CE or DCI to trigger beam reports. For semi-persistent or aperiodic downlink reference signals, network entity 104 may transmit MAC CE or DCI to trigger downlink reference signals.

[0113] The network entity transmits the 1210 downlink reference signal to the UE via this CMR set. For example, reference Figure 4 Network entity 104 transmits to UE 102 one or more sets of CMRs to be used for beam measurement. These CMRs may include downlink reference signals such as SSB and / or CSI-RS.

[0114] The network entity receives a report from the UE based on beam measurements corresponding to the downlink reference signal. For example, the reference... Figure 4 Network entity 104 receives 412 beam reports based on beam measurements from UE 102 to provide beam quality data corresponding to one or more CMR sets for supporting machine learning-based beam management. In one implementation, the beam report may be based on a configured number of reports and the received downlink reference signal.

[0115] like Figure 13 As described herein, UE device 1302 can execute the method of flowchart 1100. For example... Figure 14 As described in the diagram, one or more network entities 104 can execute the methods of flowchart 1200.

[0116] Figure 13 This is a diagram 1300 illustrating a hardware implementation of an example UE device 1302 according to some embodiments. The UE device 1302 may be a UE 102, a component of UE 102, or may implement UE functionality. The UE device 1302 may include an application processor 1306, which may have on-chip memory 1306'. In the example, the application processor 1306 may be coupled to a secure digital (SD) card 1308 and / or a display 1310. The application processor 1306 may also be coupled to a sensor module 1312, a power supply 1314, an additional memory module 1316, a camera 1318, and / or other related components. For example, the sensor module 1312 may control a barometer / altimeter, motion sensors (such as an inertial management unit (IMU), gyroscope, accelerometer), a light detection and ranging (LIDAR) device, a radio-assisted detection and ranging (RADAR) device, a sound navigation and ranging (SONAR) device, a magnetometer, an audio device, and / or other technologies for positioning.

[0117] The UE device 1302 may further include a wireless baseband processor 1326, which may be referred to as a modem. The wireless baseband processor 1326 may have on-chip memory 1326'. Together with and similarly to the application processor 1306, the wireless baseband processor 1326 may also be coupled to a sensor module 1312, a power supply 1314, an additional memory module 1316, a camera 1318, and / or other related components. The wireless baseband processor 1326 may additionally be coupled to one or more Subscriber Identity Module (SIM) cards 1320 and / or one or more transceivers 1330 (e.g., wireless RF transceivers).

[0118] Within one or more transceivers 1330, the UE device 1302 may include a Bluetooth module 1332, a WLAN module 1334, an SPS module 1336 (e.g., a GNSS module), and / or a cellular module 1338. The Bluetooth module 1332, WLAN module 1334, SPS module 1336, and cellular module 1338 may each include an on-chip transceiver (TRX), or in some cases, only a transmitter (TX) or only a receiver (RX). The Bluetooth module 1332, WLAN module 1334, SPS module 1336, and cellular module 1338 may each include a dedicated antenna and / or utilize antenna 1340 to communicate with one or more other nodes. For example, UE device 1302 can communicate with another UE (e.g., sidelink communication) and / or with network entity 104 (e.g., uplink / downlink communication) via transceiver 1330 and antenna 1340, where network entity 104 may correspond to a base station or a unit of a base station, such as RU 106, DU 108 or CU 110.

[0119] The wireless baseband processor 1326 and application processor 1306 may each include computer-readable media / memory 1326' and 1306', respectively. An additional memory module 1316 may also be considered a computer-readable media / memory. Each computer-readable media / memory 1326', 1306', and 1316 may be non-transitory. The wireless baseband processor 1326 and application processor 1306 may each be responsible for general processing, including executing software stored on the computer-readable media / memory 1326', 1306', and 1316. When executed by the wireless baseband processor 1326 / application processor 1306, this software causes the wireless baseband processor 1326 / application processor 1306 to perform the various functions described herein. The computer-readable media / memory may also be used to store data manipulated by the wireless baseband processor 1326 / application processor 1306 during software execution. The wireless baseband processor 1326 / application processor 1306 may be a component of UE 102. UE device 1302 may be a processor chip (e.g., a modem and / or an application) and includes only the wireless baseband processor 1326 and / or the application processor 1306. In other examples, UE device 1302 may be the entire UE 102 and may include additional modules for device 1302.

[0120] As in Figure 1 The discussion and about Figure 11The implemented data collection component 140 for network-side machine learning-based beam management (also known as ML data collection component 140) is configured to receive signaling from the network entity for configuring a set of channel measurement resources (CMR) and for reports on data collection associated with machine learning-based beam management; receive downlink reference signals from the network entity via the CMR set; and transmit reports to the network entity based on beam measurements corresponding to the downlink reference signals.

[0121] The ML data collection component 140 may be located within the application processor 1306 (e.g., at 140a), the wireless baseband processor 1326 (e.g., at 140b), or both the application processor 1306 and the wireless baseband processor 1326. The ML data collection components 140a to 140b may be one or more hardware components specifically configured to perform the stated process / algorithm, implemented by one or more processors configured to perform the stated process / algorithm, stored in a computer-readable medium for implementation by one or more processors, or a combination thereof.

[0122] Figure 1 This is a diagram 1400 illustrating a hardware implementation of one or more example network entities 104 according to some embodiments. The one or more network entities 104 may be a base station, a component of a base station, or may implement base station functions. The one or more network entities 104 may include or correspond to at least one of RU 106, DU 108, or CU 110. CU 110 may include a CU processor 1446, which may have on-chip memory 1446'. In some aspects, CU 110 may further include an additional memory module 1456 and / or a communication interface 1448, both of which may be coupled to the CU processor 1446. CU 110 may communicate with DU 108 via a midhaul link 162 (such as an F1 interface between the communication interface 1448 of CU 110 and the communication interface 1428 of DU 108).

[0123] DU 108 may include a DU processor 1426, which may have on-chip memory 1426'. In some aspects, DU 108 may further include an additional memory module 1436 and / or a communication interface 1428, both of which may be coupled to the DU processor 1426. DU 108 may communicate with RU 106 via a frontlink 160 between DU 108's communication interface 1428 and RU 106's communication interface 1408.

[0124] RU 106 may include an RU processor 1406, which may have on-chip memory 1406'. In some aspects, RU 106 may further include an additional memory module 1416, a communication interface 1408, and one or more transceivers 1430, all of which may be coupled to the RU processor 1406. RU 106 may further include an antenna 1440, which may be coupled to one or more transceivers 1430, enabling RU 106 to communicate with UE 102 via the antenna 1440 through one or more transceivers 1430.

[0125] On-chip memories 1406', 1426', 1446' and additional memory modules 1416, 1436, 1456 can each be considered as computer-readable media / memory. Each computer-readable medium / memory can be non-transitory. Each of processors 1406, 1426, 1446 is responsible for general processing, including executing software stored on the computer-readable medium / memory. When executed by the corresponding processor 1406, 1426, 1446, the software causes the processor 1406, 1426, 1446 to perform the various functions described herein. The computer-readable medium / memory can also be used to store data manipulated by the processors 1406, 1426, 1446 during software execution. In the example, the network-side machine learning-based beam management configuration component 150 can be located at any of one or more network entities 104, such as at CU 110; at both CU 110 and DU 108; at each of CU 110, DU 108 and RU 106; at DU 108; at both DU 108 and RU 106; or at RU 106.

[0126] like Figure 12 The discussion in the article and about ​ The implemented network-side machine learning-based beam management configuration component 150 (also known as ML data collection configuration component 150) is configured to transmit signaling to the UE for configuring a set of channel measurement resources (CMR) and reports for data collection associated with machine learning-based beam management; transmit downlink reference signals to the UE via the CMR set; and receive reports from the UE based on beam measurements corresponding to the downlink reference signals.

[0127] The ML data collection configuration component 150 may be located within one or more processors of one or more network entities 104, such as within an RU processor 1406 (e.g., at 150a), a DU processor 1426 (e.g., at 150b), and / or a CU processor 1446 (e.g., at 150c). The ML data collection configuration components 150a to 150c may be one or more hardware components specifically configured to execute the stated process / algorithm, implemented by one or more processors 1406, 1426, 1446 configured to execute the stated process / algorithm, stored in a computer-readable medium for implementation by one or more processors 1406, 1426, 1446, or a combination thereof.

[0128] The specific order or hierarchy of boxes in the processes and flowcharts disclosed herein is an example of the exemplary methods. Therefore, the specific order or hierarchy of boxes in the processes and flowcharts can be rearranged. Some boxes can also be combined or deleted. Dashed lines may indicate optional elements of the diagrams. The appended method claims present elements of various boxes in the exemplary order and are not limited to the specific order or hierarchy presented in the claims, processes, and flowcharts.

[0129] The detailed descriptions presented herein, in conjunction with accompanying drawings, depict various configurations, but do not represent the only configurations in which the concepts described herein can be practiced. These detailed descriptions include specific details used to provide a comprehensive explanation of the various concepts. However, these concepts can be practiced without using these specific details. In some cases, well-known structures and components are shown in block diagram form to avoid obscuring such concepts.

[0130] Aspects of wireless communication systems, such as telecommunications systems, are presented with reference to various devices and methods. These devices and methods are described in the following detailed description and illustrated in the accompanying drawings by various boxes, components, circuits, processes, call flows, systems, algorithms, etc. (collectively, "elements"). These elements can be implemented using electronic hardware, computer software, or a combination thereof. Whether such elements are implemented as hardware or software depends on the specific application and design constraints imposed on the overall system.

[0131] An element, or any part of an element, or any combination of elements, can be implemented as a “processing system” including one or more processors. 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, system-on-a-chip (SoCs), baseband processors, field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), state machines, gated logic, discrete hardware circuits, and other similar hardware configured to perform the various functions described throughout this disclosure. One or more processors in a processing system can execute software, which may be referred to as software, firmware, middleware, microcode, hardware description languages, or others. Software should be interpreted broadly as instructions, instruction sets, code, code segments, program code, programs, subroutines, software components, applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, or any combination thereof.

[0132] If the functions described herein are implemented in software, these functions may be stored on or encoded as one or more instructions or code on a computer-readable medium, such as a non-transitory computer-readable storage medium. Computer-readable media include computer storage media and may include random access memory (RAM), read-only memory (ROM), electrically erasable programmable ROM (EEPROM), optical disk storage, magnetic disk storage, other magnetic storage devices, combinations of these types of computer-readable media, or any other medium that can be used to store computer-executable code in the form of instructions or data structures accessible by a computer. The storage medium can be any available medium accessible to a computer.

[0133] The aspects, implementations, and / or use cases described herein can be implemented across many different platform types, devices, systems, form factors, sizes, and package arrangements. For example, aspects, implementations, and / or use cases can be generated via integrated chip implementations and other devices based on non-modular components, such as end-user devices, vehicles, communication devices, computing devices, industrial equipment, retail / procurement devices, medical devices, devices supporting artificial intelligence (AI), devices supporting machine learning (ML), etc. The scope of aspects, implementations, and / or use cases can range from chip-level or modular components to non-modular or non-chip-level implementations, and further to aggregated, distributed, or original equipment manufacturer (OEM) devices or systems incorporating one or more of the technologies described herein.

[0134] Apparatus incorporating the aspects and features described herein may also include additional components and features for implementing and practicing the claimed and described aspects and features. For example, the transmission and reception of wireless signals necessarily include numerous components for analog and digital purposes, such as hardware components, antennas, RF chains, power amplifiers, modulators, buffers, processors, interleavers, adders / summers, etc. The techniques described herein can be practiced in a wide variety of devices, chip-level components, systems, distributed arrangements, aggregated or decomposed components, end-user devices, etc., in various configurations.

[0135] The description herein is provided to enable those skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects. Therefore, the claims are not limited to the aspects described herein, but should be interpreted in light of the full scope of this disclosure consistent with the language of the claims.

[0136] Unless explicitly stated otherwise, references to singular elements do not imply "one and only one," but rather "one or more." Terms such as "if," "when," and "at" do not imply an immediate temporal relationship or response. That is, these phrases (e.g., "when") do not imply an immediate action in response to the occurrence of an action or during the occurrence of an action, but simply imply that an action will occur if a certain condition is met, without requiring a specific or immediate temporal constraint on the occurrence of the action. The terms "may," "may," and "can" as used in this disclosure generally carry certain connotations. For example, "may" refers to a permissible feature that may or may not occur, "may" refers to a feature that is likely to occur, and "can" refers to a capability (e.g., being able to). The phrase "for example" generally carries a similar connotation to "may," and therefore, "may" is sometimes excluded from sentences that include "for example" or other similar phrases.

[0137] Unless otherwise expressly stated, the term "some" means one or more. Combinations such as "at least one of A, B, or C" or "one or more of A, B, or C" include any combination of A, B, and / or C, such as A and B, A and C, B and C, or A and B and C, and may include multiple A, multiple B, and / or multiple C, or may include only A, only B, or only C. A set should be interpreted as a collection of elements having a quantity of one or more elements.

[0138] Unless otherwise explicitly indicated, ordinal terms such as “first” and “second” do not necessarily imply order in time, sequence, numerical value, etc., but are used to distinguish different instances of the term or phrase following each ordinal term. As used in the specification and figures, reference numerals are sometimes cross-referenced between figures to indicate the same or similar features. Features that are identical in multiple figures may be labeled with the same reference numerals in multiple figures. Features that are similar but not identical in multiple figures may be labeled with reference numerals that have different leading numerals but share one or more of the same trailing numerals (e.g., 206, 306, 406, etc. may refer to similar features in the figures). Sometimes, “X” is used generally to indicate multiple variations of a feature. For example, “X06” may generally refer to all reference numbers ending in “06” (e.g., 206, 306, 406, etc.).

[0139] Structural and functional equivalents of the various aspects of the elements described throughout this disclosure, known or subsequently learned by those skilled in the art, are expressly incorporated herein by reference and are covered by the claims. The terms “module,” “mechanism,” “element,” “device,” etc., may not be substitutes for the term “component.” Therefore, no claim element shall be construed as means plus function unless the phrase “component for…” is explicitly stated herein. As used herein, the phrase “based on” should not be construed as a reference to a closed set of information, one or more conditions, one or more factors, etc. In other words, unless expressly stated otherwise, the phrase “based on A” (where “A” can be information, conditions, factors, etc.) shall be construed as “at least based on A.”

[0140] The examples below are illustrative only and may be combined with other examples or teachings described herein without limitation.

[0141] Example 1 is a wireless communication method at a UE, comprising: receiving from a network entity signaling for configuring a set of channel measurement resources (CMRs) and a report for data collection associated with machine learning-based beam management; receiving a downlink reference signal from the network entity via the CMR set; and transmitting to the network entity the report based on beam measurements corresponding to the downlink reference signal.

[0142] Example 2 can be combined with Example 1 and includes: the report includes beam quality data of all or a subset of downlink reference signals from a first CMR set or a second CMR set.

[0143] Example 3 can be combined with Example 2 and includes: the report further includes a beam index that identifies one of the downlink reference signals with the best beam quality data from the first CMR set or the second CMR set.

[0144] Example 4 can be combined with Example 2 or Example 3, and includes: the report further includes multiple reports. Each report includes the beam quality data or the beam index corresponding to a corresponding CMR set in the first CMR set or the second CMR set.

[0145] Example 5 may be combined with Example 2 or 3 and includes: the report further includes the beam quality data, which corresponds to a beam index identifying the downlink reference signal with the best beam quality data from the first CMR set or the second CMR set.

[0146] Example 6 may be combined with any of Examples 1 to 5 and includes: the signaling further configuring at least one of the following: a plurality of time report instances prior to the reference time for reporting beam quality data in the report; a plurality of measurement instances prior to the reference time for measuring the downlink reference signal of each of the time report instances; or the total number of measurement instances prior to the reference time for measuring the downlink reference signal of the time report instances.

[0147] Example 7 can be combined with Example 6 and includes: the ability to transmit to the network entity the UE's capability regarding a supported configuration for data collection associated with machine learning-based beam management. The capability includes at least one of the following: a maximum value of the time reporting instances; a maximum value of the measurement instances of each of the time reporting instances; a maximum value of the total measurement instances of all the time reporting instances; a maximum number of downlink reference signals with optimal beam quality data to be reported; a maximum product of the time reporting instances and the number of downlink reference signals with optimal beam quality data; a maximum product of the measurement instances of each of the time reporting instances and the number of downlink reference signals with optimal beam quality data; a maximum product of the total measurement instances of all the time reporting instances and the number of downlink reference signals with optimal beam quality data; or a minimum processing time for the report.

[0148] Example 8 can be combined with any of Examples 1 to 7 and includes: the signaling further includes a measurement window for measuring the downlink reference signal to generate beam quality data.

[0149] Example 9 can be combined with Example 8 and includes: the measurement window includes a periodic measurement window for measuring the downlink reference signal at multiple measurement instances. The signaling further includes the periodicity of the periodic measurement window; the time slot offset at the start of each of the periodic measurement windows; and the duration of each of the periodic measurement windows.

[0150] Example 10 may be combined with any of Examples 1 to 9 and includes: receiving the downlink reference signal via the CMR set includes receiving the downlink reference signal using a UE receive beam configured by a first spatial receive parameter associated with the CMR set.

[0151] Example 11 can be combined with Example 10 and includes: configuring the UE receive beam based on a priority rule between the first spatial receive parameter and the second spatial receive parameter when the CMR set overlaps in time with a second downlink signal associated with a second spatial receive parameter that is different from the first spatial receive parameter.

[0152] Example 12 can be combined with Example 11 and includes: suppressing the use of the receive beam configured by the second spatial receive parameters to receive the downlink reference signal based on the priority rule.

[0153] Example 13 may be combined with any of Examples 1 to 12 and includes: receiving a trigger from the network entity (104) for receiving the CMR set or for transmitting the report.

[0154] Example 14 may be combined with any of Examples 2 to 13 and includes: the beam quality data in the report corresponding to the downlink reference signal includes at least one of the following: Layer 1 reference signal received power (L1-RSRP) of one of the downlink reference signals; Layer 1 signal-to-interference-plus-noise ratio (L1-SINR) of one of the downlink reference signals; or assumed measurement error of one of the downlink reference signals.

[0155] Example 15 can be combined with Example 14 and includes: reporting the beam quality data of one of the downlink reference signals in the report when one of the L1-RSRP or the L1-SINR or the assumed measurement error of one of the downlink reference signals meets the reporting criteria.

[0156] Example 16 may be combined with Example 15 and includes: the reporting criteria include at least one of the following: the L1-RSRP of the downlink reference signal is higher than a first threshold; the L1-SINR of the downlink reference signal is higher than a second threshold; or the assumed measurement error of the downlink reference signal is lower than a third threshold.

[0157] Example 17 is a wireless communication method at a network entity, comprising: transmitting to a user equipment (UE) signaling for configuring a set of channel measurement resources (CMRs) and a report on data collection associated with machine learning-based beam management at the network entity; transmitting a downlink reference signal to the UE via the CMR set; and receiving from the UE the report based on beam measurements corresponding to the downlink reference signal.

[0158] Example 18 can be combined with Example 17 and includes: the report includes beam quality data from all or a subset of the downlink reference signals from a first CMR set or a second CMR set.

[0159] Example 19 can be combined with Example 18 and includes: the report further includes a beam index that identifies one of the downlink reference signals with the best beam quality data from the first CMR set or the second CMR set.

[0160] Example 20 may be combined with Example 18 or 19 and includes: the report comprises multiple reports. Each report includes the beam quality data or the beam index corresponding to a corresponding CMR set in the first CMR set or the second CMR set.

[0161] Example 21 may be combined with any of Examples 17 to 20 and includes: the signaling further includes at least one of the following: a plurality of time report instances prior to the reference time for reporting beam quality data in the report; a plurality of measurement instances prior to the reference time for measuring the downlink reference signal of each of the time report instances; or the total number of measurement instances prior to the reference time for measuring the downlink reference signal of the time report instances.

[0162] Example 22 may be combined with any of Examples 18 to 21 and includes that the beam quality data in the report corresponding to the downlink reference signal includes at least one of the following: Layer 1 reference signal received power (L1-RSRP) of one of the downlink reference signals; Layer 1 signal-to-interference-plus-noise ratio (L1-SINR) of one of the downlink reference signals; or assumed measurement error of one of the downlink reference signals.

[0163] Example 23 is a device for wireless communication, including a memory, a transceiver, and a processor coupled to the memory and the transceiver, the device being configured to implement the method as described in any one of Examples 1 to 22.

[0164] Example 24 can be combined with Example 2 or 3 and includes carrying the first CMR set and the second CMR set on a beam having the same beam characteristics or different beam characteristics.

[0165] Example 25 can be combined with Example 6 and includes: the signaling configuring the plurality of measurement instances. The beam quality data for each of the time reporting instances includes the average value of the beam measurements performed at the measurement instance.

[0166] Example 26 can be combined with Example 6 and includes determining the reference time based on: the first symbol of the beam report and the shortest processing time for the beam report; or configuration information from the network entity.

[0167] Example 27 can be combined with Example 11 and includes: the signaling from the network entity further configuring the priority rules.

[0168] Example 28 can be combined with Example 11 and includes: using the receive beam configured by the second receive parameter to receive the one or more CMR sets based on the priority rule.

[0169] Example 29 can be combined with Example 15 and includes: the signaling further configures the reporting standard.

[0170] Example 30 may be combined with Example 18 or 19 and includes carrying the first CMR set and the second CMR set on a beam having the same or different beam characteristics.

[0171] Example 31 may be combined with Example 18 or 19 and includes: the beam report further includes the beam quality data, which corresponds to a beam index identifying the downlink reference signal with the best beam quality data from the first CMR set or the second CMR set.

[0172] Example 32 can be combined with Example 21 and includes the signaling configuration of the plurality of measurement instances. The beam quality data of each of the time reporting instances includes the average value of the beam measurements performed at the measurement instance.

[0173] Example 33 can be combined with Example 21 and includes: the signaling further configures the reference time.

[0174] Example 34 can be combined with Example 21 and includes: the ability of the UE to receive from the UE a supported configuration for data collection associated with machine learning-based beam management. The capability includes at least one of: a maximum value of the time report instances; a maximum value of the measurement instances of each of the time report instances; a maximum value of the total measurement instances of all the time report instances; a maximum number of downlink reference signals with optimal beam quality data to be reported; a maximum product of the time report instances and the number of downlink reference signals with optimal beam quality data; a maximum product of the measurement instances of each of the time report instances and the number of downlink reference signals with optimal beam quality data; a maximum product of the total measurement instances of all the time report instances and the number of downlink reference signals with optimal beam quality data; or a minimum processing time for the report.

[0175] Example 35 can be combined with Example 17 and includes: the signaling further configures the measurement window used by the UE to measure downlink reference signals to generate beam quality data.

[0176] Example 36 can be combined with Example 35 and includes: the measurement window includes a periodic measurement window used by the UE to measure the downlink reference signal at multiple measurement instances. The signaling further includes the periodicity of the periodic measurement window; the time slot offset at the start of each of the periodic measurement windows; and the duration of each of the periodic measurement windows.

[0177] Example 37 can be combined with Example 17 and includes: configuring the receive beam based on a priority rule between the first spatial receive parameter and the second spatial receive parameter when the CMR set overlaps in time with a second downlink signal associated with a second spatial receive parameter that is different from the first spatial receive parameter used by the UE to configure the receive beam for receiving the downlink reference signal.

[0178] Example 38 can be combined with Example 17 and includes: the network entity suppressing transmission of the CMR set that overlaps temporally with and is associated with a second downlink signal configured by the UE to configure the receive beam.

[0179] Example 39 can be combined with Example 17 and includes: transmitting to the UE a trigger for receiving the CMR set or for transmitting the report.

[0180] Example 40 can be combined with Example 22 and includes: when one or more of the L1-RSRP or L1-SINR or the assumed measurement error of the downlink reference signal meets the reporting criteria, the beam quality data of one or more of the downlink reference signals is included in the beam report.

[0181] Example 41 may be combined with Example 40 and includes: the reporting criteria include at least one of the following: the L1-RSRP of the downlink reference signal is higher than a first threshold; the L1-SINR of the downlink reference signal is higher than a second threshold; or the assumed measurement error of the downlink reference signal is lower than a third threshold.

[0182] Example 42 can be combined with Example 40 and includes: the signaling further configures the reporting standard.

[0183] Example 43 can be combined with Example 4 and includes: applying the beam quality data in the first report as input data to a machine learning model for beam management, and applying the beam index in the second report as output data to the machine learning model.

[0184] Example 44 can be combined with Example 20 and includes: applying the beam quality data in the first report as input data to a machine learning model for beam management, and applying the beam index in the second report as output data to the machine learning model.

Claims

1. A wireless communication method at a user equipment (UE) (102), comprising: Receive (1104) signaling from network entity (104) for configuring the following: Channel measurement resource (CMR) set; as well as Reports on data collection related to machine learning-based beam management; Receive (1110) downlink reference signal from the network entity (104) via the CMR set; as well as Transmit (1112) the report based on beam measurements corresponding to the downlink reference signal to the network entity (104).

2. The method as described in claim 1, wherein, The report includes beam quality data for all or a subset of the downlink reference signals from the first CMR set or the second CMR set.

3. The method as described in claim 2, wherein, The report further includes a beam index that identifies one of the downlink reference signals from the first CMR set or the second CMR set that has the best beam quality data.

4. The method as described in any one of claims 2 or 3, wherein, The report includes multiple reports, each of which includes beam quality data or beam index corresponding to a corresponding CMR set in the first CMR set or the second CMR set.

5. The method as described in any one of claims 2 or 3, wherein, The report further includes the beam quality data, which corresponds to the beam index that identifies the downlink reference signal with the best beam quality data from the first CMR set or the second CMR set.

6. The method according to any one of claims 1 to 5, wherein, The signaling further configures at least one of the following: Multiple time report instances prior to the reference time used to report beam quality data in the report; Multiple measurement instances of the downlink reference signal used to measure each of the time report instances prior to the reference time; or The total number of measurement instances of the downlink reference signal used to measure the time report instance prior to the reference time.

7. The method of claim 6, further comprising: The ability of the UE (102) to transmit (1102) to the network entity (104) regarding supported configurations for the data collection associated with machine learning-based beam management, the ability including at least one of the following: The maximum value of the time report instance; The maximum value of the measurement instance for each of the time report instances; The maximum value of all total measurement instances in the time report instance; The maximum number of downlink reference signals with the best beam quality data to be reported; The maximum product of the time report instance and the number of the downlink reference signals with the best beam quality data; The maximum product of the measurement instance of each of the time report instances and the number of the downlink reference signals with the best beam quality data; The maximum product of all the total measurement instances in the time report instance and the number of the downlink reference signal with the best beam quality data; or The shortest processing time for the report.

8. The method according to any one of claims 1 to 7, wherein, The signaling further configures a measurement window for measuring the downlink reference signal to generate beam quality data.

9. The method of claim 8, wherein, The measurement window includes a periodic measurement window for measuring the downlink reference signal at multiple measurement instances, and wherein the signaling is further configured as follows: The periodicity of the periodic measurement window; The time slot offset at the start of each of the periodic measurement windows; and The duration of each periodic measurement window.

10. The method according to any one of claims 1 to 9, wherein, Receiving the downlink reference signal via the CMR set includes: The downlink reference signal is received using a UE receive beam configured by a first spatial receive parameter associated with the CMR set.

11. The method of claim 10, wherein, When the CMR set overlaps in time with a second downlink signal associated with a second spatial reception parameter that is different from the first spatial reception parameter, the UE receive beam is configured based on the priority rule between the first spatial reception parameter and the second spatial reception parameter.

12. The method of claim 11, further comprising: Suppress the use of the receive beam configured by the second spatial receive parameters to receive the downlink reference signal based on the priority rule.

13. The method of any one of claims 1 to 12, further comprising: Receive (1108) a trigger from the network entity (104) for receiving the CMR set or for transmitting the report.

14. The method according to any one of claims 2 to 13, wherein, The beam quality data in the report corresponding to the downlink reference signal includes at least one of the following: The Layer 1 Reference Signal Received Power (L1-RSRP) of one of the downlink reference signals; One of the downlink reference signals is a Layer 1 signal with an interference-plus-noise ratio (L1-SINR); or The assumed measurement error of one of the downlink reference signals.

15. The method of claim 14, wherein, When one of the L1-RSRP, L1-SINR, or assumed measurement error of a downlink reference signal meets the reporting criteria, the beam quality data of one of the downlink reference signals is reported in the report.

16. The method of claim 15, wherein, The reporting criteria include at least one of the following: The L1-RSRP of the downlink reference signal is higher than a first threshold; The L1-SINR of the downlink reference signal is higher than the second threshold; or The downlink reference signal is assumed to have a measurement error below a third threshold.

17. A wireless communication method at a network entity (104), comprising: Transmit (1204) signaling to the user equipment (UE) (102) for configuring the following: Channel measurement resource (CMR) set; as well as Reports on data collection associated with machine learning-based beam management on the network entity (104); (1210) downlink reference signal is transmitted to the UE (102) via the CMR set; as well as The UE (102) receives (1212) the report based on beam measurements corresponding to the downlink reference signal.

18. The method of claim 17, wherein, The report includes beam quality data from all or a subset of downlink reference signals from a first CMR set or a second CMR set.

19. The method of claim 18, wherein, The report further includes a beam index that identifies one of the downlink reference signals from the first CMR set or the second CMR set that has the best beam quality data.

20. The method of any one of claims 18 or 19, wherein, The report includes multiple reports, each of which includes beam quality data or beam index corresponding to a corresponding CMR set in the first CMR set or the second CMR set.

21. The method according to any one of claims 17 to 20, wherein, The signaling further configures at least one of the following: Multiple time report instances prior to the reference time used to report beam quality data in the report; Multiple measurement instances of the downlink reference signal used to measure each of the time report instances prior to the reference time; or The total number of measurement instances of the downlink reference signal used to measure the time report instance prior to the reference time.

22. The method according to any one of claims 18 to 21, wherein, The beam quality data in the report corresponding to the downlink reference signal includes at least one of the following: The Layer 1 Reference Signal Received Power (L1-RSRP) of one of the downlink reference signals; One of the downlink reference signals is a Layer 1 signal with an interference-plus-noise ratio (L1-SINR); or The assumed measurement error of one of the downlink reference signals.

23. An apparatus for wireless communication, comprising a memory, a transceiver, and a processor coupled to the memory and the transceiver, the apparatus being configured to implement the method as claimed in any one of claims 1 to 22.

Images