WO2025117236A1

WO2025117236A1 - Ai/ml based rrm for serving and neighbor cells with different transmit antenna pattern implementations

Info

Publication number: WO2025117236A1
Application number: PCT/US2024/056452
Authority: WO
Inventors: Konstantinos Sarrigeorgidis; Yang Tang; Jie Cui
Original assignee: Apple Inc
Current assignee: Apple Inc
Priority date: 2023-11-30
Filing date: 2024-11-19
Publication date: 2025-06-05
Anticipated expiration: 2026-05-30

Abstract

Systems and methods are provided for AI/ML based radio resource management for serving and neighbor cells with different transmit antenna pattern implementations. A unified AI/ML model is trained using a training data set including of a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions or different Tx implementations, such as different antenna spacings. Assistance information provided to the AI/ML model is selected from beam angles of probing Tx beams, Tx codebook sizes, beamwidths, and antenna spacings.

Description

AI/ML BASED RRM FOR SERVING AND NEIGHBOR CELLS WITH DIFFERENT

TRANSMIT ANTENNA PATTERN IMPLEMENTATIONS

TECHNICAL FIELD

[0001] This application relates generally to wireless communication systems, including wireless communication systems with artificial intelligence or machine learning based radio resource management.

BACKGROUND

[0002] Wireless mobile communication technology uses various standards and protocols to transmit data between a base station and a wireless communication device. Wireless communication system standards and protocols can include, for example, 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE) (e.g., 4G), 3GPP New Radio (NR) (e.g., 5G), and Institute of Electrical and Electronics Engineers (IEEE) 802.11 standard for Wireless Local Area Networks (WLAN) (commonly known to industry groups as Wi-Fi®).

[0003] As contemplated by the 3GPP, different wireless communication systems standards and protocols can use various radio access networks (RANs) for communicating between a base station of the RAN (which may also sometimes be referred to generally as a RAN node, a network node, or simply a node) and a wireless communication device known as a user equipment (UE). 3GPP RANs can include, for example, Global System for Mobile communications (GSM), Enhanced Data Rates for GSM Evolution (EDGE) RAN (GERAN), Universal Terrestrial Radio Access Network (UTRAN), Evolved Universal Terrestrial Radio Access Network (E-UTRAN), and/or Next- Generation Radio Access Network (NG-RAN).

[0004] Each RAN may use one or more radio access technologies (RATs) to perform communication between the base station and the UE. For example, the GERAN implements GSM and/or EDGE RAT, the UTRAN implements Universal Mobile Telecommunication System (UMTS) RAT or other 3GPP RAT, the E-UTRAN implements LTE RAT (sometimes simply referred to as LTE), and NG-RAN implements NR RAT (sometimes referred to herein as 5G RAT, 5GNR RAT, or simply NR). In certain deployments, the E-UTRAN may also implement NR RAT. In certain deployments, NG-RAN may also implement LTE RAT. [0005] A base station used by a RAN may correspond to that RAN. One example of an E-UTRAN base station is an Evolved Universal Terrestrial Radio Access Network (E- UTRAN) Node B (also commonly denoted as evolved Node B, enhanced Node B, eNodeB, or eNB). One example of an NG-RAN base station is a next generation Node B (also sometimes referred to as a g Node B or gNB).

[0006] A RAN provides its communication services with external entities through its connection to a core network (CN). For example, E-UTRAN may utilize an Evolved Packet Core (EPC) while NG-RAN may utilize a 5G Core Network (5GC).

[0007] Frequency bands for 5G NR may be separated into two or more different frequency ranges. For example, Frequency Range 1 (FR1) may include frequency bands operating in sub-6 gigahertz (GHz) frequencies, some of which are bands that may be used by previous standards, and may potentially be extended to cover new spectrum offerings from 410 megahertz (MHz) to 7125 MHz. Frequency Range 2 (FR2) may include frequency bands from 24.25 GHz to 52.6 GHz. Note that in some systems, FR2 may also include frequency bands from 52.6 GHz to 71 GHz (or beyond). Bands in the millimeter wave (mmWave) range of FR2 may have smaller coverage but potentially higher available bandwidth than bands in FR1. Skilled persons will recognize these frequency ranges, which are provided by way of example, may change from time to time or from region to region.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0008] To easily identify the discussion of any particular element or act, the most significant digit or digits in a reference number refer to the figure number in which that element is first introduced.

[0009] FIG. 1 A and FIG. IB schematically illustrate a first example case, according to certain embodiments, wherein the Tx beam patterns are the same between a serving cell and a neighbor cell, but in a reshuffled order.

[0010] FIG. 2 is a flowchart of an example method for a UE, according to certain embodiments.

[0011] FIG. 3 is a flowchart of an example method for a base station, according to certain embodiments. [0012] FIG. 4 schematically illustrates a second example case, according to certain embodiments, wherein the Tx beam patterns are not the same between a serving cell and one or more neighbor cells.

[0013] FIG. 5 is a flowchart of an example method for a UE, according to certain embodiments.

[0014] FIG. 6 is a flowchart of an example method for a base station, according to certain embodiments.

[0015] FIG. 7 schematically illustrates providing assistance information to an AI/ML model in a third example case, according to certain embodiments.

[0016] FIG. 8 illustrates a flow diagram for the use of an AI/ML model at a UE to identify a best Tx-Rx beam pair between a base station and the UE, along with various corresponding diagrammatic illustrations, according to certain embodiments.

[0017] FIG. 9 illustrates a flow diagram for the use of an AI/ML model at a UE to identify a best Tx-Rx beam pair between the UE and a serving cell and one or more neighbor cells, along with various corresponding diagrammatic illustrations, according to certain embodiments.

[0018] FIG. 10 schematically illustrates an example of training different AI/ML models with different subsets of probing Tx beams, according to certain embodiments.

[0019] FIG. 11 illustrates example graphs showing performance degradation among different subsets during inference, according to certain embodiments.

[0020] FIG. 12 schematically illustrates a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions, according to certain embodiments.

[0021] FIG. 13 illustrates example graphs showing performance improvement when using assistance information, according to certain embodiments.

[0022] FIG. 14 illustrates example graphs showing performance improvement in best beam pair selection accuracy when increasing the number of Rx beams, according to certain embodiments.

[0023] FIG. 15A, FIG. 15B, FIG. 15C, and FIG. 15D are example polar pilot beamforming plots, according to certain embodiments.

[0024] FIG. 16 illustrates a flow diagram for the use of an AI/ML model at a UE to identify a best Tx-Rx beam pair between the UE and a serving cell and one or more neighbor cells, along with various corresponding diagrammatic illustrations, according to certain embodiments.

[0025] FIG. 17 schematically illustrates an example of training different AI/ML models with different subsets of probing Tx beams with different antenna spacings, according to certain embodiments.

[0026] FIG. 18 schematically illustrates a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different Tx implementations, according to certain embodiments.

[0027] FIG. 19A , FIG. 19B, FIG. 19C, and FIG. 19D illustrate example graphs showing performance improvement when using antenna spacing assistance information, according to certain embodiments.

[0028] FIG. 20 is a flowchart of an example method for a UE, according to certain embodiments.

[0029] FIG. 21 is a flowchart of an example method for a base station, according to certain embodiments.

[0030] FIG. 22 illustrates an example architecture of a wireless communication system, according to embodiments disclosed herein.

[0031] FIG. 23 illustrates a system for performing signaling between a wireless device and a network device, according to embodiments disclosed herein.

DETAILED DESCRIPTION

[0032] Various embodiments are described with regard to a UE. However, reference to a UE is merely provided for illustrative purposes. The example embodiments may be utilized with any electronic component that may establish a connection to a network and is configured with the hardware, software, and/or firmware to exchange information and data with the network. Therefore, the UE as described herein is used to represent any appropriate electronic component.

[0033] In various wireless communication systems, overhead penalties corresponding to the transmission of reference signals for radio resource management (RRM) (e.g., synchronization signal blocks (SSBs) and/or channel state information- reference signals (CSI-RSs)) are not insignificant. For example, in cases where an SSB-based RRM measurement timing configuration (SMTC) periodicity of 20 milliseconds (ms) and an SSB burst of 5 ms are used in FR2 in an NR wireless communication system, a corresponding SMTC-associated overhead is understood as 25%.

[0034] Accordingly, in some cases, it may be beneficial to reduce a number of beams used at the transmit (Tx) side and/or the receive (Rx) side in Layer 3 (L3)-related measurements (e.g., beam measurements for RRM purposes).

[0035] Embodiments disclosed herein relate enhancements for RRM functionality with artificial intelligence (AI)/machine learning (ML) models. In various embodiments, an L3 measurement delay and/or overhead reduction may be achieved by minimizing a Tx/Rx beam sweeping set through the use of spatial beam prediction by an AI/ML model. For example, an SMTC window duration can be reduced by reducing the beam sweeping factor used by /in the SMTC window (e.g., by reducing the number of Tx and/or Rx beams used by the beam sweeping mechanism of the SMTC window).

[0036] In various embodiments, L3 measurement reduction can be achieved by periodically skipping one or more Tx/Rx beam sweeping instances and using a spatial beam prediction by an AI/ML model in place of actual measurements. As one conceptual example, an SMTC periodicity may be increased, representing a conceptual skipping of various instances of SMTC-based measurements that would have otherwise occurred at the lower periodicity. The skipped measurements may be replaced by predictions for those measurements generated at an AI/ML model.

[0037] As a result of such reductions of L3 measurements, throughput can be increased. For example, scheduling restrictions at orthogonal frequency division multiplexing (OFDM) symbols that would have otherwise been used for measurements (e.g., symbols that would have otherwise been used for SSBs to be measured) can be removed.

Accordingly, as a result of using an AI/ML model and/or algorithm for RRM functionalities, RRM performance (and therefore overall system performance) can be enhanced.

[0038] For beam management in certain wireless communication systems, the UE performs measurements across a set of Tx/Rx beam pairs. When the number of Tx beams is M and the number of UE beams is N, measuring the L3 reference signal received power (RSRP) values of all MxN beam pairs and selecting the best beam leads to large measurement overhead. As used herein, a full Tx codebook represents a whole set of Tx beams at each gNB, and a probing codebook or probing Tx codebook represents a sparse subset of the full Tx codebook or another set of beams with wider beam widths used to form spatial beams and compute RSRP values and serve as input to the AI/ML model. [0039] For RRM measurements, the size of the full Tx codebook, the size of the probing codebook, as well as the beam widths could be different between a serving cell and neighbor cells. Further, a joint Tx/Rx beam determination may use a large beam sweeping factor and may involve an exhaustive search across Tx/Rx beams. AI/ML could achieve significant RRM measurement overhead (beam sweeping factor) reduction. For example, the model may be generic enough to be re-employed when the number of inputs/outputs as well as the beamwidths differ between various gNB implementations (serving and neighbor cells for L3 measurements). In some cases, a joint AI/ML based Tx/Rx beam determination may also replace the exhaustive search. [0040] Wireless communication systems that attempt RSRP measurement overhead reduction may use a full set or map of beam pair RSRP values predicted from measurements on a subset of beam pair RSRP values. An AI/ML model may be trained to interpolate or extrapolate between the subset measurements in order to recover the full set of RSRP measurements. However, different neighbor cells may use different beamforming implementations and different subsets of probing beams. Therefore, the interpolation function (e.g., for a subset of beams to a full set of beams) could be different between serving and neighbor cells. In some cases, it could be very costly for the UE to employ different AI/ML models for different interpolation functions for every possible cell Tx implementation.

[0041] Thus, certain embodiments disclosed herein provide an AI/ML architecture that can be unified and used to make spatial beam predictions across different spatial beam subsets for different cell implementations. Such embodiments describe the type of assisted signaling provided by the network in order to facilitate the employment of a unified AI/ML model for spatial beam prediction.

[0042] By way of illustration, three example cases or scenarios are discussed below. In a first example case, the Tx beam patterns are the same between the serving cell and one or more neighbor cells, but in a reshuffled order, wherein the network transmits all SSBs (i.e., full codebook). In a second example case, the Tx beam patterns are not the same between the serving cell and one or more neighbor cells, e.g., the beam widths and/or the codebook sizes differ, wherein the network transmits all SSBs (i.e., full codebook). In a third example case, the network does not transmit all SSBs but transmits only probing beams (i.e., a probing codebook is used and the probing beams are inputs to the AI/ML interpolating function).

[0043] First Example Case

[0044] FIG. 1A and FIG. IB schematically illustrate the first example case, according to certain embodiments, wherein the Tx beam patterns are the same between a serving cell 102 and a neighbor cell 104, but in a reshuffled order. In this example, the serving cell 102 uses a full Tx codebook of SSB beams (e.g., with SSB beam indexes SSB1, SSB2,..., SSB64), where each SSB beam corresponds to a respective physical angle (e.g., an azimuth angle (AZ) and an elevation angle (EL)). The neighbor cell 104 uses the same full Tx codebook of the same size (e.g., with SSB beam indexes SSB1, SSB2,..., SSB64) and the same beamwidths (e.g., half power beamwidth or 3 decibel (dB) beamwidth), but each SSB beam in the SSB beam index corresponds to some other physical angle (e.g., AZ¹, EL¹). For example, the first beam 106 (i.e., SSB1) of the SSB beam index of the serving cell 102 is transmitted at a different physical angle (i.e., is a different spatial beam) than that of the first beam 108 (i.e., SSB1) of the SSB index of the neighbor cell 104.

[0045] An AI/ML model 110 at the UE may be trained based on measured RSRP values of a subset of Tx beams corresponding to the SSB beam index of the serving cell 102. For example, the AI/ML model 110 may be trained by measuring RSRP values for SSB beam index 5, SSB beam index 8, and SSB beam index 60 of the serving cell 102 (see FIG. IB) using selected Rx beams of the UE. During measurements of the serving cell 102 for RRM, the UE may provide measured RSRP values for the same Tx/Rx beam pairs used for training to the AI/ML model 110 to obtain an accurate prediction 112 of RSRP values for each Tx beam of the serving cell 102. However, due to the different physical angles (i.e., reshuffled beam order) between the SSB beam index of the serving cell 102 and the SSB beam index of the neighbor cell 104, providing measured RSRP values corresponding to SSB beam index 5 (SSB5), SSB beam index 8 (SSB8), and SSB beam index 60 (SSB60) of the neighbor cell 104 to the AI/ML model 110 may result in an inaccurate prediction 114 of RSRP values for each Tx beam of the neighbor cell 104.

[0046] Thus, in certain embodiments, the UE selects the same subset of probing Tx beams (i.e., the same spatial beams) used to train the AI/ML model 110 for both the serving cell 102 and the neighbor cell 104. Referring to FIG. 1A, for example, the UE measures probing beam 116, probing beam 120, and probing beam 124 to provide corresponding RSRP values as inputs to the AI/ML model 110 to generate an accurate prediction 112 of RSRP values for each Tx beam of the serving cell 102. In this example, the probing beam 116 of the serving cell 102 is the same spatial beam as probing beam 118 of the neighbor cell 104, the probing beam 120 of the serving cell 102 is the same spatial beam as probing beam 122 of the neighbor cell 104, and the probing beam 124 is the same spatial beam as probing beam 126 of the serving cell 102. Thus, the UE also measures probing beam 118, probing beam 122, and probing beam 126 to provide corresponding RSRP values as inputs to the AI/ML model 110 to generate an accurate prediction 114 of RSRP values for each Tx beam of the neighbor cell 104.

[0047] In certain embodiments, the neighbor cell 104 signals a mapping between the SSB index of the neighbor cell 104 to the SSB index of the serving cell 102. For example, as shown in FIG. IB, the neighbor cell 104 signals to the UE that the SSB index = 1, 2, 3, 4, 5, etc. of the neighbor cell 104 corresponds to the SSB index = 63, 60, 62, 64, 3, etc. of the serving cell 102 that was used to train the AI/ML model 110.

[0048] FIG. IB shows that probing beam 116 corresponds to SSB3, probing beam 120 corresponds to SSB6, and probing beam 124 corresponds to SSB61 according to the SSB beam indexing of the serving cell 102. In this example, to reuse the same AI/ML model 110 (i.e., the model trained in the serving cell 102) for both the serving cell 102 and the neighbor cell 104, it is assumed that the set of physical angles (AZ¹, EL') of SSBs of the neighbor cell 104 is the same set of physical angles (AZ, EL) of SSBs of the serving cell 102 (i.e., the Tx codebooks are the same but reshuffled in an arbitrary order). In this example, it is also assumed that the serving cell 102 and the neighbor cell 104 transmit all the SSBs of the Tx codebook, the Tx codebook size is the same between the serving cell 102 and the neighbor cell 104, and the network makes a mapping between the physical angles (AZ', EL') to the physical angles (AZ, EL) per SSB. FIG. IB shows the mapped SSB index for the neighbor cell 104, in which probing beam 118 is mapped to SSB3, probing beam 122 is mapped to SSB6, and probing beam 126 is mapped to SSB61. Because the serving cell 102 and the neighbor cell 104 transmit all the SSBs of the Tx codebook, the UE can select the appropriate probing beams for the neighbor cell 104 based on the mapping to provide the corresponding RSRP values to the same AI/ML model 110. Thus, the inputs to the AI/ML model 110 are based on the same spatial beams for both the serving cell 102 and the neighbor cell 104 to obtain the prediction 112 corresponding to the Tx beams of the serving cell 102 and the prediction 114 corresponding to the Tx beams of the neighbor cell 104.

[0049] FIG. 2 is a flowchart of an example method 200 for a UE, according to certain embodiments. In block 202, the method 200 includes receiving, at the UE from a base station, an SSB index mapping between a first Tx beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell. In this example, the serving cell and the neighbor cell use a same Tx codebook, and the second Tx beam pattern is in a reshuffled order from the first Tx beam pattern. In block 204, the method 200 includes selecting, at the UE, a subset of probing Tx beams from the first Tx beam pattern and the second Tx beam pattern based on the SSB index mapping. The subset of probing Tx beams corresponds to beam angles used to train an AI/ML model. In block 206, the method 200 includes performing, at the UE, first reference signal measurements on reference signals transmitted on the subset of probing Tx beams using probed Rx beams of the UE to generate measured received signal strengths. In block 208, the method 200 includes providing the measured received signal strengths to the AI/ML model at the UE to generate predicted received signal strengths of Tx-Rx beam pairs. In block 210, the method 200 includes identifying, at the UE, one or more predicted candidate Tx-Rx beam pairs based on the predicted received signal strengths of the Tx-Rx beam pairs. In block 212, the method 200 includes performing, at the UE, second reference signal measurements on the one or more candidate Tx beams using one or more candidate Rx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs to determine a selected Tx beam. In block 214, the method 200 includes reporting, from the UE to the base station, the selected Tx beam.

[0050] FIG. 3 is a flowchart of an example method 300 for a base station, according to certain embodiments. In block 302, the method 300 includes transmitting, from the base station to a UE, Tx beams according to a first Tx beam pattern. In block 304, the method 300 includes transmitting, from the base station to the UE, an SSB index mapping between the first Tx beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell. In this example, the serving cell and the neighbor cell use a same Tx codebook for the Tx beams, and the second Tx beam pattern is in a reshuffled order from the first Tx beam pattern. In block 306, the method 300 includes receiving, at the base station from the UE, an indication of a selected Tx beam. In block 308, the method 300 includes transmitting, from the base station to the UE, downlink data using the selected Tx beam.

[0051] Second Example Case

[0052] FIG. 4 schematically illustrates the second example case, according to certain embodiments, wherein the Tx beam patterns are not the same between a serving cell and one or more neighbor cells, e.g., the beamwidths and/or the codebook sizes differ, and wherein the network transmits all SSBs (i.e., full codebook). In this example, the serving cell 402 uses a full Tx codebook of SSB beams (e.g., with SSB beam indexes SSB1, SSB2,..., SSB64), where each SSB beam corresponds to a respective physical angle (AZ, EL). The neighbor cell 404 uses a different Tx codebook with a different size (e.g., with SSB beam indexes SSB1, SSB2,..., SSBm), where each SSB beam corresponds to a different respective physical angle (AZ¹, EL') than that of the order of Tx beams used by the serving cell 402. Further, as shown in FIG. 4, the 3dB beamwidth of the Tx beams of the neighbor cell 404 is different from that of the serving cell 402.

[0053] In the example shown in FIG. 4, an AI/ML model 406 at the UE is trained in the serving cell 402, and the UE uses the same AI/ML model 406 for both the serving cell 402 and the neighbor cell 404. It is assumed that the serving cell 402 and the neighbor cell 404 each transmit all the SSBs of their respective Tx codebooks.

[0054] For the serving cell 402, the UE measures probing beam 408, probing beam 410, and probing beam 412 and provides respective RSRP values for corresponding Tx-Rx beam pairs to the AI/ML model 406. The UE also provides assistance information, which may be received from the network, to the AI/ML model 406. The assistance information includes a first probing angle p (e.g., AZ, EL) for the probing beam 408, a second probing angle <Pp for the probing beam 410, a third probing angle q)p for the probing beam 412, a j^th inquire beam angle (p- for each of one or more inquire beam 414, and a 3dB beamwidth (p_3dB for the Tx codebook of the serving cell 402. The AI/ML model 406 outputs a predicted RSRP value RSRP((p- ) for each of the one or more inquire beam 414 of the serving cell 402.

[0055] Similarly, for the neighbor cell 404, the UE measures probing beam 416, probing beam 418, and probing beam 420 and provides respective RSRP values for corresponding Tx-Rx beam pairs to the AI/ML model 406. The UE also provides assistance information, which may be received from the network, to the AI/ML model 406. The assistance information includes a first probing angle cp (e.g., AZ', EL¹) for the probing beam 416, a second probing angle <p for the probing beam 418, a third probing angle (pp for the probing beam 420, a 7^th inquire beam angle (p- for each of one or more inquire beam 422, and a 3dB beamwidth (p_3dB for the Tx codebook of the serving neighbor cell 404. As shown in FIG. 4, the assistance information of the neighbor cell 404 has different values than those of the assistance information of the serving cell 402. The AI/ML model 406 outputs a predicted RSRP value RSRP (p- ~) for each of the one or more inquire beam 422 of the neighbor cell 404.

[0056] In an alternative embodiment, the AI/ML model may be retrained with a new beam pattern for different cells. In general, an AI/ML model can be categorized by the angle difference between consecutive beams. In some cases, there can be different AI/ML models per cell Tx codebook with different angle spacing. Any re-shuffling may be handled by SSB index mapping, as described for case 1 with respect to FIG. 1A, FIG. IB, FIG. 2, and FIG. 3.

[0057] FIG. 5 is a flowchart of an example method 500 for a UE, according to certain embodiments. In block 502, the method 500 includes receiving, at the UE from a base station, assistance information for a first Tx beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell. In this example, the serving cell and the neighbor cell use different Tx codebooks. The assistance information indicates physical angles of probing Tx beams, inquire beam angles, and beamwidths of Tx beams in the first Tx beam pattern and the second Tx beam pattern. In block 504, the method 500 includes selecting, at the UE, a subset of the Tx beams in the first Tx beam pattern and the second Tx beam pattern corresponding to the physical angles of the probing Tx beams used to train an AI/ML model. In block 506, the method 500 includes performing, at the UE, reference signal measurements on reference signals transmitted on the subset of probing Tx beams using probed Rx beams of the UE to generate measured received signal strengths. In block 508, the method 500 includes providing the measured received signal strengths to the AI/ML model at the UE to generate predicted received signal strengths corresponding to the inquire beam angles. In block 510, the method 500 includes reporting, from the UE to the base station, the predicted received signal strengths corresponding to the inquire beam angles.

[0058] FIG. 6 is a flowchart of an example method 600 for a base station, according to certain embodiments. In block 602, the method 600 includes transmitting, from the base station to the UE, Tx beams according to a first Tx beam pattern of a serving cell. In block 604, the method 600 includes transmitting, from the base station to a UE, assistance information for the first Tx beam pattern of the serving cell and a second Tx beam pattern of a neighbor cell. In this example, the serving cell and the neighbor cell use different Tx codebooks. The assistance information indicates physical angles of probing Tx beams, inquire beam angles, and beamwidths of Tx beams in the first Tx beam pattern and the second Tx beam pattern. In block 606, the method 600 includes receiving, at the base station from the UE, an indication of predicted received signal strengths corresponding to the inquire beam angles. In block 608, the method 600 includes transmitting, from the base station to the UE, downlink data based on the predicted received signal strengths.

[0059] Third Example Case

[0060] As indicated above, in the third example case the network does not transmit all SSB indices all the time. Rather, each cell transmits a subset of probing Tx beams, and each subset of probing Tx beams may be different from cell to cell.

[0061] FIG. 7 schematically illustrates providing assistance information to an AI/ML model 702 in the third example case, according to certain embodiments. In this example, because each subset of probing Tx beams may be different from cell to cell, the AI/ML model 702 is presented with RSRP values corresponding to random subsets of Tx beam patterns. To use the same AI/ML model 702 for each cell, the network signals cellspecific assistance information to the UE to use as additional inputs to the AI/ML model 702. The assistance information, in certain embodiments, includes probing beam spatial directions (e.g., a first probing angle ip , a second probing angle <Pp. a third probing angle <Pp, etc.), inquire beam spatial directions (e.g., a /^th inquire beam angle (p- for each of one or more inquire beam), and a beamwidth (e.g., a 3dB beamwidth <p_3dB) for the Tx codebook of the corresponding cell. The AI/ML model 702 outputs a predicted RSRP value RSRP tp- ) for each of the one or more inquire beam of the corresponding cell.

[0062] FIG. 8 illustrates a flow diagram 800 for the use of an AI/ML model 806 at a UE 804 to identify a best Tx-Rx beam pair between a base station 802 and the UE 804 for a serving cell, along with various corresponding diagrammatic illustrations, according to embodiments provided herein. As illustrated, the base station 802 may transmit 810 reference signals on one or more probing beams 808 to the UE 804. The one or more probing beams 808 used may be a subset of all Tx beams available/useable at the base station 802 per a selected Tx codebook (may be fewer than all the Tx beams of the selected Tx codebook) of the serving cell. The particular subset of the one or more probing beams 808 used from the Tx codebook may be specific to the active serving cell. The transmission of the reference signals on the one or more probing beams 808 may occur one Tx beam at a time, according to a beam sweeping fashion.

[0063] Note that in some embodiments, the base station 802 may also transmit 810 information corresponding to the one or more probing beams 808. For example, the base station 802 may transmit 810 beam directions of the one or more probing beams 808 (e.g., in terms of angle domain information and/or SSB index information for each of the one or more probing beams 808). Further, the base station 802 may also transmit 810 an applicable Tx codebook size for the one or more probing beams 808 (e.g., a size of the Tx codebook from which the one or more probing beams 808 are taken/sampled). Beam direction and/or codebook information as provided to the UE may enable the UE to properly measure/interpret the reference signals on the one or more probing beams 808 as received.

[0064] The UE 804 scans 812 through one or more of its own Rx beams with respect to the one or more probing beams 808. In other words, the UE 804 takes reference signal measurements of the reference signals on the one or more probing beams 808 using one or more of its own Rx beams, and stores a measured reference signal strength (e.g., RSRP) corresponding to each such measurement.

[0065] The set of one or more Rx beams to be used for these measurements may be selected to correspond to the beam direction and/or codebook information for the one or more probing beams 808 as received from the base station 802. For example, the set of one or more Rx beams may be selected from an Rx codebook that is understood to correspond to a Tx codebook indicated by the base station 802, and/or the selection may be based on indicated beam directions from the base station 802.

[0066] In some cases, the UE 804 uses all available Rx beams at the UE 804 (e.g., all Rx beams of a selected Rx codebook) for this procedure. In other cases, the UE 804 may be configured to use only a subset of all the available Rx beams at the UE 804.

[0067] As a result of these measurements, the UE 804 obtains measured reference signal strengths 814 for various Tx-Rx beam pairs (with each such measurement corresponding to a unique combination of one of the one or more probing beams 808 and one of the Rx beams used to measure the reference signals on those one or more probing beams 808, as described).

[0068] These measured reference signal strengths 814 for these Tx-Rx beam pairs are then provided to the AI/ML model 806, which is trained to use the measured reference signal strengths 814 to generate predicted receive signal strengths 816 of a larger overall set (e.g., all) of Tx-Rx beam pairs between the Tx beams of the base station and the Rx beams of the UE.

[0069] The predicted receive signal strengths 816 may be understood in terms of a reference signal strength map 818 for the relationship between the base station 802 and the UE 804. FIG. 8 illustrates the reference signal strength map 818 in terms of three dimensions. The X dimension 820 and the Y dimension 822 respectively correspond to horizontal and vertical indexes that together identify applicable Tx beams of the base station 802 to which a predicted reference signal strength applies. The Z dimension 824 then contains multiple X-Y planes of such reference signal strengths, where each individual plane represents reference signal strengths for a different one of the Rx beams of the UE 804 with respect to the Tx beams of the base station 802 (as illustrated). Accordingly, the reference signal strength map 818 is understood to contain predicted reference signal strengths (as predicted by the AI/ML model 806) for the larger overall set (e.g., all) of Tx-Rx beam pairs between the base station 802 and the UE 804.

[0070] The UE 804 then identifies a number K of predicted best Tx-Rx beam pairs that correspond to the highest reference signal strengths among the Tx-Rx beam pairs represented within the reference signal strength map 818. The value of K may be (e.g., previously) configured to the UE 804 by the base station 802, or may be pre-configured per a specification for the type of wireless communication system of the base station 802 and the UE 804. Example values for K include 2, 4, 8, etc.

[0071] With respect to disclosure herein, such a number K of predicted best Tx-Rx beam pairs may sometimes be more simply referred to as the “top-X beam pairs.” Further, when, as in the case of FIG. 8, the UE 804 uses an AI/ML model 806 to generate a reference signal strength map 818, and where the applicable type of reference signal strength is an RSRP, these predicted reference signal strengths associated with these top-X beam pairs may be denoted RSRP_AI@UEtop-K.

[0072] FIG. 8 illustrates (by way of example) a case where X = 4. Accordingly, as illustrated, four highest reference signal strengths are identified from the reference signal strength map 818, which ultimately makes the UE 804 aware of the Tx beam and the Rx beam for each of the set of top-K (K = 4) beam pairs (e.g., according to the correspondence of the X dimension 820, the Y dimension 822, and the Z dimension 824 of the reference signal strength map 818 to these Tx-Rx beams, as has been described). [0073] After identifying the top-K beam pairs, the UE 804 signals 826 the K predicted best Tx beams of the predicted best Tx-Rx beam pairs (the Tx beams represented in the predicted best Tx-Rx beam pairs) to the base station 802. With respect to disclosure herein, such a number K of predicted best Tx beams may sometimes be more simply referred to as the “top-K Tx beams.” FIG. 8 accordingly illustrates one example of top-K Tx beams 828 as may be understood at the base station 802 according to the signaling 826 from the UE 804.

[0074] The base station 802 proceeds to transmit 830 reference signals on the top-K Tx beams 828. During these transmissions, the UE 804 performs reference signal measurements of the reference signals using 832 the K predicted best Rx beams of the predicted best Tx-Rx beam pairs (the Rx beams represented in the predicted best Tx-Rx beam pairs). With respect to disclosure herein, such a number K of predicted best Rx beams may sometimes be more simply referred to as the “top-K Rx beams.” At this part of the procedure, the UE 804 measures a reference signal on a Tx beam using its correspondingly paired Rx beam (as this pairing is understood per the set of the predicted best Tx-Rx beam pairs).

[0075] The UE 804 then identifies a highest received signal strength from among the received signal strengths from these (actual) reference signal measurements. The corresponding one of the predicted best Tx-Rx beam pairs (the predicted best Tx-Rx beam pair associated with the highest measured received signal strength) is identified by the UE 804 as the Tx-Rx beam pair to use for communications between the base station 802 and the UE 804 going forward. Accordingly, the UE 804 reports 834 the Tx beam of this Tx-Rx beam pair to the base station 802. The base station 802 then indicates 836 back to the UE 804 that it has determined to use this Tx beam for subsequent communication with the UE 804. The UE 804 then correspondingly determines to use 838 the Rx beam of this Tx-Rx beam pair for subsequent communication with the base station 802. [0076] In certain embodiments, the processes shown in FIG. 8 for a probing beam subset of a serving cell may be expanded to include probing beam subsets from one or more neighbor cell, where the probing beams subsets may be different from cell to cell.

[0077] For example, FIG. 9 illustrates a flow diagram 900 for the use of an AI/ML model 906 at a UE 904 to identify a best Tx-Rx beam pair between the UE 904 and a serving cell and one or more neighbor cells (e.g., neighbor cell 1 to neighbor cell n), along with various corresponding diagrammatic illustrations, according to embodiments provided herein. As illustrated, a network 902 may transmit 910 reference signals on one or more probing beams 908 to a UE 904. The one or more probing beams 908 may be a subset of all Tx beams available at the network 902 per a selected Tx codebook for the serving cell. Similarly, the UE 904 may receive one or more probing beams 940 of a subset of all Tx beams available from neighbor cell 1 and one or more probing beams 942 of a subset of all Tx beams available from neighbor cell n. The serving cell and the one or more neighbor cells may also transmit 910 information corresponding to the one or more probing beams 908, the one or more probing beams 940, and the one or more probing beams 942. For example, the serving cell may transmit 910 beam directions of the one or more probing beams 908 (e.g., in terms of angle domain information and/or SSB index information for each of the one or more probing beams 908), the neighbor cell 1 may transmit 910 beam directions of the one or more probing beams 940, and the neighbor cell n may transmit 910 beam directions of the one or more probing beams 942. Further, the serving cell and the one or more neighbor cells may also transmit 910 respective Tx codebook sizes for the one or more probing beams 908 (e.g., a size of the Tx codebook from which the one or more probing beams 908 are taken/sampled), the one or more probing beams 940, and the one or more probing beams 942.

[0078] The UE 904 scans 912 through one or more of its own Rx beams with respect to the one or more probing beams 908, the one or more probing beams 940, and the one or more probing beams 942. In other words, the UE 904 takes reference signal measurements of the reference signals on the one or more probing beams 908, the one or more probing beams 940, and the one or more probing beams 942 using one or more of its own Rx beams, and stores a measured reference signal strength (e.g., RSRP) corresponding to each such measurement.

[0079] The set of one or more Rx beams used for these measurements may be selected to correspond to the beam direction and/or codebook information for the one or more probing beams 908, the one or more probing beams 940, and the one or more probing beams 942. For example, the set of one or more Rx beams may be selected from an Rx codebook that is understood to correspond to Tx codebooks indicated by the serving cell and the one or more neighbor cells, and/or the selection may be based on beam directions indicated by the serving cell and the one or more neighbor cells. The UE 904 may use all available Rx beams at the UE 904 (e.g., all Rx beams of a selected Rx codebook) for this procedure. In other cases, however, the UE 904 may be configured to use only a subset of all the available Rx beams at the UE 904.

[0080] As a result of the measurements, the UE 904 obtains measured reference signal strengths 914 for various Tx-Rx beam pairs (with each such measurement corresponding to a unique combination of one of the one or more probing beams 908, one or more probing beams 940, and one or more probing beams 942 and one of the Rx beams used to measure the reference signals on those probing beams, as described).

[0081] The measured reference signal strengths 914 for these Tx-Rx beam pairs are then provided to the AI/ML model 906. In certain embodiments, as discussed below, the AI/ML model 906 is a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions (e.g., corresponding to the serving cell and the one or more neighbor cells). Thus, the UE 904 may provide the measured reference signal strengths 914 and assistance information 944 in the form of beam spatial directions as inputs to the AI/ML model 906 to generate predicted receive signal strengths 916 of larger respective sets of Tx-Rx beam pairs between the Tx beams of the serving cell and the one or more neighbor cells.

[0082] For each cell, the predicted receive signal strengths 916 may be understood in terms of a reference signal strength map 918 for the relationship between the cell (e.g., serving cell or neighbor cell) and the UE 904. FIG. 9 illustrates the reference signal strength map 918 in terms of three dimensions. The X dimension 920 and the Y dimension 922 respectively correspond to horizontal and vertical indexes that together identify applicable Tx beams of the network 902 to which a predicted reference signal strength applies. The Z dimension 924 then contains multiple X-Y planes of such reference signal strengths, where each individual plane represents reference signal strengths for a different one of the Rx beams of the UE 904 with respect to the Tx beams of the cell (as illustrated). Accordingly, the reference signal strength map 918 is understood to contain predicted reference signal strengths (as predicted by the AI/ML model 906) for the respective sets of Tx-Rx beam pairs between the particular cell and the UE 904.

[0083] The UE 904 then identifies a number K of predicted best Tx-Rx beam pairs that correspond to the highest reference signal strengths among the Tx-Rx beam pairs represented within the reference signal strength map 918. In certain embodiments, the UE 904 identifies K predicted best Tx-Rx beam pairs per cell. In other embodiments, the UE 904 identifies K predicted best Tx-Rx beam pairs among the serving cell and one or more neighbor cells. The value of K may be configured to the UE 904 by the network 902, or may be pre-configured per a specification for the type of wireless communication system of the network 902. Example values for K include 2, 4, 8, etc. After identifying the top-K beam pairs, the UE 904 signals 926 the K predicted best Tx beams of the predicted best Tx-Rx beam pairs (the Tx beams represented in the predicted best Tx-Rx beam pairs) to the network 902.

[0084] The network 902 proceeds to transmit 930 reference signals on the top-K Tx beams 928. In certain embodiments, each of the serving cell and the one or more neighbor cells transmits its respective top-K Tx beams 928. In other embodiments, the network 902 transmits the overall top-K Tx beams 928 from among the serving cell and the one or more neighbor cells. During these transmissions, the UE 904 performs reference signal measurements of the reference signals using 932 the K predicted best Rx beams of the predicted best Tx-Rx beam pairs (the Rx beams represented in the predicted best Tx-Rx beam pairs). At this part of the procedure, the UE 904 measures a reference signal on a Tx beam using its correspondingly paired Rx beam (as this pairing is understood per the set of the predicted best Tx-Rx beam pairs).

[0085] The UE 904 then identifies a highest received signal strength from among the received signal strengths from these (actual) reference signal measurements. The corresponding one of the predicted best Tx-Rx beam pairs (the predicted best Tx-Rx beam pair associated with the highest measured received signal strength) is identified by the UE 904 as the Tx-Rx beam pair to use for communications between the network 902 and the UE 904 going forward. Accordingly, the UE 904 reports 934 the Tx beam of this Tx-Rx beam pair to the network 902. The network 902 then indicates 936 back to the UE 904 that it has determined to use this Tx beam for subsequent communication with the UE 904. The UE 904 then correspondingly determines to use 938 the Rx beam of this Tx-Rx beam pair for subsequent communication with the network 902.

[0086] As discussed above, the AI/ML model 906 is a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions (e.g., corresponding to the serving cell and the one or more neighbor cells). Alternatively, different AI/ML models may be trained and used at the UE 904 for different subsets of probing Tx beams (e.g., from different cells). For example, FIG. 10 schematically illustrates an example of training different AI/ML models with different subsets of probing Tx beams, according to certain embodiments. In this example, the one or more probing beams 908 shown in FIG. 9 for the serving cell (SC) may be used to train a first AI/ML model 1002, the one or more probing beams 940 shown in FIG. 9 for neighbor cell 1 may be used to train a second AI/ML model 1004, and the one or more probing beams 942 shown in FIG. 9 for neighbor cell n may be used to train a third AI/ML model 1006. However, employing different AI/ML models at a UE for different subsets may be costly. Further, a performance degradation may be expected when, as shown in FIG. 10 for example, the third AI/ML model 1006 that was trained with the one or more probing beams 942 of neighbor cell n is used during inference based on the one or more probing beams 908 of the serving cell because the interpolation functions are different between the first AI/ML model 1002 and the third AI/ML model 1006.

[0087] For example, FIG. 11 illustrates example graphs showing performance degradation among different subsets during inference, according to certain embodiments. A first graph 1102 shows best beam pair selection accuracy versus number of top-K beams when the first AI/ML model 1002 shown in FIG. 10 that is trained using the one or more probing beams 908 from the serving cell is tested using the same one or more probing beams 908 from the serving cell. A second graph 1104 shows best beam pair selection accuracy versus number of top-K beams when the third AI/ML model 1006 shown in FIG. 10 that is trained using the one or more probing beams 942 from neighbor cell n is tested using the one or more probing beams 908 from the serving cell. A comparison of the first graph 1102 (approaching a 0.9 accuracy) and the second graph 1104 (not exceeding a 0.4 accuracy) shows that there may be a substantial performance degradation when an AI/ML model trained with a specific probing subset is tested with data drawn from a different subset. [0088] Thus, as shown in FIG. 12, rather than using different AI/ML models 1002, 1004, 1006, as discussed with respect to FIG. 10, the AI/ML model 906 shown in FIG. 9 is a unified AI/ML model that is trained using a training data set 1202 including of a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions (e.g., from different cells). Further, the assistance information 944 shown in FIG. 9 may be in the form of beam spatial directions (e.g., (|)Tx, 0T_X for probing Tx beams and (|)R_X for a corresponding Rx beam) provided at the input of the AI/ML model 906 to build a model that incorporates such information.

[0089] FIG. 13 illustrates example graphs showing performance improvement when using assistance information, according to certain embodiments. A first graph 1302 corresponds to an AI/ML scheme, according to embodiments disclosed herein, with trained data drawn from random subsets and assistance information with the number of training Rx beams N^_x ^ain - 4. A second graph 1304 corresponds to an AI/ML scheme, according to embodiments disclosed herein, with trained data drawn from random subsets and assistance information with the number of training Rx beams N^_x ^ain = 3. A third graph 1306 corresponds to an AI/ML scheme with trained data drawn from random subsets and no assistance information. Each of the illustrated graphs was simulated using up to a number of training Tx beams N^_x ^ain = 25.

[0090] A comparison of the first graph 1302 and the third graph 1306 shows that to achieve an 80% beam selection accuracy the UE may need to measure and signal the top- K=26 beams with no assistance information compared to the top-K=7 beams with assistance information. This is a significant performance improvement. Further, a training reduction (TR) may be given by R = 1 -

.

^NTX^NRX

[0091] It may be observed that increasing the number of Rx beams used to train and/or use the AI/ML model for RRM in serving and neighbor cells. For example, FIG. 14 illustrates example graphs showing performance improvement in best beam pair selection accuracy when increasing the number of Rx beams from Rx beams =2 to Rx beams = 4, according to certain embodiments.

[0092] AI/ML Model for Different Transmitter Configurations

[0093] In addition to using a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different spatial beam directions and corresponding assistance information, or in other embodiments, the training data set for the unified AI/ML model may include a mixture of RSRP beam pairs drawn from different Tx implementations (e.g., different Tx antenna spacings used in different cells).

[0094] For example, in some wireless communication systems, a normalized Tx antenna separation A (also referred to herein as antenna spacing A) may be represented in physical dimensions, for a Tx wavelength k, as A (e.g., antenna spacing may be expressed as half- wavelength or 0.5k spacing). Assuming n antenna elements, the normalized length L of the transmitter antenna array is L = nA. As shown in the polar pilot beamforming plots of FIG. 15A to FIG. 15D, for antenna spacings A < 1/2 (FIG. 15B, FIG. 15C, and FIG. 15D), there is a single main lobe.

[0095] For antenna spacings A > 1/2, however, other lobes start to appear. As shown in FIG. 15 A, for A = 1 there is another main lobe (at 0 angle) other than the desired main lobe at 90 degrees. For Tx implementations with antenna spacing A > 1/2, the appearance of additional lobes degrades the prediction ability of the AI/ML model (trained with A = 1/2). Therefore, while the AI/ML model trained with antenna spacing A = 1/2 may be generalized for antenna spacings A < 1/2, the AI/ML model may not be generalized for antenna spacing A > 1/2.

[0096] Thus, certain embodiments disclosed herein provide a unified AI/ML model based on antenna separation, including antenna spacing A > 1/2.

[0097] In certain embodiments, the processes shown in FIG. 8 for a probing beam subset of a serving cell may be expanded to include probing beam subsets from one or more neighbor cell with different antenna spacings A.

[0098] For example, FIG. 16 illustrates a flow diagram 1600 for the use of an AI/ML model 1606 at a UE 1604 to identify a best Tx-Rx beam pair between the UE 1604 and a serving cell and one or more neighbor cells (e.g., neighbor cell 1 to neighbor cell n), along with various corresponding diagrammatic illustrations, according to embodiments provided herein. In this example, the serving cell has an antenna spacing A = 1/2, neighbor cell 1 has an antenna spacing A = 3/4, and neighbor cell n has an antenna spacing A = 1. Skilled persons will recognize from the disclosure herein that other antenna spacings may also be used, including antenna spacing A < 1/2.

[0099] As illustrated, a network 1602 may transmit 1610 reference signals on one or more probing beams 1608 to the UE 1604. The one or more probing beams 1608 may be a subset of all Tx beams available at the network 1602 per a selected Tx codebook for the serving cell. Similarly, the UE 1604 may receive one or more probing beams 1640 of a subset of all Tx beams available from neighbor cell 1 and one or more probing beams 1642 of a subset of all Tx beams available from neighbor cell n. The serving cell and the one or more neighbor cells may also transmit 1610 information corresponding to the one or more probing beams 1608, the one or more probing beams 1640, and the one or more probing beams 1642. For example, the serving cell may transmit 1610 beam directions of the one or more probing beams 1608 (e.g., in terms of angle domain information and/or SSB index information for each of the one or more probing beams 1608), the neighbor cell 1 may transmit 1610 beam directions of the one or more probing beams 1640, and the neighbor cell n may transmit 1610 beam directions of the one or more probing beams 1642. Further, the serving cell and the one or more neighbor cells may also transmit 1610 respective Tx codebook sizes and antenna spacing for the serving cell, neighbor cell 1 , and neighbor cell n.

[0100] The UE 1604 scans 1612 through one or more of its own Rx beams with respect to the one or more probing beams 1608, the one or more probing beams 1640, and the one or more probing beams 1642. In other words, the UE 1604 takes reference signal measurements of the reference signals on the one or more probing beams 1608, the one or more probing beams 1640, and the one or more probing beams 1642 using one or more of its own Rx beams, and stores a measured reference signal strength (e.g., RSRP) corresponding to each such measurement.

[0101] The set of one or more Rx beams used for these measurements may be selected to correspond to the beam direction and/or codebook information for the one or more probing beams 1608, the one or more probing beams 1640, and the one or more probing beams 1642. For example, the set of one or more Rx beams may be selected from an Rx codebook that is understood to correspond to Tx codebooks indicated by the serving cell and the one or more neighbor cells, and/or the selection may be based on beam directions indicated by the serving cell and the one or more neighbor cells. The UE 1604 may use all available Rx beams at the UE 1604 (e.g., all Rx beams of a selected Rx codebook) for this procedure. In other cases, however, the UE 1604 may be configured to use only a subset of all the available Rx beams at the UE 1604.

[0102] As a result of the measurements, the UE 1604 obtains measured reference signal strengths 1614 for various Tx-Rx beam pairs (with each such measurement corresponding to a unique combination of one of the one or more probing beams 1608, one or more probing beams 1640, and one or more probing beams 1642 and one of the Rx beams used to measure the reference signals on those probing beams, as described). [0103] The measured reference signal strengths 1614 for these Tx-Rx beam pairs are then provided to the AI/ML model 1606. In certain embodiments, as discussed below, the AI/ML model 1606 is a unified AI/ML model that is trained using a training data set including a mixture of RSRP beam pairs drawn from different Tx antenna spacings used in different cells. In certain embodiments, as discussed in relation to FIG. 9, the training data set may also include a mixture of RSRP beam pairs drawn from different subsets or spatial beam directions (e.g., corresponding to the serving cell and the one or more neighbor cells). Thus, the UE 1604 may provide the measured reference signal strengths 1614 and assistance information 1644 in the form of antenna spacing and beam spatial directions as inputs to the AI/ML model 1606 to generate predicted receive signal strengths 1616 of larger respective sets of Tx-Rx beam pairs between the Tx beams of the serving cell and the one or more neighbor cells.

[0104] For each cell, the predicted receive signal strengths 1616 may be understood in terms of a reference signal strength map 1618 for the relationship between the cell (e g., serving cell or neighbor cell) and the UE 1 04. FIG. 16 illustrates the reference signal strength map 1618 in terms of three dimensions. The X dimension 1620 and the Y dimension 1622 respectively correspond to horizontal and vertical indexes that together identify applicable Tx beams of the network 1602 to which a predicted reference signal strength applies. The Z dimension 1624 then contains multiple X-Y planes of such reference signal strengths, where each individual plane represents reference signal strengths for a different one of the Rx beams of the UE 1604 with respect to the Tx beams of the cell (as illustrated). Accordingly, the reference signal strength map 1618 is understood to contain predicted reference signal strengths (as predicted by the AI/ML model 1606) for the respective sets of Tx-Rx beam pairs between the particular cell and the UE 1604.

[0105] The UE 1604 then identifies a number K of predicted best Tx-Rx beam pairs that correspond to the highest reference signal strengths among the Tx-Rx beam pairs represented within the reference signal strength map 1618. In certain embodiments, the UE 1604 identifies K predicted best Tx-Rx beam pairs per cell. In other embodiments, the UE 1604 identifies K predicted best Tx-Rx beam pairs among the serving cell and one or more neighbor cells. The value of K may be configured to the UE 1604 by the network 1602, or may be pre-configured per a specification for the type of wireless communication system of the network 1602. Example values for K include 2, 4, 8, etc. After identifying the top-K beam pairs, the UE 1604 signals 1626 the K predicted best Tx beams of the predicted best Tx-Rx beam pairs (the Tx beams represented in the predicted best Tx-Rx beam pairs) to the network 1602.

[0106] The network 1602 proceeds to transmit 1630 reference signals on the top-K Tx beams 1628. In certain embodiments, each of the serving cell and the one or more neighbor cells transmits its respective top-K Tx beams 1628. In other embodiments, the network 1602 transmits the overall top-K Tx beams 1628 from among the serving cell and the one or more neighbor cells. During these transmissions, the UE 1604 performs reference signal measurements of the reference signals using 1632 the K predicted best Rx beams of the predicted best Tx-Rx beam pairs (the Rx beams represented in the predicted best Tx-Rx beam pairs). At this part of the procedure, the UE 1604 measures a reference signal on a Tx beam using its correspondingly paired Rx beam (as this pairing is understood per the set of the predicted best Tx-Rx beam pairs).

[0107] The UE 1604 then identifies a highest received signal strength from among the received signal strengths from these (actual) reference signal measurements. The corresponding one of the predicted best Tx-Rx beam pairs (the predicted best Tx-Rx beam pair associated with the highest measured received signal strength) is identified by the UE 1604 as the Tx-Rx beam pair to use for communications between the network 1602 and the UE 1604 going forward. Accordingly, the UE 1604 reports 1634 the Tx beam of this Tx-Rx beam pair to the network 1602. The network 1602 then indicates 1636 back to the UE 1604 that it has determined to use this Tx beam for subsequent communication with the UE 1604. The UE 1604 then correspondingly determines to use 1638 the Rx beam of this Tx-Rx beam pair for subsequent communication with the network 1602.

[0108] As discussed above, the AI/ML model 906 is a unified AI/ML model that is trained using a training data set including at least a mixture of RSRP beam pairs drawn from different Tx antenna spacings used in different cells. Alternatively, different AI/ML models may be trained and used at the UE 904 for different subsets of probing Tx beams (e.g., from different cells with different antenna spacings). For example, FIG. 17 schematically illustrates an example of training different AI/ML models with different subsets of probing Tx beams with different antenna spacings, according to certain embodiments. In this example, the one or more probing beams 1608 shown in FIG. 16 with antenna spacing A = 1/2 for the serving cell (SC) may be used to train a first AI/ML model 1702, the one or more probing beams 1640 shown in FIG. 16 for neighbor cell 1 with antenna spacing A = 3/4 may be used to tram a second AI/ML model 1704, and the one or more probing beams 1642 shown in FIG. 16 for neighbor cell n with antenna spacing A = 1 may be used to train a third AI/ML model 1706. However, employing different AI/ML models at a UE for different subsets may be costly. Further, a performance degradation may be expected when, as shown in FIG. 17 for example, the third AI/ML model 1706 that was trained with the one or more probing beams 1642 of neighbor cell n is used during inference based on the one or more probing beams 1608 of the serving cell because the interpolation functions are different between the first AI/ML model 1702 and the third AI/ML model 1706. Or, vice versa, performance degradation is expected when the third AI/ML model 1706 is trained with antenna spacing A = 1/2 and it is used to predict spatial beams from an implementation with antenna spacing A = 1.

[0109] As shown in FIG. 18, rather than using different AI/ML models 1702, 1704, 1706 as discussed with respect to FIG. 17, the AI/ML model 1606 shown in FIG. 16 is a unified AI/ML model that is trained using a training data set 1802 including of a mixture of RSRP beam pairs drawn from different Tx implementations (e.g., antenna spacings A = 1/k, where k > 2). Further, the assistance information 1644 shown in FIG. 16 may be in the form of the cell’s antenna spacing.

[0110] FIG. 19A to FIG. 19D illustrate example graphs showing performance improvement when using antenna spacing assistance information, according to certain embodiments. For these examples, two different AI/ML models were trained and simulated. The first AI/ML model was not provided with antenna spacing information. The second AI/ML model was trained with various antenna spacing Tx implementations and received as input the antenna spacings during the simulations.

10111] FIG. 19A shows a first graph 1902 generated using the first AI/ML model with no antenna spacing information and a second graph 1904 generated using the second AI/ML model provided with an indication of X = 0.7 antenna spacing. An improvement is observed when using the second AI/ML model with the antenna spacing information. [0112] FIG. 19B shows a first graph 1906 generated using the first AI/ML model with no antenna spacing information and a second graph 1908 generated using the second AI/ML model provided with an indication of X = 0.8 antenna spacing. An improvement is observed when using the second AI/ML model with the antenna spacing information. [0113] FIG. 19C shows a first graph 1910 generated using the first AI/ML model with no antenna spacing information and a second graph 1912 generated using the second AI/ML model provided with an indication of X = 0.9 antenna spacing. An improvement is observed when using the second AI/ML model with the antenna spacing information. [0114] FIG. 19D shows a first graph 1914 generated using the first AI/ML model with no antenna spacing information and a second graph 1916 generated using the second AI/ML model provided with an indication of X = 1 antenna spacing. An improvement is observed when using the second AI/ML model with the antenna spacing information. Further, it can be seen from FIG. 19A to FIG. 19D that the performance improvement increases as the antenna spacing approaches X = 1.

[0115] FIG. 20 is a flowchart of an example method 2000 for a UE, according to certain embodiments. In block 2002, the method 2000 includes receiving, at the UE from a base station, assistance information for a first subset of probing Tx beams corresponding to a serving cell and one or more second subsets of probing Tx beams corresponding to respective neighbor cells. In block 2004, the method 2000 includes performing, at the UE, first reference signal measurements on reference signals transmitted on the first subset of probing Tx beams and the one or more second subsets of probing Tx beams using probed Rx beams of the UE to generate measured received signal strengths. In block 2006, the method 2000 includes providing the assistance information and the measured received signal strengths to an AI/ML model at the UE to generate predicted received signal strengths of Tx-Rx beam pairs. In block 2008, the method 2000 includes identifying, at the UE, one or more predicted candidate Tx-Rx beam pairs based on the predicted received signal strengths of the Tx-Rx beam pairs. In block 2010, the method 2000 includes signaling, from the UE to the base station, one or more candidate Tx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs. In block 2012, the method 2000 includes performing, at the UE, second reference signal measurements on the one or more candidate Tx beams using one or more candidate Rx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs to determine a selected Tx beam. In block 2014, the method 2000 includes reporting, from the UE to the base station, the selected Tx beam. [0116] In certain embodiments of the method 2000, signaling the one or more candidate Tx beams includes signaling the one or more candidate Tx beams in a beam spatial direction domain.

[0117] In certain embodiments of the method 2000, the one or more predicted candidate Tx-Rx beam pairs have one or more highest predicted received signal strengths of the predicted received signal strengths of the Tx-Rx beam pairs, and the selected Tx beam comprises a best Tx beam based on the second reference signal measurements.

[0118] In certain embodiments of the method 2000, the first subset of probing Tx beams is configured according to a first Tx codebook of the serving cell, the one or more second subsets of probing Tx beams are configured according to corresponding second Tx codebooks of the respective neighbor cells, and the probed Rx beams are configured according to an Rx codebook of the UE.

[0119] In certain embodiments of the method 2000, the assistance information indicates beam angles of the first subset of probing Tx beams and the one or more second subsets of probing Tx beams, and the assistance information further indicates respective sizes of the first Tx codebook and the second Tx codebooks. In certain such embodiments, the assistance information further indicates respective beamwidths of the first subset of probing Tx beams and the one or more second subsets of probing Tx beams. In addition, or in other embodiments, the assistance information further indicates respective antenna spacings for the serving cell and the respective neighbor cells.

[0120] In certain embodiments, the method 2000 further includes training the AI/ML model, at the UE, using a training dataset based on RSRP values for training Tx-Rx beam pairs with a mixture of different spatial beam directions and training assistance information indicating the spatial beam directions.

[0121] In certain embodiments, the method 2000 further includes training the AI/ML model, at the UE, using a training dataset based on RSRP values for training Tx-Rx beam pairs with a mixture of different Tx antenna spacings and training assistance information indicating the Tx antenna spacings.

[0122] FIG. 21 is a flowchart of an example method 2100 for a base station, according to certain embodiments. In block 2102, the method 2100 includes transmitting, from the base station to a UE, probing Tx beams and assistance information. The probing Tx beams are selected from Tx beams configured according to a Tx codebook. The assistance information indicates beam directions of the probing Tx beams and a size of the Tx codebook. In block 2104, the method 2100 includes receiving, at the base station from the UE, a first indication of one or more candidate Tx beams of the Tx beams configured according to the Tx codebook. In block 2106, the method 2100 includes transmitting, from the base station to the UE, the one or more candidate Tx beams. In block 2108, the method 2100 includes receiving, at the base station from the UE, a second indication of a selected Tx beam of the candidate Tx beams. In block 2110, the method 2100 includes transmitting, from the base station to the UE, downlink data using the selected Tx beam.

[0123] In certain embodiments of the method 2100, the first indication of the one or more candidate Tx beams includes beam spatial direction information.

[0124] In certain embodiments of the method 2100, the assistance information further indicates respective beamwidths of the probing Tx beams.

[0125] In certain embodiments of the method 2100, the assistance information further indicates respective antenna spacings for the Tx beams configured according to the Tx codebook.

[0126] FIG. 22 illustrates an example architecture of a wireless communication system 2200, according to embodiments disclosed herein. The following description is provided for an example wireless communication system 2200 that operates in conjunction with the LTE system standards and/or 5G or NR system standards as provided by 3GPP technical specifications.

[0127] As shown by FIG. 22, the wireless communication system 2200 includes UE 2202 and UE 2204 (although any number of UEs may be used). In this example, the UE 2202 and the UE 2204 are illustrated as smartphones (e.g., handheld touchscreen mobile computing devices connectable to one or more cellular networks), but may also comprise any mobile or non-mobile computing device configured for wireless communication.

[0128] The UE 2202 and UE 2204 may be configured to communicatively couple with a RAN 2206. In embodiments, the RAN 2206 may be NG-RAN, E-UTRAN, etc. The UE 2202 and UE 2204 utilize connections (or channels) (shown as connection 2208 and connection 2210, respectively) with the RAN 2206, each of which comprises a physical communications interface. The RAN 2206 can include one or more base stations (such as base station 2212 and base station 2214) that enable the connection 2208 and connection 2210. [0129] In this example, the connection 2208 and connection 2210 are air interfaces to enable such communicative coupling, and may be consistent with RAT(s) used by the RAN 2206, such as, for example, an LTE and/or NR.

[0130] In some embodiments, the UE 2202 and UE 2204 may also directly exchange communication data via a sidelink interface 2216. The UE 2204 is shown to be configured to access an access point (shown as AP 2218) via connection 2220. By way of example, the connection 2220 can comprise a local wireless connection, such as a connection consistent with any IEEE 802.11 protocol, wherein the AP 2218 may comprise a Wi-Fi® router. In this example, the AP 2218 may be connected to another network (for example, the Internet) without going through a CN 2224.

[0131] In embodiments, the UE 2202 and UE 2204 can be configured to communicate using orthogonal frequency division multiplexing (OFDM) communication signals with each other or with the base station 2212 and/or the base station 2214 over a multicarrier communication channel in accordance with various communication techniques, such as, but not limited to, an orthogonal frequency division multiple access (OFDMA) communication technique (e.g., for downlink communications) or a single carrier frequency division multiple access (SC-FDMA) communication technique (e.g., for uplink and ProSe or sidelink communications), although the scope of the embodiments is not limited in this respect. The OFDM signals can comprise a plurality of orthogonal subcarriers.

[0132] In some embodiments, all or parts of the base station 2212 or base station 2214 may be implemented as one or more software entities running on server computers as part of a virtual network. In addition, or in other embodiments, the base station 2212 or base station 2214 may be configured to communicate with one another via interface 2222. In embodiments where the wireless communication system 2200 is an LTE system (e.g., when the CN 2224 is an EPC), the interface 2222 may be an X2 interface. The X2 interface may be defined between two or more base stations (e.g., two or more eNBs and the like) that connect to an EPC, and/or between two eNBs connecting to the EPC. In embodiments where the wireless communication system 2200 is an NR system (e.g., when CN 2224 is a 5GC), the interface 2222 may be an Xn interface. The Xn interface is defined between two or more base stations (e.g., two or more gNBs and the like) that connect to 5GC, between a base station 2212 (e.g., a gNB) connecting to 5GC and an eNB, and/or between two eNBs connecting to 5GC (e.g., CN 2224). [0133] The RAN 2206 is shown to be communicatively coupled to the CN 2224. The CN 2224 may comprise one or more network elements 2226, which are configured to offer various data and telecommunications services to customers/subscribers (e.g., users of UE 2202 and UE 2204) who are connected to the CN 2224 via the RAN 2206. The components of the CN 2224 may be implemented in one physical device or separate physical devices including components to read and execute instructions from a machine- readable or computer-readable medium (e.g., a non-transitory machine-readable storage medium).

[0134] In embodiments, the CN 2224 may be an EPC, and the RAN 2206 may be connected with the CN 2224 via an SI interface 2228. In embodiments, the SI interface 2228 may be split into two parts, an SI user plane (Sl-U) interface, which carries traffic data between the base station 2212 or base station 2214 and a serving gateway (S-GW), and the SI -MME interface, which is a signaling interface between the base station 2212 or base station 2214 and mobility management entities (MMEs).

[0135] In embodiments, the CN 2224 may be a 5GC, and the RAN 2206 may be connected with the CN 2224 via an NG interface 2228. In embodiments, the NG interface 2228 may be split into two parts, an NG user plane (NG-U) interface, which carries traffic data between the base station 2212 or base station 2214 and a user plane function (UPF), and the SI control plane (NG-C) interface, which is a signaling interface between the base station 2212 or base station 2214 and access and mobility management functions (AMFs).

[0136] Generally, an application server 2230 may be an element offering applications that use internet protocol (IP) bearer resources with the CN 2224 (e.g., packet switched data services). The application server 2230 can also be configured to support one or more communication services (e.g., VoIP sessions, group communication sessions, etc.) for the UE 2202 and UE 2204 via the CN 2224. The application server 2230 may communicate with the CN 2224 through an IP communications interface 2232.

[0137] FIG. 23 illustrates a system 2300 for performing signaling 2334 between a wireless device 2302 and a network device 2318, according to embodiments disclosed herein. The system 2300 may be a portion of a wireless communications system as herein described. The wireless device 2302 may be, for example, a UE of a wireless communication system. The network device 2318 may be, for example, a base station (e.g., an eNB or a gNB) of a wireless communication system. [0138] The wireless device 2302 may include one or more processor(s) 2304. The processor(s) 2304 may execute instructions such that various operations of the wireless device 2302 are performed, as described herein. The processor(s) 2304 may include one or more baseband processors implemented using, for example, a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a controller, a field programmable gate array (FPGA) device, another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein.

[0139] The wireless device 2302 may include a memory 2306. The memory 2306 may be a non-transitory computer-readable storage medium that stores instructions 2308 (which may include, for example, the instructions being executed by the processor(s) 2304). The instructions 2308 may also be referred to as program code or a computer program. The memory 2306 may also store data used by, and results computed by, the processor(s) 2304.

[0140] The wireless device 2302 may include one or more transceiver(s) 2310 that may include radio frequency (RF) transmitter circuitry and/or receiver circuitry that use the antenna(s) 2312 of the wireless device 2302 to facilitate signaling (e.g., the signaling 2334) to and/or from the wireless device 2302 with other devices (e.g., the network device 2318) according to corresponding RATs.

[0141] The wireless device 2302 may include one or more antenna(s) 2312 (e.g., one, two, four, or more). For embodiments with multiple antenna(s) 2312, the wireless device 2302 may leverage the spatial diversity of such multiple antenna(s) 2312 to send and/or receive multiple different data streams on the same time and frequency resources. This behavior may be referred to as, for example, multiple input multiple output (MIMO) behavior (referring to the multiple antennas used at each of a transmitting device and a receiving device that enable this aspect). MIMO transmissions by the wireless device 2302 may be accomplished according to precoding (or digital beamforming) that is applied at the wireless device 2302 that multiplexes the data streams across the antenna(s) 2312 according to known or assumed channel characteristics such that each data stream is received with an appropriate signal strength relative to other streams and at a desired location in the spatial domain (e.g., the location of a receiver associated with that data stream). Certain embodiments may use single user MIMO (SU-MIMO) methods (where the data streams are all directed to a single receiver) and/or multi user MIMO (MU-MIMO) methods (where individual data streams may be directed to individual (different) receivers in different locations in the spatial domain).

[0142] In certain embodiments having multiple antennas, the wireless device 2302 may implement analog beamforming techniques, whereby phases of the signals sent by the antenna(s) 2312 are relatively adjusted such that the (joint) transmission of the antenna(s) 2312 can be directed (this is sometimes referred to as beam steering).

[0143] The wireless device 2302 may include one or more interface(s) 2314. The interface(s) 2314 may be used to provide input to or output from the wireless device 2302. For example, a wireless device 2302 that is a UE may include interface(s) 2314 such as microphones, speakers, a touchscreen, buttons, and the like in order to allow for input and/or output to the UE by a user of the UE. Other interfaces of such a UE may be made up of transmitters, receivers, and other circuitry (e.g., other than the transceiver(s) 2310/antenna(s) 2312 already described) that allow for communication between the UE and other devices and may operate according to known protocols (e.g., Wi-Fi®, Bluetooth®, and the like).

[0144] The wireless device 2302 may include an RRM module 2316. The RRM module 2316 may be implemented via hardware, software, or combinations thereof. For example, the RRM module 2316 may be implemented as a processor, circuit, and/or instructions 2308 stored in the memory 2306 and executed by the processor(s) 2304. In some examples, the RRM module 2316 may be integrated within the processor(s) 2304 and/or the transceiver(s) 2310. For example, the RRM module 2316 may be implemented by a combination of software components (e.g., executed by a DSP or a general processor) and hardware components (e.g., logic gates and circuitry) within the processor(s) 2304 or the transceiver(s) 2310.

[0145] The RRM module 2316 may be used for various aspects of the present disclosure, for example, aspects of FIG. 2, FIG. 5, and FIG. 20.

[0146] The network device 2318 may include one or more processor(s) 2320. The processor(s) 2320 may execute instructions such that various operations of the network device 2318 are performed, as described herein. The processor(s) 2320 may include one or more baseband processors implemented using, for example, a CPU, a DSP, an ASIC, a controller, an FPGA device, another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein. [0147] The network device 2318 may include a memory 2322. The memory 2322 may be a non-transitory computer-readable storage medium that stores instructions 2324 (which may include, for example, the instructions being executed by the processor(s) 2320). The instructions 2324 may also be referred to as program code or a computer program. The memory 2322 may also store data used by, and results computed by, the processor(s) 2320.

[0148] The network device 2318 may include one or more transceiver(s) 2326 that may include RF transmitter circuitry and/or receiver circuitry that use the antenna(s) 2328 of the network device 2318 to facilitate signaling (e.g., the signaling 2334) to and/or from the network device 2318 with other devices (e.g., the wireless device 2302) according to corresponding RATs.

[0149] The network device 2318 may include one or more antenna(s) 2328 (e.g., one, two, four, or more). In embodiments having multiple antenna(s) 2328, the network device 2318 may perform MIMO, digital beamforming, analog beamforming, beam steering, etc., as has been described.

[0150] The network device 2318 may include one or more interface(s) 2330. The interface(s) 2330 may be used to provide input to or output from the network device 2318. For example, a network device 2318 that is a base station may include interface(s) 2330 made up of transmitters, receivers, and other circuitry (e.g., other than the transceiver(s) 2326/antenna(s) 2328 already described) that enables the base station to communicate with other equipment in a core network, and/or that enables the base station to communicate with external networks, computers, databases, and the like for purposes of operations, administration, and maintenance of the base station or other equipment operably connected thereto.

[0151] The network device 2318 may include an RRM module 2332. The RRM module 2332 may be implemented via hardware, software, or combinations thereof. For example, the RRM module 2332 may be implemented as a processor, circuit, and/or instructions 2324 stored in the memory 2322 and executed by the processor(s) 2320. In some examples, the RRM module 2332 may be integrated within the processor(s) 2320 and/or the transceiver(s) 2326. For example, the RRM module 2332 may be implemented by a combination of software components (e.g., executed by a DSP or a general processor) and hardware components (e.g., logic gates and circuitry) within the processor(s) 2320 or the transceiver(s) 2326. [0152] The RRM module 2332 may be used for various aspects of the present disclosure, for example, aspects of FIG. 3, FIG. 6, and FIG. 21.

[0153] Embodiments contemplated herein include an apparatus comprising means to perform one or more elements of methods described herein for a UE (e.g., method 200, method 500, and/or method 2000). This apparatus may be, for example, an apparatus of a UE (such as a wireless device 2302 that is a UE, as described herein).

[0154] Embodiments contemplated herein include one or more non-transitory computer-readable media comprising instructions to cause an electronic device, upon execution of the instructions by one or more processors of the electronic device, to perform one or more elements of methods described herein for a UE. This non-transitory computer-readable media may be, for example, a memory of a UE (such as a memory 2306 of a wireless device 2302 that is a UE, as described herein).

[0155] Embodiments contemplated herein include an apparatus comprising logic, modules, or circuitry to perform one or more elements of methods described herein for a UE. This apparatus may be, for example, an apparatus of a UE (such as a wireless device 2302 that is a UE, as described herein).

[0156] Embodiments contemplated herein include an apparatus comprising: one or more processors and one or more computer-readable media comprising instructions that, when executed by the one or more processors, cause the one or more processors to perform one or more elements of methods described herein for a UE. This apparatus may be, for example, an apparatus of a UE (such as a wireless device 2302 that is a UE, as described herein).

[0157] Embodiments contemplated herein include a signal as described in or related to one or more elements of methods described herein for a UE.

[0158] Embodiments contemplated herein include a computer program or computer program product comprising instructions, wherein execution of the program by a processor is to cause the processor to carry out one or more elements of methods described herein for a UE. The processor may be a processor of a UE (such as a processor(s) 2304 of a wireless device 2302 that is a UE, as described herein). These instructions may be, for example, located in the processor and/or on a memory of the UE (such as a memory 2306 of a wireless device 2302 that is a UE, as described herein).

[0159] Embodiments contemplated herein include an apparatus comprising means to perform one or more elements of methods described herein for a base station (e.g., method 300, method 600, and/or method 2100). This apparatus may be, for example, an apparatus of a base station (such as a network device 2318 that is a base station, as described herein).

[0160] Embodiments contemplated herein include one or more non-transitory computer-readable media comprising instructions to cause an electronic device, upon execution of the instructions by one or more processors of the electronic device, to perform one or more elements of methods described herein for a base station. This non- transitory computer-readable media may be, for example, a memory of a base station (such as a memory 2322 of a network device 2318 that is a base station, as described herein).

[0161] Embodiments contemplated herein include an apparatus comprising logic, modules, or circuitry to perform one or more elements of methods described herein for a base station. This apparatus may be, for example, an apparatus of a base station (such as a network device 2318 that is a base station, as described herein).

[0162] Embodiments contemplated herein include an apparatus comprising: one or more processors and one or more computer-readable media comprising instructions that, when executed by the one or more processors, cause the one or more processors to perform one or more elements of methods described herein for a base station. This apparatus may be, for example, an apparatus of a base station (such as a network device 2318 that is a base station, as described herein).

[0163] Embodiments contemplated herein include a signal as described in or related to one or more elements of methods described herein for a base station.

[0164] Embodiments contemplated herein include a computer program or computer program product comprising instructions, wherein execution of the program by a processing element is to cause the processing element to carry out one or more elements of methods described herein for a base station. The processor may be a processor of a base station (such as a processor(s) 2320 of a network device 2318 that is a base station, as described herein). These instructions may be, for example, located in the processor and/or on a memory of the base station (such as a memory 2322 of a network device 2318 that is a base station, as described herein).

[0165] For one or more embodiments, at least one of the components set forth in one or more of the preceding figures may be configured to perform one or more operations, techniques, processes, and/or methods as set forth herein. For example, a baseband processor as described herein in connection with one or more of the preceding figures may be configured to operate in accordance with one or more of the examples set forth herein. For another example, circuitry associated with a UE, base station, network element, etc. as described above in connection with one or more of the preceding figures may be configured to operate in accordance with one or more of the examples set forth herein.

[0166] Any of the above described embodiments may be combined with any other embodiment (or combination of embodiments), unless explicitly stated otherwise. The foregoing description of one or more implementations provides illustration and description, but is not intended to be exhaustive or to limit the scope of embodiments to the precise form disclosed. Modifications and variations are possible in light of the above teachings or may be acquired from practice of various embodiments.

[0167] Embodiments and implementations of the systems and methods described herein may include various operations, which may be embodied in machine-executable instructions to be executed by a computer system. A computer system may include one or more general-purpose or special-purpose computers (or other electronic devices). The computer system may include hardware components that include specific logic for performing the operations or may include a combination of hardware, software, and/or firmware.

[0168] It should be recognized that the systems described herein include descriptions of specific embodiments. These embodiments can be combined into single systems, partially combined into other systems, split into multiple systems or divided or combined in other ways. In addition, it is contemplated that parameters, attributes, aspects, etc. of one embodiment can be used in another embodiment. The parameters, attributes, aspects, etc. are merely described in one or more embodiments for clarity, and it is recognized that the parameters, attributes, aspects, etc. can be combined with or substituted for parameters, attributes, aspects, etc. of another embodiment unless specifically disclaimed herein.

[0169] It is well understood that the use of personally identifiable information should follow privacy policies and practices that are generally recognized as meeting or exceeding industry or governmental requirements for maintaining the privacy of users. In particular, personally identifiable information data should be managed and handled so as to minimize risks of unintentional or unauthorized access or use, and the nature of authorized use should be clearly indicated to users.

[0170] Although the foregoing has been described in some detail for purposes of clarity , it will be apparent that certain changes and modifications may be made without departing from the principles thereof. It should be noted that there are many alternative ways of implementing both the processes and apparatuses described herein. Accordingly, the present embodiments are to be considered illustrative and not restrictive, and the description is not to be limited to the details given herein, but may be modified within the scope and equivalents of the appended claims.

Claims

1. A method for a user equipment (UE), comprising: receiving, at the UE from a base station, assistance information for a first subset of probing transmit (Tx) beams corresponding to a serving cell and one or more second subsets of probing Tx beams corresponding to respective neighbor cells; performing, at the UE, first reference signal measurements on reference signals transmitted on the first subset of probing Tx beams and the one or more second subsets of probing Tx beams using probed receive (Rx) beams of the UE to generate measured received signal strengths; providing the assistance information and the measured received signal strengths to an artificial intelligence or machine learning model (AI/ML model) at the UE to generate predicted received signal strengths of Tx-Rx beam pairs; identifying, at the UE, one or more predicted candidate Tx-Rx beam pairs based on the predicted received signal strengths of the Tx-Rx beam pairs; signaling, from the UE to the base station, one or more candidate Tx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs; performing, at the UE, second reference signal measurements on the one or more candidate Tx beams using one or more candidate Rx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs to determine a selected Tx beam; and reporting, from the UE to the base station, the selected Tx beam.

2. The method of claim 1, wherein signaling the one or more candidate Tx beams comprises signaling the one or more candidate Tx beams in a beam spatial direction domain.

3. The method of claim 1, wherein the one or more predicted candidate Tx-Rx beam pairs have one or more highest predicted received signal strengths of the predicted received signal strengths of the Tx-Rx beam pairs, and wherein the selected Tx beam comprises a best Tx beam based on the second reference signal measurements.

4. The method of claim 1, wherein the first subset of probing Tx beams is configured according to a first Tx codebook of the serving cell, wherein the one or more second subsets of probing Tx beams are configured according to corresponding second Tx codebooks of the respective neighbor cells, and wherein the probed Rx beams are configured according to an Rx codebook of the UE.

5. The method of claim 4, wherein the assistance information indicates beam angles of the first subset of probing Tx beams and the one or more second subsets of probing Tx beams, and wherein the assistance information further indicates respective sizes of the first Tx codebook and the second Tx codebooks.

6. The method of claim 5, wherein the assistance information further indicates respective beamwidths of the first subset of probing Tx beams and the one or more second subsets of probing Tx beams.

7. The method of claim 5, wherein the assistance information further indicates respective antenna spacings for the serving cell and the respective neighbor cells.

8. The method of claim 1, further comprising training the AI/ML model, at the UE, using a training dataset based on reference signal received power (RSRP) values for training Tx-Rx beam pairs with a mixture of different spatial beam directions and training assistance information indicating the spatial beam directions.

9. The method of claim 1, further comprising training the AI/ML model, at the UE, using a training dataset based on reference signal received power (RSRP) values for training Tx-Rx beam pairs with a mixture of different Tx antenna spacings and training assistance information indicating the Tx antenna spacings.

10. A method for a base station, comprising: transmitting, from the base station to a user equipment (UE), probing transmit (Tx) beams and assistance information, wherein the probing Tx beams are selected from Tx beams configured according to a Tx codebook, and wherein the assistance information indicates beam directions of the probing Tx beams and a size of the Tx codebook; receiving, at the base station from the UE, a first indication of one or more candidate Tx beams of the Tx beams configured according to the Tx codebook; transmitting, from the base station to the UE, the one or more candidate Tx beams; receiving, at the base station from the UE, a second indication of a selected Tx beam of the candidate Tx beams; and transmitting, from the base station to the UE, downlink data using the selected Tx beam.

11. The method of claim 10, wherein the first indication of the one or more candidate Tx beams includes beam spatial direction information.

12. The method of claim 10, wherein the assistance information further indicates respective beamwidths of the probing Tx beams.

13. The method of claim 10, wherein the assistance information further indicates respective antenna spacings for the Tx beams configured according to the Tx codebook.

14. A method for a user equipment (UE), comprising: receiving, at the UE from a base station, a synchronization signal block (SSB) index mapping between a first transmit (Tx) beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell, wherein the serving cell and the neighbor cell use a same Tx codebook, and wherein the second Tx beam pattern is in a reshuffled order from the first Tx beam pattern; selecting, at the UE, a subset of probing Tx beams from the first Tx beam pattern and the second Tx beam pattern based on the SSB index mapping, wherein the subset of probing Tx beams corresponds to beam angles used to train an artificial intelligence or machine learning model (AI/ML model); performing, at the UE, first reference signal measurements on reference signals transmitted on the subset of probing Tx beams using probed receive (Rx) beams of the UE to generate measured received signal strengths; providing the measured received signal strengths to the AI/ML model at the UE to generate predicted received signal strengths of Tx-Rx beam pairs; identifying, at the UE, one or more predicted candidate Tx-Rx beam pairs based on the predicted received signal strengths of the Tx-Rx beam pairs; performing, at the UE, second reference signal measurements on the one or more candidate Tx beams using one or more candidate Rx beams corresponding to the one or more predicted candidate Tx-Rx beam pairs to determine a selected Tx beam; and reporting, from the UE to the base station, the selected Tx beam.

15. A method for a base station, comprising: transmitting, from the base station to a user equipment (UE), transmit (Tx) beams according to a first Tx beam pattern; transmitting, from the base station to the UE, a synchronization signal block (SSB) index mapping between the first Tx beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell, wherein the serving cell and the neighbor cell use a same Tx codebook for the Tx beams, and wherein the second Tx beam pattern is in a reshuffled order from the first Tx beam pattern; receiving, at the base station from the UE, an indication of a selected Tx beam; and transmitting, from the base station to the UE, downlink data using the selected Tx beam.

16. A method for a user equipment (UE), comprising: receiving, at the UE from a base station, assistance information for a first transmit (Tx) beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell, wherein the serving cell and the neighbor cell use different Tx codebooks, and wherein the assistance information indicates physical angles of probing Tx beams, inquire beam angles, and beamwidths of Tx beams in the first Tx beam pattern and the second Tx beam pattern; selecting, at the UE, a subset of the Tx beams in the first Tx beam pattern and the second Tx beam pattern corresponding to the physical angles of the probing Tx beams used to train an artificial intelligence or machine learning model (AI/ML model); performing, at the UE, reference signal measurements on reference signals transmitted on the subset of probing Tx beams using probed receive (Rx) beams of the UE to generate measured received signal strengths; providing the measured received signal strengths to the AI/ML model at the UE to generate predicted received signal strengths corresponding to the inquire beam angles; and reporting, from the UE to the base station, the predicted received signal strengths corresponding to the inquire beam angles.

17. A method for a base station, comprising: transmitting, from the base station to the UE, transmit (Tx) beams according to a first Tx beam pattern of a serving cell; transmitting, from the base station to a user equipment (UE), assistance information for the first Tx beam pattern of the serving cell and a second Tx beam pattern of a neighbor cell, wherein the serving cell and the neighbor cell use different Tx codebooks, and wherein the assistance information indicates physical angles of probing Tx beams, inquire beam angles, and beamwidths of Tx beams in the first Tx beam pattern and the second Tx beam pattern; receiving, at the base station from the UE, an indication of predicted received signal strengths corresponding to the inquire beam angles; and transmitting, from the base station to the UE, downlink data based on the predicted received signal strengths.

18. A method for a user equipment (UE), comprising: receiving, at the UE from a base station, assistance information for a first transmit (Tx) beam pattern of a serving cell and a second Tx beam pattern of a neighbor cell, wherein the serving cell and the neighbor cell use different Tx codebooks, and wherein the assistance information indicates physical angles of probing Tx beams, inquire beam angles, and beamwidths of Tx beams in the first Tx beam pattern and the second Tx beam pattern; selecting, at the UE, a first artificial intelligence or machine learning model (AI/ML model) for the first Tx beam pattern of the serving cell and a second AI/ML model for the second Tx beam pattern of the neighbor cell based on the assistance information; selecting, at the UE, a first Tx beam subset from the first Tx beam pattern and a second Tx beam subset from the second Tx beam pattern corresponding to the physical angles of the probing Tx beams; performing, at the UE, reference signal measurements on reference signals transmitted on the first Tx beam subset to generate first measured received signal strengths and on the second Tx beam subset to generate second measured received signal strengths; providing the first measured received signal strengths to the AI/ML model at the UE to generate first predicted received signal strengths corresponding to the inquire beam angles; providing the second measured received signal strengths to the AI/ML model at the UE to generate second predicted received signal strengths corresponding to the inquire beam angles; and reporting, from the UE to the base station, the first predicted received signal strengths and the second predicted received signal strengths corresponding to the inquire beam angles.

19. An apparatus comprising means to perform the method of any of claim 1 to claim 18.

20. A computer-readable media comprising instructions to cause an electronic device, upon execution of the instructions by one or more processors of the electronic device, to perform the method of any of claim 1 to claim 18.

21. An apparatus comprising logic, modules, or circuitry to perform the method of any of claim 1 to claim 18.

22. A baseband processor of a user equipment (UE) that is configured to perform the method of any of claim 1 to claim 9, claim 14, claim 16 and claim 18.