Flexible design of type 1 precoding codebook for channel state information (CSI) framework

Abstract

Embodiments of the present disclosure relate to a method (700) for channel state information (CSI) reference signal processing and feedback reporting at a user equipment (UE) (500) The method (700) includes receiving (702), from a base station (gNB) (600), a plurality of non-precoded non- zero power (NZP) CSI reference signal (CSI-RS) resources. Based on the CSI-RS resources, UE (500) determines beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension, derived using array splitting in horizontal and vertical combinations. Co-phasing phase shift indicators are applied to the DFT SD vectors to construct candidate narrow SD beams. A polarization co-phasing phase shift is applied to a subset of these beams to generate a beam for each transmission layer. UE (500) then generates a CSI report comprising information on the selected beam indicators, co-phasing shift indicators, and polarization co-phasing for transmission to the gNB (600).

Description

TITLE: FLEXIBLE DESIGN OF TYPE 1 PRECODING CODEBOOK FOR CHANNEL STATE INFORMATION (CSI) FRAMEWORK FIELD OF TECHNOLOGY

[0001] The present disclosure relates to methods and apparatus for channel state information (CSI) reference signal processing and feedback reporting in large antenna MIMO systems supporting rank 3 and above. BACKGROUND

[0002] The evolution of fifth generation (5G) and future wireless communication systems involves deployments with large-scale antenna arrays and operation at higher frequency bands such as millimeter wave (mmWave) and terahertz (THz). These developments give rise to two key scenarios (i) use of massive MIMO with a large number of antennas at conventional frequencies, and (ii) use of high-frequency bands with relatively narrow directional beams — both of which challenge the current precoding and feedback frameworks.

[0003] In closed-loop MIMO systems, the user equipment (UE) estimates the downlink channel from channel state information reference signals (CSI-RS) transmitted by the base station (gNodeB or gNB). This estimated channel state information (CSI) is reported via uplink control information (UCI), which includes parameters such as the rank indicator (RI), channel quality indicator (CQI), and precoding matrix indicator (PMI). The gNB uses this information for selecting the downlink precoding matrix and performing beamforming for both SU-MIMO and MU-MIMO transmissions.

[0004] In current 3GPP implementations of Type-I codebooks, different strategies are employed based on the number of CSI-RS ports (P₍CSI-RS₎). For P₍CSI-RS₎ < 16, layers 1 and 2 may be assigned with same or different DFT beams, while layers 3 and 4 are separated from the 1stand 2ndlayers, respectively, through polarization co-phasing. For P₍CSI-RS₎ ≥ 16, a split array codebook is used, where a reduced-dimension spatial domain vector is selected from a horizontally split sub-array. This reduced-dimension spatial domain vector is reused across all layers using predefined phase and polarization shifts, limiting adaptability. Although this approach helps reduce UCI size, the approach applies for ranks 3 and 4 alone and also restricts flexibility in selecting beams per layer.

[0005] Precoding in such 3GPP systems often relies on Type-I codebooks. The efficiency and scalability of these split array codebooks are assessed based on: (i) the size of the UCI payload required to report CSI, (ii) the computational complexity of PMI selection at the UE, and (iii) the codebook's ability to support orthogonal beam selection for multi-layer transmissions.

[0006] With the increasing number of antenna elements (e.g., 128 ports and beyond in Release 19 and above), the size of the codebook and corresponding feedback parameters (e.g., ^^1,1, ^^1,2for beam direction) also increases. The current framework, which supports wideband-only reporting of these indicators, becomes insufficient in high-frequency deployments due to beam squint effects — where beam directions deviate across sub-bands. This may create a need for sub-band level feedback of beam parameters to preserve beam alignment and maintain throughput.

[0007] Additionally, current 3GPP implementations only support one-dimensional array splitting, typically in the horizontal dimension, which reduces the beam index space by a single bit. While effective for smaller arrays, this strategy becomes inadequate as antenna array sizes increase, and the number of supported layers grows. For higher-rank configurations such as ranks 5 through 8, legacy codebooks extend existing techniques using fixed spatial offsets to derive orthogonal beams. These approaches may significantly increase CSI feedback overhead and UE-side complexity, especially in systems with narrow beamwidths and fine angular resolution.

[0008] Moreover, as antenna arrays grow, the compute complexity at the UE increases due to the expanded search space for optimal PMI selection, particularly for high-rank scenarios (rank ≥ 3). Simultaneously, existing codebooks limit beam selection flexibility for higher-layer transmissions by enforcing fixed offsets or restricted candidate sets for secondary beams, which is inadequate in scenarios with narrow beamwidths and high spatial resolution.

[0009] To enable scalable high-rank MIMO in such contexts, there is a pressing need to redesign codebooks and feedback mechanisms that reduce UCI overhead, simplify UE-side computations, and allow flexible beam combinations — including support for sub-band level feedback and adaptable array splitting strategies. SUMMARY

[0010] In an embodiment, a method for channel state information (CSI) reference signal processing and feedback reporting performed at a user equipment (UE) is disclosed. The method comprises receiving, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement. Each NZP CSI-RS resource is associated with one of a plurality of CSI-RS ports. Further, the method comprises determining from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions. The set of DFT SD vectors corresponds to wider segmented beams. The method also comprises determining co- phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions. Moreover, the method comprises determining a polarization co-phasing phase shift applied to a subset of the set of candidate narrow SD beams to generate a beam corresponding to each layer of transmission. Lastly, the method comprises generating a CSI report for transmission to the gNB. The CSI report comprises information on the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing.

[0011] In another embodiment, a method for processing channel state information (CSI) report performed at a gNode B (gNB) is disclosed. The method comprises receiving, from a user equipment (UE), a CSI report comprising a precoding matrix indicator (PMI). The PMI including beam indicators corresponding to a set of wide-beam discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension, wherein the wide-beam DFT SD vectors are derived based on an array split configuration in horizontal and vertical antenna dimensions. The method also comprises determining, based on the received beam indicators, a set of DFT SD vectors representing wide segmented beams with reduced directivity. The array split enables segmentation of the antenna array into sub-arrays in horizontal and vertical domains to reduce dimensionality. Further, the method comprises constructing a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD vectors. The co-phasing phase shift indicators specify phase values to combine horizontal and vertical wide-beam segments. Moreover, the method comprises generating a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams. The polarizationco-phasing aligns dual-polarized antenna elements to enhance per-layer beamforming performance. Lastly, the method comprises performing downlink beamforming of data streams based on the generated transmit beams for each transmission layer.

[0012] In an embodiment, a user equipment comprising a memory and at least one processor is disclosed. The at least one processor is communicatively coupled to the memory and is configured to receive, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement. Each NZP CSI-RS resource is associated with one of a plurality of CSI-RS ports. Further, the at least one processor is configured to determine from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions. The set of DFT SD vectors corresponds to a wider segmented beam. Moreover, the at least one processor is configured to determine co-phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions. The at least one processor is also configured to determine a polarization co- phasing phase shift applied to a subset of the set of candidate narrow SD beams to generate a beam corresponding to each layer of transmission. Lastly, the at least one processor is configured to generate a CSI report for transmission to the gNB. The CSI report comprises information on: the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing.

[0013] In another embodiment, a gNodeB (gNB) comprising a plurality of antennas and at least one processor communicatively coupled to the plurality of antennas is disclosed. The at least one processor is configured to receive, from a user equipment (UE), a CSI report comprising a precoding matrix indicator (PMI), the PMI including beam indicators corresponding to a set of near-optimal wide-beam discrete Fourier transform (DFT) spatial domain (SD) basis vectors of reduced dimension. The wide-beam DFT SD vectors are derived based on an array split configuration in horizontal and vertical antenna dimensions. Further, the at least one processor is configured to determine, based on the received beam indicators, a set of DFT SD vectors representing near-optimal wide segmented beams with reduced directivity. The array split enables segmentation of the antennaarray into sub-arrays in horizontal and vertical domains to reduce dimensionality. Moreover, the at least one processor is configured to construct a set of candidate narrow SD beams by applying co- phasing phase shift indicators to the determined wide-beam DFT SD vectors. The co-phasing phase shift indicators specify phase values to combine horizontal and vertical wide-beam segment. The at least one processor is also configured to generate a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams. The polarization co-phasing aligns dual-polarized antenna elements to enhance per-layer beamforming performance. Lastly, the at least one processor is configured to perform downlink beamforming of data streams based on the generated transmit beams for each transmission layer.

[0014] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description. For a better understanding of exemplary embodiments of the present disclosure, together with other and further features and advantages thereof, reference is made to the following description, taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The embodiments of the disclosure itself, as well as a preferred mode of use, further objectives, and advantages thereof, will best be understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings. One or more embodiments are now described, by way of example only, with reference to the accompanying drawings in which:

[0016] FIG. 1 illustrates an environment for channel state information (CSI) reference signal processing, feedback reporting and downlink beamforming, in accordance with some embodiments of the present disclosure;

[0017] FIG. 2A illustrates an antenna panel layout for spatial domain basis vector computation through array splitting at a user equipment (UE), in accordance with some embodiments of the present disclosure;

[0018] FIG. 2B illustrates an example of beam co-phasing performed by a user equipment (UE) based on spatial domain basis vectors obtained through array splitting of an antenna panel, in accordance with some embodiments of the present disclosure;

[0019] FIG. 3A illustrates a schematic representation of antenna array segmentation and wide beam generation at a user equipment (UE) in accordance with some embodiments of the present disclosure;

[0020] FIG.3B illustrates an example of polarization-based beam grouping across segmented antenna subarrays at the gNodeB, in accordance with some embodiments of the present disclosure;

[0021] FIG.4 illustrates orthogonal beam generation and layer-specific precoder construction at a gNodeB, in accordance with some embodiments of the present disclosure;

[0022] FIG.5 illustrates a block diagram of a user equipment (UE) for channel state information (CSI) reference signal processing and feedback reporting, in accordance with some embodiments of the present disclosure;

[0023] FIG.6 illustrates a block diagram of a base station (gNB) for channel state information (CSI) reference signal processing and downlink beamforming, in accordance with some embodiments of the present disclosure;

[0024] FIG. 7 illustrates a flowchart for a method for channel state information (CSI) reference signal processing and feedback reporting performed at a user equipment (UE), in accordance with some embodiments of the present disclosure; and

[0025] FIG. 8 illustrates a flowchart for a method for processing channel state information (CSI) reports at a gNodeB, in accordance with some embodiments of the present disclosure.

[0026] The figures depict embodiments of the disclosure for purposes of illustration only. One skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein. DESCRIPTION OF THE DISCLOSURE

[0027] In the present document, the word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any embodiment or implementation of the present subject matter described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

[0028] While the disclosure is susceptible to various modifications and alternative forms, specific embodiment thereof has been shown by way of example in the drawings and will be described in detail below. It should be understood, however that it is not intended to limit the disclosure to the particular forms disclosed, but on the contrary, the disclosure is to cover all modifications, equivalents, and alternative falling within the spirit and the scope of the disclosure.

[0029] The terms “comprises”, “comprising”, or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a setup, device, or method that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such setup or device or method. In other words, one or more elements in a device or system or apparatus proceeded by “comprises… a” does not, without more constraints, preclude the existence of other elements or additional elements in the device or system or apparatus.

[0030] In the following detailed description of the embodiments of the disclosure, reference is made to the accompanying drawings that form a part hereof, and in which are shown by way of illustration specific embodiments in which the disclosure may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosure, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the present disclosure. The following description is, therefore, not to be taken in a limiting sense.

[0031] Embodiments of the present disclosure relate to methods, a user equipment and a base station for CSI reference signal processing and feedback reporting, and for performing downlink beamforming based on flexible Type-I codebook designs using array splitting in horizontal and / or vertical dimensions, beam co-phasing, and polarization coefficient. The disclosed method addresses three core challenges in next-generation MIMO systems with large antenna arrays: (i) increased uplink control information (UCI) payload due to expansion of beam direction parameters (^^1,1, ^^1,2)with growing antenna dimensions (N1, N2), (ii) high computational complexity in UE-side PMI selection caused by large beam search spaces and the need for layer-specific basis vectors, especially for ranks 3 and above, and (iii) limited flexibility in current codebook designs to select orthogonal beams for higher-rank transmissions, as current standards constrain secondary beam offsets (e.g., via k1, k2). The present disclosure enables sub-band feedback of beam indicators to mitigate beam squint effects in mmWave and THz bands, reduces UE-side computation by enabling wide-beam derivation through reduced-dimension array splits, and increases spatial reuse through support for multiple narrow beams and polarization-based layer multiplexing. These enhancements support scalable, high-rank beamforming in evolving 5G and future cellular standards.

[0032] FIG.1 illustrates an environment 100 for channel state information (CSI) reference signal processing, feedback reporting, and downlink beamforming, in accordance with some embodiments of the present disclosure. The environment 100 may include a user equipment (UE) 101 and a base station (gNB) 103 operating in a wireless communication network supporting multiple antenna ports.

[0033] The UE 101 may refer to the device(s) used by end-users to access a 3GPP network system and its services. These devices, such as smartphones, IoT gadgets, or other wireless-enabled endpoints, may act as the entry point for communication with the network. However, the UE 101 is not limited to the above examples and any other device having 5G communication capability or higher may be well within the scope of present disclosure. The user equipment 101 may be configured to receive a CSI reference signal (RS) transmitted by the gNB 103, wherein the CSI-RS is associated with a plurality of antenna ports mapped to physical antenna elements. The UE 101 may be further configured to determine, based on the received CSI reference signal, one or more spatial domain basis vectors of reduced dimension representing wide beams. The UE 101 may also determine one or more phase shift indicators for beam co-phasing across segmented arrays, and one or more polarization co-phasing indicators for generating multiple beams per layer. The user equipment 101 may be configured to generate a CSI report including at least the wide-beam indicators, the beam co-phasing phase shift indicators, and the polarization co-phasing indicators, and transmit the CSI report to the gNB 103. The generation of the CSI report by the UE is described in further detail in the embodiments below.

[0034] The gNB 103 may be configured to transmit a CSI-RS mapped to its physical antenna elements to the UE 101. Moreover, the gNB 103 may also be configured to receive, from the UE101, the CSI report comprising the beam-related indicators. The gNB 103 may be further configured to generate a precoding matrix. The gNB 103 may then apply the generated precoding matrix to one or more downlink data streams for transmission to the user equipment 101 using transmit beamforming. The generation of the precoding matrix and performing downlink transmission by the gNB 103 are described in further detail in the embodiments below.

[0035] FIG. 2A illustrates an antenna panel layout 205 for spatial domain basis vector computation through array splitting at a user equipment (UE) 201, in accordance with some embodiments of the present disclosure.

[0036] The antenna panel 205 may include a plurality of physical antenna elements arranged in a two-dimensional rectangular grid comprising N1 antenna elements along a horizontal dimension and N2 antenna elements along a vertical dimension. Each physical antenna element may support two orthogonal polarizations, resulting in a total of PCSI−RS = 2 N1 N2 CSI reference signal ports.

[0037] The UE 201 may be configured to receive a channel state information reference signal (CSI-RS) transmitted by a gNB 203 across the CSI ports associated with the antenna panel 205. The CSI-RS may be used by the UE 201 to estimate the downlink channel conditions over the antenna ports.

[0038] Based on the received CSI-RS, the UE 201 may perform an array splitting operation on the antenna panel 205 in at least one of the horizontal or vertical dimensions. The array splitting may divide the panel into sub-arrays, where each sub-array may comprise a subset of antenna elements used to compute a spatial domain basis vector of reduced dimension. The spatial domain basis vectors obtained from the sub-arrays may represent wide beams with relatively broad coverage characteristics. The array splitting may be applied to reducecomputational complexity in beam selection and feedback generation.

[0039] Further, the UE 201 may select one or more spatial domain basis vectors from among those determined by the array splitting. Each selected basis vector may be represented by one or more feedback indicators. These indicators may include i1,1which may specify the horizontal spatial domain basis vector, and i1,2,which may specify the vertical spatial domain basis vector. The values of i1,1 and i1,2 may be selected from a grid of oversampled beamdirections, where the horizontal and vertical grids may contain N1O1and N2O2discrete beamdirections, respectively. O1 and O2 may denote horizontal and vertical oversampling factors.

[0040] The spatial domain basis vectors may be selected to reduce the size of the CSI feedback and limit the computational burden associated with searching over dense beam grids. As the values of N1 and N2 increase, the size of the beam grid may grow significantly, leading to an increase in receiver-side processing complexity. The use of array-split wide-beam vectorsmay address this issue by providing a reduced set of candidate beams with sufficient spatialcoverage, particularly for systems supporting higher-rank transmissions.

[0041] The wide-beam indicators may be included in a CSI report generated by the UE 201 and transmitted to the gNB 203. The gNB 203 may use the reported indicators, along with additional beam-related indicators, to generate a precoding matrix for downlink data transmission. The reported spatial domain basis vectors may serve as inputs to further processing such as beam co-phasing and polarization co-phasing, discussed in detail in the embodiments below.

[0042] The procedure for computing spatial domain basis vectors using array splitting, selecting appropriate feedback indicators, and transmitting the CSI report is described in greater detail in the embodiments below.

[0043] FIG. 2B illustrates an example of beam co-phasing performed by a user equipment (UE) 201 based on spatial domain basis vectors obtained through array splitting of an antennapanel 205, in accordance with some embodiments of the present disclosure.

[0044] As described in FIG.2A, the UE 201 may apply an array split to an antenna panel 205, resulting in multiple sub-arrays. Each sub-array may be associated with a reduced-dimension spatial domain basis vector, which may correspond to a wide beam. The sub-arrays may bederived in at least one of the horizontal or vertical dimensions.

[0045] The UE 201 may apply one or more phase shifts to these spatial domain basis vectors to perform beam co-phasing. For instance, a first phase shift indicator θp may be used to apply a relative phase offset between the basis vectors from different sub-arrays. These co-phasedbasis vectors may then be combined to synthesize a beam with narrower directivity than the individual wide beams.

[0046] The resulting co-phased beam ^̃^^^,^^may enable the UE 201 to generate directional beams suitable for higher-layer MIMO transmission without performing a full-dimensional search. By combining wide-beam basis vectors through phase-aligned addition, the UE 201 may control beam shape while maintaining reduced-dimensional complexity.

[0047] The UE 201 may also encode the selected phase shift indicator(s) into a CSI feedback report and transmit the report to the gNB 203. The gNB 203 may use the received indicator(s) to reconstruct the intended beam structure for downlink precoding. This process allows more flexible and adaptive beam generation based on UE-side computation and configuration.

[0048] The procedure for determining phase shift values and combining spatial domain basis vectors through co-phasing to be used in conjunction with polarization co-phasing, is discussed in further detail in the embodiments below.

[0049] FIG.3A illustrates a schematic representation of antenna array segmentation and wide beam generation at a user equipment (UE) 301 in accordance with some embodiments of the present disclosure.

[0050] The UE 301 may receive channel state information reference signals (CSI-RS) from a gNB 303 via a plurality of antenna ports 305 mapped to physical antenna elements. The gNB antenna panel may be configured as a dual-polarized two-dimensional grid comprising N1 horizontal and N2 vertical antenna elements. The UE 301 may be configured to performantenna array splitting in the horizontal and / or vertical direction, producing reduced -dimension spatial domain (SD) basis vectors that facilitate wide beam generation with reduced directivity and feedback overhead.

[0051] To enable codebook-based CSI acquisition, the UE 301 may generate a 1D discreteFourier transform (DFT) structure of reduced dimension (^^1⁄ ^^) x 1 for beamforming weightsin the horizontal, N1direction, may be, 2^^^^^ ⁄^^ ^̃^2^^^^^(^^ ^^ −1)^^ =^^ ^^ ^^1]

[0052] Similarly, a reduced-dimension DFT structure of (^^2⁄ ^^) x 1 f or vertical weights maybe defined as: 2^2^^^^^^(^^ ⁄ ^^ −1)^^ ^̃^^^ = [1 ^^^^ ^^^^^ ^^2^^2 … ^^^^2^^2^^2]

[0053] Here, r and s may represent the order of antenna array split in the horizontal andvertical domains, respectively, and may be such that ^^^^^^ℎ ^^ℎ^^^^ ^^ = 2,3,4, … ^^1 / 2, ^^ =2,3,4, … ^^2 / 2. The values of r and s may be selected by the gNB 303 and signaled to the UE301 via higher-layer RRC configuration.

[0054] Further, the beam indicating parameters l and m may specify the reduced-dimensionSD vectors corresponding to the subarrays, with ^^ = ^^1,1 ∈ 0, .. , (^^1^^1⁄ r − 1) ) and ^^ = ^^1,2 ∈0, … , (^^2^^2⁄ s − 1) ). The UE 301 may transmit these indices as feedback parameters, therebyreducing CSI reporting overhead to: log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋.

[0055] This may be the foundation of the wide beam feedback design, wherein array-split based wide beams are derived for subsequent co-phasing and layer generation.

[0056] FIG.3B illustrates an example of polarization-based beam grouping across segmented antenna subarrays at the gNodeB 303, in accordance with some embodiments of the present disclosure.

[0057] As shown in FIG. 3B, a 64T64R antenna configuration where the antenna array maybe divided into N1 = 8 and N2 = 4, with vertical and horizontal splits set to ^^ = 2, ^^ = 2. Theresulting four subarrays per polarization may be denoted as G₁–G₄ for the first polarization 305a and G₅–G₈ for the second polarization 305b.

[0058] Each group may be applied with beam co-phasing using a phase shift:^^^^ = ^^ ^^^^^^ / 4, where ^^ ∈ {0, 1, 2, 3} to generate vertically combined vectors:= [^̃^^^ ^^ ^^^^^̃^^^ ]^̃̃̃^ = [^̃^ −^^ ^̃^^^ ^^ ^^ ^^ ^^]

[0059] These vectors may then be combined with the corresponding horizontal basis vectors via Kronecker product: ^̃^^^,^^ = ^̃^^^ ^ ^̃̃^^^^̃̃^^^,^^ = ^̃^^^ ^ ^̃̃̃^^^

[0060] Next, a horizontal co-phasing phase shift:^^ ^^^^^^ / 4^^ = ^^ , where ^^ ∈ {0, 1, 2, 3} may be applied to produce the final beams:^^1 = [^̃^^^ ^^,^^ ^^^^^̃^^^,^^]^^ ^^ ^^

[0061] These four beams maytransmission layers. The values of p and q may be reported by the UE 301 as part of feedback indicator ^^1,3. For further compression, the system may be configured such that p = q, reducing the bitlength of ^^1,3indicator from 4 bits to 2 bits.

[0062] FIG.4 illustrates orthogonal beam generation and layer-specific precoder constructionat a gNodeB 401, in accordance with some embodiments of the present disclosure.

[0063] Based on the received feedback indicators ^^1,1, ^^1,2, ^^1,3 and the inter-polarization co-phasing parameter n, the gNB 401 may reconstruct up to four distinct spatially and ^^^^^^polarization-separated beams (B1–B4), and apply inter-polarization phase shifts ^^^^ = ^^2to transmit up to eight layers using these four SD beams.

[0064] The beams B1 and B2 will belong to the same azimuth plane and either of B1 and B2 may be a split beam. And, the beams B3 and B4 will also belong to the same azimuth plane and either of B3 and B4 may be a split beam. Likewise, B1 and B3 will belong to the sameelevation plane and either of B1 and B3 may be a split beam. And, the beams B2 and B4 willalso belong to the same elevation plane and either of B2 and B4 may be a split beam.

[0065] For rank 3 and 4 transmissions, layer matrices may be constructed based on deployment environment:

[0066] Case 1: N1 > N2; N2 ≥ 4(wide panel):

[0067] Case 1 may be realized for deployments with a large number of scatterers in the azimuth domain as compared to the elevation domain, for example, in a rural or an urban macro or an urban micro open-square deployment. The example realization employs ^^1 and ^^2 beams, without precluding the choice of employing ^^3 and ^^4.

[0068] In 3-layer transmission: 1 ^^ (3)^^1 ^^2 ^^1^^,^^,^^,^^= [^^^^^^1 ^^^^^^2 −^^^^^^1]

[0069] In 4-layer1 ^^ (4)^^1 ^^2 ^^1 ^^2^^,^^,^^,^^= [^^ ^^^^ −^^^^ −^^]

[0070] Case 2: ^^

[0071] Case 2 may be realized for deployments with a large number of scatterers in the elevation domain as compared to azimuth domain, for example, in an urban micro street canyon or an indoor deployment with high rise reflectors. The example realization employs ^^1 and ^^3 beams, without precluding the choice of employing ^^2 and ^^4.

[0072] In 3-layer transmission: 1 ^^ (3) =^^1 ^^3 ^^1^^ ^^]

[0073] In 4-layer1 ^^ (4) = [^^1 ^^3 ^^1 ^^3^^,^^,^^,^^ ^^ ^^1 ^^ ^^] √4^^CSI-RS ^^ ^^ 3 −^^^^^^1 −^^^^^^3

[0074] Case 3:

[0075] Case 3 may be realized for deployments with adequate channel scatterers in theazimuth and elevation domains, without restricting to any deployment scenario.

[0076] In 7-layer transmission: ( 1^^1 ^^2 ^^3 ^^4 ^^1 ^^2^^3 ^^ 7) ^^,^^,^^,^^= [^^^^^^1 ^^^^^^2 ^^^^^^3 ^^^^^^4 −^^^^^^1 −^^^^^^2−^^^^^^3]with any 5 beam combinations that may be selected appropriately.

[0079] These matrices may be configured adaptively depending on whether beam directionality is driven more by horizontal, vertical, or polarization dimensions. The methodology supports high-rank beam generation with scalable feedback and reduced computational overhead, providing flexibility across a range of antenna panel sizes and propagation scenarios.

[0080] The choice of best beams for rank 3, 4, 5 and 6 transmissions can be either a static decision based on the deployment characteristics or mutually agreed upon by the UE and gNB through specific signaling.

[0081] The present disclosure may be applied to panels with ^^1 < 4 or ^^2 < 4, wherein thenumber of antennas in the other dimension is large enoughallow multiple splits creating beams in multiple directions (a minimum of 4 directions to support rank 7 or 8). In that case,either ^^ or ^^ takes a value of 1, where ^^ = 1 indicates no beam segmentation in the horizontaldirection and ^^ = 1 indicates no beam segmentation in the vertical direction. The resulting1D SD vector can be realized with beam co-phasing on each group in the chosen direction,also aligned with the disclosed solution. Hence, the design is equally applicable for antennaarrays with uniform linear array (ULA) structure or uniform planar array (UPA) structure.

[0082] Though the beam squint problem is typically associated with high-frequency bands such as millimeter wave (mmWave) and terahertz, the methodology disclosed herein is not limited to these scenarios. It may also be applicable for communications in microwave andsub-6 GHz frequency bands where large antenna arrays and beamforming are implemented.

[0083] Furthermore, the present disclosure may be applied across a wide range of antenna configurations, including but not limited to 32T32R, 48T48R, 64T64R, 128T128R, and 256T256R arrays. These configurations may be deployed in planar layouts where the vertical dimension (N₂) exceeds the horizontal dimension (N₁), or vice versa. The flexibility to choose different values of N₁ and N₂ may be exploited to meet the spatial resolution needs of different deployment scenarios such as street canyons, open squares, indoor high-rise environments, or rural macro deployments.

[0084] Table 1 below illustrates sample mappings between the number of CSI-RS antenna ports (P₍CSI-RS₎) and the corresponding (N₁, N₂) values for dual-polarized panels. Thisprovides guidance for configuring array splits and beam directions:Number of CSI-RS (^^1, ^^2) ports ^^CSI−RS(6, 4), (4, 6) 48 (8, 3), (3, 8) (12, 2), (2, 12) 64 (8, 4), (4, 8)(16, 2), (2, 16) (8, 6), (6, 8) 96 (12, 4), (4, 12) (16,3), (3, 16) (8, 8) 128 (16, 4), (4, 16) (16, 8), (8, 16) 256 (32,4), (4, 32) (64, 2), (2, 64) Table 1

[0085] This guidance for configuring array splits and beam directions provides a scalable, implementation-friendly approach to generate directional beams and support higher transmission ranks (up to 8 layers), with significantly reduced CSI feedback and PMI search complexity. In particular, the feedback for the strongest SD basis vector may be compressedto log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋.^^ ^^ ^^ ^

[0086] Furthermore, PMI search complexity may be reduced to1 1 2^2^^^^ .

[0087] This reduction in complexity may be especially to side implementation in terms of search latency and power consumption. Even when beam co-phasing indicators p and q are distinct (requiring 4 bits), the major feedback overhead is still dominated by ^^1,1and^^1,2, which specify the spatial beam directions. Therefore, overall performance remains robust while minimizing complexity.

[0088] The number of bits required for co-phasing indicator ^^1,3does not scale with N₁ and N₂, remaining at either 2 bits (when p = q) or 4 bits (when p ≠ q), regardless of array size. This ensures scalability without linear feedback growth. In scenarios where PMI indicators ofhorizontal and vertical beams (i.e., ^^1,1, ^^1,2, ^^1,3) are reported as wideband parameters, ^^ = ^^may be used for overhead reduction. Hence, ^^1,3may be indicated with a single feedbackindicator of 2 bits for ^^ (^^) when ^^ = ^^ is chosen. However, for the subband reporting case,overhead / complexity can be traded off with performance by using distinct ^^ and ^^ thereby having separate feedback indicators for horizontal and vertical co-phase phase shifting between the split arrays.

[0089] Unlike prior approaches such as the 3GPP Rel-15 Type-I codebook, which restrict array splitting to the horizontal direction and to a factor of 2, the present disclosure introduces a flexible array segmentation mechanism. The methodology enables splitting in either or both dimensions, and with arbitrary valid factors, thereby allowing dynamic tailoring of beam patterns as per deployment geometry.

[0090] As a result, the present methodology enables dynamic beam selection from a larger candidate set, customizable via configuration of split size r, s and co-phasing parameters p, q. These configurations may be signaled via higher-layer RRC signaling or via low-layer signaling such as downlink control information (DCI) or medium access control (MAC) control element (MAC CE).

[0091] Therefore, the present disclosure provides improvements in cost-efficiency, component simplification, and low-power operation for CSI reference signal processing and feedback. Further, the methodology supports next-generation wireless deployments with enhanced directional granularity, reduced UCI overhead, and scalable support for large antenna panels and higher frequency operation.

[0092] FIG. 5 illustrates a block diagram of a user equipment (UE) 500 for channel state information (CSI) reference signal processing and feedback reporting, in accordance withsome embodiments of the present disclosure. In an embodiment, the UE 500 may be similar to the UEs 101, 201, 301 of figs.1, 2A and 3A.

[0093] In an embodiment of the present disclosure, the UE 500 may comprise a memory 503, at least one processor 501, a transceiver 505, and an input / output (IO) unit 507 and a communication interface 509 communicatively coupled with each other.

[0094] It may be noted that, in some embodiments, the UE 500 may include more or fewer components than those depicted herein. The various components of the UE 500 may be implemented using hardware, software, firmware or any combinations thereof. Further, the various components of the UE 500 may be operably coupled with each other. More specifically, various components of the UE 500 may be capable of communicating with each other using communication channel media (such as buses, interconnects, etc.).

[0095] In one embodiment, the at least one processor 501 may be embodied as a multi-core processor, a single core processor, or a combination of one or more multi-core processors and one or more single core processors. For example, the at least one processor 501 may be embodied as one or more of various processing devices, such as a coprocessor, a microprocessor, a controller, a digital signal processor (DSP), a processing circuitry with or without an accompanying DSP, or various other processing devices including, a microcontroller unit (MCU), a hardware accelerator, a special- purpose computer chip, or the like.

[0096] The processor 501 may include one or a plurality of processors. At this time, one or a plurality of processors may be a general-purpose processor, such as a central processing unit (CPU), an application processor (AP), or the like, a graphics-only processing unit such as a graphics processing unit (GPU), a visual processing unit (VPU), and / or an AI-dedicated processor such as a neural processing unit (NPU).

[0097] The one or a plurality of processors control the processing of the input data in accordance with a predefined operating rule or artificial intelligence (AI) model stored in the non-volatile memory and the volatile memory. The predefined operating rule or artificial intelligence model is provided through training or learning.

[0098] In one embodiment, the memory 503 is capable of storing machine executable instructions, referred to herein as instructions. In an embodiment, the at least one processor 501 is embodied as an executor of software instructions. As such, the at least one processor 501 is capable of executing the instructions stored in the memory 503 to perform one or more operations described herein.

[0099] The memory 503 can be any type of storage accessible to the at least one processor 501 to perform respective functionalities. For example, the memory 503 may include one or more volatile or non-volatile memories, or a combination thereof. For example, the memory 503 may be embodied as semiconductor memories, such as flash memory, mask ROM, PROM (programmable ROM), EPROM (erasable PROM), RAM (random access memory), etc. and the like.

[0100] In an embodiment of the present disclosure, the at least one processor 501 may be configured to receive, from a gNB, a plurality of non-precoded, non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement. Each NZP CSI-RS resource may be associated with one of a plurality of CSI-RS ports. The at least one processor 501 may be further configured to determine from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions. The set of DFT SD vectors may correspond to a wider segmented beam.

[0101] The at least one processor 501 may be configured to determine co-phasing phase shift indicators on the set of SD basis vectors to generate a plurality of candidate narrow SD beams by combining the wider segmented beams in the horizontal and vertical dimensions. The at least one processor 501 may also be configured to determine a polarization co-phasing phase shift applied to a subset of the candidate narrow SD beams to generate a beam corresponding to each transmission layer. These beams may be applied to support one or more layers for varying transmission ranks.

[0102] The at least one processor 501 may also be configured to generate a CSI report for transmission to the gNB, wherein the CSI report comprises information on: the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing.

[0103] The at least one processor 501 may be further configured to derive a precoding codeword for each layer associated with using a flexible split array design, wherein the precoding codeword extends beam segmentation in plurality of combinations of horizontal and vertical directions.

[0104] The at least one processor 501 may also be configured to enable adaptable splitting of gNB antennas using split factors r and s in the horizontal and vertical domains, respectively, where r and s may be selected by the gNB and signaled to the UE 501 via higher layer signaling. Additionally, the at least one processor 501 may receive such r and s values from the network via either higher layer signaling or MAC control element (CE) signaling.

[0105] The at least one processor 501 may be further configured to compute and report reduced feedback bits for wide-beam SD basis vector selection, wherein the total feedback overhead may be: log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋.

[0106] The at least one processor 501 may be configured to apply beam co-phasing and polarization co-phasing to the wide-beam SD basis vectors in order to select a beam configuration that maximizes a post-equalization metric, such as signal-to-interference-plus- noise ratio (SINR), channel capacity, or throughput.

[0107] In some embodiments, the beam indicators for the wide-beam SD basis vector entities may be reported at a sub-band granularity and may be selected from a predefined codebook. The at least one processor 501 may be further configured to generate and report a feedback parameter i₁,₃, which may indicate co-phasing phase shifts applied to the horizontal and vertical split arrays via sub-band parameters p and q, respectively, to generate the candidate narrow SD beams

[0108] The at least one processor 501 may configure values of p and q through the feedback parameter i1,3 to select a subset of narrow beams from a larger candidate pool. For high- resolution narrow-beam configurations (e.g., large antenna panels), the at least one processor501 may encode distinct p and q values into the bits of the feedback parameter i1,3. Inwideband feedback scenarios, the at least one processor 501 may configure feedback with distinct values of p and q to allow per-beam tuning across layers. Alternatively, for reduced-antenna, broad-beam configurations, the at least one processor 501 may report i1,3as a 2-bitindicator with p = q, thereby trading off beam search complexity for reduced overhead.

[0109] The at least one processor 501 may further configure CSI reports to include the number of dual cross-polarized CSI-RS ports as PCSI−RS=2 N1N2, where N1and N2may represent the number of antenna elements in the horizontal and vertical dimensions, respectively. The values of N1 and N2 may be selected to suit deployment-specific requirements, such that N1 > N2,N1< N2or N1= N2, depending on whether azimuthal, elevational, or balanced scatterer coverage is desired.

[0110] The at least one processor 501 may be configured to apply the plurality of candidate narrow SD beams to different transmission layers for ranks greater than or equal to one. The beam selection for higher layers may be flexible and may not be constrained by fixed offsets or complex layer-specific codebooks. Additional narrow beams created through multi-level array splits may also be used without restriction.

[0111] The at least one processor 501 may also be configured to select beams either through predefined logic or based on higher layer signaling. Furthermore, the beam configuration for higher ranks may be determined using N1, N2 values, or may be separately indicated throughhigher layer signaling for specific UEs to provide maximum adaptability.

[0112] In some embodiments, the at least one processor 501 may be configured to reduce precoding matrix indicator (PMI) search complexity for wide-beam SD vector selection. This reduction may be achieved by limiting the search space to ^^1^^1^^2^^2 / ^^^^ via smaller values of i1,1 and i1,2, corresponding to horizontal and vertical beamforming basis indices, respectively.

[0113] The at least one processor 501 may also be configured to reduce feedback reporting overhead by a factor of 2, 3, or more, based on the dimension and configuration of the split array. This may apply particularly to sub-band-level SD basis vector reporting for digital precoding scenarios.

[0114] Finally, the beam indicators corresponding to the reduced-dimension SD vectors may be reported per sub-band, with at least one of the horizontal or vertical indices determinedand transmitted independently for each sub-band to support frequency-selective CSI feedback.

[0115] Thus, the UE 501 may provide a highly adaptable and efficient mechanism for CSI feedback reporting in high-rank MIMO systems. By supporting flexible array splitting in both horizontal and vertical dimensions, the UE 501 may enable reduced-dimension SD vector selection with significantly lower feedback overhead. The use of configurable co-phasing and polarization co-phasing may allow the UE to generate a rich set of narrow beams from wide- beam candidates, facilitating optimal beam selection across multiple layers without reliance on fixed-offset or layer-specific codebooks. Additionally, the UE 501 may support sub-band granularity in feedback reporting, scalable signaling of beamforming parameters, and reduced precoding matrix search complexity, making it well-suited for deployment in systems with large antenna arrays, such as 64T64R or 256T256R, operating across mmWave, sub-6 GHz, or microwave bands. Thus, the present disclosure may result in improved spectral efficiency, reduced UE-side computational load, and enhanced adaptability across diverse propagation scenarios.

[0116] FIG.6 illustrates a block diagram of a base station (gNB) 600 for channel state information (CSI) reference signal processing and downlink beamforming, in accordance with some embodiments of the present disclosure. In an embodiment, the gNB 600 may be similar to the gNBs 103, 203, 303 of figs.1, 2A and 3A.

[0117] In an embodiment of the present disclosure, the gNB 600 may comprise a memory 603, at least one processor 601, a transceiver 605, a database 607, and a communication interface 609 communicatively coupled with each other.

[0118] It may be noted that, in some embodiments, the gNB 600 may include more or fewer components than those depicted herein. The various components of the gNB 600 may be implemented using hardware, software, firmware or any combinations thereof. Further, the various components of the gNB 600 may be operably coupled with each other. More specifically, various components of the gNB 600 may be capable of communicating with each other using communication channel media (such as buses, interconnects, etc.).

[0119] In one embodiment, the at least one processor 601 may be embodied as a multi-core processor, a single core processor, or a combination of one or more multi-core processors and one or more single core processors. For example, the at least one processor 601 may be embodied as one or more of various processing devices, such as a coprocessor, a microprocessor, a controller, a digital signal processor (DSP), a processing circuitry with or without an accompanying DSP, or various other processing devices including, a microcontroller unit (MCU), a hardware accelerator, a special- purpose computer chip, or the like.

[0120] The processor 601 may include one or a plurality of processors. At this time, one or a plurality of processors may be a general-purpose processor, such as a central processing unit (CPU), an application processor (AP), or the like, a graphics-only processing unit such as a graphics processing unit (GPU), a visual processing unit (VPU), and / or an AI-dedicated processor such as a neural processing unit (NPU).

[0121] The one or a plurality of processors control the processing of the input data in accordance with a predefined operating rule or artificial intelligence (AI) model stored in the non-volatile memory and the volatile memory. The predefined operating rule or artificial intelligence model is provided through training or learning.

[0122] In one embodiment, the memory 603 is capable of storing machine executable instructions, referred to herein as instructions. In an embodiment, the at least one processor 601 is embodied as an executor of software instructions. As such, the at least one processor 601 is capable of executing the instructions stored in the memory 603 to perform one or more operations described herein.

[0123] The memory 603 can be any type of storage accessible to the at least one processor 601 to perform respective functionalities. For example, the memory 603 may include one or more volatile or non-volatile memories, or a combination thereof. For example, the memory 603 may be embodied as semiconductor memories, such as flash memory, mask ROM, PROM (programmable ROM), EPROM (erasable PROM), RAM (random access memory), etc. and the like.

[0124] In an embodiment, the at least one processor 601 may be configured to receive, from a user equipment (UE), a channel state information (CSI) report comprising a precodingmatrix indicator (PMI). The PMI may include beam indicators corresponding to a set of near-optimal wide-beam discrete fourier transform (DFT) spatial domain (SD) basis vectors ofreduced dimension, wherein the wide-beam DFT SD vectors may be derived by the UE based on an array split configuration in the horizontal and vertical antenna dimensions of the base station.

[0125] The at least one processor 601 may be further configured to determine, based on the received beam indicators, a set of DFT SD basis vectors representing near-optimal wider beams with reduced directivity. The array split configuration may enable segmentation of the antenna array into sub-arrays along both the horizontal and vertical domains, thereby reducing the dimensionality of the codebook while preserving angular resolution.

[0126] The at least one processor 601 may also be configured to construct a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD basis vectors. These co-phasing indicators may specify relative phase values that enable coherent combination of segmented beam components across horizontal and vertical sub-arrays.

[0127] The at least one processor 601 may be further configured to generate a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams. The polarization co-phasing may align the phases of orthogonally polarized antenna elements to form effective per-layer beams with enhanced isolation and directional control.

[0128] Based on the generated transmit beams for each transmission layer, the at least one processor 601 may be configured to perform downlink beamforming of one or more datastreams, such that the transmission per layer is aligned with the channel conditions ind icatedby the UE feedback.

[0129] In some embodiments, the determined DFT SD vectors may correspond to wider beams with reduced directivity, constructed using array split in a plurality of horizontal and vertical combinations. The choice of array split configuration may be based on the deployment layout of the gNB 600, including but not limited to azimuth-dominant, elevation-dominant, or balanced scatterer environments.

[0130] Thus, the gNB 600 may support scalable and deployment-adaptive beamforming based on reduced-dimension CSI feedback received from the UE. By reconstructing wide-beam DFT SD vectors using flexible array splits and combining them through configurable co-phasing and polarization alignment, the gNB 600 may efficiently generate transmission beams for high-rank MIMO scenarios without requiring full-dimensional codebook searches. Thus, the methodology of the present disclosure may reduce computational complexity and signaling overhead at the gNB 600 while enabling fine-grained directional transmission across multiple layers. The gNB 600 may dynamically adapt beamforming strategy based on antenna geometry, frequency band, and propagation environment, thereby improving spectral efficiency, per-layer beam isolation, and overall throughput in next-generation wireless networks.

[0131] FIG. 7 illustrates a flowchart for a method for channel state information (CSI) reference signal processing and feedback reporting performed at a user equipment (UE), inaccordance with some embodiments of the present disclosure.

[0132] At step 701, the method 700 discloses receiving, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement. Each NZP CSI-RS resource may be associated with one of a plurality of CSI-RS ports. The CSI report may include a number of dual cross-polarized CSI-RS ports defined as PCSI-RS=2N1N2 where N1 corresponds to antenna elements in the horizontal dimension and N2 corresponds to antenna elements in the vertical dimension. The values of N1and N2may be selected to enhance the formation of channel scatterers in theazimuth or elevation domain or both, depending on deployment conditions.

[0133] At step 703, the method 700 discloses determining from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT)spatial domain (SD) vectors of reduced dimension using array split ting in a plurality ofcombinations of horizontal and vertical dimensions. The DFT SD vectors may correspond to wider segmented beams. The method 700 further discloses enabling adaptable splitting ofgNB antennas, where the order of antenna array split is into horizontal (r) and vertical (s)domains. The values of r and s may be configured by the network and indicated to the UE through higher layer signaling. Additionally, the method 700 may include receiving, at a UE,from the network, an indication comprising the choice of r and s values for the gNB, which may be received via higher layer signaling or medium access control (MAC) control element (CE) signaling.

[0134] At step 705, the method 700 discloses determining co-phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions. A precoding codeword may be derived for each layer is associated with using a flexible split array design. The precoding codeword extends beam segmentation in plurality of combinations of horizontal and vertical directions. The UE may be configured to report reduced feedback bits for indicating a wide-beam SD basis vector entity, computed as log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋.

[0135] The method 700 also discloses applying a first beam co-phasing to the wide-beam SD basis vector entity and next a polarization co-phasing to the narrow-beam SD basis vector entity to select a configuration that maximizes post-equalization signal-to-interference-plus- noise ratio (SINR), channel capacity, user throughput or a linked metric. The beam indicators corresponding to the wide-beam SD basis vector entity may be reported with either a wide- band granularity or a sub-band granularity.

[0136] The method 700 further discloses configuring the UE to report an indicator i 1,3 whichspecifies beam co-phasing phase shifts applied to the horizontal and vertical splits using parameters p and q, respectively. The parameters p and q may be used to generate a plurality of candidate narrow SD beams. Further, the method 700 discloses configuring the UE to set values of at least one of the parameters p or q via the feedback parameter i1,3 thereby selecting a subset of narrow beams from a larger pool of candidate beams. In configurations using larger antenna arrays and narrow-beam designs, the method 700 discloses configuring the UE to report distinct values of the sub-band parameters p and q in the feedback parameter i1,3. In wideband feedback reporting scenarios, distinct values of p and q may be encoded in i1,3to enable layer-specific optimization. In contrast, for reduced-antenna and broader-beam systems, the UE may report i1,3 using 2-bit encoding when p=q, thereby offering a trade-off between beam search complexity and feedback overhead.

[0137] At step 707, the method 700 discloses determining a polarization co-phasing phase shift applied to a subset of the candidate narrow SD beams to generate a beam corresponding to each layer of transmission. The polarization co-phasing phase shift may be used to align dual-polarized antenna elements, thereby enhancing per-layer beamforming performance. The method 700 further discloses applying the plurality of candidate narrow SD beams totransmission layers for different ranks. The array split design may be f lexible and may offerincreased choices for creating directional beams, such that additional beams generated from a multi-level split are not precluded. This may be achieved without restricting higher-layer beam selection through fixed offsets or requiring complex layer-specific beam selection logic.

[0138] The method 700 also discloses that the UE may be configured to select a plurality of narrow beams, or a subset thereof, from the beams generated through the split array. This selection may be predefined or may be signaled to the UE through higher layer signaling. The method 700 also comprises configuring the adaptation of the candidate narrow SD beams for higher transmission ranks, where such configuration may be linked to the deployment scenario and determined using the N1 and N2 values configured by higher layer RRC signaling. Alternately, the configuration may be signaled for specific UEs through higher layer RRC signaling as a separate indication in addition to N1and N2, or via low-layer signaling in order to provide increased flexibility.

[0139] At step 709, method 700 discloses generating a CSI report for transmission to the gNB. The CSI report may include information on the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators across one or more dimensions, and the polarization co-phasing. The CSI report may further include basis vector indices i1,1 and i1,2, which specify the horizontal and vertical beamforming vectors respectively. The reporting of beam indicators may be performed per sub-band, such that at least one of a horizontal beam index and a vertical beam index is determined and transmitted at a sub-band granularity. The precoding matrix indicator (PMI) search complexity for identifying the strongest wide-beam ^^ ^^ ^^ ^^ SD basis vector may be reduced to1 1 2 2^^^^ as a result of the reduction in feedback search space through compressed representation of i1,1 and i1,2. Additionally, the feedback bits may be further reduced by a factor of 2, 3, or more as a function of the split array dimension to minimize sub-band-level reporting overhead for digital precoding.

[0140] Thus, method 700 enables efficient CSI feedback at the UE by supporting adaptable horizontal and vertical array splits, co-phasing, and polarization alignment. The method 700 also reduces feedback overhead through compressed indicators while allowing flexible beam selection across sub-band and wideband domains. The method 700 also simplifies UE-side processing, supports high-rank transmission, and improves spectral efficiency in diverse deployment scenarios.

[0141] FIG. 8 illustrates a flowchart for a method 800 for processing channel state information (CSI) reports at a gNB, in accordance with some embodiments of the present disclosure.

[0142] At step 801, the method 800 discloses receiving, from a UE, a CSI report comprising a precoding matrix indicator (PMI). The PMI may include beam indicators corresponding to a set of near-optimal wide-beam DFT SD basis vectors of reduced dimension. The wide-beam DFT SD vectors may be derived based on an array split configuration in horizontal and vertical antenna dimensions, as configured or signaled to the UE by the gNB.

[0143] At step 803, the method 800 discloses determining, based on the received beam indicators, a set of DFT SD vectors representing near-optimal wide segmented beams with reduced directivity. The array split may enable segmentation of the antenna array into sub- arrays in horizontal and vertical domains to reduce dimensionality. This spatial segmentation may reflect the feedback structure configured by the network and encoded in the UE's CSI report. In some embodiments, the DFT SD basis vectors may correspond to near-optimal wider beams with reduced directivity using array split in a plurality of combinations of horizontal and vertical dimensions. The choice of horizontal and vertical segmentation may be determined based on the deployment layout of the gNB, such as whether channel scatterersare predominantly present in azimuth, elevation, or both spatial domains. This flexibilityenables the gNB to reconstruct optimized transmission beams across diverse deployment scenarios, including wide, tall, or square array configurations.

[0144] The method 800 also discloses constructing a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD vectors, atstep 805. The co-phasing indicators may specify phase values required to combine horizontal and vertical wide-beam segments derived from split antenna arrays.

[0145] Further, at step 807, the method 800 discloses generating a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams. The polarization co-phasing phase shift may align dual-polarized antenna elements within the gNB array to enhance per-layer beamforming performance and isolation.

[0146] Moreover, at step 809, the method 800 discloses performing downlink beamforming of data streams based on the generated transmit beams for each transmission layer. The transmission beam for each layer may be adapted to the spatial channel feedback received from the UE and constructed using the SD basis and co-phasing logic described above.

[0147] Thus, the method 800 allows the gNB to reconstruct directional transmit beams using reduced-dimension feedback from the UE. By combining segmented beams with co-phasing and polarization shifts, the method 800 supports accurate per-layer beamforming. The method 800 enables the gNB to adapt to different array layouts and deployment environments, enhancing downlink performance with lower complexity.

[0148] The present disclosure may further include the below embodiments:1. A user equipment comprising: a memory; at least one processor, wherein the at least one processor is communicatively coupled to the memory and is configured to: receive, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement, wherein each NZP CSI-RS resource is associated with one of a plurality of CSI-RS ports; determine from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions, wherein the set of DFT SD vectors corresponds to a wider segmented beam;determine co-phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions; determine a polarization co-phasing phase shift applied to a subset of the set of candidate narrow SD beams to generate a beam corresponding to each layer of transmission; and generate a CSI report for transmission to the gNB, wherein the CSI report comprises information on: the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing. 2. The UE of embodiment 1, wherein the at least one processor is further configured to: derive a precoding codeword for each layer associated with using a flexible split array design, wherein the precoding codeword extends beam segmentation in plurality of combinations of horizontal and vertical directions. 3. The UE of embodiment 1, wherein the at least one processor is further configured to: enable adaptable splitting of gNB antennas, wherein an order of antenna array split is into horizontal (r) and vertical (s) domains respectively, and wherein an r antenna array split and an s antenna array split are configured by a network and indicated to the UE through higher layer signaling. 4. The UE of embodiment 1, wherein the at least one processor is further configured to: receive, at a UE, from a network, an indication comprising a choice of horizontal split (r) and vertical split (s) for the gNB, wherein the indication is received via higher layer signaling or a medium access control (MAC) control element (CE) signaling. 5. The UE of embodiment 1, wherein the at least one processor is further configured to:configure the UE to report reduced feedback bits of log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋ for indicating a wide-beam SD basis vector basisentity; and apply first a beam co-phasing to the wide-beam SD basis vector entity and next a polarization co-phasing to the narrow-beam SD basis vector entity to select a configuration that maximizes a post-equalization signal-to-interference-plus-noise ratio (SINR), channel capacity, user throughput, or a linked metric. 6. The UE of embodiment 5, wherein the beam indicators for the wide-beam SD basis vector entity are reported with either a wide-band granularity or a sub-band granularity. 7. The UE of embodiment 1, wherein the at least one processor is further configured to: configure the UE to report indicator ^^1,3, wherein ^^1,3indicates beam co-phasing phase shifts applied to horizontal and vertical splits through parameters ^^ and ^^, respectively, for generating a plurality of candidate narrow SD beams. 8. The UE of embodiment 7, wherein the at least one processor is further configured to: configure the UE by setting values of at least one of the parameters p and q via a feedback parameter ^^1,3to select a subset of narrow beams from a larger set of available beam options. 9. The UE of embodiment 7, wherein the at least one processor is further configured to: configure feedback parameter ^^1,3with more bits comprising distinct values of the parameters ^^ and ^^ for larger antenna, narrow-beam systems for improved beam selection management. 10. The UE of embodiment 9, wherein the at least one processor is further configured to: configure the feedback parameter ^^1,3with more bits comprising distinct values of theparameters ^^, ^^,wherein a choice of SD basis vector (^^1) reporting is wideband.11. The UE of embodiment 7, wherein the at least one processor is further configured to:configure the feedback parameter ^^1,3with reduced bits such that parameters p and q are equal to each other for reduced antenna, broader-beam systems, and for trading-off a beam search complexity. 12. The UE of embodiment 1, wherein the CSI report includes a number of dual cross polarizedCSI-RS ports ^^CSI−RS = 2^^1^^2, wherein ^^1 corresponds to antenna elements in a horizontaldimension and ^^2 to antenna elements in a vertical dimension and wherein ^^1 > ^^2or ^^1 < ^^2 or ^^1 = ^^2are toformation of channel scatterers in orelevation or both the domains. 13. The UE of embodiment 7, wherein the at least one processor is further configured to: apply the plurality of candidate narrow SD beams to layers for different ranks. 14. The UE of embodiment 13, wherein an array split design is flexible with increased choices of creating a set of directional beams, wherein additional beams created from the multi-level split are not precluded, without restricting beam selection for higher layers through fixed offsets, and without enhancing the beam selection for higher layers though complex layer-specific free selection of beams. 15. The UE of embodiment 13, wherein the at least one processor is further configured to: configure the UE for choosing a plurality of narrow beams or part of the plurality of narrow beams created from split array, wherein the choosing of the plurality of narrow beams is predefined or signaled using higher layer signaling. 16. The UE of embodiment 14, wherein the at least one processor is further configured to: configure adaptation of the plurality of candidate narrow SD beams for higher ranks, wherein: the configuring is linked to deployment scenarios and determined using the ^^1, ^^2values configured by higher layer RRC configuration; or the configuring by a network is for specific UEs and signaled through higher layer RRC configuration as a separate indication in addition to the ^^1, ^^2 values or through lowlayer signaling in order to provide higher flexibility.17. The UE of embodiment 15, wherein the pre-coding matrix indicator (PMI) search complexity for identifying strongest wide beam SD basis vector is reduced to ^^1^^1^^2^^2 / ^^^^ due to reduction in ^^1,1and ^^1,2, and wherein ^^1,1corresponds to bits that specify basis vectors for horizontal beam forming and ^^1,2corresponds to bits that specify basis vectors for vertical beam forming. 18. The UE of embodiment 15, wherein the at least one processor is further configured to: provision for reducing feedback bits by a factor of 2, 3 or more as a function of split array dimension to reduce the reporting overhead for sub-band-level SD basis reporting through digital precoding. 19. The UE of embodiment 1, wherein the beam indicators corresponding to the set of DFT SD vectors are reported per sub-band, and wherein at least one of a horizontal beam index and a vertical beam index is determined and transmitted at a sub-band granularity. 20. A gNodeB (gNB) comprising: a plurality of antennas; at least one processor communicatively coupled to the plurality of antennas, wherein the at least one processor is configured to: receive, from a user equipment (UE), a CSI report comprising a precoding matrix indicator (PMI), the PMI including beam indicators corresponding to a set of near-optimal wide-beam discrete Fourier transform (DFT) spatial domain (SD) basis vectors of reduced dimension, wherein the wide-beam DFT SD vectors are derived based on an array split configuration in horizontal and vertical antenna dimensions; determine, based on the received beam indicators, a set of DFT SD vectors representing near-optimal wide segmented beams, wherein the array split enables segmentation of the antenna array into sub-arrays in horizontal and vertical domains to reduce dimensionality; construct a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD vectors, wherein the co-phasing phaseshift indicators specify phase values to combine horizontal and vertical wide-beam segments; generate a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams, wherein the polarization co-phasing aligns dual-polarized antenna elements to enhance per-layer beamforming performance; and perform downlink beamforming of data streams based on the generated transmit beams for each transmission layer. 21. The gNB of embodiment 20, wherein the determined DFT SD basis vectors correspond to near-optimal wider beams with reduced directivity using array split in a plurality of combinations of horizontal and vertical dimensions based on deployment layout of the gNB.

Claims

Claims:

1. A method for channel state information (CSI) reference signal processing and feedback reporting performed at a user equipment (UE), the method comprising: receiving, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement, wherein each NZP CSI-RS resource is associated with one of a plurality of CSI-RS ports; determining from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions, wherein the set of DFT SD vectors corresponds to a wider segmented beam; determining co-phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions; determining a polarization co-phasing phase shift applied to a subset of the set of candidate narrow SD beams to generate a beam corresponding to each layer of transmission; and generating a CSI report for transmission to the gNB, wherein the CSI report comprises information on: the beam indicators for the near optimal wide beam SD basis vectors, the co- phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing.

2. The method of claim 1, further comprising: deriving a precoding codeword for each layer is associated with using a flexible split array design, wherein the precoding codeword extends beam segmentation in plurality of combinations of horizontal and vertical directions.

3. The method of claim 1, further comprising: enabling adaptable splitting of gNB antennas, wherein an order of antenna array split is into horizontal (r) and vertical (s) domains respectively, and wherein an r antenna array split andan s antenna array split are configured by a network and indicated to the UE through higher layer signaling.

4. The method of claim 1, further comprising: receiving at a UE, from a network, an indication comprising a choice of horizontal split (r) and vertical split (s) for the gNB, wherein the indication is received via higher layer signaling or a medium access control (MAC) control element (CE) signaling.

5. The method of claim 1, further comprising: configuring the UE to report reduced feedback bits of log2⌊(^^1^^1 / ^^)⌋ + log2⌊(^^2^^2 / ^^)⌋ for indicating a wide-beam SD basis vector basisentity; applying first a beam co-phasing to the wide-beam SD basis vector entity and next a polarization co-phasing to the narrow-beam SD basis vector entity to select a configuration that maximizes a post-equalization signal-to-interference-plus-noise ratio (SINR), channel capacity, user throughput, or a linked metric, and wherein the beam indicators for the wide-beam SD basis vector entity are reported with either a wide-band granularity or a sub-band granularity.

6. The method of claim 1, further comprising: configuring the UE to report indicator ^^1,3, wherein ^^1,3indicates beam co-phasing phase shifts applied to horizontal and vertical splits through parameters ^^ and ^^, respectively, for generating a plurality of candidate narrow SD beams.

7. The method of claim 6, further comprising: configuring the UE by setting values of at least one of the parameters p and q via a feedback parameter ^^1,3to select a subset of narrow beams from a larger set of available beam options; and configuring feedback parameter ^^1,3with more bits comprising distinct values of the parameters ^^ and ^^ for larger antenna, narrow-beam systems for improved beam selection management.

8. The method of claim 7, further comprising: configuring the feedback parameter ^^1,3with more bits comprising distinct values of theparameters ^^, ^^, wherein a choice of SD basis vector (^^1) reporting is wideband.

9. The method of claim 6, further comprising: configuring the feedback parameter ^^1,3with reduced bits such that parameters p and q are equal to each other, for reduced antenna, broader-beam systems, and for trading-off a beam search complexity.

10. The method of claim 1, wherein the CSI report includes a number of dual cross polarizedCSI-RS ports ^^CSI−RS = 2^^1^^2, wherein ^^1 corresponds to antenna elements in a horizontaldimension and ^^2 corresponds to antenna elements in a vertical dimension and wherein ^^1 > ^^2or ^^1 < ^^2 or ^^1 = ^^2 are choices to enhance formation of channel scatterers in azimuth ordomains.

11. The method of claim 6, further comprising: applying the plurality of candidate narrow SD beams to layers for different ranks.

12. The method of claim 11, wherein an array split design is flexible with increased choices of creating a set of directional beams, wherein additional beams created from the multi-level split are not precluded, without restricting beam selection for higher layers through fixed offsets, and without enhancing the beam selection for higher layers through complex layer-specific free selection of beams.

13. The method of claim 11, comprising: configuring the UE for choosing a plurality of narrow beams or part of the plurality of narrow beams created from split array, wherein the choosing the plurality of narrow beams is predefined or signaled using higher layer signaling.

14. The method of claim 12, further comprising:configuring adaptation of the plurality of candidate narrow SD beams for higher ranks, wherein: the configuring is linked to deployment scenarios and determined using the ^^1, ^^2values configured by higher layer RRC configuration; or the configuring by a network is for specific UEs and signaled through higher layer RRC configuration as a separate indication in addition to the ^^1, ^^2 values or through lowlayer signaling in order to provide higher flexibility.

15. The method of claim 13, wherein the pre-coding matrix indicator (PMI) search complexity for identifying strongest wide beam SD basis vector is reduced to ^^1^^1^^2^^2 / ^^^^ due to reduction in ^^1,1and ^^1,2, and wherein ^^1,1corresponds to bits that specify basis vectors for horizontal beam forming and ^^1,2corresponds to bits that specify basis vectors for vertical beam forming.

16. The method of claim 14, further comprising: provisioning for reducing feedback bits by a factor of 2, 3 or more as a function of split array dimension to reduce the reporting overhead for sub-band-level SD basis reporting through digital precoding.

17. The method of claim 1, wherein the beam indicators corresponding to the set of DFT SD vectors are reported per sub-band, such that at least one of a horizontal beam index and a vertical beam index is determined and transmitted at a sub-band granularity.

18. A method for processing channel state information (CSI) report performed at a gNode B (gNB), the method comprising: receiving, from a user equipment (UE), a CSI report comprising a precoding matrix indicator (PMI), the PMI including beam indicators corresponding to a set of wide-beam discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension, wherein the wide- beam DFT SD vectors are derived based on an array split configuration in horizontal and vertical antenna dimensions;determining, based on the received beam indicators, a set of DFT SD vectors representing wide segmented beams with reduced directivity, wherein the array split enables segmentation of the antenna array into sub-arrays in horizontal and vertical domains to reduce dimensionality; constructing a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD vectors, wherein the co-phasing phase shift indicators specify phase values to combine horizontal and vertical wide-beam segments; generating a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams, wherein the polarization co-phasing aligns dual-polarized antenna elements to enhance per-layer beamforming performance; and performing downlink beamforming of data streams based on the generated transmit beams for each transmission layer.

19. The method of claim 18, wherein the determined DFT SD vectors correspond to wider beams with reduced directivity using array split in a plurality of combinations of horizontal and vertical dimensions based on deployment layout of the gNB.

20. A user equipment comprising: a memory; at least one processor, wherein the at least one processor is communicatively coupled to the memory and is configured to: receive, from a gNodeB (gNB), a plurality of non-precoded non-zero power (NZP) channel state information reference signal (CSI-RS) resources for channel measurement, wherein each NZP CSI-RS resource is associated with one of a plurality of CSI-RS ports; determine from the received CSI-RS resources a plurality of beam indicators corresponding to a set of discrete Fourier transform (DFT) spatial domain (SD) vectors of reduced dimension using array splitting in a plurality of combinations of horizontal and vertical dimensions, wherein the set of DFT SD vectors corresponds to a wider segmented beam;determine co-phasing phase shift indicators on the set of DFT SD vectors to create a set of candidate narrow SD beams by combining the wider segmented beams in horizontal and vertical dimensions; determine a polarization co-phasing phase shift applied to a subset of the set of candidate narrow SD beams to generate a beam corresponding to each layer of transmission; and generate a CSI report for transmission to the gNB, wherein the CSI report comprises information on: the beam indicators for the near optimal wide beam SD basis vectors, the co-phasing shift indicators in one or more combinations of horizontal and vertical dimensions, and the polarization co-phasing.

21. A gNodeB (gNB) comprising: a plurality of antennas; at least one processor communicatively coupled to the plurality of antennas, wherein the at least one processor is configured to: receive, from a user equipment (UE), a CSI report comprising a precoding matrix indicator (PMI), the PMI including beam indicators corresponding to a set of near-optimal wide-beam discrete Fourier transform (DFT) spatial domain (SD) basis vectors of reduced dimension, wherein the wide-beam DFT SD vectors are derived based on an array split configuration in horizontal and vertical antenna dimensions; determine, based on the received beam indicators, a set of DFT SD vectors representing near-optimal wide segmented beams, wherein the array split enables segmentation of the antenna array into sub-arrays in horizontal and vertical domains to reduce dimensionality; construct a set of candidate narrow SD beams by applying co-phasing phase shift indicators to the determined wide-beam DFT SD vectors, wherein the co-phasing phase shift indicators specify phase values to combine horizontal and vertical wide-beam segments; generate a transmit beam corresponding to each transmission layer by applying a polarization co-phasing phase shift to a subset of the candidate narrow SD beams, whereinthe polarization co-phasing aligns dual-polarized antenna elements to enhance per-layer beamforming performance; and perform downlink beamforming of data streams based on the generated transmit beams for each transmission layer.

Images