WO2022060825A1

WO2022060825A1 - Device and method for performing beamforming in angle-delay domains

Info

Publication number: WO2022060825A1
Application number: PCT/US2021/050439
Authority: WO
Inventors: Nadisanka Rupasinghe; Haralabos Papadopoulos; Yuki Matsumura
Original assignee: Ntt Docomo, Inc.; Docomo Innovations, Inc.
Priority date: 2020-09-15
Filing date: 2021-09-15
Publication date: 2022-03-24

Abstract

A terminal is disclosed that includes a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain (SD) beam selection. The terminal also includes a processor that considers a plurality of Discrete Fourier Transform (DFT) basis vectors for downlink (DL) CSI-RS beamforming as SD beams and selects the SD beams based on the plurality of DFT basis vectors, the SD beams considering UL spatial covariance.

Description

DEVICE AND METHOD FOR PERFORMING BEAMFORMING IN ANGLE-DELAY DOMAINS

TECHNICAL FIELD

[0001] One or more embodiments disclosed herein relate to a device and a method for performing assisted beamforming in angle-delay domains.

BACKGROUND

[0002] New Radio (NR) supports Type II channel state information (CSI) feedback for rank 1 and rank 2 (Release 15 of NR). In the Type II CSI feedback, an amplitude scaling mode is configured.

[0003] In the amplitude scaling mode, a user equipment (UE) may be configured to report a wideband (WB) amplitude with subband (SB) amplitude and SB phase information. In the conventional scheme, considerable fraction of the total overhead may be occupied by overhead for the SB amplitude and phase reporting. The equation below shows the SB precoder generation in NR Release 15 Type II CSI for single layer transmission.

[0004] W = W_spaceW_coeff (1)

[0005] Here, the matrix W (N_t X N_SB) captures precoding vectors for N_SB subbands. N_t denotes a number of available TXRU ports. W_space (N_t x 2L) consists of a 2L wideband spatial 2D-Discrete Fourier Transform (DFT) beams. The matrix captures the SB combination coefficients as represented in (1) by W_coeff. The SB amplitude and phase information that needs to be reported is captured in W_coeff .

[0006] Further, the NR supports Type II CSI reporting for precoding downlink transmissions on a Physical Downlink Shared Channel (PDSCH). In this regard, Type II solutions focus on providing detailed CSI for the purposes of Multi-User Multiple- Input Multiple-Output (MIMO). In NR Release 15, these solutions support a maximum Rank of 2 corresponding to a maximum of 2 layers per UE (i.e., hereinafter referred to as terminal or device). In NR Release 15, 2x2 MIMO offers two spatial streams of wirelessly transmitting and receiving data on the same channel or frequency. For this implementation, a maximum number of layers per cell is higher to allow multiple UE to use 2x2 MIMO simultaneously while sharing a common Resource Block allocation. Type II reports are based upon selecting a set of beams and then specifying relative amplitudes and phases to generate a weighted combination of beams for each layer of transmission. As such, Type II Port Selection solution relies on a Base Station having some advance information to allow beamforming of the CSI Reference Signal (RS) transmissions. This advance information can originate from uplink measurements if channel reciprocity is available. Otherwise it can originate from Beam Management reports or it can use wideband reports from different Precoding Matrix Indicator (PMI) reporting solutions (i.e., a hybrid solution is when a combination of PMI reporting solutions is used).

[0007] In NR, the majority of parameters associated with PMI reporting are configured using a CodebookConfig parameter structure. This parameter structure uses the combination of codebookType and subtype to identify any relevant PMI reporting solutions. Each PMI reporting solution and the corresponding relevant parameter sets for the Type II Port Selection solution.

CITATION LIST

NON-PATENT REFERENCE

[0008]

[Non-Patent Reference 1] 3GPP RP 193133, “New WID: Further enhancements on MIMO for NR”, Dec., 2019.

[Non-Patent Reference 2] 3GPP TS 38.214, “NR; Physical layer procedures for data (Release 16)”.

SUMMARY [0009] In general, in one aspect, embodiments disclosed herein relate to a terminal that includes a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain (SD) beam selection. The terminal also includes a processor that considers a plurality of Discrete Fourier Transform (DFT) basis vectors for downlink (DL) CSI-RS beamforming as SD beams and selects the SD beams based on the plurality of DFT basis vectors, the SD beams considering UL spatial covariance.

[0010] In general, in one aspect, embodiments disclosed herein relate to a terminal that includes a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain-Frequency Domain (SD- FD) pair selection; and a processor that: considers a plurality of eigen vectors in an uplink (UL) dominant sub-space, the plurality of eigen vectors capturing a set of multipath components (MPCs), and selects SD-FD pairs based on the plurality of eigen vectors.

[0011] In general, in one aspect, embodiments disclosed herein relate to a terminal that includes a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain (SD) beam selection and Frequency Domain (FD) basis selection; and a processor that: considers a bitmap, the bitmap associating SD beams and FD bases to one another, and freely selects SD beams and FD bases based on a rule using on the bitmap.

[0012] In general, in one aspect, embodiments disclosed herein relate to method for performing Sounding Reference Signal (SRS) assisted Channel State Information - Reference Signal (CSI-RS) beamforming in angle-delay domains, the method comprising obtaining beamforming information relating to one or more beamformed CSI-RSs, the beamforming information corresponding to Spatial Domain (SD) beam selection; considering a plurality of Discrete Fourier Transform (DFT) basis vectors for downlink DL CSI-RS beamforming as SD beams; selecting the SD beams based on the plurality of DFT basis vectors, the SD beams considering UL spatial covariance.

[0013] In general, in one aspect, embodiments disclosed herein relate to a method for performing SRS assisted CRI-RS beamforming in angle-delay domains. The method includes obtaining beamforming information relating to one or more CSI-RS, the beamforming information corresponding to SD-FD pair selection. The method includes considering a plurality of eigen vectors in an UL dominant sub-space, the plurality of eigen vectors capturing set of multi-path components (MPCs). The method includes selecting SD-FD pairs based on the plurality of eigen vectors.

[0014] In general, in one aspect, embodiments disclosed herein relate to a method for performing SRS assisted CSI-RS beamforming in angle-delay domains. The method includes obtaining beamforming information relating to one or more CSI-RS, the beamforming information corresponding to SD beam selection and FD basis selection. The method includes considering a bitmap, the bitmap associating SD beams and FD bases to one another. The method includes freely selecting SD beams and FD bases based on a rule using on the bitmap.

[0015] Advantageously, enhancements on CSI measurement and reporting are being discussed in the development of Release 17 of NR. One such enhancement includes evaluating and, if needed, specifying CSI reporting for Downlink (DL) multiTransmission Reception Points (TRP) and/or multi-panel transmission to enable more dynamic channel/interference hypotheses for non-coherent joint transmission (NCJT), targeting both Frequency Range 1 (FR1) (i.e., 410 MHz to 7,125 MHz, sub-6 GHz) and Frequency Range 2 (FR2) (i.e., 24,250 MHz to 52,600 MHz, mmWaves). Another such enhancement includes evaluating and, if needed, specifying Type II port selection codebook enhancements (based on Rel.15/16 Type II port selection) where information related to angle(s) and delay(s) are estimated at a gNB based on Sound Reference Signal (SRS) by utilizing DL/Uplink (UL) reciprocity of angle and delay. The remaining DL CSI is reported by the UE, mainly targeting Frequency Division Duplex (FDD) FR1 to achieve better trade-off among UE complexities, performance, and reporting overhead. [0016] In view of the above enhancements, the present invention describes how Type II port selection codebook can be further enhanced by taking into consideration the inherent angle-delay reciprocity of propagation channel and using UL SRS transmission.

[0017] Other aspects of the disclosure will be apparent from the following description and the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0018] FIG. 1 shows a diagram showing a configuration of a wireless communication system according to one or more embodiments of the present invention.

[0019] FIG. 2 shows a diagram showing a configuration of a wireless communication system according to one or more embodiments of the present invention.

[0020] FIG. 3 A shows an example in accordance with one or more embodiments.

[0021] FIG. 3B shows an example in accordance with one or more embodiments.

[0022] FIG. 4 a diagram showing a configuration of a wireless communication system according to one or more embodiments of the present invention.

[0023] FIG. 5A shows an SD beam selection example in accordance with one or more embodiments.

[0024] FIG. 5B shows an SD beam selection example in accordance with one or more embodiments.

[0025] FIG. 6A shows an SD beam selection example in accordance with one or more embodiments.

[0026] FIG. 6B shows an SD beam selection example in accordance with one or more embodiments.

[0027] FIG. 7A shows an example bitmap in accordance with one or more embodiments. [0028] FIG. 7B shows an example bitmap in accordance with one or more embodiments.

[0029] FIG. 8 shows an example covariance matrix in accordance with one or more embodiments.

[0030] FIG. 9A shows an SD-FD beam pair selection example in accordance with one or more embodiments.

[0031] FIG. 9B shows an SD-FD beam pair selection example in accordance with one or more embodiments.

[0032] FIG. 10 shows an example of a bitmap in accordance with one or more embodiments.

[0033] FIG. 11 shows an example of a bitmap in accordance with one or more embodiments.

[0034] FIG. 12 shows an example covariance matrix in accordance with one or more embodiments.

[0035] FIG. 13 shows a block diagram of an assembly in accordance with one or more embodiments.

[0036] FIG. 14 shows a block diagram of an assembly in accordance with one or more embodiments.

DETAILED DESCRIPTION

[0037] Specific embodiments of the invention will now be described in detail with reference to the accompanying figures. Like elements in the various figures are denoted by like reference numerals for consistency.

[0038] In the following detailed description of embodiments of the invention, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to one of ordinary skill in the art that the invention may be practiced without these specific details. In other instances, well-known features have not been described in detail to avoid unnecessarily complicating the description.

[0039] Throughout the application, ordinal numbers (e.g., first, second, third, etc.) may be used as an adjective for an element (i.e., any noun in the application). The use of ordinal numbers is not to imply or create any particular ordering of the elements nor to limit any element to being a single element unless expressly disclosed, such as by the use of the terms “before”, “after”, “single”, and other such terminology. Rather, the use of ordinal numbers is to distinguish between the elements. By way of an example, a first element is distinct from a second element, and the first element may encompass more than one element and succeed (or precede) the second element in an ordering of elements.

[0040] A wireless communication system 100 according to one or more embodiments of the present invention will be described below with reference to FIG. 1.

[0041] As shown in FIG. 1, the wireless communication system 100 includes a User Equipment (UE) 10, a Base Station (BS) 20, and a core network 30. The wireless communication system 100 may be a New Radio (NR) system or a Long Term Evolution (LTE)/LTE- Advanced (LTE-A) system.

[0042] The BS 20 communicates with the UE 10 via multiple antenna ports using a multiple-input and multiple-output (MIMO) technology. The BS 20 may be a gNodeB (gNB) or an Evolved NodeB (eNB). The BS 20 receives downlink packets from a network equipment such as upper nodes or servers connected on the core network 30 via the access gateway apparatus, and transmits the downlink packets to the UE 10 via the multiple antenna ports. The BS 20 receives uplink packets from the UE 10 and transmits the uplink packets to the network equipment via the multiple antenna ports.

[0043] The BS 20 includes antennas for MIMO to transmit radio signals between the UE 10, a communication interface to communicate with an adjacent BS 20 (for example, X2 interface), a communication interface to communicate with the core network (for example, SI interface), and a CPU (Central Processing Unit) such as a processor or a circuit to process transmitted and received signals with the UE 10. Functions and processing of the BS 20 described below may be implemented by the processor processing or executing data and programs stored in a memory. However, the BS 20 is not limited to the hardware configuration set forth above and may include any appropriate hardware configurations. Generally, a plurality of the BSs 20 may be disposed so as to cover a broader service area of the wireless communication system 1.

[0044] The UE 10 communicates with the BS 20 using MIMO technology. The UE 10 transmits and receives radio signals such as data signals and control signals between the BS 20 and the UE 10. The UE 10 may be a mobile station, a smartphone, a cellular phone, a tablet, a radio terminal, a mobile router, or information processing apparatus having a radio communication function such as a wearable device.

[0045] The UE 10 includes a CPU such as a processor, a RAM (Random Access Memory), a flash memory, and a radio communication device to transmit/receive radio signals to/from the BS 20 and the UE 10. For example, functions and processing of the UE 10 described below may be implemented by the CPU processing or executing data and programs stored in a memory. The UE 10 is not limited to the hardware configuration set forth above and may be configured with, e.g., a circuit to achieve the processing described below.

[0046] The wireless communication 1 supports Type II CSI feedback. As shown in FIG. 1, at step SI, the BS 20 transmits CSI-Reference Signals (RSs). When the UE 10 receives the CSI-RSs from the BS 20, the UE 10 performs measurements of the received CSI-RSs. Then, at step S2, the UE 10 performs CSI reporting to notify the BS 20 of the CSI as CSI feedback. For example, the CSI includes at least one of rank indicator (RI), precoding matrix indicator (PMI), channel quality information (CQI), CSI-RS resource indicator (CRI), a wideband (WB) amplitude, and a subband (SB) amplitude. In one or more embodiments of the present invention, the CSI reporting that reports the SB amplitude may be referred to as SB amplitude reporting. For example, rather than reporting the SB amplitude every time when the CSI reporting takes place, the periodicity of reporting the SB amplitude may be dynamically adjusted using higher layer signaling from the BS 20. The SB amplitude reporting may be performed for K leading coefficients. For example, if K is small, the number of coefficients reporting SB amplitudes is small.

[0047] If the SB amplitudes are small compared to an amplitude of the strongest coefficient, achievable gains with SB amplitude reporting may be marginal. That may occur when a user channel is highly sparse in an environment with very few scatterers, for example.

[0048] Furthermore, in one or more embodiments, while Type II CSI feedback may allow layer handling up to layers with RI of 1 and 2, by altering the scheme, Type II CSI feedback may also be implemented in ranks greater than 2. As such, by extending Type II CSI feedback scheme for rank > 2, spectral efficiency can be further enhanced. Extending the Type II CSI feedback scheme to ranks greater than 2 may reduce the overhead generally associated with the scheme.

[0049] To this point and as indicated above, Type II CSI precoding vector generation for N₃ precoding matrix indicator (PMI) sub-bands (SBs) considering RI = v, layer I E {1,2, ••• υ] transmission may be evaluated. For example,

[0050] (2)

[0051] In the above equation, is a matrix consisting of L SD 2D- DFT basis for layer I, L is a Beam number, N_tis a Number of ports, and W_{coeff z} (2L X N₃) is an SB complex combination coefficient matrix for layer I.

[0052] In the above equations, the SD 2D-DFT basis subset may be given as 2D DFT basis vector corresponding to

the /-th layer.

[0053] In one or more embodiments, frequency domain compression must be accounted for as information within W_{coeff l}, which may be compressed. As such, corresponding overhead may be further reduced. For example, Type II CSI precoding vectors of layer I for N₃ SBs considering FD compression can be given by expanding W_{coeff, i} from rule (2). |0054| (3)

[0055] In the above equation, is a matrix consisting of M FD

DFT basis vectors for layer I and

is a matrix consisting of complex combination coefficients for layer I. Furthermore, the frequency domain DFT basis subset may be given as where f_{l i} is the i-th (G {1, DFT basis

vector corresponding to the l-th layer. Additionally, M is calculated as,

where R E {1,2}. Given L and p, the number of SD and FD basis subsets for layer I can be identified.

[0056] In one or more embodiments, in order to achieve a proper balance between performance and overhead, it is important to identify SD and FD bases across layers appropriately.

[0057] A wireless communication system 200 according to one or more embodiments of the present invention will be described below with reference to FIG. 2.

[0058] As shown in FIG. 2, the wireless communication system 200 includes the BS 20 that communicates with the UE 10 via multiple antenna ports using the MIMO technology. A Type II port selection codebook does not require the UE 10 to derive spatial domain (SD) beams considering 2D-DFT basis as in regular Type II codebook. In this case, the BS 20 may be a gNB that transmits a number K of beamformed (BF) CSI-RS ports as a candidate set of SD beams. The UE 10 has to identify a number L (< K) of best CSI-RS ports (i.e., beams) and to report their corresponding indices within

[0059] SB-wise precoding vector generation with NR Release 16 Type II port selection codebook for layer I G {1, 2, 3, 4} by further modifying (2) and (3) can be given as:

[0060] (4)

[0061] In this case, Q(N_t x K) may represent a number K of SD beams used for

CSI-RS beamforming, W₁( KX2L) may represent a block diagonal matrix,

M) may represent a linear combination (LC) coefficient matrix, and may

be used for DFT basis vectors (i.e., FD bases).

[0062] A number of CSI-RS ports P_{CSI -RS} may include the number K of BF CSI- RS ports being configured by higher layer signaling. In this case, P_{CSI -RS} ^∈ {4, 8, 12, 16, 24, 32}. consists of column vectors of an identity matrix. As such, each such vector signifies a selected beam. To this point, a number of ports is selected, and these ports may include the number L being configured by higher layer signaling. In this case, L E {2, 3, 4} when P_{CSI -RS} > 4.

[0063] The selection of SD beams within Q is transparent to the UE 10. Specifically, the SD beams can be determined based on sounding reference signals (SRS) or Uplink (UL) Demodulation Reference Signal (DMRS) transmission. In this case, even though the UL dominant sub-space is not the same as that of the downlink (DL), port selection in the DL allows the UE 10 to approximately select ports covering the DL dominant sub-space.

[0064] FIG. 3A and FIG. 3B show reporting of W₁ based on selected beams. In some embodiments, a parameter d may be configured by the BS 20 determines a sampling granularity for both groups. In this case, a CSI report in NR may be made up of two granularities (e.g., categories). In some embodiments, d may be configured as d E {1, 2, 3, 4} and d < L. In such case, the UE 10 may report i_1;1 as part of PMI to select L beams following:

[0065] (5)

[0066] For example, as shown in FIG. 3A and FIG. 3B, (5) may be used to determine available port-pairs for selection out of the beams for polarization. In FIG. 3 A, letting K = 8 and L = 2, the calculation may result in a first case where d = 2 such that i_1;1 = {0, 1}. In this regard, the available port-pairs for selection may be equal to {B0, Bl} and {B2, B3}. In FIG. 3B, letting K= 8 and L = 2, the calculation may result in a second case where d = 1 such that i_1;1 = {0, 1, 2, 3}. In this regard, the available port-pairs for selection may be equal to {B0, Bl}, {Bl, B2}, {B2, B3}, and {B3, B0}. [0067] In FIG. 3A and FIG. 3B, W₁ is set as follow:

[0068]

[

¹0069] J In this case, , where ’ represents a vector with all zeros except 1 at I^th location.

[0070] In view of the aforementioned calculations, a Type II port selection codebook can be further enhanced by taking into consideration the port selection of both SD beams and frequency domain (FD) basis vectors in of (4). As such, in some embodiments, a general structure for SB-wise precoder generation for port selection codebook can be given by:

[0071] (6)

[0072] In this case, Q(N_t x K) may represent a number K of beamformed CSI- RS ports for SD beam selection such that b_i is i ∈ {1, 2, ... K} for an z-th SD basis vector, S(N_t X K') may represent a number K’ of beamformed CSI-RS ports for FD basis selection such that fj is j E {1, 2, ... K'} for an /-th FD basis vector, W^K x 2L) may represent a block diagonal matrix where each matrix block consists of L columns of an (K X K) identity matrix. W_{f l}(K' X M) may represent a matrix consisting of columns of an (K' X K') identity matrix, and W_l(2L X M) may represent a linear combination (LC) coefficient matrix. Further, in this case, Q and S may be, for example,

[0073] In (6), beamforming is done both in SD and FD, and FD bases for beamforming can be determined considering delay reciprocity. The BS 20 transmits (K x K’) beamformed CSI-RS ports. The selection of K SD beams and K’ FD bases is transparent to the UE. In this case, the UE selects a number 2L of SD beams and a number M of FD bases and report them back to the BS 20 as part of the PMI. Here, the W₁ and the may capture selected SD and FD bases. In addition, is reported along with the LC coefficients by the UE 10. [0074] Based on the above, as shown in FIG. 4, (6) may be implemented in a communication system 400 including the UE 10 and the BS 20. In communication system 400, (1) UL SRS are transmitted from the UE 10 to the BS 10. In response, the BS 20 may exchange (3) beamformed CSI-RS to the UE 10, which causes the UE 10 to report at least UL CSI including values related to W₁ The reporting

provided by the UE 10 may be transmitted in (4) UL CSI reporting. To this point, by using reported SD beams, FD bases, and LC coefficients, the BS 20 determines the DL precoders as captured in (4).

[0075] Consistent with the above, FIG. 5A and FIG. 5B show the UE 10 assisting in identifying DL dominant sub-space by selecting a number L of ports out of a number K of beamformed CSI-RS ports. In the figures, Q is considered in the context of (6) to include DFT basis vectors for DL CSI-RS beamforming as SD beams. As shown in FIG. 5A, the selected DFT basis vectors at least entirely cover a dominant sub-space of UL spatial covariance. For example, as shown in FIG. 5B, DFT beam identification for DL CSI-RS beamforming at the BS 20 are based on UL spatial covariance matrix having a dominant sub-space with 3 eigen vectors. In FIG. 5B, the selected DFT beams (i.e., b₁ b₂, b₃, b₄) can be considered in Q for CSI-RS beamforming.

[0076] FIG. 6A and FIG. 6B show the UE 10 assisting in identifying DFT beams in a case in which Q consists of DFT beams. A difference between FIG. 5 A and FIG. 6A is that, in FIG. 5A, there are 3 eigen vectors in the considered dominant sub-space. However, in FIG. 6A, only the dominant eigen vector of UL spatial covariance is considered. As such, FIG. 6A may be a special case of FIG. 5A. In the figures, only the dominant eigen vector of UL spatial covariance may be considered. As discussed above, the identified DFT vectors in Q entirely cover the dominant UL eigen direction. As shown in FIG. 6A, DFT beam identification for DL CSI-RS beamforming considering dominant eigen vector of UL spatial covariance may be limited to one dominant eigen vector. For example, as shown in FIG. 6B, DFT beam identification for DL CSI-RS beamforming at the BS 20 are based on UL spatial covariance matrix having a dominant sub-space with 1 eigen vector. In FIG. 6B, the selected DFT beams (i.e., b₁ b₂) can be considered in Q for CSI-RS beamforming.

[0077] FIG. 7A and FIG. 7B illustrate bitmaps in which SD-beam selection/reporting based on beamformed CSI-RS is being implemented. The dominant sub-space of UL spatial covariance matrix is different than that of DL when considering phased arrays (with fixed inter-element spacing) for transmission. As such, it is better to let the UE 10 pick a number of beams, L, out of K(≥ L), the number of beamformed CSI-RS ports. At this point, the BS 20 can either configure the number L, or let the UE 10 to freely choose the number of SD beams based on channel observation. In a case in which the BS 20 configures the number L, the BS 20 takes into consideration a size of the dominant sub-space of UL covariance matrix for determining/configuring L.

[0078] Further, the UE 10 can freely select any number of SD beams out of the number K of beamformed CSI-RS ports. In this regard, the UE 10 freely selects a set of SD beams and reports the selected SD beams using a bitmap such as that one shown in FIG. 7A. In the figure, letting K = 10, the UE 10 may use a 10-bit length bitmap to report selected SD beams. In this example, the UE 10 freely selects b₁, b₂, b₆, b₇ SD beams. Under this case, in some embodiments, the K -length bitmap is then encoded using a Huffman coding scheme. The codewords associated with the Huffman encoding may be previously determined or predefined. As shown in FIG. 7B, the UE 10 may report the number L of selected SD beams and a starting position of the selected SD beams. In the case of FIG. 7A, letting K = 10, the UE 10 reports L = 4 and starting position as b₃ to report selected SD beams. At this stage, the UE 10 may use combinatorial signaling to reports the selected SD beams as

[0079] FIG. 8 shows a diagram illustrating SRS enhancements for UL covariance estimation. As discussed above, a UL spatial covariance matrix may be required for determining DFT basis vectors in Q. Then, to determine the UL covariance, the UL SRS transmission may be considered. In particular, for estimating and tracking the UL spatial covariance matrix, periodic/semi-persistent SRS transmission along with aperiodic-SRS triggering only when necessary can be considered. As shown in FIG. 8, periodic/SP-SRS may be considered for estimating and tracking UL spatial covariance. Further, in some embodiments, since the channel is wide-sense stationary, SRS transmission only across part of the BW is sufficient. In this regard, when NW detects a change in UL spatial covariance compared to previous estimation, NW can trigger an A-SRS to quickly re-estimate the UL spatial covariance. As shown in FIG. 8, the UE 10 may not report the number L of SD ports when there is no A-SRS triggering. In this case, the UE 10 may only report LC coefficients in

and FD based in W_{f z} as PMI. For calculating LC coefficients and FD bases, the UE 10 may use the previously reported number L SD ports. Similarly, when there is an A-SRS trigger by the NW, as shown in FIG. 8, during a subsequent CSI reporting, the UE 10 may include a newly selected number L SD ports in addition to the LC coefficients and the FD bases.

[0080] Consistent with the above, FIG. 9A and FIG. 9B show the UE 10 assisting in identifying SD-FD beam pairs for CSI-RS Beamforming in DL. In the figures, Q and S are considered in the context of (6) to include SD and FD beamforming vectors, respectively. As shown in FIG. 9A, the selected DFT basis vectors at least entirely cover a dominant sub-space of UL spatial covariance. In this regard, each eigen vector in the UL dominant sub-space captures a set of multi-path components (MPCs). For example, as shown in FIGs. 9A and 9B, the three eigen vectors, u₂ and u₃ of the dominant sub-space of UL covariance capture MPCs with delays, ,

respectively. In FIG. 9A, the selected SD beam in Q captures MPCs with delays

In this regard, for SD-FD based CSI-RS beamforming, each SD beam needs to be paired with a FD basis vector (corresponding to a particular delay).

[0081] FIG. 9A shows that SD beams in Q may capture MPCs with different delays. If an SD beam captures MPCs with multiple delays, weighted interpolation of those delays may be considered to determine a delay value (and corresponding FD basis for that delay). For example, as shown in FIG. 9B, letting an SD beam b_i capturing MPCs with delays a resultant delay value is calculated considering weighted

interpolation of . In this regard, such that w₁ and w₂ are

'etermined by considering an overlapping ratio of b_i with u₁ and u₂. Alternatively, w₁ and w₂ may be pre-defined or previously determined by the UE 10 and the BS 20. In this example, if w₁ = w₂ = 0.5, then = . Other possibilities for w₁ and

w₂ determination may not precluded. The solution shown in FIG. 9B is especially useful when UE is not allowed to select and report FD bases.

[0082] In some embodiments, if an SD beam captures MPCs with multiple delays, the SD beam may be separately paired with all FD bases corresponding to those delays and let the UE 10 select the most appropriate FD bases. For example, assume an SD beam b_i captures MPCs with delays . Then, it can be paired with f

₁ and f₂ FD bases which correspond to delays , respectively. The UE 10 may select

an FD basis out of f₁ and f₂ and report that to the BS 20. In this case, the UE 10 is allowed to select FD bases for a given SD beam. The NW can determine multiple FD bases for a given SD beam considering other criteria as well and achieve CSI-RS beamforming by pairing said SD beam with those FD bases. As discussed above, it is still necessary to ensure that the identified SD beams (in SD-FD pairs) for CSI-RS beamforming cover entirely the UL dominant sub-space. As such, in this case, FD bases can be paired with each of those identified SD beams.

[0083] FIG. 10 shows a bitmap illustrating SD-FD beam selection/reporting based on beamformed CSI-RS. In FIG. 10, the UE 10 may freely select any number of SD beams out of the number K of beamformed CSI-RS ports and any number of FD beams out of the number K' of beamformed CSI-RS ports. The UE 10 freely selects a set of SD beams and FD beams and the UE 10 reports the selected SD beams using a bitmap and associated FD bases considering some other approach. For example, letting K = 10 and K ^' = 5, then the UE 10 may use a 10-bit length bitmap to report the selected SD beams and the associated FD bases directly. In FIG. 10, the SD-FD pairing for this reporting is such that {b₁,f₁}, {b₂,f₂}, {b₆,f₄},{b-₇,f_s}. Further, the UE 10 reports more than one FD basis per SD beam. The same number of FD bases may be reported for all SD beams. For example, SD-FD pairing when 2 FD bases per SD beam are reported may be given by {b₄,f₄}, {b₁,f₂}, {b₂,f₂} , {b₂,f₁}, {b₆,f₄} , {b₆,f₅} , . For reporting the bitmap, which identifies the selected SD beams, all

options discussed above may be used. For reporting the bitmap, which identifies the selected SD beams, all options discussed with respect to FIGs. 7 A and 7B can be used.

[0084] In FIG. 10, the UE 10 may freely select a set of SD beams and FD beams and report selected SD beams using a bitmap as discussed above and the associated FD bases using combinatorial signaling. For example, letting K = 10 and K' = 5, the UE

10 may use a 10-bit length bitmap to report selected SD beams. By looking at the number of SD beams, the NW may process the number of FD beams. In FIG. 10, the UE 10 selects b₁, b₂, b₆, b₇. In some embodiments, each SD beam is paired with a single FD basis such that the BS 20 knows that the number of FD bases is equal to the number of 1’s in the bitmap. For example, let’s say the number of 1’s is M. The associated permutation for FD bases can then may be reported using In

some embodiments, each SD beam is paired with multiple FD bases where the number of FD bases is equal to the number of 1’s in the bitmap. In view of the above, FD bases are beam common. In this regard, assuming that the number of 1’s is

, then the associated FD bases may be reported using combinatorial signaling as using

’ bits, where [x] denotes the smallest integer at least as large as x. Similarly, each

SD beam is paired with multiple FD bases such that these FD bases are beam common.

The number of FD bases M may be reported by the UE 10 or derived using some other way. Then, the FD bases can be reported using combinatorial signaling as,

^usin

[0085] The UE 10 may freely select a set of SD beams and FD bases and the UE 10 reports the selected FD bases using a bitmap including the associated SD bases directly considering some other approach as described above. In this case, all solutions discussed above may be used for FD bases bitmap reporting and for the associated SD beams reporting. [0086] FIG. 11 shows a 2D bitmap illustrating the UE 10 selection of SD-FD pairs. In FIG. 11, the UE 10 has selected b₁, b₂, b₆, b₇ such that each selected SD beam is associated with SD-beam specific FD bases. The UE 10 freely selects a set of SD beams and FD bases and the UE 10 reports the selected SD beams and the associated FD bases using a 2D bitmap. In this regard, letting K = 8 and K' = 5, then the UE 10 may use a 40-bit length 2D-bitmap to report the selected SD beams and the associated FD bases.

[0087] In FIG. 11, the UE 10 freely selects the set of SD beams and FD beams and the UE 10 reports the selected SD beams and the associated FD beams using a compressed representation of the 2D bitmap. In this regard, the UE 10 may capture selected SD and FD bases as discussed above within a KK' -bits size 2D bitmap. The /f/C-bits size bitmap may then be encoded using a Huffman coding scheme. The codewords associated with the Huffman encoding may be previously determined or predefined, or shared between the BS 20 and the UE 10 using higher layers signaling or DCI. Other ways to compress the bitmap are not precluded.

[0088] FIG. 12 shows a diagram illustrating SRS enhancements for UL covariance estimation. As discussed above, a UL spatial covariance matrix may be required for determining the DFT basis vectors in Q . Then, to determine angle(s)/delay(s) associated with a propagation channel, SRS transmission across the entire CSI-RS bandwidth (BW) may be required. In particular, for estimating and tracking angle(s)/delay(s), periodic/semi-persistent SRS transmission along with aperiodic-SRS triggering only when necessary can be considered. As shown in FIG. 12, periodic/SP-SRS is considered for estimating and tracking the UL spatial covariance (corresponding to tracking the angle(s)). Further, for determining the delay(s) associated with the propagation channel, SRS transmission at least across CSI-RS BW may be required. In this case, when the NW detects a change in angle(s)/delay(s) compared to the aforementioned embodiments, the NW may trigger an A-SRS to quickly re-estimate the angle(s)/delay(s). [0089] In FIG. 12, the UE 10 may not report a number L of SD ports and a number M of FD ports when there is no A-SRS triggering. In this regard, the UE 10 may only report LC coefficients in as PMI. For calculating LC coefficients, the UE

10 may use the previously reported numbers L and M, signifying the number of SD ports and the number of FD ports, respectively. In this regard, when there is an A-SRS trigger by the NW, during the subsequent CSI reporting, the UE 10 may include a newly selected number L of SD ports and the number M of FD ports in addition to LC coefficients.

[0090] The BS 20 according to one or more embodiments of the present invention will be described below with reference to the FIG. 13.

[0091] As shown in FIG. 13, the BS 20 may include an antenna 201 for 3D MIMO, an amplifier 202, a transmitter/receiver circuit 203 (hereinafter referred as including a CSI-RS scheduler), a baseband signal processor 204 (hereinafter referred as including a CS-RS generator), a call processor 205, and a transmission path interface 206. The transmitter/receiver 202 includes a transmitter and a receiver.

[0092] The antenna 201 may comprise a multi-dimensional antenna that includes multiple antenna elements such as a 2D antenna (planar antenna) or a 3D antenna such as antennas arranged in a cylindrical shape or antennas arranged in a cube. The antenna 201 includes antenna ports having one or more antenna elements. The beam transmitted from each of the antenna ports is controlled to perform 3D MIMO communication with the UE 10.

[0093] The antenna 201 allows the number of antenna elements to be easily increased compared with a linear array antenna. MIMO transmission using a large number of antenna elements is expected to further improve system performance. For example, with 3D beamforming, high beamforming gains are also expected from increasing the number of antennas. Furthermore, MIMO transmission is also advantageous in terms of interference reduction, for example, by null point control of beams, and effects such as interference rejection among users in multi-user MIMO can be expected. [0094] The amplifier 202 generates input signals to the antenna 201 and performs reception processing of output signals from the antenna 201.

[0095] The transmitter included in the transmitter/receiver circuit 203 transmits data signals (e.g., reference signals and precoded data signals) via the antenna 201 to the UE 10. The transmitter transmits CSI-RS resource information that indicates a state of the determined CSI-RS resources (e.g., subframe configuration ID and mapping information) to the UE 20 via higher layer signaling or lower layer signaling. The transmitter transmits the CSI-RS allocated to the determined CSI-RS resources to the UE 10.

[0096] The receiver included in the transmitter/receiver circuit 203 receives data signals (i.e., reference signals and the CSI feedback information) via the antenna 201 from the UE 10.

[0097] The CSI-RS scheduler 203 determines CSI-RS resources allocated to the CSI-RS. For example, the CSI-RS scheduler 203 determines a CSI-RS subframe that includes the CSI-RS in subframes. The CSI-RS scheduler 203 determines at least an RE that is mapped to the CSI-RS.

[0098] The CSI-RS generator 204 generates CSI-RS for estimating the downlink channel states. The CSI-RS generator 204 may generate reference signals defined by the LTE standard, dedicated reference signal (DRS) and Cell-specific Reference Signal (CRS), synchronized signals such as Primary synchronization signal (PSS) and Secondary synchronization signal (SSS), and newly defined signals in addition to CSI- RS.

[0099] The call processor 205 determines a precoder applied to the downlink data signals and the downlink reference signals. The precoder is called a precoding vector or more generally a precoding matrix. The call processor 205 determines the precoding vector (precoding matrix) of the downlink based on the CSI indicating the estimated downlink channel states and the decoded CSI feedback information inputted. [00100] The transmission path interface 206 multiplexes CSI-RS on REs based on the determined CSI-RS resources by the CSI-RS scheduler 203.

[00101] The transmitted reference signals may be Cell-specific or UE-specific. For example, the reference signals may be multiplexed on the signal such as PDSCH, and the reference signal may be precoded. Here, by notifying a transmission rank of reference signals to the UE 10, estimation for the channel states may be realized at the suitable rank according to the channel states.

[00102] The BS 20 further, in one or more embodiments, comprising hardware configured for reducing the feedback overheads associated with bitmap reporting between a user equipment and a base station. For example, the BS 20 may include the capabilities described above for reducing feedback overhead when communicating with the UE 10.

[00103] The UE 10 according to one or more embodiments of the present invention will be described below with reference to the FIG. 14.

[00104] As shown in FIG. 14, the UE 10 may comprise a UE antenna 101 used for communicating with the BS 20, an amplifier 102, a transmitter/receiver circuit 103, a controller 104, the controller including a CSI feedback controller and a codeword generator, and a CSI-RS controller. The transmitter/receiver circuit 103 includes a transmitter and a receiver 1031.

[00105] The transmitter included in the transmitter/receiver circuit 103 transmits data signals (for example, reference signals and the CSI feedback information) via the UE antenna 101 to the BS 20.

[00106] The receiver included in the transmitter/receiver circuit 103 receives data signals (for example, reference signals such as CSI-RS) via the UE antenna 101 from the BS 20.

[00107] The amplifier 102 separates a PDCCH signal from a signal received from the BS 20. [00108] The controller 104 estimates downlink channel states based on the CSI- RS transmitted from the BS 20, and then outputs a CSI feedback controller.

[00109] The CSI feedback controller generates the CSI feedback information based on the estimated downlink channel states using the reference signals for estimating downlink channel states. The CSI feedback controller outputs the generated CSI feedback information to the transmitter, and then the transmitter transmits the CSI feedback information to the BS 20. The CSI feedback information may include at least one of Rank Indicator (RI), PMI, CQI, BI and the like.

[00110] The CSI-RS controller determines whether the specific user equipment is the user equipment itself based on the CSI-RS resource information when CSI-RS is transmitted from the BS 20. When the CSI-RS controller 16 determines that the specific user equipment is the user equipment itself, the transmitter that CSI feedback based on the CSI-RS to the BS 20.

[00111] The UE 10 further, in one or more embodiments, comprising hardware configured for reducing feedback overhead associated with bitmap reporting between a user equipment and a base station. For example, the UE 10 may include the capabilities described above for reducing feedback overhead when communicating with the BS 20.

[00112] The above examples and modified examples may be combined with each other, and various features of these examples can be combined with each other in various combinations. The invention is not limited to the specific combinations disclosed herein.

[00113] Although the disclosure has been described with respect to only a limited number of embodiments, those skilled in the art, having benefit of this disclosure, will appreciate that various other embodiments may be devised without departing from the scope of the present invention. Accordingly, the scope of the invention should be limited only by the attached claims.

Claims

CLAIMS What is claimed is:

1. A terminal comprising : a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain (SD) beam selection; and a processor that: considers a plurality of Discrete Fourier Transform (DFT) basis vectors for downlink (DL) CSI-RS beamforming as SD beams, and selects the SD beams based on the plurality of DFT basis vectors, the SD beams considering uplink (UL) spatial covariance.

2. The terminal according to claim 1, wherein the selected SD beams comprise coverage of dominant sub-space of the UL spatial covariance.

3. The terminal according to claim 1, further comprising: a transmitter that transmits a message comprising a set of the SD beams based on the one or more beamformed CSI-RSs.

4. The terminal according to claim 1, wherein the processor further: considers periodic/semi-persistent-SRS transmission along with Aperiodic-SRS for uplink (UL) covariance tracking and estimation.

5. A terminal comprising: a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RSs), the beamforming information corresponding to Spatial Domain-Frequency Domain (SD-FD) pair selection; and a processor that: considers a plurality of eigen vectors in an uplink (UL) dominant sub-space, the plurality of eigen vectors capturing a set of multi-path components (MPCs), and selects SD-FD pairs based on the plurality of eigen vectors. The terminal according to claim 5, wherein the processor further: when one SD beam out of the SD-FD pairs captures MPCs with multiple delays, considers a weighted interpolation of the multiple delays to determine a delay and an associated FD basis for CSI-RS beamforming. The terminal according to claim 5, wherein the processor further: when one SD beam out of the SD-FD pairs captures MPCs with multiple delays, separately pairs the one SD beam with all FD bases corresponding to the multiple delays. The terminal according to claim 5, wherein the processor further: considers periodic/semi-persistent-SRS transmission along with Aperiodic-SRS for UL covariance tracking and estimation. A terminal comprising: a receiver that receives beamforming information relating to one or more beamformed Channel State Information - Reference Signals (CSI-RS s), the beamforming information corresponding to Spatial Domain (SD) beam selection and Frequency Domain (FD) basis selection; and a processor that: considers a bitmap, the bitmap associating SD beams and FD bases to one another, and freely selects SD beams and FD bases based on a rule using on the bitmap. . The terminal according to claim 9, wherein the bitmap is based on reporting of SD beams/FD bases.

. The terminal according to claim 9, wherein the bitmap is a 2D-bitmap based on joint reporting of SD beams/FD bases. . A method for performing Sounding Reference Signal (SRS) assisted Channel State Information - Reference Signal (CSI-RS) beamforming in angle-delay domains, the method comprising: obtaining beamforming information relating to one or more beamformed CSI-RSs, the beamforming information corresponding to Spatial Domain (SD) beam selection; considering a plurality of Discrete Fourier Transform (DFT) basis vectors for downlink DL CSI-RS beamforming as SD beams; and selecting the SD beams based on the plurality of DFT basis vectors, the SD beams considering UL spatial covariance. . The method according to claim 12, wherein the selected SD beams comprise coverage of dominant sub-space of the uplink (UL) spatial covariance. . The method according to claim 12, further comprising: reporting a message comprising a set of the SD beams based on beamformed CSI- RS. . The method according to claim 12, further comprising: considering periodic/semi-persistent-SRS transmission along with Aperiodic-SRS for UL covariance tracking and estimation.