CN112889224A

CN112889224A - Beam-based preprocessing in MU-MIMO systems

Info

Publication number: CN112889224A
Application number: CN201880098919.5A
Authority: CN
Inventors: 刘皓; 赵岩; 孙欢
Original assignee: Nokia Shanghai Bell Co Ltd; Nokia Networks Oy
Current assignee: Nokia Shanghai Bell Co Ltd; Nokia Oyj
Priority date: 2018-10-23
Filing date: 2018-10-23
Publication date: 2021-06-01
Anticipated expiration: 2038-10-23
Also published as: CN112889224B; WO2020082244A1

Abstract

Embodiments of the present disclosure provide an apparatus and method for beam-based preprocessing in a transmitting device in a multi-user multiple-input multiple-output (MU-MIMO) system. The transmitting device also includes a transceiver unit (TRX) array and a digital precoding module. The apparatus is configured to map a plurality of TRXs of a TRX array to a plurality of transmit ports of a digital precoding module according to a precoding matrix, wherein a number of transmit ports is less than a number of the plurality of TRXs. Determining the precoding matrix by: the method includes determining a channel covariance matrix for a plurality of receiving devices based on channel information between the plurality of receiving devices and a transmitting device, performing an eigen decomposition operation on the channel covariance matrix to derive an eigenvector matrix, and selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix.

Description

Beam-based preprocessing in MU-MIMO systems

Technical Field

The non-limiting and exemplary embodiments of the present disclosure relate generally to precoding solutions for multi-user multiple-input multiple-output (MU-MIMO) systems, and in particular, to an apparatus and method for beam-based pre-processing precoding in MU-MIMO systems.

Background

Massive MIMO technology has been identified as one of the key technologies of the 5G New Radio (NR) system. Massive MIMO technology meets the ever-increasing performance requirements of 5G NR, but the large number of Antenna Elements (AE) or transceiver units (TRX) deployed in massive MIMO systems (e.g., MU-MIMO systems) inevitably introduce significant baseband processing complexity, hardware implementation costs, channel measurement and feedback signaling overhead, etc. How to effectively solve these problems is the key to successfully guarantee massive MIMO applications.

Hybrid precoding is a low-cost high-capacity solution for massive MIMO systems. It may include two stages of precoding, such as analog beamforming 130 and Digital Precoding (DP)110 shown in fig. 1. In the first stage, analog beamforming 130 may be implemented by a Phase Shift Network (PSN) in the Radio Frequency (RF) domain. For example, the PSN may include a plurality of splitters 131. With PSN, multiple AEs (e.g., N AEs) may be mapped to multiple transceiver units (TRXs) 120 (e.g., S TRXs), where the limited number of TRXs is much smaller than the number of AEs, i.e., S < N. In the second stage, DP 110 may be implemented in the baseband domain to mitigate MU interference and achieve greater MU multiplexing gain for the MU-MIMO system based on the effective channels weighted by analog beamforming 130. Since the effective channel has only S TRXs, which is much smaller than the number N of AEs, the baseband processing complexity for digital precoding can be greatly reduced. Taking the case of transmit precoding as an example, the DP 110 may map multiple data streams (e.g., M data streams) into multiple TRXs 120 in a TRX array through digital precoding. The data streams in the TRX array may then be weighted by the analog beamforming module 130 and transmitted via a plurality of Power Amplifiers (PAs) 140 and a plurality of AEs, respectively, as shown in fig. 1.

However, in practical product implementations for existing hybrid precoding schemes as shown in fig. 1, digital precoding still faces the challenge of baseband processing complexity due to the large number of TRXs (e.g., S-64) for massive MIMO as shown in fig. 1. As shown in fig. 2, the challenges may be manifested in several aspects, for example, a gsdeb (gnb)201 may implement a feature-based beamforming (EBB) operation for each co-scheduled User Equipment (UE) to determine transmit precoding for single-user MIMO (SU-MIMO) or MU-MIMO. EBB calculations may include Singular Value Decomposition (SVD), feature decomposition (ED), or zero-forcing (ZF) precoding, among others. Assuming that channel information for each UE is acquired in the gNB through an Uplink (UL) Sounding Reference Signal (SRS), EBB calculations may be performed on a Wideband (WB), Subband (SB), or Physical Resource Block (PRB) level. EBB processing complexity depends on the dimension of the channel matrix (e.g., the number of TRXs, e.g., S TRXs) and the frequency granularity of the matrix computation (e.g., per PRB computation). That is, the gNB may implement EBB calculations based on S TRXs (e.g., S-64) for each UE in each PRB. In addition, as shown in fig. 2, the UE 202 may measure Channel State Information (CSI) based on a CSI Reference Signal (RS) having, for example, S ports (which correspond to S TRX, e.g., S-64). Accordingly, the UE may perform EBB calculations based on S-TRX (e.g., S-64) channel information in the baseband in each PRB to determine Channel Quality Indication (CQI) feedback.

In summary, the large number of TRX configurations significantly increases CSI-RS signaling overhead and baseband processing complexity on the gbb and UE sides. How to solve this problem also becomes an increasing concern for massive MIMO applications in 5G NR.

Disclosure of Invention

The present disclosure will solve the above-mentioned problems by proposing an apparatus for beam-based pre-processing (BPP) in a transmitting device in a MU-MIMO system in order to reduce CSI-RS signaling overhead and baseband processing complexity in both transmitting and receiving sides. Other features and advantages of embodiments of the present disclosure will be understood from the following description of specific embodiments, which illustrate, by way of example, the principles of the embodiments of the disclosure.

In a first aspect of the present disclosure, an apparatus for beam-based preprocessing (BPP) in a transmitting device in a MU-MIMO system is provided. The transmitting device further includes a TRX array and a digital precoding module. The apparatus is configured to map a plurality of TRXs of the TRX array to a plurality of transmit ports of the digital precoding module according to a precoding matrix, wherein a number of the transmit ports is less than a number of the plurality of TRXs. Determining the precoding matrix by: determining a channel covariance matrix for a plurality of receiving devices based on channel information between the plurality of receiving devices and the transmitting device; performing an eigen decomposition operation on the channel covariance matrix to derive an eigenvector matrix; and selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix.

In one embodiment, selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix may include: and selecting an eigenvector corresponding to the largest multiple eigenvalues from the eigenvector matrix as a column vector of the precoding matrix.

In one embodiment, each transmit port may be connected to all TRXs of the plurality of TRXs. Determining the channel covariance matrix for the plurality of receiving devices may include: calculating a channel covariance matrix of the receiving device in the plurality of active subcarriers; and summing the channel covariance matrices of the plurality of receiving devices to obtain a channel covariance matrix for the plurality of receiving devices, and wherein the number of selected eigenvectors is equal to the number of transmit ports of the digital precoding module.

In one embodiment, the plurality of TRXs and the plurality of transmission ports are divided into a plurality of groups. Each of the plurality of groups includes two or more TRXs and two or more transmission ports. Each transmit port in each group may be connected to all TRXs in the same group. Determining a channel covariance matrix for a plurality of receiving devices may include: calculating channel covariance matrices of the receiving devices in the plurality of active subcarriers for the TRXs in each group; and summing the channel covariance matrices of the plurality of receiving devices for the TRXs in each group to obtain a channel covariance matrix for each group. Performing the feature decomposition operation may include: an eigen decomposition operation is performed on the channel covariance matrix for each group to derive an eigenvector matrix for each group. Selecting a plurality of eigenvectors from an eigenvector matrix as column vectors of the precoding matrix may include: an eigenvector is selected from the eigenvector matrix for each group as a column vector of subblocks, wherein a plurality of subblocks are arranged along a diagonal of the precoding matrix, and the number of eigenvectors selected for each group is equal to the number of transmission ports of the digital precoding modules in the same group.

In one embodiment, the plurality of TRXs and the plurality of transmission ports are divided into a plurality of groups. Each of the plurality of groups includes two or more TRXs and two or more transmission ports. Each transmit port in each group may be connected to all TRXs in the same group. Determining a channel covariance matrix for a plurality of receiving devices may include: calculating channel covariance matrices for receiving devices in the plurality of active subcarriers for the TRXs in each group; calculating an average of channel covariance matrices over a plurality of groups; and summing the average channel covariance matrices of the plurality of receiving devices to obtain a channel covariance matrix for the plurality of receiving devices. Selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix may include: the eigenvectors are selected as column vectors of all sub-blocks along the diagonal of the precoding matrix, where the number of eigenvectors selected is equal to the number of transmit ports of the digital precoding module in each group.

In one embodiment, the groups may be grouped based on polarization direction.

In one embodiment, the channel information for the TRX may be obtained based on an uplink sounding reference signal.

In an embodiment, the precoding matrix may be adjusted periodically.

In one embodiment, the adjustment period of the precoding matrix may be equal to or greater than the adjustment period of the digital precoding module.

In a second aspect of the present disclosure, a transmitting device in a MU-MIMO system is provided. The transmitting device includes a digital precoding module and a transceiver unit (TRX) array including a plurality of TRXs for transmitting an output data signal of the digital precoding module. The transmitting device further comprises an apparatus for beam-based preprocessing according to the first aspect of the present disclosure.

In a third aspect of the present disclosure, a method for configuring an apparatus for beam-based preprocessing in a transmitting device in a MU-MIMO system is provided. The transmitting device further includes a TRX array and a digital precoding module. The method includes mapping a plurality of TRXs of the TRX array to a plurality of transmit ports of the digital precoding module according to a precoding matrix, wherein a number of transmit ports is less than a number of the plurality of TRXs. The precoding matrix is determined by: determining a channel covariance matrix for a plurality of receiving devices based on channel information between the plurality of receiving devices and a transmitting device; performing eigen decomposition operation on the channel covariance matrix to derive an eigenvector matrix; and selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix.

In one embodiment, the step of selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix may comprise: and selecting an eigenvector corresponding to the largest multiple eigenvalues from the eigenvector matrix as a column vector of the precoding matrix.

In one embodiment, each transmit port may be connected to all TRXs of the plurality of TRXs. The step of determining a channel covariance matrix for a plurality of receiving devices may comprise: calculating a channel covariance matrix of a receiving device in the plurality of active subcarriers; and summing the channel covariance matrices of the plurality of receiving devices to obtain a channel covariance matrix for the plurality of receiving devices, and wherein the number of selected eigenvectors is equal to the number of transmit ports of the digital precoding module.

In an embodiment, the plurality of TRXs and the plurality of transmission ports are divided into a plurality of groups. Each of the plurality of groups includes two or more TRXs and two or more transmission ports. Each transmit port in each group may be connected to all TRXs in the same group. The step of determining the channel covariance matrices of the plurality of receiving devices may comprise: calculating channel covariance matrices of the receiving devices in the plurality of active subcarriers for the TRXs in each group; and summing the channel covariance matrices of the plurality of receiving devices for the TRXs in each group to obtain a channel covariance matrix for each group. The step of performing a feature decomposition operation may comprise: an eigen decomposition operation is performed on the channel covariance matrix for each group to derive an eigenvector matrix for each group. The step of selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of a precoding matrix may comprise: an eigenvector is selected from the eigenvector matrix for each group as a column vector of subblocks, wherein a plurality of subblocks are arranged along a diagonal of the precoding matrix, and the number of eigenvectors selected for each group is equal to the number of transmission ports of the digital precoding modules in the same group.

In one embodiment, the plurality of TRXs and the plurality of transmission ports are divided into a plurality of groups. Each of the plurality of groups includes two or more TRXs and two or more transmission ports. Each transmit port in each group may be connected to all TRXs in the same group. The step of determining a channel covariance matrix for a plurality of receiving devices may comprise: calculating channel covariance matrices of the receiving devices in the plurality of active subcarriers for the TRXs in each group; calculating an average of channel covariance matrices over a plurality of groups; and summing the average channel covariance matrices of the plurality of receiving devices to obtain the channel covariance matrices of the plurality of receiving devices. The step of selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix may comprise: selecting the eigenvectors as column vectors of all sub-blocks along a diagonal of the precoding matrix, wherein the number of eigenvectors selected is equal to the number of transmit ports of the digital precoding module in each group.

In one embodiment, the plurality of groups may be grouped based on polarization direction.

In one embodiment, the precoding matrix may be adjusted periodically.

In a fourth aspect of the disclosure, a computer-readable storage medium having stored thereon computer-executable instructions is provided. The computer executable instructions, when executed on the at least one processor, cause the at least one processor to perform a method according to the third aspect of the present disclosure.

According to the various aspects and embodiments as described above, the problem of BPP precoding in massive MIMO systems can be solved.

Drawings

The above and other aspects, features and benefits of various embodiments of the present disclosure will become more apparent from the following detailed description, by way of example, with reference to the accompanying drawings in which like reference numerals or letters are used to designate similar or equivalent elements. The accompanying drawings, which are included to provide a further understanding of the embodiments of the disclosure, are not necessarily drawn to scale and:

FIG. 1 shows a block diagram of an existing hybrid precoding scheme 100 for MU-MIMO systems;

fig. 2 schematically illustrates the challenges of baseband processing for the existing hybrid precoding scheme as illustrated in fig. 1;

FIG. 3 shows a block diagram of a transmitting device including a BPP module according to an embodiment of the invention;

FIGS. 4A and 4B illustrate BPP precoding in full-connected mode and hierarchical full-connected mode, respectively;

fig. 5 shows a block diagram of a hybrid precoding scheme 500 including a BPP module according to an embodiment of the disclosure;

FIG. 6 shows a flow diagram of a method 600 for configuring a BPP module according to an embodiment of the present disclosure;

FIG. 7 shows a flowchart of a process 700 for determining a precoding matrix for a BPP module, according to an embodiment of the present disclosure;

FIG. 8 shows a flowchart of a process 800 for determining a precoding matrix for a BPP module in accordance with an embodiment of the present disclosure;

FIG. 9 shows a flowchart of a process 900 for determining a precoding matrix for a BPP module in accordance with an embodiment of the present disclosure;

FIG. 10 illustrates the benefits of BPP precoding; and

fig. 11 shows a simplified block diagram of an apparatus according to an embodiment of the present disclosure.

Detailed Description

Hereinafter, the principle and spirit of the present disclosure will be described with reference to illustrative embodiments. It is to be understood that all of these examples are given solely for the purpose of enabling those skilled in the art to better understand and further practice the present invention, and are not intended to limit the scope of the present invention. For instance, features illustrated or described as part of one embodiment, can be used with another embodiment to yield a still further embodiment. In the interest of clarity, not all features of an actual implementation are described in this specification.

References in the specification to "one embodiment," "an example embodiment," etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may not necessarily include the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

In the following description and claims, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. For example, the term "transmitting device" as used herein may refer to any device having wireless communication capabilities, including but not limited to a base station device, a terminal equipment, or a User Equipment (UE). The term "receiving device" as used herein may refer to any device having wireless communication capabilities, including but not limited to a base station device, a terminal equipment, or a UE. The terminal device or UE may be a mobile phone, a cellular phone, a smart phone or a Personal Digital Assistant (PDA), a portable computer, etc. Furthermore, non-mobile user equipment may also readily employ embodiments of the present invention. In the following description, the terms "user equipment", "UE" and "terminal equipment" may be used interchangeably. Similarly, the term "base station apparatus" may denote a Base Station (BS), a node B (NodeB or NB), an evolved NodeB (eNodeB or eNB), a gnodeb (gnb), a Relay Node (RN), and the like.

For purposes of illustration, several embodiments of the present disclosure will be described in the context of an NR MU MIMO system. However, those skilled in the art will appreciate that the concepts and principles of several embodiments of the present disclosure may be more generally applicable to other wireless networks, such as third generation long term evolution (3G-LTE) networks, fifth generation (4G) networks, 4.5G LTE, or future networks (e.g., 5G networks).

Non-limiting and exemplary embodiments of the present disclosure relate to a hybrid precoding scheme for MU-MIMO systems. As described above, the conventional hybrid precoding scheme still needs to implement a large number of EBB calculations due to a large number of TRXs. No solution to this problem is proposed. The non-limiting and exemplary embodiments of the present disclosure propose an apparatus for beam-based preprocessing in a transmitting device in a MU-MIMO system. The main concept of the invention is shown in fig. 3. A baseband-domain BPP precoding scheme is proposed in which a dynamic precoding technique may be performed to reduce the number of actual transmission ports and reduce the signaling overhead and baseband processing complexity for digital precoding. As shown in fig. 3, the BPP module 340 may be embedded in a hybrid precoding structure, particularly in the transmitting device 300. The transmitting device 300 also includes a TRX array and DP module 330. The BPP module 340 may be located between the TRX array and the DP module 330. The BPP module 340 may be configured to map a plurality of TRXs 310 of the TRX array to a plurality of Transmission Ports (TPs) 320 of the DP module 330 according to a precoding matrix, where the number of TPs is less than the number of the plurality of TRXs. Assume that there are S TRXs in the TRX array and T TPs in the DP block 330. With BPP precoding, S TRXs can be mapped to T TPs, where T < S. Thus, for digital precoding, the complexity of the baseband processing may be reduced as the number of TPs T decreases.

To ensure an optimal configuration, the precoding matrix may be determined based on a eigen decomposition operation. Fig. 6 shows a flow diagram of a method 600 for configuring a BPP module according to an embodiment of the disclosure. As shown in step 620 in fig. 6, the precoding matrix may be determined by: determining 6201 channel covariance matrices for a plurality of receiving devices (e.g., UEs) based on channel information between the plurality of receiving devices and transmitting device 300; performing an eigen decomposition operation 6202 on the channel covariance matrix to derive an eigenvector matrix; and selecting 6203 a plurality of eigenvectors from the eigenvector matrix as column vectors of a precoding matrix.

Details of these steps will now be described with respect to some specific embodiments. In a further embodiment, the eigenvector corresponding to the largest multiple eigenvalues may be selected as the column vector of the precoding matrix. In particular embodiments, channel information for the TRX may be obtained based on an uplink sounding reference signal (UL SRS). In another particular embodiment, the precoding matrix may be adjusted periodically. For example, the adjustment period of the precoding matrix may be equal to or greater than the adjustment period of the DP block. It should be appreciated that in addition to or in lieu of the above implementations, the BPP module may be configured based on other precoding matrices derived in a similar manner. Moreover, the precoding matrix may be designed based on other similar, available operations and steps.

Two implementations for the BPP module are set forth below, such as a fully connected mode and a hierarchical fully connected mode:

BPP precoding in full connected mode

Fig. 4a shows BPP precoding in full-connected mode. Assume that there are proposed BPP modules 410a, S TRXs 420a, N TPs 430a for DP modules in proposed transmitting devices in the MU-MIMO system. As shown in fig. 4a, BPP precoding in full-connection mode can be implemented by using adaptive precoding technique in baseband to connect each TP to all S TRXs.

BPP precoding in full-connection mode may be determined at the wideband level using S-TRX channel information of all receiving devices (e.g., all UEs), which may be obtained in the gNB through UL SRS transmission. Such BPP precoding may be achieved by the following steps, as shown in fig. 7:

step 710: a channel covariance matrix for a receiving device (e.g., a UE) in a plurality of active subcarriers is calculated. For example, the channel covariance matrix for user equipment k is calculated:

wherein H_k(w) is the dimension N in subcarrier w_rChannel matrix of xS UE k, N_rIs the number of antenna ports, N, at the UE side_scIs the total number of active subcarriers in the bandwidth, thus R_kThe dimension is S × S.

Step 720: summing the channel covariance matrices of all UEs to obtain the channel covariance matrices of all UEs:

R＝∑_k R_k 2)

where the dimension of R is S × S.

Step 730: performing an Eigen Decomposition (ED) operation on the covariance matrix R to derive an eigenvector matrix U:

R＝U∑U^H 3)

wherein the eigenvector matrix U may be formed from eigenvectors and the eigenvalue matrix Σ may be composed of eigenvalues along the main diagonal. For example, U ═ U₁ … u_S]And

step 740: and selecting an eigenvector corresponding to the largest multiple eigenvalues from the eigenvector matrix U as a column vector of the BPP precoding matrix. That is, the BPP precoding matrix may be formed of T dominant eigenvectors from the eigenvector matrix U having the largest multiple eigenvalues. For example, the BPP precoding matrix W may be written as:

W＝[u₁ … u_T] 4)

wherein u is_t(T1, …, T) is a feature vector selected from the feature vector matrix U, and the corresponding feature value satisfies₁＞…＞λ_T。

In the fully connected mode, BPP precoding may introduce negligible EBB computation complexity at the broadband level, since only a single EBB computation may be performed to determine BPP precoding based on the S TRXs for all UEs in the broadband. But since the dimensionality of the equivalent channel is reduced to T < S after BPP precoding, the complexity of digital precoding can be significantly reduced.

BPP precoding in hierarchical fully connected mode

Fig. 4b shows BPP precoding in hierarchical fully connected mode. Assume that there are proposed BPP modules 410b, S TRXs 420b, N TPs 430b for DP modules in proposed transmitting devices in the MU-MIMO system. To obtain BPP precoding in the hierarchical full-connection mode, S TRXs and N TPs are divided into a plurality of groups (also referred to as "layers"), where each of the plurality of groups includes two or more TRXs and two or more transmission ports. That is, two or more of the S TRXs 420b may be grouped with two or more of the N TPs 430b to form a group. The plurality of groups may be grouped based on polarization direction. It should be appreciated that in alternative embodiments, the plurality of groups may be grouped based on any other similar characteristic parameter. Each TP 430b in each group may be connected to all TRXs 420b in the same group. The BPP precoding in the hierarchical fully connected mode in fig. 4b is a two-layer structure, for example, where one group (also referred to as "layer" or "pole") corresponds to the polarization direction. As shown in FIG. 4b, group 440 may consist of T/2 TPs and S/2 TRXs. By using adaptive beamforming in baseband, each TP can be connected to all S/2 TRXs in the same group (i.e. belonging to a common polarity).

BPP precoding in hierarchical full-connected mode may be determined at the wideband level using S-TRX channel information for all receiving devices (e.g., all UEs), which may be obtained in the gNB through UL SRS transmission. Such BPP precoding may be implemented by a method in the following steps, as shown in fig. 8:

step 810: for the TRXs in each group, a channel covariance matrix of a receiving device (e.g., UE) in multiple active subcarriers is calculated, which may be calculated similar to equation 1.

Step 820: the channel covariance matrices of all UEs are summed for the TRXs in each group to obtain a channel covariance matrix for each group, which may be calculated similar to equation 2).

Step 830: the ED operation is performed on the channel covariance matrix for each group to derive the eigenvector matrix for each group, and the calculation may be similar to equation 3).

Step 840: the eigenvectors corresponding to the largest multiple eigenvalues are selected from the eigenvector matrix for each group as column vectors for the sub-blocks to form all sub-blocks (e.g., 2 sub-blocks for 2 groups in fig. 4 b), which may be arranged along the diagonal of the BPP precoding matrix. Each sub-matrix for a sub-block may be written similar to equation 4).

As an alternative embodiment, the BPP precoding in the layered full-connected mode may be implemented by another method of the following steps, as shown in fig. 9:

step 910: for the TRXs in each group, a channel covariance matrix of a receiving device (e.g., UE) in multiple active subcarriers is calculated, which may be similar to equation 1.

Step 920: the mean of the channel covariance matrices for all groups (e.g., 2 groups in fig. 4 b) is calculated. For example, the channel covariance matrix for UEk in group i is calculated:

wherein H_k，i(w)Is dimension N_rX S/2 channel matrix of UEk in subcarrier w, N_rIs the number of antenna ports on the UE side, N_scIs the total number of active subcarriers in the bandwidth, thus R_kHas a dimension of S/2 × S/2.

Step 930: summing the average channel covariance matrices of all the UEs to obtain a channel covariance matrix for all the UEs:

R＝∑_k R_k 6)

wherein the dimension of R is S/2 multiplied by S/2.

Step 940: performing ED operation on the covariance matrix R to derive an eigenvector matrix U:

R＝U∑U^H 7)

where the eigenvector matrix U may be formed from the eigenvectors and the eigenvalue matrix sigma may be composed of the eigenvalues along the main diagonal. For example, U ═ U₁ … u_S/2]And

step 950: the eigenvector corresponding to the largest multiple eigenvalues is selected from the eigenvector matrix U as all column vectors for all sub-blocks (e.g., 2 sub-blocks for 2 groups in fig. 4 b), which may be along the diagonal of the BPP precoding matrix. That is, the BPP precoding matrix for each group may be formed by T/2 dominant eigenvectors from the eigenvector matrix U having the largest plurality of eigenvalues. For example, for 2 groups, the BPP precoding matrix W may be written in block diagonal form:

wherein _tu(T1, …, T/2) is an eigenvector selected from the eigenvector matrix U, and is common to both groups, and the corresponding eigenvalue satisfies λ₁＞…＞λ_T/2。

Compared with the full-connection mode, the BPP pre-coding in the layered full-connection mode can further reduce the computation complexity of EBB in the BPP pre-coding from dimension S to S/2, and simultaneously, the complexity of digital pre-coding after the BPP pre-coding can be reduced.

Fig. 5 shows a block diagram of a hybrid precoding scheme 500 including a BPP module according to an embodiment of the disclosure. As shown in fig. 5, it is noted that BPP precoding may introduce an additional precoding module 510 between the analog beamforming 540 and the DP module 530 for hybrid precoding. It should be appreciated that the proposed BPP-based hybrid precoding may reduce EBB computational complexity by half for MU precoding in the gNB and CQI feedback in the UE, and may only require half of the CSI-RS signaling overhead compared to conventional hybrid precoding schemes, whereas BPP-based hybrid precoding may only introduce negligible EBB computational complexity for BPP precoding, as shown in fig. 10.

Further detailed complexity statistics and performance comparisons for BPP precoding schemes are provided below. From the above algorithm description, as shown in fig. 10, it can be seen that the BPP-based hybrid precoding scheme can greatly reduce the EBB complexity and CSI-RS signaling overhead for MU precoding in the gNB and CQI feedback in the UE compared to the conventional hybrid precoding scheme, while it may only bring negligible EBB computation complexity for BPP precoding. Detailed complexity statistics and comparisons are shown in table 1.

Table 1: complexity statistics for different hybrid precoding schemes

Assume that the number S of TRXs is 64 and the number T of TPs is 32 after BPP precoding as shown in fig. 3, 4a and 4 b. For digital precoding and CQI feedback, EBB calculations are performed in each PRB for a total of 50 PRBs. 10 UEs are located in one cell.

For digital precoding in the gNB shown in table 1, the baseline hybrid precoding scheme may require channel matrix-based EBB operation with 64 TRXs in each PRB for each UE, so it may have 50 × 10-500 EBB operations in the gNB with dimension 64. However, BPP-based hybrid precoding may utilize a reduced number of TPs (e.g., T-32) to compute EBBs in the baseband. The BPP-based hybrid precoding can perform 500 EBB operations of only 32 dimensions for digital precoding, and thus it can greatly reduce the baseband processing complexity by half compared to the general hybrid precoding scheme.

For CQI feedback in the UE, as shown in table 1, the baseline hybrid precoding scheme may require EBB operations based on a channel matrix with 64 TRXs in each PRB, so the UE may perform 50 EBB operations of dimension 64 to determine CQI feedback. However, BPP-based hybrid precoding may calculate EBBs in the baseband with a reduced number of TPs (e.g., T-32). The BPP-based hybrid precoding can perform 50 EBB operations with a dimension of only 32 for CQI feedback, and thus it can also greatly reduce the baseband processing complexity by half compared to the conventional hybrid precoding scheme.

For CSI-RS measurements shown in table 1, the baseline hybrid precoding scheme may require CSI-RS signaling with 64 ports, whereas BPP-based hybrid precoding may measure CSI with a reduced number of TPs (e.g., T-32), and thus may save CSI-RS signaling overhead by half.

For the third column in table 1, BPP may introduce an additional precoding module for hybrid precoding, but BPP precoding may perform only a single EBB calculation based on the total variance matrix of all UEs over the wideband, which may only take 1/500-0.2% of the EBB calculation amount in digital precoding. Thus, BPP precoding may only bring negligible EBB computational complexity for hybrid precoding. Furthermore, the hierarchical fully connected mode may require only half the computational complexity of the EBB compared to the fully connected mode.

In summary, BPP-based hybrid precoding may reduce EBB computation complexity for MU precoding in the gNB and CQI feedback in the UE by a factor of two significantly and may only require half of the CSI-RS signaling overhead compared to conventional hybrid precoding schemes, which may introduce negligible EBB computation complexity only for hybrid precoding.

The performance of the proposed BPP precoding scheme for MU-MIMO can be evaluated and compared to existing solutions. For performance evaluation of the hybrid precoding scheme, full buffer system level evaluation is performed in the LTE 3D UMa scenario and dynamic switching between SU and MU-MIMO is considered in the user scheduling process. Results are provided for 64 antenna ports that employ (N1, N2) ═ 8, 4 in the horizontal and vertical dimensions, respectively. Table 2 lists the relevant simulation parameters. The conventional hybrid precoding scheme in fig. 1 is used as a performance reference. The simulation results are shown in table 3.

Table 2: simulation assumptions for System level evaluation

Table 3: system level evaluation of different hybrid precoding schemes

As shown in table 3, BPP can achieve system performance close to that of conventional hybrid precoding, for example, with a loss of up to 4% to 15%, while complexity of baseband processing and CSI-RS signaling overhead can be significantly reduced. The performance of the BPP system in the hierarchical fully-connected mode may be about 2% -13% worse than that of the BPP system in the fully-connected mode, but it may further reduce the complexity of the BPP precoding.

Therefore, since BPP precoding can significantly reduce baseband processing complexity and signaling overhead while system performance loss is very limited, it can be recommended as a promising solution for hybrid precoding application in 5G NR.

Referring now to fig. 11, a simplified block diagram of an apparatus 1100 is shown, according to some embodiments of the present disclosure. The apparatus may be embodied in/as a base station in a MIMO system, which may communicate with multiple UEs simultaneously. For example, the base station may be a gNB operating in a MU-MIMO system. In another embodiment, the apparatus 1100 may be embodied in/as another entity (e.g., a UE) at the user side, which may be communicatively connected to a base station. Apparatus 1100 is operable to perform

exemplary methods

600, 700, 800, and/or 900 described with reference to fig. 6, 7, 8, and/or 9, and may perform any other process or method. It should also be understood that any of the

methods

600, 700, 800, and/or 900 need not be performed entirely by the apparatus 1100. Some of the steps of

methods

600, 700, 800, and/or 900 may be performed by one or more other entities.

The apparatus 1100 may include at least one processor 1101, such as a Data Processor (DP) and at least one memory (MEM)1102 coupled to the processor 1101. The apparatus 1100 may further include a transmitter TX and a receiver RX 1103 coupled to the processor 1101. The MEM 1102 stores a Program (PROG) 1104. The PROG 1104 may include instructions that, when executed on the associated processor 1101, enable the apparatus 1100 to operate according to embodiments of the present disclosure, such as to perform the

methods

600, 700, 800, or 900. The combination of the at least one processor 1101 and the at least one MEM 1102 may form a processing module 1105 suitable for implementing various embodiments of the present disclosure.

Various embodiments of the disclosure may be implemented by computer programs, software, firmware, hardware, or a combination thereof executable by the processor 1101. By way of non-limiting example, the processor 1101 may be of any type suitable to the local technical environment, and may include one or more of general purpose computers, special purpose computers, microprocessors, digital signal processors, DSPs, and processors based on a multi-core processor architecture. By way of non-limiting example, the MEM 1102 may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor-based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory and removable memory.

The transmitter TX and receiver RX 1103 may have multiple antennas that support MU-MIMO techniques with various transmit diversity schemes. For example, apparatus 1100 may include two transmit antennas or four transmit antennas that support beamforming.

Additionally, the present disclosure may also provide a carrier containing a computer program as described above, wherein the carrier is one of an electronic signal, optical signal, radio signal, or computer readable storage medium. The computer readable storage medium may be, for example, an optical disk or an electronic storage device such as a RAM (random access memory), a ROM (read only memory), a flash memory, a magnetic tape, a CD-ROM, a DVD, a blu-ray disk, and the like.

The techniques described herein may be implemented by various means so that a device implementing one or more functions of the corresponding device described in the embodiments includes not only the means of the prior art but also means for implementing one or more functions of the device described in the embodiments and it also includes separate means for each separate function or means that can be configured to perform two or more functions. For example, these techniques may be implemented in hardware (one or more devices), firmware (one or more devices), software (one or more modules), or a combination thereof. For firmware or software, the functions described herein may be performed by modules (e.g., procedures, functions, and so on).

Exemplary embodiments herein have been described above with reference to block diagrams and flowchart illustrations of methods and apparatus. It will be understood that each block of the block diagrams and flowchart illustrations, and combinations of blocks in the block diagrams and flowchart illustrations, respectively, can be implemented by various means including computer program instructions. These computer program instructions may be loaded onto a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions which execute on the computer or other programmable data processing apparatus create a means for implementing the functions specified in the flowchart block or blocks.

Further, while operations are described in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In some cases, multitasking and parallel processing may be advantageous. Also, while the above discussion contains several specific implementation details, these should not be construed as limitations on the scope of the subject matter described herein, but rather as descriptions of features of possible specific embodiments. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination.

It is obvious to a person skilled in the art that with the advancement of technology, the inventive concept may be implemented in various ways. The above-described embodiments are given for the purpose of illustration and not limitation of the present disclosure, and it is to be understood that modifications and variations may be made without departing from the spirit and scope of the disclosure, as will be readily understood by those skilled in the art. Such modifications and variations are considered to be within the scope of the disclosure and the appended claims. The scope of the disclosure is defined by the appended claims.

Claims

1. An apparatus for beam-based preprocessing in a transmitting device in a multi-user multiple-input multiple-output (MU-MIMO) system, the transmitting device further comprising a transceiver unit (TRX) array and a digital precoding module, the apparatus configured to:

mapping a plurality of TRXs of the TRX array to a plurality of transmission ports of the digital precoding module according to a precoding matrix, wherein the number of the transmission ports is less than the number of the plurality of TRXs;

wherein the precoding matrix is determined by:

determining a channel covariance matrix for a plurality of receiving devices based on channel information between the plurality of receiving devices and the transmitting device;

performing an eigen decomposition operation on the channel covariance matrix to derive an eigenvector matrix; and

selecting a plurality of eigenvectors from the eigenvector matrix as column vectors of the precoding matrix.

2. The apparatus of claim 1, wherein selecting a plurality of eigenvectors comprises selecting an eigenvector corresponding to a largest plurality of eigenvalues.

3. The apparatus of claim 2, wherein each transmit port is connected to all TRXs of the plurality of TRXs, and determining a channel covariance matrix for the plurality of receiving devices comprises:

calculating a channel covariance matrix of the receiving device in the plurality of active subcarriers;

summing the channel covariance matrices for the plurality of receiving devices to obtain a channel covariance matrix for the plurality of receiving devices,

and wherein the number of selected eigenvectors is equal to the number of transmit ports of the digital pre-coding module.

4. The apparatus of claim 2, wherein the plurality of TRXs and the plurality of transmit ports are divided into a plurality of groups, wherein each of the plurality of groups comprises two or more TRXs and two or more transmit ports, and each transmit port in each group is connected to all TRXs in the same group, and

determining a channel covariance matrix for the plurality of receiving devices comprises:

calculating channel covariance matrices of the receiving devices in the plurality of active subcarriers for the TRXs in each group;

summing channel covariance matrices of the plurality of receiving devices for the TRXs in each group to obtain a channel covariance matrix for each group;

performing a feature decomposition operation includes:

performing an eigen decomposition operation on the channel covariance matrix for each group to derive an eigenvector matrix for each group; and is

Selecting a plurality of feature vectors comprises:

selecting eigenvectors from an eigenvector matrix for each group as column vectors of subblocks, wherein a plurality of subblocks are arranged along a diagonal of the precoding matrix, and the number of eigenvectors selected for each group is equal to the number of transmission ports of the digital precoding modules in the same group.

5. The apparatus of claim 2, wherein the plurality of TRXs and the plurality of transmit ports are divided into a plurality of groups, wherein each of the plurality of groups comprises two or more TRXs and two or more transmit ports, and each transmit port in each group is connected to all TRXs in the same group, and

determining a channel covariance matrix for a plurality of receiving devices comprises:

calculating an average of channel covariance matrices over a plurality of groups;

summing the average channel covariance matrices for the plurality of receiving devices to obtain a channel covariance matrix for the plurality of receiving devices; and is

Selecting a plurality of feature vectors comprises:

the eigenvectors are selected as column vectors of all sub-blocks along the diagonal of the precoding matrix, where the number of eigenvectors selected is equal to the number of transmit ports of the digital precoding module in each group.

6. The apparatus of claim 4 or 5, wherein the plurality of groups are grouped based on polarization direction.

7. The apparatus of claim 1, wherein the channel information for the TRX is obtained based on an uplink sounding reference signal.

8. The apparatus of claim 1, wherein the precoding matrix is adjusted periodically.

9. The apparatus of claim 8, wherein an adjustment period of the precoding matrix is equal to or greater than an adjustment period of the digital precoding module.

10. A transmitting apparatus in a multi-user multiple-input multiple-output (MU-MIMO) system, the transmitting apparatus comprising:

a digital pre-coding module;

a transceiver unit (TRX) array comprising a plurality of TRXs for transmitting output data signals of the digital precoding module;

an analog beamforming module for weighting the data signals in the TRX array; and is

The transmitting device further comprising an apparatus for beam-based preprocessing according to any of claims 1 to 9.

11. A method for configuring an apparatus for beam-based preprocessing in a transmitting device in a multi-user multiple-input multiple-output (MU-MIMO) system, the transmitting device further including a transceiver unit (TRX) array and a digital precoding module, and the method comprising:

mapping a plurality of TRXs of the TRX array to a plurality of transmit ports of the digital precoding module according to a precoding matrix, wherein the number of transmit ports is less than the number of the plurality of TRXs;

wherein the precoding matrix is determined by:

performing characteristic decomposition operation on the channel covariance matrix to obtain an eigenvector matrix; and

12. The method of claim 11, wherein the step of selecting a plurality of eigenvectors comprises selecting an eigenvector corresponding to the largest plurality of eigenvalues.

13. The method of claim 12, wherein each transmit port is connected to all TRXs of the plurality of TRXs, and determining a channel covariance matrix for the plurality of receiving devices comprises:

calculating a channel covariance matrix of a receiving device in the plurality of active subcarriers;

14. The method of claim 12, wherein the plurality of TRXs and the plurality of transmit ports are divided into a plurality of groups, wherein each of the plurality of groups includes two or more TRXs and two or more transmit ports, and each transmit port in each group is connected to all TRXs in the same group, and

the step of determining a channel covariance matrix for the plurality of receiving devices comprises:

calculating channel covariance matrices of receiving devices in the plurality of active subcarriers for the TRXs in each group;

the step of performing a feature decomposition operation includes:

The step of selecting a plurality of feature vectors comprises:

the eigenvectors are selected from the eigenvector matrix for each group as column vectors of the sub-blocks, wherein a plurality of sub-blocks are arranged along a diagonal of the precoding matrix, and the number of eigenvectors selected for each group is equal to the number of transmission ports of the digital precoding modules in the same group.

15. The method of claim 12, wherein the plurality of TRXs and the plurality of transmit ports are divided into a plurality of groups, wherein each of the plurality of groups includes two or more TRXs and two or more transmit ports, and each transmit port in each group is connected to all TRXs in the same group, and

the step of determining a channel covariance matrix for a plurality of receiving devices comprises:

The step of selecting a plurality of feature vectors comprises:

selecting the eigenvectors as column vectors of all sub-blocks along a diagonal of the precoding matrix, wherein the number of eigenvectors selected is equal to the number of transmit ports of the digital precoding module in each group.

16. The method of claim 14 or 15, wherein the plurality of groups are grouped based on polarization direction.

17. The method of claim 11, wherein the channel information for the TRX is obtained based on an uplink sounding reference signal.

18. The method of claim 11, wherein the precoding matrix is adjusted periodically.

19. The method of claim 18, wherein an adjustment period of the precoding matrix is equal to or greater than an adjustment period of the digital precoding module.

20. A computer-readable storage medium having stored thereon computer-executable instructions that, when executed by at least one processor, cause performance of the method recited in any one of claims 11-19.