CN109196789A

CN109196789A - A kind of downlink transmission method and the network equipment

Info

Publication number: CN109196789A
Application number: CN201680086187.9A
Authority: CN
Inventors: 杨非; 陈凯; 王智鹰
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2016-08-19
Filing date: 2016-08-19
Publication date: 2019-01-11
Also published as: WO2018032492A1

Abstract

The embodiment of the invention discloses a kind of downlink transmission method and the network equipments, for solving the problems, such as that the TM4 of existing multi-input multi-output system is not suitable for carrying out downlink transfer to MU.It include: that the network equipment receives N number of pre-coding matrix instruction PMI, N number of PMI is N number of user equipment (UE) according to determined by the first weighted pilot signal that the network equipment is sent, and the first weighted pilot signal is weighted according to the first pilot weighted matrix to the first pilot signal by the network equipment；The network equipment determines the main feature vector of N number of reconstruct channel according to N number of PMI and the first pilot weighted matrix；The network equipment determines M scheduled UE from N number of UE, and determines the second pilot weighted matrix according to the main feature vector of reconstruct channel of each of M scheduled UE UE；Second weighted pilot signal is sent to N number of UE by the network equipment, wherein the second weighted pilot signal is weighted according to the second pilot weighted matrix to the second pilot signal by the network equipment.

Description

Downlink transmission method and network equipment

Technical Field

The present invention relates to the field of communications, and in particular, to a downlink transmission method and a network device.

Background

In a multiple-input multiple-output (MIMO) system, since reciprocity (channel reliability) of uplink and downlink channels in a Time Division Duplex (TDD) system is established, a network device obtains accurate instantaneous downlink Channel State Information (CSI) by detecting an uplink reference signal sent by a User Equipment (UE), so that accurate beamforming transmission can be performed. The reciprocity of uplink and downlink channels in a Frequency Division Duplex (FDD) system is generally not established, a network device cannot directly use an uplink channel estimation result to send downlink beamforming, but adopts a fixed codebook for channel characteristic vector quantization, and the UE selects an optimal precoding codeword based on a certain criterion from the codebook according to the downlink channel estimation result and feeds back an index of the optimal precoding codeword to the network device, that is: after detecting the PMI, the network device uses a precoding codeword corresponding to the PMI to perform weighted transmission on the downlink data signal.

In a Transmission Mode (TM) 4 of the MIMO system, a network device and a UE in an FDD system store a same set of prescribed precoding codebook, where the precoding codebook includes a plurality of precoding matrices. In the communication process, as shown in fig. 1, the network device sends a pilot signal to the UE, for example: the method comprises the steps that a cell specific reference signal (CRS) is detected by UE (user equipment), a downlink channel is estimated by detecting the CRS, an optimal quantization result of a current downlink channel estimation result in a precoding codebook is selected according to a criterion set in the UE, the optimal quantization result is used as a PMI to be sent out to network equipment, the network equipment detects the PMI sent by the UE, and a precoding matrix corresponding to the PMI is set as a single-user (SU) precoding matrix.

Since TM4 employs a precoding matrix based on a precoding codebook, wherein a codebook of 2 antenna ports is defined in a 3GPP related protocol, including 4 2-dimensional complex vectors as PMI codewords of rank (rank) ═ 1; and a codebook of 4 antenna ports including 16 4-dimensional complex vectors as PMI codewords of Rank ═ 1. For SU, the UE can select a PMI codeword matched with an actual downlink channel from the 16 PMI codewords, and then the network device determines a precoding matrix according to the PMI codeword selected by the UE, but for a multi-user (MU), a quantization error of the 16 PMI codewords and the actual downlink channel corresponding to each UE is large, and a PMI codeword selected by each UE from the 16 PMI codewords is not matched with the actual downlink channel corresponding to each UE, so that the network device determines the precoding matrix according to the PMI codeword selected by each UE and performs weighted transmission on a downlink data signal, which may cause severe user interference.

Disclosure of Invention

The embodiment of the invention provides a downlink transmission method and network equipment, which are used for solving the problem that a TM4 of the existing multi-input multi-output system is not suitable for downlink transmission of MU.

The first aspect of the present invention provides a downlink transmission method, which is applied to a mimo system, and with the technical solution of the present invention, a TM4 of the mimo system is not only suitable for downlink transmission of SU, but also suitable for downlink transmission of MU, and the method includes: the network equipment receives N PMIs, wherein N is an integer larger than 1, the N PMIs are determined by N UEs according to the downlink channel estimation of a first weighted pilot signal sent by the network equipment and the result of the downlink channel estimation, the first weighted pilot signals are weighted by the network device in accordance with a first pilot weighting matrix, the network device determines N reconstructed channel principal eigenvectors from the N PMIs and the first pilot weighting matrix, the network device determines M scheduled UEs from the N UEs and determines a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each of the M scheduled UEs, wherein M is an integer no greater than N, the network device transmitting second weighted pilot signals to the N UEs, wherein the second weighted pilot signal is obtained by the network device weighting the second pilot signal according to the second pilot weighting matrix.

Different from the prior art, the network device receives N PMIs, where N is an integer greater than 1, the N PMIs are determined by N UEs according to a first weighted pilot signal sent by the network device, the first weighted pilot signal is obtained by the network device by weighting a first pilot signal according to the first pilot weighting matrix, when the network device weights the first pilot signal, what the UE really sees is not the first pilot signal, nor determines the PMI through the first pilot signal, and what the UE really sees is the first weighted pilot signal, and determines the PMI after downlink channel estimation is performed through the first weighted pilot signal, so as to expand the selection range of the PMIs, because the N PMIs not the quantization of a real downlink channel, but the quantization of the downlink channel after the first pilot signal is weighted through the first pilot weighting matrix, therefore, the network device needs to perform corresponding inverse transformation to correctly restore the real downlink channel, and then determines N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix, then determines M scheduled UEs from the N UEs, and then determines a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, so that a second weighted pilot signal can be obtained by weighting the second pilot signal through the second weighting matrix, and the second weighted pilot signal is sent to the N UEs, thereby reducing quantization errors of the downlink channel, better suppressing interference between users, and improving the accuracy of the downlink channel.

In some possible implementations, the first pilot weighting matrix is a target pilot weighting matrix determined by the previous pilot weighting, and is used to weight the pilot signal at this time, since the pilot signal is not weighted in the prior art, the first pilot weighting matrix for weighting the pilot signal for the first time is at least one of a unit weighting matrix and a preset pilot weighting matrix, the target pilot weighting matrix determined after weighting the pilot signal for the second time is used as a first pilot weighting matrix for the third time, and the pilot signal for the third time is weighted according to the first pilot weighting matrix, and so on, which is not limited herein.

In some possible implementations, there are many ways for the network device to determine the N reconstructed channel principal eigenvectors according to the N PMIs and the first pilot weighting matrix, and one possible way includes: the network device determines N precoding code words corresponding to the N PMI first according to the N PMI, then determines a conjugate transpose matrix corresponding to the N precoding code words and a conjugate transpose matrix corresponding to the first pilot weighting matrix, and then determines the N reconstructed channel main eigenvectors according to the N precoding code words, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix. Since the N PMIs are not quantization of the true downlink channel, but quantization of the downlink channel weighted by the first pilot weighting matrix, the network device correctly restores the true downlink channel by determining the N reconstructed channel main eigenvectors.

In some possible implementation manners, the determining, by the network device, the N reconstructed channel main eigenvectors according to the N precoding code words, the first pilot weighting matrix, the conjugate transpose matrices corresponding to the N precoding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix includes: the network device determines the N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is expressed as:

the method comprises the steps of representing a reconstructed channel main characteristic vector of user equipment u determined on a subframe t, PrimEigVec representing a main characteristic vector of an acquisition matrix, L representing the total number of measurement subframes within a preset length, m being a measurement subframe sequence number and representing that the measurement subframe is s_mRepresents the user equipment u to the measurement sub-frame s_mThe first pilot signal of (a) is selected as a PMI after channel estimation, and the PMI is expressed as a pairThe corresponding precoding code word represents the corresponding conjugate transpose matrix and represents the corresponding conjugate transpose matrix.

In order to make the pilot power weighted by the first pilot signal unchanged and keep the correlation of the PMI before and after rotation unchanged, the network device constrains the first pilot weighting matrix to be a unitary matrix, and then the network device locally maintains a queue for storing the first pilot weighting matrix and another queue for storing the PMI fed back by each UE, and determines N reconstructed channel main eigenvectors through the first formula.

In some possible implementation manners, before determining the second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, a candidate pilot weighting matrix set is determined, and then a certain candidate pilot weighting matrix is selected from the candidate pilot weighting matrix set as the second pilot weighting matrix according to a certain manner, where the specific implementation process includes the following possible manners:

the first possible way is: if the first pilot weighting matrix is a unit weighting matrix, before determining a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, if M is 1, the network device determines an alternative pilot weighting matrix set according to the reconstructed channel main eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE.

In some possible implementations, the determining, by the network device, the candidate pilot weighting matrix set according to the reconstructed channel main eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE includes: the network equipment determines an alternative pilot weighting matrix set by using a second formula according to a reconstructed channel main eigenvector corresponding to one UE and a precoding code word corresponding to the one UE, wherein the second formula is represented as:

wherein, W_i∈w₁Representing a set of candidate pilot weighting matrices, W_iA precoding codeword with PMI i indicating rank 1 represents a reconstructed channel principal eigenvector corresponding to one UE, and w₁A PMI codebook representing rank 1.

It can be seen that, if the scheduled user is a single user and the first pilot weighting matrix is a unit weighting matrix, the network device determines the candidate pilot weighting matrix set according to the second formula directly according to the reconstructed channel dominant eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE, and the calculation amount is small.

A second possible way is: if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, before determining a second pilot weighting matrix according to a main eigenvector of a reconstructed channel of each UE of the M scheduled UEs, if M is 2, the network device determines a target weight matrix according to the main eigenvectors of the reconstructed channels respectively corresponding to the two UEs, and then when the correlation of the main eigenvectors of the reconstructed channels respectively corresponding to the two UEs is smaller than a first preset threshold, the network device determines an alternative pilot weighting matrix set according to a pre-coded code word corresponding to the two UEs and the target weight matrix.

In some possible implementation manners, the determining, by the network device, the target weight matrix according to the reconstructed channel main eigenvectors respectively corresponding to the two UEs includes: the network equipment determines the target weight matrix by using a third formula according to the main eigenvectors of the reconstructed channel corresponding to the two UEs respectively, wherein the third formula is expressed as:

wherein, represents the target weight matrix and represents u₁Corresponding reconstructed channel principal eigenvector, representing u₂Corresponding reconstructed channel principal eigenvector, representing u₁Corresponding weight vector, representing u₂The corresponding weight vector, the expressed conjugate transpose matrix and the expressed diagonal matrix.

In some possible implementation manners, the determining, by the network device, the candidate pilot weighting matrix set according to the precoding code words and the target weight matrix corresponding to the two UEs includes: the network equipment determines an alternative pilot frequency weighting matrix set by using a fourth formula according to the precoding code words and the target weight matrix corresponding to the two UEs, wherein the fourth formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iThe PMI of rank 2 is a precoding codeword of i, which represents a target weight matrix.

It can be seen that, if two users are scheduled, and if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, the network device first determines the target weight matrices corresponding to the two UEs, where the specific determination manner is as the third formula, and generally, the power corresponding to the target weight matrix corresponding to each UE is the same. Then, the correlations of the main eigenvectors of the reconstructed channels corresponding to the two UEs are compared, when the correlations of the main eigenvectors of the reconstructed channels corresponding to the two UEs are smaller than a first preset threshold, the downlink channels corresponding to the two UEs have no interference or have small interference, wherein the first preset threshold is generally 1, the first preset threshold can be determined according to actual conditions, and is not specifically limited here, the network device determines the corresponding candidate pilot weighting matrix set according to the target weight matrix and the precoding code words corresponding to the two UEs, and the specific determination mode is as in the fourth formula.

A third possible way is: if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, before determining a second pilot weighting matrix according to a main eigenvector of a reconstructed channel of each UE of the M scheduled UEs, if M is greater than or equal to 2, the network equipment determines a target weight matrix according to the main eigenvector of the reconstructed channel corresponding to each of the at least two UEs; and when the correlation of the target weight matrixes corresponding to any two pieces of UE is smaller than a second preset threshold value, the network equipment determines an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the at least two pieces of UE and the target weight matrixes.

In some possible implementation manners, the determining, by the network device, the candidate pilot weighting matrix set according to the precoding code words corresponding to the at least two UEs and the target weight matrix includes: the network device determines a candidate pilot weighting matrix set by using a fifth formula according to the precoding code words and the target weight matrix corresponding to the at least two UEs, wherein the fifth formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iAnd representing precoding code words with PMI i, wherein rank is more than or equal to 2, and representing the target weight matrix.

It can be seen that, if at least two users are scheduled, and if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, the network device first determines the target weight matrices corresponding to the at least two UEs, where the specific determination manner is as the third formula, and generally, the power corresponding to the target weight matrix corresponding to each UE is the same. When the correlation between the target weight matrices corresponding to any two UEs in the at least two UEs is smaller than a second preset threshold, the candidate pilot weighting matrix set is determined according to the target weight matrix and the precoding code words corresponding to the at least two UEs, where the specific determination manner is, for example, a fifth formula, and in practical application, the second preset threshold is determined by the network device according to an actual situation, for example, the second preset threshold is 1, and is not specifically limited here.

After determining the candidate pilot weighting matrix set, selecting a certain candidate pilot weighting matrix from the candidate pilot weighting matrix set according to a certain preset mode as the second pilot weighting matrix, wherein the specific implementation process includes the following possible modes:

the first possible way is: if the first pilot weighting matrix is at least one of a unity weighting matrix and a preset pilot weighting matrix, determining a second pilot weighting matrix according to a reconstructed channel dominant eigenvector of each UE of the M scheduled UEs comprises: the network equipment randomly selects a target pre-coding code word, selects an alternative pilot weighting matrix corresponding to the target pre-coding code word from the alternative pilot weighting matrix set according to the target pre-coding matrix, and determines the alternative pilot weighting matrix corresponding to the target pre-coding code word as the second pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, the network device randomly selects a certain pre-coding codeword from N pre-coding codewords corresponding to N UEs as the target pre-coding codeword, selects an alternative pilot weighting matrix corresponding to the target pre-coding codeword from the alternative pilot weighting matrix set according to the target pre-coding codeword, and then uses the alternative pilot weighting matrix corresponding to the target pre-coding codeword as the second pilot weighting matrix, where the method is used for determining the second pilot weighting matrix under the non-measured subframe, and the second pilot weighting matrix is used for performing weighted transmission on a next pilot signal, thereby improving the accuracy of the downlink channel.

A second possible way is: if the first pilot weighting matrix is a unit weighting matrix, determining a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs comprises: the network equipment determines a target pre-coding code word according to a preset rule, selects an alternative pilot weighting matrix corresponding to the target pre-coding code word from the alternative pilot weighting matrix set according to the target pre-coding matrix, and then determines the alternative pilot weighting matrix corresponding to the target pre-coding code word as the second pilot weighting matrix.

In some possible implementations, the determining, by the network device, the target precoding codeword according to a preset rule includes: the network device determines the target precoding codeword using a sixth formula, wherein the sixth formula is expressed as:

wherein, representing a target pre-coding code word, n represents a measurement subframe s_mThe sequence number of the previous measurement subframe indicates that the measurement subframe is s_mA corresponding candidate pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is a unit weighting matrix, the network device determines a target precoding codeword according to a sixth formula, and selects an alternative pilot weighting matrix corresponding to the target precoding codeword from the alternative pilot weighting matrix set according to the target precoding codeword as the second pilot weighting matrix, where the method is used to measure the second pilot weighting matrix determined under the subframe, the second pilot weighting matrix is used to weight and transmit a next pilot signal, and the target second pilot weighting matrix and the target precoding codeword are used to weight and transmit a data signal once, thereby effectively improving the accuracy of the downlink channel.

In practical application, the network device may further determine the second pilot weighting matrix according to the first pilot weighting matrix, and the specific implementation process includes the following possible manners:

the first possible way is: if the first pilot weighting matrix is a preset pilot weighting matrix, determining the second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs comprises: if M is 1, the network device first obtains a target weight matrix of one UE, and then the network device determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the one UE.

In some possible implementations, the determining, by the network device, the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the UE includes: the network device first determines whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, wherein the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁A PMI codebook representing rank 1, a conjugate transpose matrix represented, Qmod (m, L) represents a first pilot weighting matrix with a measurement subframe number m in L measurement subframes, and δ represents a constraint threshold value;

the network device determines the first pilot weighting matrix satisfying the seventh formula as the second pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is a preset pilot weighting matrix and when the scheduled user is a single user, the network device first obtains the target weight matrix of the UE, and determines the target pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the UE, where the specific determination method is as in the seventh formula, where the method is used to measure the target pilot weighting matrix determined under the subframe, and the target pilot weighting matrix is used to perform weighted transmission on the next pilot signal, so as to effectively improve the accuracy of the downlink channel.

In some possible implementation manners, the network device further needs to determine a target precoding codeword, and weight the next data signal according to the target precoding codeword and the second pilot weighting matrix to improve the accuracy of the downlink channel, because the network device may determine the second pilot weighting matrix first and does not determine the target precoding codeword, after the network device determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix, the network device determines the target precoding codeword by using an eighth formula, where the eighth formula is represented as:

wherein the target precoding codeword is represented.

In some possible implementations, after determining the second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, the network device sends a first weighted data signal to the N UEs, where the first weighted data signal is obtained by weighting, by the network device, the first data signal according to the target precoding codeword and the second pilot weighting matrix.

As can be seen, the network device weights the first data signal according to the target precoding code word and the second pilot weighting matrix, so as to obtain a first weighted data signal, where the first data signal includes control information and the like, and sends the first weighted data signal to the N UEs, so as to improve the accuracy of a downlink channel, and also enable the TM4 in the MIMO system to be suitable for downlink transmission of MUs, thereby effectively reducing signal interference between MUs.

A second aspect of the present invention provides a network device configured to implement the functionality of the method provided by the first aspect above. The functions may be implemented by hardware, or by hardware executing corresponding software, where the hardware or software includes one or more modules corresponding to the above functions.

Wherein, this network equipment includes: the device comprises a receiving module, a first determining module, a second determining module, a third determining module and a sending module.

A receiving module, configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, where the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by a network device, and the first weighted pilot signal is obtained by weighting, by the network device, the first pilot signal according to a first pilot weighting matrix.

And the first determining module is used for determining N reconstructed channel main eigenvectors according to the N PMIs received by the receiving module and the first pilot weighting matrix.

A second determining module to determine M scheduled UEs from the N UEs.

And a third determining module, configured to determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs determined by the second determining module, where M is an integer no greater than N.

A sending module, configured to send the second weighted pilot signal determined by the third determining module to the N UEs, where the second weighted pilot signal is obtained by weighting, by the network device, a second pilot signal according to the second pilot weighting matrix.

A third aspect of the present invention provides a network device, which includes a processor, a memory, a bus system, and an input/output I/O device, the processor, the memory, and the I/O device being connected via the bus system, wherein the memory stores one or more programs, the memory is used to provide the processor with operation instructions and data included in the one or more programs, and the processor executes the programs stored in the memory to implement the steps in the method provided by the first aspect.

The I/O device is configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by a network device, and the first weighted pilot signal is obtained by weighting, by the network device, the first pilot signal according to a first pilot weighting matrix.

The processor 301 is configured to determine N reconstructed channel primary eigenvectors according to the N PMIs and the first pilot weighting matrix.

The processor 301 is further configured to determine M scheduled UEs from the N UEs, and determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, where M is an integer no greater than N.

The I/O device 304 is further configured to transmit second weighted pilot signals to the N UEs, wherein the second weighted pilot signals are obtained by the network device weighting second pilot signals according to the second pilot weighting matrix.

A fourth aspect of the present invention provides a downlink transmission system, including a network device and N user equipments UEs, where N is an integer greater than 1, and the network device is in communication connection with the N UEs;

the network device is the network device according to the second aspect or any optional implementation manner of the second aspect.

A fifth aspect of the present invention provides a computer storage medium, where the computer storage medium stores a program for downlink transmission according to the first aspect or any one of the optional implementations of the first aspect.

Different from the prior art, the network device receives N PMIs, the N PMIs are determined by N UEs according to a first weighted pilot signal sent by the network device, the first weighted pilot signal is obtained by the network device by weighting a first pilot signal according to a first pilot weighting matrix, when the network device weights the first pilot signal, the UE really sees the first pilot signal but not the first pilot signal, and the UE really sees the first weighted pilot signal, and determines the PMI after downlink channel estimation is performed through the first weighted pilot signal, so as to expand the selection range of the PMIs, because the N PMIs not the quantization of the real downlink channel, but the PMI is the quantization of the downlink channel weighted by the first pilot weighting matrix, therefore, the network device needs to perform corresponding inverse transformation to correctly restore the real downlink channel, the network device determines N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix, determines M scheduled UEs from the N UEs, and determines a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, so that a second weighted pilot signal can be obtained by weighting the second pilot signal by the second weighting matrix, and the second weighted pilot signal is sent to the N UEs, thereby reducing quantization errors of downlink channels, better suppressing interference between users, and improving accuracy of the downlink channels.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a diagram of an embodiment of downlink transmission in the prior art;

fig. 2 is a diagram of a network architecture of a MIMO system in accordance with an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of a network device according to an embodiment of the present invention;

fig. 4 is a schematic diagram of an embodiment of a downlink transmission method in an embodiment of the present invention;

fig. 5 is a schematic diagram of an application scenario of a downlink transmission method in the embodiment of the present invention;

fig. 6 is a schematic diagram of another application scenario embodiment of a downlink transmission method in the embodiment of the present invention;

fig. 7 is another schematic structural diagram of a network device in an embodiment of the present invention;

fig. 8 is another schematic structural diagram of a network device in the embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Before describing embodiments of the present invention, terms that may be involved in embodiments of the present invention are described:

the term "precoding codebook": a matrix or vector for quantizing a MIMO channel set in a Long Term Evolution (LTE) protocol, wherein the precoding codebook is represented as each precoding codebook including a plurality of precoding matrices, wherein the precoding matrix is represented as B representing the precoding codebook size, and the number of rows of the matrix is N_TThe number of antennas of the network device is represented, the number of columns r represents the rank of the precoding matrix, and i represents the index of each codeword W in the precoding codebook.

The term "rank": is based on the rank of the precoding codebook w. Each PMI codeword is N when the rank is 1_TEach codeword is an N with rank of 2_TA complex matrix of x 2. The rank can be determined by the UE according to the downlink signaling sent by the network device, or can be selected by the UE according to the current channel, and the UE feeds back the rank used by the UE while feeding back the PMI.

The term "pilot weighting matrix": the matrix for weighting the pilot signals of the multiple antennas is N_TA dimensional unitary matrix.

The term "CSI": for accurate instantaneous channel state information, for the MIMO channel, the channel coefficient matrix, the channel correlation matrix, the channel eigenvector, etc. are included, and the CSI fed back by the UE includes a PMI and a Channel Quality Indicator (CQI).

The term "channel principal eigenvector": the channel coefficient matrix is subjected to Singular Value Decomposition (SVD), and the right singular vector corresponding to the maximum singular value is the main eigenvector of the channel.

The term "measurement subframe": the network equipment sends CRS in each subframe, UE carries out downlink channel estimation on the CRS in each subframe, and the downlink channel estimation result is used for data channel demodulation of the subframe. Those subframes in which the downlink channel estimation result is only used for demodulation are called non-measurement subframes, and the measurement subframes are configured by the network equipment according to the LTE protocol and are signaled to all UEs served by the network equipment.

It should be understood that the technical solution of the embodiment of the present invention may be applied to a MIMO system, where the MIMO technology refers to that a plurality of transmitting antennas are used at a transmitting end, a plurality of receiving antennas are used at a receiving end, and a signal is transmitted through a plurality of antennas at the transmitting end and received through a plurality of antennas at the receiving end, thereby effectively improving the transmission efficiency of the signal. As shown in fig. 2, under the framework of the MIMO system, the MIMO system includes: the network equipment sends data to the user equipment in a downlink transmission mode, and the user equipment sends data to the network equipment in an uplink transmission mode.

User equipment, which may be referred to as a User Equipment (UE), access user equipment, a subscriber unit, a subscriber station, a mobile station, a remote station, remote user equipment, a mobile device, a wireless communication device, a user agent, or a user equipment, may communicate with one or more core networks via a Radio Access Network (RAN). The access user equipment may be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA), a handheld device with wireless communication capability, a computing device or other processing device connected to a wireless modem, a vehicle mounted device, a wearable device, a user equipment in a future 5G network, etc.

The network device may be a device for communicating with the user device, for example: the network device may be a Base Transceiver Station (BTS) in a GSM system or a CDMA system, a network device (NB) in a WCDMA system, an evolved node B (eNB or eNodeB) in an LTE system, or a relay station, an access point, a vehicle-mounted device, a wearable device, a network side device in a future 5G network, or a network device in a future evolved PLMN network.

Next, a structure of a network device according to an embodiment of the present invention is described, and as shown in fig. 3, the network device 300 includes: a processor 301, a memory 302, a bus system 303 and an input/output (I/O) device 304, wherein the processor 301, the memory 302 and the I/O device 304 are connected via the bus system 303, the memory 302 stores one or more programs, and the memory 302 is used for providing operating instructions and data included in the one or more programs to the processor 301;

the I/O device 304 is configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by a network device, and the first weighted pilot signal is obtained by the network device by weighting the first pilot signal according to a first pilot weighting matrix;

the processor 301 is configured to determine N reconstructed channel dominant eigenvectors according to the N PMIs and the first pilot weighting matrix;

the processor 301 is further configured to determine M scheduled UEs from the N UEs, and determine a second pilot weighting matrix according to a reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, where M is an integer no greater than N;

In some possible implementations, the first pilot weighting matrix is at least one of an identity weighting matrix and a preset pilot weighting matrix.

In some possible implementations, the processor 301 is specifically configured to determine precoding codewords corresponding to the N PMIs according to the N PMIs; determining a conjugate transpose matrix corresponding to the N precoding codewords and a conjugate transpose matrix corresponding to the first pilot weighting matrix; and determining the N reconstructed channel main eigenvectors according to the precoding code words corresponding to the N PMIs, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix.

In some possible implementations, the processor 301 is specifically configured to determine the N reconstructed channel main eigenvectors according to the precoding code words corresponding to the N PMIs, the first pilot weighting matrix, the conjugate transpose matrices corresponding to the N precoding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is expressed as:

the method comprises the steps of representing a reconstructed channel main characteristic vector of user equipment u determined on a subframe t, PrimEigVec representing a main characteristic vector of an acquisition matrix, L representing the total number of measurement subframes within a preset length, m being a measurement subframe sequence number and representing that the measurement subframe is s_mRepresents the user equipment u to the measurement sub-frame s_mThe selected PMI after the channel estimation of the first pilot signal represents that the PMI is a corresponding precoding code word, represents a corresponding conjugate transpose matrix and represents a corresponding conjugate transpose matrix.

In practical application, before determining the second pilot weighting matrix, the processor 301 may determine a candidate pilot weighting matrix set, select the second pilot weighting matrix from the candidate pilot weighting matrix set, weight a second pilot signal according to the second pilot weighting matrix to obtain a second weighted pilot signal, and send the second weighted pilot signal to the N UEs, so as to improve the accuracy of the downlink channel, where the second pilot signal is a next pilot signal used for weighting after the first pilot signal is sent in a weighted manner. In practical applications, there are many ways to determine the candidate pilot weighting matrix set, and the following describes several possible ways:

in some possible implementations, the processor 301 is further configured to, if the first pilot weighting matrix is a unit weighting matrix, before determining a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, if M is 1, determine the candidate pilot weighting matrix set according to the reconstructed channel main eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE.

In some possible implementations, the processor 301 is specifically configured to determine the candidate pilot weighting matrix set by using a second formula according to a reconstructed channel main eigenvector corresponding to one UE and a precoding codeword corresponding to the one UE, where the second formula is represented as:

It can be seen that, if the scheduled user is a single user and the first pilot weighting matrix is a unit weighting matrix, the processor 301 determines the candidate pilot weighting matrix set directly according to the reconstructed channel dominant eigenvector corresponding to one UE and the pre-coding codeword corresponding to one UE, and the specific determination manner is as the second formula, which has a small calculation amount.

In some possible implementations, the processor 301 is further configured to determine, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, a target weight matrix according to the main eigenvector of the reconstructed channel corresponding to each of the two UEs if M is 2 before determining the second pilot weighting matrix according to the main eigenvector of the reconstructed channel of each of the M scheduled UEs; and when the correlation of the main characteristic vectors of the reconstructed channels corresponding to the two UEs is smaller than a first preset threshold, determining the candidate pilot weighting matrix set according to the precoding code words corresponding to the two UEs and the target weight matrix.

In some possible implementations, the processor 301 is specifically configured to determine the target weight matrix by using a third formula according to the primary eigenvectors of the reconstructed channels corresponding to the two UEs, where the third formula is expressed as:

wherein, represents the target weight matrix and represents u₁Corresponding reconstructed channel principal eigenvector, representing u₂Corresponding reconstructed channel principal eigenvector, representing u₁Corresponding weight vector, representing u₂The corresponding weight vector, the expressed conjugate transpose matrix, the expressed diagonal matrix, are used to make the modulo square of the sum 1/2.

In some possible implementations, the processor 301 is specifically configured to determine the candidate pilot weighting matrix set by using a fourth formula according to the precoding code words corresponding to the two UEs and the target weight matrix, where the fourth formula is represented as:

It can be seen that, if two users are scheduled, and the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, the processor 301 first determines target weight matrices corresponding to the two UEs, and determines the target weight matrices in a manner such as a third formula. And when the correlation of the main characteristic vectors of the reconstructed channels respectively corresponding to the two UEs is smaller than a first preset threshold, determining the candidate pilot weighting matrix set according to the target weight matrix and the pre-coding code words respectively corresponding to the two UEs, wherein the specific determination mode is as a fourth formula.

In some possible implementations, the processor 301 is further configured to determine, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, a target weight matrix according to the main eigenvector of the reconstructed channel corresponding to each of the at least two UEs if M is greater than or equal to 2 before determining the second pilot weighting matrix according to the main eigenvector of the reconstructed channel of each of the M scheduled UEs; and when the correlation of the target weight matrixes corresponding to any two UEs is smaller than a second preset threshold, determining the alternative pilot frequency weighting matrix set according to the pre-coding code words corresponding to the at least two UEs and the target weight matrixes.

In some possible implementations, the processor 301 is specifically configured to determine the candidate pilot weighting matrix set by using a fifth formula according to the precoding code words corresponding to the at least two UEs and the target weight matrix, where the fifth formula is represented as:

It can be seen that, when at least two users are scheduled, the processor 301 first determines the target weight matrices corresponding to the at least two UEs, and the manner of determining the target weight matrices is as the third formula. And when the correlation of the target weight matrix corresponding to any two UEs in the at least two UEs is smaller than a second preset threshold, determining an alternative pilot weighting matrix set according to the target weight matrix and the precoding code words corresponding to the at least two UEs, wherein the specific determination mode is as a fifth formula.

After the network device determines the candidate pilot weighting matrix set, a second pilot weighting matrix is determined from the candidate pilot weighting matrix set, and in practical applications, there are many ways to determine the second pilot weighting matrix, and several possible ways are described below:

in some possible implementation manners, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, the processor 301 is specifically configured to randomly select a target pre-coding codeword, and select an alternative pilot weighting matrix corresponding to the target pre-coding codeword from the alternative pilot weighting matrix set according to the target pre-coding matrix; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.

It can be seen that, when the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, the processor 301 randomly selects a precoding code word corresponding to any one UE as the target precoding code word, and uses an alternative pilot weighting matrix corresponding to the target precoding code word as the second pilot weighting matrix, where the method is used for the second pilot weighting matrix determined under the non-measured subframe, and the second pilot weighting matrix is used for weighting and sending the next pilot signal, for example: and weighting a second pilot signal (a pilot signal subjected to downlink weighting next after the first pilot signal is subjected to downlink weighting transmission) according to the second pilot weighting matrix to obtain a second weighted pilot signal, and transmitting the second weighted pilot signal to the N UEs, so that the accuracy of a downlink channel is improved.

In some possible implementation manners, if the first pilot weighting matrix is a unit weighting matrix, the processor 301 is specifically configured to determine a target precoding code word according to a preset rule, and select an alternative pilot weighting matrix corresponding to the target precoding code word from the alternative pilot weighting matrix set according to the target precoding matrix; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.

In some possible implementations, the processor 301 is specifically configured to determine the target precoding codeword using a sixth formula, where the sixth formula is expressed as:

wherein, representing a target pre-coding code word, n represents a measurement subframe s_mThe sequence number of the previous measurement subframe indicates that the measurement subframe is s_mA corresponding candidate pilot weighting matrix. Correspondingly, as a second pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is a unit weighting matrix, the processor 301 determines a target precoding code word from precoding code words corresponding to N UEs according to a sixth formula, and uses an alternative pilot weighting matrix corresponding to the target precoding code word as the second pilot weighting matrix, where the method is used to measure a second pilot weighting matrix determined under a subframe, the second pilot weighting matrix is used to weight and send a next pilot signal, and the second pilot weighting matrix and the target precoding code word are used to weight and send a next data signal, so as to effectively improve the accuracy of a downlink channel.

In practical applications, the second pilot weighting matrix may also be determined according to the first pilot weighting matrix, and several possible ways are described as follows:

in some possible implementations, if the first pilot weighting matrix is a preset pilot weighting matrix, the processor 301 is specifically configured to obtain a target weight matrix of one UE if M is 1; and determining the second pilot frequency weighting matrix according to the first pilot frequency weighting matrix and the target weight matrix of the UE.

In some possible implementations, the processor 301 is specifically configured to determine whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, where the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁A PMI codebook representing rank 1, a conjugate transpose matrix represented, Qmod (m, L) represents a first pilot weighting matrix with a measurement subframe number m in L measurement subframes, and δ represents a constraint threshold value; determining the first pilot weighting matrix satisfying the seventh formula as the second pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is the preset pilot weighting matrix and the scheduled user is a single user, the processor 301 first obtains the target weight matrix of the UE, and determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the UE, where the specific determination method is as in the seventh formula, where the method is used to measure the second pilot weighting matrix determined under the subframe, and the second pilot weighting matrix is used to perform weighted transmission on the next pilot signal, so as to effectively improve the accuracy of the downlink channel.

In some possible implementations, the processor 301 is further configured to determine a target precoding code word by using an eighth formula after determining the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix, where the eighth formula is represented as:

wherein the target precoding codeword is represented.

In some possible implementations, the processor 301 is further configured to determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, and then transmit a first weighted data signal to the N UEs, where the first weighted data signal is obtained by weighting, by the network device, the first data signal according to the target precoding codeword and the second pilot weighting matrix.

It can be seen that, if the first pilot weighting matrix is the preset pilot weighting matrix, after the processor 301 determines the second pilot weighting matrix first, the processor determines the target precoding code word according to the eighth formula, where the second pilot weighting matrix and the target precoding code word are used for performing weighted transmission on the next data signal, so as to effectively improve the accuracy of the downlink channel.

Different from the prior art, the I/O device 304 receives N PMIs, where N is an integer greater than 1, the N PMIs are determined by N UEs according to a first weighted pilot signal sent by the I/O device 304, the first weighted pilot signal is obtained by the processor 301 by weighting the first pilot signal according to a first pilot weighting matrix, when the processor 301 weights the first pilot signal, the UE really sees the first weighted pilot signal and does not see the first pilot signal, and does not see the PMI according to the first pilot signal, and the UE really sees the first weighted pilot signal and determines the PMI after performing downlink channel estimation by the first weighted pilot signal, so as to expand the selection range of PMIs, because the N PMIs not the quantization of the real downlink channel, but the quantization of the downlink channel after weighting the first pilot signal by the first pilot weighting matrix, therefore, the processor 301 needs to perform corresponding inverse transformation to correctly restore the real downlink channel, and then the processor 301 determines N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix, determines M scheduled UEs from the N UEs, and then determines a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, so that a second weighted pilot signal can be obtained by weighting the second pilot signal by the second weighting matrix, and the second weighted pilot signal is sent to the N UEs by the I/O device 304, thereby reducing quantization errors of the downlink channel, better suppressing interference between users, and improving the accuracy of the downlink channel.

In an embodiment of the present invention, the processor 301 may also be referred to as a Central Processing Unit (CPU). Memory 302 may include both read-only memory and random-access memory, and provides instructions and data to processor 301. A portion of the memory 302 may also include non-volatile random access memory (NVRAM). In a particular application, the various components of network device 300 are coupled together by a bus system 303, where bus system 303 may include a power bus, a control bus, a status signal bus, and the like, in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 303 in FIG. 3.

The method disclosed in the above embodiments of the present invention may be applied to the processor 301, or implemented by the processor 301. The processor 301 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware or instructions in the form of software in the processor 301. The processor 301 may be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in the memory 302, and the processor 301 reads the information in the memory 302 and completes the steps of the method in combination with the hardware.

Referring to fig. 4, an embodiment of a downlink transmission method in the embodiment of the present invention is schematically illustrated, where the downlink transmission method is applied to a MIMO system, and the specific flow is as follows:

step 401, the network device determines a first pilot signal for downlink channel estimation.

In this embodiment of the present invention, the first pilot signal is used for the UE to perform channel estimation on the downlink channel, so that the network device first determines the first pilot signal, generally, the first pilot signal is a single frequency, and the network device randomly selects a part of the communication signal as the first pilot signal or determines the first pilot signal according to a preset rule, for example: the network equipment side has a pilot signal specially used for downlink channel estimation.

Step 402, the network device determines a first pilot weighting matrix according to the first pilot signal.

In contrast to the prior art, the network device does not directly send the first pilot signal to the UE, but first determines a first pilot weighting matrix according to the first pilot signal, where the first pilot weighting matrix is a precursor pilot weighting matrix for weighting the first pilot signal. In practical applications, the first pilot weighting matrix may be an identity matrix or a unitary matrix preset for the first pilot signal, and since the pilot signal is not weighted by using the pilot weighting matrix in the prior art, the first pilot weighting matrix used when the pilot signal is weighted for the first time may be at least one of the identity matrix and a preset pilot weighting matrix, and after the first pilot signal is weighted by using the first pilot weighting matrix for the first time, the next pilot weighting matrix is determined as the weighting for the next pilot signal, which is not limited herein.

Step 403, the network device weights the first pilot signal according to the first pilot weighting matrix to obtain a first weighted pilot signal, and sends the first weighted pilot signal to the N UEs.

The first weighted pilot signal is used for the N UEs to perform downlink channel estimation, and determines N PMIs according to a result of the downlink channel estimation, and then sends the determined N PMIs to the network device, where N is an integer greater than 1.

The network equipment weights the first pilot signal according to the first pilot weighting matrix to obtain a first weighted pilot signal, and the N UEs perform channel estimation on the first weighted pilot signal according to the existing channel estimation method, so that the effect is that the downlink channel seen by the UE is transformed by the first pilot weighting matrix, the UE really sees the first pilot signal but not determines the PMI through the first pilot signal, and the UE really sees the first weighted pilot signal but determines the PMI through the downlink channel estimation of the first weighted pilot signal, thereby expanding the selection range of the PMI.

Step 404, the N UEs determine N PMIs according to the first weighted pilot signal.

And the N UEs perform downlink channel estimation according to the received first weighted pilot signal and determine the corresponding PMI according to the result of the downlink channel estimation, wherein in order to improve the accuracy of the PMI, each UE selects the PMI corresponding to the precoding code word with the optimal downlink channel quality according to the received first weighted pilot signal.

Step 405, the N UEs send the N PMIs to the network device.

Since the N PMIs are used by the network device to determine the N reconstructed channel principal eigenvectors, the N UEs send the respective determined PMIs to the network device.

Step 406, the network device receives the N PMIs sent by the N UEs.

The network device receives the N PMIs sent by the N UEs, and may first locally cache the N PMIs, so as to obtain the N PMIs in time during subsequent use.

Step 407, the network device determines N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix.

Different from the prior art, after the first pilot signal is weighted, the N PMIs determined by the N UEs are not quantization of the true downlink channel but quantization of the downlink channel after transformation by the first pilot weighting matrix, so that the true downlink channel can be correctly restored by performing corresponding inverse transformation on the network device side.

In some possible implementation manners, after obtaining the N PMIs and the first pilot weighting matrix, the network device determines, according to the N PMIs, N precoding codewords corresponding to the N PMIs, then the network device determines a conjugate transpose matrix corresponding to the N precoding codewords and a conjugate transpose matrix corresponding to the first pilot weighting matrix, and determines, according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix, the N reconstructed channel main eigenvectors.

Because a corresponding relationship exists between the PMI and the precoding code word, the network device firstly determines corresponding N precoding code words according to the N PMIs, and determines a conjugate transpose matrix corresponding to the N precoding code words and a conjugate transpose matrix corresponding to the first pilot weighting matrix in order to keep the pilot power after the first pilot signal weighting unchanged and keep the correlation of the N PMI code words before and after rotation unchanged, and restrict the first pilot weighting matrix to be a unitary matrix.

In practical applications, there are many ways for the network device to determine the N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix, where one of the ways is as follows:

the network device determines N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is expressed as:

Step 408, the network device determines M scheduled UEs from the N UEs.

Wherein M is an integer not greater than N. In practical application, the quality of the downlink channel corresponding to each UE is not good, and therefore, in order to improve the accuracy of the downlink channel, the network device determines the scheduled UE according to the N reconstructed channel principal eigenvectors and the N CQIs of the N UEs, where the CQIs of the N UEs are stored locally in the network device, and therefore, the network device may directly obtain the CQIs of the N UEs locally.

Step 409, the network device determines a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs.

Before the network device determines the second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, the network device generally determines a candidate pilot weighting matrix set, and selects a certain candidate pilot weighting matrix from the candidate pilot weighting matrix set as the second pilot weighting matrix according to a preset rule, wherein there are many ways to determine the candidate pilot weighting matrix set, and the following ways are possible to implement:

in some possible implementation manners, if the first pilot weighting matrix is a unit weighting matrix, and if M is 1, the network device determines an alternative pilot weighting matrix set according to a reconstructed channel main eigenvector corresponding to one UE and a precoding codeword corresponding to the one UE.

In practical application, the network device may determine the second pilot weighting matrix according to a reconstructed channel dominant eigenvector corresponding to one UE and a precoding codeword corresponding to the first UE by using a second formula, where the second formula is represented as:

It can be seen that, if the scheduled user is a single user and the first pilot weighting matrix is a unit weighting matrix, the network device determines the candidate pilot weighting matrix set according to the second formula directly according to the reconstructed channel dominant eigenvector corresponding to one UE and the pre-coding codeword corresponding to one UE, and the calculation amount is small.

In other possible implementation manners, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, and if M is 2, the network device determines a target weight matrix according to the main eigenvectors of the reconstructed channel corresponding to the two UEs respectively;

and when the correlation of the main characteristic vectors of the reconstructed channels corresponding to the two UEs is smaller than a first preset threshold, the network equipment determines the candidate pilot weighting matrix set according to the pre-coding code words corresponding to the two UEs and the target weight matrix.

In practical application, the network device may determine the target weight matrix according to the reconstructed channel dominant eigenvectors corresponding to the two UEs respectively by using a third formula, where the third formula is expressed as:

The network device determines a candidate pilot weighting matrix set by using a fourth formula according to the precoding code words corresponding to the two UEs and the target weight matrix, wherein the fourth formula is represented as:

In other possible implementation manners, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, and if M is greater than or equal to 2, the network device determines a target weight matrix according to the reconstructed channel main eigenvectors corresponding to the at least two UEs, respectively;

and when the correlation of the target weight matrixes corresponding to any two pieces of UE is smaller than a second preset threshold value, the network equipment determines an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the at least two pieces of UE and the target weight matrixes.

In practical application, the network device may determine the candidate pilot weighting matrix set by using a fifth formula according to the precoding code words corresponding to the at least two UEs and the target weight matrix, where the fifth formula is represented as:

It can be seen that, if at least two users are scheduled, and if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, the network device first determines the target weight matrices corresponding to the at least two UEs, where the specific determination manner is as the third formula, and generally, the power corresponding to the target weight matrix corresponding to each UE is the same. When the correlation between the target weight matrices corresponding to any two UEs in the at least two UEs is smaller than a second preset threshold, the network device determines the candidate pilot weighting matrix set according to the target weight matrix and the pre-coding code words corresponding to the at least two UEs, where the specific determination manner is, for example, a fifth formula, and in practical application, the second preset threshold is determined by the network device according to an actual situation, for example, the second preset threshold is 1, and is not specifically limited here.

In the embodiment of the present invention, after determining the candidate pilot weighting matrix set, a certain candidate pilot weighting matrix may be selected from the pilot weighting matrix set to determine as the second pilot weighting matrix, where there are many ways to determine the second pilot weighting matrix, and the following are several ways that may be implemented:

in some possible implementation manners, if the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix, the network device first randomly selects a target pre-coding codeword, and selects an alternative pilot weighting matrix corresponding to the target pre-coding codeword from the alternative pilot weighting matrix set according to the target pre-coding codeword, and then the network device determines the alternative pilot weighting matrix corresponding to the target pre-coding codeword as the second pilot weighting matrix.

In other possible implementation manners, if the first pilot weighting matrix is a unit weighting matrix, the network device selects a precoding code word from N corresponding N precoding code words according to a preset rule to determine as a target precoding code word, selects an alternative pilot weighting matrix corresponding to the target precoding code word from an alternative pilot weighting matrix set according to the target precoding code word, and then determines the alternative pilot weighting matrix corresponding to the target precoding code word as the second pilot weighting matrix.

In practical applications, the network device may determine the target precoding codeword by using a sixth formula, where the sixth formula is expressed as:

In practical applications, the network device may further determine the second pilot weighting matrix according to the first pilot weighting matrix, where there are many ways to determine the second pilot weighting matrix, and the following are several ways that may be implemented:

in some possible implementation manners, if the first pilot weighting matrix is a preset pilot weighting matrix, and if M is 1, the network device obtains a target weight matrix of one UE, and then determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the one UE.

In practical application, the network device determines whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, where the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁PMI codebook representing rank 1, conjugate transpose matrix represented, Qmod (m, L) representing first pilot weighting with measurement subframe number m among L measurement subframesMatrix, δ represents a constraint threshold;

In some possible implementation manners, the network device further needs to determine a target precoding codeword, and weights a next data signal according to the target precoding codeword and the second pilot weighting matrix to improve the accuracy of the downlink channel, and since the network device may determine the second pilot weighting matrix first and does not determine the target precoding codeword, the network device determines the target precoding codeword by using an eighth formula, where the eighth formula is expressed as:

wherein the target precoding codeword is represented.

It can be seen that, if the first pilot weighting matrix is a preset pilot weighting matrix, after the network device determines the target pilot weighting matrix first, the network device determines the target precoding code word according to the eighth formula, where the second pilot weighting matrix and the target precoding code word are used for performing weighted transmission on a next data signal, so as to effectively improve the accuracy of the downlink channel.

Step 410, the network device weights the second pilot signal according to the second pilot weighting matrix to obtain a second weighted pilot signal.

Step 411, the network device sends the second weighted pilot signal to the N UEs.

After determining the second pilot weighting matrix, the network device weights the second pilot signal according to the second pilot weighting matrix, so as to obtain a second weighted pilot signal, and sends the second weighted pilot signal to the N UEs, thereby improving the accuracy of the downlink channel.

The embodiment shown in fig. 4 describes a process in which the network device first determines a second pilot weighting matrix, and then weights a second pilot signal according to the second pilot weighting matrix to obtain a second weighted pilot signal, and in some possible embodiments, the network device weights a first data signal according to a target pre-coding code word and the second pilot weighting matrix to obtain a first weighted data signal, where the first data signal includes control information and the like, and sends the first weighted data signal to N UEs to improve the accuracy of a downlink channel, so that the TM4 in the MIMO system is also suitable for downlink transmission of MUs, and signal interference between MUs is effectively reduced.

The downlink transmission method in the embodiment of the present invention is described below with reference to a specific example, and as shown in fig. 5, an application scenario embodiment of the downlink transmission method in the embodiment of the present invention is schematically illustrated, and a specific flow includes:

in practical application, taking a network device as a base station, the first pilot signal and the second pilot signal are both CRS as an example:

the base station determines a CRS used for downlink channel estimation, generates a first pilot weighting matrix according to the CRS, and caches the generated first pilot weighting matrix if the first pilot weighting matrix is a unit weighting matrix, wherein the first pilot weighting matrix is used for weighting the CRS.

The base station performs weighting processing on the CRS according to the first pilot weighting matrix to obtain a weighted pilot signal, that is: the first weighted pilot signal is sent to N UEs, wherein N is an integer greater than 1, so that the N UEs perform downlink channel estimation according to the first weighted pilot signal, the N UEs determine N PMIs and N CQIs of corresponding downlink channels according to the result of the downlink channel estimation, then the N UEs feed back the N PMIs and the N CQIs to a base station, and the base station performs local caching on the N PMIs and the N CQIs received from the N UEs, so as to be directly utilized later.

The base station determines a first pilot weighting matrix and a second pilot weighting matrixN PMIs received by N UEs are used for channel reconstruction, and because N PMIs determined by N UEs are not the quantization of a real channel but the quantization of a downlink channel after the transformation of a first pilot weighting matrix, the real downlink channel can be correctly restored only by performing corresponding inverse transformation on a base station side, a specific method is that a base station determines N main eigenvectors of the reconstructed channel according to the N PMIs and the first pilot weighting matrix, wherein the specific process of determining the N main eigenvectors of the reconstructed channel is as follows: note UE_uThe downlink channel is H_u(N_R×N_TComplex matrix, N_RFor UE receiving antenna port number, N_TNumber of antenna ports for base station transmission), the first pilot weighting matrix Q used by the base station in sub-frame t_T(N_R×N_TComplex matrix) is transmitted in a weighted manner with respect to the CRS, but the first pilot weighting matrix is constrained to be a unitary matrix in order to keep the pilot power after CRS weighting unchanged and the correlation before and after PMI codeword rotation unchanged. UE (user Equipment)_uTaking the CRS of the subframe t as the channel estimation and selecting rank 1PMI as i_u,tThe corresponding rank-1 precoding code word is recorded as the measurement subframe as s₁,s₂…s_mBase station for each UE_uMaintaining a queue with length L for storing PMI fed back by UE and maintaining a queue Que with length L^WgtFor storing the first pilot weighting matrix used, the queue is from head to tail, every measurement sub-frame s_mThen the first pilot weighting matrix used by the sub-frame t is queued into the queue quee^WgtThe base station receives the UE_uFor measurement subframe s_mIf calculated, enqueue it

Suppose that in subframe t, the measurement subframe corresponding to the last L times of feedback PMIs received by the base station is: s₁,s₂…s_mThen queue Que^WgtWherein the element from beginning to end is Q₁,Q₂…Q_mThe elements from the top to the end of the queue are: the base station determines N reconstructed channel main characteristic vectors (hereinafter abbreviated as reconstructed channels) as follows:

wherein, the main characteristic vector of the reconstructed channel of the user equipment u determined on the subframe t is represented, and the PrimEigVec represents that the matrix is obtainedA main characteristic vector, L represents the total number of the measuring subframes in a preset length, m is the measuring subframe sequence number and represents that the measuring subframe is s_mRepresents the user equipment u to the measurement sub-frame s_mThe selected PMI after the CRS is used for channel estimation indicates that the PMI is a corresponding precoding code word, indicates a corresponding conjugate transpose matrix and indicates a corresponding conjugate transpose matrix.

The specific method comprises the following steps: and (4) obtaining the SVD, and then obtaining a right singular vector corresponding to the maximum singular value of the SVD. Wherein, the setting criterion of the queue length L (i.e. the reconstruction window length) is: the time used by L times of feedback is within the coherence time of the channel, and L is as large as possible. For example: and for the carrier frequency of 2GHz, the mobility of the user is 3km/h according to a coherent time calculation formula, the coherent time is about 76ms, and if the feedback period is 10ms, L is less than or equal to 8. For the sake of conservation, L ═ 5 may be taken, but is not limited to this value, and the calculation of the reconstructed channel is only performed when the base station receives the PMI fed back by the UE. In the subframe between two PMI feedbacks of a UE, the reconstructed channel is equal to the result of the latest channel reconstruction calculation, and the reconstructed channel and the fed-back CQI are used as the input of a downlink scheduling module (used for determining the number of scheduled users).

After the base station determines to reconstruct the channel, the base station obtains the user CQI cached, and the base station determines single-user scheduling or multi-user scheduling according to the CQI and the result of reconstructing the channel, namely determining M scheduled UEs from the N UEs, wherein M is an integer not more than N. And if the scheduling module determines that only one UE is scheduled in the subframe t, taking the reconstructed channel of one UE as a target weight of downlink transmission. Since TM4 only allows weight matrix selection from the precoding codebook, it is necessary to find a PMI codeword W with rank 1_i∈w₁(w₁Is a rank-1 PMI codebook, each codeword being N_TX 1 complex vector, defined in 3GPP TS36.211) such that there is an alternative set of pilot weighting matrices to W_iThe unitary matrix with rotation satisfying the above formula exists and is not unique, and a method for solving the candidate pilot frequency weighting matrix set is as follows:

first, two sets of complex vectors are randomly generated, each set containing (N)_T-1) N_TX 1 complex vector: and wherein the first set of auxiliary vectors is the second setAn auxiliary vector.

Then, the method performs Gram-Schmidt orthogonalization operation (detailed process is as follows) on the signals respectively, thereby forming two sets of standard orthogonal bases (due to W)_iAnd are both unit vectors, themselves in orthonormal basis):

for the unitary matrix composed of the above orthonormal basis, the candidate pilot weighting matrix set is: the Gram-Schmidt orthogonalization method comprises the following steps: a set of N-dimensional vectors { X₁,X₂,X_NThe transformation into a set of standard orthogonal sets

If the scheduling module decides to determine the subset U ═ U by the UE within the subframe t₁,…u_rR users are paired, and a downlink sending weight matrix calculated according to the current reconstruction channels of the users is recorded as (N)_TThe xr complex matrix, which will be abbreviated as Channel Reconstruction-BeamForming (abbreviated as CR-BF) weight hereinafter), each column represents the transmission weight of each paired user, and the weight power of each user is divided equally, that is: since TM4 only allows weight matrix selection from the codebook, a PMI codeword W of rank r is found_i∈w_r(w_rIs a rank-r PMI codebook, each codeword being N_TXr complex matrix, defined in 3GPP TS36.211) such that there is a set of alternative pilot weighting matrices that exactly fit W to W_iRotate to become

The PMI codebook defined by the LTE protocol has the property that each column of each codeword matrix is orthogonal, and since unitary matrix transformation keeps the correlation between vectors constant, each column that must be guaranteed is also orthogonal, and there is a unitary matrix that strictly satisfies the above formula. However, in practical application, it cannot be guaranteed that weights of multiple scheduled users are strictly orthogonal, so the present invention proposes the following alternative pilot weighting matrix set solving method based on scheduling constraint:

first, add a constraint on the existing user pairing rules: the correlation between the CR-BF weights of any two paired users must be lower than a preset threshold z e (0,1), i.e.:

for example: the value of z is 0.1, and the actual value is not limited to this value.

In some possible implementation manners, in a scenario where at most 2 users are paired, it is directly determined whether a pairing requirement is met according to a reconstruction channel of each user, and it is assumed that u is when₁And u₂During pairing, the sending weight is calculated according to the reconstruction channels of the two users:

The diagonal matrix has the effect that the weight vectors of the weight matrix are both 0.5, and the weight vector sent by two users is 1, namely: the correlation between the weights of two users is equal to the correlation between their reconstructed channels, since the weights have the following important properties:

therefore, when two users are paired, whether the pairing condition is met is directly judged according to the reconstruction channels of the two users, and therefore the calculation complexity in the user scheduling process is reduced.

Then, two sets of complex vectors are randomly generated, each set containing N_T-r N_TX 1 complex vector: and

next, note W_i＝[W_i，1，…，W_i，r]For the orthogonalization operation performed with Gram-Schmidt, respectively, two sets of orthonormal bases are constructed:

finally, the unitary matrix formed by the standard orthogonal base determines the candidate pilot frequency weighting matrix set as follows:

in the above procedure, the calculation of the set of alternative pilot weighting matrices depends on the selected PMI. Precoding words W corresponding to different PMIs_i(rank 1 or rank)>1) The corresponding candidate pilot weighting matrix set can be obtained, but the influence of the first pilot weighting matrix on the channel reconstruction performance is not considered in the process. In principle, it is desirable that the first pilot weighting matrix be rotated as uniformly as possible within the channel reconstruction windowThe channel can make the PMI quantization error be evenly distributed, and can reduce the quantization error through the reconstruction operation, that is, the second pilot weighting matrix is selected from the candidate pilot weighting matrix set to be used as the weighted transmission of the next CRS, and the target precoding code word corresponding to the second pilot weighting matrix and the second pilot weighting matrix are used as the weighted transmission of the next data signal. Therefore, the present invention provides a method for selecting a second pilot weighting matrix as follows:

the first method comprises the following steps: in the initial state of the subframe t, the subframe s is measured for the downlink₀Recording the set of scheduled users as U₀The number of scheduling users is r₀The CR-BF weight is given by₀Randomly selecting a pre-coding code word in the pre-coding codebook to calculate as a sub-frame s₀The second pilot weighting matrix of the CRS is:

and the second method comprises the following steps: measuring subframe s in downlink_mM is 1,2, …, and the set of scheduled users is U_mThe number of scheduling users is r_mThe CR-BF weight is given by_mThe target precoding codeword is selected according to the following formula in the precoding codebook:

the meaning of this formula is: in the measurement subframe s_mCalculating all possible precoding words W_iThe corresponding rotation matrices are then compared one by one with the determined pilot weighting matrices of the preceding measurement subframes, the precoding codeword that maximizes the minimum distance is selected for subframe s_mAs sub-frame s, the corresponding candidate pilot weighting matrix_mThe second pilot weighting matrix of the CRS is:

where dist (a, B) is defined as two unitary matrices a ═ a₁,a₂,…a_N]And B ═ B₁,b₂,…b_N]The distance between the unitary matrix and the rotation matrix shows the transformation effect of the unitary matrix on the channel, and the larger the distance is, the more uniform the rotation on the channel is. One specific definition that may be employed is:

and the third is that: in a non-measurement subframe s, recording a CR-BF weight calculated according to a subframe scheduling result as W selected in a precoding codebook in an arbitrary mode_i∈w_r(r is the number of scheduled users in sub-frame s), calculating the number of users referred to hereinIncluding random selection, or the same selection as the measurement subframe, or keeping the same PMI as used in the previous measurement subframe.

Therefore, the accuracy of the downlink channel is improved by performing weighting processing and channel reconstruction on the CRS. And obtaining an alternative pilot weighting matrix set through calculation, and selecting a certain pre-coding code word from the pre-coding code words corresponding to the alternative pilot weighting matrix set to rotate to a BF weight corresponding to SU or MU, so that the pre-coding code word based on the codebook can realize the effect of BF which is not based on the codebook. And the second pilot frequency weighting matrix of the most uniform rotation channel is selected from the alternative pilot frequency weighting matrices, so that the accuracy of channel reconstruction is improved to the maximum extent.

As shown in fig. 6, another application scenario embodiment of the downlink transmission method in the embodiment of the present invention is schematically illustrated, and the specific flow includes:

unlike the embodiment shown in fig. 5, in the embodiment of fig. 6, the base station sets a first pilot weighting matrix in advance and weights the first pilot signals according to the preset first pilot weighting matrix, wherein the first pilot weighting matrix used in the measurement sub-frame is set off-line, and a group of L N pilot weighting matrices is preset_T×N_TSet of unitary matrix components, denoted as p ═ Q₀,Q₁,…,Q_L-1L is equal to the reconstruction window length, see the description in fig. 5. Specific first pilot weighting matrix forms include, but are not limited to, the following three:

the first is the MUB weighting matrix:

for a d-dimensional complex vector space, its two sets of orthonormal bases E ═ E₀,e₁,…,e_d-1F ═ F₀,f₁,…,f_d-1Called "Bases Unbiased with respect to each other" (overall in english: Mutually unabridged Bases, abbreviation: MUB), if and only if the modulo square of the inner product between any basis vectors within two groups of Bases is equal to the reciprocal of the spatial dimension:

it has been theoretically demonstrated that: if the space dimension d is exactly an integer power of a certain prime number, then (d +1) sets of orthonormal bases must be found to constitute the MUBs of the space.

In the current LTE system design, the number of antenna ports N of the base station_TAlways an integer power of 2, so that always a containment can be constructed (N)_T+1) MUB of the set of orthonormal bases:

l＝0,1,…,N_Tarranging the base vectors in each set of orthonormal base into a matrix according to columns to obtain (N)_T+1) unitary matrices:

in general, the reconstruction window length is set to L ≦ (N)_T+1), so the set Q of the first L matrices is taken as { Q ═ Q₀,Q₁,…,Q_L-1As the first set of pilot weighting matrices.

The second is a Kerdock weighting matrix:

for a base station with 2 antenna ports or 4 antenna ports, a simplified MUB weighting matrix exists, and the characteristic is that each element in the matrix only takes values in {0, +/-1, +/-j }. The characteristic is beneficial to reducing the storage overhead of the off-line weighting matrix and the operation overhead during weighting.

The first set of pilot weighting matrices for a base station with 2 antenna ports is Q ═ Q₀,Q₁,Q₂And (c) the step of (c) in which,

the first set of pilot weighting matrices for a base station with 4 antenna ports is Q ═ Q₀,Q₁,Q₂,Q₃,Q₄And (c) the step of (c) in which,

the third is a phase rotation weighting matrix:

currently, cross polarization antennas are mostly adopted in a base station of LTE, that is, an antenna array can be divided into two groups of polarization directions, and the number of antenna ports is generally 0 to (N)_T/2-1) to antennas of one polarization direction, the number of antenna port N_T/2～(N_T-1) an antenna assigned to the other polarization direction.

When the base station is a 2-antenna port, the first pilot weighting matrix is of the following form:

an example of a rotational phase value is L-3, θ₀＝0,θ₁＝π/2,θ₂Pi/4, but not necessarily limited thereto, θ_lRepresenting the rotation angle between the two polarization directions.

When the base station is a 4-antenna port, the first pilot weighting matrix is of the following form:

where M is 0,1, …, M-1, N is 0,1, …, N-1 are indices of two groups of phase rotation angles, M × N is L, and the total index L is M + (N-1) M. Indicating the rotation angle between adjacent antennas in the same polarization direction. An example of a rotational phase value is:

but is not necessarily limited thereto.

In practical application, no matter what kind of off-line setting is used for the first pilot weighting matrix set, each matrix in the set is used for weighting the first pilot signals in the measurement subframe cycle. Suppose the measurement subframe is s₀,s₁,s₂…, in sub-frame s_mUsing a first pilot weighting matrix Q_lAnd L is mod (m, L), where mod (m, L) denotes the integer m times the integer L.

And only SU scheduling is carried out in the measurement subframe. The reasons for not doing MU are: the preset first pilot weighting matrix can not ensure that the sending weight is consistent with the pre-coding code word, the demodulation performance of the UE is lost, the robustness of the SU is strong, the influence of the demodulation loss on the user rate is small, and the user rate loss is possibly large if the MU is adopted.

And adding a constraint on the basis of the existing SU scheduling criterion, wherein the requirement is that a pre-coding matrix can be found for a scheduled user, and the correlation between the pre-coding matrix and the CR-BF weight is not lower than a certain threshold after the pre-coding matrix is rotated by the first pilot weighting matrix. In particular, in the measurement subframe s_mAssuming that the CR-BF weight of UEu is Qmod (m, L) as the first pilot weighting matrix used, the added scheduling constraint is:

the value range of the threshold is delta epsilon (0,1), and the larger the value range is, the stricter the constraint is. An example is δ — 0.8, but is not limited thereto. It can be seen that the first pilot weighting matrix satisfying the above formula is used as the second pilot weighting matrix, where the second pilot weighting matrix is used for weighting and transmitting the next pilot signal.

Users not meeting the constraint cannot be scheduled in the current subframe, and if the constraint is met, the selected precoding matrix is as follows: wherein the target precoding codeword is represented. And after the target pre-coding code word is determined, the target pre-coding code word and the second pilot frequency weighting matrix are used for carrying out weighting sending on the next data signal.

MU scheduling is performed on a non-measured subframe, and scheduling criterion modification, CR-BF weight calculation, second pilot weighting matrix calculation and PMI selection are similar to those of the embodiment shown in fig. 5, except that the influence of the first pilot weighting matrix on channel reconstruction does not need to be considered on the non-measured subframe, so that PMI can be selected arbitrarily.

To facilitate a better implementation of the above-described related methods of embodiments of the present invention, a network device for cooperating with the above-described methods is also provided below.

Referring to fig. 7, a schematic structural diagram of a network device 700 according to an embodiment of the present invention, where the network device is a device in a mimo system, the network device 700 includes: a receiving module 701, a first determining module 702, a second determining module 703, a third determining module 704 and a sending module 705.

A receiving module 701, configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, where the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by the network device, and the first weighted pilot signal is obtained by weighting, by the network device, a first pilot signal according to a first pilot weighting matrix;

a first determining module 702, configured to determine N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix received by the receiving module 701;

a second determining module 703 for determining M scheduled UEs from the N UEs;

a third determining module 704, configured to determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs determined by the second determining module 703, where M is an integer no greater than N;

a sending module 705, configured to send the second weighted pilot signals determined by the third determining module 704 to the N UEs, where the second weighted pilot signals are obtained by the network device through weighting second pilot signals according to the second pilot weighting matrix.

Different from the prior art, the receiving module 701 receives N PMIs, where N is an integer greater than 1, where the N PMIs determined by N UEs according to a first weighted pilot signal sent by the network device, where the first weighted pilot signal is obtained by the network device by weighting a first pilot signal according to the first pilot weighting matrix, and after the network device weights the first pilot signal, the UE really sees the first weighted pilot signal and does not determine the PMI according to the first pilot signal, and the UE really sees the first weighted pilot signal and determines the PMI after performing downlink channel estimation according to the first weighted pilot signal, so as to expand the selection range of PMIs, because the N PMIs not the quantization of a real downlink channel, but the quantization of the downlink channel after weighting the first pilot signal according to the first pilot weighting matrix, therefore, the network device needs to perform corresponding inverse transformation to correctly restore the real downlink channel, and then the network device determines N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix, the second determining module 703 determines M scheduled UEs from the N UEs, then the third determining module 704 determines the second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs determined by the second determining module 703, so that the second pilot signal can be weighted by the second weighting matrix to obtain a second weighted pilot signal, and the sending module 705 sends the second weighted pilot signal to the N UEs, thereby reducing quantization errors of the downlink channel, better suppressing interference between users, and improving the accuracy of the downlink channel.

Referring to fig. 8, another structural diagram of a network device 700 is shown, where the network device 700 includes: a receiving module 701, a first determining module 702, a second determining module 703, a third determining module 704, a sending module 705 and a fourth determining module 706.

A receiving module 701, configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, where the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by the network device, and the first weighted pilot signal is obtained by weighting, by the network device, a first pilot signal according to a first pilot weighting matrix.

The first pilot weighting matrix is a target pilot weighting matrix determined by previous pilot weighting, and is used for weighting the pilot signal at this time, because the pilot signal is not weighted in the prior art, the first pilot weighting matrix for weighting the pilot signal at the first time is at least one of a unit weighting matrix and a preset pilot weighting matrix, the target pilot weighting matrix determined after weighting the pilot signal at the second time is used as a first pilot weighting matrix for the third time, and the pilot signal at the third time is weighted according to the first pilot weighting matrix, and so on, which is not specifically limited herein.

A first determining module 702, configured to determine N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix received by the receiving module 701.

There are many ways for the first determining module 702 to determine the N reconstructed channel main eigenvectors according to the N PMIs and the first pilot weighting matrix, and one possible way includes:

the first determining module 702 is specifically configured to determine, according to the N PMIs, N precoding codewords corresponding to the N PMIs; determining a conjugate transpose matrix corresponding to the N precoding codewords and a conjugate transpose matrix corresponding to the first pilot weighting matrix; and determining the N reconstructed channel main eigenvectors according to the N pre-coding code words, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N pre-coding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix.

In some possible implementations, the first determining module 702 is specifically configured to determine the N reconstructed channel main eigenvectors according to the N precoding code words, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is represented as:

A second determining module 703 is configured to determine M scheduled UEs from the N UEs.

A third determining module 704, configured to determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs determined by the second determining module 703, where M is an integer no greater than N.

In some possible implementation manners, before the third determining module 704 determines the second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, a candidate pilot weighting matrix set is determined, and then a certain candidate pilot weighting matrix is selected from the candidate pilot weighting matrix set as the second pilot weighting matrix according to a certain manner, where the specific implementation process includes the following possible manners:

the first possible way is: a fourth determining module 706, configured to, if the first pilot weighting matrix is the unit weighting matrix, before the third determining module 704 determines the second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs, if M is 1, determine an alternative pilot weighting matrix set according to the reconstructed channel main eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE.

In some possible implementations, the fourth determining module 706 is specifically configured to determine, according to a reconstructed channel dominant eigenvector corresponding to one UE and a precoding codeword corresponding to the one UE, a candidate pilot weighting matrix set by using a second formula, where the second formula is represented as:

A second possible way is: a fourth determining module 706, configured to, if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, before the third determining module 704 determines a second pilot weighting matrix according to a reconstructed channel main eigenvector of each UE of the M scheduled UEs, if M is 2, determine a target weight matrix according to the reconstructed channel main eigenvectors respectively corresponding to the two UEs; and when the correlation of the main characteristic vectors of the reconstructed channels corresponding to the two UEs is smaller than a first preset threshold, determining a candidate pilot weighting matrix set according to the precoding code words corresponding to the two UEs and the target weight matrix.

In some possible implementation manners, the fourth determining module 706 is specifically configured to determine the target weight matrix according to the primary eigenvectors of the reconstructed channel corresponding to the two UEs by using a third formula, where the third formula is represented as:

In some possible implementation manners, the fourth determining module 706 is specifically configured to determine, according to the precoding code words and the target weight matrix corresponding to the two UEs, a candidate pilot weighting matrix set by using a fourth formula, where the fourth formula is represented as:

A third possible way is: a fourth determining module 706, configured to determine, if the first pilot weighting matrix is the unit weighting matrix and the preset pilot weighting matrix, a target weight matrix according to the reconstructed channel main eigenvector corresponding to each of the at least two UEs if M is greater than or equal to 2 before the third determining module determines the second pilot weighting matrix according to the reconstructed channel main eigenvector of each of the M scheduled UEs; and when the correlation of the target weight matrixes corresponding to any two UEs is smaller than a second preset threshold, determining an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the at least two UEs and the target weight matrixes.

In some possible implementations, the fourth determining module 706 is specifically configured to determine, according to the precoding code words corresponding to the at least two UEs and the target weight matrix, a candidate pilot weighting matrix set by using a fifth formula, where the fifth formula is represented as:

After the fourth determining module 706 determines the candidate pilot weighting matrix set, a candidate pilot weighting matrix is selected from the candidate pilot weighting matrix set according to a preset manner as the second pilot weighting matrix, and the specific implementation process includes the following possible manners:

the first possible way is: if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, the third determining module 704 is specifically configured to randomly select a target pre-coding codeword, and select an alternative pilot weighting matrix corresponding to the target pre-coding codeword from the alternative pilot weighting matrix set according to the target pre-coding matrix; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.

A second possible way is: if the first pilot weighting matrix is the unit weighting matrix, the third determining module 704 is specifically configured to determine a target precoding code word according to a preset rule, and select an alternative pilot weighting matrix corresponding to the target precoding code word from the alternative pilot weighting matrix set according to the target precoding matrix; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.

In some possible implementations, the third determining module 707 is specifically configured to determine the target precoding codeword by using a sixth formula, where the sixth formula is expressed as:

the first possible way is: if the first pilot weighting matrix is the preset pilot weighting matrix, the third determining module 704 is specifically configured to obtain a target weight matrix of the UE if M is 1; and determining the second pilot frequency weighting matrix according to the first pilot frequency weighting matrix and the target weight matrix of the UE.

In some possible implementations, the third determining module 704 is specifically configured to determine whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, where the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁A PMI codebook representing rank 1, a conjugate transpose matrix represented, Qmod (m, L) represents a first pilot weighting matrix with a measurement subframe number m in L measurement subframes, and δ represents a constraint threshold value; determining a first pilot weighting matrix satisfying the seventh formula as the second pilot weighting matrix.

In some possible implementation manners, the network device further needs to determine a target precoding codeword, and weight the next data signal according to the target precoding codeword and the second pilot weighting matrix to improve the accuracy of the downlink channel, because the network device may determine the second pilot weighting matrix first and does not determine the target precoding codeword, the fourth determining module 706 is further configured to determine the target precoding codeword by using an eighth formula after determining the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix, where the eighth formula is represented as:

wherein the target precoding codeword is represented.

In some possible implementations, the sending module 705 is further configured to, after the third determining module 704 determines a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, send a first weighted data signal to the N UEs, where the first weighted data signal is obtained by weighting, by the network device, a first data signal according to the target precoding codeword and the second pilot weighting matrix.

In the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

The above detailed description is provided for the operation method and terminal of the application program group provided by the present invention, and the principle and implementation of the present invention are explained in this document by applying specific examples, and the description of the above embodiments is only used to help understanding the method and core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.

Claims

A downlink transmission method is applied to a multiple-input multiple-output system, and the method comprises the following steps:

the network equipment receives N precoding matrix indicator PMIs, wherein N is an integer greater than 1, the N PMIs are determined by N user equipment UE according to a first weighted pilot signal sent by the network equipment, and the first weighted pilot signal is obtained by weighting the first pilot signal by the network equipment according to a first pilot weighting matrix;

the network equipment determines N reconstructed channel main characteristic vectors according to the N PMIs and the first pilot weighting matrix;

the network equipment determines M scheduled UEs from the N UEs, and determines a second pilot weighting matrix according to the main characteristic vector of the reconstructed channel of each UE of the M scheduled UEs, wherein M is an integer not greater than N;

and the network equipment sends a second weighted pilot signal to the N UEs, wherein the second weighted pilot signal is obtained by weighting a second pilot signal by the network equipment according to the second pilot weighting matrix.
The method of claim 1,

the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix.
The method of claim 1, wherein the network device determining N reconstructed channel primary eigenvectors from the N PMIs and the first pilot weighting matrix comprises:

the network equipment determines N precoding code words corresponding to the N PMIs according to the N PMIs;

the network device determines a conjugate transpose matrix corresponding to the N precoding codewords and a conjugate transpose matrix corresponding to the first pilot weighting matrix;

and the network equipment determines the N reconstructed channel main eigenvectors according to the N pre-coding code words, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N pre-coding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix.
The method of claim 3, wherein the network device determining the N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix comprises:

the network device determines the N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is expressed as:

the method comprises the steps of representing a reconstructed channel main characteristic vector of user equipment u determined on a subframe t, PrimEigVec representing a main characteristic vector of an acquisition matrix, L representing the total number of measurement subframes within a preset length, m being a measurement subframe sequence number and representing that the measurement subframe is s_mRepresents the user equipment u to the measurement sub-frame s_mThe selected PMI after the channel estimation of the first pilot signal represents that the PMI is a corresponding precoding code word, represents a corresponding conjugate transpose matrix and represents a pairThe corresponding conjugate transpose matrix.
The method of claim 2, wherein before determining a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each of the M scheduled UEs if the first pilot weighting matrix is the unity weighting matrix, the method further comprises:

and if the M is 1, the network equipment determines an alternative pilot frequency weighting matrix set according to a reconstructed channel main characteristic vector corresponding to one UE and a precoding code word corresponding to the UE.
The method of claim 5, wherein the network device determining the candidate pilot weighting matrix set according to the reconstructed channel dominant eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE comprises:

the network equipment determines an alternative pilot weighting matrix set by using a second formula according to a reconstructed channel main eigenvector corresponding to one UE and a precoding code word corresponding to the one UE, wherein the second formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iA precoding codeword with PMI i indicating rank 1 represents a reconstructed channel principal eigenvector corresponding to one UE, and w₁A PMI codebook representing rank 1.
The method of claim 2, wherein before determining a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each of the M scheduled UEs if the first pilot weighting matrix is at least one of the unity weighting matrix and the preset pilot weighting matrix, the method further comprises:

if M is 2, the network equipment determines a target weight matrix according to the main eigenvectors of the reconstructed channels corresponding to the two UEs respectively;

and when the correlation of the main characteristic vectors of the reconstructed channels corresponding to the two UEs is smaller than a first preset threshold, the network equipment determines an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the two UEs and the target weight matrix.
The method of claim 7, wherein the network device determining the target weight matrix according to the primary eigenvectors of the reconstructed channel corresponding to the two UEs respectively comprises:

the network equipment determines the target weight matrix by using a third formula according to the main eigenvectors of the reconstructed channel corresponding to the two UEs respectively, wherein the third formula is expressed as:

wherein, represents the target weight matrix and represents u₁Corresponding reconstructed channel principal eigenvector, representing u₂Corresponding reconstructed channel principal eigenvector, representing u₁Corresponding weight vector, representing u₂The corresponding weight vector, the expressed conjugate transpose matrix and the expressed diagonal matrix.
The method of claim 7, wherein the network device determining the candidate pilot weighting matrix set according to the precoding code words corresponding to the two UEs and the target weight matrix comprises:

the network device determines a candidate pilot weighting matrix set by using a fourth formula according to the precoding code words corresponding to the two UEs and the target weight matrix, wherein the fourth formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iThe PMI of rank 2 is a precoding codeword of i, which represents a target weight matrix.
The method of claim 2, wherein before determining a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each of the M scheduled UEs if the first pilot weighting matrix is at least one of the unity weighting matrix and the preset pilot weighting matrix, the method further comprises:

if M is greater than or equal to 2, the network equipment determines a target weight matrix according to the main eigenvectors of the reconstructed channel corresponding to the at least two pieces of UE respectively;

and when the correlation of the target weight matrixes corresponding to any two UEs is smaller than a second preset threshold, the network equipment determines an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the at least two UEs and the target weight matrixes.
The method of claim 10, wherein the network device determining the candidate pilot weighting matrix set according to the precoding code words corresponding to the at least two UEs and the target weight matrix comprises:

the network device determines an alternative pilot weighting matrix set by using a fifth formula according to the precoding code words corresponding to the at least two UEs and the target weight matrix, wherein the fifth formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iAnd representing precoding code words with PMI i, wherein rank is more than or equal to 2, and representing the target weight matrix.
The method according to any of claims 5 to 11, wherein if the first pilot weighting matrix is at least one of the unitary weighting matrix and the predetermined pilot weighting matrix, the determining a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs comprises:

the network equipment randomly selects a target pre-coding code word and selects an alternative pilot weighting matrix corresponding to the target pre-coding code word from the alternative pilot weighting matrix set according to the target pre-coding matrix;

and the network equipment determines the candidate pilot frequency weighting matrix corresponding to the target pre-coding code word as the second pilot frequency weighting matrix.
The method according to any of claims 5 to 11, wherein if the first pilot weighting matrix is the unitary weighting matrix, the determining a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs comprises:

the network equipment determines a target pre-coding code word according to a preset rule, and selects an alternative pilot weighting matrix corresponding to the target pre-coding code word from the alternative pilot weighting matrix set according to the target pre-coding matrix;

and the network equipment determines the candidate pilot frequency weighting matrix corresponding to the target pre-coding code word as the second pilot frequency weighting matrix.
The method of claim 13, wherein the network device determining the target precoding codeword according to a preset rule comprises:

the network device determines the target precoding codeword using a sixth formula, where the sixth formula is expressed as:

wherein, representing a target pre-coding code word, n represents a measurement subframe s_mThe sequence number of the previous measurement subframe indicates that the measurement subframe is s_mA corresponding candidate pilot weighting matrix.
The method according to any of claims 2 to 11, wherein if the first pilot weighting matrix is the predetermined pilot weighting matrix, the determining a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs comprises:

if M is 1, the network equipment acquires a target weight matrix of the UE;

and the network equipment determines the second pilot frequency weighting matrix according to the first pilot frequency weighting matrix and the target weight matrix of the UE.
The method of claim 15, wherein the network device determining the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix of the UE comprises:

the network device determines whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, wherein the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁A PMI codebook representing rank 1, a conjugate transpose matrix represented, Q mod (m, L) represents a first pilot weighting matrix with a measurement subframe number m in L measurement subframes, and δ represents a constraint threshold value;

the network device determines a first pilot weighting matrix satisfying the seventh formula as the second pilot weighting matrix.
The method of claim 16, wherein after the network device determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix, the method further comprises:

the network device determines a target precoding codeword using an eighth formula, where the eighth formula is expressed as:

wherein the target precoding codeword is represented.
The method according to any of claims 11 to 17, wherein after determining the second pilot weighting matrix according to the reconstructed channel dominant eigenvector for each of the M scheduled UEs, the method further comprises:

and the network equipment sends a first weighted data signal to the N UEs, wherein the first weighted data signal is obtained by weighting the first data signal by the network equipment according to the target precoding code word and the second pilot weighting matrix.
A network device, wherein the network device is a device in a multiple-input multiple-output system, the network device comprising:

a receiving module, configured to receive N precoding matrix indicator PMIs, where N is an integer greater than 1, where the N PMIs are determined by N user equipments UEs according to a first weighted pilot signal sent by the network device, and the first weighted pilot signal is obtained by weighting, by the network device, a first pilot signal according to a first pilot weighting matrix;

a first determining module, configured to determine N reconstructed channel primary eigenvectors according to the N PMIs and the first pilot weighting matrix received by the receiving module;

a second determining module for determining M scheduled UEs from the N UEs;

a third determining module, configured to determine a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs determined by the second determining module, where M is an integer no greater than N;

a sending module, configured to send the second weighted pilot signal determined by the third determining module to the N UEs, where the second weighted pilot signal is obtained by the network device by weighting a second pilot signal according to the second pilot weighting matrix.
The network device of claim 19, comprising:

the first pilot weighting matrix is at least one of a unit weighting matrix and a preset pilot weighting matrix.
The network device according to claim 20, wherein the first determining module is specifically configured to determine, according to the N PMIs, N precoding codewords corresponding to the N PMIs; determining a conjugate transpose matrix corresponding to the N precoding codewords and a conjugate transpose matrix corresponding to the first pilot weighting matrix; and determining the N reconstructed channel main eigenvectors according to the N pre-coding code words, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N pre-coding code words and the conjugate transpose matrix corresponding to the first pilot weighting matrix.
The network device of claim 21, wherein the first determining module is specifically configured to determine the N reconstructed channel main eigenvectors according to the N precoding codewords, the first pilot weighting matrix, the conjugate transpose matrix corresponding to the N precoding codewords and the conjugate transpose matrix corresponding to the first pilot weighting matrix by using a first formula, where the first formula is expressed as:

the method comprises the steps of representing a reconstructed channel main characteristic vector of user equipment u determined on a subframe t, PrimEigVec representing a main characteristic vector of an acquisition matrix, L representing the total number of measurement subframes within a preset length, m being a measurement subframe sequence number and representing that the measurement subframe is s_mRepresents the user equipment u to the measurement sub-frame s_mThe selected PMI after the channel estimation of the first pilot signal represents that the PMI is a corresponding precoding code word, represents a corresponding conjugate transpose matrix and represents a corresponding conjugate transpose matrix.
The network device of claim 20, wherein the network device further comprises:

a fourth determining module, configured to determine, if the first pilot weighting matrix is the unity weighting matrix, before the third determining module determines the second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, if M is 1, a candidate pilot weighting matrix set according to the reconstructed channel dominant eigenvector corresponding to one UE and the precoding codeword corresponding to the one UE.
The network device of claim 23, wherein the fourth determining module is specifically configured to determine the candidate pilot weighting matrix set according to a reconstructed channel dominant eigenvector corresponding to one UE and a precoding codeword corresponding to the one UE by using a second formula, where the second formula is expressed as:

wherein, W_i∈w₁Representing a set of candidate pilot weighting moments, W_iA precoding codeword with PMI i indicating rank 1 represents a reconstructed channel principal eigenvector corresponding to one UE, and w₁A PMI codebook representing rank 1.
The network device of claim 20, wherein the network device further comprises:

a fourth determining module, configured to determine, if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, a target weight matrix according to the corresponding reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, if M is 2, before the third determining module determines a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs; and when the correlation of the main characteristic vectors of the reconstructed channels corresponding to the two UEs is smaller than a first preset threshold, determining a candidate pilot weighting matrix set according to the precoding code words corresponding to the two UEs and the target weight matrix.
The network device of claim 25, wherein the fourth determining module is specifically configured to determine the target weight matrix according to the reconstructed channel dominant eigenvectors corresponding to the two UEs by using a third formula, where the third formula is expressed as:

wherein, represents the target weight matrix and represents u₁Corresponding reconstructed channel principal eigenvector, representing u₂Corresponding reconstructed channel principal eigenvector, representing u₁Corresponding weight vector, representing u₂The corresponding weight vector, the expressed conjugate transpose matrix and the expressed diagonal matrix.
The network device of claim 25, wherein the fourth determining module is specifically configured to determine the candidate pilot weighting matrix set by using a fourth formula according to the precoding code words and the target weight matrices corresponding to the two UEs, where the fourth formula is represented as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iThe PMI of rank 2 is a precoding codeword of i, which represents a target weight matrix.
The network device of claim 20, wherein the network device further comprises:

a fourth determining module, configured to determine, if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, a target weight matrix according to the reconstructed channel main eigenvector corresponding to at least two UEs, if M is greater than or equal to 2, before the third determining module determines a second pilot weighting matrix according to the reconstructed channel main eigenvector of each UE of the M scheduled UEs; and when the correlation of the target weight matrixes corresponding to any two UEs is smaller than a second preset threshold, determining an alternative pilot frequency weighting matrix set according to the precoding code words corresponding to the at least two UEs and the target weight matrixes.
The network device of claim 28, wherein the fourth determining module is specifically configured to determine the candidate pilot weighting matrix set according to the precoding code words and the target weight matrices corresponding to the at least two UEs by using a fifth formula, where the fifth formula is expressed as:

wherein, a set of candidate pilot weighting matrices, W, is represented_iAnd representing precoding code words with PMI i, wherein rank is more than or equal to 2, and representing the target weight matrix.
The network device according to any one of claims 23 to 29, wherein if the first pilot weighting matrix is at least one of the unit weighting matrix and the preset pilot weighting matrix, the third determining module is specifically configured to randomly select a target precoding code word, and select an alternative pilot weighting matrix corresponding to the target precoding code word from the alternative pilot weighting matrix set according to the target precoding matrix; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.
The network device according to any one of claims 23 to 29, wherein if the first pilot weighting matrix is the unit weighting matrix, the third determining module is specifically configured to determine a target precoding codeword according to a preset rule, and select, according to the target precoding matrix, an alternative pilot weighting matrix corresponding to the target precoding codeword from the alternative pilot weighting matrix set; and determining the candidate pilot frequency weighting matrix corresponding to the target precoding code word as the second pilot frequency weighting matrix.
The network device of claim 31, wherein the third determining module is specifically configured to determine the target precoding codeword using a sixth formula, where the sixth formula is expressed as:

wherein, representing a target pre-coding code word, n represents a measurement subframe s_mThe sequence number of the previous measurement subframe indicates that the measurement subframe is s_mA corresponding candidate pilot weighting matrix.
The network device according to any one of claims 20 to 29, wherein if the first pilot weighting matrix is the preset pilot weighting matrix, the third determining module is specifically configured to obtain a target weight matrix of a UE if M is 1; and determining the second pilot frequency weighting matrix according to the first pilot frequency weighting matrix and the target weight matrix of the UE.
The network device of claim 33, wherein the third determining module is specifically configured to determine whether the first pilot weighting matrix and the target weight matrix satisfy a seventh formula, where the seventh formula is expressed as:

wherein, W_iPrecoding codeword with PMI i, w, representing rank ═ 1₁A PMI codebook representing rank 1, a conjugate transpose matrix represented, Q mod (m, L) represents a first pilot weighting matrix with a measurement subframe number m in L measurement subframes, and δ represents a constraint threshold value; determining a first pilot weighting matrix satisfying the seventh formula as the second pilot weighting matrix.
The network device of claim 34, wherein the fourth determining module is further configured to determine a target precoding codeword using an eighth formula after the third determining module determines the second pilot weighting matrix according to the first pilot weighting matrix and the target weight matrix, wherein the eighth formula is expressed as:

wherein the target precoding codeword is represented.
The network device of any one of claims 29 to 35,

the sending module is further configured to send a first weighted data signal to the N UEs after the third determining module determines a second pilot weighting matrix according to the reconstructed channel dominant eigenvector of each UE of the M scheduled UEs, where the first weighted data signal is obtained by weighting, by the network device, a first data signal according to the target precoding codeword and the second pilot weighting matrix.