CN103888213A - Precoding method and device - Google Patents

Abstract

The embodiments of the invention disclose a precoding method and device, and relates to the field of communication. The precoding method and device of the invention does not limited by the antenna number of receiving end devices, and at the same time, the computation complexity of channel precoding can be reduced, and the channel transmitting performance gain can be ensured. The method comprises the following steps: a downlink channel matrix is acquired; an initial weight value corresponding to the downlink channel matrix is acquired; a balanced equivalent channel matrix is acquired according to the downlink channel matrix and the initial weight value and a weight value corresponding to the next equivalent channel matrix is acquired according to the equivalent channel matrix; and iterative computation is performed according to the equivalent channel matrix and the weight value corresponding to the next equivalent channel matrix to obtain a weight value corresponding to the downlink channel matrix until a default iterative time is satisfied or the weight value corresponding to the downlink channel matrix that obtained through computation satisfies a default convergence condition. The embodiments of the invention are applied in channel precoding.

Description

Precoding method and equipment

Technical Field

The present invention relates to the field of communications, and in particular, to a precoding method and apparatus.

Background

In a downlink Multi-User Multiple-input Multiple-Output (MU-MIMO) system, a base station transmits different data to Multiple users on the same time-frequency domain resource, at this time, Co-Channel Interference (CCI) exists between cooperative users, and theoretically, it is desirable to design a transmission signal to eliminate inter-User Interference (MUI). If the sending end can obtain the downlink CSI, the interference situation of each user can be known, and the interference among multiple users can be eliminated through a reasonable precoding mode.

The first is direct channel inversion (ZF), and the idea of the method is to directly Zero the interference between the receiving side and the user terminal, but the method has a limit on the number of receiving antennas at the receiving side, and the performance is seriously degraded when the number of receiving antennas exceeds a threshold; the second method is a Signal-to-noise-Ratio (SLNR) method, which does not limit the number of receiving antennas at the receiving side, but needs to acquire the noise power of the user terminal at the receiving side, and the performance gain is severely reduced when the Signal-to-noise Ratio is high; the third method is an iteration scheme based on zero forcing, and is to perform iteration calculation according to the initial weight of the downlink channel matrix of the system to obtain the precoding matrix adopted by the downlink channel corresponding to the next transmission.

In the channel precoding method provided in the prior art, the inventor finds that in the prior art, the number of antennas of the transmitting end device is limited in the channel precoding process, the calculation complexity is high, and the channel transmission performance gain cannot be ensured.

Disclosure of Invention

Embodiments of the present invention provide a precoding method and device, which are not limited by the number of antennas of a receiving end device, and meanwhile, can reduce the computational complexity of channel precoding, and ensure the transmission performance gain of a channel.

In order to achieve the above purpose, the embodiment of the invention adopts the following technical scheme:

in a first aspect, a precoding method is provided, including:

acquiring a downlink channel matrix;

acquiring an initial weight corresponding to the downlink channel matrix;

obtaining an equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix;

and carrying out iterative computation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until a preset iteration number is met or the calculated weight corresponding to the downlink channel matrix meets a preset convergence condition.

In a first possible implementation manner, with reference to the first aspect, the obtaining an initial weight corresponding to the downlink channel matrix includes:

performing singular value decomposition on the downlink channel matrix;

zero forcing is carried out on the decomposed right singular vector of the downlink channel matrix to obtain an initial weight corresponding to the downlink channel matrix;

or,

and taking the decomposed non-zero right singular vector of the downlink channel matrix as an initial weight corresponding to the downlink channel matrix.

In a second possible implementation manner, with reference to the first aspect, the obtaining an initial weight corresponding to the downlink channel matrix includes:

acquiring a downlink channel covariance matrix according to the downlink channel matrix;

performing eigenvalue decomposition on the downlink channel covariance matrix;

zero forcing is carried out on the eigenvalue vector of the decomposed downlink channel covariance matrix to obtain an initial weight corresponding to the downlink channel matrix;

or,

and taking the decomposed non-zero eigenvalue vector of the downlink channel covariance matrix as an initial weight corresponding to the downlink channel matrix.

In a third possible implementation manner, with reference to the first aspect, the obtaining an equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix includes:

obtaining an equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

In a fourth possible implementation manner, with reference to the second possible implementation manner, the obtaining the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix includes:

acquiring a downlink channel covariance matrix according to the downlink channel matrix;

obtaining an equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

In a fifth possible implementation manner, with reference to the first aspect, the obtaining an equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix includes:

calculating an equivalent channel matrix of each user terminal according to the initial weight;

obtaining the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal and the downlink channel matrix;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

In a sixth possible implementation manner, with reference to the first aspect, the obtaining an equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix includes:

calculating an equivalent channel matrix of each user terminal according to the initial weight;

obtaining the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal and the interference noise covariance matrix fed back by each user terminal;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

In a seventh possible implementation manner, with reference to the first aspect or any one of the first to sixth possible implementation manners, performing iterative computation according to the equivalent matrix and a weight corresponding to the next equivalent channel matrix to obtain a weight corresponding to the next downlink channel matrix, and before a preset iteration number is met or a weight corresponding to the next downlink channel matrix obtained through computation converges to meet a preset convergence condition, the method further includes:

acquiring the type of a terminal receiver sent by a user terminal;

and setting the iteration times corresponding to the type of the terminal receiver as the preset iteration times.

In an eighth possible implementation manner, in combination with the seventh possible implementation manner, if the terminal receiver type cannot be obtained, the preset iteration number is set as the default iteration number according to the default terminal receiver type of the user terminal.

In a second aspect, an encoding apparatus is provided, which includes:

the channel acquisition unit is used for acquiring a downlink channel matrix;

the initialization unit is used for acquiring an initial weight corresponding to the downlink channel matrix forwarded by the channel acquisition unit;

the iteration unit is used for acquiring the equalized equivalent channel matrix according to the downlink channel matrix forwarded by the channel acquisition unit and the initial weight forwarded by the initialization unit and acquiring a weight corresponding to the next equivalent channel matrix according to the equivalent channel matrix; and carrying out iterative computation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until a preset iteration number is met or the calculated weight corresponding to the downlink channel matrix meets a preset convergence condition.

In a first possible implementation manner, with reference to the second aspect, the initialization unit includes:

a singular value decomposition subunit, configured to perform singular value decomposition on the downlink channel matrix forwarded by the channel acquisition unit;

a zero forcing calculation subunit, configured to zero-force a right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit to obtain an initial weight corresponding to the downlink channel matrix;

or,

and the zero forcing calculation subunit is configured to use the decomposed non-zero right singular vector of the downlink channel matrix forwarded by the singular value decomposition subunit as an initial weight corresponding to the downlink channel matrix.

In a second possible implementation manner, with reference to the second aspect, the initialization unit includes:

the covariance calculation subunit is used for acquiring a downlink channel covariance matrix according to the downlink channel matrix;

the eigenvalue decomposition subunit is further configured to perform eigenvalue decomposition on the downlink channel covariance matrix forwarded by the covariance calculation subunit;

the zero forcing calculation subunit is further configured to zero-force the eigenvalue vector of the downlink channel covariance matrix decomposed by the eigenvalue decomposition subunit to obtain an initial weight corresponding to the downlink channel matrix;

or,

the zero-forcing calculation subunit is further configured to use the non-zero eigenvalue vector of the downlink channel covariance matrix after the eigenvalue decomposition subunit decomposes as an initial weight corresponding to the downlink channel matrix.

In a third possible implementation manner, with reference to the second aspect, the iteration unit includes:

the equalizing subunit is further configured to obtain an equalized equivalent channel matrix from the downlink channel covariance matrix forwarded by the covariance calculating subunit and the initial weight forwarded by the initialization unit;

and the weight calculation subunit is further configured to obtain a weight corresponding to a next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

In a fourth possible implementation manner, with reference to the second possible implementation manner, the iteration unit includes:

the covariance calculation subunit is also used for acquiring a downlink channel covariance matrix according to the downlink channel matrix forwarded by the channel acquisition unit;

the equalization subunit is configured to obtain an equalized equivalent channel matrix according to the downlink channel covariance matrix forwarded by the covariance evaluation subunit and the initial weight forwarded by the initialization unit;

and the weight calculation subunit is used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

In a fifth possible implementation manner, with reference to the second aspect, the iteration unit includes:

an equivalent matrix obtaining subunit, configured to calculate an equivalent channel matrix of each user terminal according to the initial weight forwarded by the initialization unit;

the equalizing subunit is further configured to obtain the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix obtaining subunit and the downlink channel matrix forwarded by the channel acquiring unit;

and the weight calculation subunit is further configured to obtain a weight corresponding to a next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

In a sixth possible implementation manner, with reference to the second aspect, the iteration unit includes:

an equivalent matrix obtaining subunit, configured to calculate an equivalent channel matrix of each user terminal according to the initial weight forwarded by the initialization unit;

the equalizing subunit is configured to obtain the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix obtaining subunit and the interference noise covariance matrix fed back by each user terminal;

and the weight calculation subunit is used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

In a seventh possible implementation manner, with reference to the second aspect or any one of the first to sixth possible implementation manners, the encoding device further includes:

the iteration number setting unit is used for acquiring the type of the terminal receiver sent by the user terminal; and setting the iteration times corresponding to the type of the terminal receiver as the preset iteration times.

In an eighth possible implementation manner, in combination with the seventh possible implementation manner, the iteration number setting unit is further configured to set the preset iteration number as a default iteration number according to a default terminal receiver type of the user terminal if the iteration number setting unit cannot acquire the terminal receiver type.

According to the precoding method and the device provided by the embodiment of the invention, only one eigenvalue decomposition is needed in the process of calculating the initial weight corresponding to the current downlink channel matrix, and then the subsequent iterative calculation is carried out according to the initial weight and the weight corresponding to the next downlink channel matrix is obtained, so that the limitation of the number of antennas of receiving end equipment is avoided, the calculation complexity of channel precoding can be reduced, and the emission performance gain of a channel is ensured.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a flowchart illustrating a precoding method according to an embodiment of the present invention;

fig. 2 is a flowchart illustrating another precoding method according to an embodiment of the present invention;

fig. 3 is a flowchart illustrating a precoding method according to another embodiment of the present invention;

fig. 4 is a flowchart illustrating a precoding method according to another embodiment of the present invention;

fig. 5 is a flowchart illustrating another precoding method according to an embodiment of the present invention;

fig. 6 is a schematic diagram of a performance simulation curve of a precoding method according to an embodiment of the present invention;

fig. 7 is a schematic diagram of a performance simulation curve of a precoding method according to another embodiment of the present invention;

fig. 8 is a schematic structural diagram of an encoding apparatus according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of another encoding apparatus provided in an embodiment of the present invention;

fig. 10 is a schematic structural diagram of another encoding apparatus according to an embodiment of the present invention;

fig. 11 is a schematic structural diagram of a further encoding apparatus according to an embodiment of the present invention;

fig. 12 is a schematic structural diagram of another encoding apparatus provided in the embodiment of the present invention;

fig. 13 is a schematic structural diagram of another encoding apparatus according to an embodiment of the present invention;

fig. 14 is a schematic structural diagram of a further encoding apparatus according to an embodiment of the present invention;

fig. 15 is a schematic structural diagram of another encoding apparatus according to an embodiment of the present invention;

fig. 16 is a schematic structural diagram of an encoding apparatus according to another 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.

Taking a downlink MU-MIMO system as an example, a sending end sends different data to multiple user terminals on the same time-frequency domain resource, and at this time, co-channel interference CCI exists between the cooperating user terminals, theoretically, it is desirable to reasonably design a sending signal to eliminate interference MUI between the user terminals, and if the sending end can obtain downlink channel state information, it is possible to know the interference situation of each user terminal, and eliminate interference between the user terminals through a reasonable precoding manner; here, the precoding device at the transmitting end may be a base station, and may also be a module or a functional unit on the base station, and the following embodiment takes the base station as an example for description. Firstly, signals based on MU-MIMO system are givenAnd (4) modeling. Suppose N_tIs the number of antennas at the base station that occur,the number of receiving antennas of user k, the transmission channel of user k isThe weight matrix of user k is <math> <mrow> <msub> <mi>W</mi> <mi>k</mi> </msub> <mo>=</mo> <mo>[</mo> <msub> <mi>W</mi> <mrow> <mi>k</mi> <mo>,</mo> <mn>1</mn> </mrow> </msub> <mo>,</mo> <msub> <mi>W</mi> <mrow> <mi>k</mi> <mo>,</mo> <mn>2</mn> </mrow> </msub> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <msub> <mi>W</mi> <mrow> <mi>k</mi> <mo>,</mo> <msubsup> <mi>N</mi> <mi>s</mi> <mi>k</mi> </msubsup> </mrow> </msub> <mo>]</mo> <mo>&Element;</mo> <msup> <mi>C</mi> <mrow> <msub> <mi>N</mi> <mi>t</mi> </msub> <mo>&times;</mo> <msubsup> <mi>N</mi> <mi>s</mi> <mi>k</mi> </msubsup> </mrow> </msup> <mo>,</mo> </mrow> </math> The signal vector transmitted by user k is <math> <mrow> <msub> <mi>X</mi> <mi>k</mi> </msub> <mo>&Element;</mo> <msup> <mi>C</mi> <mrow> <msubsup> <mi>N</mi> <mi>s</mi> <mi>k</mi> </msubsup> <mo>&times;</mo> <mn>1</mn> </mrow> </msup> <mo>,</mo> </mrow> </math> Assuming that k users are paired, the signal model of the receiving end can be represented as follows:

$y={[y_1^T,...,y_K^T]}^T={\mathrm{HW}}_X+n$

wherein, $H={[H_1^T,...,H_K^T]}^T,$ $X={[X_1^T,...,X_K^T]}^T,$ W_k＝[W₁，W₂，...，W_k]， $n={[n_1^T,...,n_K^T]}^T,$ <math> <mrow> <msub> <mi>n</mi> <mi>k</mi> </msub> <mo>&Element;</mo> <msup> <mi>C</mi> <mrow> <msub> <mi>N</mi> <mi>r</mi> </msub> <mo>&times;</mo> <mn>1</mn> </mrow> </msup> </mrow> </math> is gaussian white noise corresponding to user k.

Based on the above system model, referring to fig. 1, an embodiment of the present invention provides a precoding method, including:

101. the encoding device acquires a downlink channel matrix.

102. And acquiring an initial weight corresponding to the downlink channel matrix.

103. And acquiring the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and acquiring the weight corresponding to the next equivalent channel matrix according to the equivalent channel matrix.

104. And carrying out iterative calculation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until the preset iteration times are met or the weight corresponding to the calculated downlink channel matrix meets the preset convergence condition.

The precoding method provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

Specifically, referring to fig. 2, an embodiment of the present invention provides a precoding method, including:

201. the encoding device acquires a downlink channel matrix H.

Of course, the downlink channel matrix H is obtained by the base station according to the feedback from the user terminal side or the uplink and downlink channel reciprocity.

202. Calculating covariance matrix R of downlink channel_hh，k。

Wherein $R_{\mathrm{hh},k}=H_k^H{H}_K.$

203. After singular value decomposition is carried out on a downlink channel matrix H, an initial weight T corresponding to the downlink channel matrix is obtained by adopting a zero-forcing algorithm₁。

In step 203, firstly, H is measured_kK is formed by Singular value decomposition (SVD for short) decomposition of k belonging to {1, 2

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> Equation 1

At this point in front of the user terminal k

The non-zero right singular vectors can be expressed as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^{\mathrm{<figure-callout\; id="2"\; label="H\; equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">H</figure-callout>}}$ equation 2

Order to

To pair

Zero forcing can be carried out to obtain an initial weight T₁，

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H{\left(\tilde{H}{\tilde{H}}^H\right)}^{-1}\mathrm{diag}{\left(p_1\right)}^{1/2}$ Equation 3

Wherein,

is a power adjustment factor. P₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

Of course, step 202 and step 203 are not sequential, and may be performed simultaneously.

204. And acquiring the equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight.

The specific calculation process of the equivalent matrix is as follows:

let T_j，kRepresenting the weight value corresponding to the jth iteration of the user k, wherein the expression is as follows:

<math> <mrow> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>T</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mo>:</mo> <mo>,</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mi>k</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>:</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math> equation 4

Of course, when j is 1, the initial weight can be obtained.

Then, using R_hh，kObtaining the equalized equivalent channel matrix Z_j

$Z_j={[{\left(T_{j,1}^H{R}_{\mathrm{hh},1}\right)}^T,{\left(T_{j,2}^H{R}_{\mathrm{hh},2}\right)}^T,...,{\left(T_{j,k}^H{R}_{\mathrm{hh},k}\right)}^T]}^{\mathrm{<figure-callout\; id="5"\; label="T\; Equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}$ Equation 5

205. And acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix.

Finally, a weight matrix is calculated by utilizing a zero-forcing algorithm,

$T_{j+1}=Z_j^H{\left(Z_j{Z}_j^H\right)}^{-1}\mathrm{diag}{\left(p_j\right)}^{1/2}$ equation 6

Is a power adjustment factor. Let j be j +1 and W be T_j+1。

206. And carrying out iterative calculation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until the preset iteration times are met or the weight corresponding to the calculated downlink channel matrix meets the preset convergence condition.

Here, the processes of equations 4 to 6 are repeated until a preset number of iterations is satisfied or a weight corresponding to the next downlink channel matrix obtained through calculation satisfies a preset convergence condition.

The convergence condition in the present invention is set according to whether the user equipment UE feeds back the receiver type. And if the UE feeds back the receiver type, the base station sets the iteration times according to the receiver type fed back by the UE. At this time, the embodiment of the present invention further includes:

206a, obtaining the terminal receiver type sent by the user terminal.

206b, setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Of course, if the UE does not feed back the receiver type, the embodiment further includes:

206c, setting the preset iteration times as default iteration times or setting convergence conditions according to the default terminal receiver type of the user terminal.

Of course, the number of iterations here is an empirical value set according to the terminal receiver type of the user terminal, and common reception types mainly include: a maximum-ratio Combining (MRC) receiver, a Minimum Mean Square Error (MMSE) receiver, an Interference Rejection Combining (IRC) receiver, and a Maximum Likelihood Detection (MLD) receiver are respectively set to have an iteration number N_MRC、N_MMSE、N_IRCAnd N_MLDOne possible value is shown in table 1. If the paired users all adopt the same receiver, the iteration times required by the corresponding receiver are set, and if the paired users adopt different receivers, the maximum iteration times required by the paired users are selected for setting.

	NMRC	NMMSE	NIRC	NMLD
					2 user pairing	1	1	1	1
3 user pairing	3	3	1	1
					4 user pairing	6	6	1	1

TABLE 1

And if the UE does not feed back the receiver type, the base station sets default iteration times or convergence conditions. If default iteration times are set, according to N_MRCOr N_MMSEThe setting is carried out by the following steps,

or convergence conditions are set, wherein the expression of one convergence condition is as follows:

||T_j+1-T_jequation 7

Where ε is a constant.

In this embodiment, the initial weight T corresponding to the downlink channel matrix₁The acquisition is based on the calculation of zero forcing of right singular vectors after singular value decomposition of a downlink channel matrix HMethod, of course, for the initial weight T corresponding to the downlink channel matrix₁The obtaining of (b) may also be based on a calculation method of zero-forcing a non-zero right singular vector after singular value decomposition of the downlink channel matrix H.

Specifically, step 203 may also be implemented by:

adopting the non-zero right singular vector of the channel as the initial weight T of the MU-MIMO system₁. The specific description is as follows:

to H_kK is in SVD decomposition of {1, 2., k },

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> equation 8

Front of user k

The non-zero right singular vectors can be represented as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^H$ equation 9

Order to

Taking the non-zero right singular vector of the channel as the initial weight of the MU-MIMO system,

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H\mathrm{diag}{\left(p_1\right)}^{1/2}$ equation 10

Wherein,

is a power adjustment factor. p is a radical of₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

The precoding method provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

Specifically, referring to fig. 3, an embodiment of the present invention provides a precoding method, including:

301. the encoding device acquires a downlink channel matrix H.

Of course, the downlink channel matrix H is obtained by the base station according to the feedback from the user terminal side or the uplink and downlink channel reciprocity.

302. Calculating a downlink channel covariance matrix R_hh，k。

Wherein $R_{\mathrm{hh},k}=H_k^H{H}_K.$

303. For downlink channel covariance matrix R_hh，kPerforming eigenvalue decomposition, and zero forcing on the eigenvalue vector of the decomposed downlink channel covariance matrixAnd acquiring an initial weight corresponding to the downlink channel matrix.

<math> <mrow> <msub> <mi>R</mi> <mrow> <mi>hh</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msubsup> <mi>&Sigma;</mi> <mi>k</mi> <mn>2</mn> </msubsup> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> Equation 11

At this point in front of the user terminal k

The vector of non-zero eigenvalues may be represented as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^{\mathrm{<figure-callout\; id="12"\; label="H\; equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">H</figure-callout>}}$ equation 12

Order toTo pairZero forcing can be carried out to obtain an initial weight T₁，

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H{\left(\tilde{H}{\tilde{H}}^H\right)}^{-1}\mathrm{diag}{\left(p_1\right)}^{1/2}$ Equation 13

Wherein,

is a power adjustment factor. P₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

304. And acquiring the equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight.

The specific calculation process of the equivalent matrix is as follows:

let T_j，kRepresenting the weight value corresponding to the jth iteration of the user k, wherein the expression is as follows:

<math> <mrow> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>T</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mo>:</mo> <mo>,</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mi>k</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>:</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math> equation 14

Of course, when j is 1, the initial weight can be obtained.

Then, using R_hh，kObtaining the equalized equivalent channel matrix Z_j

$Z_j={[{\left(T_{j,1}^H{R}_{\mathrm{hh},1}\right)}^T,{\left(T_{j,2}^H{R}_{\mathrm{hh},2}\right)}^T,...,{\left(T_{j,k}^H{R}_{\mathrm{hh},k}\right)}^T]}^{\mathrm{<figure-callout\; id="15"\; label="T\; Equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}$ Equation 15

305. And acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix.

Finally, a weight matrix is calculated by utilizing a zero-forcing algorithm,

$T_{j+1}=Z_j^H{\left(Z_j{Z}_j^H\right)}^{-1}\mathrm{diag}{\left(p_j\right)}^{1/2}$ equation 16

Is a power adjustment factor. Let j be j +1 and W be T_j+1。

306. And carrying out iterative calculation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until the preset iteration times are met or the weight corresponding to the calculated downlink channel matrix meets the preset convergence condition.

Here, the processes of equations 14 to 16 are repeated until a preset number of iterations is satisfied or a weight corresponding to the next downlink channel matrix obtained through calculation satisfies a preset convergence condition.

The convergence condition in the present invention is set according to whether the user equipment UE feeds back the receiver type. And if the UE feeds back the receiver type, the base station sets the iteration times according to the receiver type fed back by the UE. At this time, the embodiment of the present invention further includes:

306a, obtaining the terminal receiver type sent by the user terminal.

And 306b, setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Of course, if the UE does not feed back the receiver type, the embodiment further includes:

and 306c, setting the preset iteration times as default iteration times or setting a convergence condition according to the default terminal receiver type of the user terminal.

Of course, the iteration number here is an empirical value set according to the terminal receiver type of the user terminal, and the terminal receiver type, the iteration number corresponding to different terminal receivers, and the setting of the default convergence condition are not described again with specific reference to the previous embodiment.

In this embodiment, the initial weight T corresponding to the downlink channel matrix₁Is obtained based on the covariance matrix R of the downlink channel_hh，kAfter the eigenvalue decomposition, the method of zero forcing calculation is performed on the eigenvalue vector of the decomposed downlink channel covariance matrix, certainly for the initial weight T corresponding to the downlink channel matrix₁The obtaining may also be based on characterizing the downlink channel covariance matrixTaking the non-zero eigenvalue vector as an initial weight T after value decomposition₁。

Specifically, step 303 may also be implemented by:

after eigenvalue decomposition is carried out on the downlink channel covariance matrix, the non-zero eigenvalue vector is used as the initial weight T of the MU-MIMO system₁. The specific description is as follows:

to H_kK is in SVD decomposition of {1, 2., k },

<math> <mrow> <msub> <mi>R</mi> <mrow> <mi>hh</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msubsup> <mi>&Sigma;</mi> <mi>k</mi> <mn>2</mn> </msubsup> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> equation 17

Front of user k

The vector of non-zero eigenvalues may be represented as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^{\mathrm{<figure-callout\; id="18"\; label="H\; equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">H</figure-callout>}}$ equation 18

Order to

Taking the non-zero right singular vector of the channel as the initial weight of the MU-MIMO system,

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H\mathrm{diag}{\left(p_1\right)}^{1/2}$ equation 19

Wherein,

is a power adjustment factor. p is a radical of₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

The precoding method provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of precoding and ensure the emission performance gain of a channel.

An embodiment of the present invention provides a precoding method, which is shown in fig. 4 and includes:

401. the encoding device acquires a downlink channel matrix H.

Of course, the downlink channel matrix H is obtained by the base station according to the feedback from the user terminal side or the uplink and downlink channel reciprocity.

402. After singular value decomposition is carried out on a downlink channel matrix H, an initial weight T corresponding to the downlink channel matrix is obtained by adopting a zero-forcing algorithm₁。

In step 402, H is first aligned_kK belongs to {1, 2.,. k } to be processed by SVD decomposition to obtain,

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> equation 20

At this point in front of the user terminal k

The non-zero right singular vectors can be expressed as follows,

order to

To pair

Zero forcing can be carried out to obtain an initial weight T₁，

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H{\left(\tilde{H}{\tilde{H}}^H\right)}^{-1}\mathrm{diag}{\left(p_1\right)}^{1/2}$ Equation 21

Wherein,

is a power adjustment factor. P₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

403. And calculating the equivalent channel matrix of each user terminal according to the initial weight.

Calculating the equivalent channel R of the user_k，jFirst, the received signal of user k can be expressed as follows,

<math> <mrow> <msub> <mi>y</mi> <mrow> <mi>k</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>H</mi> <mi>k</mi> </msub> <msub> <mi>T</mi> <mi>J</mi> </msub> <mi>x</mi> <mo>+</mo> <msub> <mi>n</mi> <mi>k</mi> </msub> <mo>=</mo> <msub> <mi>H</mi> <mi>k</mi> </msub> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>i</mi> </mrow> </msub> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>+</mo> <msub> <mi>n</mi> <mi>k</mi> </msub> </mrow> </math> equation 22

Then the equivalent channel matrix for user k is:

R_k，j＝H_kT_j，kequation 23

Wherein, T_j，kRepresenting the weight corresponding to the jth iteration of user k,

<math> <mrow> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>T</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mo>:</mo> <mo>,</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mi>k</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>:</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math> equation 24

404. And acquiring the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal.

By means of R_k，jObtain a matrix Z_jAt this time, no matter what receiver the UE employs, it is assumed that the UE employs the MRC receiver, and Z at this time_jThe expression of (a) is:

$Z_j={[{\left({\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">R</figure-callout>}}_{1,j}^H{H}_1\right)}^T,{\left({\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">R</figure-callout>}}_{2,j}^H{H}_2\right)}^T,...,{\left(R_{K,j}^H{H}_k\right)}^T]}^{\mathrm{<figure-callout\; id="25"\; label="T\; equation"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}$ equation 25

405. And acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix.

$T_{j+1}=Z_j^H{\left(Z_j{Z}_j^H\right)}^{-1}\mathrm{diag}{\left(p_j\right)}^{1/2}$ Equation 26

For power adjustment factor, let j be j +1 and W be T_j+1。

406. And carrying out iterative calculation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the next downlink channel matrix until the preset iteration times are met or the weight corresponding to the next downlink channel matrix obtained by calculation meets the preset convergence condition.

Repeating the processes from the formula 16 to the formula 17 until a preset iteration number is satisfied or a weight corresponding to the next downlink channel matrix obtained through calculation satisfies a preset convergence condition.

The convergence condition in the present invention is set according to whether the user equipment UE feeds back the receiver type. And if the UE feeds back the receiver type, the base station sets the iteration times according to the receiver type fed back by the UE. At this time, the embodiment of the present invention further includes:

406a, obtaining the terminal receiver type sent by the user terminal.

And 406b, setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Of course, if the UE does not feed back the receiver type, the embodiment further includes:

and 406c, setting the preset iteration times as default iteration times or setting a convergence condition according to the type of the terminal receiver of the user terminal.

Of course, the iteration number here is an empirical value set according to the terminal receiver type of the user terminal, and the terminal receiver type, the iteration number corresponding to different terminal receivers, and the setting of the default convergence condition are not described again with specific reference to the previous embodiment.

In this embodiment, the initial weight T corresponding to the downlink channel matrix₁The obtaining is based on a calculation method of zero forcing of right singular vectors after singular value decomposition of a downlink channel matrix H, and certainly for an initial weight T corresponding to the downlink channel matrix₁The obtaining of (b) may also be based on a calculation method of zero-forcing a non-zero right singular vector after singular value decomposition of the downlink channel matrix H.

Specifically, the step 402 can also be implemented by the following steps:

taking a non-zero right singular vector of a channel as an MU-MIMO initial weight T₁. The specific description is as follows.

To H_kK is in SVD decomposition of {1, 2., k },

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> equation 27

Front of user k

The non-zero right singular vectors can be represented as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^H$ equation 28

Order to

Taking the non-zero right singular vector of the channel as the initial weight of the MU-MIMO system,

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H\mathrm{diag}{\left(p_1\right)}^{1/2}$ equation 29

Wherein,

is a power adjustment factor. p is a radical of₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

In addition, step 402 can be replaced by obtaining the initial weight T by performing eigenvalue decomposition on the downlink channel covariance matrix as described in steps 302 and 303₁。

The precoding method provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

Referring to fig. 5, an embodiment of the present invention provides a precoding method, including the following steps:

501. the encoding device acquires a downlink channel matrix H.

Of course, the downlink channel matrix H is obtained by the base station according to the feedback from the user terminal side or the uplink and downlink channel reciprocity.

502. After singular value decomposition is carried out on a downlink channel matrix H, an initial weight T corresponding to the downlink channel matrix is obtained by adopting a zero-forcing algorithm₁。

In step 502, H is first aligned_KAnd k belongs to {1, 2, 3

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> Equation 30

At this point in front of the user terminal k

The non-zero right singular vectors can be expressed as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^H$ equation 31

Order to

To pairZero forcing can be carried out to obtain an initial weight T₁，

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H{\left(\tilde{H}{\tilde{H}}^H\right)}^{-1}\mathrm{diag}{\left(p_1\right)}^{1/2}$ Equation 32

Wherein,

is a power adjustment factor. P₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

503. And calculating the equivalent channel matrix of each user terminal according to the initial weight.

Calculating the equivalent channel R of the user_k，jFirst, the received signal of user k can be expressed as follows,

<math> <mrow> <msub> <mi>y</mi> <mrow> <mi>k</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>H</mi> <mi>k</mi> </msub> <msub> <mi>T</mi> <mi>J</mi> </msub> <mi>x</mi> <mo>+</mo> <msub> <mi>n</mi> <mi>k</mi> </msub> <mo>=</mo> <msub> <mi>H</mi> <mi>k</mi> </msub> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>i</mi> </mrow> </msub> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>+</mo> <msub> <mi>n</mi> <mi>k</mi> </msub> </mrow> </math> equation 33

Then the equivalent channel matrix for user k is:

R_k，j＝H_kT_j，kequation 34

Wherein, T_j，kRepresenting the weight corresponding to the jth iteration of user k,

<math> <mrow> <msub> <mi>T</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>k</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>T</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mo>:</mo> <mo>,</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mi>k</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>:</mo> <mrow> <mo>(</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msubsup> <mi>N</mi> <mi>s</mi> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math> equation 35

504. And acquiring the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal and the interference noise covariance matrix fed back by each user terminal.

By means of R_k，jObtain a matrix Z_jAt this time, no matter what receiver the UE employs, it is assumed that the UE employs the IRC receiver, and Z at this time_jThe expression of (a) is:

for user k, the jth iteration, if it is assumed that the UE employs an IRC receiver:

${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">G</figure-callout>}}_{1,j}^H=R_{k,j}^H(R_{k,j}{R}_{k,j}^H+R_{\mathrm{uu},k})$ equation 36

Wherein R is_uu，kInterference noise covariance matrix representing user k. This Z is_jThe expression of (a) is as follows:

$Z_j={[{\left({\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">G</figure-callout>}}_{1,j}^H{H}_1\right)}^T,{\left({\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">G</figure-callout>}}_{2,j}^H{H}_2\right)}^T,...,{\left(G_{K,j}^H{H}_K\right)}^T]}^T$ equation 37

505. And acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix.

$T_{j+1}=Z_j^H{\left(Z_j{Z}_j^H\right)}^{-1}\mathrm{diag}{\left(p_j\right)}^{1/2}$ Equation 38

For power adjustment factor, let j be j +1 and W be T_j+1。

506. And carrying out iterative calculation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the next downlink channel matrix until the preset iteration times are met or the weight corresponding to the next downlink channel matrix obtained by calculation meets the preset convergence condition.

Specifically, step 502 may also be implemented by:

adopting the non-zero right singular vector of the channel as the initial weight T of the MU-MIMO system₁. The specific description is as follows.

To H_kK is in SVD decomposition of {1, 2., k },

<math> <mrow> <msub> <mi>H</mi> <mi>k</mi> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>U</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>&Sigma;</mi> <mi>k</mi> </msub> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msubsup> </mtd> <mtd> <msubsup> <mi>V</mi> <mi>k</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msubsup> </mtd> </mtr> </mtable> </mfenced> <mi>H</mi> </msup> </mrow> </math> equation 39

Front of user kThe non-zero right singular vectors can be represented as follows,

${\tilde{H}}_k={\left(V_k^{\left(1\right)}(:,1:N_s^k)\right)}^H$ equation 40

Order to

And adopting a non-zero right singular vector of a channel as an initial weight of the MU-MIMO system.

${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00002618658700041.png"\; state="\{\{state\}\}">T</figure-callout>}}_1={\tilde{H}}^H\mathrm{diag}{\left(p_1\right)}^{1/2}$ Equation 41

Wherein,

is a power adjustment factor. p is a radical of₁The method can be set according to various principles, including water filling principles, average power distribution principles and the like.

The convergence condition in the present invention is set according to whether the user equipment UE feeds back the receiver type. And if the UE feeds back the receiver type, the base station sets the iteration times according to the receiver type fed back by the UE. At this time, the embodiment of the present invention further includes:

506a, acquiring the terminal receiver type sent by the user terminal.

And 506b, setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Of course, if the UE does not feed back the receiver type, the embodiment further includes:

506c, setting the preset iteration times as default iteration times or setting convergence conditions according to the default terminal receiver type of the user terminal.

Of course, the iteration number here is an empirical value set according to the terminal receiver type of the user terminal, and details of the terminal receiver type, the iteration number corresponding to different terminal receivers, and the setting of the default convergence condition are not described here with specific reference to the above embodiments.

In addition, step 502 may be replaced by obtaining the initial weight T by performing eigenvalue decomposition on the downlink channel covariance matrix as described in steps 302 and 303₁。

The mathematical operation notation used in the above examples is as follows: (.)^HRepresents the conjugate transpose of (·); (.)^-1Represents inverting (·); (.)^TRepresents transposing (·); (.)^*Represents conjugating (-) to; tr (-) represents (-) tracing; diag (-) denotes a diagonal matrix composed of the elements of the vector of (-); i is_nAn identity matrix representing n × n; a (: m: n) represents the m-th to n-th columns of the selection matrix A.

In addition, as shown in fig. 6, a transmitting antenna of a transmitting end device 4, a receiving antenna model of each user terminal 2 of a user terminal 2, and a receiving antenna model of each user terminal 2 of a transmitting end device 4 are provided, and fig. 7 is a diagram of a curve relationship between a Signal-to-Noise ratio (SNR) and a performance (capacity) of each user terminal during single-stream pairing when the channel precoding method provided by the embodiment shown in fig. 2 of the present invention is adopted, where fig. 6 and fig. 7 are graphs of a curve relationship between a Signal-to-Noise ratio (SNR/Noise, SNR) and a performance (capacity) of each user terminal, and a curve relationship between a Signal-to-Noise ratio (SNR/Noise, SNR) and a performance (capacity) of a corresponding user terminal during a simulation process is randomly generated, where a scheme 1, a scheme 2, and a scheme 3 provided by the prior art and a curve relationship between a Signal-to-Noise ratio (SNR/Noise, SNR) and a performance (capacity) under the scheme provided by the present, The performance gain effect is the best along with the increase of the Signal-to-noise Ratio in the modes provided by the scheme 2 and the scheme 3, wherein the scheme 1 is a direct channel inversion (ZF for short), and the scheme 2 is a Signal-to-leakage-and-noise Ratio (SLNR for short) method; scheme 3 is an iterative scheme based on zero forcing, where the total number of receiving antennas of the user terminal corresponding to fig. 6 is greater than the number of transmitting antennas of the base station, so that scheme 1 cannot be adopted, and therefore the graph of scheme 1 is not given.

The precoding method provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

An embodiment of the present invention provides an encoding apparatus, as shown in fig. 8, the encoding apparatus 6 including: a channel acquisition unit 61, an initialization unit 62 and an iteration unit 63, wherein:

and the channel acquisition unit 61 is configured to acquire a downlink channel matrix.

The initialization unit 62 is configured to obtain an initial weight corresponding to the downlink channel matrix forwarded by the channel acquisition unit 61.

An iteration unit 63, configured to obtain the equalized equivalent channel matrix according to the downlink channel matrix forwarded by the channel acquisition unit 61 and the initial weight forwarded by the initialization unit 62, and obtain a weight corresponding to the next equivalent channel matrix according to the equivalent channel matrix; and performing iterative computation according to the equivalent channel matrix forwarded by the equalization unit 63 and the weight corresponding to the next equivalent channel matrix to obtain a weight corresponding to the downlink channel matrix until a preset iteration number is met or the weight corresponding to the calculated downlink channel matrix meets a preset convergence condition.

Further optionally, as shown in fig. 9, the initialization unit 62 includes:

the singular value decomposition subunit 621a is configured to perform singular value decomposition on the downlink channel matrix forwarded by the channel acquisition unit 61.

A zero forcing calculation subunit 622a, configured to zero-force a right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit 621a to obtain an initial weight corresponding to the downlink channel matrix;

or,

the zero forcing calculation subunit 622a is configured to use the non-zero right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit 621a as an initial weight corresponding to the downlink channel matrix.

Optionally, as shown in fig. 10, the initialization unit 62 includes:

a covariance calculation subunit 621b configured to obtain a downlink channel covariance matrix according to the downlink channel matrix,

an eigenvalue decomposition subunit 622b, further configured to perform eigenvalue decomposition on the downlink channel covariance matrix forwarded by the covariance calculation subunit 621b,

the zero forcing calculation subunit 623b is further configured to use the eigenvalue vector of the downlink channel covariance matrix decomposed by the eigenvalue decomposition subunit 622b as an initial weight corresponding to the downlink channel matrix;

or,

the zero forcing calculation subunit 623b is further configured to use the non-zero eigenvalue vector of the downlink channel covariance matrix decomposed by the eigenvalue decomposition subunit 622b as an initial weight corresponding to the downlink channel matrix.

Further optionally, as shown in fig. 11, the iteration unit 63 further includes:

the equalizing subunit 631a is further configured to obtain an equalized equivalent channel matrix from the covariance matrix of the downlink channel forwarded by the covariance evaluation subunit 621a and the initial weight forwarded by the initialization unit 62,

the weight value calculating operator unit 632a is further configured to obtain a weight value corresponding to the next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalizing sub unit 631 a.

Optionally, as shown in fig. 12, the iteration unit 63 includes:

the covariance calculation subunit 631b obtains a downlink channel covariance matrix according to the downlink channel matrix forwarded by the channel acquisition unit 61.

The equalizing subunit 632b is configured to obtain an equalized equivalent channel matrix from the downlink channel covariance matrix forwarded by the covariance value subunit 631b and the initial weight forwarded by the initialization unit 62.

The weight calculating subunit 633b is configured to obtain a weight corresponding to the next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalizing subunit 632 b.

Optionally, as shown in fig. 13, the iteration unit 63 further includes:

an equivalent matrix obtaining sub-unit 631c for calculating an equivalent channel matrix of each user terminal according to the initial weight forwarded by the initialization unit 62,

the equalizing subunit 632c is further configured to obtain an equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix obtaining subunit 631c and the downlink channel matrix forwarded by the channel acquiring unit 61,

the weight value calculating operator unit 633c is further configured to obtain a weight value corresponding to the next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalizing subunit 632 c.

Optionally, as shown in fig. 14, the iteration unit 63 further includes:

the equivalent matrix acquiring subunit 631d is further configured to calculate an equivalent channel matrix of each ue according to the initial weight forwarded by the initialization unit 62,

the equalizing subunit 632d is further configured to acquire an equalized equivalent channel matrix according to the equivalent channel matrix of each ue forwarded by the equivalent matrix acquiring subunit 631b and the interference noise covariance matrix fed back by each ue,

the weight value calculating sub-unit 633d is further configured to obtain a weight value corresponding to the next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalizing sub-unit 632 d.

Optionally, as shown in fig. 15, the encoding device 6 further includes: an iteration number setting unit 64, configured to obtain a terminal receiver type sent by the user terminal; and setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Optionally, if the iteration number setting unit 64 cannot acquire the terminal receiver type, the preset iteration number is set as the default generation number according to the default terminal receiver type of the user terminal.

The coding device provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

Fig. 16 is a schematic structural diagram of an encoding apparatus according to yet another embodiment of the present invention, where the encoding apparatus 7 includes at least one processor 71, a memory 72, a communication bus 73, and at least one communication interface 74.

The communication bus 73 is used for realizing connection and communication among the above components, and the communication interface 74 is used for connecting and communicating with an external device.

The memory 72 stores program codes to be executed, and the program codes may specifically include: a channel acquisition unit 721, an initialization unit 722 and an iteration unit 723.

The processor 71 is configured to execute the units stored in the memory 72, and when the units are executed by the processor 71, the following functions are realized:

the channel acquisition unit 721 is configured to acquire a downlink channel matrix.

The initialization unit 722 is configured to obtain an initial weight corresponding to the downlink channel matrix forwarded by the channel acquisition unit 721.

An iteration unit 723, configured to obtain the equalized equivalent channel matrix according to the downlink channel matrix forwarded by the channel acquisition unit 721 and the initial weight forwarded by the initialization unit 722, and obtain a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix; and carrying out iterative computation according to the equivalent matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until the preset iteration times are met or the weight corresponding to the downlink channel matrix obtained by computation meets the preset convergence condition.

Further optionally, the initialization unit 722 includes:

and the singular value decomposition subunit is configured to perform singular value decomposition on the downlink channel matrix forwarded by the channel acquisition unit 721.

A zero forcing calculation subunit, configured to zero-force a right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit to obtain an initial weight corresponding to the downlink channel matrix;

or,

the zero-forcing calculation subunit is configured to use the non-zero right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit as an initial weight corresponding to the downlink channel matrix.

Optionally, the initialization unit 722 includes:

and the covariance calculation subunit is used for acquiring the downlink channel covariance matrix according to the downlink channel matrix.

And the eigenvalue decomposition subunit is also used for performing eigenvalue decomposition on the downlink channel covariance matrix forwarded by the covariance calculation subunit.

The zero forcing calculation subunit is also used for taking the eigenvalue vector of the downlink channel covariance matrix after the eigenvalue decomposition subunit decomposition as the initial weight corresponding to the downlink channel matrix;

or,

the zero-forcing calculation subunit is further configured to use the non-zero eigenvalue vector of the downlink channel covariance matrix decomposed by the eigenvalue decomposition subunit as an initial weight corresponding to the downlink channel matrix.

Further optionally, the iteration unit 723 further includes:

the equalization subunit is further configured to obtain an equalized equivalent channel matrix from the downlink channel covariance matrix forwarded by the covariance evaluation subunit and the initial weight forwarded by the initialization unit 722.

And the weight calculation subunit is also used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

Optionally, the iteration unit 723 includes:

and a covariance calculation subunit configured to obtain a downlink channel covariance matrix according to the downlink channel matrix forwarded by the channel acquisition unit 721.

And the equalizing subunit is configured to obtain an equalized equivalent channel matrix from the downlink channel covariance matrix forwarded by the covariance evaluation subunit and the initial weight forwarded by the initialization unit 722.

And a weight calculating subunit, configured to obtain a weight corresponding to the next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalizing subunit 632 b.

Optionally, the iteration unit 723 further includes:

and the equivalent matrix obtaining subunit is configured to calculate an equivalent channel matrix of each ue according to the initial weight forwarded by the initialization unit 722.

The equalizing subunit is further configured to obtain an equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix obtaining subunit and the downlink channel matrix forwarded by the channel acquiring unit 721.

And the weight calculation subunit is also used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

Optionally, the iteration unit 723 further includes:

the equivalent matrix obtaining subunit is further configured to calculate an equivalent channel matrix of each ue according to the initial weight forwarded by the initialization unit 722.

And the equalizing subunit is further configured to acquire the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix acquiring subunit and the interference noise covariance matrix fed back by each user terminal.

And the weight calculation subunit is also used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

Optionally, the memory 72 further comprises: an iteration number setting unit 724 configured to obtain a terminal receiver type sent by the user terminal; and setting the iteration times corresponding to the type of the terminal receiver as preset iteration times.

Optionally, if the iteration number setting unit 724 cannot acquire the terminal receiver type, the preset iteration number is set as the default generation number according to the default terminal receiver type of the user terminal.

The coding device provided by the embodiment of the invention only needs to carry out one-time eigenvalue decomposition in the process of calculating the initial weight corresponding to the current downlink channel matrix, then carries out subsequent iterative calculation according to the initial weight and obtains the weight corresponding to the downlink channel matrix, is not limited by the number of antennas of receiving end equipment, and simultaneously can reduce the calculation complexity of channel precoding and ensure the emission performance gain of a channel.

Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (18)

1. A precoding method, comprising:

acquiring a downlink channel matrix;

acquiring an initial weight corresponding to the downlink channel matrix;

obtaining an equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix;

and carrying out iterative computation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until a preset iteration number is met or the calculated weight corresponding to the downlink channel matrix meets a preset convergence condition.

2. The method according to claim 1, wherein the obtaining the initial weight corresponding to the downlink channel matrix comprises:

performing singular value decomposition on the downlink channel matrix;

zero forcing is carried out on the decomposed right singular vector of the downlink channel matrix to obtain an initial weight corresponding to the downlink channel matrix;

or,

and taking the decomposed non-zero right singular vector of the downlink channel matrix as an initial weight corresponding to the downlink channel matrix.

3. The method according to claim 1, wherein the obtaining the initial weight corresponding to the downlink channel matrix comprises:

acquiring a downlink channel covariance matrix according to the downlink channel matrix;

performing eigenvalue decomposition on the downlink channel covariance matrix;

zero forcing is carried out on the eigenvalue vector of the decomposed downlink channel covariance matrix to obtain an initial weight corresponding to the downlink channel matrix;

or,

and taking the decomposed non-zero eigenvalue vector of the downlink channel covariance matrix as an initial weight corresponding to the downlink channel matrix.

4. The method of claim 3, wherein the obtaining the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix comprises:

obtaining an equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

5. The method of claim 1, wherein the obtaining the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix comprises:

acquiring a downlink channel covariance matrix according to the downlink channel matrix;

obtaining an equalized equivalent channel matrix according to the downlink channel covariance matrix and the initial weight;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

6. The method of claim 1, wherein the obtaining the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix comprises:

calculating an equivalent channel matrix of each user terminal according to the initial weight;

obtaining the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal and the downlink channel matrix;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

7. The method of claim 1, wherein the obtaining the equalized equivalent channel matrix according to the downlink channel matrix and the initial weight and obtaining a weight corresponding to a next equivalent channel matrix according to the equivalent channel matrix comprises:

calculating an equivalent channel matrix of each user terminal according to the initial weight;

obtaining the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal and the interference noise covariance matrix fed back by each user terminal;

and acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero forcing algorithm according to the equalized equivalent channel matrix.

8. The method according to any one of claims 1 to 7, wherein performing iterative computation according to the equivalent matrix and the weight corresponding to the next equivalent channel matrix to obtain a weight corresponding to a next downlink channel matrix until a preset iteration number is met or a weight corresponding to the next downlink channel matrix obtained by the iterative computation converges to meet a preset convergence condition, further comprises:

acquiring the type of a terminal receiver sent by a user terminal;

and setting the iteration times corresponding to the type of the terminal receiver as the preset iteration times.

9. The method of claim 8, wherein if the terminal receiver type cannot be obtained, the preset number of iterations is set as a default number of iterations according to a default terminal receiver type of the user terminal.

10. An encoding device, characterized by comprising:

the channel acquisition unit is used for acquiring a downlink channel matrix;

the initialization unit is used for acquiring an initial weight corresponding to the downlink channel matrix forwarded by the channel acquisition unit;

the iteration unit is used for acquiring the equalized equivalent channel matrix according to the downlink channel matrix forwarded by the channel acquisition unit and the initial weight forwarded by the initialization unit and acquiring a weight corresponding to the next equivalent channel matrix according to the equivalent channel matrix; and carrying out iterative computation according to the equivalent channel matrix and the weight corresponding to the next equivalent channel matrix to obtain the weight corresponding to the downlink channel matrix until a preset iteration number is met or the calculated weight corresponding to the downlink channel matrix meets a preset convergence condition.

11. The encoding device according to claim 10, wherein the initialization unit includes:

a singular value decomposition subunit, configured to perform singular value decomposition on the downlink channel matrix forwarded by the channel acquisition unit;

a zero forcing calculation subunit, configured to zero-force a right singular vector of the decomposed downlink channel matrix forwarded by the singular value decomposition subunit to obtain an initial weight corresponding to the downlink channel matrix;

or,

and the zero forcing calculation subunit is configured to use the decomposed non-zero right singular vector of the downlink channel matrix forwarded by the singular value decomposition subunit as an initial weight corresponding to the downlink channel matrix.

12. The encoding device according to claim 10, wherein the initialization unit includes:

the covariance calculation subunit is used for acquiring a downlink channel covariance matrix according to the downlink channel matrix;

the eigenvalue decomposition subunit is further configured to perform eigenvalue decomposition on the downlink channel covariance matrix forwarded by the covariance calculation subunit;

the zero forcing calculation subunit is further configured to use the eigenvalue vector of the downlink channel covariance matrix after the eigenvalue decomposition subunit decomposes as an initial weight corresponding to the downlink channel matrix;

or,

the zero-forcing calculation subunit is further configured to use the non-zero eigenvalue vector of the downlink channel covariance matrix after the eigenvalue decomposition subunit decomposes as an initial weight corresponding to the downlink channel matrix.

13. The encoding device according to claim 12, wherein the iteration unit includes:

the equalizing subunit is further configured to obtain an equalized equivalent channel matrix from the downlink channel covariance matrix forwarded by the covariance calculating subunit and the initial weight forwarded by the initialization unit;

and the weight calculation subunit is further configured to obtain a weight corresponding to a next equivalent channel matrix by using a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

14. The encoding device according to claim 10, wherein the iteration unit includes:

the covariance calculation subunit is also used for acquiring a downlink channel covariance matrix according to the downlink channel matrix forwarded by the channel acquisition unit;

the equalization subunit is configured to obtain an equalized equivalent channel matrix according to the downlink channel covariance matrix forwarded by the covariance evaluation subunit and the initial weight forwarded by the initialization unit;

and the weight calculation subunit is used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

15. The encoding device according to claim 10, wherein the iteration unit includes:

an equivalent matrix obtaining subunit, configured to calculate an equivalent channel matrix of each user terminal according to the initial weight forwarded by the initialization unit;

the equalization subunit is configured to acquire the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix acquisition subunit and the downlink channel matrix forwarded by the channel acquisition unit;

and the weight calculation subunit is used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

16. The encoding device according to claim 10, wherein the iteration unit includes:

an equivalent matrix obtaining subunit, configured to calculate an equivalent channel matrix of each user terminal according to the initial weight forwarded by the initialization unit;

the equalizing subunit is configured to obtain the equalized equivalent channel matrix according to the equivalent channel matrix of each user terminal forwarded by the equivalent matrix obtaining subunit and the interference noise covariance matrix fed back by each user terminal;

and the weight calculation subunit is used for acquiring a weight corresponding to the next equivalent channel matrix by adopting a zero-forcing algorithm according to the equalized equivalent channel matrix forwarded by the equalization subunit.

17. The encoding apparatus according to any one of claims 10 to 16, characterized by further comprising:

the iteration number setting unit is used for acquiring the type of the terminal receiver sent by the user terminal; and setting the iteration times corresponding to the type of the terminal receiver as the preset iteration times.

18. The encoding device according to claim 17, wherein the iteration number setting unit is further configured to set the preset iteration number as a default iteration number according to a terminal receiver type of a default user terminal if the iteration number setting unit cannot acquire the terminal receiver type.

Images