CN101931450B - Method for selecting codebooks and device thereof - Google Patents

Abstract

The embodiment of the invention discloses a method for selecting codebooks and a device thereof. The method comprises the following steps of: obtaining an element on a diagonal line in an inverse matrix of a measurement matrix which corresponds to each codebook according to an adjoint matrix method; obtaining the measurement values of each codebook according to the element on the diagonal line in the inverse matrix of the measurement matrix; selecting an optimal measurement value from the measurement values of each codebook according to a presetting principle, and taking the code book which corresponds to the optimal measurement value as an optimal codebook. The embodiment of the invention can reduce the selection time of the codebooks, thereby saving the system resources.

Description

Method and device for selecting codebook

Technical Field

The present invention relates to the field of wireless communication technologies, and in particular, to a method and an apparatus for codebook selection.

Background

In the closed-loop transmission technology of LTE (Long Term Evolution), after a terminal estimates a channel, an optimal codebook is selected and fed back to a base station, and the base station linearly adjusts a signal on a transmitting antenna according to the received optimal codebook. The method for selecting the optimal codebook by the terminal mainly comprises a codebook selection method based on measurement and a codebook selection method based on singular value decomposition. Among them, in the codebook selection method based on the measurementIn the method, the metric value corresponding to each codebook in the codebook set needs to be obtained first, the optimal metric value is selected from the metric values of all codebooks according to a preset principle, and then the codebook corresponding to the optimal metric value is selected as the optimal codebook. Here, in the process of obtaining the metric value corresponding to each codebook in the codebook set, the metric matrix W needs to be solved^HH^HHW+σ²I_MDiagonal elements in the inverse matrix of (c). In the prior art, an inverse matrix of a measurement matrix is solved by a cyclic variable numbering method, and then elements on a diagonal line in the inverse matrix are extracted.

However, the inventor has found in research that, in the process of obtaining the metric value corresponding to each codebook in the codebook set, only the element on the diagonal in the inverse matrix of the metric matrix needs to be used. However, the cyclic renumbering method of variables requires that the inverse matrix of the entire metric matrix is completely calculated and then the diagonal elements are extracted, which makes the process of obtaining the metric values computationally expensive. For example, when the metric matrix is a 4 × 4 matrix, the computation of the remaining 12 elements is actually wasted, except for the 4 elements on the diagonal of the inverse matrix. Finally, the whole codebook selection process wastes system resources greatly due to time consumption.

Disclosure of Invention

The embodiment of the invention provides a method and a device for selecting a codebook, which are used for saving system resources.

The embodiment of the invention discloses a method for selecting a codebook, which comprises the following steps: obtaining elements on the diagonal line in the inverse matrix of the measurement matrix corresponding to each codebook according to an adjoint matrix method; obtaining the metric value of each codebook according to the elements on the diagonal line in the inverse matrix of the metric matrix; and selecting an optimal metric value from the metric values of each codebook according to a preset principle, and taking the codebook corresponding to the optimal metric value as an optimal codebook.

Preferably, the obtaining, according to the adjoint matrix method, elements on a diagonal line in an inverse matrix of the metric matrix corresponding to each codebook includes: when RI corresponding to a codebook is a non-full rank, determinants of measurement matrixes respectively corresponding to the codebook are obtained according to input noise, a channel estimation matrix and the codebook, algebraic remainder quotients of diagonal elements in the measurement matrixes are subjected to quotient calculation with the determinants of the measurement matrixes, and elements on diagonals in an inverse matrix of the measurement matrix corresponding to the codebook are obtained; when RI corresponding to a codebook is a full rank, obtaining a determinant of a measurement matrix corresponding to the codebook according to input noise and a channel estimation matrix, and carrying out quotient calculation on an algebraic residue sub-formula of diagonal elements in the measurement matrix and the determinant of the measurement matrix to obtain diagonal elements in an inverse matrix of the measurement matrix corresponding to the codebook.

Preferably, the obtaining the metric value of each codebook according to the diagonal elements in the inverse matrix of the metric matrix includes: obtaining the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix; setting the signal-to-noise ratio of each codebook as a metric value of the codebook; or, the mean square error of each codebook is obtained according to the elements on the diagonal line in the inverse matrix of the measurement matrix, and the mean square error of each codebook is set as the measurement value of the codebook.

Preferably, the selecting an optimal metric value from the metric values of each codebook according to a preset principle, and using the codebook corresponding to the optimal metric value as an optimal codebook includes: selecting the maximum signal-to-noise ratio from the signal-to-noise ratios of all codebooks according to the signal-to-noise ratio maximization principle; taking the codebook corresponding to the maximum signal-to-noise ratio as an optimal codebook; or selecting the minimum mean square error from the mean square errors of each codebook according to the mean square error minimization principle, and taking the codebook corresponding to the minimum mean square error as the optimal codebook.

Preferably, the method further comprises: and feeding back the optimal codebook to the base station.

Preferably, the feeding back the optimal codebook to the base station includes: setting a codebook number of each codebook together with the base station in advance; and sending the code book number of the optimal code book to the base station, and obtaining the optimal code book by the base station according to the corresponding relation between the stored code book number and the code book.

The embodiment of the invention also discloses a device for selecting the codebook, which comprises the following steps: an element obtaining unit, configured to obtain, according to an adjoint matrix method, elements on a diagonal line in an inverse matrix of a metric matrix corresponding to each codebook; a metric value obtaining unit, configured to obtain a metric value of each codebook according to an element on a diagonal line in an inverse matrix of the metric matrix; and the selection unit is used for selecting an optimal metric value from the metric values of each codebook according to a preset principle, and taking the codebook corresponding to the metric value as an optimal codebook.

Preferably, the element obtaining unit includes: the first obtaining subunit is configured to, when an RI corresponding to a codebook is a non-full rank, obtain, according to input noise, a channel estimation matrix, and the codebook, a determinant of a metric matrix corresponding to each codebook, and quotient an algebraic residue of a diagonal element in the metric matrix with the determinant of the metric matrix to obtain an element on a diagonal in an inverse matrix of the metric matrix corresponding to each codebook; and the second acquiring subunit is used for acquiring a determinant of the measurement matrix corresponding to each codebook according to the input noise and the channel estimation matrix when the RI corresponding to the codebook is a full rank, and obtaining elements on a diagonal line in an inverse matrix of the measurement matrix corresponding to each codebook by taking a quotient of an algebraic remainder of diagonal line elements in the measurement matrix and the determinant of the measurement matrix.

Preferably, the metric value obtaining unit includes: the signal-to-noise ratio obtaining subunit obtains the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix; the signal-to-noise ratio setting subunit is used for setting the signal-to-noise ratio of each codebook as the metric value of the codebook; or, the mean square error obtaining subunit is configured to obtain a mean square error of each codebook according to an element on a diagonal line in the inverse matrix of the measurement matrix, and the mean square error setting subunit is configured to set the mean square error of each codebook as a measurement value of the codebook.

Preferably, the selection unit includes: a maximum signal-to-noise ratio selecting subunit, configured to select a maximum signal-to-noise ratio from the signal-to-noise ratios of each codebook according to a signal-to-noise ratio maximization principle; a maximum signal-to-noise ratio setting subunit, configured to use the codebook corresponding to the maximum signal-to-noise ratio as an optimal codebook; or, the minimum mean square error selecting subunit is used for selecting the minimum mean square error from the mean square errors of each codebook according to a mean square error minimization principle; and the minimum mean square error setting subunit is used for taking the codebook corresponding to the minimum mean square error as the optimal codebook.

Preferably, the apparatus further comprises: and the feedback unit is used for feeding back the optimal codebook to the base station.

Preferably, the feedback unit includes: a codebook number setting subunit, configured to set a codebook number of each codebook together with the base station in advance; and the sending subunit is configured to send the codebook number of the optimal codebook to the base station, and the base station obtains the optimal codebook according to the stored correspondence between the codebook number and the codebook.

It can be seen from the above embodiments that, the adjoint matrix method is used to solve the diagonal elements in the inverse matrix of the metric matrix, and each element in the inverse matrix can be relatively independently calculated by the adjoint matrix method, without calculating all elements of the whole inverse matrix as in the variable cyclic renumbering method, thereby reducing the amount of calculation, saving the time for selecting the codebook, and further saving the system resources.

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, and 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 these drawings without creative efforts.

FIG. 1 is a flow diagram of one embodiment of a method of codebook selection of the present invention;

FIG. 2 is a diagram illustrating a mean square error estimation based codebook selection method according to the present invention;

FIG. 3 is a flow chart of another embodiment of a method of codebook selection of the present invention;

FIG. 4 is a block diagram of an apparatus for codebook selection according to an embodiment of the present invention.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.

Example one

Please refer to fig. 1, which is a flowchart illustrating a codebook selection method according to an embodiment of the present invention, the method includes the following steps:

step 101: obtaining elements on the diagonal line in the inverse matrix of the measurement matrix corresponding to each codebook according to an adjoint matrix method;

wherein, the obtaining of the diagonal elements in the inverse matrix of the metric matrix corresponding to each codebook according to the adjoint matrix method includes:

when an RI (rank indicator) corresponding to a codebook is a non-full rank, obtaining determinants of measurement matrices respectively corresponding to the codebook according to input noise, a channel estimation matrix and the codebook, and taking a quotient of an algebraic remainder of diagonal elements in the measurement matrices and the determinants of the measurement matrices to obtain diagonal elements in an inverse matrix of the measurement matrix corresponding to the codebook;

when RI corresponding to a codebook is a full rank, obtaining a determinant of a measurement matrix corresponding to the codebook according to input noise and a channel estimation matrix, and carrying out quotient calculation on an algebraic residue sub-formula of diagonal elements in the measurement matrix and the determinant of the measurement matrix to obtain diagonal elements in an inverse matrix of the measurement matrix corresponding to the codebook.

Step 102: obtaining the metric value of each codebook according to the elements on the diagonal line in the inverse matrix of the metric matrix;

wherein the obtaining the metric value of each codebook according to the diagonal elements in the inverse matrix of the metric matrix comprises: obtaining the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix; setting the signal-to-noise ratio of each codebook as a metric value of the codebook; or, the mean square error of each codebook is obtained according to the elements on the diagonal line in the inverse matrix of the measurement matrix, and the mean square error of each codebook is set as the measurement value of the codebook.

Step 103: and selecting an optimal metric value from the metric values of each codebook according to a preset principle, and taking the codebook corresponding to the optimal metric value as an optimal codebook.

Selecting an optimal metric value from the metric values of each codebook according to a preset principle, wherein taking the codebook corresponding to the optimal metric value as an optimal codebook comprises: selecting the maximum signal-to-noise ratio from the signal-to-noise ratios of all codebooks according to the signal-to-noise ratio maximization principle; taking the codebook corresponding to the maximum signal-to-noise ratio as an optimal codebook; or selecting the minimum mean square error from the mean square errors of each codebook according to the mean square error minimization principle, and taking the codebook corresponding to the minimum mean square error as the optimal codebook.

Further, the method further comprises: and feeding back the optimal codebook to the base station. The method specifically comprises the following steps: setting a codebook number of each codebook together with the base station in advance; and sending the code book number of the optimal code book to the base station, and obtaining the optimal code book by the base station according to the corresponding relation between the stored code book number and the code book.

It can be seen from the above embodiments that, the adjoint matrix method is used to solve the diagonal elements in the inverse matrix of the metric matrix, and each element in the inverse matrix can be relatively independently calculated by the adjoint matrix method, without calculating all elements of the whole inverse matrix as in the variable cyclic renumbering method, thereby reducing the amount of calculation, saving the time for selecting the codebook, and further saving the system resources.

In addition, when the RI corresponding to the codebook is a full rank, the determinant of the measurement matrix of each codebook can be directly obtained by the channel estimation matrix and the input noise, thereby further reducing the calculation amount, saving the time for selecting the codebook and saving the system resources.

Example two

Please refer to fig. 2, which is a schematic diagram of a codebook selection method based on mean square error estimation of the present invention, and referring to fig. 3 in combination with fig. 2, which is a structural diagram of another embodiment of a codebook selection method of the present invention, and takes mean square error estimation as an example to describe in detail the method for a terminal to select a codebook, including the following steps:

step 301: calculating a metric matrix W corresponding to each codebook according to the input noise, the channel estimation matrix and the codebook^HH^HHW+σ²I_M；

If the LTE system has 2 antennas, when RI corresponding to the codebook is 1, the metric matrix of the codebook is reduced to a real number, and when RI corresponding to the codebook is 2, the metric matrix of the codebook is a 2 × 2 matrix.

If the LTE system has 4 antennas, in order to reduce the calculation amount in the process of calculating the codebook matrix, the following process may be performed to calculate the metric matrices when RI is 1, RI is 2, RI is 3, and RI is 4 in sequence, and a is used below^(R)Metric matrix when RI ═ R, a_ij ^(R)Representation matrix A^(R)Row i and column j of (1), let B^(R)＝W^HH^HHW, where codebook W is not normalized, u_iN-th representing a codebook W₁The ith element of the column, h_iThe ith column of the matrix H is denoted, i 1, 2, 3, 4. Wherein,

when the RI is 1, the first signal is,

$B^{\left(1\right)}=W^H{H}^H\mathrm{HW}={\left|\right|h_1\left|\right|}^2-\mathrm{real}\left(h_1^H\left(\mathrm{Hu}\right)u_1^{*}\right)+\left|\right|\mathrm{Hu}\left|\right|{\left|u_1\right|}^2/4;$ A⁽¹⁾＝B⁽¹⁾+σ²

when the RI is 2, the first signal is,

$B^{\left(2\right)}=W^H{H}^H\mathrm{HW}=\left[\begin{array}{cc}{\left|\right|h_{n_1}\left|\right|}^2& h_{n_1}^H{h}_{n_2}\\ h_{n_2}^H{h}_{n_1}& {\left|\right|h_{n_2}\left|\right|}^2\end{array}\right]-\mathrm{real}\left(\begin{array}{c}\left[\begin{array}{cc}{u}_{n_1}^{*}{h}_{n_1}^H\mathrm{Hu}& u_{n_2}^{*}{h}_{n_1}^H\mathrm{Hu}\\ u_{n_1}^{*}{h}_{n_2}^H\mathrm{Hu}& u_{n_2}^{*}{h}_{n_2}^H\mathrm{Hu}\end{array}\right]\end{array}\right)+\left|\right|\mathrm{Hu}\left|\right|\left[\begin{array}{cc}{\left|\right|u_{n_1}\left|\right|}^2& u_{n_1}{u}_{n_2}^{*}\\ u_{n_2}{u}_{n_1}^{*}& {\left|\right|u_{n_2}\left|\right|}^2\end{array}\right]/4$

from the nested structure of the matrices $A_{11}^{\left(2\right)}=A_{11}^{\left(1\right)},$ Considering the conjugate symmetry again, only calculation is needed

<math> <mrow> <msup> <mi>A</mi> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mfrac> <mn>1</mn> <mn>2</mn> </mfrac> <msup> <mi>B</mi> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </msup> <mo>+</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>2</mn> </msub> </mrow> </math>

The element of the run-up is already calculated when RI is 1, and can be directly used without being determined again. Beat and beat

Are respectively conjugated with elements symmetrical about the diagonal.

When the RI is 3, the number of bits is 3,

$B^{\left(3\right)}=\left[\begin{array}{ccc}{\left|\right|h_{n_1}\left|\right|}^2& h_{n_1}^H{h}_{n_2}& h_{n_1}^H{h}_{n_3}\\ h_{n_2}^H{h}_{n_1}& {\left|\right|h_{n_2}\left|\right|}^2& h_{n_2}^H{h}_{n_3}\\ h_{n_3}^H{h}_{n_1}& h_{n_3}^H{h}_{n_2}& {\left|\right|h_{n_3}\left|\right|}^2\end{array}\right]-\mathrm{real}\left(\left[\begin{array}{c}{h}_{n_1}^H\\ h_{n_2}^H\\ h_{n_3}^H\end{array}\right]\left(\mathrm{Hu}\right)\left[\begin{array}{ccc}{u}_{n_1}^{*}& u_{n_2}^{*}& u_{n_3}^{*}\end{array}\right]\right)+\left|\right|\mathrm{Hu}\left|\right|\left[\begin{array}{ccc}{\left|\right|u_{n_1}\left|\right|}^2& u_{n_1}{u}_{n_2}^{*}& u_{n_1}{u}_{n_3}^{*}\\ u_{n_2}{u}_{n_1}^{*}& {\left|\right|u_{n_2}\left|\right|}^2& u_{n_2}{u}_{n_3}^{*}\\ u_{n_3}{u}_{n_1}^{*}& u_{n_3}{u}_{n_2}^{*}& {\left|\right|u_{n_3}\left|\right|}^2\end{array}\right]/4,$

the element of the run-up is already calculated when RI is 2, and can be directly used without being determined again.

Beat and beat

Are respectively conjugated with elements symmetrical about the diagonal.

<math> <mrow> <msup> <mi>A</mi> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mfrac> <mn>1</mn> <mn>3</mn> </mfrac> <msup> <mi>B</mi> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </msup> <mo>+</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>3</mn> </msub> </mrow> </math>

If the two columns of elements of the codebook W corresponding to the case where RI is 2 are not the first two columns of the codebook W corresponding to the case where RI is 3, only the corresponding elements need to be selected.

When the RI is 4, the number of bits is 4,

<math> <mrow> <msup> <mi>A</mi> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mfrac> <mn>1</mn> <mn>4</mn> </mfrac> <msup> <mi>B</mi> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </msup> <mo>+</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> </mrow> </math>

the element of the run-up is already calculated when RI is 3, and can be directly used without being obtained again. Beat and beat

Are respectively conjugated with elements symmetrical about the diagonal.

Step 302: judging whether RI corresponding to each codebook is a full rank, if so, entering step 303, and if not, entering step 304;

step 303: obtaining a determinant of a metric matrix corresponding to the codebook according to the input noise and the channel estimation matrix, and entering step 305;

wherein, when the RI corresponding to the codebook is a full rank, the sum is given by | I + AB | ═ I + BA |, and $\frac{1}{4}{\mathrm{WW}}^H=I_4$ it is possible to obtain,

<math> <mrow> <mo>|</mo> <mi>A</mi> <mo>|</mo> <mo>=</mo> <mo>|</mo> <mfrac> <mn>1</mn> <mn>4</mn> </mfrac> <msup> <mi>B</mi> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </msup> <mo>+</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> <mo>|</mo> </mrow> </math>

<math> <mrow> <mo>=</mo> <mo>|</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> <mo>+</mo> <mfrac> <mn>1</mn> <mn>4</mn> </mfrac> <msup> <mi>W</mi> <mi>H</mi> </msup> <msup> <mi>H</mi> <mi>H</mi> </msup> <mi>HW</mi> <mo>|</mo> </mrow> </math>

<math> <mrow> <mo>=</mo> <mo>|</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> <mo>+</mo> <mfrac> <mn>1</mn> <mn>4</mn> </mfrac> <mi>HW</mi> <msup> <mi>W</mi> <mi>H</mi> </msup> <msup> <mi>H</mi> <mi>H</mi> </msup> <mo>|</mo> </mrow> </math>

<math> <mrow> <mo>=</mo> <mo>|</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> <mo>+</mo> <msup> <mi>HH</mi> <mi>H</mi> </msup> <mo>|</mo> </mrow> </math>

<math> <mrow> <mo>=</mo> <mo>|</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mn>4</mn> </msub> <mo>+</mo> <msup> <mi>H</mi> <mi>H</mi> </msup> <mi>H</mi> <mo>|</mo> </mrow> </math>

therefore, when the RI is a full rank, the determinant of the metric matrix is independent of the codebook, and the determinants of the metric matrices of all codebooks are the same and only need to be calculated once.

Step 304: obtaining determinants of measurement matrixes respectively corresponding to the codebooks according to the input noise, the channel estimation matrix and the codebooks;

when RI is not full rank, the determinant of the metric matrix is related to the codebook, and when the codebooks are different, the determinant of the codebook metric matrix is also different, and therefore, the determinant of the metric matrix of each codebook needs to be calculated.

Step 305: the algebraic remainder formula of diagonal elements in the measurement matrix and the determinant of the measurement matrix are subjected to quotient calculation to obtain diagonal elements in an inverse matrix of the measurement matrix corresponding to the codebook;

step 306: obtaining the mean square error of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix;

wherein, when the codebook selection is carried out based on the mean square error, the codebook selection is carried out by a formula <math> <mrow> <mover> <mi>MSE</mi> <mo>&OverBar;</mo> </mover> <mrow> <mo>(</mo> <mi>W</mi> <mo>)</mo> </mrow> <mo>=</mo> <msup> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>M</mi> </msub> <mo>+</mo> <mfrac> <mrow> <msup> <mi>W</mi> <mi>H</mi> </msup> <msup> <mi>H</mi> <mi>H</mi> </msup> <mi>HW</mi> </mrow> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> </mfrac> <mo>)</mo> </mrow> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> </mrow> </math> The mean square error of each codebook is obtained.

When codebook selection is performed based on signal-to-noise ratio, the codebook selection is performed by formula <math> <mrow> <msubsup> <mi>SNR</mi> <mi>k</mi> <mi>MMSE</mi> </msubsup> <mo>=</mo> <mfrac> <mn>1</mn> <mrow> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msubsup> <mrow> <mo>[</mo> <msup> <mi>W</mi> <mi>H</mi> </msup> <msup> <mi>H</mi> <mi>H</mi> </msup> <mi>HW</mi> <mo>+</mo> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <msub> <mi>I</mi> <mi>M</mi> </msub> <mo>]</mo> </mrow> <mrow> <mi>k</mi> <mo>,</mo> <mi>k</mi> </mrow> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msubsup> </mrow> </mfrac> <mo>-</mo> <mn>1</mn> </mrow> </math> And acquiring the signal-to-noise ratio of each codebook, wherein k is the index of each subcarrier in the granularity range.

Step 307: comparing the mean square error of each codebook, and taking the codebook corresponding to the minimum mean square error as an optimal codebook;

wherein the codebook selection based on mean square error MSE can be expressed as:

<math> <mrow> <msub> <mi>W</mi> <mi>opt</mi> </msub> <mo>=</mo> <munder> <mrow> <mi>arg</mi> <mi>min</mi> </mrow> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>&Element;</mo> <mi>W</mi> </mrow> </munder> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>0</mn> </mrow> <mrow> <mi>K</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <mi>tr</mi> <mrow> <mo>(</mo> <msub> <mi>MSE</mi> <mi>k</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow> </math>

the codebook selection based on SNR can be expressed as:

<math> <mrow> <msub> <mi>W</mi> <mi>opt</mi> </msub> <mo>=</mo> <munder> <mrow> <mi>arg</mi> <mi>max</mi> </mrow> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>&Element;</mo> <mi>W</mi> </mrow> </munder> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>0</mn> </mrow> <mrow> <mi>K</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msubsup> <mi>SNR</mi> <mi>k</mi> <mi>MMSE</mi> </msubsup> <mrow> <mo>(</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> </mrow> </math>

wherein K is the number of subcarriers included in the feedback granularity, K is the index of each subcarrier in the granularity range, W is the codebook, and tr (·) is the trace.

Step 308: and sending the code book number of the optimal code book to the base station.

The method comprises the steps that a base station and a terminal are jointly provided with a codebook number of each codebook in advance, and after the terminal sends the codebook number of the optimal codebook to the base station, the base station obtains the optimal codebook according to the corresponding relation between the stored codebook numbers and the codebooks.

It can be seen from the above embodiments that, the adjoint matrix method is used to solve the diagonal elements in the inverse matrix of the metric matrix, and each element in the inverse matrix can be relatively independently calculated by the adjoint matrix method, without calculating all elements of the whole inverse matrix as in the variable cyclic renumbering method, thereby reducing the amount of calculation, saving the time for selecting the codebook, and further saving the system resources.

In addition, when the RI corresponding to the codebook is a full rank, the determinant of the measurement matrix of each codebook can be directly obtained by the channel estimation matrix and the input noise, thereby further reducing the calculation amount, saving the time for selecting the codebook and saving the system resources.

EXAMPLE III

Corresponding to the method for selecting the codebook, the embodiment of the invention also provides a device for selecting the codebook. Please refer to fig. 4, which is a block diagram of an apparatus for codebook selection according to an embodiment of the present invention, the apparatus includes: an element obtaining unit 401, a metric value obtaining unit 402 and a selecting unit 403. The internal structure and connection relationship of the device will be further described below in conjunction with the working principle of the device.

An element obtaining unit 401, configured to obtain, according to an adjoint matrix method, elements on a diagonal line in an inverse matrix of a metric matrix corresponding to each codebook;

a metric value obtaining unit 402, configured to obtain a metric value of each codebook according to an element on a diagonal line in an inverse matrix of the metric matrix;

a selecting unit 403, configured to select an optimal metric value from the metric values of each codebook according to a preset principle, and use a codebook corresponding to the metric value as an optimal codebook.

Wherein the element obtaining unit 401 includes: a first acquisition sub-unit 4011 and a second acquisition sub-unit 4012,

a first obtaining subunit 4011, configured to, when RI corresponding to a codebook is a non-full rank, obtain, according to input noise, a channel estimation matrix and the codebook, determinants of a metric matrix corresponding to each codebook, and quotient an algebraic residue of a diagonal element in the metric matrix with the determinant of the metric matrix, to obtain an element on a diagonal in an inverse matrix of the metric matrix corresponding to each codebook;

a second obtaining subunit 4012, configured to, when the RI corresponding to the codebook is a full rank, obtain a determinant of a metric matrix corresponding to each codebook according to the input noise and the channel estimation matrix, and quotient an algebraic residue of a diagonal element in the metric matrix with the determinant of the metric matrix, to obtain an element on a diagonal in an inverse matrix of the metric matrix corresponding to each codebook.

The metric value obtaining unit 402 may include: a noise ratio obtaining sub-unit 4021 and a noise ratio setting sub-unit 4022,

a signal-to-noise ratio obtaining subunit 4021, which obtains the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix;

a signal-to-noise ratio setting subunit 4022, configured to set a signal-to-noise ratio of each codebook as a metric of the codebook.

Alternatively, the metric value obtaining unit 402 may also include: a mean square error obtaining subunit, configured to obtain a mean square error of each codebook according to an element on a diagonal line in an inverse matrix of the measurement matrix, and a mean square error setting subunit, configured to set the mean square error of each codebook as a metric value of the codebook.

The selecting unit 403 may include: a maximum signal-to-noise ratio selection subunit 4031 and a maximum signal-to-noise ratio setting subunit 4032,

a maximum snr selecting subunit 4031, configured to select a maximum snr from the snrs of each codebook according to a snr maximization principle;

a maximum snr setting subunit 4032, configured to use the codebook corresponding to the maximum snr as an optimal codebook;

alternatively, the selection unit 403 may include: the minimum mean square error selection subunit is used for selecting the minimum mean square error from the mean square errors of each codebook according to a mean square error minimization principle; and the minimum mean square error setting subunit is used for taking the codebook corresponding to the minimum mean square error as the optimal codebook.

Further, the apparatus further comprises: a feedback unit 404, configured to feed back the optimal codebook to the base station. Among them, the feedback unit 404 includes: a codebook number setting sub-unit 4041 and a transmitting sub-unit 4042,

a codebook number setting subunit 4041, configured to set a codebook number of each codebook together with the base station in advance;

the sending subunit 4042 is configured to send the codebook number of the optimal codebook to the base station, and the base station obtains the optimal codebook according to the correspondence between the stored codebook number and the codebook.

It can be seen from the above embodiments that, the adjoint matrix method is used to solve the diagonal elements in the inverse matrix of the metric matrix, and each element in the inverse matrix can be relatively independently calculated by the adjoint matrix method, without calculating all elements of the whole inverse matrix as in the variable cyclic renumbering method, thereby reducing the amount of calculation, saving the time for selecting the codebook, and further saving the system resources.

In addition, when the RI corresponding to the codebook is a full rank, the determinant of the measurement matrix of each codebook can be directly obtained by the channel estimation matrix and the input noise, thereby further reducing the calculation amount, saving the time for selecting the codebook and saving the system resources.

It should be noted that, as will be understood by those skilled in the art, all or part of the processes in the methods of the above embodiments may be implemented by a computer program, which may be stored in a computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

The method and apparatus for selecting a codebook provided by the present invention are described in detail above, and the principles and embodiments of the present invention are explained herein by applying specific embodiments, and the descriptions of the above embodiments are only used to help understanding the method and core ideas 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 (10)

1. A method for codebook selection in an LTE system, comprising:

obtaining elements on the diagonal line in the inverse matrix of the measurement matrix corresponding to each codebook according to an adjoint matrix method;

obtaining the metric value of each codebook according to the elements on the diagonal line in the inverse matrix of the metric matrix;

selecting an optimal metric value from the metric values of each codebook according to a preset principle, and taking the codebook corresponding to the optimal metric value as an optimal codebook;

the obtaining of the diagonal elements in the inverse matrix of the metric matrix corresponding to each codebook according to the adjoint matrix method includes:

when RI corresponding to a codebook is a non-full rank, determinants of measurement matrixes respectively corresponding to the codebook are obtained according to input noise, a channel estimation matrix and the codebook, algebraic remainder quotients of diagonal elements in the measurement matrixes are subjected to quotient calculation with the determinants of the measurement matrixes, and elements on diagonals in an inverse matrix of the measurement matrix corresponding to the codebook are obtained;

when RI corresponding to a codebook is a full rank, obtaining a determinant of a measurement matrix corresponding to the codebook according to input noise and a channel estimation matrix, and carrying out quotient calculation on an algebraic residue sub-formula of diagonal elements in the measurement matrix and the determinant of the measurement matrix to obtain diagonal elements in an inverse matrix of the measurement matrix corresponding to the codebook.

2. The method of claim 1, wherein deriving the metric value for each codebook from diagonal elements in an inverse matrix of a metric matrix comprises:

obtaining the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix;

setting the signal-to-noise ratio of each codebook as a metric value of the codebook;

or,

obtaining the mean square error of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix,

setting the mean square error of each codebook as the metric value of the codebook.

3. The method of claim 2, wherein the selecting an optimal metric value from the metric values of each codebook according to a preset rule, and the using the codebook corresponding to the optimal metric value as the optimal codebook comprises:

selecting the maximum signal-to-noise ratio from the signal-to-noise ratios of all codebooks according to the signal-to-noise ratio maximization principle;

taking the codebook corresponding to the maximum signal-to-noise ratio as an optimal codebook;

or,

selecting the minimum mean square error from the mean square errors of each codebook according to the mean square error minimization principle,

and taking the codebook corresponding to the minimum mean square error as an optimal codebook.

4. The method according to any one of claims 1-3, further comprising:

and feeding back the optimal codebook to the base station.

5. The method of claim 4, wherein the feeding back the optimal codebook to a base station comprises:

setting a codebook number of each codebook together with the base station in advance;

and sending the code book number of the optimal code book to the base station, and obtaining the optimal code book by the base station according to the corresponding relation between the stored code book number and the code book.

6. An apparatus for codebook selection in an LTE system, comprising:

an element obtaining unit, configured to obtain, according to an adjoint matrix method, elements on a diagonal line in an inverse matrix of a metric matrix corresponding to each codebook;

a metric value obtaining unit, configured to obtain a metric value of each codebook according to an element on a diagonal line in an inverse matrix of the metric matrix;

a selecting unit, configured to select an optimal metric value from the metric values of each codebook according to a preset principle, and use a codebook corresponding to the metric value as an optimal codebook;

the element obtaining unit includes:

the first obtaining subunit is configured to, when an RI corresponding to a codebook is a non-full rank, obtain, according to input noise, a channel estimation matrix, and the codebook, a determinant of a metric matrix corresponding to each codebook, and quotient an algebraic residue of a diagonal element in the metric matrix with the determinant of the metric matrix to obtain an element on a diagonal in an inverse matrix of the metric matrix corresponding to each codebook;

and the second acquiring subunit is used for acquiring a determinant of the measurement matrix corresponding to each codebook according to the input noise and the channel estimation matrix when the RI corresponding to the codebook is a full rank, and obtaining elements on a diagonal line in an inverse matrix of the measurement matrix corresponding to each codebook by taking a quotient of an algebraic remainder of diagonal line elements in the measurement matrix and the determinant of the measurement matrix.

7. The apparatus of claim 6, wherein the metric value obtaining unit comprises:

the signal-to-noise ratio obtaining subunit obtains the signal-to-noise ratio of each codebook according to the elements on the diagonal line in the inverse matrix of the measurement matrix;

the signal-to-noise ratio setting subunit is used for setting the signal-to-noise ratio of each codebook as the metric value of the codebook;

or,

a mean square error obtaining subunit, configured to obtain a mean square error of each codebook according to elements on a diagonal line in an inverse matrix of the measurement matrix,

and the mean square error setting subunit is used for setting the mean square error of each codebook as the metric value of the codebook.

8. The apparatus of claim 7, wherein the selection unit comprises:

a maximum signal-to-noise ratio selecting subunit, configured to select a maximum signal-to-noise ratio from the signal-to-noise ratios of each codebook according to a signal-to-noise ratio maximization principle;

a maximum signal-to-noise ratio setting subunit, configured to use the codebook corresponding to the maximum signal-to-noise ratio as an optimal codebook;

or,

the minimum mean square error selection subunit is used for selecting the minimum mean square error from the mean square errors of each codebook according to a mean square error minimization principle;

and the minimum mean square error setting subunit is used for taking the codebook corresponding to the minimum mean square error as the optimal codebook.

9. The apparatus of any one of claims 6-8, further comprising:

and the feedback unit is used for feeding back the optimal codebook to the base station.

10. The apparatus of claim 9, wherein the feedback unit comprises:

a codebook number setting subunit, configured to set a codebook number of each codebook together with the base station in advance;

and the sending subunit is configured to send the codebook number of the optimal codebook to the base station, and the base station obtains the optimal codebook according to the stored correspondence between the codebook number and the codebook.

Images