JP5536884B2 - Beamforming using base codebook and differential codebook - Google Patents

Description

  Embodiments of the present disclosure generally relate to wireless communication systems, and more particularly, to a beamforming method and apparatus that utilizes a base codebook and a differential codebook.

  In general, a mobile station in a closed-loop multiple-input and / or multiple-input (MIMO) system transmits channel state information to the base station via a feedback path. Channel state information is used for beamforming at the base station to compensate for current channel conditions. In some conventional systems, a mobile station transmits a channel covariance matrix to a base station, whereby the base station determines a beamforming matrix to be used for beamforming at the base station. In another conventional system, a mobile station generates a beamforming matrix based on channel conditions. The beam forming matrix generated in this way is provided to the base station as feedback. However, transmitting the channel covariance matrix and / or beamforming matrix from the mobile station to the base station involves a relatively high bandwidth consumption that would otherwise be available for data traffic.

  Embodiments of the present invention will now be described by way of example, which are not limiting, and like reference numerals represent like components in the accompanying drawings.

1 is a schematic diagram of a MIMO system. An example of how to determine and quantize the beamforming matrix is shown. An example method for estimating a beamforming matrix based on feedback received from a mobile station is shown. 1 illustrates an example system that can implement a communication device, all in accordance with various embodiments of the present disclosure.

  Exemplary embodiments of the present invention include, but are not limited to, methods and apparatus for generating and / or estimating a beamforming matrix that utilizes a base codebook and a differential codebook.

  Various aspects of the exemplary embodiments are described using terminology commonly used by one of ordinary skill in the art to convey his work to others skilled in the art. However, as will be appreciated by those skilled in the art, other embodiments may be implemented utilizing only some of these described aspects. For purposes of illustration, specific numbers, materials, and configurations are set forth to facilitate a thorough understanding of the exemplary embodiments. However, it will be apparent to one skilled in the art that alternative embodiments may be practiced without these specific details. On the other hand, there are also places where well-known features are omitted or simplified so as not to obscure the exemplary embodiments.

  In addition, various operations are described as a plurality of separate operations to help understanding of the Rie-Suitable embodiment. However, the order of description depends on this order. It should not be interpreted as meaning. In particular, these operations do not have to be performed in the order presented.

  The phrase “some embodiments” is used repeatedly. This phrase does not always represent the same embodiment, but it may. The terms “comprising”, “having” and “including” are synonymous unless otherwise specified. The phrase “A and / or B” means (A), (B), or (A and B). The phrase “A / B”, like “A and / or B”, also means (A), (B), or (A and B). The phrase “at least one of A, B, and C” means (A), (B), (C), (A and B), (A and C), (B and C), or (A, B and C). The phrase “(A) B” means (B) or (A and B), that is, A is an optional member.

  While specific embodiments are illustrated and described herein, those skilled in the art will recognize a wide variety of alternative and / or equivalent implementations without departing from the scope of the embodiments of the present invention. It will be apparent that the particular embodiments illustrated and described may be replaced. This application is intended to cover all adaptations and variations of the embodiments described herein. Therefore, it is to be understood that the embodiments of the invention are intended to be limited only by the claims and the equivalents thereof.

  In this disclosure, unless otherwise specified, the conjugate transpose of an m × n matrix A that may contain complex numbers is n × m, expressed as A *, transpose the matrix A, and It is obtained by performing complex conjugate of each entry of the matrix formed by transposing. In this disclosure, unless otherwise specified, the unitary matrix is an n × n complex matrix B, which satisfies the condition (B *) B = B (B *) = In, where In is n-dimensional And B * is the conjugate transpose of B. The matrix B is therefore unitary only if B is an inverse, where the inverse of B is equal to the conjugate transpose of B. In this disclosure, unless otherwise specified, two vectors are orthogonal if they are perpendicular to each other (eg, two vectors form a right angle and the dot product of the two vectors is zero) Is). In this disclosure, unless otherwise specified, Hermitian matrix C is a square matrix that may contain complex numbers, and each entry is equal to its conjugate transpose (e.g., i for matrix C for all indices i and j). The elements of the th row and j th column are equal to the complex conjugate of the elements of the j th row and i th column of the matrix C). Therefore, when the matrix C is a Hermitian matrix, C * = C. In this disclosure, unless otherwise stated, for an m × n matrix M, the singular value decomposition of the matrix M is a factorization of the form M = E∧F *, where E is an m × m unitary matrix. ∧ is an m × n diagonal matrix with a non-negative real number on the diagonal, F * represents the conjugate transpose of the matrix F, and the matrix F is an n × n unitary matrix.

  Embodiments of the present disclosure include IEEE 802.16-2009 (approved May 13, 2009), all amendments, updates, and / or revisions (eg, pre-draft stages) In the multi-carrier transmission scheme presented in the 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE) Project, Ultra Mobile Broadband (UMB) Project (also referred to as “3GPP2”), etc. It can be used in radio access networks that utilize orthogonal frequency division multiple access (OFDMA) communications. In other embodiments, the communication may be compliant with communication standards and / or specifications in addition to or in lieu of this.

  FIG. 1 is a schematic diagram of a communication system 100 in various embodiments of the present disclosure. In various embodiments, the communication system 100 includes a base station 104 that communicates with a mobile station 140 via a wireless channel 130. In various embodiments, base station 104 and / or mobile station 140 may be a MIMO device. In various embodiments, the communication system 100 may be a closed loop system that utilizes beamforming to increase the signal to noise ratio (SNR) of signals transmitted from the base station 104 to the mobile station 140.

  In various embodiments, the base station 104 may transmit one or more data streams to the mobile station 140. For example, FIG. 1 illustrates how the data stream S1 is transmitted from the base station 104 to the mobile station 140, but in various other embodiments, any other suitable number of data streams may be provided. it can. Prior to transmission, the data stream S1 may be appropriately weighted by one or more components of the base station 104 as described below.

  In various embodiments, the base station 104 may include a beamforming module 112 that weights data signals (eg, data signals in the data stream S1) using a beamforming matrix. The term beamforming here means to apply a beamforming factor or weight to the signal in the frequency domain of the data stream before transmission. In various embodiments, the beamforming coefficients or weights may be determined from a beamforming matrix.

  Base station 104 may include multiple transmit antennas 108a, 108b, 108c, and 108d for transmitting weighted data streams. In FIG. 1, four transmit antennas are shown, but in various other embodiments, the base station 104 may include any other suitable number of transmit antennas.

  Base station 104 may further include one or more receive antennas (eg, receive antenna 110) that may also receive feedback regarding channel conditions from mobile station 140, as well as other information.

  In various embodiments, the base station 104 may further include a beamforming matrix estimation module 116 that may be configured to estimate a beamforming matrix based at least in part on feedback received from the mobile station 140.

The order of the beamforming matrix (eg, the number of rows and / or columns) may be based on the number of data streams transmitted by the base station 104 and the number of transmit antennas included in the base station 104. In various embodiments, the beamforming matrix may be of order N _t × N _s , where N _t and N _s may be the number of transmit antennas and the number of data streams of base station 104, respectively. For example, FIG. 1 shows that N _t is 4 (four transmit antennas 108a, 108b, 108c, and 108d) and N _s is 1 (since there is only one data stream S1), so the beamforming matrix B is 4 X1 vector. In various embodiments, the signal transmitted by the base station 104 is x = B. It is expressed as S (Equation 1). Here, S represents the Ns data stream of the base station 104 (such as the data stream S1 of FIG. 1), B is an N _t × N _s beam matrix determined by the beamforming matrix estimation module 116, and x is N _t × 1 vector corresponding to the weighted data signal transmitted by the four transmit antennas 108a, 108b, 108c, and 108d.

In various embodiments, base station 104 may include at least as many transmit antennas as the number of data streams transmitted by base station 104, although the scope of this disclosure is not limited in this respect. In these embodiments, N _t has the same height as at least N _s.

  Referring back to FIG. 1, in various embodiments, the mobile station 140 includes one or more receive antennas (eg, receive antennas 144a and 144b) configured to receive signals transmitted by the base station 104 via the channel 130. May include. Although two receive antennas are shown in FIG. 1, any other suitable number of receive antennas can be utilized in various other embodiments. In various embodiments, mobile station 140 may include at least as many receive antennas as the number of data streams transmitted by base station 104, although the scope of this disclosure is not limited in this respect.

In various embodiments, the mobile station may further include a channel estimation module 148 for estimating the channel condition of the channel 130 based at least in part on signals received from one or more of the transmit antennas 108a,. . For example, the channel estimation module 148 can determine a channel matrix that describes the current state of the channel 130. In various embodiments, the channel matrix H may indicate the conditions of the subchannels between each of the transmit antennas 108a, ..., 108d, and each of the receive antennas 144a and 144b. In various embodiments, the channel matrix H may be of order N _r × N _t , where Nr may be the number of receive antennas for the mobile station 140. FIG. 1 represents four transmit antennas 108a,..., 108d (ie, N _t = 4) of the base station 104 and two receive antennas 144a and 144b (ie, N _r = 2) of the mobile station 140, Channel matrix H may be a 2 × 4 matrix in MIMO system 100.

The channel estimation module 148 can also construct a channel covariance matrix R from the channel matrix H. For example, the channel covariance matrix R may be equal to R = E [(H *) H] (Equation 2). Here, H * is a conjugate transpose of the channel matrix H, and E [] is an expectation operation. In various embodiments, the channel matrix H may represent the instantaneous condition of the channel 130 and the channel covariance matrix R may represent the relatively long-term statistics of the channel 130. Accordingly, the channel matrix H changes faster than the channel covariance matrix R in time and frequency. In various embodiments, the channel covariance matrix R may be a Hermite matrix of order N _t × N _t , and N _t (eg, the number of transmit antennas of the base station 104) is 4 in the MIMO system 100.

In various embodiments, the mobile station 140 may further include a matrix decomposition module 152 configured to decompose the channel covariance matrix R using, for example, singular value decomposition. Singular value decomposition of a matrix is to factor the matrix into three different matrices. For example, the singular value decomposition of the channel covariance matrix R may be in the form of R = U∧V * (Equation 3). Where U is a unitary square matrix of the order of N _t , ∧ is a diagonal matrix of N _t × N _t having real numbers that are not negative on the diagonal, and V * is a unitary matrix of the order of N _t This is a conjugate transpose of the square matrix V. In various embodiments, the columns of matrix V may be eigenvectors of matrix (R *) R, and the diagonal values of matrix ∧ may be singular values of R.

In various embodiments, the matrix V may include a beamforming matrix, and the portion of the matrix V that represents the beamforming matrix may be represented by _Vb . For example, as described above, for a single data stream (eg, N _s = 1) and four transmit antennas (eg, N _t = 4) of the base station 104 (see example in FIG. 1), the beam The forming matrix is a 4 × 1 vector. In this case, the matrix V is a 4 × 4 square matrix, and the first column of V (eg, the principal eigenvector of the matrix (R *) R) forms the beamforming matrix V _b. Can do. In other words, in this case, the beamforming matrix _Vb may be composed of the first column of the matrix V.

In another example, for two data streams (eg, N _s = 2) and four transmit antennas (eg, N _t = 4) of the base station 104 (not shown in FIG. 1), the matrix V is It becomes a 4 × 4 square matrix, and the beamforming matrix V _b becomes a 4 × 2 matrix. In this case, the first two columns of V (eg, two eigenvectors of the matrix (R *) R) can form the beamforming matrix _Vb .

In various embodiments, the mobile station 140 may further include a quantization module 156. When beamforming matrix V _b is generated from the matrix V, the quantization module 156, a beamforming matrix V _b may be quantized by using the base codebook C _b. In these embodiments, the base codebook C _b is to populate the surface of the manifold, it may be utilized to efficiently encoding or quantization of the beamforming matrix. Base codebook C _b may include a plurality of candidate matrices each having a dimension (dimension The) similar to the beamforming matrix V _b. Among the plurality of candidate matrices, most beams one that matches the forming matrix V _b to select the base codebook C _b, may be fed back from the mobile station 140 a codeword corresponding to the selected candidate matrix to the base station 104 . Here, the selected candidate matrix may represent the beamforming matrix V _b (eg, the selected candidate matrix may be a quantized version of the beamforming matrix V _b ) and the selected candidate The matrix will be referred to herein as a quantized beamforming matrix.

ここで、

は、量子化されたビームフォーミング行列であり、Ｃ_ｂはベースコードブックであり、ＶｉはベースコードブックＣ_ｂの中の候補行列を表し、

は、ビームフォーミング行列Ｖ_ｂの複素共役であり、

は、フロベニウスノルム演算である。フロベニウスノルム演算が数４では利用されているが、様々な他の実施形態では、他の任意の適切な行列ノルムまたはベクトルノルムを利用することもできる（例えば、スペクトルノルム、ユークリッドノルム等）。数４は、ベースコードブックＣ_ｂに含まれている複数の候補行列の中から、ビームフォーミング行列Ｖ_ｂを最もよく表している、量子化されたビームフォーミング行列

を選択するものである。 For example, referring again to FIG. 1, the quantization module 156 can quantize the beamforming matrix _Vb as follows.

here,

Is a beamforming matrix is quantized, C _b is the base codebook, Vi represents a candidate matrix in the base codebook C _b,

Is the complex conjugate of the beamforming matrix V _b ,

Is the Frobenius norm operation. Although the Frobenius norm operation is utilized in Equation 4, in various other embodiments, any other suitable matrix norm or vector norm may be utilized (eg, spectral norm, Euclidean norm, etc.). Number 4, from among a plurality of candidate matrices included in the base codebook C _b, best describe the beamforming matrix V _b, beamforming matrix quantized

Is to select.

がビームフォーミング行列Ｖ_ｂを表しているが、量子化されたビームフォーミング行列

が、ベースコードブックＣ_ｂから選択される場合には、量子化エラーも生じうる（例えば、量子化されたビームフォーミング行列

と、ビームフォーミング行列Ｖ_ｂとの差異等）。量子化エラーは、これらに限定はされないが、ベースコードブックＣ_ｂに含まれる候補行列数、および、ビームフォーミング行列Ｖ_ｂがどのくらい、ベースコードブックＣ_ｂから選択された候補行列と整合しているか（例えば量子化されたビームフォーミング行列

と整合しているか）を含む、幾つかの要因に基づいて生じうる。 In various embodiments, a quantized beamforming matrix

Represents the beamforming matrix V _b , but the quantized beamforming matrix

But if it is selected from a base codebook C _b can occur quantization error (for example, beamforming matrix is quantized

And the difference from the beam forming matrix V _b ). Or quantization errors, but are not limited to, the number of candidate matrices included in the base codebook C _b, and, how long beamforming matrix V _b, is consistent with the selected candidate matrix from the base codebook C _b (Eg quantized beamforming matrix

Can be based on several factors, including:

との間の差異を表す差異行列を決定することができる。例えば差異行列は、以下のように形成されてよい。

ここで、

は、Ｎ_ｔ×Ｎ_ｔの行列

の共役転置である。さらに、

は、量子化されたビームフォーミング行列

の列に直交する列を含む行列であってよく、行列

の次数は、Ｎ_ｔ×（Ｎ_ｔ−Ｎ_ｓ）であってよい。例えば、Ｎ_ｓ＝１であり、Ｎ_ｔ＝４の場合には、

は、４×１のベクトルであり、

は、

の列それぞれがベクトル

に直交するように選択された４×３のベクトルであってよい。別の例として、Ｎ_ｓ＝２であり、Ｎ_ｔ＝４の場合には、

は、４×２のベクトルであり、

は、

の列それぞれがベクトル

の列それぞれに直交するように選択された４×２のベクトルであってよい。様々な実施形態では、

は、

がユニタリ行列となるように選択されてよい。様々な実施形態では、差異行列Ｄおよび／または行列

は、例えば、量子化されたビームフォーミング行列

に対してハウスホルダー変換を行うことで算出することができる。差異行列Ｄは、ビームフォーミング行列Ｖ_ｂと行列

との間の差異を表していてよい。 In various embodiments, to reduce this quantization error, the mobile station 140 may use a beamforming matrix _Vb and a quantized beamforming matrix.

A difference matrix representing the difference between and can be determined. For example, the difference matrix may be formed as follows.

here,

Is a matrix of N _t × N _t

Is a conjugate transpose of further,

Is the quantized beamforming matrix

Matrix containing columns orthogonal to the columns of

May be N _t × (N _t −N _s ). For example, if N _s = 1 and N _t = 4,

Is a 4 × 1 vector,

Is

Each column is a vector

It may be a 4 × 3 vector selected to be orthogonal to As another example, if N _s = 2 and N _t = 4,

Is a 4 × 2 vector,

Is

Each column is a vector

4 × 2 vectors selected to be orthogonal to each of the columns. In various embodiments,

Is

May be chosen to be a unitary matrix. In various embodiments, the difference matrix D and / or the matrix

Is, for example, a quantized beamforming matrix

It can be calculated by performing a householder conversion on. The difference matrix D is the beam forming matrix V _b and the matrix

May represent the difference between

ここで、量子化された差異行列

は、差異行列Ｄの量子化であり、Ｃ_ｄは差分コードブックであり、Ｄ_ｉは、差分コードブックＣ_ｄ内の候補行列を表し、

は、フロベニウスノルム演算である。フロベニウスノルム演算が数６では利用されているが、様々な他の実施形態では、他の任意の適切な行列ノルムまたはベクトルノルムを利用することもできる。数６は、差分コードブックＣ_ｄから、差異行列Ｄを最もよく表している候補行列

を選択するものである。 In various embodiments, the difference matrix D may be quantized using a differential codebook C _d. Differential codebook C _d may include a plurality of candidate matrix having a dimension, each similar to the difference matrix D. For example, referring again to FIG. 1, the quantization module 156 can quantize the difference matrix D as follows.

Where the quantized difference matrix

Is the quantization of the difference matrix D, C _d is the difference codebook, D _i represents the candidate matrix in the difference codebook C _d ,

Is the Frobenius norm operation. Although the Frobenius norm operation is utilized in Equation 6, in various other embodiments, any other suitable matrix norm or vector norm may be utilized. 6 is a differential codebook C _d, best describes that the candidate matrices a difference matrix D

Is to select.

と関連付けられていてよく、第２のコードワード（例えば差分コードブックＣ_ｄからのもの）は、量子化された差異行列

と関連付けられていてよい。移動局１４０は、第１のコードワードおよび第２のコードワードを、基地局１０４に送信して（実際の行列

および

を送信する代わりに）、基地局１０４側で、これらの送信された第１および第２のコードワードからビームフォーミング行列Ｖ_ｂを推定させてもよい。 In various embodiments, the mobile station 140 may transmit the first codeword and the second codeword to the base station 104 via the transmit antenna 160. In various embodiments, the first codeword (such as those from the base codebook C _b), the beam forming matrix is quantized

May be associated with the second codeword (such as those from the difference codebook C _d) is the difference matrix is quantized

May be associated with. The mobile station 140 transmits the first codeword and the second codeword to the base station 104 (actual matrix).

and

Instead, the base station 104 may cause the beamforming matrix V _b to be estimated from these transmitted first and second codewords.

が、ベースコードブックＣ_ｂ内の複数の候補行列のｎ番目の行列である場合、且つ、量子化された差異行列

が、差分コードブックＣ_ｄ内の複数の候補行列のｍ番目の行列である場合には、ｎおよびｍがそれぞれ第１および第２のコードブックとなる。様々な他の実施形態では、量子化されたビームフォーミング行列

と関連付けられているコードワードおよび／または量子化された差異行列

は、任意の他の適切な方法で生成されてもよい。 For example, a quantized beamforming matrix

But when n-th matrix of a plurality of candidate matrices based codebook C _b, and, quantized difference matrix

But if the m-th matrix of a plurality of candidate matrices differential codebook C _d is, n and m is the first and second codebooks, respectively. In various other embodiments, a quantized beamforming matrix

Codewords associated with and / or quantized difference matrix

May be generated in any other suitable manner.

および

をそれぞれ決定することができる。様々な他の実施形態では、基地局１０４が、受信されたコードワードから、行列

および

を、任意の他の適切な方法で決定することもできる。 In various embodiments, when the base station 104 receives the first and second codewords from the mobile station 140, the base station 104 determines the base codebook _Cb and the difference from the received first and second codewords. Using a saved copy of codebook _Cd , the matrix

and

Can be determined respectively. In various other embodiments, the base station 104 can generate a matrix from the received codeword.

and

Can be determined in any other suitable manner.

および

を決定すると、基地局１０４のビームフォーミング行列推定モジュール１１６は、元のビームフォーミング行列Ｖ_ｂを推定してよい。例えば、ビームフォーミング行列推定モジュール１１６は、行列

から、

を生成してもよい。これに続き、ビームフォーミング行列推定モジュール１１６は、以下のようにして、推定されたビームフォーミング行列

を決定することができる。

Base station 104 is a queue

and

, The beamforming matrix estimation module 116 of the base station 104 may estimate the original beamforming matrix _Vb . For example, the beamforming matrix estimation module 116 may

From

May be generated. Following this, the beamforming matrix estimation module 116 determines the estimated beamforming matrix as follows:

Can be determined.

が、元のビームフォーミング行列Ｖ_ｂの推定となりうる。様々な実施形態では、基地局１０４（例えばビームフォーミングモジュール１１２）は、データストリーム（例えば数１を参照して説明したようなもの）に、推定されたビームフォーミング行列

で重み付けを行ってもよく、基地局１０４の送信アンテナ１０８ａ、…、１０８ｄが、この重み付けされたデータストリームを送信してもよい。 Therefore, the estimated beamforming matrix

Can be an estimate of the original beamforming matrix _Vb . In various embodiments, the base station 104 (eg, the beamforming module 112) transmits an estimated beamforming matrix to a data stream (eg, as described with reference to Equation 1).

The transmission antennas 108a, ..., 108d of the base station 104 may transmit this weighted data stream.

  FIG. 2 illustrates an example method 200 for beamforming matrix determination and quantization in various embodiments of the invention. One or more operations of method 200 may be performed by one or more modules of mobile station 140. With reference to FIGS. 1 and 2, in various embodiments, method 200 is received at 204 (“determining channel covariance matrix R”) by base station 104, eg, by channel estimation module 148 of mobile station 140. Determining a channel covariance matrix R based on the processed signals. In various embodiments, the channel covariance matrix R may be formed from the channel matrix H as described with reference to Equation 2. As already described, the channel matrix H represents the condition of subchannels between one or more transmitting antennas 108a,..., 108d of the base station 104 and one or more receiving antennas 144a, 144b of the mobile station 140. It may be.

  The method 200 further includes at 208 (“decompose the channel covariance matrix”), eg, by using the singular value decomposition as described above with reference to Equation 3 by the matrix decomposition module 152 of the mobile station 140. This includes decomposing the covariance matrix R into complex conjugates of a left matrix U, a diagonal matrix ∧, and a right matrix V.

The method 200 further includes at 212 (“determine the beamforming matrix V _b ”), for example by the matrix decomposition module 152 of the mobile station 140, where the beam forming matrix V _b includes one or more columns of the right matrix V. Determining the beamforming matrix _Vb . In various embodiments, the beamforming matrix V _b may include the first N _s columns of the right matrix V.

を選択することを含んでよく、この量子化されたビームフォーミング行列

は、ビームフォーミング行列Ｖ_ｂを表すものであってよい。量子化されたビームフォーミング行列

は、数４を参照して説明したように、量子化されたビームフォーミング行列

が、ベースコードブックＣ_ｂ内の複数の第１の候補行列の中で、ビームフォーミング行列、および、量子化されたビームフォーミング行列

の複素共役の積のフロベニウスノルムを最大化するよう選択されてよい。 The method 200 further includes at 216 (“selecting a quantized beamforming matrix”), eg, a plurality of first candidate matrices included in the base codebook C _b by the quantization module 156 of the mobile station 140. To the quantized beamforming matrix

This quantized beamforming matrix may include selecting

May represent the beamforming matrix _Vb . Quantized beamforming matrix

Is a quantized beamforming matrix as described with reference to Equation 4.

But among the plurality of first candidate matrix in the base codebook C _b, beamforming matrix, and, beamforming matrix is quantized

May be chosen to maximize the Frobenius norm of the complex conjugate product of

との間の差異を表す差異行列Ｄを決定することを含んでよい。例えば、差異行列Ｄを決定するためには、行列

を、行列

の１以上の列それぞれが、量子化されたビームフォーミング行列

の１以上の列それぞれに直交するように、量子化されたビームフォーミング行列

に少なくとも一部基づいて、（例えば量子化されたビームフォーミング行列

のハウスホルダー変換利用することで）、形成することができる。次いで、量子化されたビームフォーミング行列

と、行列

との組み合わせであってよい

を形成してよい。様々な実施形態では、行列

は、

が、ビームフォーミング行列Ｖ_ｂの行数に等しい次数を有するユニタリ行列となるように形成されてよい。差異行列Ｄは、数５を参照して説明したように、差異行列Ｄが、行列

の複素共役と、ビームフォーミング行列Ｖ_ｂの積となるように決定されてよい。 The method 200 further includes at 220 (“determining the difference matrix”), for example, by the quantization module 156 of the mobile station 140, as described with reference to Equation 5, and the beam forming matrix V _b and the quantized beam. Forming matrix

Determining a difference matrix D representing the difference between. For example, to determine the difference matrix D, the matrix

The matrix

Each of one or more columns of is a quantized beamforming matrix

A beamforming matrix quantized to be orthogonal to each of one or more columns of

(E.g., quantized beamforming matrix) based at least in part

Can be formed by using house holder conversion). Then the quantized beamforming matrix

And the matrix

May be combined with

May be formed. In various embodiments, the matrix

Is

May be formed to be a unitary matrix having an order equal to the number of rows of the beamforming matrix _Vb . As described with reference to Equation 5, the difference matrix D is the same as the difference matrix D.

And the beam forming matrix _Vb may be determined.

を選択することを含んでよいが、この量子化された差異行列

は、差異行列Ｄを表すものであってよい。量子化された差異行列

は、数６を参照して説明したように、量子化された差異行列

が、差分コードブックＣ_ｄ内の複数の第２の候補行列の中で、差異行列の複素共役と、量子化された差異行列

との積のフロベニウスノルムを最大化するよう選択することができる。 The method 200 further includes at 224 (“selecting a quantized beamforming matrix”), eg, by the quantization module 156 of the mobile station 140, from a plurality of second candidate matrices included in the difference codebook Cd. , Quantized difference matrix

This quantized difference matrix may include

May represent the difference matrix D. Quantized difference matrix

Is the quantized difference matrix as described with reference to Equation 6.

Are the complex conjugate of the difference matrix and the quantized difference matrix among the plurality of second candidate matrices in the difference codebook C _d

Can be chosen to maximize the Frobenius norm of the product.

  The method 200 is further quantized at 228 (“transmitting the first codeword and the second codeword”), eg, as described above, from the transmit antenna 160 of the mobile station 140 to the base station 104. Transmitting a first codeword and a second codeword associated with the beamforming matrix and the quantized difference matrix, respectively.

および量子化された差異行列

にそれぞれ関連付けられている第１のコードワードおよび第２のコードワードを受信することを含む。 FIG. 3 illustrates an example method 300 for the base station 104 to estimate the beamforming matrix based on feedback received from the mobile station 140 in various embodiments of the invention. One or more operations of method 300 may be performed by one or more modules of base station 104. With reference to FIGS. 1 and 3, in various embodiments, the method 300 may be performed at 304 (“receive a first codeword and a second codeword”) from a mobile station 140 (eg, from a transmit antenna 160). ), Receiving antenna 110 is quantized beamforming matrix

And quantized difference matrix

Receiving a first codeword and a second codeword respectively associated with the.

および量子化された差異行列

を、それぞれ、受信した第１のコードワードおよび第２のコードワードに少なくとも一部基づいて決定することを含んでよい。 The method 300 may further include a quantized beamforming matrix at 308 (“determine quantized beamforming matrix and quantized difference matrix”), eg, by the beamforming matrix estimation module 116.

And quantized difference matrix

May be determined based at least in part on the received first codeword and second codeword, respectively.

）を、数７を参照して説明したように、決定された量子化されたビームフォーミング行列

および量子化された差異行列

から推定することを含んでよい。 The method 300 is further performed at 312 (“estimating the beamforming matrix”) by the beamforming matrix estimation module 116 (eg, an estimated beamforming matrix).

) As described with reference to Equation 7, the determined quantized beamforming matrix

And quantized difference matrix

From the estimation.

を利用して、１以上のデータストリーム（例えばデータストリームＳ１）に重み付けを行うことを含んでよい。方法３００はさらに、３２０で（「重み付けがなされたデータストリームを送信する」）、基地局１０４の１以上の送信アンテナ（例えば送信アンテナ１０８ａ、…、１０８ｄ）により、重み付けがなされたデータストリームを移動局１４０に送信することを含んでよい。 The method 300 further includes an estimated beamforming matrix at 316 (“weight one or more data streams”), eg, by the beamforming module 112.

May be used to weight one or more data streams (eg, data stream S1). The method 300 further moves the weighted data stream at 320 (“transmit weighted data stream”) by one or more transmit antennas (eg, transmit antennas 108a,..., 108d) of the base station 104. Transmitting to the station 140 may be included.

と、量子化された差異行列

とに量子化することで、量子化エラーを低減させることができる。従って、基地局１０４で形成されるビームフォーミング行列の推定が、より正確なものとなる。 Quantizing the beamforming matrix using two codebooks C _b and C _d has several advantages over quantizing the beamforming matrix using a single codebook. is there. For example, a beamforming matrix is converted into a quantized beamforming matrix.

And the quantized difference matrix

The quantization error can be reduced by performing the quantization. Therefore, the estimation of the beam forming matrix formed by the base station 104 becomes more accurate.

が生成されると、差異行列Ｄと量子化された差異行列

との間の差異を表していてよい第２の差異行列を（例えば数５に少なくとも一部類似している数式を利用することで）生成することができる。第２の差異行列は、その後、第２の差分コードブックを用いて量子化され、第２の量子化された差異行列が生成される。移動局１４０は基地局１０４に、量子化されたビームフォーミング行列と量子化された差異行列とに対応するコードワードの送信に加えて、第２の量子化された差異行列に対応するコードワードを送信することができる。基地局１０４は、量子化されたビームフォーミング行列、量子化された差異行列、および第２の量子化された差異行列に対応するコードワードを用いて、ビームフォーミング行列

を推定することができる。 In the various embodiments described so far, the beamforming matrix is quantized using one base codebook and one difference codebook. However, in various embodiments, more than one difference codebook can be utilized. For example, the difference matrix D and the quantized difference matrix

Is generated, the difference matrix D and the quantized difference matrix

A second difference matrix may be generated that may represent the difference between (eg, by using a mathematical formula that is at least partially similar to Equation 5). The second difference matrix is then quantized using the second difference codebook to generate a second quantized difference matrix. In addition to transmitting codewords corresponding to the quantized beamforming matrix and the quantized difference matrix, the mobile station 140 sends a codeword corresponding to the second quantized difference matrix to the base station 104. Can be sent. The base station 104 uses the codeword corresponding to the quantized beamforming matrix, the quantized difference matrix, and the second quantized difference matrix to generate a beamforming matrix.

Can be estimated.

および量子化された差異行列

を含む）を、移動局１４０から基地局１０４へと送信していた。様々な実施形態では、移動局１４０はさらに、チャネル共分散行列Ｒの量子化された形態を、（ビームフォーミング行列の量子化された形態の送信の代わりに、またはこれに加えて）基地局１０４に送信して、基地局１０４側で、チャネル共分散行列を再構築または推定させて、この推定されたチャネル共分散行列を分解することでビームフォーミング行列を決定させることも可能である。例えば、移動局１４０は、チャネル共分散行列Ｒ（例えば数２から得られたもの）を、ベースコードブックを利用して量子化して（数４に少なくとも一部類似した方法を利用して）、量子化されたチャネル共分散行列を得ることもできる。次いで移動局１４０は、チャネル共分散行列と、量子化されたチャネル共分散行列との間の差異を表す、対応する差異行列を（数５に少なくとも一部類似した方法を利用して）、算出することができる。移動局１４０は、差分コードブックを利用して（数６に少なくとも一部類似した方法を利用して）、対応する差異行列を量子化して、対応する量子化された差異行列を得ることができる。移動局１４０は、生成された量子化された行列に対応するコードワードを基地局１０４に送信し、これにより基地局１０４は、チャネル共分散行列を推定することができる。基地局１０４は、次いで、推定されたチャネル共分散行列を（例えば特異値分解を利用することで）分解することで、ビームフォーミング行列を推定することができる。 In the embodiments described so far, the quantized form of the beamforming matrix (quantized beamforming matrix)

And quantized difference matrix

Are transmitted from the mobile station 140 to the base station 104. In various embodiments, the mobile station 140 further converts the quantized form of the channel covariance matrix R to the base station 104 (instead of or in addition to the transmission of the quantized form of the beamforming matrix). It is also possible to determine the beamforming matrix by reconstructing or estimating the channel covariance matrix and decomposing the estimated channel covariance matrix on the base station 104 side. For example, the mobile station 140 quantizes the channel covariance matrix R (eg, obtained from Equation 2) using a base codebook (using a method at least partially similar to Equation 4), and It is also possible to obtain a quantized channel covariance matrix. Mobile station 140 then calculates a corresponding difference matrix (using a method at least partially similar to Equation 5) that represents the difference between the channel covariance matrix and the quantized channel covariance matrix. can do. The mobile station 140 can quantize the corresponding difference matrix using the difference codebook (using a method at least partially similar to Equation 6) to obtain the corresponding quantized difference matrix. . The mobile station 140 transmits a codeword corresponding to the generated quantized matrix to the base station 104 so that the base station 104 can estimate the channel covariance matrix. The base station 104 can then estimate the beamforming matrix by decomposing the estimated channel covariance matrix (eg, using singular value decomposition).

  The communication device described herein may be implemented in the system so as to perform a desired configuration using any appropriate hardware and / or software as appropriate. FIG. 4 illustrates, in one embodiment, one or more processors 404, system control logic 408 coupled to at least one of the processors 404, system memory 412 coupled to system control logic 408, and system control logic 408. 1 illustrates an example system 400 that includes a non-volatile memory (NVM) / storage 416 and one or more communication interfaces 420 coupled to a system control logic 408.

  The system control logic 408 of an embodiment is any that provides any suitable interface to at least one of the processors 404 and / or to any suitable device or component that communicates with the system control logic 408. A suitable interface controller may be included.

  The system control logic 408 of one embodiment may include one or more memory controllers that provide an interface to the system memory 412. System memory 412 can be utilized, for example, to load or store data and / or instructions for system 400. In one embodiment, system memory 412 may include any suitable volatile memory, such as, for example, any suitable dynamic random access memory (DRAM).

  The system control logic 408 of one embodiment may include one or more memory controllers that provide an interface to the NVM / storage 416 and the communication interface 420.

  NVM / storage 416 may be utilized, for example, to store data and / or instructions. NVM / storage 416 may include any suitable non-volatile memory, such as flash memory, and / or one or more hard disk drives (HDDs), one or more compact disk (CD) drives, and / or 1 A non-volatile storage device such as the above DVD drive may be included.

  The NVM / storage 416 may include storage resources that are a physical part of the device on which the system 400 is installed, or may be accessible without necessarily being part of the device. For example, the NVM / storage 416 may be accessed via a network via the communication interface 420.

  The system memory 412 and NVM / storage 416 may in particular be temporary and permanent copies of the beamforming matrix logic 424, respectively. In various embodiments, the system 400 may be part of the mobile station 140, and the beamforming matrix logic 424, when executed by at least one of the processors 404, includes a beam matrix as described herein. And / or quantize the beamforming matrix (eg, by using a base codebook and a difference codebook). In various other embodiments, the system 400 may be part of the base station 104 and once the beamforming matrix logic 424 is executed by at least one of the processors 404, the system 400 may now be Instructions may be included that cause a beamforming matrix to be estimated from received codewords (eg, codewords corresponding to the beamforming matrix) of the described base codebook and differential codebook.

  In some embodiments, beamforming matrix logic 424 may be additionally (or alternatively) placed in system control logic 408.

  Communication interface 420 may provide an interface to system 400 for communicating via one or more networks and / or with any other suitable device. Communication interface 420 may include any suitable hardware and / or software. In one embodiment, the communication interface 420 may include, for example, a network adapter, a wireless network adapter, a telephone modem, and / or a wireless modem. For wireless communication, the communication interface 420 in one embodiment may utilize one or more antennas.

  In one embodiment, at least one of the processors 404 may be packaged with one or more controller logic of the system control logic 408. In one embodiment, at least one of the processors 404 may be packaged with the logic of one or more controllers of the system control logic 408 to form a system in package (SiP). In one embodiment, at least one of the processors 404 may be integrated on the same die as the logic of one or more controllers of the system control logic 408. In one embodiment, at least one of the processors 404 may be integrated on the same die as the logic of one or more controllers of the system control logic 408 to form a system on chip (SoC).

  In various embodiments, system 400 may have more or less than the number of components described and / or different architectures.

  Although some examples of manufacturing methods, apparatuses, and articles have been shown so far, the scope of the present disclosure is not limited thereto. The present disclosure includes all manufacturing methods, devices, and articles to the extent that they are protected literally and within the scope of the equivalent claims. For example, while the above discloses exemplary systems that include software or firmware running on hardware in addition to other components, these systems are intended for illustration only and are to be taken as limitations Should not. In particular, any or all of the disclosed hardware, software, and / or firmware components may be embodied only in hardware, embodied only in software, embodied only in firmware, or hardware. Hardware, software, and / or some combination of firmware.

Claims (15)

Determining, by the mobile station, a channel matrix representing a channel state between the base station and the mobile station in a first period;
Determining, from the channel matrix, a channel covariance matrix representing the state of the channel in a second period longer than the first period by the mobile station;
Determining a beamforming matrix based at least in part on the channel covariance matrix determined by the mobile station;
Selecting by the mobile station a first matrix representing the beamforming matrix from a plurality of first candidate matrices included in a first codebook;
Determining a difference matrix representing a difference between the beamforming matrix and the first matrix by the mobile station;
Selecting a second matrix representing the difference matrix from a plurality of second candidate matrices included in a second codebook by the mobile station;
Transmitting from the mobile station to the base station a first codeword and a second codeword associated with the first matrix and the second matrix, respectively .
The step of determining the difference matrix further comprises:
Determining the third matrix based at least in part on the first matrix such that each of the one or more columns of the third matrix is orthogonal to each of the one or more columns of the first matrix; And the stage of
Forming a fourth matrix that is a combination of the first matrix and the third matrix and is a unitary matrix having an order equal to the number of rows of the beamforming matrix;
Determining the difference matrix such that the difference matrix is a product of a complex conjugate of the fourth matrix and the beamforming matrix;
How having a. The step of determining the third matrix further comprises:
The performed Householder transformation in the first matrix, the method of claim 1 including the step of determining the third matrix. Selecting the first matrix comprises:
Among all the plurality of first candidate matrices, the first matrix maximizes the Frobenius norm of the product of the complex conjugate of the beamforming matrix and the first matrix. The method of claim 1, comprising selecting a matrix. Selecting the second matrix comprises:
Among all of the plurality of second candidate matrices, the second matrix is such that the second matrix maximizes the Frobenius norm of the product of the complex conjugate of the difference matrix and the second matrix. The method of claim 1, comprising the step of selecting: The step of determining the beamforming matrix further comprises:
Decomposing the channel covariance matrix into complex conjugates of a left matrix, a diagonal matrix, and a right matrix using singular value decomposition;
The method according to any one of claims 1 to 4 having the steps of the beamforming matrix to determine the beamforming matrix to contain one or more columns of the right matrix. The mobile station receives, from the base station, data signals associated with N data streams, which is an integer greater than or equal to 1,
The step of determining the beamforming matrix further comprises:
6. The method of claim 5 , comprising determining the beamforming matrix such that the beamforming matrix includes N columns of the right matrix.   The method according to claim 1, wherein the channel matrix represents a condition of a subchannel between one or more transmit antennas of the base station and one or more receive antennas of the mobile station. Determining a channel matrix representing a channel state between a device and another device in a first period, and a channel covariance matrix representing the channel state in a second period longer than the first period A channel estimation module to determine;
A matrix decomposition module for determining a beamforming matrix based on the channel covariance matrix;
A quantization module for determining a quantized beamforming matrix by using a base codebook and determining a quantized difference matrix by using a difference codebook based at least in part on the beamforming matrix;
A transmission module for transmitting a first codeword and a second codeword respectively associated with the quantized beamforming matrix and the quantized difference matrix to the another device ;
The quantization module further includes
Selecting the quantized beamforming matrix from a plurality of first candidate matrices included in the base codebook such that the quantized beamforming matrix represents the beamforming matrix;
Determining a difference matrix representing a difference between the beamforming matrix and the quantized beamforming matrix;
Selecting the quantized difference matrix from a plurality of second candidate matrices included in the difference codebook such that the quantized difference matrix represents the difference matrix;
Determining a first matrix such that each of the one or more columns of the first matrix is orthogonal to each of the one or more columns of the quantized beamforming matrix;
Forming a second matrix that is a combination of the quantized beamforming matrix and the first matrix;
Wherein such difference matrix is the product of said beamforming matrix and the complex conjugate of the second matrix, that determine the difference matrix unit. The quantization module further includes
9. The apparatus of claim 8 , wherein a Householder transformation is performed on the quantized beamforming matrix to determine the first matrix. The quantization module further includes
Among all of the plurality of first candidate matrices, the quantized beamforming matrix maximizes a matrix norm of a product of a complex conjugate of the beamforming matrix and the quantized beamforming matrix. 9. The apparatus of claim 8 , wherein the quantized beamforming matrix is selected as follows. The quantization module further includes
The quantized difference matrix, among all the plurality of second candidate matrices, maximizes the matrix norm of the product of the complex conjugate of the difference matrix and the quantized difference matrix. 9. The apparatus according to claim 8 , wherein the apparatus selects a quantized difference matrix. The channel estimation module further includes:
Decomposing the channel covariance matrix into complex conjugates of a left matrix, a diagonal matrix, and a right matrix using singular value decomposition;
Determining the beamforming matrix such that the beamforming matrix includes N columns of the right matrix that are integers of 1 or more;
12. Apparatus according to any one of claims 8 to 11 , receiving a data signal associated with N data streams from said another apparatus. A receiving module for receiving a first codeword and a second codeword associated with the quantized beamforming matrix and the quantized difference matrix, respectively, from the mobile station;
A beamforming matrix estimation module, and
The beamforming matrix estimation module includes:
Determining the quantized beamforming matrix and the quantized difference matrix based at least in part on the received first codeword and the second codeword, respectively.
Estimating a beamforming matrix from the determined quantized beamforming matrix and the quantized difference matrix;
The beamforming matrix estimation module further includes:
Determining the first matrix such that each of the one or more columns of the first matrix is orthogonal to each of the one or more columns of the quantized beamforming matrix;
Forming a second matrix that is a combination of the quantized beamforming matrix and the first matrix;
A base station that estimates the beamforming matrix so that the estimated beamforming matrix is a product of a complex conjugate of the second matrix and the quantized difference matrix. The base station according to claim 13 , wherein the beamforming matrix estimation module further performs householder transformation on the quantized beamforming matrix to determine the first matrix. A beamforming module that weights one or more data streams using the estimated beamforming matrix;
The base station according to claim 13 or 14 , further comprising one or more transmit antennas for transmitting the weighted data streams.

Images