KR20130116366A - Beamforming using base and differential codebooks - Google Patents

Abstract

Embodiments of methods and apparatus for determining and / or quantizing a beamforming matrix are disclosed. In some embodiments, the determination and / or quantization of the beamforming matrix may include use of a basic codebook and a difference codebook. Additional variations and embodiments are also disclosed.

Description

[0001] BEAM FORMING USING BASE AND DIFFERENTIAL CODEBOOKS [0002]

BACKGROUND OF THE INVENTION [0002] Embodiments of the present invention generally relate to wireless communication systems, and more particularly to methods and apparatus for forming beams using a base codebook and a difference codebook.

In a closed loop multiple input and / or multiple output (MIMO) system, a mobile station typically transmits channel state information to a base station via a feedback path. The channel state information is used to compensate for current channel conditions using beamforming at the base station. In some of the traditional systems, the mobile station transmits a channel covariance matrix to the base station, which determines the beamforming matrix used to utilize beamforming at the base station from this matrix. In some other traditional systems, a beamforming matrix is generated based on channel conditions at the mobile station. The generated beamforming matrix is then provided as feedback to the base station. However, the transmission of the channel covariance matrix and / or the beamforming matrix from the mobile station to the base station may otherwise consume a relatively large bandwidth that can be used for data traffic.

BRIEF DESCRIPTION OF THE DRAWINGS Embodiments of the present invention will now be described, by way of non-limiting example embodiments, in the accompanying drawings, wherein like reference numerals designate like elements and in which:
Figure 1 schematically shows a MIMO system,
Figure 2 shows an exemplary method for the determination and quantization of a beamforming matrix,
Figure 3 shows an exemplary method for estimating a beamforming matrix based on feedback received from a mobile station,
4 illustrates an exemplary system capable of implementing a communication device, in accordance with various embodiments of the present invention.

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

Various aspects of the illustrative embodiments are described using terms commonly used by engineers in the art to communicate their research to other engineers in the field. However, it will be apparent to those skilled in the art that alternative embodiments may be practiced using only some of the described aspects. For purposes of explanation, specific numbers, materials, and configurations are set forth in order to provide a thorough understanding of the exemplary embodiments. However, it will be apparent to those skilled in the art that alternative embodiments may be practiced without such specific details. In other instances, well known features may be omitted or simplified in order not to obscure the exemplary embodiments.

In addition, while the various operations are described in sequence as a number of separate operations in a manner that is most advantageous to understanding the exemplary embodiments, the order of description should not be construed to imply that these operations are necessarily order dependent. In particular, these operations need not be performed in the order of description.

The phrase "in some embodiments" is used repeatedly. This expression does not generally refer to the same embodiment, but it may be. The terms "comprising", "having" and "having" are synonymous unless the context requires otherwise. The expression "A and / or B" means (A), (B) or (A and B). The expression "A / B" means (A), (B) or (A and B) similarly to the expression "A and / or B ". At least one of A, B, and C "means that at least one of A, B, C, A and B, A and C, B and C, The expression "(A) B" means (B) or (A and B), that is, A is an option.

Although specific embodiments have been shown and described herein, it will be appreciated by those of ordinary skill in the art that various alternatives and / or equivalent implementations may be substituted for the specific embodiments shown and described without departing from the scope of the embodiments of the invention You will know. This application is intended to cover any adaptations or variations of the embodiments described herein. Accordingly, it is manifestly intended that embodiments of the invention be limited only by the claims and their equivalents.

For purposes of this disclosure and unless otherwise stated, the conjugate transpose matrix of the mxn matrix A, possibly with complex entries, is an nxm matrix denoted by A *, taking the transpose matrix of matrix A and then taking the transpose matrix of matrix A Is obtained by taking a complex conjugate of each entry of the formed matrix. For purposes of this disclosure and unless otherwise stated, a unitary matrix is an nxn complex matrix B satisfying the condition (B *) (B) = B (B *) = I _n , where I _n Is an n-order identity matrix, and B * is a conjugate transpose matrix of B. Thus, the matrix B is a unitary matrix only if B has an inverse matrix, where the inverse of B is the same as the conjugate transpose of B. For purposes of this disclosure and unless otherwise stated, the two vectors are orthogonal if they are perpendicular to each other (e.g., when two vectors form a right angle and the inner product of the two vectors is zero) . For purposes of this disclosure and unless otherwise stated, a Hermitian matrix C is a square matrix with possible complex entries, each of which is identical to its own conjugate transpose matrix (e.g., all For exponents i and j, the elements in column i and row j of matrix C are the same as the complex conjugates of elements in column j and row i of matrix C). Therefore, when the matrix C is an Hermitian matrix, C * = C. For the purposes of this disclosure and for the mxn matrix M, the singular value decomposition of the matrix M represents the factorization of the form M = E ∧ F *, where E is an m × m unitary matrix and ∧ Mxn diagonal matrix with non-negative real numbers on its diagonal, and F * represents the conjugate transpose matrix of matrix F, which is an nxn unitary matrix.

Embodiments of the present invention may be used in conjunction with, for example, IEEE (Institute of Electrical and Electronics Engineers) approved on May 13, 2009, along with any revisions, updates and / or modifications And multicarrier transmission schemes, such as those provided by the Institute of Electrical and Electronics Engineers 802.16-2009, 3rd Generation Partnership Project (3GPP) long-term evolution (LTE) project, ultra mobile broadband (UMB) project (OFDMA) communication as is known in the art. In other embodiments, the communications may conform to additional / alternative communication standards and / or specifications.

Figure 1 schematically illustrates a communication system 100 in accordance with various embodiments of the present invention. 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 MIMO devices. In various embodiments, communication system 100 may be a closed loop system that utilizes beamforming to improve the signal-to-noise ratio (SNR) of signals transmitted by base station 104 to mobile station 140.

In various embodiments, base station 104 may transmit one or more data streams to mobile station 140. For example, Figure 1 shows one data stream S1 transmitted by the base station 104 to the mobile station 140, but in various other embodiments any other suitable number of data streams may be provided . Prior to transmission, the data stream S1 may be suitably weighted by one or more components of the mobile station 104 as described below.

In various embodiments, base station 104 may include a beamformer module 112 for weighting data signals (e.g., data signals of data stream S1) by a beamforming matrix. The term beamforming is used herein to describe applying beamforming coefficients or weights to frequency domain signals in the data stream (s) prior to transmission. In various embodiments, the beamforming coefficients or weights may be determined from a beamforming matrix.

The base station 104 may comprise a plurality of transmit antennas 108a, 108b, 108c, 108d for transmitting a weighted data stream. Although four transmit antennas are shown in FIG. 1, any other suitable number of transmit antennas may be included in base station 104 in various other embodiments.

Base station 104 may further include one or more receive antennas (e.g., receive antenna 110) that may receive feedback from a mobile station 140 on channel conditions, among other information.

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

(E.g., the number of rows and / or columns) of the beamforming matrix may be based on the number of data streams (s) transmitted by base station 104 and the number of transmit antennas included in base station 104. [ In various embodiments, the beamforming matrix may have an order of N _t x N _s , where N _t and N _s are the number of transmit antennas and the number of data stream (s) of the base station 104, respectively. For example, in Figure 1, N _t is (because the four transmit antennas (108a, 108b, 108c, 108d ) present) 4, and an N _s is 1 (because one data stream (S1) is present), Thus, the beamforming matrix B is a 4 x 1 vector. In various embodiments, the signal transmitted by base station 104 may be expressed as:

Where S denotes the N _s data streams (e.g., data stream S 1 in FIG. 1) of the base station 104 and B denotes the N _t x N _s beamforming matrix determined by the beamforming matrix estimation module 116 And x is an N _t x 1 vector corresponding to the weighted data signals transmitted by the four transmit antennas 108a, 108b, 108c, and 108d.

In various embodiments, the base station 104 may include at least as many transmit antennas as the number of data stream (s) transmitted by the base station 104, although the scope of the present invention may not be limited in this regard . In these embodiments, N _t is at least as large as N _s .

1, mobile station 140 may include one or more receive antennas configured to receive signals transmitted via channel 130 by base station 104, e.g., receive antennas (e.g., 144a, 144b. Although two receive antennas are shown in Figure 1, any other suitable number of receive antennas may be used in various other embodiments. In various embodiments, the mobile station 140 may include at least as many receive antennas as the number of data stream (s) transmitted by the base station 104, although the scope of the present invention may not be limited in this regard .

In various embodiments, the mobile station 140 may be configured to estimate the channel conditions of the channel 130 based, at least in part, on signals received from one or more of the transmit antennas 108a, ..., Module 148 may also be included. For example, channel estimation module 148 may determine a channel matrix H that describes the current state of channel 130. In various embodiments, the channel matrix H may represent conditions of sub-channels between each of transmit antennas 108a, ..., 108d and receive antennas 144a, 144b, respectively. In various embodiments, the channel matrix H may have an order of N _r x N _t , where N _r may be the number of receive antennas of the mobile station 140. 4 transmit antennas of Figure 1 is base station (104) ((108a, ... , 108d) ( i.e., N _t = 4) and two reception of the mobile station 140, an antenna (144a, 144b) (that is, N _r = 2), and thus the channel matrix H may be a 2 x 4 matrix for the MIMO system 100. [

The channel estimation module 148 may also construct a channel covariance matrix R from the channel matrix H. For example, the channel covariance matrix R may be:

Here, H * is the conjugate transpose matrix of the channel matrix H, and E [] is an expectation operation. In various embodiments, the channel matrix H may represent an instantaneous condition of the channel 130, while the channel covariance matrix R may represent relatively long-term statistics of the channel 130. [ Thus, the channel matrix H may change more rapidly over time and frequency than the channel covariance matrix R. [ In various embodiments, the channel covariance matrix R may be an Hermitian matrix with a degree of N _t x N _t , where N _t (e.g., the number of transmit antennas of the base station 104) is provided to the MIMO system 100 Is about 4.

In various embodiments, the mobile station 140 may also include a matrix decomposition module 152 configured to decompose the channel covariance matrix R using, for example, singular value decomposition. The singular value decomposition of a matrix is a factorization of a matrix into three different matrices. For example, the singular value decomposition of the channel covariance matrix R may take the following form.

Where U is a unitary square matrix of degree N _t , Λ is an N _t x N _t diagonal matrix with non-negative real numbers on its diagonal, V * is the conjugate transpose of unitary square matrix V of degree N _t It is a matrix. In various embodiments, the columns of the matrix V may be eigenvectors of the matrix R * R, and the diagonal values in the matrix A may be the singular values of R.

In various embodiments, the matrix V may comprise a beamforming matrix, a portion of the matrix V represent the beamforming matrix may be represented by V _b. For example, as described above, a single data stream (e.g., N _s = 1) and four transmit antennas (e.g., N _t = 4) of mobile station 104 (e.g., In this case, the beamforming matrix is a 4 x 1 vector. In this case, the matrix V is a 4 x 4 square matrix, the first column of V (e.g., matrix (R *) key-specific vector R) may form a beamforming matrix V _b. That is, in this case, the beamforming matrix V _b may be of the first column of the matrix V.

(Not shown in Figure 1) in another example, the mobile station 104, the two data streams (e. G., N _s = 2) and four transmit antennas (e. G., N _t = 4), the matrix V is 4 x 4 is a square matrix, the beamforming matrix V is a _b 4 x 2 matrix. In this case, the first two columns of V (e.g., matrix (R *) 2 eigenvectors of R) can form a beam forming matrix V _b.

In various embodiments, the mobile station 140 may also include a quantization module 156. When the V matrix from the beamforming matrix V _b is generated, the quantizer module 156 may quantize the beamforming matrix V _b by the base codebook C _b. In such embodiments, the base codebook C _b can be used to populate the surface of the manifold (manifold) to efficiently encoded or quantized beamforming matrix. Base codebook C _b may include a plurality of the candidate matrixes having similar dimensions and the beam forming matrix V _b, respectively. A candidate matrix that best matches the beamforming matrix V _b among the plurality of candidate matrices may be selected from the basic codebook C _b and a codeword corresponding to the selected candidate matrix may be fed back to the base station 104 by the mobile station 140 . Here, the selected candidate matrix may represent the beamforming matrix V _b (for example, the selected candidate matrix may be a quantized version of the beamforming matrix V _b), the selected candidate matrix is referred to as the beam forming matrix quantization herein .

For example, referring again to Figure 1, the quantization module 156 is a beam forming matrix V _b may be quantized as follows.

는 양자화된 빔 형성 행렬이고, C_b는 기본 코드북이고, Vi는 기본 코드북 C_b 내의 후보 행렬들을 나타내고,

는 빔 형성 행렬 V_b의 복소 켤레이고,

는 프로베니우스-놈(Frobenius-norm) 연산이다. 프로베니우스-놈이 수학식 4에서 사용되지만, 다양한 다른 실시예들에서는 임의의 다른 적절한 행렬 놈 또는 벡터 놈(예컨대, 스펙트럼 놈, 유클리드 놈 등)도 사용될 수 있다. 수학식 4는 기본 코드북 C_b에 포함된 복수의 후보 행렬 중에서 빔 형성 행렬 V_b를 최상으로 표현하는 양자화된 빔 형성 행렬

를 선택한다.here,

Is a quantized beamforming matrix, C _b is a basic codebook, Vi denotes the 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 Provenius-norm is used in Equation 4, any other suitable matrix or vector norm (e.g., spectral norm, Euclidean norm, etc.) may be used in various other embodiments. Equation (4) is quantized beamforming matrix representing the beamforming matrix V _b of the plurality of the candidate matrixes included in the base codebook C _b the best

.

가 빔 형성 행렬 V_b를 나타내지만, 양자화된 빔 형성 행렬

가 기본 코드북 C_b로부터 선택될 때, 가능한 양자화 에러(예컨대, 양자화된 빔 형성 행렬

와 빔 형성 행렬 V_b 간의 차이)가 존재할 수 있다. 양자화 에러는 기본 코드북 C_b에 포함된 후보 행렬들의 수, 및 빔 형성 행렬 V_b가 기본 코드북 C_b로부터 선택된 후보 행렬(예컨대, 양자화된 빔 형성 행렬

)과 얼마나 가깝게 매칭되는지를 포함하지만, 이에 한정되지 않는 여러 인자에 기초할 수 있다.In various embodiments, a quantized beamforming matrix

It represents the beam forming matrix V _b, the quantized beamforming matrix

Time is selected from a base codebook C _b, possible quantization error (e.g., the quantized beamforming matrix

And the difference between the beam forming matrix V _b) may be present. Quantization error is the number of candidate matrix, and the beamforming matrix V _b a candidate matrix selected from a base codebook C _b (e.g., the quantized beamforming matrix included in the base codebook C _b

Quot;), < / RTI > as well as how close it is to a match.

사이의 차이를 나타내는 차분 행렬을 결정할 수 있다. 예를 들어, 차분 행렬 D는 다음과 같이 형성될 수 있다.In various embodiments, to reduce this quantization error, the mobile station 140 is a beam forming matrix V _b and the quantized beamforming matrix

Lt; RTI ID = 0.0 > a < / RTI > For example, the difference matrix D may be formed as follows.

은 N_t x N_t 행렬

의 켤레 전치 행렬이다. 또한,

는 양자화된 빔 형성 행렬

의 열들에 직교하는 열들을 포함하는 행렬일 수 있으며, 행렬

의 차수는 N_t x (N_t - N_s)일 수 있다. 예를 들어, N_s = 1 및 N_t = 4의 경우,

는 4 x 1 벡터이고,

는

의 열들 각각이 벡터

에 직교하도록 선택된 4 x 3 행렬일 수 있다. 다른 예에서, N_s = 2 및 N_t = 4의 경우,

는 4 x 2 벡터이고,

는

의 열들 각각이 행렬

의 열들 각각에 직교하도록 선택된 4 x 2 행렬일 수 있다. 다양한 실시예들에서,

는

가 유니터리 행렬이도록 선택될 수 있다. 다양한 실시예들에서, 차분 행렬 D 및/또는 행렬

는 예를 들어 양자화된 빔 형성 행렬

에 대한 하우스홀더 리플렉션 연산(householder reflection operation)에 의해 계산될 수 있다. 차분 행렬 D는 빔 형성 행렬 V_b와 행렬

사이의 차이를 나타낼 수 있다.here,

N _t x N _t matrix

≪ / RTI > Also,

A quantized beamforming matrix < RTI ID = 0.0 >

Lt; / RTI > may be a matrix including columns orthogonal to the columns of matrix < RTI ID =

May be N _t x (N _t - N _s ). For example, for N _s = 1 and N _t = 4,

Is a 4 x 1 vector,

The

≪ / RTI >

Lt; RTI ID = 0.0 > 4x3 < / RTI > In another example, for N _s = 2 and N _t = 4,

Is a 4 x 2 vector,

The

≪ / RTI >

Lt; / RTI > may be a 4 x 2 matrix selected to be orthogonal to each of the columns of the matrix. In various embodiments,

The

Can be selected to be a unitary matrix. In various embodiments, the difference matrix D and / or matrix < RTI ID = 0.0 >

For example, a quantized beamforming matrix < RTI ID = 0.0 >

Can be computed by a householder reflection operation on the < / RTI > Difference matrix D is beam forming matrix V _b and the matrix

. ≪ / RTI >

In various embodiments, the difference matrix D may be quantized using the difference codebook C _d . The difference codebook C _d may include a plurality of candidate matrices each having dimensions similar to the difference matrix D. [ For example, referring back to FIG. 1, the quantization module 156 may quantize the differential matrix D as follows.

는 차분 행렬 D의 양자화이고, C_d는 차분 코드북이고, Di는 차분 코드북 C_d 내의 후보 행렬들을 나타내고,

는 프로베니우스-놈 연산이다. 프로베니우스-놈이 수학식 6에서 사용되지만, 다양한 다른 실시예들에서는 임의의 다른 적절한 행렬 놈 또는 벡터 놈도 사용될 수 있다. 수학식 6은 차분 코드북 C_d로부터 차분 행렬 D를 최상으로 표현하는 후보 행렬

를 선택한다.Here, the quantized difference matrix

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

Is a Probenius-Nom operation. Although Provenius-Nom is used in Equation 6, any other suitable matrix or vector object may be used in various other embodiments. Equation (6) represents a candidate matrix _< RTI ID = 0.0 _>

.

와 연관될 수 있으며, (예를 들어, 차분 코드북 C_d로부터의) 제2 코드워드는 양자화된 차분 행렬

와 연관될 수 있다. 이동국(140)은 (예를 들어, 실제 행렬들

및

를 송신하는 대신에) 제1 코드워드 및 제2 코드워드를 기지국(104)으로 송신하여, 기지국(104)이 송신된 제1 및 제2 코드워드들로부터 빔 형성 행렬 V_b를 추정하게 할 수 있다.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, (e. G., From the base codebook C _b) the first code word is quantized beamforming matrix

And the second code word (e.g., from the difference codebook C _d ) may be associated with a quantized difference matrix

May be associated with The mobile station 140 may determine (e.g.,

And

Transmitted in place of the transmission), the first codeword and the second codeword to the base station 104, the base station 104 is able to estimate a beamforming matrix V _b from the transmitted first and second code words have.

가 기본 코드북 C_b 내의 복수의 후보 행렬 중 n번째 행렬이고, 양자화된 차분 행렬

가 차분 코드북 C_d 내의 복수의 후보 행렬 중 m번째 행렬인 경우, 숫자 n 및 m은 각각 제1 및 제2 코드워드일 수 있다. 다양한 다른 실시예들에서, 양자화된 빔 형성 행렬

및 양자화된 차분 행렬

와 연관된 코드워드들은 임의의 다른 적절한 방식으로도 형성될 수 있다.For example, a quantized beamforming matrix

And a second matrix of a plurality of candidate n matrix in the base codebook C _b, the quantized difference matrix

Is the m-th matrix among the plurality of candidate matrices in the difference codebook C _d , the numbers n and m may be the first and second codewords, respectively. In various other embodiments, a quantized beamforming matrix

And a quantized difference matrix

May be formed in any other suitable manner.

및

를 각각 결정할 수 있다. 다양한 다른 실시예들에서, 기지국(104)은 수신된 코드워드들로부터 임의의 다른 적절한 방식으로도 행렬들

및

를 결정할 수 있다.In various embodiments, when the base station 104 receives the first and second code word from the mobile station 140, the base station 104 using a stored copy of the base codebook C _b and a differential codebook C _d, the received From the first and second codewords,

And

Respectively. In various other embodiments, the base station 104 may receive the codewords from the received codewords in any other suitable manner,

And

Can be determined.

및

를 결정하면, 기지국(104) 내의 빔 형성 행렬 추정 모듈(116)은 오리지널 빔 형성 행렬 V_b를 추정할 수 있다. 예를 들어, 빔 형성 행렬 추정 모듈(116)은 결정된 행렬

로부터 행렬

를 생성할 수 있다. 이어서, 빔 형성 행렬 추정 모듈(116)은 추정된 빔 형성 행렬

를 다음과 같이 결정할 수 있다.When the base station 104 determines

And

When determining, the beamforming matrix estimation module 116 in the base station 104 may estimate the original beamforming matrix V _b. For example, the beamforming matrix estimation module 116 may determine

From the matrix

Lt; / RTI > Then, the beamforming matrix estimation module 116 estimates the beamforming matrix < RTI ID = 0.0 >

Can be determined as follows.

는 오리지널 빔 형성 행렬 V_b의 추정치일 수 있다. 다양한 실시예들에서, 기지국(104)(예컨대, 빔 형성기 모듈(112))은 추정된 빔 형성 행렬

를 이용하여 (예컨대, 수학식 1과 관련하여 설명된 바와 같이) 데이터 스트림(들)을 가중화할 수 있고, 기지국(104)의 송신 안테나들(108a, ..., 108d)은 가중화된 데이터 스트림(들)을 송신할 수 있다.Thus, the estimated beamforming matrix

May be an estimate of the original beamforming matrix V _b . In various embodiments, the base station 104 (e.g., beamformer module 112) may use an estimated beamforming matrix

(E.g., as described in connection with Equation 1), and the transmit antennas 108a, ..., 108d of the base station 104 may be weighted with the weighted data Stream (s).

Figure 2 illustrates an exemplary method 200 for determination and quantization of a beamforming matrix, in accordance with various embodiments of the present invention. One or more operations of the method 200 may be performed by one or more modules of the mobile station 140. Referring to Figures 1 and 2, in various embodiments, the method 200 may be performed by the channel estimation module 148 of, for example, the mobile station 140, at 204 ("determining the channel covariance matrix R" And determining a channel covariance matrix R based on the signals received from the antenna 104. [ In various embodiments, the channel covariance matrix R may be formed from the channel matrix H as described in connection with equation (2). The channel matrix H is defined by the condition of the subchannels between one or more transmit antennas 108a ... 108d of the base station 104 and one or more receive antennas 144a and 144b of the mobile station 140 Lt; / RTI >

The method 200 may be performed by the matrix decomposition module 152 of the mobile station 140, for example, at 208 ("decomposing the channel covariance matrix"), using singular value decomposition as described above in connection with equation Decomposing the channel covariance matrix R into a complex conjugate of a left matrix U, a diagonal matrix A and a right matrix V. [

The method 200, for example mobile station 140 is a matrix decomposition module 152 is a beam forming matrix V _b comprises one or more columns on the right side matrix V by the on ( "for determining the beamforming matrix V _b") 212 that may further comprise the step of determining the beamforming matrix V _b. In various embodiments, the beamforming matrix V _b may include the first N _s columns of the columns of the right matrix V.

를 선택하는 단계를 더 포함할 수 있으며, 양자화된 빔 형성 행렬 는 빔 형성 행렬 V_b를 나타낼 수 있다. 양자화된 빔 형성 행렬

는 기본 코드북 C_b 내의 제1 복수의 후보 행렬 중에서 양자화된 빔 형성 행렬

가 수학식 4와 관련하여 설명된 바와 같이 빔 형성 행렬의 복소 켤레 및 양자화된 빔 형성 행렬

의 내적의 프로베니우스 놈을 최대화하도록 선택될 수 있다.The method 200 may be performed by a quantization module 156 of the mobile station 140, for example, at 216 ("selecting a quantized beamforming matrix"), quantized from a first plurality of candidate matrices included in a base codebook C _b Beamforming matrix

, And the quantized beamforming matrix < RTI ID = 0.0 > It may represent the beamforming matrix V _b. Quantized beamforming matrix

It is the quantized beamforming matrix from the first plurality of candidate matrix in the base codebook C _b

≪ / RTI > is the complex conjugate of the beamforming matrix and the quantized beamforming matrix < RTI ID = 0.0 >

Can be chosen to maximize the inner Benvenutoid of.

사이의 차이를 나타내는 차분 행렬 D를 결정하는 단계를 더 포함할 수 있다. 예를 들어, 차분 행렬 D를 결정하기 위해, 양자화된 빔 형성 행렬

에 적어도 부분적으로 기초하여 (예컨대, 양자화된 빔 형성 행렬

에 대한 하우스홀더 리플렉션을 이용하여), 행렬

의 하나 이상의 열들 각각은 양자화된 빔 형성 행렬

의 하나 이상의 열들 각각과 직교하도록, 행렬

가 형성될 수 있다. 이어서, 양자화된 빔 형성 행렬

와 행렬

의 콤비네이션일 수 있는 행렬

가 형성될 수 있다. 다양한 실시예들에서, 행렬

는 행렬

가 빔 형성 행렬 V_b의 행들의 수와 동일한 차수를 갖는 유니터리 행렬이 되도록 형성될 수 있다. 차분 행렬 D는 차분 행렬 D가 수학식 5와 관련하여 설명된 바와 같이 행렬

의 복소 켤레 및 빔 형성 행렬 V_b의 내적이 되도록 결정될 수 있다.The method 200, for example, beam forming matrix V _b and the quantized beam, as described in connection with Equation (5) by the quantization module 156 of the mobile station 140 in a ( "to determine a difference matrix") 220 Formation matrix

And determining a difference matrix D that represents the difference between the two. For example, to determine the difference matrix D, a quantized beamforming matrix < RTI ID = 0.0 >

(E. G., A quantized beamforming matrix < RTI ID = 0.0 >

Using house-holder reflection for < RTI ID = 0.0 >

Each of the one or more columns of the quantized beamforming matrix < RTI ID = 0.0 >

To be orthogonal to each of the one or more columns of the matrix

Can be formed. The quantized beamforming matrix < RTI ID = 0.0 >

And matrix

A matrix that can be a combination of

Can be formed. In various embodiments,

The matrix

May be formed so that the unitary matrix of the same order as the row number of the beamforming matrix V _b. The difference matrix D is a matrix in which the difference matrix D is a matrix < RTI ID = 0.0 >

A can be determined such that the inner product of the complex conjugate and the beam forming matrix V _b.

를 선택하는 단계를 더 포함할 수 있으며, 양자화된 차분 행렬

는 차분 행렬 D를 나타낼 수 있다. 양자화된 차분 행렬

는 차분 코드북 C_d 내의 제2 복수의 후보 행렬 중에서 양자화된 차분 행렬

가 수학식 6과 관련하여 설명된 바와 같이 차분 행렬의 복소 켤레 및 양자화된 차분 행렬

의 내적의 프로베니우스 놈을 최대화하도록 선택될 수 있다.The method 200 ( "for selecting the quantized beamforming matrix"), for example, in the 224 quantized from the second plurality of candidate matrices included in the differential codebook C _d by the quantization module 156 of the mobile station 140 Differential matrix

, And the quantized difference matrix < RTI ID = 0.0 >

Can represent a differential matrix D. Quantized differential matrix

Of the second plurality of candidate matrices in the difference codebook C _d ,

Lt; RTI ID = 0.0 > Equation 6, < / RTI > the complex conjugate of the difference matrix and the quantized difference matrix &

Can be chosen to maximize the inner Benvenutoid of.

The method 200 is performed by the transmit antenna 160 of, for example, the mobile station 140, at 228 ("transmitting a first codeword and a second codeword") using a quantized beamforming matrix and quantization And transmitting a first codeword and a second codeword, respectively, associated with the difference matrix to the base station 104, respectively.

및 양자화된 차분 행렬

와 각각 관련된 제1 코드워드 및 제2 코드워드를 수신하는 단계를 포함한다.Figure 3 illustrates an exemplary method 300 for estimating the beamforming matrix by base station 104 based on feedback received from mobile station 140, in accordance with various embodiments of the present invention. One or more of the operations of method 300 may be performed by one or more modules of base station 104. Referring to Figures 1 and 3, in various embodiments, a method 300 may be performed by a receiving antenna 110, e.g., at 304 ("receiving a first codeword and a second codeword"), 140 from, for example, transmit antenna 160, a quantized beamforming matrix

And a quantized difference matrix

And receiving a first code word and a second code word, respectively, associated with the first code word and the second code word, respectively.

및 양자화된 차분 행렬

를 결정하는 단계를 더 포함할 수 있다.The method 300 is performed by the beamforming matrix estimation module 116, for example, at 308 ("determining a quantized beamforming matrix and a quantized difference matrix") using the received first and second codewords Lt; RTI ID = 0.0 > quantized < / RTI > beamforming matrix

And a quantized difference matrix

Determining may be further included.

및 양자화된 차분 행렬

로부터 빔 형성 행렬(예컨대, 추정된 빔 형성 행렬

)을 추정하는 단계를 더 포함할 수 있다.The method 300 may be performed by a beamforming matrix estimation module 116, for example, at 312 ("Estimating a beamforming matrix"), using a quantized beamforming matrix determined as described in connection with equation (7)

And a quantized difference matrix

(E.g., an estimated beamforming matrix < RTI ID = 0.0 >

) Of the received signal.

를 이용하여 하나 이상의 데이터 스트림들(예컨대, 데이터 스트림(S1))을 가중화하는 단계를 더 포함할 수 있다. 방법(300)은 ("가중화된 데이터 스트림들을 송신하는") 320에서 예를 들어 기지국(104)의 하나 이상의 송신 안테나들(예컨대, 송신 안테나들(108a, ..., 108d))에 의해, 가중화된 데이터 스트림(들)을 이동국(140)으로 송신하는 단계를 더 포함할 수 있다.The method 300 may be performed by, for example, beamformer module 112 at 316 ("weighting one or more data streams"),

To weight one or more data streams (e.g., data stream S1) using the data stream. The method 300 may be performed by one or more transmit antennas (e.g., transmit antennas 108a, ..., 108d) of the base station 104, for example, at 320 to transmit the weighted data streams , And transmitting the weighted data stream (s) to the mobile station 140.

및 양자화된 차분 행렬

로 양자화하는 것은 양자화 에러를 줄일 수 있다. 따라서, 기지국(104)에서 형성된 빔 형성 행렬의 추정은 더 정확할 수 있다.Second quantization of the beamforming matrix using a codebook of C _b C _d and has a number of advantages over the quantized beamforming matrix using a single codebook. For example, if the beamforming matrix is a quantized beamforming matrix

And a quantized difference matrix

Quantization error can be reduced. Thus, the estimate of the beamforming matrix formed at base station 104 may be more accurate.

가 생성되면, (예를 들어, 수학식 5와 적어도 부분적으로 유사한 수학식을 이용하여) 차분 행렬 D와 양자화된 차분 행렬

사이의 차이를 나타낼 수 있는 제2 차분 행렬이 생성될 수 있다. 이어서, 제2 차분 코드북을 이용하여 제2 차분 행렬을 양자화하여 제2의 양자화된 차분 행렬을 생성할 수 있다. 이동국(140)은 양자화된 빔형성 행렬 및 양자화된 차분 행렬에 대응하는 코드워드들을 송신하는 것에 더하여 제2의 양자화된 차분 행렬에 대응하는 코드워드도 기지국(104)으로 송신할 수 있다. 기지국(104)은 양자화된 빔 형성 행렬, 양자화된 차분 행렬 및 제2의 양자화된 차분 행렬에 대응하는 코드워드들을 이용하여 빔 형성 행렬

를 추정할 수 있다.In the various embodiments described so far, a beamforming matrix is quantized using a basic codebook and a differential codebook. However, in various embodiments, more than one difference codebook may be used. For example, the difference matrix D and the quantized difference matrix

(E.g., using equations at least partially similar to Equation 5), the difference matrix D and the quantized difference matrix < RTI ID = 0.0 >

A second difference matrix can be generated which can represent the difference between the first and second differential matrixes. Then, the second difference matrix may be quantized using the second difference codebook to generate a second quantized difference matrix. The mobile station 140 may transmit to the base station 104 a codeword corresponding to the second quantized difference matrix in addition to transmitting the codewords corresponding to the quantized beamforming matrix and the quantized difference matrix. The base station 104 uses the codewords corresponding to the quantized beamforming matrix, the quantized difference matrix and the second quantized difference matrix to form a beamforming matrix

Can be estimated.

및 양자화된 차분 행렬

를 포함하는) 양자화된 형태의 빔 형성 행렬이 이동국(140)에 의해 기지국(104)으로 송신된다. 다양한 실시예들에서, 이동국(140)은 또한 (예컨대, 양자화된 형태의 빔 형성 행렬을 송신하는 대신에 또는 그에 더하여) 양자화된 형태의 채널 공분산 행렬 R을 기지국(104)으로 송신하여, 기지국(104)이 채널 공분산 행렬을 재구성 또는 추정한 후에 추정된 채널 공분산 행렬을 분해함으로써 빔 형성 행렬을 결정하게 할 수 있다. 예를 들어, 이동국(140)은 기본 코드북을 이용하여(수학식 4와 적어도 부분적으로 유사한 방법을 이용하여) (예컨대, 수학식 2로부터 얻은) 채널 공분산 행렬 R을 양자화하여 양자화된 채널 공분산 행렬을 얻을 수 있다. 이어서, 이동국(140)은 채널 공분산 행렬과 양자화된 채널 공분산 행렬 사이의 차이를 나타내는 대응하는 차분 행렬을 (수학식 5와 적어도 부분적으로 유사한 방법을 이용하여) 계산할 수 있다. 이동국(140)은 차분 코드북을 이용하여(수학식 6과 적어도 부분적으로 유사한 방법을 이용하여) 대응하는 차분 행렬을 양자화하여 대응하는 양자화된 차분 행렬을 얻을 수 있다. 이동국(140)은 생성된 양자화된 행렬들에 대응하는 코드워드들을 기지국(104)으로 송신할 수 있고, 기지국(104)은 이들로부터 채널 공분산 행렬을 추정할 수 있다. 이어서, 기지국(104)은 (예컨대, 특이값 분해를 이용하여) 추정된 채널 공분산 행렬을 분해함으로써 빔 형성 행렬을 추정할 수 있다.In the embodiments described so far, (a quantized beamforming matrix

And a quantized difference matrix

Is transmitted by the mobile station 140 to the base station 104. The beamforming matrix < RTI ID = 0.0 > In various embodiments, the mobile station 140 also transmits a quantized form of channel covariance matrix R to the base station 104 (e.g., in addition to or in addition to transmitting the beamforming matrix of quantized form) 104 may reconstruct or estimate the channel covariance matrix and then determine the beamforming matrix by decomposing the estimated channel covariance matrix. For example, the mobile station 140 may quantize the channel covariance matrix R (e.g., from Equation 2) using a base codebook (using a method at least partially similar to Equation 4) to obtain a quantized channel covariance matrix Can be obtained. The mobile station 140 may then calculate a corresponding differential 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. The mobile station 140 may quantize the corresponding difference matrix (using a method at least partially similar to Equation 6) using the difference codebook to obtain a corresponding quantized difference matrix. The mobile station 140 may transmit the codewords corresponding to the generated quantized matrices to the base station 104 and the base station 104 may estimate the channel covariance matrix therefrom. The base station 104 may then estimate the beamforming matrix by decomposing the estimated channel covariance matrix (e.g., using singular value decomposition).

The communication devices described herein may be implemented in a system using any suitable hardware and / or software to configure as needed. 4 illustrates system control logic 408 coupled to at least one of one or more processor (s) 404, processor (s) 404 in one embodiment, system memory 412 coupled to system control logic 408, ), A non-volatile memory (NVM) / storage device 416 coupled to the system control logic 408, and one or more communication interface (s) 420 coupled to the system control logic 408 (400).

The system control logic 408 is coupled to at least one of the processor (s) 404 and / or to any suitable device or component that communicates with the system control logic 408, in one embodiment, to provide any suitable interface Lt; RTI ID = 0.0 > controller. ≪ / RTI >

System control logic 408 may include one or more memory controller (s) for providing an interface to system memory 412 in one embodiment. The system memory 412 may be used, for example, to load and store data and / or instructions for the system 400. The system memory 412 may comprise any suitable volatile memory, such as, for example, a suitable dynamic random access memory (DRAM) in one embodiment.

The system control logic 408 may include one or more input / output (I / O) controller (s) for providing an interface to the NVM / storage device 416 and the communication interface (s) 420 in one embodiment .

The NVM / storage device 416 may be used, for example, to store data and / or instructions. The NVM / storage device 416 may comprise any suitable non-volatile memory, such as, for example, a flash memory and / or may include, for example, one or more hard disk drives CD) drive (s) and / or one or more digital versatile disk (DVD) drive (s).

The NVM / storage device 416 may comprise a storage resource physical part of the device in which the system 400 is installed, or may be accessed by the device, but may not be part of it. For example, the NVM / storage device 416 may be accessed via the network via the communication interface (s) 420.

The system memory 412 and the NVM / storage device 416 may each include a temporary copy and a permanent copy of the beamforming matrix logic 424, respectively. In various embodiments, system 400 may be part of mobile station 140 and beamforming matrix logic 424 may be implemented as part of system 400 as described herein when executed by at least one of processor (E.g., using a base codebook and a difference codebook) to generate a beamforming matrix and / or to quantize the beamforming matrix. In various other embodiments, system 400 may be part of base station 104 and beamforming matrix logic 424 may enable system 400 to determine when beamforming matrix logic 424 is executed by at least one of processor And may include instructions to estimate the beamforming matrix from the received code words of the base codebook and the difference codebook (e.g., codewords corresponding to the beamforming matrix) as described in the specification.

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

The communication interface (s) 420 may provide an interface for the system 400 to communicate with one or more network (s) and / or with any other suitable device. The communication interface (s) 420 may comprise any suitable hardware and / or firmware. The communication interface (s) 420 may include, for example, a network adapter, a wireless network adapter, a telephone modem, and / or a wireless modem in one embodiment. For wireless communication, communication interface (s) 420 in one embodiment may use more than one antenna.

In one embodiment, at least one of the processor (s) 404 may be packaged with logic for one or more controller (s) of the system control logic 408. In one embodiment, at least one of the processor (s) 404 may be packaged with logic for one or more controllers of system control logic 408 to form a System in Package (SiP). In one embodiment, at least one of the processor (s) 404 may be integrated on the same die as the logic for one or more controller (s) of the system control logic 408. In one embodiment, at least one of the processor (s) 404 may be integrated on the same die as the logic for one or more controller (s) of the system control logic 408 to form a System on Chip (SoC) have.

In various embodiments, the system 400 may have more or fewer components, and / or different architectures.

Although certain exemplary methods, apparatus, and articles of manufacture have been described herein, the scope of protection of the present invention is not so limited. Rather, the invention encompasses all methods, apparatus and articles of manufacture, which literally or equivalently fall within the scope of the appended claims. For example, it should be noted that the above describes exemplary systems including software or firmware that runs on many of the components, particularly on hardware, but such systems are exemplary and should not be considered to be limiting. In particular, it is contemplated that any or all of the hardware, software, and / or firmware components disclosed may be implemented by hardware only, software only, firmware alone, or any combination of hardware, software, and / or firmware.

Claims (26)

One or more non-transitory computer readable media comprising instructions that, when executed by one or more processors, cause the base station to:
Processing the feedback transmission from the mobile station to determine a first codebook index and a second codebook index,
One or more non-transitory computer readable media for determining a beamforming matrix based on the first codebook index and the second codebook index to facilitate formation of a beam to transmit data to the mobile station in one or more streams. . The method of claim 1,
The instructions also cause, when executed,
And to apply beamforming coefficients to frequency domain signals within the one or more streams based on the determined beamforming matrix. The method of claim 1,
The instructions also cause, when executed,
One or more non-transitory computer readable media to cause the one or more streams to be transmitted by a plurality of antennas of the base station. The method of claim 3,
Wherein the beamforming matrix comprises a plurality of rows each corresponding to the plurality of antennas. The method of claim 1,
Wherein the first codebook index is associated with a first quantization error and the second codebook index is associated with a second quantization error that is less than the first quantization error. The method of claim 1,
Wherein the beamforming matrix is a first beamforming matrix and the instructions further cause the base station to:
Determine a second beamforming matrix based on the first codebook index, and determine the first beamforming matrix for the second beamforming matrix. The method of claim 1,
Wherein the feedback transmission comprises an indication of the first codebook index and an indication of the second codebook index. As an apparatus employed in a base station,
A first circuit receiving an indication of a first codebook index and a second codebook index and determining a beamforming matrix based on the first codebook index and the second codebook index upon transmission of feedback from the mobile station; And
A second circuit for weighting signals of a data stream to be transmitted to the mobile station based on the beamforming matrix;
/ RTI > 9. The method of claim 8,
And further comprising four or more antennas for transmitting the weighted signals. 10. The method of claim 9,
Wherein the beamforming matrix comprises four or more rows each corresponding to the four or more antennas. 9. The method of claim 8,
Wherein the first codebook index is associated with a first quantization error and the second codebook index is associated with a second quantization error that is less than the first quantization error. 9. The method of claim 8,
Wherein the beamforming matrix is a first beamforming matrix,
Wherein the first circuit determines a second beamforming matrix based on the first codebook index and determines the first beamforming matrix for the second beamforming matrix. 9. The method of claim 8,
Wherein the first circuit and the second circuit are programmable. At a base station, receiving a feedback transmission from a mobile station, the feedback transmission comprising a first codebook index and a second codebook index;
Determining a beamforming matrix based on the first codebook index and the second codebook index;
Weighting the data signals based on the beamforming matrix; And
Transmitting the weighted data signals
≪ / RTI > 15. The method of claim 14,
Transmitting the weighted signals by four or more antennas. 15. The method of claim 14,
Wherein the first codebook index is associated with a first quantization error and the second codebook index is associated with a second quantization error less than the first quantization error. One or more non-transitory computer readable media comprising instructions that, when executed by one or more processors, cause the mobile station to:
Determine a desired beamforming matrix to be used by the base station to transmit data to the mobile station in one or more streams based on the channel conditions,
Determining a first codebook index and a second codebook index to be used together to identify the desired beamforming matrix,
One or more non-transitory computer readable media adapted to transmit the first codebook index and the second codebook index to the base station following determination of the first codebook index and the second codebook index. 18. The method of claim 17,
Wherein the desired beamforming matrix comprises beamforming coefficients to be applied to frequency domain signals in one or more streams to be transmitted to the mobile station. 18. The method of claim 17,
The instructions also cause, when executed,
One or more non-transitory computer readable media for causing the one or more antennas of the mobile station to receive the one or more streams. 18. The method of claim 17,
Wherein the first codebook index is associated with a first quantization error and the second codebook index is associated with a second quantization error that is less than the first quantization error. As an apparatus employed in a mobile station,
A first circuit for determining a desired beamforming matrix for downlink transmissions, and for determining a first codebook index and a second codebook index to identify the desired beamforming matrix; And
At the time of feedback transmission, a second circuit for transmitting the first codebook index and the second codebook index
/ RTI > The method of claim 21,
And a plurality of antennas for receiving weighted downlink transmissions based on the desired beamforming matrix. The method of claim 21,
Wherein the first codebook index is associated with a first quantization error and the second codebook index is associated with a second quantization error that is less than the first quantization error. The method of claim 21,
Further comprising a system on a chip (SoC) comprising the first circuit and the second circuit. Determining a channel matrix representing states of a channel between a base station and a mobile station;
Determining a beamforming matrix based on the channel matrix;
Determining a first codebook index and a second codebook index to identify the beamforming matrix; And
Transmitting a feedback transmission including the first codebook index and the second codebook index to the base station
≪ / RTI > 26. The method of claim 25,
Further comprising receiving one or more streams by a plurality of antennas, wherein the one or more streams are beamformed based on the beamforming matrix.

Images