WO2011005677A2

WO2011005677A2 - Beamforming using base and differential codebooks

Info

Publication number: WO2011005677A2
Application number: PCT/US2010/040887
Authority: WO
Inventors: Yuan Zhu; Qinghua Li; Xintian E. Lin; Huaning Niu; Alexei Davydov
Original assignee: Intel Corporation
Priority date: 2009-07-06
Filing date: 2010-07-02
Publication date: 2011-01-13
Also published as: KR101345950B1; JP2012532567A; KR101442452B1; EP2452445A4; EP2452445A2; BR112012000427B1; CN102474320A; BR112012000427A2; KR20120033348A; CN102474320B; JP5536884B2; WO2011005677A3; KR20130116366A

Abstract

Embodiments of methods and apparatus for determining and/or quantizing a beamforming matrix are disclosed. In some embodiments, the determining and/or quantizing of the beamforming matrix may include the use of a base codebook and a differential codebook. Additional variants and embodiments are also disclosed.

Description

BEAMFORMING USING BASE AND DIFFERENTIAL CODEBOOKS

TECHNICAL FIELD

Embodiments of the present disclosure relate generally to wireless communication systems, and more particularly, to methods and apparatuses for beamforming using a base codebook and a differential codebook.

BACKGROUND

A mobile station in a closed-loop multi input and/or multi output (MIMO) system generally transmits channel state information to a base station over a feedback path. The channel state information is used to employ beamforming at the base station, to compensate for the current channel conditions. In some of the conventional systems, the mobile station transmits a channel covariance matrix to the base station, from which the base station determines a beamforming matrix that is used to employ beamforming at the base station. In some other conventional systems, a beamforming matrix is generated at the mobile station based on the channel conditions. The generated beamforming matrix is then provided to the base station as feedback. However, transmitting the channel covariance matrix and/or the beamforming matrix from the mobile station to the base station may consume relatively high bandwidth that might otherwise be available for data traffic.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will be described by way of exemplary

embodiments, but not limitations, illustrated in the accompanying drawings in which like references denote similar elements, and in which:

Fig. 1 schematically illustrates a MIMO system;

Fig. 2 illustrates an exemplary method for determination and quantization of a beamforming matrix;

Fig. 3 illustrates an exemplary method for estimating a beamforming matrix based on feedback received from a mobile station; and

Fig. 4 illustrates an example system capable of implementing a communication device, all in accordance with various embodiments of the present disclosure. DETAILED DESCRIPTION

Illustrative embodiments of the invention include, but are not limited to, methods and apparatuses for generating and/or estimating a beamforming matrix using a base codebook and a differential codebook.

Various aspects of the illustrative embodiments will be described using terms commonly employed by those skilled in the art to convey the substance of their work to others skilled in the art. However, it will be apparent to those skilled in the art that alternate embodiments may be practiced with 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 illustrative embodiments. However, it will be apparent to one skilled in the art that alternate embodiments may be practiced without the specific details. In other instances, well-known features are omitted or simplified in order not to obscure the illustrative embodiments.

Further, various operations will be described as multiple discrete operations, in turn, in a manner that is most helpful in understanding the illustrative embodiments; however, the order of description should not be construed as to imply that these operations are necessarily order dependent. In particular, these operations need not be performed in the order of presentation.

The phrase "in some embodiments" is used repeatedly. The phrase generally does not refer to the same embodiments; however, it may. The terms "comprising," "having," and "including" are synonymous, unless the context dictates otherwise. The phrase "A and/or B" means (A), (B), or (A and B). The phrase "A/B" means (A), (B), or (A and B), similar to the phrase "A and/or 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 optional.

Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a wide variety of alternate 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. This application is intended to cover any adaptations or variations of the embodiments discussed herein. Therefore, it is manifestly intended that the embodiments of the invention be limited only by the claims and the equivalents thereof.

For the purpose of this disclosure and unless otherwise mentioned, a conjugate transpose of an m by n matrix A with possibly complex entries is an n by m matrix, represented by A*, obtained by taking the transpose of the matrix A, and then taking the complex conjugate of each entry of the matrix formed by taking the transpose of the matrix A. For the purpose of this disclosure and unless otherwise mentioned, a unitary matrix is an n by n complex matrix B satisfying the condition (B*)B = B(B*) = I_n, where I_n is the identity matrix in n dimensions, and B* is the conjugate transpose of B. Thus, a matrix B is unitary if and only if B has an inverse, wherein the inverse of B is equal to the conjugate transpose of B. For the purpose of this disclosure and unless otherwise mentioned, two vectors are orthogonal if the two vectors are perpendicular to each other (e.g., the two vectors form a right angle, the dot product of the two vectors is 0). For the purpose of this disclosure and unless otherwise mentioned, a hermitian matrix C is a square matrix with possible complex entries, each of which is equal to its own conjugate transpose (e.g., the element in the i^l row and j^l column of matrix C is equal to the complex conjugate of the element in the j^th row and 1^th column of matrix C, for all indices i and j). Thus, if a matrix C is a hermitian matrix, then C* = C. For the purpose of this disclosure and unless otherwise mentioned, for an m by n matrix M, a singular value decomposition of matrix M refers to a factorization of the form M = E A F*, where E is an m by m unitary matrix, A is a m by n diagonal matrix with nonnegative real numbers on its diagonal, and F* denotes the conjugate transpose of matrix F, where matrix F is an n by n unitary matrix.

Embodiments of the present disclosure may be used in wireless access networks that employ orthogonal frequency division multiple access (OFDMA) communications as used by multicarrier transmission schemes presented in, e.g., the Institute of Electrical and Electronics Engineers (IEEE) 802.16 - 2009, approved May 13, 2009, along with any amendments, updates, and/or revisions (e.g., 802.16m, which is presently at predraft stage), 3rd Generation Partnership Project (3GPP) long-term evolution (LTE) project, ultra mobile broadband (UMB) project (also referred to as "3GPP2"), etc. In other

embodiments, communications may be compatible with additional/alternative

communication standards and/or specifications.

FIG. 1 schematically illustrates a communication system 100, in accordance with 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 over wireless channel 130. In various embodiments, the base station 104 and/or the mobile station 140 may be MIMO devices. In various embodiments, the communication system 100 may be a closed-loop system that employs beamforming to increase a signal to noise ratio (SNR) of signals transmitted by base station 104 to 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 a data stream Sl being transmitted by the base station 104 to the mobile station 140, although in various other embodiments, any other suitable number of data streams may also be provided. Prior to transmission, the data stream Sl may be appropriately weighted by one or more components of the base station 104 as discussed hereinafter.

In various embodiments, the base station 104 may include a beamformer module 112 to weight data signals (e.g., data signals of data stream Sl) by a beamforming matrix. The term beamforming is used herein to describe the application of 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 the beamforming matrix.

Base station 104 may comprise a plurality of transmit antennas 108a, 108b, 108c and 108d, to transmit the weighted data stream. In Fig. 1, four transmit antennas are illustrated, although in various other embodiments, any other suitable number of transmit antennas may be included in the base station 104.

The base station 104 may also include one or more receive antennas (e.g., receive antenna 110) that may receive, among other information, feedback about the channel condition from the mobile station 140.

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

An order (e.g., number of rows and/or columns) of the beamforming matrix may be based on a number of data stream(s) transmitted by the base station 104 and a number of transmit antennas included in the base station 104. In various embodiments, the beamforming matrix may be of order N_t by N₈, where N_t and N_s are the number of transmit antennas and the number of data stream(s), respectively, of the base station 104. For example, in Fig. 1, N_t is 4 (as there are four transmit antennas 108a, 108b, 108c and 108d) and Ns is 1 (as there is one data stream Sl), and hence, a beamforming matrix B is a 4 by 1 vector. In various embodiments, the signal transmitted by the base station 104 may be represented by

x = B. S, Equation (1)

where S represents N_s data stream(s) (e.g., data stream Sl of Fig. 1) of the base station 104, B is the N_t by N_s beamforming matrix determined by the beamforming matrix estimation module 116, and x is an N_t by 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) being transmitted by base station 104, although the scope of this disclosure may not be limited in this respect. In these embodiments, N_t is at least as high as N_s.

Referring again to Fig. 1, in various embodiments, the mobile station 140 may include one or more receive antennas, e.g., receive antennas 144a and 144b, configured to receive signals transmitted through channel 130 by the base station 104. In Fig. 1, two receive antennas are illustrated, although in various other embodiments, any other suitable number of receive antennas may be used. In various embodiments, the mobile station 140 may include at least as many receive antennas as the number of data stream(s) being transmitted by base station 104, although the scope of this disclosure may not be limited in this respect.

In various embodiments, the mobile station 140 may also include a channel estimation module 148 to estimate channel conditions of the channel 130, based at least in part on signals received from one or more of the transmit antennas 108a, ..., 108d. For example, the channel estimation module 148 may determine a channel matrix H which describes the current state of channel 130. In various embodiments, the channel matrix H may be indicative of conditions of sub-channels 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 the order N_r by N_t, where N_r may be a number of receive antennas in the mobile station 140. Fig. 1 illustrates four transmit antennas 108a, ..., 108d (i.e., N_t = 4) of the base station 104 and two receive antennas 144a and 144b (i.e., N_r = 2) of the mobile station 140, and accordingly, the channel matrix H may be a 2 by 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 equal to

R = £ [(H*)H], Equation (2)

wherein H* is the conjugate transpose of the channel matrix H, and E[ ] is an expectation operation. In various embodiments, the channel matrix H may be

representative of an instantaneous condition of the channel 130, whereas the channel covariance matrix R may be representative of relatively long-term statistics of the channel 130. Thus, the channel matrix H may change faster over time and frequency as compared to the channel covariance matrix R. In various embodiments, the channel covariance matrix R may be a Hermitian matrix of order N_t by N_t, where N_t (e.g., the number of transmit antennas of the base station 104) is 4 for the MIMO system 100.

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. Singular value decomposition of a matrix refers to factorizing the matrix in three different matrices. For example, the singular value decomposition of the channel covariance matrix R may be of the form:

R = U A V*, Equation (3)

where U is an unitary square matrix of order Nt, Λ is a N_t by N_t diagonal matrix with nonnegative real numbers on its diagonal, and V* is the conjugate transpose of an unitary square matrix V of order Nt. In various embodiments, the columns of matrix V may be the eigenvectors of matrix (R*)R, and the diagonal values in matrix Λ may be the singular values of R.

In various embodiments, the matrix V may include the beamforming matrix, and the part of the matrix V that represents the beamforming matrix may be represented by Vb. For example, as previously discussed, for a single data stream (e.g., N_s = 1) and four transmit antennas (e.g., N_t = 4) of the base station 104 (e.g., as illustrated in Fig. 1), the beamforming matrix is a 4 by 1 vector. In this case, the matrix V is a 4 by 4 square matrix, and the first column of V (e.g., the principal eigenvector of matrix (R*)R) may form the beamforming matrix V_b. That is, in this case, the beamforming matrix V_b may consist of the first column of the matrix V.

In another example (not illustrated in Fig. 1), for two data streams (e.g., N_s = 2) and four transmit antennas (e.g., N_t = 4) of the base station 104, the matrix V is a 4 by 4 square matrix and the beamforming matrix V_b is a 4 by 2 matrix. In this case, the first two columns of V (e.g., two eigenvectors of matrix (R*)R) may form the beamforming matrix V_b.

In various embodiments, the mobile station 140 may also include a quantization module 156. Once the beamforming matrix Vb is generated from the matrix V, the quantization module 156 may quantize the beamforming matrix Vb using a base codebook Cb. In these embodiments, the base codebook Cb may be used to populate the surface of a manifold to efficiently encode or quantize the beamforming matrix. The base codebook Cb may include a plurality of candidate matrices, each having dimensions similar to the beamforming matrix Vb. A candidate matrix, among the plurality of candidate matrices, that best matches the beamforming matrix Vb may be selected from the base codebook Cb, and a codeword corresponding to the selected candidate matrix may be feedback by the mobile station 140 to the base station 104. Here, the selected candidate matrix may be representative of the beamforming matrix Vb (e.g., the selected candidate matrix may be a quantized version of the beamforming matrix Vb), and the selected candidate matrix may be referred herein as quantized beamforming matrix.

For example, referring again to Fig. 1, the quantization module 156 may quantize the beamforming matrix Vb as follows:

Equation (4)

where V is the quantized beamforming matrix, Cb is the base codebook, Vi represents the candidate matrices in the base codebook Cb, V_b ^* is a complex conjugate of the beamforming matrix V_b, and | | is the Frobenius-norm operation. Although

Frobenius-norm is used in equation 4, in various other embodiments, any other appropriate matrix norm or vector norm (e.g., the spectral norm, the Euclidean norm, or the like) may also be used. Equation 4 selects, from among the plurality of candidate matrices included in the base codebook Cb, the quantized beamforming matrix V that is best representative of the beamforming matrix Vb.

In various embodiments, although quantized beamforming matrix V is

representative of the beamforming matrix Vb, there may be possible quantization error

(e.g., difference between the quantized beamforming matrix V and the beamforming matrix Vb) when the quantized beamforming matrix V is selected from the base codebook C_b. The quantization error may be based on several factors including, but not limited to, a number of candidate matrices included in the base codebook Cb and how closely the beamforming matrix Vb matches the selected candidate matrix (e.g., the quantized beamforming matrix V ) from the base codebook C_b.

In various embodiments, to reduce this quantization error, the mobile station 140 may determine a difference matrix that is representative of a difference between the beamforming matrix Vb and the quantized beamforming matrix V. For example, a difference matrix D may be formed such that

D = [V V^] * V_b, Equation (5)

wherein [V V¹] ^* is the conjugate transpose of N_t by N_t matrix [V V¹] . Also, V¹- may be a matrix that includes columns that are orthogonal to the columns of the quantized beamforming matrix V, and the order of the matrix V^L may be N_t by (N_t - N₈). For example, for N_s = 1 and Nt = 4, V is a 4 by 1 vector, and V^L may be a 4 by 3 matrix that is selected such that each of the columns of V^ are orthogonal to the vector V. In another example, for N_s = 2 and N_t = 4, V is a 4 by 2 matrix, and V¹ may be a 4 by 2 matrix that is selected such that each of the columns of V¹ are orthogonal to each of the columns of the matrix V. In various embodiments, V¹ may be selected such that [V V¹] is a unitary matrix. In various embodiments, the difference matrix D and/or the matrix V¹ may be calculated by, for example, a householder reflection operation on the quantized beamforming matrix V. The difference matrix D may be representative of the difference between the beamforming matrix V_b and the matrix V.

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

Z) = argmax |D^*Z⁾J , Equation (6)

A ε O

where the quantized difference matrix D is a quantization of the difference matrix D, Ca is the differential codebook, Di represents the candidate matrices in the differential codebook Ca, and | | is the Frobenius-norm operation. Although Frobenius-norm is used in equation 6, in various other embodiments, any other appropriate matrix norm or vector norm may also be used. Equation 6 selects, from the differential codebook Ca, a candidate matrix D that is best representative of the difference matrix D.

In various embodiments, the mobile station 140 may transmit, by way of the transmit antenna 160, a first codeword and a second codeword to the base station 104. In various embodiments, the first codeword (e.g., from the base codebook Cb) may be associated with the quantized beamforming matrix V and the second codeword (e.g., from the differential codebook Ca) may be associated with the quantized difference matrix D . The mobile station 140 may transmit the first codeword and the second codeword to the base station 104 (e.g., instead of sending the actual matrices V and D ), to enable the base station 104 to estimate the beamforming matrix Vb from the transmitted first and second codewords.

For example, if the quantized beamforming matrix V is an n^th matrix of the plurality of candidate matrices in the base codebook Cb and if the quantized difference matrix D is a m^th matrix of the plurality of the candidate matrices in the differential codebook Ca, then the numbers n and m may be the first and second codewords, respectively. In various other embodiments, codewords associated with the quantized beamforming matrix V and/or the quantized difference matrix D may be formed in any other appropriate manner as well.

In various embodiments, once the base station 104 receives the first and second codewords from the mobile station 140, the base station 104 may determine, from the received first and second codewords, the matrices D and V , respectively, using saved copies of the base codebook Cb and differential codebook Ca. In various other embodiments, the base station 104 may determine, from the received codewords, the matrices D and V in any other appropriate manner as well.

Once the base station 104 determines the matrices D and V , 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 generate the matrix V^ from the determined matrix V . Subsequently, the beamforming matrix estimation module 116 may determine an estimated beamforming matrix V_b as follows: V_b = [V V^]D . Equation (7)

Thus, the estimated beamforming matrix J^ may be an estimate of the original beamforming matrix Vb. In various embodiments, the base station 104 (e.g., the beamformer module 112) may weight data stream(s) (e.g., as discussed with respect to equation 1) with the estimated beamforming matrix V_b , and the transmit antennas 108a,

..., 108d of the base station 104 may transmit the weighted data stream(s).

Fig. 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 carried by one or more modules of the mobile station 140. Referring to Figs. 1 and 2, in various embodiments, the method 200 includes, at 204 ("Determining a channel covariance matrix R"), determining, e.g., by the channel estimation module 148 of the mobile station 140, a channel covariance matrix R based on signals received from the base station 104. In various embodiments, the channel covariance matrix R may be formed from the channel matrix H, as discussed with respect to equation 2. As previously discussed, the channel matrix H may be representative of conditions of sub-channels between one or more transmit antennas 108a, ... 108d of the base station 104 and one or more receive antennas 144a, 144b of the mobile station 140.

The method 200 may further include, at 208 ("Decomposing the channel covariance matrix"), decomposing, e.g., by the matrix decomposition module 152 of the mobile station 140, the channel covariance matrix R into a left matrix U, a diagonal matrix

Λ, and a complex conjugate of a right matrix V using singular value decomposition, as previously discussed with respect to equation 3.

The method 200 may further include, at 212 ("Determining beamforming matrix

Vb"), determining, e.g., by the matrix decomposition module 152 of the mobile station 140, the beamforming matrix Vb such that the beamforming matrix Vb comprises one or more columns of the right matrix V. In various embodiments, the beamforming matrix Vb may comprise the first N_s number of columns of the right matrix V.

The method 200 may further include, at 216 ("Selecting a quantized beamforming matrix"), selecting, e.g., by the quantization module 156 of the mobile station 140, a quantized beamforming matrix V from a first plurality of candidate matrices included in base codebook Cb, where the quantized beamforming matrix V may be representative of the beamforming matrix Vb. The quantized beamforming matrix V may be selected such that the quantized beamforming matrix V , among the first plurality of candidate matrices in the base codebook Cb, maximizes a Frobenius norm of a product of a complex conjugate of the beamforming matrix and the quantized beamforming matrix V , as discussed with respect to equation 4.

The method 200 may further include, at 220 ("Determining a difference matrix"), determining, e.g., by the quantization module 156 of the mobile station 140, a difference matrix D that is representative of a difference between the beamforming matrix Vb and the quantized beamforming matrix V , as discussed with respect to equation 5. For example, to determine the difference matrix D, matrix F^xmay be formed (e.g., using householder reflection on the quantized beamforming matrix V ) based at least in part on the quantized beamforming matrix V , such that each of one or more columns of the matrix F^xis orthogonal to each of one or more columns of the quantized beamforming matrix V . Subsequently, matrix [V v ] may be formed, which may be a combination of the quantized beamforming matrix V and the matrix V^ . In various embodiments, the matrix V^L may be formed such that the matrix [V K ] is an unitary matrix having an order that is equal to a number of rows of the beamforming matrix Vb. The difference matrix D may be determined such that the difference matrix D is a product of a complex conjugate of the matrix [V V^±] and the beamforming matrix V_b, as discussed with respect to equation 5.

The method 200 may further include, at 224 ("Selecting a quantized beamforming matrix"), selecting, e.g., by the quantization module 156 of the mobile station 140, a quantized difference matrix D from a second plurality of candidate matrices included in differential codebook Ca, where the quantized difference matrix D may be representative of the difference matrix D. The quantized difference matrix D may be selected such that the quantized difference matrix D , among all the second plurality of candidate matrices in the difference codebook Ca, maximizes a Frobenius norm of a product of a complex conjugate of the difference matrix and the quantized difference matrix D , as discussed with respect to equation 6.

The method 200 may further include, at 228 ("Transmitting a first codeword and a second codeword"), transmitting, e.g., by the transmit antenna 160 of the mobile station 140, to the base station 104, a first codeword and a second codeword associated with the quantized beamforming matrix and the quantized difference matrix, respectively, as previously discussed.

Fig. 3 illustrates an exemplary method 300 for estimating, by base station 104, the beamforming matrix based on feedback received from the mobile station 140, in accordance with various embodiments of the present invention. One or more operations of the method 300 may be carried by one or more modules of the base station 104. Referring to Figs. 1 and 3, in various embodiments, the method 300 includes, at 304 ("Receiving a first codeword and a second codeword"), receiving, e.g., by the receive antenna 110, from the mobile station 140, e.g., from the transmit antenna 160, a first codeword and a second codeword associated with the quantized beamforming matrix V and the quantized difference matrix D , respectively.

The method 300 may further include, at 308 ("Determining the quantized beamforming matrix and the quantized difference matrix"), determining, e.g., by the beamforming matrix estimation module 116, the quantized beamforming matrix V and the quantized difference matrix D based at least in part on the received first codeword and the second codeword, respectively.

The method 300 may further include, at 312 ("Estimating a beamforming matrix"), estimating, e.g., by the beamforming matrix estimation module 116, the beamforming matrix (e.g., the estimated beamforming matrix V_b) from the determined quantized beamforming matrix V and the quantized difference matrix!⁾ , as discussed with respect to equation 7.

The method 300 may further include, at 316 ("Weighing one or more data streams"), weighing, e.g., by the beamformer module 112, one or more data streams (e.g., data stream Sl) using the estimated beamforming matrix V_b . The method 300 may further include, at 320 ("Transmitting the weighed data streams"), transmitting, e.g. by one or more transmit antennas (e.g., transmit antennas 108a, ..., 108d) of the base station 104, the weighed data stream(s) to the mobile station 140.

Quantizing the beamforming matrix using two codebooks Cb and Ca has several advantages relative to quantizing a beamforming matrix using a single codebook. For example, quantizing the beamforming matrix into the quantized beamforming matrix V and the quantized difference matrix D may reduce a quantization error. Accordingly, the estimation of the beamforming matrix, formed at the base station 104, may be more accurate.

In various embodiments discussed so far, a base codebook and a differential codebook is used to quantize the beamforming matrix. However, in various embodiments, more than one differential codebook may also be used. For example, once the difference matrix D and the quantized difference matrix D are generated, a second difference matrix may be generated (e.g., using an equation that is at least in part similar to equation 5), which may be representative of a difference between the difference matrix D and the quantized difference matrix D . The second difference matrix may then be quantized using a second differential 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 codewords

corresponding to the quantized beamforming matrix and the quantized difference matrix. The base station 104 may estimate the beamforming matrix f using codewords corresponding to the quantized beamforming matrix, the quantized difference matrix, and the second quantized difference matrix.

In the embodiments discussed so far, quantized form of the beamforming matrix

(which includes the quantized beamforming matrix V and the quantized difference matrix Z) ) is transmitted by the mobile station 140 to the base station 104. In various embodiments, the mobile station 140 may also transmit a quantized form of the channel covariance matrix R (e.g., instead of or in addition to transmitting the quantized form of the beamforming matrix) to the base station 104, to enable the base station 104 to reconstruct or estimate the channel covariance matrix and subsequently 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., obtained from equation 2) using a base codebook (using a method that is at least in part similar to equation 4) to obtain a quantized channel covariance matrix. The mobile station 140 may then calculate a corresponding difference matrix (using a method that is at least in part similar to equation 5) that is representative of a difference between the channel covariance matrix and the quantized channel covariance matrix. The mobile station 140 may quantize the corresponding difference matrix using a differential codebook (using a method that is at least in part similar to equation 6) to obtain a corresponding quantized difference matrix. The mobile station 140 may transmit codewords corresponding to the generated quantized matrices to the base station 104, from which the base station 104 may estimate the channel covariance matrix. The base station 104 may then estimate the beamforming matrix by decomposing (e.g., using singular value decomposition) the estimated channel covariance matrix.

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

System control logic 408 for one embodiment may include any suitable interface controllers to provide for any suitable interface to at least one of the processor(s) 404 and/or to any suitable device or component in communication with system control logic 408.

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

System control logic 408 for one embodiment may include one or more input/output (I/O) controller(s) to provide an interface to NVM/storage 416 and communications interface(s) 420.

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

The NVM/storage 416 may include a storage resource physically part of a device on which the system 400 is installed or it may be accessible by, but not necessarily a part of, the device. For example, the NVM/storage 416 may be accessed over a network via the communications interface(s) 420.

System memory 412 and NVM/storage 416 may include, in particular, temporal and persistent copies of beamforming matrix logic 424, respectively. In various embodiments, the system 400 may be a part of the mobile station 140, and the beamforming matrix logic 424 may include instructions that when executed by at least one of the processor(s) 404 result in the system 400 generating a beamforming matrix and/or quantizing the beamforming matrix (e.g., using a base codebook and a differential codebook), as described herein. In various other embodiments, the system 400 may be a part of the base station 104, and the beamforming matrix logic 424 may include instructions that when executed by at least one of the processor(s) 404 result in the system

400 estimating a beamforming matrix from received codewords (e.g., corewords corresponding to the beamforming matrix) of the base codebook and the differential codebook, as described herein.

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

Communications interface(s) 420 may provide an interface for system 400 to communicate over one or more network(s) and/or with any other suitable device.

Communications interface(s) 420 may include any suitable hardware and/or firmware.

Communications interface(s) 420 for one embodiment may include, for example, a network adapter, a wireless network adapter, a telephone modem, and/or a wireless modem. For wireless communications, communications interface(s) 420 for one embodiment may use one or more antennae.

For one embodiment, at least one of the processor(s) 404 may be packaged together with logic for one or more controller(s) of system control logic 408. For one embodiment, at least one of the processor(s) 404 may be packaged together with logic for one or more controllers of system control logic 408 to form a System in Package (SiP).

For one embodiment, at least one of the processor(s) 404 may be integrated on the same die with logic for one or more controller(s) of system control logic 408. For one embodiment, at least one of the processor(s) 404 may be integrated on the same die with logic for one or more controller(s) of system control logic 408 to form a System on Chip

(SoC).

In various embodiments, system 400 may have more or less components, and/or different architectures. Although certain example methods, apparatus, and articles of manufacture have been described herein, the scope of coverage of this disclosure is not limited thereto. On the contrary, this disclosure covers all methods, apparatus, and articles of manufacture fairly falling within the scope of the appended claims either literally or under the doctrine of equivalents. For example, although the above discloses example systems including, among other components, software or firmware executed on hardware, it should be noted that such systems are merely illustrative and should not be considered as limiting. In particular, it is contemplated that any or all of the disclosed hardware, software, and/or firmware components could be embodied exclusively in hardware, exclusively in software, exclusively in firmware or in some combination of hardware, software, and/or firmware.

Claims

CLAIMS What is claimed is:

1. A method comprising:

determining, by a mobile station, a channel covariance matrix based on signals received from a base station;

determining, by the mobile station, a beamforming matrix based at least in part on the determined channel covariance matrix;

selecting, by the mobile station, a first matrix from a first plurality of candidate matrices included in a first codebook, wherein the first matrix is representative of the beamforming matrix;

determining, by the mobile station, a difference matrix that is representative of a difference between the beamforming matrix and the first matrix;

selecting, by the mobile station, a second matrix from a second plurality of candidate matrices included in a second codebook, wherein the second matrix is representative of the difference matrix; and

transmitting, by the mobile station to the base station, a first codeword and a second codeword associated with the first matrix and the second matrix, respectively.

2. The method of claim 1, wherein determining the difference matrix further comprises:

determining a third matrix based at least in part on the first matrix, such that each of one or more columns of the third matrix is orthogonal to each of one or more columns of the first matrix.

3. The method of claim 2, wherein determining the difference matrix further comprises:

forming a fourth matrix that is a combination of the first matrix and the third matrix, wherein the fourth matrix is a unitary matrix having an order that is equal to a number of rows of the beamforming matrix; and

determining the difference matrix such that the difference matrix is a product of: a complex conjugate of the fourth matrix, and

the beamforming matrix.

4. The method of claim 2, wherein determining the third matrix further comprises: determining the third matrix using householder reflection on the first matrix

5. The method of claim 1, wherein selecting the first matrix further comprises:

selecting the first matrix such that the first matrix, among all the first plurality of candidate matrices, maximizes a Frobenius norm of a product of:

a complex conjugate of the beamforming matrix; and

the first matrix.

6. The method of claim 1, wherein selecting the second matrix further comprises: selecting the second matrix such that the second matrix, among all the second plurality of candidate matrices, maximizes a Frobenius norm of a product of:

a complex conjugate of the difference matrix; and

the second matrix.

7. The method of claim 1, wherein determining the beamforming matrix further comprises:

decomposing the channel covariance matrix into a left matrix, a diagonal matrix, and a complex conjugate of a right matrix using singular value decomposition; and

determining the beamforming matrix such that the beamforming matrix comprises one or more columns of the right matrix.

8. The method of claim 7, wherein the mobile station receives, from the base station, data signals associated with N data streams, where N is an integer value of one or more, and wherein determining the beamforming matrix further comprises:

determining the beamforming matrix such that the beamforming matrix comprises N columns of the right matrix.

9. The method of claim 1, wherein determining the channel covariance matrix further comprises:

determining a channel matrix that is representative of conditions of sub-channels between one or more transmit antennas of the base station and one or more receive antennas of the mobile station; and determining the channel covariance matrix from the channel matrix.

10. An apparatus comprising:

a matrix decomposition module configured to determine a beamforming matrix; a quantization module configured to determine, based at least in part on the beamforming matrix, a quantized beamforming matrix using a base codebook and quantized difference matrix using a differential codebook; and

a transmission module configured to transmit, to another apparatus, a first codeword and a second codeword associated with the quantized beamforming matrix and the quantized difference matrix, respectively.

11. The apparatus of claim 10, wherein the quantization module is further configured to:

select, from a first plurality of candidate matrices included in the base codebook, the quantized beamforming matrix such that the quantized beamforming matrix is representative of the beamforming matrix;

determine a difference matrix that is representative of a difference between the beamforming matrix and the quantized beamforming matrix; and

select, from a second plurality of candidate matrices included in the differential codebook, the quantized difference matrix such that the quantized difference matrix is representative of the difference matrix.

12. The apparatus of claim 11, wherein the quantization module is further configured to:

determine a first matrix such that each of one or more columns of the first matrix is orthogonal to each of one or more columns of the quantized beamforming matrix;

form a second matrix that is a combination of the quantized beamforming matrix and the first matrix; and

determine the difference matrix such that the difference matrix is a product of: a complex conjugate of the second matrix, and

the beamforming matrix.

13. The apparatus of claim 12, wherein the quantization module is further configured to determine the first matrix using householder reflection on the quantized beamforming matrix.

14. The apparatus of claim 10, wherein the quantization module is further configured to:

select the quantized beamforming matrix such that the quantized beamforming matrix, among all the first plurality of candidate matrices, maximizes a matrix norm of a product of:

a complex conjugate of the beamforming matrix, and

the quantized beamforming matrix.

15. The apparatus of claim 10, wherein the quantization module is further configured to:

select the quantized difference matrix such that the quantized difference matrix, among all the second plurality of candidate matrices, maximizes a matrix norm of a product of:

a complex conjugate of the difference matrix, and

the quantized difference matrix.

16. The apparatus of claim 10, further comprising a channel estimation module configured to:

determine a channel covariance matrix;

decompose the channel covariance matrix into a left matrix, a diagonal matrix, and a complex conjugate of a right matrix using singular value decomposition; and

determine the beamforming matrix such that the beamforming matrix comprises N columns of the right matrix, where N is an integer value of one or more, wherein the apparatus receives, from the another apparatus, data signals associated with N data streams.

17. A base station comprising:

a receiving module configured to receive, from a mobile station, a first codeword and a second codeword associated with a quantized beamforming matrix and a quantized difference matrix, respectively; and a beamforming matrix estimation module configured to:

determine the quantized beamforming matrix and the quantized difference matrix based at least in part on the received first codeword and the second codeword, respectively; and

estimate a beamforming matrix from the determined quantized beamforming matrix and the quantized difference matrix.

18. The base station of claim 17, wherein the beamforming matrix estimation module is configured to:

determine a first matrix such that each of one or more columns of the first matrix is orthogonal to one or more columns of the quantized beamforming matrix;

estimate the beamforming matrix, such that the estimated beamforming matrix is a product of a complex conjugate of the second matrix and the quantized difference matrix.

19. The base station of claim 17, wherein the beamforming matrix estimation module is further configured to determine the first matrix using householder reflection on the quantized beamforming matrix.

20. The base station of claim 17, further comprising:

a beamformer module configured to weight one or more data streams using the estimated beamforming matrix; and

one or more transmit antennas configured to transmit the weighed data streams.