CN102811111A

CN102811111A - Channel information feedback method and system

Info

Publication number: CN102811111A
Application number: CN2011101472951A
Authority: CN
Inventors: 姜静; 郁光辉; 朱常青
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2011-06-02
Filing date: 2011-06-02
Publication date: 2012-12-05
Also published as: WO2012163125A1

Abstract

The invention discloses a channel information feedback method and a system. The channel information feedback method comprises the following steps: a pre-coding matrix required to be fed back can be decomposed into the product of a plurality of Givens rotating matrixes, wherein each rotating matrix is only related to a rotating angle; and the corresponding rotating angle of the rotating matrix after being decomposed is fed back. The method and the system provided by the invention are a channel compression quantized feedback solution based on the Givens transformation, wherein the pre-coding matrix required to be fed back can be transformed into a limited angle value through continuous Givens rotation, and a base station can rebuild a channel matrix by using the fed-back quantitative angles so as to calculate a pre-coding vector, so that the feedback cost can be effectively reduced. Compared with the traditional scalar quantization, the channel information feedback method can effectively reduce the number of the quantitative elements, thereby decreasing the feedback quantity. With respect to the vector quantization, the only requirement is to carry out the operation of decomposition and quantization, so that the system complexity can be effectively lowered under the condition of ensuring the system performance.

Description

Channel information feedback method and system

Technical Field

The present invention relates to the field of communications, and in particular, to a channel information feedback method and system.

Background

The coordinated multi-point transmission (CoMP) technology can effectively improve the frequency utilization rate of the system, and can effectively suppress the inter-user interference through the inter-cell coordination processing. However, as the number of antennas increases and multiple cells cooperate, a user needs to simultaneously feed back channel information of a local cell and a cooperative cell thereof, and the feedback amount is large. Therefore, the user performs quantization compression on the estimated Channel State Information (CSI), and then feeds back the quantized bit information to the base station. And the base station reconstructs a channel according to the quantization information obtained by feedback and designs a precoding matrix. Therefore, interference among multiple users can be effectively inhibited through cooperation among cells, and the edge throughput and the average throughput of the cells are improved.

Conventional channel quantization compression schemes are mainly classified into two types: scalar quantization and vector quantization. One approach in scalar quantization is to quantize each element in the matrix to a fixed number of bits; another approach is to perform a fixed number of bits quantization on the diagonal elements and the upper triangular elements of the channel covariance matrix. Scalar quantization has a large relationship with the number of elements to be quantized, and when the number of elements to be quantized is large, a large feedback amount is required. Vector quantization can effectively reduce the feedback quantity, and the idea is as follows: firstly, selecting a code word with the maximum energy reaching the user from a predetermined codebook, and then feeding back a code word index. Compared with scalar quantization, vector quantization can effectively reduce the number of quantization bits and improve the system performance. However, in the CoMP system, since the codebook of each cell needs to be jointly optimized and searched, the complexity of the system is increased to a great extent.

Disclosure of Invention

In view of the above, the main objective of the present invention is to provide a method and a system for feeding back channel information, which effectively reduce the complexity of the system on the basis of ensuring the performance of the system.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

a channel information feedback method, the method comprising:

decomposing a precoding matrix needing to be fed back into a product of a plurality of Givens rotation matrixes, wherein each rotation matrix is only related to a rotation angle; feeding back the corresponding rotation angle of the rotation matrix which is decomposed.

The rotation angle of the feedback is

Phi value sum

A value of psi.

Phi is in the range of [0, 2 pi ], phi is in the range of [0, pi/2 ];

phi and psi are quantized with different bit numbers, respectively.

The rotation angle of the feedback is represented in the form of a feedback matrix having the format: v ═ V₁，V₂，...，V_N】；

Wherein, V₁Is a channel information feedback matrix, V, from the terminal to the first cooperative base station₂Is a channel information feedback matrix, V, from the terminal to the second cooperative base station_NIs the channel information feedback matrix from the terminal to the nth cooperative base station.

Each channel information feedback matrix adopts the same feedback format as that of a single cell; or,

and taking the first feedback matrix as a reference, and feeding back the last N-1 feedback matrices in a differential mode in sequence.

A channel information feedback system comprises a decomposition unit and a feedback unit; wherein,

the decomposition unit is used for decomposing the precoding matrix needing to be fed back into a product of a plurality of Givens rotation matrixes, and each rotation matrix is only related to the rotation angle;

the feedback unit is used for feeding back the corresponding rotation angle of the rotation matrix decomposed by the decomposition unit.

The rotation angle fed back by the feedback unit is

Phi value sum

A value of psi.

Phi is in the range of [0, 2 pi ], phi is in the range of [0, pi/2 ];

phi and psi are quantized with different bit numbers, respectively.

The rotation angle fed back by the feedback unit is represented in the form of a feedback matrix, and the format of the feedback matrix is as follows: v ═ V₁，V₂，...，V_N】；

Therefore, the channel information feedback technology of the invention is a channel compression quantization feedback scheme based on Givens transformation. The M-dimensional vector x can be converted into a vector with only one nonzero element through continuous Given rotation, a precoding matrix needing to be fed back can be converted into a limited angle value through the continuous Given rotation, the base station can reconstruct a channel matrix by using the feedback quantization angles, a precoding vector can be calculated, and the feedback overhead can be effectively reduced. Compared with the traditional scalar quantization, the method can effectively reduce the number of quantization elements, thereby reducing the feedback quantity; compared with vector quantization, only decomposition quantization operation is needed, and joint optimization search is not needed, so that the system complexity can be effectively reduced under the condition of ensuring the system performance.

Drawings

Figure 1 is a system model for CoMP;

fig. 2 is a simplified channel information feedback process according to an embodiment of the present invention;

fig. 3 is a system diagram of channel information feedback according to an embodiment of the present invention.

Detailed Description

In practical application, a channel compression quantization feedback scheme based on Givens rotation can be proposed. The precoding matrix can be decomposed into a series of Givens rotation matrix products by a correlation transformation, and each rotation matrix is only related to the rotation angle. Therefore, only by quantizing each rotation angle, the sending end can obtain a corresponding Givens rotation matrix by using the feedback information so as to reconstruct a channel matrix. It can be seen that the feedback content is the Given rotation matrix dependent rotation angle, and the finite rotation angle can be quantized with a fixed number of bits. Then, the base station side reconstructs a channel by using the quantized rotation angle obtained from the feedback link, and calculates a precoding matrix. Compared with the traditional scalar quantization, the method can effectively reduce the number of quantization elements, thereby reducing the feedback quantity; compared with vector quantization, only decomposition quantization operation is needed, and joint optimization search is not needed, so that the system complexity can be effectively reduced on the basis of ensuring the system performance.

In the system model under the multi-cell multi-user joint transmission scenario, it is assumed that there are two base stations (eNB)1 and eNB2 in the system, which jointly perform cooperative transmission for two User Equipments (UE)1 and UE 2, as shown in fig. 1. The number of antennas of the base station is M, and the number of antennas of the UE is N.

The channel from the base station eNBj (j 1, 2) to UE i (i 1, 2) is

It is assumed here that the UE can estimate accurate channel matrix information from two base stations to the UE through a downlink reference signal. In the present system, F may be used₁And F₂Representing the linear precoding matrices of UE 1 and UE 2, respectively, the received signal can be represented by the following equation:

y₁＝H₁₁F₁x₁+H₁₂F₂x₂+n₁

y₂＝H₂₁F₁x₁+H₂₂F₂x₂+n₂ (1)

said x₁And x₂Are data symbols, n, sent to UE 1 and UE 2, respectively₁And n₂Represents the additive Gaussian zero mean white noise (the variance of the noise is sigma)²). If the ue is an MRC (maximum ratio combining) receiver, the ue receiving channel ratio can be expressed as:

where i ∈ {1, 2}, j ∈ {1, 2} and i ≠ j. The UE rate may be represented by:

R_i＝log₂(1+r_i) (3)

the systematic sum rate can be derived from the symmetry of the UE:

C＝R₁+R₂ (4)

suppose that the precoding vector fed back by UE 1 is

And

the precoding vector fed back by UE 2 is

And

the precoding vectors respectively correspond to the local channel matrix and the right singular matrix of the cooperative channel matrix. In a cooperative beamforming scenario, if zero-forcing precoding is adopted, the precoding vector in the above equation is:

where null { A } represents the null space of matrix A, and svd { A } represents the eigenspace corresponding to the dominant singular values of matrix A.

Taking two-dimensional vector x ═ x₁ x₂]]^TAnd taking the argument psi of x as arctan (x)₂/x₁) Then there is a Given matrix:

then, G (ψ) x corresponds to the x falling on the x-axis after rotating x clockwise by an angle ψ in the x-y plane, and the modulus of the vector after rotation remains unchanged, i.e.:

the Givens matrix can be extended to a rank 2 correction matrix of the unit matrix:

G_li(ψ_l，j) Can rotate vector x ═ x₁ x₂ … x_M]^TI and l elements of (2), such that

And [ x ]]_l0, and the remaining positional elements of x are unchanged. This allows one element of x to be eliminated by a Givens rotation, and multiple elements to be eliminated by multiple rotations. If 1 is taken for i, and 2, 3, a, M in that order, x can be transformed into:

it should be noted that the precoding matrix to be fed back may be decomposed into a product of several Givens rotation matrices, and channel information may be reconstructed at the receiving end only by feeding back the corresponding rotation angle. The specific quantization compression steps are as follows:

(1) for local channel H at UE i (i ═ 1, 2)₁₁Or H₂₂SVD (singular value) decomposition is done (in case no ambiguity is caused, the subscript i is omitted in the following analysis):

H＝UAV^H (10)

the feedback content is required to be V_p＝V(：，1:N_s)(N_sM) where N is_sCorresponding to the number of data streams for the UE.

(2) Will V_pThe last row of elements becomes positive and real. Taking a matrix:

wherein theta is_i＝angle([V_p]_M，i) (i-1, 2) to obtain

(3) By D₁Will be provided withIs transformed into a positive real number to obtain

Wherein D₁The values are as follows:

(4) using Givens matrix G₂₁(ψ_2，1)，G₃₁(ψ_3，1)，...，G_M1(ψ_M，1) In turn will

The 2, 3.., M row elements of the first column are eliminated, resulting in:

the symbol "+" in the above formula represents any element or matrix.

V obtained after transformation₂The first element of the second column is 0, and the value is represented by V_pIs determined by the column orthogonality of (a).

(5) Repeating the steps (3) and (4) to process V₂The other columns of (2). For the ith column (i ═ 2, 3.., min { M-1, N }), there are:

the required Givens matrix is G in turn_i+1，i(ψ_i+1，i)，G_i+2，i(ψ_i+2，i)，...，G_M-1，i(ψ_M-1，i). Each column needs

A value of phi to change the column of elements to real numbers, then

The psi value is used to make a Givens rotation.

(6) Through the above transformation, V_pUnit array for M × N:

d for the above conversion_iAnd G_li(ψ_l，i) Are unitary matrices and can therefore be obtained using the inverse of the above transformation

As can be seen, the UE only needs to feed back

Phi value sum

A psi value, then the phi and psi can be used to reconstruct at the base station side

Similarly, the right singular matrix of the cooperative base station to the UE channel matrix may be fed back.

In practical applications, the above-mentioned value range of phi is [0, 2 pi ], and the value range of psi is [0, pi/2 ], which can be quantized by different bit numbers, for example:

k＝0，1，...，2^b+2-1

k＝0，1，...，2^b-1

(17)

where b is the number of quantization bits of ψ and b +2 is the number of quantization bits of φ.

3.4 feedback Format

When feedback is performed, the specific format of the feedback matrix is as follows:

V＝【V₁，V₂，...，V_N】 (18)

wherein V₁Is a channel information feedback matrix, V, from the terminal to the first cooperative base station₂Is a channel information feedback matrix, V, from the terminal to the second cooperative base station_NIs the channel information feedback matrix from the terminal to the nth cooperative base station. Each channel information feedback matrix can adopt the same feedback format as that of a single cell, and the last N-1 feedback matrixes can be fed back in a differential mode sequentially by taking the first feedback matrix as a reference.

The first embodiment is as follows:

the specific format of the feedback matrix is as follows:

V＝【V₁，V₂，...，V_N】 (18)

wherein V₁Is a channel information feedback matrix, V, from the terminal to the first cooperative base station₂Is a channel information feedback matrix, V, from the terminal to the second cooperative base station_NIs the channel information feedback matrix from the terminal to the nth cooperative base station. Each channel information feedback matrix may use the same feedback format as a single cell, i.e., the same quantization method as equation (17).

Example two:

the specific format of the feedback matrix is as follows:

V＝【V₁，V₂，...，V_N】 (18)

wherein V₁Is a channel information feedback matrix, V, from the terminal to the first cooperative base station₂Is a channel information feedback matrix, V, from the terminal to the second cooperative base station_NIs the channel information feedback matrix from the terminal to the nth cooperative base station. The first feedback matrix of each channel information feedback matrix is taken as a reference, and the last N is-1 feedback matrices are fed back in differential form in turn. I.e. the first matrix V₁That is, the quantization method is the same as the equation (17), where b is the number of quantization bits of ψ and b +2 is the number of quantization bits of Φ; the feedback values of the second matrix are Φ 2- Φ 1, ψ 2- ψ 1, and the differential values Φ 2- Φ 1 are fed back with b-2 bits and ψ 2- ψ 1 is fed back with b bits. The feedback values of the Nth matrix are phi N-phi 1 and psi N-psi 1, and the differential values phi N-phi 1 are fed back by b-2 bits and psi N-psi 1 are fed back by b bits.

As can be seen from the above description, the operation idea of performing channel information feedback in the present invention can be represented as a flow shown in fig. 2, where the flow includes the following steps:

step 210: the precoding matrix that needs to be fed back is decomposed into the product of multiple Givens rotation matrices, each rotation matrix being related to the rotation angle only.

Step 220: feeding back the corresponding rotation angle of the rotation matrix which is decomposed.

In order to ensure that the technical description and the operational idea described above can be implemented smoothly, an arrangement as shown in fig. 3 can be made. Referring to fig. 3, fig. 3 is a diagram of a channel information feedback system according to an embodiment of the present invention, where the system includes a decomposition unit and a feedback unit connected to each other.

In practical application, the decomposition unit can decompose the precoding matrix needing to be fed back into a product of a plurality of Givens rotation matrixes, and each rotation matrix is only related to a rotation angle; the feedback unit can feed back the corresponding rotation angle of the rotation matrix decomposed by the decomposition unit.

In summary, the channel information feedback technology of the present invention is a channel compression quantization feedback scheme based on Givens transform, regardless of the method or system. The M-dimensional vector x can be converted into a vector with only one nonzero element through continuous Given rotation, a precoding matrix needing to be fed back can be converted into a limited angle value through the continuous Given rotation, the base station can reconstruct a channel matrix by using the feedback quantization angles, a precoding vector can be calculated, and the feedback overhead can be effectively reduced. Compared with the traditional scalar quantization, the method can effectively reduce the number of quantization elements, thereby reducing the feedback quantity; compared with vector quantization, only decomposition quantization operation is needed, and joint optimization search is not needed, so that the system complexity can be effectively reduced under the condition of ensuring the system performance.

The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.

Claims

1. A channel information feedback method, comprising:

2. The method of claim 1, wherein the rotation angle fed back is

Phi value sum

A value of psi.

3. The method of claim 2, wherein phi is in the range of [0, 2 pi ], and psi is in the range of [0, pi/2 ];

phi and psi are quantized with different bit numbers, respectively.

4. A method according to any one of claims 1 to 3, wherein the rotation angle of the feedback is represented in the form of a feedback matrix having the format: v ═ V₁，V₂，...，V_N】；

5. The method of claim 4,

6. A channel information feedback system is characterized in that the system comprises a decomposition unit and a feedback unit; wherein,

7. The system of claim 6, wherein the rotation angle fed back by the feedback unit isPhi value sumA value of psi.

8. The system of claim 7, wherein phi is in the range of [0, 2 pi ], and psi is in the range of [0, pi/2 ];

phi and psi are quantized with different bit numbers, respectively.

9. The system according to any one of claims 6 to 8, wherein the rotation angle fed back by the feedback unit is represented in the form of a feedback matrix having the format: v ═ V₁，V₂，...，V_N】；

10. The system of claim 9,