JP2018152623A - Radio communication system, base station device, precoder determination method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a precoder capable of achieving high throughput with a low computation amount in MU-MIMO.SOLUTION: A method includes: determining at least one first precoder for each of a plurality of terminals S1; calculating a first index in consideration of interfering power on the basis of the first precoder S2; sorting the terminals on the basis of the first index S3; sequentially selecting a transmission precoder determination vector from a predefined set on the basis of a second index in consideration of interference suffered from other terminals at the terminals S4; and determining a second precoder to be used for transmission on the basis of the transmission precoder determination vector S5.SELECTED DRAWING: Figure 16

Description

  The present invention relates to a radio communication system, a base station apparatus, a precoder determination method, and a program.

  In order to improve frequency utilization efficiency, MIMO (Multiple Input Multiple Output) that spatially multiplexes data is attracting attention. In MIMO, antennas are arranged on both the transmitter and the receiver, and data is spatially multiplexed to improve frequency utilization efficiency. Each spatially multiplexed data is called a layer. Spatial multiplexing transmission is also called multi-layer transmission. The maximum number of layers is determined by the smaller of the number of antennas of the base station and the terminal. In communication between a single base station and a single terminal, the small number of antennas of the terminal becomes a bottleneck, and the improvement of frequency utilization efficiency is limited.

  To this problem. In MU-MIMO (Multi-User MIMO), a plurality of terminals are virtually regarded as one receiver, and the number of antennas on the receiver side is increased, thereby greatly improving frequency utilization efficiency. In MU-MIMO, a base station precodes transmission symbols to reduce interference between terminals and between layers.

  In order to improve SINR (Signal to Interference and Noise Ratio) in a terminal that receives a signal from a base station, a precoder in the base station is important. The SINR at each terminal is directly linked to the throughput characteristics of the system (note that the throughput is evaluated based on the SINR using, for example, the Shannon capacity formula (BW · log2 (1 + SINR): BW is the bandwidth). ).

  For this reason, it is preferable to determine a precoder that achieves the highest possible SINR in the base station.

  Hereinafter, a typical example of related technology relating to determination of a precoder will be described. In related technology, as a precondition for precoder determination, in a TDD (Time Division Duplex) system using the same carrier frequency in uplink and downlink, the channel of the propagation path between the base station and each terminal by the base station by the following procedure Get a matrix.

  The base station estimates an uplink channel based on an uplink reference signal, and further converts uplink channel information into downlink channel information using uplink and downlink channel reciprocity in TDD.

  In Zero Forcing (ZF) described in Non-Patent Document 1 and Block Diagonalization (BD) described in Non-Patent Document 2, a plurality of terminals are spatially orthogonalized so that To eliminate the interference. For this reason, high system throughput can be obtained.

ただし、肩のHは行列の複素共役をとり転置させるエルミート演算子である。 As an orthogonalization method, the ZF method uses a precoder (also called precoding matrix) as the inverse matrix of the channel matrix, so that a plurality of terminals are spatially orthogonalized, and data destined for one terminal interferes with other terminals. Control not to give. In ZF, the channel matrix is H, and the precoder W is given as follows, for example.

The shoulder H is a Hermitian operator that takes the complex conjugate of the matrix and transposes it.

のV_i ^(０)である。U_i (i=1,2)はM×Mのユニタリ行列、D_iは、L_i個(L_iはH_iのrank)の正特異値の対角成分を有し残りの対角成分及び非対角成分に０を有するM×N_Tの行列、V_i ⁽¹⁾ (i=1,2)は、L_i個の正の特異値に対応するL_i個のN_T次元の正規直交べクトルからなるN_T×L_iの行列、Vi^(０) (i=1,2)はH_i (i=1,2)のゼロ特異値に対応する（N_T-L_i）個のN_T次元の正規直交ベクトルを列ベクトルとするN_T×(N_T -L_i)の行列である。各端末では、基地局を複数のアンテナ（M本）で受信し受信信号に対して受信マトリクス（ポストコーダ）を演算することで複数（M）レイヤから成るシンボルを得る。 Also, in the block diagonalization (BD) method, multiple terminals are spatially separated by multiplying a precoder that multiplies a transmission signal addressed to a certain terminal by a singular vector corresponding to a zero singular value of a channel matrix to other terminals. They are orthogonal. For example, referring to FIG. 18, the base station 10 maps one or more code words (Code Word) as a complex symbol to a plurality of layers (in FIG. 18, the number of layers is M), and each layer is multiplied by the precoder W. And assigned to each antenna. A signal for each antenna is mapped to a time-frequency resource element and transmitted. In the BD method, H × W is block diagonalized with respect to the channel matrix H to obtain a matrix W that makes interference with other terminals zero. In MU-MIMO of a base station (number of antennas: N _T ) and two terminals (number of antennas: N _R ) (the number of layers M is equal to or less than the number of antennas N _T and N _R , in FIG. 18, N _R = M) , with respect to the channel matrix H ₁ between the base station and the terminal _{1,2, H 2 (N T ×} M), V 2 (0) in the precoder W ₁ for terminal 1 (zero singular channel matrix of H ₂ to the terminal 2 is multiplied a set of singular vectors corresponding to the value) is a set of singular vectors corresponding to the zero singular values of the channel matrix H ₁ to V ₁ ⁽⁰⁾ (terminal 1 in the precoder W ₂ for the terminal 2) is multiplied by The Where V _i ⁽⁰⁾ (i = 1,2) is the singular value decomposition (SVD) of the M × N _T channel matrix H _i :

V _i ⁽⁰⁾ . U _i (i = 1,2) is an M × M unitary matrix, D _i is L _i (L _i is the rank of H _i ) positive singular value diagonal components and the remaining diagonal components matrix M × N _T with 0 in the non-diagonal ^{_{components, V i (1) (i}} = 1,2) is, L _i number of N _T dimensional orthonormal corresponding to L _i number of positive singular values N _T × L _i matrix consisting of vectors, Vi ⁽⁰⁾ (i = 1,2) is (N _T -L _i ) N corresponding to zero singular values of H _i (i = 1,2) An N _T × (N _T -L _i ) matrix having a _T- dimensional orthonormal vector as a column vector. In each terminal, a base station is received by a plurality of antennas (M), and a reception matrix (post coder) is calculated for the received signal to obtain a symbol composed of a plurality of (M) layers.

  However, the precoder determination method of the related technique requires high-cost operations such as matrix singular value decomposition and inverse matrix operation. For this reason, the block diagonalization method is not practical.

  Further, in the block diagonalization method, when there is a channel estimation error, the orthogonality between terminals is lost, and the characteristics are greatly deteriorated.

に設定する。すなわち、チャネル行列Hの複素共役転置をプリコーダに設定するという簡単な演算によって、受信端における所望信号電力を最大にすることができる。 In MRT (Maximum Ratio Transmission) described in Non-Patent Document 3, a precoder is used.

Set to. That is, the desired signal power at the receiving end can be maximized by a simple operation of setting the complex conjugate transpose of the channel matrix H to the precoder.

  However, MRT cannot take into account interference between terminals. For this reason, in MRT, there is a possibility that SINR at the receiving end is greatly deteriorated due to interference.

  In TM5 (Transmission Mode 5) (see Non-Patent Document 4) in LTE (Long Term Evolution) adopted as one of the so-called 3.9G (3.9G) mobile radio communication systems, a precoder determined in advance is used. Select a precoder from the table (codebook). A plurality of precoders registered in the codebook are generally orthogonal to each other in order to suppress interference between beams. Hereinafter, an example of the operation of TM5 will be described.

  As described above, the base station estimates the uplink channel based on the uplink reference signal from the terminal, and estimates the downlink channel matrix using the channel reciprocity.

  Thereafter, the base station searches for a combination that maximizes the system throughput from all combinations of all terminals and precoders in the codebook.

  In TM5, since the precoder is selected from the code book, the amount of calculation for determining the precoder is small.

  However, the assignable precoders are limited to precoders defined in advance in the codebook. For this reason, TM5 has a possibility that the optimality of the selected precoder is lower and the throughput characteristic is lower than the method of calculating the precoder directly from the channel matrix like BD, ZF, and MRT. .

Jinsu Kim, Sungwoo Park, et al, “A Scheduling Algorithm Combined with Zero-forcing Beamforming for a Multiuser MIMO Wireless System,” in Proc. IEEE VTC 2005 Fall, Sep. 2005. (page 212) Q. Spencer and M. Haardt, “Capacity and downlink transmission algorithms for a multi-user MIMO channel,” in Proc. 36th Asilomar Conf. On Signals, Systems, and Computers, Pacific Grove, CA, IEEE Computer Society Press, Nov. 2002. (1385 pages) T. Lo, “Maximum ratio transmission,” IEEE Transactions on Communications, vol. 47, no. 10, pp. 1458-1461, 1999. (page 1459) E. Dahlman, S. Parkvall, and J. Skold, “4G LTE / LTE-Advanced for Mobile Broadband,” Academic Press, 2011.

  In the LTE system, the TM5 method is adopted instead of ZF and BD. As described above, as described above, for example, the calculation amount of ZF and BD is enormous, and the channel error tolerance of ZF and BD is low.

  On the other hand, in recent years, the number of antennas has been increased due to factors such as cost reduction of RF (Radio Frequency) circuits. In a multi-element antenna, the array gain and the number of spatially multiplexed terminals increase due to an increase in antenna freedom. For this reason, system throughput can be significantly increased.

  However, if TM5 is simply applied as a method for determining a precoder for a multi-element antenna, the amount of calculation required for selecting a precoder is greatly increased.

As the cause, for example,
・ Increase of precoder matrix size by increasing the number of transmitting antennas, the number of multiple terminals, and the number of multiple layers,
・ Increase the number of combinations of terminals and precoders
Etc.

The matrix size of the precoder is
(Number of transmit antennas) x (Total number of multiplexed layers)
It is.

  In order to cope with an increase in the amount of computation necessary for selecting a precoder in a base station, it is necessary to improve the computation processing capability of the base station. For this reason, the hardware cost of a base station increases.

  Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to provide a system, apparatus, method, and program that can reduce the amount of calculation required for determining a precoder.

According to one aspect of the present invention, a base station that precodes a transmission signal and transmits a plurality of antennas to a plurality of terminals determines a transmission precoder,
A first step of obtaining at least one first precoder for each of the plurality of terminals in a predetermined manner;
A second step of calculating a first index considering the interference power based on the first precoder;
A third step of sorting the plurality of terminals based on the first index;
A fourth step of sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal;
A fifth step of determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A transmission precoder determination method is provided.

According to another aspect of the present invention, a base station apparatus that precodes a transmission signal and transmits the signal to a plurality of terminals from a plurality of antennas,
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
There is provided a base station apparatus including a second precoder calculation unit that determines a second precoder to be used for transmission based on at least one transmission precoder determination vector.

According to still another aspect of the present invention, the base station comprises: a plurality of terminals; and a base station that precodes a transmission signal and transmits the signals from a plurality of antennas to a plurality of terminals to be transmitted.
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
A second precoder calculation unit for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A wireless communication system is provided.

According to still another aspect of the present invention, a computer of a radio station that precodes a transmission signal and transmits it from a plurality of antennas to a plurality of terminals,
A first precoder calculation process for determining at least one first precoder for the terminal by a predetermined method;
A first index calculation process for calculating a first index considering interference power based on the first precoder;
A terminal sorting process for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. Selection process,
A second precoder calculation process for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A program for executing is provided. According to the present invention, a computer-readable recording medium (semiconductor memory, magnetic recording medium, CD (Compact Disk) / DVD (Digital Versatile Disk), etc.) on which the program is recorded is provided.

  According to the present invention, it is possible to reduce the amount of calculation for precoder determination, and it can be applied to a base station of a multi-element antenna.

It is a figure which illustrates the structure of the base station of the 1st exemplary embodiment of this invention. It is a figure which illustrates the directivity which each vector in the predefined set of the 1st exemplary embodiment of this invention shows. 3 is a flowchart illustrating the operation of the base station of the first exemplary embodiment of the present invention. It is a figure which illustrates the structure of the base station of the 2nd exemplary embodiment of this invention. 6 is a flowchart illustrating an operation of a base station according to the second exemplary embodiment of the present invention. It is a figure which shows the structure of the base station of the 3rd exemplary embodiment of this invention. 7 is a flowchart illustrating the operation of a base station according to the third exemplary embodiment of the present invention. It is a figure which illustrates the structure of the base station of the 4th exemplary embodiment of this invention. 6 is a flowchart illustrating an operation of a base station according to the fourth exemplary embodiment of the present invention. It is a figure which illustrates the structure of the base station of the 5th exemplary embodiment of this invention. 7 is a flowchart illustrating the operation of a base station according to a fifth exemplary embodiment of the present invention. It is a figure which illustrates the structure of the base station of the 6th exemplary embodiment of this invention. 12 is a flowchart illustrating an operation of a base station according to the sixth exemplary embodiment of the present invention. It is a figure which illustrates the structure of the base station of the 7th exemplary embodiment of this invention. 12 is a flowchart illustrating an operation of a base station according to the seventh exemplary embodiment of the present invention. It is a figure explaining the basic form of the present invention. It is a figure explaining embodiment of this invention. It is a figure explaining one of related technology.

  A mode for carrying out the present invention will be described below.

  According to the present invention, a base station (wireless station) having a plurality of antennas selects at least one vector from a set of predefined vectors (predefined set) and calculates based on the selected vector. (Second precoder) multiplexes and transmits data to a plurality of terminals at the same time and the same frequency.

According to one aspect of the present invention, referring to FIG. 16, a radio station such as a base station that communicates wirelessly with a terminal determines a transmission precoder by the following steps.
(S1) At least one first precoder is obtained for each of a plurality of terminals by a predetermined method (algorithm).
(S2) Based on the first precoder, a first index considering the interference power is calculated.
(S3) The plurality of terminals are sorted based on the first index.
(S4) Based on a second index considering inter-terminal interference at the terminal, a vector for transmission precoder determination is sequentially selected for the terminal from the vector set of the predefined set.
(S5) A second precoder used for transmission is determined based on at least one transmission precoder determination vector. It goes without saying that these processes and functions may be realized by a program executed by a computer constituting a component of the radio station. In the following, some exemplary embodiments will be described.

<First Exemplary Embodiment>
According to the first exemplary embodiment of the present invention, a base station having a plurality of antennas spatially multiplexes a plurality of terminals having a plurality of antennas, and performs data transmission in a plurality of layers. A single layer transmission is performed for each terminal.

  FIG. 1 is a diagram illustrating the system configuration of the first exemplary embodiment. Referring to FIG. 1, a base station 10_1 having a function of a precoder determination method includes a communication unit 101 for communicating with a terminal, an effective channel vector calculation unit 102, an effective channel vector storage unit 103, and a predefined set storage unit. 104, first precoder calculation unit 105, first index calculation unit 106, terminal sort unit 107, transmission precoder determination vector selection unit 108, second precoder calculation unit 109, and transmission precoder calculation unit 110 and a control unit 111. Each unit 101-110 is configured to be able to exchange necessary information via the control unit 111. Of course, the processing of at least one of the units 101-111 of the base station 10_1 in FIG. 1 may be realized by a program executed on the computer of the base station 10_1. The same applies to other embodiments described below. Hereinafter, each part of FIG. 1 will be described.

  The effective channel vector calculation unit 102 receives the uplink reference signal received by the communication unit 101 from the communication unit 101, calculates the effective channel vector of each terminal from the estimated downlink channel matrix, and the effective channel vector storage unit 103. To store. A method for calculating the effective channel vector will be described later.

が格納されている。 The predefined set storage unit 104 includes a predefined set.

Is stored.

は、次式（１）で表され、そのｉ_set番目の要素であるベクトル：

は、次式（２）で表され、複素数体Ｃ上に定義されたN_ｔ行×１列の行列、すなわちN_ｔ次元複数ベクトル空間上の複素ベクトルである。 Predefined set

Is expressed by the following equation (1), and is a vector that is the i _set- th element:

Is a matrix of N _t rows × 1 column defined by the following equation (2) and defined on the complex number field C, that is, a complex vector on an N _t -dimensional multi-vector space.

内のベクトルの数である。 N _set is a predefined set

Is the number of vectors in.

（2)

(2)

  In equation (1) and the like, {p, q, r, s...} Means a set having p, q, r, s. Moreover, in Formula (2), T of the shoulder represents transposition.

に対して

は行列の正規化を意味し、

である。ただし

は行列

のフロベニウスノルムである。すなわち、

をm×nの行列とすると、

Any matrix

Against

Means matrix normalization,

It is. However,

Is a matrix

The Frobenius norm. That is,

Is an m × n matrix,

である。ただし、N_tは基地局のアンテナ素子数である。 In the exemplary embodiment, the number of vectors in the predefined set:

It is. Here, _Nt is the number of antenna elements of the base station.

の任意の２つの要素（ベクトル）は互いに直交しているものとする。

Also, predefined set

Any two elements (vectors) are orthogonal to each other.

の場合に、事前定義セット

の各ベクトルが示す指向性の一例を図２に示す。図２において、縦軸はビームゲイン、横軸はブロードサイド（１次元配列の軸に垂直な方向）からの角度オフセットである。 For example,

The predefined set

An example of the directivity indicated by each vector is shown in FIG. In FIG. 2, the vertical axis represents the beam gain, and the horizontal axis represents the angle offset from the broad side (direction perpendicular to the one-dimensional array axis).

  The first precoder calculation unit 105 inputs an effective channel vector stored in the effective channel vector storage unit 103. The first precoder calculation unit 105 calculates the first precoder from the effective channel vector based on the MRT precoder generation method. The calculation of the first precoder will be described later.

  The first index calculation unit 106 receives an effective channel vector from the effective channel vector storage unit 103. Further, the first index calculation unit 106 inputs the first precoder calculated by the first precoder calculation unit 105.

  The first index calculation unit 106 calculates the first index from the input effective channel vector and the first precoder based on the interference power. The calculation of the first index will be described later.

The terminal sort unit 107 inputs the calculation result of the first index from the first index calculation unit 106. The terminal sort unit 107 sorts the terminals 20_1 to 20_N _users in ascending order of the first index.

  The transmission precoder determination vector selection unit 108 inputs the terminal sort result by the terminal sort unit 107. Further, transmission precoder determination vector selection section 108 receives an effective channel vector from effective channel vector storage section 103. Further, the transmission precoder determination vector selection unit 108 inputs a predefined set from the predefined set storage unit 104.

  The transmission precoder determination vector selection unit 108 calculates a transmission precoder determination vector based on the input information (terminal sort result, effective channel vector, predefined set).

  Second precoder calculation unit 109 inputs a transmission precoder determination vector from transmission precoder determination vector selection unit 108. Second precoder calculation unit 109 calculates a second precoder based on the transmission precoder determination vector.

  The transmission precoder calculation unit 110 inputs the second precoder calculated by the second precoder calculation unit 109. The transmission precoder calculation unit 110 calculates a transmission precoder based on the input second precoder.

  The control unit 111 performs communication control according to the first exemplary embodiment and controls each unit 101-110.

The control unit 111 controls transmission of data from the communication unit 101 to a plurality of (N _user ) terminals (20_1 to 20_N _user ) based on the transmission precoder calculated by the transmission precoder calculation unit 110.

  FIG. 3 is a flowchart illustrating the operation of the first exemplary embodiment. Next, the operation of the base station in the first exemplary embodiment will be described with reference to FIGS. 1 and 3.

<Calculation of effective channel vector>
The effective channel vector calculation unit 102 estimates the channel matrix of the propagation path between the base station 10_1 and the terminals 20_1 to 20_N _user from the reception signal of the uplink reference signal acquired by the communication unit 101.

  The effective channel vector calculation unit 102 calculates an effective channel vector based on the channel matrix of the propagation path between the base station 10_1 and each terminal, and stores the calculated effective channel vector in the effective channel vector storage unit 103 (S101). .

  An example of an effective channel vector calculation procedure in the effective channel vector calculation unit 102 of FIG. 1 will be described below.

FIG. 17 shows a channel matrix (H _{1 to} H _Nuser ) of a transmission path between the base station 10 and N _user terminals 20, a precoder (W _{2,1 to} W _{2, Nuser} ) in the base station 10, the post coder (reception ^{_{matrix) (U H 1 ~U H Nuser}} ) is illustrated FIG schematically. The channel matrix _{H i (i = 1~N user)} , the left singular matrix U _i and the right singular matrix V _i obtained by singular value decomposition, as follows, respectively, and U _i of postcoder U ^H _i of the terminal 20, The precoder W _{2, i} of the base station 10 is assumed.

The received signal (vector) at the terminal i (20_i) with respect to the transmission data s _i for the terminal _i is expressed by, for example, the following (noise component is omitted).

In the first exemplary embodiment, in FIG. 17, the number of layers N _layer is 1, and in the base station 10, symbols are mapped to data of one layer (1 input N _t output: N _t- dimensional complex vector). And assigned to each of Nt antennas. Each terminal 20 multiplies the signal received from the N _r antennas by a post decoder (for example, N _r input 1 output: complex conjugate transpose of N _r dimensional complex vector), and outputs one layer of data.

と、端末インデックス＝i_userの端末２０における推定ポストコーダ（Nt次元複素ベクトル）

はそれぞれ次式（３）、（４）で算出される。なお、推定ポストコーダは、基地局１−＿１で推定された端末２０のポストコーダである。Nt、Nrはそれぞれ基地局１０＿１のアンテナ数と端末２０のアンテナ数である。 Effective channel vector (Nt-dimensional complex vector)

And the estimated postcoder (Nt-dimensional complex vector) in the terminal 20 with terminal index = i _user

Are calculated by the following equations (3) and (4), respectively. The estimated postcoder is a postcoder of the terminal 20 estimated by the base station 1-_1. Nt and Nr are the number of antennas of the base station 10_1 and the number of antennas of the terminal 20, respectively.

(3)

(3)

(4)

(Four)

は、基地局１０＿１とi_user番目(i_user＝１,…,N_user)の端末２０の伝搬路のチャネル行列（N_ｒ×N_ｔの複素数値の行列）である。N_userは同時に送信する対象となる端末２０の数である。 here,

Is the channel matrix (N _r × N _t complex value matrix) of the propagation path of the base station 10_1 and the i _user- th (i _user = 1,..., N _user ) terminal 20. N _user is the number of terminals 20 to be simultaneously transmitted.

はチャネル行列：

の第１の左特異ベクトルである。

はN_ｒ次元複素ベクトルであり、以下の式（６）の１番目の要素（ベクトル）で与えられる。

Is the channel matrix:

Is the first left singular vector.

Is an _Nr- dimensional complex vector, and is given by the first element (vector) of the following equation (6).

は、次式（５）にように特異値分解（Singular Value Decomposition: SVD）される。 Channel matrix

Is subjected to singular value decomposition (SVD) as shown in the following equation (5).

(5)

(Five)

はN_r×N_ｒのユニタリ行列であり、以下のように、N_R個のN_R次元のベクトルからなる。 In the above equation (5),

Is a unitary matrix of N _r × N _r, as follows, consisting of the N _R N _{R-dimensional} vector.

(6)

(6)

はN_t×N_tのユニタリ行列である。

Is an N _t × N _t unitary matrix.

はN_r×N_ｔの対角行列である。

(7)

Is a diagonal matrix of N _r × N _t .

(7)

は、p, q, r, s・・・を対角要素に持つ対角行列を表す。 However,

Represents a diagonal matrix having p, q, r, s... As diagonal elements.

は、a, b, c, d・・・の中で最小の値を表す演算子である。

Is an operator representing the smallest value among a, b, c, d.

はチャネル行列

の特異値：

を対角成分に持つ対角行列であり、特異値行列ともいう。 As shown in the above equation (7),

Is the channel matrix

Singular value of:

Is a diagonal matrix, and is also called a singular value matrix.

は、行列：

の固有値：

の平方根である。

Singular value:

The matrix:

Eigenvalues of:

Is the square root of

は左から順に大きな値とされる。

Above singular values:

Is a large value in order from the left.

In the above equation (5),

はそれぞれチャネル行列

Is the channel matrix

の特異値

に対応する左特異ベクトルと右特異ベクトルを順に並べたものである。 Left singular matrix and right singular matrix of

Singular values of

The left singular vector and the right singular vector corresponding to are arranged in order.

と

のｊ番目の固有値の固有ベクトルであり、それぞれ以下の式を満たす。
The j-th left singular vector u _j and the right singular vector v _j are respectively

When

Are the eigenvectors of the j-th eigenvalue, and satisfy the following expressions respectively.

は、以下を満たす。 Therefore, the first left singular vector of equation (4)

Satisfies the following.

を読み出し、ｉ_user番目の第１のプリコーダ <Calculation of the first precoder>
Next, the first precoder calculation unit 105 receives the effective channel vector from the effective channel vector storage unit 103:

I _user- th first precoder

を次式(８)により算出する（Ｓ１０２）。

Is calculated by the following equation (8) (S102).

  However, * on the shoulder is an operator that takes the complex conjugate of the original value of each element.

と、第１のプリコーダ計算部１０５で算出された第１のプリコーダ

を用いて第１の指標を計算する（Ｓ１０３）。 <Calculation of the first index>
Subsequently, the first index calculation unit 106 reads the effective channel vectors of all the transmission target terminals read from the effective channel vector storage unit 103.

And the first precoder calculated by the first precoder calculation unit 105

Is used to calculate the first index (S103).

は、基地局１０＿１が第１のプリコーダ計算部１０５で算出された第１のプリコーダ First indicator

Is the first precoder calculated by the first precoder calculation unit 105 by the base station 10_1.

を用いてデータを送信した際のｉ_user番目以外の端末への与干渉電力の和である。第１の指標は、次式（９）で計算される。

Is the sum of the interference power to terminals other than the i _userth when data is transmitted using. The first index is calculated by the following equation (9).

Further, as a first index calculation method, the first precoder calculation unit 105 of the base station 10_1 uses the first precoder.

May be an arithmetic average value, harmonic average value, weighted average value, and median value of interference to terminals other than the i _user- th terminal when data is transmitted using.

  The magnitude of interference power from one link between a base station and a terminal to another link is related to the propagation path and the transmission weight.

  On the other hand, one of the purposes of the precoder is to increase the reception power at the reception end.

  In the first exemplary embodiment, the interference is estimated with the MRT precoder that can transmit power most strongly to a certain terminal as the first precoder. As a result, the base station precoder is not specifically assumed, and the interference closer to the precoder used in actual transmission is estimated than when the base station assumes omni directivity (omnidirectional directivity). be able to.

As the calculation of the first index, i _user-th terminal is the arithmetic mean value of the interference to i _user th other terminal when sending the data using the first precoder, harmonic mean, weighted It may be an average value or a median value.

を算出する（Ｓ１０４）。 The first precoder calculation unit 105 and the first index calculation unit 106 perform steps S102 and S103 on all N _user terminals (20_1 to 20_N _user ), respectively, so that the first index

Is calculated (S104).

に基づいて、端末のソートを行う（Ｓ１０５）。具体的には、例えば、第１の指標が小さい端末から大きい端末へ順に、端末インデックスｉ'_userを再割り当てする（例えばソートの結果、第１の指標が最も小さな端末に対して端末インデックス＝１を再割り当てし、第１の指標が次に大きな端末に対して端末インデックス＝２を再割り当てする、という割り当て手法が用いられる）。 <Sort terminal>
Further, the terminal sorting unit 107 includes a first index of each terminal calculated by the first index calculation unit 106.

Based on the above, the terminals are sorted (S105). Specifically, for example, the terminal index i ′ _user is reassigned in order from the terminal having the smallest first index to the terminal having the largest first index (for example, the terminal index = 1 for the terminal having the smallest first index as a result of sorting). Is assigned and terminal index = 2 is reassigned to the terminal having the next largest first index).

と、
事前定義セット記憶部１０４から読み出した事前定義セットと、
に基づいて、第２の指標を計算する。 <Selection of transmission precoder determination vector>
The transmission precoder deciding vector selection unit 108 is the effective channel vector of all transmission target terminals read from the effective channel vector storage unit 103.

When,
A predefined set read from the predefined set storage unit 104;
Based on the above, a second index is calculated.

を逐次的に選択する（Ｓ１０６）。 Furthermore, the transmission precoder determination vector selection unit 108 maximizes the second index in ascending order of the terminal index i ^′ _user acquired from the terminal sort unit 107.

Are sequentially selected (S106).

  Specific calculation of the precoder determination vector for maximizing the second index by the transmission precoder determination vector selection unit 108 is given by, for example, the following equation (10).

を最大にする引数（argmax）

を表している。 The above formula (10) is

Argument that maximizes (argmax)

Represents.

は、事前定義セットのｉ_set番目の要素であるベクトル

とされる。 However, the transmission precoder decision vector to be searched

Is a vector that is the i _set th element of the predefined set

It is said.

は、逐次探索の中で既に割り当てが決まった送信プリコーダ決定用ベクトル

を要素とする集合である。

Is a transmission precoder determination vector that has already been assigned in a sequential search.

Is a set whose elements are

は、

すなわち、過去の逐次探索で既に割り当てが決まった送信プリコーダ決定用ベクトルの集合

に含まれないベクトルとされる。 Vector for determining transmission precoder to be searched (vector of i _set- th element of predefined set)

Is

That is, a set of transmission precoder determination vectors that have already been assigned in the past sequential search

Is not included in the vector.

は１番目からｉ^' _user番目の端末の間での端末間干渉を考慮した、１番目からｉ^' _user番目の端末（k_user=i'_user）までのレートの和である。すなわち、Shannonの容量式に従う到達可能な情報レート：log₂（1＋SINR_ｋuser）のk_user=1からk_user=i'_userまでの和（Σ）である。 Of formula (10)

The ^'considering inter-terminal interference between _user-th terminal, the first ^i' from the first i is the sum of rates up _user-th terminal (k _user = i _'user). That, Shannon capacity reachable accordance expression information rate: the sum of the k _user = 1 of log ₂ (1 + SINR _kuser) to _{k user = i 'user (Σ} ).

As shown in Expression (10), in the rate sum, the rate of a certain terminal (k _user- th) includes the SINR of the terminal.

は、基地局１０がプリコーダ

を用いて、ｋ_user番目の端末宛に送信した信号電力である。 Moreover, in the said rate sum, the electric power of the signal transmitted to the said terminal 20 from the base station 10 is considered as a desired signal. Molecule of formula (10):

The base station 10 is a precoder

_Is the signal power transmitted to the k _user- th terminal.

は、基地局１０がプリコーダ

を用いてｋ_user番目の端末以外（ｌ_user≠ｋ_user）の端末２０に送信したときの端末間の干渉電力の和である。 Further, in the rate sum, as the interfered signal, the power of the signal transmitted to the terminal 20 other than the k _user- th terminal is considered. That is, the denominator term in equation (10):

The base station 10 is a precoder

_Is the sum of the interference power between the terminals when transmitted to the terminals 20 other than the k _user th terminal (l _user ≠ k _user ).

  In this specification, the operation of sequentially selecting the precoder determination vector of each user is referred to as “sequential search” in this specification.

を選択する（Ｓ１０７）。 The transmission precoder determination vector selection unit 108 repeats step S106 for all terminal indexes (i _user = 1 to N _user ), whereby transmission precoder determination vectors for all terminals (20_1 to 20_N _user ).

Is selected (S107).

  In the following, when searching for a vector corresponding to a transmission precoder determination vector, the degradation of system characteristics due to the use of sequential search instead of full search is suppressed by sorting in ascending order of the first index. Explain why this is possible.

In general, the influence of a certain terminal vector (transmission precoder determination vector) on the rate sum, which is a characteristic of the entire system, is
(A) the desired signal power used to calculate the rate of itself (own terminal);
(B) Interference power used to calculate the rate of terminals other than itself (own terminal),
There are two.

の探索の仕方によって、これら２つの要素(A)、（B）を考慮できる度合いが異なる。このため、通信特性にも差が生じる。 Transmission precoder decision vector

The degree to which these two elements (A) and (B) can be considered differs depending on the search method. For this reason, a difference also arises in communication characteristics.

を決定している。 In the first exemplary embodiment, in order to reduce the amount of calculation required for the calculation of the transmission precoder, the transmission precoder determination vector is not a full search but a sequential search.

Is determined.

  Expression (11) shows a search expression for an optimal vector for determining a transmission precoder in the full search.

を探索する。 As shown in Expression (11), in the case of searching for an optimal vector for determining a transmission precoder by full search, the rate sum of all terminals to be transmitted is used as an optimization index. That is, a collection of N _user transmission precoder determination vectors that maximize the rate sum:

Explore.

  For this reason, in the full search, it is possible to perform a search in consideration of all the two effects (A) and (B) described above. As described above, in the search for the optimal vector for determining the transmission precoder by the full search, the optimal solution for the entire system is determined.

  On the other hand, in the case of searching for an optimal vector for transmission precoder determination by sequential search, as shown in the above equation (10), when searching for a certain transmission precoder determination vector, some terminals among the transmission target terminals Is the optimization index.

  Here, a part of terminals are a terminal to which a transmission precoder determination vector is already assigned in a sequential search and a terminal to be optimized.

For example, it is assumed that there are first to third terminals in ascending order of the terminal index i ^′ _user , and the predefined set is A = {a1, a2, a3}. At the time of determining the transmission precoder determining vector of the first terminal, B _alloc = φ (empty set), and the transmission precoder determining vector of the first terminal is
The optimization metric is the rate of the first device,
Search range for all elements of predefined set A,
Then, optimization is performed (the element that maximizes the rate of the first terminal is determined from the search range). As a result of this optimization, it is assumed that the vector a1 in the predefined set A is assigned to the first terminal as a transmission precoder determination vector. Next, when determining the transmission precoder determination vector of the second terminal, B _alloc = {a1}, so the transmission precoder determination vector of the second terminal is
The optimization indicator is the rate sum of the first terminal and the second terminal,
The search range is optimized as A \ B _alloc = {a2, a3} (\ is a difference set). That is, at the time of optimization of the second terminal, the first terminal corresponds to the above-described “terminal to which a transmission precoder determination vector has already been assigned in the sequential search”. Further, the rate sum of the first terminal and the second terminal corresponds to the “rate sum of some terminals”.

  The earlier the step of the sequential search, the smaller the number of terminals that can be considered in the optimization index (second index) of the above equation (10).

  For this reason, the optimal transmission precoder determination vector determined by the sequential search may deviate from the optimal solution for the entire system.

  In the present embodiment, in view of the characteristics of the sequential search described above, the early stage of the sequential search, that is, the step that the optimization index can consider only a part of the terminals, that is, the terminals that affect the fewer terminals, that is, the interference Search for small terminals.

  Conversely, the end of the sequential search, that is, the step in which the optimization index can consider more terminals, the search for terminals that are expected to affect the characteristics of more terminals, that is, terminals with higher interference is performed. I do.

  According to the first exemplary embodiment, an optimization index is obtained by searching a transmission precoder determination vector for a terminal with a small amount of interference, for example, at the beginning of a sequential search, and for a terminal with a large amount of interference, for example, at the end. Inconsistency between the number of terminals to be considered and the number of terminals that can be considered is resolved.

  As a result, according to the first exemplary embodiment, in the first half of the search, the influence due to the shortcoming of the sequential search that only the optimization index for some terminals can be used is reduced. For this reason, it is possible to suppress the deterioration of the system characteristics due to the adoption of the sequential search instead of the full search.

に基づいて、次式（１２）により第２のプリコーダ

を算出する（Ｓ１０８）。 <Calculation of second precoder>
Further, the second precoder calculation unit 109 transmits the transmission precoder determination vector acquired from the transmission precoder determination vector selection unit 108.

On the basis of the second precoder by the following equation (12)

Is calculated (S108).

は端末あたりのレイヤ数である。第１の例示的な実施形態では、

である。この場合、式（１２）の第２のプリコーダ

はN_t次元複素ベクトルである。

Is the number of layers per terminal. In the first exemplary embodiment,

It is. In this case, the second precoder of equation (12)

Is an N _t -dimensional complex vector.

を計算する（Ｓ１０９）。送信プリコーダW^Txは、N_t行、（N_layer・N_user）列の複素行列である（N_layer＝１の場合、N_t行×N_user列の複素行列）。 <Calculation of transmission precoder>
Finally, the transmission precoder calculation unit 110 uses the following equation (13) based on the second precoder acquired from the second precoder calculation unit 109 to transmit a transmission precoder (transmission precoding matrix).

Is calculated (S109). The transmission precoder W ^Tx is a complex matrix of N _t rows and (N _layer · N _user ) columns (when N _layer = 1, N _t rows × N _user columns complex matrix).

The communication unit 101, transmission precoder multiplies the transmission symbol of each terminal is transmitted from the N _t antennas to the terminals 20_1~20_N _user.

は、式（１４）で与えられ、N_ｔ次元の複素ベクトルである。 Signal transmitted from communication unit 101

Is an N _t -dimensional complex vector given by equation (14).

はｉ_user番目の端末宛のシンボルである。 here

Is a symbol addressed to the i _user- th terminal.

のｎ_t(ｎ_t=1, ・・・, Ｎ_t)番目の要素は、基地局１０＿１のｎ_t番目のアンテナから送信される信号に対応する。 Signal (N _t- dimensional complex vector)

N _t (n _t = 1,..., N _t ) th element corresponds to a signal transmitted from the n _t th antenna of the base station 10_1.

  According to the first exemplary embodiment, the transmission precoder determination vector is sequentially determined. For this reason, the amount of calculation is reduced as compared with the case where the transmission precoder determination vector is determined by full search.

  Furthermore, according to the first exemplary embodiment, a potential pre-interference defined by the pre-interference power of the first precoder is estimated, and a vector for determining a transmission precoder is determined in order from the terminal having the small presumed interference. The sequential search is performed. According to the first exemplary embodiment having such a configuration, it is possible to suppress deterioration of characteristics due to the adoption of sequential search instead of full search.

<Second Exemplary Embodiment>
The difference between the second exemplary embodiment of the present invention and the first exemplary embodiment is how the first precoder is calculated.

Instead of calculating the first precoder based on the MRT precoder generation method as in the first exemplary embodiment, in the second exemplary embodiment,
-Select a vector that is likely to be selected as a transmission precoder determination vector from the predefined set,
Determine a first precoder based on the selected vector.

  As a result, according to the second exemplary embodiment, it is possible to improve the estimation accuracy of the interference power by the first index. Hereinafter, the configuration and operation of the second exemplary embodiment will be described.

  FIG. 4 is a diagram illustrating the configuration of the second exemplary embodiment. In FIG. 4, the configuration of the base station 10_2 of the second exemplary embodiment is partially different from the base station 10_1 of the first exemplary embodiment of FIG. 4, the same elements as those in FIG. 1 are denoted by the same reference numerals. In the following, regarding the second exemplary embodiment, the description of the same parts as the configuration of the first exemplary embodiment will be omitted, and the structural differences will be described.

  As described above, in the first exemplary embodiment of FIG. 1, the first precoder calculation unit 105 inputs the effective channel vector from the effective channel vector storage unit 103 and performs the first calculation based on the MRT precoder generation method. 1 precoder (complex conjugate of effective channel vector) is calculated.

On the other hand, the first precoder calculation unit 105b of the second exemplary embodiment of FIG.
The effective channel vector is input from the effective channel vector storage unit 103, and the predefined set is input from the predefined set storage unit 104. The first precoder calculation unit 105b calculates a first precoder based on the effective channel vector and the predefined set.

  A specific method for calculating the first precoder in the second exemplary embodiment will be described later. Other configurations are the same as those of the base station 10_1 in FIG.

  FIG. 5 is a flow diagram illustrating an example of operation of the base station of the second exemplary embodiment. The operation of the base station 10_2 in the second exemplary embodiment will be described with reference to FIGS.

  Steps S101 and S104 to S109 in FIG. 5 are the same as those in the first exemplary embodiment in FIG. 3, but steps S102b and S103b in FIG. 5 are the steps in the first exemplary embodiment in FIG. This is different from S102 and S103.

<Calculation of effective channel vector>
As in the first exemplary embodiment, the effective channel vector calculation unit 102 estimates a channel matrix based on the received signal of the uplink reference signal acquired from the communication unit 101, and calculates the effective channel vector based on the estimated channel matrix. The effective channel vector calculation result is stored in the effective channel vector storage unit 103 (S101).

<Calculation of the first precoder>
In step S102 of the first exemplary embodiment of FIG. 3, the MRT precoder is the first precoder, whereas in step S102b of the second exemplary embodiment of FIG. Then, a vector is selected on the basis of a predetermined standard to be a first precoder. Hereinafter, an example of a specific operation will be described.

と、
事前定義セット記憶部１０４から読み出した事前定義セット

と、に基づいて、第１のプリコーダを計算する。 The first precoder calculation unit 105b of the base station 10_2 reads the effective channel vector read from the effective channel vector storage unit 103 (see the above equations (3) and (4)).

When,
Predefined set read from the predefined set storage unit 104

And calculating a first precoder.

の各要素とする。 Although not particularly limited, as a specific calculation example of the first precoder, for example, a predetermined number is selected in descending order of the desired signal, and the first precoder group is selected.

Each element of

を用いた際の所望信号電力

は、次式（１５）で算出される。 In step S102b of FIG. 5, the i _set- th (1 ≦ i _set ≦ N _set ) vector of the predefined set A for the i _user- th (1 ≦ i _user ≦ N _user ) terminal.

Desired signal power when using

Is calculated by the following equation (15).

  By using a vector close to a transmission precoder determination vector used for actual transmission from the predefined set as the first precoder used for estimating the interference power, it is possible to improve the estimation accuracy of the interference power.

  Furthermore, if the estimation accuracy of the interference power increases, the effect of reducing the communication characteristic deterioration due to the successive search (characteristic deterioration caused by using the successive search instead of the full search) increases.

  Therefore, in the second exemplary embodiment, by selecting, as the first precoder, a vector having a high desired signal power that is estimated to be highly likely to be selected as a transmission precoder determination vector, interference The power estimation accuracy is increased.

  The reference for selecting the first precoder may be the interference power of each terminal instead of the above-described desired signal power, or may be SLR (Signal to Leakage Ratio) or SLNR (Signal to Leakage). and Noise Ratio).

<Calculation of the first index>
In step S103 of the first exemplary embodiment of FIG. 3, a first index (interference index) is calculated for a single first precoder. On the other hand, in the second exemplary embodiment, in step S103b of FIG. 5, the interference power index is calculated for a plurality of first precoders. Hereinafter, step S103b of FIG. 5 will be described in more detail.

と、
実効チャネルベクトル記憶部１０３から読み出した実効チャネルベクトル

と、に基づいて、第１の指標を計算する。 The first index calculation unit 106 is a first precoder group acquired from the first precoder calculation unit 105b.

When,
Effective channel vector read from effective channel vector storage unit 103

Based on the above, the first index is calculated.

は少なくとも一つの第１のプリコーダを要素とする集合である。 Here, the first precoder group

Is a set having at least one first precoder as an element.

の各要素

に基づき計算した与干渉電力指標

の算術平均値を、第１の指標

とする（Ｓ１０３ｂ）。与干渉電力指標と第１の指標は、それぞれ次式（１６）、（１７）で算出される。 Specifically, the first index calculation unit 106 uses the first precoder group.

Each element

Interference power index calculated based on

The arithmetic mean of the first index

(S103b). The interference power index and the first index are calculated by the following equations (16) and (17), respectively.

に関して、

は集合

の要素数を表す。 However, any set

With respect to

Is a set

Represents the number of elements of.

の算出方法は、第１のプリコーダ群

の各要素に基づき計算した与干渉電力指標

の算術平均値に制限されるものではなく、例えば中央値であってもよいし、調和平均や重み付け平均であってもよい。 The first indicator

The calculation method of the first precoder group

Interference power index calculated based on each element of

Is not limited to the arithmetic average value, and may be, for example, a median value, a harmonic average, or a weighted average.

はｉ_user番目の端末が第１のプリコーダを用いてデータを送信した際のｉ_user番目以外の端末への与干渉電力の算術平均値、調和平均値、重み付け平均値、中央値のいずれであってもよい。 In the second exemplary embodiment, as in the first exemplary embodiment, the interference power index

There are i _user-th terminal arithmetic average of Interference power to the i _user th other terminal when sending the data using the first precoder, harmonic mean, weighted mean value, in either median May be.

Further, as in the first exemplary embodiment, the first precoder calculation unit 105b and the first index calculation unit 106 repeat steps S102b and 103b in FIG. 5 for all terminal indexes, respectively. Thus, the first index of all terminals (20_1 to 20_N _user ) is calculated (S104).

<Terminal sorting, selection of transmission precoder determination vector, calculation of second precoder, calculation of transmission precoder>
Hereinafter, similarly to the first exemplary embodiment whose operation has been described with reference to FIG. 3, the terminal sort unit 107 is based on the first index calculated by the first index calculation unit 106. Are sorted (S105).

  The transmission precoder determination vector selection unit 108 selects a transmission precoder determination vector for all terminal indexes based on the result of terminal sorting by the terminal sorting unit 107 (S106, S107).

に基づいて、上式（１２）より、第２のプリコーダ

を計算する（Ｓ１０８）。 Second precoder calculation unit 109 transmits transmission precoder determination vector selected by transmission precoder determination vector selection unit 108.

From the above equation (12), the second precoder

Is calculated (S108).

Finally, the transmission precoder calculation unit 110 calculates the transmission precoder W ^Tx of Expression (13) based on the second precoder calculated by the second precoder calculation unit 109 (S109).

  According to the second exemplary embodiment of the present invention, in addition to the effects of the first exemplary embodiment, the precoder used in actual transmission is more than that of the first exemplary embodiment. By using the first precoder that is closer, the estimation accuracy of the interference can be improved. For this reason, according to the second exemplary embodiment, it is possible to more appropriately suppress the deterioration of characteristics due to the adoption of the sequential search instead of the full search.

<Third exemplary embodiment>
The difference between the third exemplary embodiment and the second exemplary embodiment is how the first precoder is calculated. According to the third exemplary embodiment, the first precoder is determined based on the second precoder used when data was transmitted in the past instead of the channel information of the latest time.

  FIG. 6 is a diagram illustrating the configuration of the third exemplary embodiment. Referring to FIG. 6, the base station 10_3 of the third exemplary embodiment is partially different in configuration from the base station 10_2 of the second exemplary embodiment of FIG. In FIG. 6, the same elements as those in FIG. 4 are denoted by the same reference numerals. Hereinafter, regarding the base station 10_3 of the third exemplary embodiment, description of the same elements as those of the base station 10_2 of the second exemplary embodiment will be omitted, and different components will be described.

  The base station 10_3 of the third exemplary embodiment includes a first precoder calculation unit 105c instead of the first precoder calculation unit 105b of FIG. 4 as compared with the second exemplary embodiment. 4, a control unit 111b is provided instead of the control unit 111, and a second precoder storage unit 112, which is not provided in FIG. 4, is added.

  In the second precoder storage unit 112, in the data transmission at the time before the generation of the transmission precoder, for example, the first exemplary embodiment (the first precoder is generated by the MRT precoder generation method), or A second precoder (precoding matrix) calculated in the same manner as in the second exemplary embodiment (generating a first precoder based on a predefined set) is stored.

In the second exemplary embodiment, the first precoder calculation unit 105b includes
-The effective channel vector is input from the effective channel vector storage unit 103,
-Input a predefined set from the predefined set storage unit 104,
A first precoder is calculated based on the effective channel vector and the predefined set.

On the other hand, the first precoder calculation unit 105c of the third exemplary embodiment
-Input the second precoder used for the previous transmission from the second precoder storage unit 112,
The first precoder is determined based on the input second precoder information.

The main difference between the control unit 111b in the third exemplary embodiment and the control unit 111 in the second exemplary embodiment is as follows.
The control unit 111b stores the information of the second precoder calculated by the second precoder calculation unit 109 in the second precoder storage unit 112.

The control unit 111b controls the units 101 to 110 and transmits data to the transmission target terminals (20_1 to 20_N _user ) based on the precoder input by the transmission precoder calculation unit 110.

  Further, the control unit 111 b receives the second precoder calculated by the second precoder calculation unit 109 and stores it in the second precoder storage unit 112.

The second precoder storage unit 112 stores, for each user (for each terminal), the second precoder determined before the time of determining the transmission precoder defined in the third exemplary embodiment. _{Stores reg} .

The N _reg second precoders stored in the second precoder storage unit 112 are all determined according to the first exemplary embodiment or according to the second exemplary embodiment. It may be determined. Alternatively, the first N _reg second precoders are determined by the first exemplary embodiment or the second exemplary embodiment, and the remaining N _reg −1 is the third one. May be determined according to the exemplary embodiment.

  Other configurations are the same as those of the second exemplary embodiment.

  FIG. 7 is a flow diagram illustrating an example of the operation of the third exemplary embodiment. In FIG. 7, the same processing steps as those of the second exemplary embodiment described with reference to FIG. In the following, with reference to FIGS. 6 and 7, the operation of the third exemplary embodiment will be described mainly with respect to the differences from the operation of the second exemplary embodiment of FIG. 5.

  In FIG. 7, steps S101, S103b, and S104 to S109 are the same as those in the second exemplary embodiment of FIG. 5, but steps S110 and S102c are different from those in the second exemplary embodiment. Yes.

<Calculation of effective channel vector>
As in the second exemplary embodiment, the effective channel vector calculation unit 102
-Estimate the channel matrix based on the received signal of the uplink reference signal acquired from the communication unit 101,
Calculate the effective channel vector based on the estimated channel matrix,
The effective channel vector calculation result is stored in the effective channel vector storage unit 103 (S101).

<Calculation of the first precoder>
In step S102 of the second exemplary embodiment described with reference to FIG. 5, the first precoder selects a vector selected from a predefined set of vectors based on a predetermined index, such as desired signal power, for example. It is said.

  In step S102c of the third exemplary embodiment of FIG. 7, the first precoder is determined based on the second precoder used for transmission precoder determination before the previous time.

Specifically, the first precoder calculation unit 105 c reads N _reg second precoders from the second precoder storage unit 112. The N _reg second precoders are used for generating transmission precoders before and after the previous time for each terminal 20 to be transmitted from the communication unit 101.

を生成する（Ｓ１０２ｃ）。 Then, the first precoder calculation unit 105c includes a first precoder group including the read N _reg second precoders as elements.

Is generated (S102c).

<Calculation of first index, sorting of terminals, selection of transmission precoder determination vector, calculation of second precoder, calculation of transmission precoder>

を計算してもよい。 Hereinafter, similarly to the second exemplary embodiment, the first index calculation unit 106 calculates the first index based on the first precoder group generated by the first precoder calculation unit 105c. (S103b). For example, the first index calculation unit 106 uses the expressions (16) and (17) of the second exemplary embodiment to calculate the first index.

May be calculated.

The first precoder calculation unit 105c and the first index calculation unit 106 repeat steps S102c and S103b for all terminal indexes, respectively, so that the first precoders of all the transmission target terminals (20_1 to 20_N _user ) An index is calculated (S104).

  The terminal sort unit 107 sorts the terminals based on the first index acquired from the first index calculation unit 106 (S105). For example, the terminal sort unit 107 performs reassignment of the terminal index in order from a terminal having a small first index to a terminal having a large first index.

Based on the terminal sort result acquired from the terminal sort unit 107, the transmission precoder determination vector selection unit 108 searches the transmission precoder determination vector by a sequential search (the above formula ( 10)), selection is made for all terminals (20_1 to 20_N _user ) (S106, S107).

に基づいて、上式（１２）より、第２のプリコーダ

を計算する（Ｓ１０８）。 Further, the second precoder calculation unit 109 transmits the transmission precoder determination vector acquired from the transmission precoder determination vector selection unit 108.

From the above equation (12), the second precoder

Is calculated (S108).

Finally, the transmission precoder calculation unit 110 calculates the transmission precoder W ^Tx of Expression (13) based on the second precoder acquired from the second precoder calculation unit 109 (S109).

<Storage of second precoder information>
As described above, step S110 is an operation added in the third exemplary embodiment, not in the second exemplary embodiment.

  The controller 111b stores the second precoder (precoding matrix information) calculated by the second precoder calculator 109 in the second precoder storage unit 112 (S110).

Further, the control unit 111b stores the data of the second precoder stored in the second precoder storage unit 112 so that the number of second precoders stored in the second precoder storage unit 112 is N _reg. May be deleted.

Note that a predetermined fixed value may be used as the number N _reg of the second precoders held in the second precoder storage unit 112.

Alternatively, the number N _reg of second precoders stored in the second precoder storage unit 112 is, for example,
-The magnitude of the fluctuation of the propagation environment between the base station 10_3 and the terminal 20,
・ The size of the terminal movement speed,
It may be set variably according to the above.

  According to the third exemplary embodiment of the present invention, similar to the second exemplary embodiment, compared to the first exemplary embodiment, it is closer to a precoder used during actual transmission. By using the first precoder, it is possible to improve the estimation accuracy of the interference. According to the third exemplary embodiment, it is possible to more appropriately suppress the deterioration of characteristics by adopting the sequential search instead of the full search.

<Fourth Exemplary Embodiment>
According to the fourth exemplary embodiment of the present invention, unlike the first exemplary embodiment, the number of layers per terminal can be plural. Therefore, in the first exemplary embodiment, one transmission precoder determination vector is assigned to each terminal, whereas in the fourth exemplary embodiment, transmission is performed for the number of layers of each terminal. A precoder determination vector is allocated to each terminal.

In the operation of the fourth exemplary embodiment, the operation of assigning transmission precoder determination vectors to all terminals in the first exemplary embodiment is repeated for the number of layers (= N _layer ).

  FIG. 8 is a diagram illustrating the configuration of the fourth exemplary embodiment of the present invention. In FIG. 8, the configuration of the base station 10_4 of the fourth exemplary embodiment is the same as the configuration of the base station 10_1 of the first exemplary embodiment of FIG. 8, the same elements as those in FIG. 1 are denoted by the same reference numerals.

FIG. 9 is a flowchart illustrating an operation example of the base station 10_4 of the fourth exemplary embodiment. With reference to FIG. 9 and FIG. 10, the operation of the base station 10_4 of the fourth exemplary embodiment will be described. In the fourth exemplary embodiment, the base station 10_4 maps symbols to multiple layers (N _layer ) data and allocates them to multiple (N _t ) antennas, as shown in FIG. Each terminal 20 generates a plurality of layers (N _layer ) data by post-coding signals received from a plurality (N _r ) of antennas. Note that the number of layers N _layer is equal to or less than the number of antennas (N _layer ≦ min (N _t , N _r )).

As a basic operation,
Determining a first precoder;
-Calculating a first index;
・ Steps to sort terminals,
A series of steps of selecting a transmission precoder determination vector,
One vector is assigned to each terminal.

  Further, in the fourth exemplary embodiment, the above-described series of steps is repeated a plurality of times in order to assign a plurality of transmission precoder determination vectors to each terminal.

  However, when the MRT precoder generation method is used as the first precoder calculation as in the fourth exemplary embodiment, the same first precoder is calculated between the series of steps. Is done.

Thus, in the operation of the fourth exemplary embodiment, as shown in the flowchart in FIG. 9, step S103c for determining the first precoder is executed only once, and a transmission precoder determination vector is selected in step S106. _Is repeated a plurality of times for the number of layers (= N _layer ).

<Calculation of effective channel vector>
In step S101 of the first exemplary embodiment described with reference to FIG. 3, a single effective channel vector is calculated for each terminal 20.

  In contrast, in step S101b of the fourth exemplary embodiment of FIG. 9, a plurality of effective channel vectors corresponding to a plurality of layers are calculated for each terminal 20.

  As a specific operation of step S101b, the effective channel vector calculation unit 102 determines each layer based on the channel matrix of the propagation path between the base station 10_4 and the terminal, which is estimated based on the uplink reference signal received by the communication unit 101. An effective channel vector is generated and stored in the effective channel vector storage unit 103.

  As the generation of the effective channel vector, an execution channel vector obtained by multiplying the channel matrix by the estimated postcoder is calculated and stored in the effective channel vector storage unit 103 (S101b). An example of a specific effective channel vector calculation method will be described below.

と推定ポストコーダ

は、基地局１０＿４からｉ_user番目の端末２０へのチャネル行列

を用いて、それぞれ次式（１８）、（１９）で算出される。

Effective channel vector corresponding to the base station and the i _layer th _layer of the i _user th terminal 20

And estimated postcoder

Is the channel matrix from the base station 10_4 to the i _user th terminal 20

Are calculated by the following equations (18) and (19), respectively.

Here, N _layer is the number of multiple layers for each terminal 20.

は、チャネル行列

のｉ_layer番目の特異値に対応する左特異ベクトルである。 Estimated postcoder of terminal 20

Is the channel matrix

_Is the left singular vector corresponding to the i _layer- th singular value.

は、以下を満たす。

Satisfies the following.

は

のｉ_layer番目の固有値であり、

はその固有ベクトルである。 here,

Is

I _layer- th eigenvalue of

Is its eigenvector.

は

のｉ_layer番目の特異値である。

Is

I _layer- th singular value.

<Calculation of the first precoder>
In step S102 of the first exemplary embodiment described with reference to FIG. 3, a single first precoder generated by the MRT precoder generation method is used per terminal 20. In contrast, in step S102d of the fourth exemplary embodiment of FIG. 9, per terminal 20, each first precoder plurality in MRT precoder generation method in response to a plurality of layers (N _layer) Generate.

を読み出し、端末（２０＿i_user）あたり、複数の第１のプリコーダ

を次式(２０)により算出する（Ｓ１０２ｄ）。 The first precoder calculation unit 105 receives the effective channel vector from the effective channel vector storage unit 103.

And a plurality of first precoders per terminal (20_i _user )

Is calculated by the following equation (20) (S102d).

  However, the calculation of the first precoder according to the equation (20) may use the first precoder determination method of the second exemplary embodiment or the third exemplary embodiment.

<Calculation of the first index>
In step 103 of the first exemplary embodiment described with reference to FIG. 3, a first index is calculated for a single first precoder per terminal 20. On the other hand, in step S103c in the fourth exemplary embodiment of FIG. 9, the first index is calculated for the first precoders for a plurality of layers.

  A specific operation in step S103c will be described.

と、
第１のプリコーダ計算部１０５から取得した第１のプリコーダ

を用いて第１の指標を計算する（Ｓ１０３ｃ）。 The first index calculation unit 106 in FIG. 8 reads the effective channel vector of each layer of all the transmission target terminals read from the effective channel vector storage unit 103.

When,
First precoder acquired from first precoder calculation unit 105

Is used to calculate the first index (S103c).

として、各端末の各レイヤの与干渉電力指標

のレイヤに関する和(ｊ_layer=1からN_layerまでの和)を用いる。 First indicator

As an interference power index of each layer of each terminal

The sum (j _layer = 1 to N _layer ) is used.

と

は、例えば、それぞれ次式（２１）、（２２）で算出される。

When

Are calculated by the following equations (21) and (22), for example.

はｉ_user番目の端末が第１のプリコーダを用いてデータを送信した際のｉ_user番目以外の端末への与干渉電力（式（２２）の右辺第１項）と、ｉ_user番目のｉ_layer番目のレイヤ以外のレイヤへの与干渉電力（式（２２）の右辺第２項）の和である。 Interference power index of each layer of each terminal

And i _user-th terminal is the interfering power to i _user th other terminal when sending data using a first precoder (first term of the right side of the equation (22)), i _user th i _layer It is the sum of the interference power to the layers other than the th layer (the second term on the right side of Expression (22)).

は、ｉ_user番目の端末が第１のプリコーダを用いてデータを送信した際のｉ_user番目以外の端末への与干渉と、ｉ_user番目のｉ_layer番目のレイヤ以外のレイヤへの与干渉電力の算術平均、調和平均、重み付け平均、中央値であってもよい。 Here, the interference power index of each layer of each terminal

Is, i _user-th terminal and interference to the first i _user th other terminal when sending data using precoder, i _user th i _layer th interfering power to other layers Layer May be an arithmetic average, harmonic average, weighted average, or median.

は、与干渉で電力指標

の算術平均、調和平均、重み付け平均、中央値であってもよい。 The first indicator

Is a power indicator with interference

May be an arithmetic average, harmonic average, weighted average, or median.

  Furthermore, the first precoder calculation unit 105 and the first index calculation unit 106 calculate the first index of all terminals by repeating S102d and S103c for all terminal indexes, respectively (S104).

に基づいて、例えば、端末のソートを行う（Ｓ１０５）。第１の指標が小さい端末から大きい順に端末インデックスi'userを割り当てる。 <Sort terminal>
Further, the terminal sorting unit 107 uses the first index acquired from the first index calculation unit 106 as in the first exemplary embodiment.

For example, the terminals are sorted (S105). The terminal index i′user is assigned in ascending order from the terminal with the first index being small.

と、
事前定義セット記憶部１０４から読み出した事前定義セット

と、
端末ソート部１０７から取得したユーザインデックス

と、に基づいて、第２の指標

を最大化するベクトルを、事前定義セットから、逐次的に探索し、プリコーダ決定用ベクトル

をi'_user番目の端末のｉ_layer番目のレイヤに割り当てる（Ｓ１０６）。 <Selection of transmission precoder determination vector>
Further, the transmission precoder determining vector selection unit 108 reads the effective channel vectors of all the transmission target terminals (20_1 to 20_N _user ) read from the effective channel vector storage unit 103.

When,
Predefined set read from the predefined set storage unit 104

When,
User index acquired from the terminal sort unit 107

And a second indicator based on

The vector for maximizing is pre-determined by sequentially searching from the predefined set.

_Is assigned to the i _layer th _layer of the i ' _user th terminal (S106).

  A specific calculation is as the following formula (23).

を逐次探索する際の第２の指標として、各端末の各レイヤのうち、すでに送信プリコーダ決定用ベクトルが割り当てられたレイヤのレート和を用いている。

As the second index when sequentially searching for, the rate sum of the layers to which the transmission precoder determining vector has already been assigned among the layers of each terminal is used.

は全端末の１番目のレイヤからｉ_layer番目のレイヤのレート和（式(23)の第１項ΣΣlog₂[1+SINR]）と、1番目の端末からi'_user番目のｉ_layer番目のレート和（式（23）の第２項Σlog₂[1+SINR]）の和である。 here,

The i _layer th rate sum of the layer from the first layer of all the terminals (the first term ΣΣlog ₂ [1 + SINR] of the formula (23)), from the first terminal i _'user-th i _layer th This is the sum of the rate sums (the second term Σlog ₂ [1 + SINR] in equation (23)).

The transmission precoder determining vector selection unit 108 repeats step S106 with respect to all terminal indexes (loop processing from i ′ _user = 1 to N _user ), whereby the i _layer th _{layer of} all terminals (20_1 to 20_N _user ). The transmission precoder determination vector is calculated (S107).

Further, the transmission precoder determination vector selection unit 108 repeats the operations of steps S106 and S107 for all layer indexes N _layer (loop processing from i _layer = 1 to N _layer ), so that all terminals (20_1 to 20_N). _user ), transmission precoder determination matrices for all layers (N _layer ) are determined (S112).

<Calculation of second precoder>
In step S108 of the first exemplary embodiment described with reference to FIG. 3, since a single layer is transmitted per terminal 20, a single transmission precoder determination vector becomes the second precoder. . On the other hand, in step S108b in the fourth exemplary embodiment of FIG. 9, the second precoder is determined by integrating the precoder determination vectors for a plurality of layers.

に基づいて、例えば次式（２４）により第２のプリコーダ

を算出する（Ｓ１０８ｂ）。 As a specific operation of step S108b, the second precoder calculating unit 109 performs transmission precoder determination vector of i _layer (i _layer = 1 to N _layer ) acquired from the transmission precoder determination vector selection unit 108.

For example, the second precoder by the following equation (24)

Is calculated (S108b).

<Calculation of transmission precoder>
Step S109 of the first exemplary embodiment described with reference to FIG. 3 corresponds to a single layer per terminal 20, whereas the fourth exemplary embodiment of FIG. Step S109b in corresponds to a plurality of layers per terminal 20.

に基づいて、次式（２５）を用いて送信プリコーダW^Txを計算する（Ｓ１０９ｂ）。 As a specific operation, the transmission precoder calculation unit 110 includes N _user (i ′ _user = 1 to N _user ) second precoders calculated by the second precoder calculation unit 109.

Based on the above, the transmission precoder W ^Tx is calculated using the following equation (25) (S109b).

  According to the fourth exemplary embodiment, in addition to the effects of the first exemplary embodiment, it is possible to transmit in a plurality of layers for each terminal. For this reason, the rate sum (throughput) can be further improved as compared with the first exemplary embodiment.

<Fifth Exemplary Embodiment>
The difference between the fifth exemplary embodiment of the present invention and the first exemplary embodiment is the calculation method of the second precoder. According to the fifth exemplary embodiment, instead of selecting one vector from the predefined set, a plurality of vectors are selected and a vector obtained by combining is used as the precoder. This is different from the typical embodiment.

  As a vector synthesis method, a linear synthesis method is used in which a linear sum using a predetermined weight coefficient is calculated for a plurality of vectors, and this is used as a precoder per layer. In this specification, a precoder generated by the linear synthesis method is referred to as a “linear synthesis precoder”.

  FIG. 10 is a diagram illustrating the configuration of the fifth exemplary embodiment. 10, the same elements as those in FIG. 1 are given the same reference numerals. In the following, the difference of the configuration of the base station 10_5 of the fifth exemplary embodiment from the base station 10_1 of FIG. 1 will be mainly described. Referring to FIG. 10, the base station 10_5 of the fifth exemplary embodiment includes a second precoder calculation unit 109b instead of the second precoder calculation unit 109, as compared with the base station 10_1 of FIG. ing.

  The second precoder calculation unit 109 of the first exemplary embodiment of FIG. 1 inputs a single transmission precoder determination vector from the transmission precoder determination vector selection unit 108, and calculates a second precoder. ing.

  On the other hand, the second precoder calculation unit 109b of the fifth exemplary embodiment receives a plurality of transmission precoder determination vectors from the transmission precoder determination vector selection unit 108, and a plurality of transmission precoder determination vectors. A second precoder is calculated using the linear synthesis precoder determined based on.

  FIG. 11 is a flowchart illustrating the operation of the fifth exemplary embodiment. In FIG. 11, steps S101 to S107 and S109 are the same as steps S101 to S107 and S109 of FIG. Step S108c in FIG. 11 is different from step S108 in FIG. 3, and S113 in FIG. 11 is added.

<Calculation of effective channel vector, calculation of first precoder, calculation of first index, sort of terminals>
The operations from the calculation of the effective channel vector to the sorting of the terminals (S101 to S105 in FIG. 11) are the same as those in the first exemplary embodiment described with reference to FIG.

  The effective channel vector calculation unit 102 estimates the channel matrix based on the received uplink reference signal received from the communication unit 101 to calculate the effective channel vector (see, for example, the above equation (3)), and calculates the effective channel vector. The result is stored in the effective channel vector storage unit 103 (S101).

  The first precoder calculation unit 105 calculates the first precoder based on the effective channel vector read from the effective channel vector storage unit 103 by the MRT precoder generation method (see, for example, the above equation (8)) (S102).

  The first index calculation unit 106 calculates a first index (see, for example, the above formula (9)) based on the first precoder acquired from the first precoder calculation unit 105 (S103).

  Furthermore, the first precoder calculation unit 105 and the first index calculation unit 106 repeat steps S102 and S103, respectively, to calculate the first index of all terminals (S104).

  The terminal sort unit 107 sorts the terminals based on the first index based on the first index acquired from the first index calculation unit 106 (S105).

と、
事前定義セット記憶部１０４から読み出した事前定義セット

と、
端末ソート部１０７から取得したユーザインデックス

と、に基づいて、第２の指標を最大化するベクトルを、事前定義セットから、逐次的に探索し、送信プリコーダ決定用ベクトル

を

番目のベクトルとして割り当てる。 <Selection of transmission precoder determination vector>
The transmission precoder deciding vector selection unit 108 is the effective channel vector of all the transmission target terminals (20_1 to 20_N _user ) read from the effective channel vector storage unit 103.

When,
Predefined set read from the predefined set storage unit 104

When,
User index acquired from the terminal sort unit 107

And a vector for maximizing the second index is sequentially searched from the predefined set to determine a transmission precoder determination vector.

The

Assign as the th vector.

は同一ユーザに同一レイヤに割り当てられ、線形合成の対象となる送信プリコーダ決定用ベクトルの数である（Ｓ１０６）。具体的な計算は，次式（２６）の通りである。 However,

Is the number of transmission precoder determination vectors to be assigned to the same user for the same layer and subjected to linear synthesis (S106). A specific calculation is as the following equation (26).

は、少なくとも一つの送信プリコーダ決定用ベクトルの割り当てが行われた端末のレート和である。レート和算出の際の最適化対象以外の端末は、送信プリコーダ用として割り当て済みの送信プリコーダ決定用ベクトルの線形合成プリコーダを用いている。 here,

Is the rate sum of the terminals to which at least one transmission precoder determination vector is assigned. Terminals other than those to be optimized at the time of rate sum calculation use a linear synthesis precoder of a transmission precoder determination vector that has already been allocated for the transmission precoder.

は、括弧内（）の集合の要素に対する線形和で定義され（LC: Linear Combination）、次式（２７）で表される。なお、式（２７）において、

は、

と表記されている。 Also,

Is defined as a linear sum (LC: Linear Combination) with respect to the elements in the parentheses (), and is expressed by the following equation (27). In equation (27),

Is

It is written.

は

に対応する線形合成重み係数である。 here

Is

Is a linear composite weighting factor corresponding to.

はi'_user番目の端末の実効チャネルベクトル

のＭＲＴプリコーダ生成法で生成されたプリコーダの直交分解係数に基づき計算され、次式(２８)の条件を満たす。なお、あるベクトルを、ベクトル空間の直交基底の１次結合として表すことを、ベクトルの直交分解(orthogonal decompose)といい、その係数を直交分解係数という。

Is the effective channel vector of the i ' _user- th terminal

And is calculated based on the orthogonal decomposition coefficient of the precoder generated by the MRT precoder generation method, and satisfies the condition of the following equation (28). Note that expressing a certain vector as a linear combination of orthogonal bases in a vector space is called an orthogonal decomposition of the vector, and the coefficient is called an orthogonal decomposition coefficient.

は、ＭＲＴプリコーダ法で生成されたプリコーダ（ＭＲＴプリコーダ）である。 here,

Is a precoder (MRT precoder) generated by the MRT precoder method.

は事前定義セット

内のベクトルの数である。

Is a predefined set

Is the number of vectors in.

の線形合成重み係数として用いることにより、複数の送信プリコーダ決定用ベクトルを、受信端で受信電力が最大となるように、合成することが可能である。 The orthogonal decomposition coefficient of the MRT precoder (vector) is converted into a plurality of transmission precoder determination vectors.

By using these as the linear combination weighting coefficients, it is possible to combine a plurality of transmission precoder determination vectors so that the reception power becomes maximum at the receiving end.

The reason for the maximum received signal is given below, for example.
First, the MRT precoder is a precoder that can achieve the maximum received power,
The orthogonal decomposition coefficient of the MRT precoder is a complex coefficient for causing the signal to be in phase at the receiving end when a signal is transmitted using each orthogonal base as a precoder.

  Therefore, by combining a plurality of transmission precoder determination vectors using the orthogonal decomposition coefficient of the MRT precoder as a linear weighting coefficient, the signals of the transmission precoder determination vectors are in phase at the receiving end. For this reason, the maximum received power can be realized.

番目の送信プリコーダ決定用ベクトル

を決定する（Ｓ１０７）。 The transmission precoder determination vector selection unit 108 repeats step S106 for the user, so that all terminals

Vector for determining the th transmission precoder

Is determined (S107).

が決定される（Ｓ１１３）。 Further, by repeating steps S106 and S107 from 1 to Nset with respect to the linear composite index (i _LC ), transmission precoder determination vectors corresponding to all linear composite indexes of all terminals (20_1 to 20_N _user ).

Is determined (S113).

<Calculation of second precoder>
In step S108 of the first exemplary embodiment described with reference to FIG. 3, the single transmission precoder determination vector is directly used as the second precoder.

  In contrast, in step S108c of the fifth exemplary embodiment of FIG. 11, a precoder obtained by linearly synthesizing a plurality of transmission precoder determination vectors is set as a second precoder.

を取得し、該複数の送信プリコーダ決定用ベクトルに基づいて、次式（２９）を用いて線形合成し、第２のプリコーダ（線形合成プリコーダ）

を計算する（Ｓ１０８ｃ）。 The second precoder calculation unit 109b includes a plurality of transmission precoder determination vectors calculated by the transmission precoder determination vector selection unit 108.

, And linearly synthesizing using the following equation (29) based on the plurality of transmission precoder determination vectors, and a second precoder (linear synthesis precoder)

Is calculated (S108c).

に基づいて、前記第１の例示的な実施形態と同様に、式（１３）により、送信プリコーダ（ＭＵ−ＭＩＭＯプリコーダ）

を計算する（Ｓ１０９）。 Finally, the transmission precoder calculation unit 110 receives the second precoder acquired from the second precoder calculation unit 109b.

As in the first exemplary embodiment, a transmission precoder (MU-MIMO precoder) is obtained according to equation (13).

Is calculated (S109).

内の複数のベクトルの組み合わせにより、プリコーダが決まる。 In the fifth exemplary embodiment, the second precoder obtained by linearly synthesizing the transmission precoder determination vector is a predefined set.

A precoder is determined by a combination of a plurality of vectors.

  For this reason, according to the fifth exemplary embodiment, the number of precoder patterns that can be realized is larger than in the case where one vector is selected from the predefined set as the precoder. Therefore, a precoder more suitable for the propagation environment can be selected, and the communication quality is improved.

  By applying the operation of the fifth exemplary embodiment to each layer of the fourth exemplary embodiment, it is possible to determine a transmission precoder that uses a linear synthesis precoder in each layer.

  According to the fifth exemplary embodiment, in addition to the effects of the first exemplary embodiment, the second precoder is generated by linearly combining a plurality of transmission precoder determination vectors. . As a result, signal quality can be improved. For this reason, there is an advantage that the rate sum can be improved.

<Sixth exemplary embodiment>
In the sixth exemplary embodiment of the present invention, transmission of multiple layers per terminal is performed as in the fourth exemplary embodiment. The difference between the fourth exemplary embodiment and the sixth exemplary embodiment is the method for determining the first precoder.

  In the fourth exemplary embodiment, the same first precoder determination algorithm is used when allocating transmission precoder determination vectors of a plurality of layers to each terminal. In this embodiment, an algorithm different from that of the first layer is applied when assigning transmission precoder determination vectors for the second and subsequent layers.

  Specifically, when the transmission precoder determination vector of the first layer is used for each terminal, the transmission precoder is set by using the MRT precoder as the first precoder as in the first exemplary embodiment. A decision vector is determined.

  For the second and subsequent layers, when a transmission precoder determination vector of a certain layer is allocated, the first precoder is determined based on the transmission precoder determination vector allocated to the previous layer.

  FIG. 12 is a diagram illustrating the configuration of the sixth exemplary embodiment. Referring to FIG. 12, a base station 10_6 of the sixth exemplary embodiment is partially different from the base station 10_4 of the fourth exemplary embodiment of FIG. 12, the same elements as those in FIG. 8 are denoted by the same reference numerals. In the following description, differences of the configuration of the sixth exemplary embodiment from the fourth exemplary embodiment will be mainly described.

  Compared to the fourth exemplary embodiment, the sixth exemplary embodiment includes a first precoder calculation unit 105d and a transmission precoder determination vector storage unit 113, and a control unit 111. The point which is provided with the control part 111c instead of is different.

  The first precoder calculation unit 105 in the fourth exemplary embodiment calculates the first precoder based on the effective channel vector.

On the other hand, the first precoder calculation unit 105d in the sixth exemplary embodiment is based on the transmission precoder determination vector already assigned to the terminal during the sequential search (repetition regarding the layer i _layer ). Compute the first precoder.

  As a specific configuration, the first precoder calculation unit 105d inputs a transmission precoder determination vector from the transmission precoder determination vector storage unit 113, and calculates a first precoder.

  The control unit 111c in the sixth exemplary embodiment has a function of storing a transmission precoder determination vector in addition to the function of the control unit 111 in the fourth exemplary embodiment.

  As a specific configuration, the control unit 111 c receives a transmission precoder determination vector from the transmission precoder determination vector selection unit 108 and stores it in the transmission precoder determination vector storage unit 113.

  FIG. 13 is a flowchart illustrating an example of the operation of the sixth exemplary embodiment. In FIG. 13, the operation of the base station 10_6 of the sixth exemplary embodiment is partially different from the operation of the base station 10_4 of the fourth exemplary embodiment of FIG. In FIG. 13, the same reference numerals are assigned to steps of the same operation in FIG. In the following, description of the same steps as in the fourth exemplary embodiment will be omitted, and only different steps will be described. Steps S101b, S102d, S104, S105, S106, S107, S108b, and S109b of the base station 10_6 are the same as the steps in FIG. Steps S102e, 103d, S112b, and S115 in FIG. 13 are different from those in FIG.

<Effective channel vector calculation, first precoder calculation, first index calculation, terminal sorting, transmission precoder determination vector selection>
Steps S101b, S102d, S115, S103c, S104, S105, S106, and S107 are the same operations as those in the fourth exemplary embodiment of FIG.

  First, the effective channel vector calculation unit 102b estimates a channel matrix based on the received signal of the uplink reference signal acquired from the communication unit 101, calculates an effective channel vector based on the estimated channel matrix, and calculates the effective channel vector calculation result. It stores in the effective channel vector storage unit 103 (S101b).

Next, for the first layer (i _layer = 1), the first precoder (for example, the first precoder calculation unit 105 uses the effective channel vector acquired from the effective channel vector storage unit 103 by the MRT precoder generation method, for example, (See the above equation (8)) (S115, S102d).

  The first index calculation unit 106 reads the effective channel vectors of all terminals from the effective channel vector calculation unit 102, and uses the first precoder calculated by the first precoder calculation unit 105 to use the first index of each terminal. (See, for example, the above equation (9)) (S103d, S104).

The terminal sort unit 107 sorts the terminals based on the first index acquired from the first index calculation unit 106 (S105). The terminal sort unit 107 assigns terminal indexes i ′ _user (= 1 to N _user ) in order from the terminal with the smallest first index.

The transmission precoder determination vector selection unit 108
A terminal sort result acquired from the terminal sort unit 107;
An effective channel vector read from the effective channel vector storage unit 103;
A predefined set read from the predefined set storage unit 104;
From
A transmission precoder determination vector for the first layer (i _layer = 1) of all terminals is selected (S106, S107).

<Saving Precoder Determination Vector>
Step S114 in FIG. 13 is an operation added from the first exemplary embodiment described with reference to FIG. The control unit 111c stores the transmission precoder determination vector acquired from the transmission precoder determination vector selection unit 108 in the transmission precoder determination vector storage unit 113 (S114).

が決定される。 <Calculation of the first precoder>
In the first exemplary embodiment, the order of terminals once sorted by the terminal sort unit 107 (terminal index i ′ _user ) is maintained as it is, and all transmission precoder determination vectors are maintained.

Is determined.

In contrast, in the sixth exemplary embodiment, once the transmission precoder determination vector is searched for all the terminals (20_1 to 20_N _user ), the terminals are sorted again. That is, after step S106 of FIG. 13 is executed for all terminals (step S107), the transmission precoder determination vector is stored in the storage unit 113, and then the determination of step S115 is performed, and the second and subsequent layers (i _{If layer} ≧ 2), the process branches to step S102e.

Step S102e in FIG. 13 is a process of determining the first precoder that is a precondition for searching for a transmission precoder determination vector for terminals in the second and subsequent layers (i _layer ≧ 2). The process of step S102e will be specifically described below.

を次式（３０）を用いて計算する（Ｓ１０２ｅ）。 Based on the transmission precoder determination vector read from transmission precoder determination vector storage unit 113, first precoder calculation unit 105d determines transmission precoder determination vectors for the second and subsequent layers (i _layer ≧ 2) of each terminal. First precoder when making a decision

Is calculated using the following equation (30) (S102e).

が考慮する範囲にある。 <Calculation of the first index>
The difference between step S103c of the fourth exemplary embodiment described with reference to FIG. 9 and step S103d of the sixth exemplary embodiment in FIG.

Is in the range to consider.

を計算するときに、ｉ_user番目以外の全ての端末２０の全てのレイヤと、ｉ_user番目の端末２０のｉ_layer番目以外の全てのレイヤの与干渉すべてを考慮している。 In step S103c of the fourth exemplary embodiment, first precoders of all layers of all terminals are determined. For this reason, in step S103c, as expressed by the above equation (22), for example, the interference index of the i _layer th layer of the i _user th terminal

_Is calculated, all the interferences of all the layers of all the terminals 20 other than the i _user- th terminal and all the interferences of all the layers other than the i _layer- th of the i _user- th terminal 20 are taken into consideration.

  On the other hand, in step S103d of the sixth exemplary embodiment, the transmission precoder determination vector that has already been assigned to the terminal 20 is used as the first precoder. For this reason, in the step in the middle of the sequential search, the terminals and layers for which the first precoder (transmission weight) is determined are limited to a part.

を計算する。 Specifically, in step S103d of FIG. 13, the first index calculation unit 106 acquires the first precoder calculated by the first precoder calculation unit 105d, and based on the first precoder, The first index using equations (31) and (32)

Calculate

  Regarding the calculation of the first precoder and the method of calculating the first index, in equations (31) and (32), only the transmission precoder determination vector that has already been assigned is set as the first precoder, Layers that are not assigned when the index is calculated are not considered.

  However, as another method, the transmission precoder determination vector is used as the first precoder for the assigned layer, and the first precoder determined in step S102d in FIG. 13 is used for the unassigned layer. The interference given to all other layers by the precoder may be used as the first index.

  Subsequently, the first precoder calculation unit 105d repeats step S102e of FIG. 13 and the first index calculation unit 106 calculates the first index of all terminals by repeating step S103d of FIG. 13 (S104). ).

  When the first index is determined in this manner, the terminal sort unit 107 obtains the first index acquired from the first index calculation unit 106, as in the fourth exemplary embodiment 4 in FIG. Based on the index, the terminals are sorted (S105).

  The transmission precoder determination vector selection unit 108 selects a transmission precoder determination vector according to the terminal sort result by the terminal sort unit 107 (S106, S107).

In this way, transmission precoder determination vectors for the first to i _layer layers of all terminals are assigned.

  Further, the control unit 111c stores the transmission precoder determination vector acquired from the transmission precoder determination vector selection unit 108 in the transmission precoder determination vector storage unit 113 (S114).

The control unit 111c repeats the above-described steps S102e, S103c, S104, S105, S106, S107, and S114, thereby transmitting precoders for all terminals (20_1 to 20_N _user ) and all layers (i _layer = 1 to N _layer ). A decision vector is assigned (S112).

から、例えば上式（２４）により第２のプリコーダ

を算出する（ステップＳ１０８ｂ）。 <Calculation of second precoder, calculation of transmission precoder>
Thus, when the transmission precoder determination vector is allocated to all terminals and all layers, the second precoder calculation unit 109 reads from the transmission precoder determination vector storage unit 113, as in the fourth exemplary embodiment. The second precoder is determined using the read transmission precoder determination vector. That is, the second precoder calculation unit 109 transmits a transmission precoder determination vector.

From, for example, the second precoder by the above equation (24)

Is calculated (step S108b).

を計算する（Ｓ１０９ｂ）。 Based on the second precoder calculated by the second precoder calculation unit 109, the transmission precoder calculation unit 110 uses the transmission precoder using the equation (25) as in the fourth exemplary embodiment.

Is calculated (S109b).

  According to the sixth exemplary embodiment, in addition to the effects of the fourth exemplary embodiment, the first exemplary embodiment is closer to the precoder used in actual transmission than the fourth exemplary embodiment. By using this precoder, the accuracy of interference estimation can be improved. For this reason, it is possible to more appropriately suppress the deterioration of characteristics caused by adopting the sequential search instead of the full search.

<Seventh exemplary embodiment>
In the seventh exemplary embodiment of the present invention, as in the fifth exemplary embodiment, a transmission precoder is determined by performing linear synthesis on a plurality of transmission precoder determination vectors.

  On the other hand, in the sixth exemplary embodiment, the transmission precoder determination vector that has been assigned in a step of the sequential search is used for the calculation of the first precoder in the subsequent steps.

  In the seventh exemplary embodiment, the first precoder determination method in the sixth exemplary embodiment is applied to the fifth exemplary embodiment.

  FIG. 14 is a diagram illustrating the configuration of the seventh exemplary embodiment. Referring to FIG. 14, a base station 10_7 of the seventh exemplary embodiment is partially different from the base station 10_5 of the fifth exemplary embodiment of FIG. 14, the same reference numerals are assigned to the same elements as those in FIG. In the following description, components of the seventh exemplary embodiment that are different from the fifth exemplary embodiment will be mainly described.

  Referring to FIG. 14, the configuration of the seventh exemplary embodiment is different from that of the fifth exemplary embodiment of FIG. 10 in that the first precoder calculation unit 105e and the transmission precoder determination vector storage unit 113 are used. And a point provided with a control unit 111c instead of the control unit 111.

The first precoder calculation unit 105 in the fifth exemplary embodiment described with reference to FIG.
-The effective channel vector is input from the effective channel vector storage unit 103,
Calculate a first precoder based on the MRT precoder generation method.

  On the other hand, the first precoder calculation unit 105d in the seventh exemplary embodiment reads the transmission precoder determination vector from the transmission precoder determination vector storage unit 113, and based on the transmission precoder determination vector, Calculate the precoder.

  The control unit 111 c receives the transmission precoder determination vector from the transmission precoder determination vector selection unit 108 and stores it in the transmission precoder determination vector storage unit 113.

  FIG. 15 is a flowchart illustrating the operation of the base station 10_7 of the seventh exemplary embodiment. In FIG. 15, steps S101, S104, S105, S106, S107, S108c, and S109 are the same as in the fifth exemplary embodiment of FIG. Steps S102d, S102f, S103d, S114, and S116 in FIG. 15 are different from those in FIG. Steps S102d, S103d, S104 to S107, and S114 in FIG. 15 are the same as those in the sixth exemplary embodiment in FIG.

  First, the effective channel vector calculation unit 102 estimates a channel matrix based on the received signal of the uplink reference signal acquired from the communication unit 101, calculates an effective channel vector based on the estimated channel matrix, and stores the calculation result in an effective channel vector storage. The data is stored in the unit 103 (S101).

Next, for the first layer (i _LC = 1), the first precoder calculation unit 105 calculates the first precoder using the effective channel vector acquired from the effective channel vector storage unit 103 (S116, S102d). ).

  The first index calculation unit 106 calculates the first index using the first precoder acquired from the first precoder calculation unit 105 (S103d, S104).

  Further, the terminal sort unit 107 sorts terminals based on the first index acquired from the first index calculation unit 106 (S105).

The transmission precoder determination vector selection unit 108
A terminal sort result acquired from the terminal sort unit 107;
An effective channel vector read from the effective channel vector storage unit 103;
A predefined set read from the predefined set storage unit 104;
The transmission precoder determination vector for the first layer of all terminals is selected (S106, S107).

<Saving Precoder Determination Vector>
Saving the transmission precoder determination vector is a processing step added in the seventh exemplary embodiment of FIG. 15 from the fifth exemplary embodiment of FIG. The control unit 111c stores the transmission precoder determination vector acquired from the transmission precoder determination vector selection unit 108 in the transmission precoder determination vector storage unit 113 (S114).

<Calculation of the first precoder>
In step S102 of the fifth exemplary embodiment of FIG. 11, the first precoder is generated from the effective channel vector by the MRT precoder generation method.

  On the other hand, in step S102f of the seventh exemplary embodiment, a precoder obtained by linearly synthesizing a transmission precoder determination vector that has already been allocated in the sequential search is set as the first precoder.

Based on the transmission precoder determination vector read from the transmission precoder determination vector storage unit 113, the first precoder calculation unit 105e executes the first precoder for the second and subsequent layers with the layer index (i _LC ). Calculation is performed using Expression (33) (S116, S102f).

<Calculation of the first index>
Subsequently, as in the fifth exemplary embodiment, the first index calculation unit 106 uses the following equation based on the first precoder of the equation (33) acquired from the first precoder calculation unit 105e. The first index is calculated using (34) (S103d).

を計算する。 Further, the first precoder calculation unit 105e and the first index calculation unit 106 repeat S102f and S103, respectively, for all terminal indexes (i _user = 1 to N _user ) to thereby determine all terminals (20_1 to 20_N _user). ) 1st indicator

Calculate

  When the first index is calculated, the terminal sorting unit 107 sorts the terminals (S105), as in the fifth exemplary embodiment.

The transmission precoder determination vector selection unit 108 selects a transmission precoder determination vector according to the terminal sort result by the terminal sort 107 (S106, S107). Transmitting precoder determination vector layers from the first all terminals i until _LC-th are assigned.

  Further, the control unit 111c stores the transmission precoder determination vector selection result in the transmission precoder determination vector storage unit 113 (S114).

を割り当てる。 The control unit 111c repeats the above-described S102f, S103d, S104, S105, S106, S107, and S114 to thereby determine transmission precoder determination vectors for all terminals (20_1 to 20_N _user ) and all layers (N _LC ).

Assign.

<Calculation of second precoder, calculation of transmission precoder>
When the transmission precoder determination vector is assigned to all terminals (N _user ) and all layers (N _LC ), the second precoder calculation unit 109b, similar to the fifth exemplary embodiment, transmits the transmission precoder. The second precoder is determined using the transmission precoder determination vector read from the determination vector storage unit 113.

に基づいて、前記第５の例示的な実施形態と同様、例えば、式（２９）を用いて線形合成し、第２のプリコーダ（線形合成プリコーダ）

を計算する（ステップＳ１０８ｃ）。 Second precoder calculation unit 109b transmits transmission precoder determination vectors for all layers read from transmission precoder determination vector storage unit 113.

As in the fifth exemplary embodiment, the second precoder (linear synthesis precoder) is linearly synthesized using, for example, Expression (29).

Is calculated (step S108c).

に基づいて、前記第５の例示的な実施形態と同様、上式（１３）により、送信プリコーダ

を計算する（Ｓ１０９）。 Second precoder acquired by transmission precoder calculation unit 110 from second precoder calculation unit 109

As in the fifth exemplary embodiment, the transmission precoder is expressed by the above equation (13).

Is calculated (S109).

  According to the seventh exemplary embodiment, in addition to the effects of the fifth exemplary embodiment, the interference estimation accuracy is improved by using the first precoder closer to the precoder used during actual transmission. Can do. For this reason, it is possible to further suppress characteristic deterioration caused by adopting sequential search instead of full search.

  The above-described exemplary embodiments and specific examples are appended as follows (however, they are not limited to the following).

(Appendix 1)
A method for determining a transmission precoder by a base station that precodes a transmission signal and transmits it to a plurality of terminals from a plurality of antennas,
A first step of determining at least one first precoder for each of the plurality of terminals in a predetermined manner;
A second step of calculating a first index considering the interference power based on the first precoder;
A third step of sorting the plurality of terminals based on the first index;
Based on a second index in consideration of inter-terminal interference at the terminal, a vector for transmission precoder determination is sequentially selected for the terminal from a predefined set that is a predefined vector set. The fourth step;
A fifth step of determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A transmission precoder determination method comprising:

(Appendix 2)
The transmission precoder determination vector is
The transmission precoder determination method according to supplementary note 1, wherein each time the series of steps of the first to fourth steps are executed once, one is determined for each terminal.

(Appendix 3)
The transmission precoder determination method according to appendix 1, wherein the predetermined method in the first step is the same between a series of steps of the first to fourth steps performed at least once. .

(Appendix 4)
Including a series of steps of at least two times, wherein the predetermined method in the first step is different between the previous step and the current step between a series of steps of the first to fourth steps performed a plurality of times. The transmission precoder determination method according to appendix 1 or 2, characterized in that:

(Appendix 5)
In the first step,
The predetermined method is:
(a) determining the first precoder based on an MRT (Maximum Ratio Transmission) precoder generation method;
(b) determining the first precoder based on at least one vector of the predefined set;
The transmission precoder determination method according to supplementary note 3 or 4, characterized in that any one of the above is included.

(Appendix 6)
In the first step,
The predetermined method is:
(c) The transmission precoder determining method according to appendix 3 or 4, comprising determining the first precoder based on at least one second precoder used for transmission before the previous time.

(Appendix 7)
In the first step,
The predetermined method is:
(d) When performing the series of steps, for determining at least one of the transmission precoders determined in the series of steps of the first to fourth steps before the previous time in a single transmission precoder decision. The transmission precoder determining method according to appendix 3 or 4, comprising determining the first precoder based on a vector.

(Appendix 8)
In the first step, all the terminals to be transmitted and the first precoder corresponding to a plurality of layers to which the symbols to be transmitted are mapped are obtained,
In the second step, the first index considering the interference power of each layer of each terminal is calculated,
The transmission precoder determination method according to supplementary note 1, wherein the process of assigning the transmission precoder determination vector to all transmission target terminals in the fourth step is repeated for the number of layers.

(Appendix 9)
The transmission precoder according to any one of appendices 1 to 7, wherein the second precoder is obtained by linearly combining a plurality of transmission precoder determination vectors selected from the vectors of the predefined set. Decision method.

(Appendix 10)
In the fourth step, a process of searching for a vector that maximizes the rate sum from the vectors of the predefined set, using the rate sum of some of the terminals to be transmitted as the second index. The transmission precoder determination method according to any one of appendices 1 to 7, wherein the transmission precoder repeats sequentially until all terminals to be transmitted.

(Appendix 11)
A base station apparatus that pre-codes a transmission signal and transmits it from a plurality of antennas to a plurality of terminals,
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
A second precoder calculation unit for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A base station apparatus comprising:

(Appendix 12)
The transmission precoder determination vector is
Each time a series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit are executed, one is determined for each terminal. The base station apparatus according to appendix 11, wherein

(Appendix 13)
During the series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit that are performed at least once or more, the first precoder calculation unit The base station apparatus according to Supplementary Note 11 or 12, wherein the predetermined method is the same.

(Appendix 14)
During a series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit, which are performed a plurality of times, the first and the previous 13. The base station apparatus according to appendix 11 or 12, characterized in that the predetermined method in the precoder calculation unit includes at least two series of processes different from each other.

(Appendix 15)
In the first precoder calculation unit,
The predetermined method is:
(a) determining the first precoder based on an MRT (Maximum Ratio Transmission) precoder generation method;
(b) determining the first precoder based on at least one vector of the predefined set;
The base station apparatus according to appendix 13 or 14, characterized in that any one of the following is included.

(Appendix 16)
In the first precoder calculation unit,
The predetermined method is:
15. The base station apparatus according to appendix 13 or 14, comprising determining the first precoder based on at least one second precoder used for transmission before the previous time.

(Appendix 17)
In the first precoder calculation unit,
The predetermined method is:
In performing the series of operations, the first precoder is determined based on at least one transmission precoder determination vector determined in the series of operations before the previous time in a single transmission precoder determination. 15. The base station apparatus according to supplementary note 13 or 14, characterized by comprising:

(Appendix 18)
The first precoder calculation unit obtains the first precoder corresponding to a plurality of layers to which all terminals to be transmitted and symbols to be transmitted are mapped,
The first index calculation unit calculates the first index considering the interference power of each layer of each terminal,
The base station apparatus according to supplementary note 11, wherein the second precoder calculation unit repeats the process of assigning the transmission precoder determination vector to all transmission target terminals by the number of layers.

(Appendix 19)
The second precoder calculating unit obtains the second precoder by linearly synthesizing a plurality of transmission precoder determination vectors selected from the vectors of the predefined set. The base station apparatus according to any one of 17.

(Appendix 20)
In the second precoder calculation unit, a rate sum of a part of all terminals to be transmitted is set as the second index, and a vector that maximizes the rate sum is searched from the vector of the predefined set. The base station apparatus according to any one of appendices 11 to 17, wherein the processing to be performed is sequentially repeated up to all terminals to be transmitted.

(Appendix 21)
Multiple devices,
A base station that pre-codes a transmission signal and transmits it from a plurality of antennas to a plurality of terminals to be transmitted;
With
The base station
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
A second precoder calculation unit for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A wireless communication system comprising:

(Appendix 22)
The transmission precoder determination vector is
Each time a series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit are executed, one is determined for each terminal. The wireless communication system according to appendix 21, wherein

(Appendix 23)
During the series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit that are performed at least once or more, the first precoder calculation unit The wireless communication system according to appendix 21 or 22, wherein the predetermined method is the same.

(Appendix 24)
During a series of operations by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit, which are performed a plurality of times, the first and the previous 23. The wireless communication system according to appendix 21 or 22, wherein the predetermined method in the precoder calculating unit includes at least two series of operations different from each other.

(Appendix 25)
In the first precoder calculation unit,
The predetermined method is:
(a) determining the first precoder based on an MRT (Maximum Ratio Transmission) precoder generation method;
(b) determining the first precoder based on at least one vector of the predefined set;
The wireless communication system according to appendix 23 or 24, including any of the following.

(Appendix 26)
In the first precoder calculation unit,
The predetermined method is:
25. The wireless communication system according to appendix 23 or 24, comprising determining the first precoder based on at least one of the second precoders used for previous transmissions.

(Appendix 27)
In the first precoder calculation unit,
The predetermined method is:
In performing the series of operations, the first precoder is determined based on at least one transmission precoder determination vector determined in the series of operations before the previous time in a single transmission precoder determination. The wireless communication system according to supplementary note 23 or 24, further comprising:

(Appendix 28)
The first precoder calculation unit obtains the first precoder corresponding to a plurality of layers to which all terminals to be transmitted and symbols to be transmitted are mapped,
The first index calculation unit calculates the first index considering the interference power of each layer of each terminal,
The wireless communication system according to appendix 21, wherein the second precoder calculation unit repeats the process of assigning the transmission precoder determination vector to all the transmission target terminals by the number of layers.

(Appendix 29)
Supplementary notes 21 to 21 wherein the second precoder calculation unit obtains the second precoder by linearly synthesizing a plurality of transmission precoder determination vectors selected from the vectors of the predefined set. 27. The wireless communication system according to any one of 27.

(Appendix 30)
In the second precoder calculation unit, a rate sum of a part of all terminals to be transmitted is set as the second index, and a vector that maximizes the rate sum is searched from the vector of the predefined set. The wireless communication system according to any one of appendices 21 to 27, wherein the processing to be performed is sequentially repeated up to all terminals to be transmitted.

(Appendix 31)
To the base station computer that pre-codes the transmission signal and transmits it from multiple antennas to multiple terminals,
A first precoder calculation process for determining at least one first precoder for the terminal by a predetermined method;
A first index calculation process for calculating a first index considering interference power based on the first precoder;
A terminal sorting process for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. Selection process,
A second precoder calculation process for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A program that executes

(Appendix 32)
The transmission precoder determination vector is
Each time a series of processes by the first precoder calculation process, the first index calculation process, the terminal sort process, and the transmission precoder determination vector selection process is executed once, one is determined for each terminal. The program according to supplementary note 31, wherein

(Appendix 33)
In the first precoder calculation process, a series of processes by the first precoder calculation process, the first index calculation process, the terminal sort process, and the transmission precoder determination vector selection process performed at least once. The program according to appendix 31 or 32, wherein the predetermined method is the same.

(Appendix 34)
The first precoder calculation process, the first index calculation process, the terminal sort process, and the transmission precoder determination vector selection process that are performed a plurality of times, between the previous process and the current process. The program according to appendix 31 or 32, wherein the predetermined method in the precoder calculation process includes at least two series of processes different from each other.

  Each disclosure of Non-Patent Documents 1-4 above is incorporated herein by reference. Within the scope of the entire disclosure (including claims) of the present invention, the embodiment or the exemplary embodiment can be changed or adjusted based on the basic technical concept. In addition, various combinations or selections of various disclosed elements (including each element of each supplementary note, each element of each exemplary embodiment, each element of each drawing, etc.) are possible within the scope of the claims of the present invention. It is. That is, the present invention of course includes various variations and modifications that could be made by those skilled in the art according to the entire disclosure including the claims and the technical idea.

10_1 to 10_7 base stations 20_1 to 20_N _user terminal 101 communication unit 102 effective channel vector calculation unit 103 effective channel vector storage unit 104 predefined set storage units 105, 105b, 105c, 105d, 105e first precoder calculation unit 106 first Index calculation unit 107 Terminal sorting unit 108 Transmission precoder determination vector selection unit 109, 109b Second precoder calculation unit 110 Transmission precoder calculation unit 111, 111b, 111c Control unit 112 Second precoder storage unit 113 Transmission precoder determination vector storage Part

Claims (10)

A method for determining a transmission precoder by a base station that precodes a transmission signal and transmits it to a plurality of terminals from a plurality of antennas,
A first step of determining at least one first precoder for each of the plurality of terminals in a predetermined manner;
A second step of calculating a first index considering the interference power based on the first precoder;
A third step of sorting the plurality of terminals based on the first index;
Based on a second index in consideration of inter-terminal interference at the terminal, a vector for transmission precoder determination is sequentially selected for the terminal from a predefined set that is a predefined vector set. The fourth step;
A fifth step of determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A transmission precoder determination method comprising:   2. The transmission precoder determination vector is determined for each of the terminals one by one every time a series of steps of the first to fourth steps is executed once. Transmission precoder determination method. In the first step,
The predetermined method is:
(a) determining the first precoder based on an MRT (Maximum Ratio Transmission) precoder generation method;
(b) determining the first precoder based on at least one vector of the predefined set;
(c) determining the first precoder based on at least one second precoder used for previous transmissions;
(d) When performing the series of steps of the first to fourth steps, based on at least one transmission precoder determination vector determined in the series of steps of the first to fourth steps before the previous time. Determining the first precoder,
The transmission precoder determination method according to claim 1, comprising at least one of the following. The predetermined method in the first step is the same between a series of steps of the first to fourth steps performed at least once, or
The transmission according to any one of claims 1 to 3, wherein the predetermined method in the first step in the first step includes at least two series of steps different from each other. Precoder determination method. In the fourth step,
The rate sum of some terminals among all the terminals to be transmitted is set as the second index,
The selection of the transmission precoder determination vector includes sequentially repeating a process of searching for a vector that maximizes the rate sum from the vectors of the predefined set up to all terminals to be transmitted. Item 5. The transmission precoder determination method according to any one of Items 1 to 4. A base station apparatus that pre-codes a transmission signal and transmits it from a plurality of antennas to a plurality of terminals,
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
A second precoder calculation unit for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A base station apparatus comprising: The transmission precoder determination vector is
Each time a series of processing by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit is executed, one is determined for each terminal. The base station apparatus according to claim 6, wherein: The first precoder calculator is
Estimating a channel matrix of a propagation path between the base station and each terminal from an uplink reference signal, calculating an effective channel vector based on the channel matrix,
(a) taking the complex conjugate transpose of the effective channel vector to be the first precoder;
(b) determining the first precoder from at least one vector of the predefined set based on a computation result of the effective channel vector and at least one vector of the predefined set;
(c) determining the first precoder based on at least one of the second precoders used for previous transmissions;
(d) In a series of processes by the first precoder calculation unit, the first index calculation unit, the terminal sort unit, and the transmission precoder determination vector selection unit, at least one of the processes determined in the previous process Determining the first precoder based on at least one of transmission precoder determination vectors;
The base station apparatus according to claim 6 or 7, wherein at least one of the following is performed. Multiple devices,
A base station that pre-codes a transmission signal and transmits it from a plurality of antennas to a plurality of terminals to be transmitted;
With
The base station
A first precoder calculator that determines at least one first precoder for the terminal in a predetermined manner;
A first index calculation unit that calculates a first index considering interference power based on the first precoder;
A terminal sort unit for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. A selection section;
A second precoder calculation unit for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A wireless communication system comprising: To the base station computer that pre-codes the transmission signal and transmits it from multiple antennas to multiple terminals,
A first precoder calculation process for determining at least one first precoder for the terminal by a predetermined method;
A first index calculation process for calculating a first index considering interference power based on the first precoder;
A terminal sorting process for sorting the plurality of terminals based on the first index;
A transmission precoder determination vector for sequentially selecting a transmission precoder determination vector for the terminal from a predefined vector set based on a second indicator considering inter-terminal interference at the terminal. Selection process,
A second precoder calculation process for determining a second precoder to be used for transmission based on at least one transmission precoder determination vector;
A program that executes

Images