JP4698346B2 - Wireless transmission system, base station, and wireless transmission method - Google Patents

Description

  The present invention relates to a radio transmission system, a base station, and a radio transmission method used in next-generation radio communication (advanced radio technology) represented by MIMO (Multiple-Input Multiple-Output).

  In recent years, a technique called MIMO (Multiple-Input Multiple-Output) has attracted attention as a communication method for improving effective speed in wireless network technology. In this MIMO technology, a plurality of antennas used for transmission and reception of wireless communication are spatially multiplexed for both transmission and reception, thereby increasing the substantial communication speed without increasing the wireless frequency bandwidth.

  Currently, as a transmission method in MIMO systems, the SDM (Space Division Multiplexing) method that transmits multiple substreams with equal power without requiring channel information on the transmission side, transmission that differs for each substream based on channel information on the transmission side There are eigenbeam-SDM (Eigenbeam-SDM) systems that form beams, and application to various wireless transmission systems (cellular mobile communication systems, IEEE802.11 wireless LAN systems, etc.) is being studied (non-patented) Reference 1).

  An example will be described in detail. For example, in the downlink in the eigenmode transmission scheme, if the number of spatial multiplexing (the number of data streams) exceeds the degree of freedom of array of the receiving terminal antenna, the co-channel interference cannot be removed on the receiving terminal side. The antenna weight control for reducing or eliminating the interference in advance at the base station (transmission side) is required.

  Conventionally, as such an antenna weight control method, a method of obtaining a high transmission diversity gain while eliminating co-channel interference by null-forming only channel response vectors of the allocated spatial multiplexing number of interfering terminals has been proposed. .

A method for determining the antenna weight in this antenna weight control will be described below.
(1) Allocate spatial multiplexing number to each terminal (2) Null formation for channel response vector of allocated spatial multiplexing number of other terminals (3) Form eigen beam using remaining transmission array degrees of freedom Such antenna weight According to this determination method, since the channel response vectors of all assigned spatial multiplexing numbers can be orthogonalized, it is possible to remove all the same channel interference after arbitrary receiving antenna weight control on the receiving side. Further, since the interference terminal does not form a null for a partial space that is not used in spatial multiplexing, the transmission array flexibility can be effectively used to the maximum, and a high diversity gain can be obtained.

This transmission antenna weight determination method is a method called a so-called Zero Forcing norm, specifically, in the system shown in FIG. 4, in the base station, so that the interference component in the received signal at the mobile station becomes 0, An antenna weight V _i for eigenmode transmission and a transformation matrix W _i that is an element of the transmission antenna weight are superimposed on the transmission signal.

ここで、基地局から移動局MS#iに対するチャネル行列をH_i、受信機雑音をn_i(t)、送信シンボルをs_i(t)とする。 On the other hand, on the mobile station side, a plurality of streams are separated and detected from the received signal by the receiving antenna weight U _i ^H. At this time, the received symbol r _i (t) in the mobile station MS # i is expressed by the following equation. Here, it is assumed that the channel information (channel matrix H _i ) is completely known on both the transmission and reception sides.

Here, the channel matrix from the base station to the mobile station MS # i is H _i , the receiver noise is n _i (t), and the transmission symbol is s _i (t).

とおき、上記式(1)における送信アンテナウェイトW_iを、

を満たすように定める。このとき、W_iはR_uu ⁽ⁱ⁾の零固有値に対応する固有ベクトルから構成される行列となる。 The antenna weight determination method based on the Zero Forcing criterion, the condition of W _i by other user interference components in the received signal at each mobile station becomes 0, the spatial correlation matrix for the interference component

And the transmission antenna weight W _i in the above equation (1),

It is determined to satisfy. At this time, W _i is a matrix composed of eigenvectors corresponding to zero eigenvalues of R _uu ⁽ⁱ⁾ .

By using the transformation matrix W _i is an element of such a transmission antenna weight, a signal component for the other party in each of the mobile station can be zero, it is possible to avoid the interference to other users caused Yes (see Non-Patent Document 2).

By the way, the eigenmode transmission scheme requires channel information on the transmission side. When the above eigenmode transmission method is applied in an environment where there is a propagation path fluctuation, if the transmission path fluctuation occurs, the frequency division duplex (FDD) scheme affects the effects of feedback delay, control delay, etc. In the duplex system (TDD), the channel state changes at the time of transmission antenna weight determination and at the time of data reception due to the difference in transmission timing between the uplink and the downlink, and the characteristics deteriorate. Therefore, conventionally, methods for reducing this characteristic deterioration have been studied (see Non-Patent Document 3).
Takeo Ohgane, Toshihiko Nishimura, Yasutaka Ogawa, “Space Division Multiplexing and its Characteristics in MIMO Channels,” IEICE (B), vol.J87-B, no.9, Sept.2004. (And references) Yosuke Fujino, Daisei Uchida, Osamu Kagami, Masahiro Umehira, “Effects of Adaptive Spatial Multiplex Allocation in Multiuser MIMO,” May 2005, IEICE General Conference, B-5-81, p.530 Tsutsumi Takahiko, Nishimura Toshihiko, Ogane Takeo, Ogawa Yasutaka, “A Study on the Effect of Channel Information Error in Various Space Division Multiplexing Systems,” IEICE (B), vol.J87-B, no.9, pp.1496-1504 , Sept. 2004.

  However, in the interference cancellation based on the above-mentioned Zero Forcing norm, receiver noise is not taken into consideration, and therefore, when the interference wave direction is approaching, the received SNR accompanying the gain reduction in the desired wave direction. Degradation occurs.

  Therefore, the present invention has been made in view of the above points, and in next-generation wireless communication in which a plurality of mobile stations communicate at the same time at the same time using a plurality of antennas such as MIMO technology, It is an object of the present invention to provide a wireless transmission system, a base station, and a wireless transmission method capable of improving the effective speed in a wireless network by applying a whitening filter to suppress the intra-cell interference in the downlink.

  In order to solve the above problems, the present invention provides a frequency division duplex between a base station having a plurality of antennas and a mobile station having a plurality of antennas at the same carrier frequency from the plurality of antennas at the same time. When transmitting / receiving data to / from each other by transmission or time division duplexing, the base station acquires channel information in the downlink from the mobile station, and based on the acquired channel information, transmission signals and noise to other mobile stations Generates a transmission antenna weight whose element is a conversion matrix for converting the interference noise component generated by the mobile station side into a signal equivalent to spatiotemporally uncorrelated Gaussian noise, and superimposes it on the transmission signal And transmit to the mobile station through the downlink. Then, in the mobile station, a desired signal in the downlink is separated and detected from the received signal in which the interference noise component is converted into a signal equivalent to Gaussian noise, using a receiving antenna weight generated based on the channel information.

  According to the present invention, the interference noise component on the mobile station side generated by the transmission signal and noise for the other mobile station is converted into a space-time space as a transformation matrix that is an element of the transmission antenna weight for other-user interference suppression. A so-called whitening filter, which is a conversion matrix that converts the signal into a signal equivalent to an uncorrelated Gaussian noise, is used. This whitening filter converts the interference wave signal into a signal equivalent to uncorrelated Gaussian noise rather than completely reducing the interference component to zero, so that mutual interference between mobile stations can be achieved without degrading the received SNR. Can be suppressed.

ここで、基地局から移動局MS#iに対するチャネル情報を表す行列（チャネル行列）をH_i、受信機雑音をn_i(t)、ガウス雑音と等価な信号に変換するための変換行列をW_i、送信シンボルをs_i(t)とする。また、上記合成アンテナウェイトW_iV_iは、そのノルムが常に1になるとは限らないことから、送信電力を規定するために、基地局側で、スケーリング係数α_iを各移動局に対し重畳する。 The whitening filter will be specifically described. In the system shown in FIG. 1, when a plurality of streams are separated and detected from the received signal by the receiving antenna weight U _i ^H on the mobile station side, the received symbol r in the mobile station MS # i is detected. _i (t) is expressed by the following equation.

Here, the matrix (channel matrix) representing the channel information from the base station to the mobile station MS # i is H _i , the receiver noise is n _i (t), and the transformation matrix for transforming the signal equivalent to Gaussian noise is W _{Let i be} the transmitted symbol, s _i (t). Further, since the norm of the combined antenna weight W _i V _i is not always 1, the base station side superimposes the scaling factor α _i on each mobile station in order to define the transmission power. .

ここで，P_nは受信機雑音電力を表す。このとき，受信シンボルr_i(t)は次式(5)及び(6)で表され、他ユーザ干渉のない場合と等価に扱うことができる。

ここで，L_iはR_uu ⁽ⁱ⁾の固有値を対角要素とする対角行列であり、Q_iはR_uu ⁽ⁱ⁾の固有値展開の結果得られる固有ベクトルから構成される固有ベクトル行列を表す。 The whitening filter in the present invention corresponds to W _i which satisfies the following equation (3) and (4) instead of the equation (2).

Here, P _n represents the receiver noise power. At this time, the received symbol r _i (t) is expressed by the following equations (5) and (6), and can be handled equivalently to the case where there is no interference from other users.

Here, L _i is a diagonal matrix whose diagonal elements are eigenvalues of R _uu ⁽ⁱ⁾ , and Q _i represents an eigenvector matrix composed of eigenvectors obtained as a result of eigenvalue expansion of R _uu ⁽ⁱ⁾ .

In the present invention, the transmit antenna weight includes an equivalent channel matrix (A _i = H _i W _i ) and a Hermitian transpose matrix (A _i ^{H of the} equivalent channel matrix). ) And an eigenvector matrix V _i obtained by eigenvalue expansion. That is, from the above formulas (5) to (6), the transmission / reception antenna weights V _i and U _i are eigenvector matrices composed of eigenvectors obtained as a result of eigenvalue expansion of A _i ^H A _i and A _i A _i ^H , respectively. By doing so, it is possible to perform eigenmode transmission in the downlink when other-user interference in the same cell exists.

が成立する。特に、式(7)におけるS_i （= L_i ^1/2）の対角要素は特異値と呼ばれ、また、式(7)のような展開は特異値分解と呼ばれている（例えば、「唐沢好男,“MIMO伝搬チャネルモデリング,”信学論(B), vol.J86-B, no.9, pp.1706-1720, Sept. 2003」参照）。 In this case, the matrix A _i, the matrix ^{_{A i H A i, A i}} A i diagonal matrix L _i for the diagonal elements eigenvalues of ^H, A _i A _i ^H matrix composed of eigenvectors (unit vector) of In general, between U _i and a matrix V _i composed of eigenvectors (unit vectors) of A _i ^H A _i ,

Is established. In particular, the diagonal elements of S _i (= L _i ^1/2 ) in equation (7) are called singular values, and the expansion like equation (7) is called singular value decomposition (for example, “See Yoshio Karasawa,“ MIMO Propagation Channel Modeling, ”Science Review (B), vol. J86-B, no. 9, pp. 1706-1720, Sept. 2003).

そして、基地局が有する複数のアンテナは、前記固有ベクトル行列V_iを用いて、前記固有ベクトル行列に含まれる固有ベクトル毎にビームフォーミングを行うことにより、サブストリーム毎に固有の伝搬路を形成し、複数のサブストリーム毎に個々の変調処理及び送信電力を制御するようにしてもよい。 As a result, the transmission antenna weight has a singular value of an equivalent channel matrix A _i (= H _i W _i ) having a product of the channel information H _i and the transformation matrix W _i as shown in the following equation. The eigenvector matrix V _i obtained by the decomposition is included.

A plurality of antennas which the base station has, using the eigenvector matrix V _i, by performing beamforming for each eigenvector contained in the eigenvector matrix, form a unique channel for each sub-stream, multiple Individual modulation processing and transmission power may be controlled for each substream.

In the present invention, channel information _Hi is known at least on the transmission side (base station). As a method for obtaining channel information H _i of the mobile station by the base station, e.g., sending and receiving using different carrier frequencies in the uplink and downlink simultaneously FDD (Frequency Division Duplex: Frequency Division Duplex), the mobile station In TDD (Time Division Duplex), which divides the communication path on the time axis and switches between transmission and reception and performs simultaneous transmission / reception by feedback from, the channel estimation result on the uplink is used. .

In the above invention, similar results can be obtained by using the reception antenna weights of the above-mentioned Zero Forcing norm or MMSE norm as the receive antenna weight. For example, instead of the receiving antenna weight U _i ^H, the transformation matrix W _i and eigenvector matrix V synthesis including _i antenna weights W _i V _i and, extended channel matrix H is the product of the channel information H _i (synthetic matrix) _i W _i to V _i, acquired by using the dedicated pilot symbols at the reception side, the matrix and obtained by use of this extended channel matrix H _i W _i V _i, based on the Zero Forcing criterion, the extension channel matrix H _i W _i V _i pseudo inverse matrix of the _{(H i W i V i)} -1 and the like can be used as a reception antenna weight. (Note that if a square matrix X does not have zero eigenvalues, an inverse matrix exists, and the inverse matrix of X and the pseudo-inverse of X are the same. Therefore, H _i W _i V _i does not have zero eigenvalues. (If it is a square matrix, the inverse matrix of H _i W _i V _i can be used as the receiving antenna weight.)
With such a receiving antenna weights, due to the mismatch of the actual channel information and the channel information H _i on the transmission side than the case of using as the A _i A _i ^H eigenvector matrix U _i ^H, caused by channel variation Characteristic deterioration can be reduced.

  As described above, according to the present invention, even when a plurality of mobile station directions are close to each other, the whitening filter described above suppresses an interference signal to an adjacent mobile station and Transmission characteristics can be improved.

(Outline of wireless transmission method)
Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is an explanatory diagram showing an overview of the wireless transmission method according to the present embodiment.

  As shown in the figure, in this embodiment, a signal is transmitted between a base station BS having Nt antennas and a mobile station MS # i (# 1, # 2) having Nr antennas. A MIMO wireless transmission system that transmits and receives from the antenna at the same time and the same carrier frequency is used. Here, for simplicity of explanation, the case where the number of mobile stations is 2 will be described as an example. However, the present invention is not limited to this embodiment, and the number of mobile stations is 3 or more. Can also be applied.

In the figure, channel information from the base station BS to the mobile station #i is H _i and transmission symbols are s _i (t). The base station BS side operates as an eigenmode transmission antenna weight (eigenvector matrix) V _i and a spatial filter for suppressing other-user interference for the transmission symbols s _i (t) to each mobile station MS # i. A transformation matrix (antenna weight) _Wi is superimposed on each other.

In the present embodiment, the transformation matrix _Wi is an antenna weight (whitening filter) for converting an interference noise component on the mobile station MS side into a signal equivalent to uncorrelated Gaussian noise. The antenna weight V _i for eigenmode transmission included in this antenna weight is Hermitian transpose matrix A _{i of} equivalent channel matrix A _i (= H _i W _i ) having channel information H _i and transformation matrix W _i as elements. and ^H, and the eigenvector matrix obtained by eigenvalue expansion using a the channel matrix a _i (see Equation 8).

Then, the plurality of antennas Nt included in the base station BS forms a unique propagation path for each substream by performing beamforming for each eigenvector included in the eigenvector matrix V _i using the antenna weight V _i . Individual modulation processing and transmission power are controlled for each of a plurality of substreams. Since these combined antenna weights W _i V _i do not always have a norm of 1, the scaling factor α _i for the transmission symbol s _i (t) is different for each movement in order to define the transmission power. It is superimposed for each station.

This signal sequence is received by each mobile station MS # i as x _i (t). On the mobile station MS # i side, the received signal sequence x _i (t) is multiplied by the reception antenna weight W _ri . Thus, the received symbols r _i which are a plurality of (t) streams are separated and detected. At this time, the channel information (channel matrix H _i ) is always acquired on the base station BS side by being notified from the mobile station MS # i side, and is completely known on both sides.

As the receiving antenna weight W _ri , the eigenvector matrix of A _i A _i ^H satisfying the above equations (3) to (5) U _i ^H can be used, but the same result can be obtained by using the receiving antenna weights of the above-mentioned Zero Forcing norm or MMSE norm.

For example, in transmission in the eigenmode transmission of the FDD scheme as in the present embodiment, the channel matrix H _i obtained by using individual pilot symbols on the reception side as the reception antenna weight W _ri of the above-mentioned Zero Forcing standard And a pseudo-inverse matrix (H _i W _i V _i ) ⁻¹ of a transmission antenna weight composite matrix (extended channel matrix) H _i W _i V _i can be used. This pseudo-inverse matrix is also called a generalized inverse matrix, and its details are explained in "G. Strang, Masaya Yamaguchi (supervised), Linear Algebra and its Applications, Sangyo Tosho, 1978" etc. )

Using such a receiving antenna weight reduces the deterioration in characteristics caused by mismatch between the channel information on the transmitting side and the actual channel information caused by channel fluctuations compared to the case of using the eigenvector matrix of A _i A _i ^H. Can do. As the transformation matrix W _i included in the composite matrix, Equation (3) as well as transmission antenna weight (conversion matrix) W _i that satisfies, is calculated based on the Zero Forcing criterion, the above equation (2) Given the use of elements in a transformation matrix W _i of the transmitting antenna weight that satisfies, it is also possible to calculate the reception antenna weight.

(Configuration of wireless transmission system)
Next, the configuration of the wireless transmission system according to the present embodiment will be described in the case where it is applied to FDD (Frequency Division Duplex). By operating this wireless transmission system, the above-described wireless transmission method of the present invention can be implemented. FIG. 2 is a block diagram showing a configuration of the wireless transmission system according to the present embodiment.

  As shown in the figure, a base station BS transmits a transmission signal to a mobile station through a downlink, a transmission signal generation unit 11 that generates a transmission signal, a transmission signal division unit 12 that divides the transmission signal into one or a plurality of substreams, and a downlink. A plurality of transmission processing units 131 (132) that perform modulation processing and individual known symbol addition processing on the transmission signal transmitted in this manner, a transmission antenna weight multiplication unit 16 that multiplies or adds various coefficients to the transmission signal, and Nt And a common known symbol adding unit 17t for adding a common pilot symbol using a common control channel for each antenna. Each transmission antenna is provided with a device for multiplexing transmission signals for other mobile stations, a frequency conversion device (not shown) that is an up-converter that converts baseband signals into RF band signals, and the like. Yes.

The transmission antenna weight multiplication unit 16 is a transmission antenna weight superimposing unit that superimposes a transmission antenna weight on a transmission signal. Specifically, the above-mentioned coefficient and antenna weight (scaling coefficient α _i , antenna weight for eigenmode transmission) The transmission symbol s _i (t) is multiplied by V _i , the transmission antenna weight W _i ) for suppressing interference with other users.

  Each of the transmission processing units 131 and 132 includes a transmission signal modulation unit 141 (142) and individual known symbol addition units 151 and 152 for adding known symbols.

The base station BS transmits a module for generating antenna weights, the mobile station MS # i channel information acquisition unit 18 for acquiring channel information H _i in, based on the obtained channel information H _i, moving Transmit antenna weight for calculating combined antenna weight W _i V _i for realizing eigenmode transmission while converting the interference noise component at the station side into a signal equivalent to Gaussian noise uncorrelated with the transmit signal for other mobile stations And a generation unit 19.

Transmitting antenna weight multiplying unit 16, based on the acquired channel information in the channel information acquisition unit 18, a transformation matrix W _i and for converting the interference noise component at the mobile station side to Gaussian noise equivalent signal, singular Alternatively, V _i for forming a unique propagation path is multiplied for each of a plurality of substreams. Specifically, the transmission antenna weight generation unit 19 is based on the channel information H _i notified from the channel information notification unit 28, and the combined antenna weight W so as to satisfy the above formulas (1 ′) and (3) to (6). _i V _i is calculated.

On the other hand, the mobile station MS # i includes Nr reception antennas, a common known symbol component separation unit 21 that separates and detects a known symbol component from the received signal x _i (t), and an individual signal that is a known signal in each substream. The individual known symbol component separation unit 22 that separates individual pilot symbols using the control channel and inputs them to the reception antenna weight generation unit 26, and the interference noise component equivalent to Gaussian noise using the antenna weight (reception antenna weight) _Wri Receive antenna weight multiplier 23 that separates and detects a desired received symbol r _i (t) from the received signal converted into a simple signal, received signal demodulators 241 and 242 that demodulate the received signal for each substream, And a reception signal combining unit 25 for combining signals. Each receiving antenna is provided with a frequency converter (not shown) such as a down converter that converts an RF band signal into a baseband signal.

Further, the mobile station MS # i includes a channel estimation unit 27 that estimates channel information between the transmission and reception antennas from the reception signal of the common pilot symbol part, the channel information H _i from the reception signal of the individual pilot symbol part, and the combined antenna An extended channel matrix estimation unit 29 that estimates an extended channel matrix including the weights W _i V _i , a reception antenna weight generation unit 26 that generates reception antenna weights W _ri based on known symbols, and channel information of received signals And a channel information notification unit 28 for extracting the information and notifying the base station BS. In the present embodiment, the base station BS uses the channel information acquisition unit 18 to acquire channel information from the channel information notification unit 28 of the mobile station MS # i using a feedback line.

Further, the base station BS obtains the propagation path gain specific to each substream from the channel information acquired by the channel information acquisition unit 18 and the antenna weight generated by the transmission antenna weight generation unit 19, thereby obtaining an appropriate number of substreams. And a substream number / modulation scheme control unit 20 that determines a modulation scheme for each substream for each mobile station, and controls the transmission signal division unit 12 and the transmission signal modulation units 141 and 142, respectively. Note that the propagation path gain unique to each substream is obtained from the singular value of the equivalent channel matrix A _i = H _i W _i which is the product of the channel information H _i and the transformation matrix Wi.

(Wireless transmission method)
By operating the wireless transmission system having the above configuration, the wireless transmission method of the present invention can be implemented.

First, the transmission signal generated by the transmission signal generation unit 11 is divided into a plurality of signal sequences (substreams) S ₁ (t) and S ₂ (t) by the transmission signal division unit 12. Each of the divided substreams S ₁ (t) and S ₂ (t) is subjected to modulation processing for each of the plurality of substreams by the transmission signal modulation unit 141 (142), and the reception antenna weight _Wri is calculated. The individual pilot symbols are added by the individual known symbol adding unit 151 (152).

Next, after the scaling factor α _i corresponding to the sub-stream of each mobile station is superimposed in the transmission antenna weight multiplication unit 16, eigenmode transmission antenna weight (eigenvector matrix) V _i and other user interference suppression transformation matrix (whitening filter, antenna weights) that operates as a spatial filter W _i are respectively superimposed. The product of the whitening filter W _i multiplied by the transmission antenna weight multiplication unit 16 and the antenna weight V _i for eigenmode transmission (the combined antenna weight W _i V _i ) is the channel information notification unit 28 and the channel information acquisition unit 18. Is generated based on the channel information H _i acquired through.

  Thereafter, each transmission signal sequence is added and superimposed for each corresponding transmission antenna Nt, further added with a common pilot symbol, and transmitted through a plurality of antennas Nt through a unique propagation path formed for each substream. .

In the mobile station MS, the signal transmitted from the base station is received through the wireless transmission path. The signal x _i (t) received by each antenna of the mobile station #i is separated into a common pilot symbol component and a transmission information symbol component by the common known symbol component separation unit 21, and the transmission information symbol component is individually known. The common pilot symbol component is input to the symbol estimation unit 22 and the common pilot symbol component is input to the channel estimation unit 27.

  Channel estimation unit 27 estimates channel characteristics between the respective transmitting and receiving antennas based on the common pilot symbol components, calculates channel information indicating the channel characteristics, and outputs the channel information to channel information notification unit 28. The base station BS uses the channel information acquisition unit 18 to acquire channel information from the channel information notification unit 28 of the mobile station MS using a feedback line.

Further, the number of substreams and the modulation method are selected for each substream from the channel information acquired by the channel information acquisition unit 18 and the antenna weight generated by the transmission antenna weight generation unit 19 by the substream number / modulation method control unit 20. A unique channel gain is obtained, and an appropriate number of substreams and a modulation scheme for each substream are determined for each mobile station so that throughput is maximized or error rate is minimized. Controls transmission signal modulators 141 and 142. Note that the propagation path gain unique to each substream is obtained from the singular value of the equivalent channel matrix A _i = H _i W _i which is the product of the channel information H _i and the transformation matrix Wi.

The individual known symbol component separation unit 22 separates the individual pilot symbol components and inputs the separated individual pilot symbols to the extended channel matrix estimation unit 29 and the reception antenna weight generation unit 26. The extended channel matrix estimation unit 29 estimates the extended channel matrix H _i W _i V _i based on the dedicated pilot symbols, and inputs the estimation result to the reception antenna weight generation unit. The reception antenna weight generating unit 26, the estimated expanded channel matrix H _i W _i V _i received based on the antenna weight W _ri (where is the pseudo-inverse of the extension channel matrix _{H i W i V i (H} i W _i V _i ) ⁻¹ ) is calculated and input to the receiving antenna weight multiplier 23 of each signal separation detector. (Note that if a square matrix X does not have zero eigenvalues, an inverse matrix exists, and the inverse matrix of X and the pseudo-inverse of X are the same. Therefore, H _i W _i V _i does not have zero eigenvalues. (If it is a square matrix, the inverse matrix of H _i W _i V _i can be used as the receiving antenna weight.)
On the other hand, the transmission information symbol component from which the individual pilots are separated is multiplied by the reception antenna weight multiplier 23 and multiplied by the reception antenna weight _Wri , and further demodulated by the reception signal demodulation unit 241 (242) for each substream. Is done. Then, the received signals after the demodulation processing are combined in the reception signal combining unit 25 by a method corresponding to the dividing process of the transmission signal dividing unit 12, and the transmission signal sequence is restored.

(Operations and effects in this embodiment)
According to the radio transmission system and radio transmission method according to the present embodiment described above, the interference noise component on the mobile station side as a transmission antenna weight for other-user interference suppression is a Gauss that is uncorrelated with the transmission signal for other mobile stations. Since a so-called whitening filter that converts noise is used, mutual interference between mobile stations can be suppressed without degrading the received SNR.

Next, the simulation model and simulation results of the present invention will be described. Table 1 shows the specifications of this simulation.

As shown in Table 1, in this embodiment, the array antenna shape is a half-wavelength equally spaced linear array for both transmission and reception, the number of transmitting antennas Nt is 4, and the number of receiving antennas Nr is 2. The number of mobile stations is 2, and the direction of mobile stations is θ _t (= θ _t1 , θ _t2 ). Here, for simplicity, it is assumed that the transmission power assigned to each mobile station is the same, the average reception power is the same between the mobile stations, and the channel information is completely known for both transmission and reception.

  Further, in this embodiment, transmission is based on 4-bit transmission, and the modulation scheme is an average error rate of two modes of (1) QPSK × 2 stream or (2) 16QAM × 1 stream shown in Table 1. The smaller one is selected. In the mode (1), the transmission power allocated to each substream is optimally distributed so that the average bit error rate is minimized.

As a result of the above simulation, the average bit error rate characteristics when the direction θ _t of the two mobile stations is (−40 °, 60 ° (CaseI)) and (40 °, 60 ° (CaseII)) are shown in FIG. Respectively. FIG. 3 shows that when the whitening filter of the present invention is used (with WF), the error rate characteristic can be greatly improved as compared with the case of using a conventional Zero Forcing standard transmission antenna weight (with ZF).

In addition, when compared with the average received SNR with an average error rate of 10 ^-2 , when using the whitening filter of the present invention, it is about 2 dB for Case I, about 6 dB for Case II, and the transmission antenna weight of the conventional Zero Forcing standard Compared to Although the characteristics of both methods deteriorate as the direction difference between mobile stations narrows, the amount of deterioration is smaller when using a whitening filter than when using a transmitting antenna weight of the Zero Forcing standard. .

It is explanatory drawing which shows the outline | summary of the wireless transmission method which concerns on embodiment. 1 is a block diagram illustrating a configuration of a wireless transmission system according to an embodiment. It is a graph which shows the average bit error rate characteristic as a simulation result in an Example. It is explanatory drawing which shows the outline | summary of the conventional wireless transmission method.

Explanation of symbols

BS ... Base station
H _i ... Channel information
MS ... Mobile station
Nr ... Number of receiving antennas
Nt ... Number of transmitting antennas
U _i ^H Receive antenna weight
W _i ... Transmit antenna weight α _i ... Scaling factor
r _i (t) ... Received symbol
s _i (t): Transmission symbol
x _i (t): Received signal
11 ... Transmission signal generator
12 ... Transmission signal divider
16 ... Transmitting antenna weight multiplier
17t ... Common known symbol addition part
18 Channel information acquisition unit
19 ... Transmitting antenna weight generator
20 ... Substream number / modulation method controller
21 ... Common known symbol component separation unit
22: Individual known symbol component separation unit
23 ... Receiving antenna weight multiplier
25 ... Received signal coupling part
26 ... Receiving antenna weight generator
27 ... Channel estimation section
28 ... Channel information notification section
29 ... Extended channel matrix estimation unit
131, 132 ... transmission processing section
141, 142 ... Transmission signal modulation section
151, 152 ... Individual known symbol adding section
241, 242 ... Received signal demodulator

Claims (10)

Between a base station having a plurality of antennas and a plurality of mobile stations having a plurality of antennas, signals are mutually transmitted from the plurality of antennas at the same time at the same carrier frequency by frequency division duplex or time division duplex. A wireless transmission system for transmitting and receiving, wherein the base station is
Channel information acquisition means for acquiring each channel information between the base station and the plurality of mobile stations in the downlink based on signals transmitted from the plurality of mobile stations ;
Based on the acquired channel information, an interference noise component on the mobile station side generated by a transmission signal and noise to other mobile stations that transmit and receive signals to and from the base station is converted into a signal equivalent to Gaussian noise. Transformation matrix generation means for generating a transformation matrix for conversion;
Transmit antenna weight superimposing means for superimposing transmit antenna weights each having the transform matrix as an element on a substream of each mobile station generated from transmission signals transmitted to the plurality of mobile stations via a downlink;
Each substream on which the transmission antenna weight is superimposed by the transmission antenna weight superimposing means is added for each corresponding antenna and superimposed, and transmitted by the plurality of antennas through a unique propagation path formed for each substream. And means for
The mobile station separates and detects a desired signal in the downlink based on a reception antenna weight generated based on the channel information from a reception signal in which an interference noise component is converted into a signal equivalent to Gaussian noise. wireless transmission system comprising means. The transmission antenna weight includes an eigenvector matrix obtained by eigenvalue expansion of an equivalent channel matrix that is a product of the channel information and the transformation matrix and a Hermitian transpose matrix of the equivalent channel matrix,
The base station uses the eigenvector matrix to perform beamforming for each eigenvector included in the eigenvector matrix, thereby forming a unique propagation path for each substream or a plurality of substreams. 2. The radio transmission system according to claim 1, wherein modulation processing and transmission power control are performed on the stream. The signal separation detection unit of the mobile station receives an extended channel matrix that is a product of the transmission antenna weight and the channel information including the transformation matrix and the eigenvector matrix using an individual pilot symbol set for each mobile station. 3. The radio transmission system according to claim 2, wherein a matrix obtained on the side and obtained by using the extended channel matrix is used as the reception antenna weight.   4. The radio transmission system according to claim 3, wherein the reception antenna weight of the mobile station is a pseudo inverse matrix of the extended channel matrix. A base station that transmits / receives signals to / from a plurality of antennas at the same time at the same carrier frequency using frequency division duplexing or time division duplexing,
Channel information acquisition means for acquiring each channel information between the base station and the plurality of mobile stations in the downlink based on signals transmitted from the plurality of mobile stations ;
Based on the acquired channel information, an interference noise component on the mobile station side generated by a transmission signal and noise to other mobile stations that transmit and receive signals to and from the base station is converted into a signal equivalent to Gaussian noise. Transformation matrix generation means for generating a transformation matrix for conversion;
Transmit antenna weight superimposing means for superimposing transmit antenna weights each having the transform matrix as an element on a substream of each mobile station generated from transmission signals transmitted to the plurality of mobile stations via a downlink;
Each substream on which the transmission antenna weight is superimposed by the transmission antenna weight superimposing means is added for each corresponding antenna and superimposed, and transmitted by the plurality of antennas through a unique propagation path formed for each substream. base station, characterized in that it comprises a a <br/> means. The transmit antenna weight includes an eigenvector matrix obtained by eigenvalue expansion using an equivalent channel matrix having the channel information as an element and a Hermitian transpose matrix of the equivalent channel matrix,
By performing beam forming for each eigenvector included in the eigenvector matrix using the eigenvector matrix, a unique propagation path is formed for each substream or each substream, and modulation processing for each substream is performed. 6. The base station according to claim 5, wherein transmission power control is performed. A wireless transmission method for transmitting and receiving signals to and from a base station having a plurality of antennas and a mobile station having a plurality of antennas at the same carrier frequency at the same time from the plurality of antennas,
In the base station, obtaining each channel information between the base station and the plurality of mobile stations in the downlink , based on signals transmitted from the plurality of mobile stations, (1),
In the base station, based on the acquired channel information, an interference noise component on the mobile station side generated by a transmission signal and noise to other mobile stations that transmit and receive signals to and from the base station is expressed as Gaussian noise. (2) generating a transformation matrix for generating a transformation matrix for converting to a signal equivalent to
In the base station, the sub-stream of each mobile station that is generated from the transmission signal, the transmission antenna weight to the transformation matrix element superimposed respectively, to the mobile station, and transmitted via downlink, the transmit antenna weights Is superimposed for each corresponding antenna, and is transmitted by the plurality of antennas through a unique propagation dew formed for each substream; and
The mobile station comprises a step (4) of separating and detecting a desired signal from a reception signal in which an interference noise component is converted into a signal equivalent to Gaussian noise, using a reception antenna weight generated based on the channel information. A wireless transmission method. The transmission antenna weight includes an eigenvector composed of an eigenvector obtained by eigenvalue expansion or singular value decomposition using an equivalent channel matrix that is a product of the channel information and the transformation matrix, and a Hermitian transpose matrix of the equivalent channel matrix. Contains a matrix,
In the step (3), the base station uses the eigenvector matrix to perform a beamforming for each eigenvector included in the eigenvector matrix, thereby forming a unique propagation path for each substream or a plurality of substreams. 8. The radio transmission method according to claim 7, wherein modulation processing and transmission power for each substream are controlled for each substream. In the step (4), an extended channel matrix that is a product of the transmission antenna weight and the channel information including the transformation matrix and the eigenvector matrix is acquired on the reception side using an individual pilot symbol set for each mobile station. 9. The radio transmission method according to claim 8, wherein a matrix obtained by using the extended channel matrix is used as the reception antenna weight.   10. The radio transmission method according to claim 9, wherein the reception antenna weight of the mobile station is a pseudo inverse matrix of the extended channel matrix.

Images