CN1729634B - Channel estimation and spatial processing for TDD MIMO systems - Google Patents

Abstract

The present invention provides a channel estimation and spatial processing for a TDD MIMO system. Calibration may be performed to account for differences in the responses of transmit/ receive chains at the access point and user terminal (512). During normal operation, a MIMO pilot is transmitted on a first link (522) and used to derive an estimate of the first link channel response (524), which is decomposed to obtain a diagonal matrix of singular values and a first unitary matrix containing both left eigenvectors of the first link and right eigenvectors of a second link (526). A steered reference is transmitted on the second link using the eigenvectors in the first unitary matrix (530), and is processed to obtain the diagonal matrix and a second unitary matrix containing both left eigenvectors of the second link and right eigenvectors of the first link (532). Each unitary matrix may be used to perform spatial processing for data transmission/reception via both links (540, 542, 550, 552).

Description

The channel estimating of TDD mimo system and spatial manipulation

Prioity claim according to 35U.S.C. § 119

The application requires the 60/421st, No. 428, the 60/421st, the 462 and the 60/421st, 309 U.S. Provisional Application No., all three application submission dates are on October 25th, 2002, described three applications are entitled as " Channel estimation and Spatial Processing for TDD MIMOSystems " in order respectively, " Channel Calibration for a Time Division DuplexedCommunication System " and " MIMO WLAN System ", all are transferred to assignee of the present invention and are incorporated herein by reference fully at this.

Background

The field

The present invention relates generally to communication, relate in particular to the technology that in time division duplex (TDD) multiple-input and multiple-output (MIMO) communication system, realizes channel estimating and spatial manipulation.

Background

Multiple-input and multiple-output (MIMO) communication system is used a plurality of (N _T) transmitting antenna and a plurality of (N _R) reception antenna carries out transfer of data, and be called as (N _T, N _R) system.By N _TIndividual transmitting antenna and N _RThe mimo channel that individual reception antenna forms may be broken down into N _SIndividual independent channel, wherein N _S≤ min{N _T, N _R.N _SEach of individual independent channel also is called as the spatial sub-channel or the eigenmodes of mimo channel, and corresponding one dimension.If utilize a plurality of additional dimension that antenna is set up that transmit and receive, then can improve mimo system performance (for example transmission capacity of Zeng Jiaing)

For N at mimo channel _SOne or more the going up of individual eigenmodes sends data, must also handle in implementation space, transmitter place at the receiver place and usually.From N _TThe data flow that individual transmitting antenna sends is disturbed at the reception antenna place mutually.Spatial manipulation attempts to separate the data flow at receiver place, and it can be resumed separately.

Handle for the implementation space, generally need the accurate channel response between the transmitter and receiver to estimate.For the TDD system, down link between access point and user terminal (being forward link) and up link (being reverse link) are generally shared same frequency band.In this case, realizing calibration (as described below) to consider the transmitting and receiving after the chain difference of access point and user terminal place, down link and uplink channel responses are supposed (reciprocal) reciprocal each other.If promptly HThe channel response matrix of expression from aerial array A to aerial array B, channel then reciprocal represent from array B to array A coupling by H ^TProvide, wherein M ^TExpression MTransposition.

The major part of the channel estimating of mimo system and the general consume system resources of spatial manipulation.Therefore need a kind of technology to realize channel estimating and spatial manipulation that the TDD mimo system is interior effectively in the field.

General introduction

In being provided at the TDD mimo system, this realizes the technology of channel estimating and spatial manipulation with effective means.For the TDD mimo system, the characteristic of channel reciprocal can be used to simplify the channel estimating and the spatial manipulation at transmitter and receiver place.During beginning, intrasystem access point and user terminal can be realized calibrating with the difference of determining the response that it transmits and receives chain and the correction factor that obtains to be used to consider this difference.Realize that calibration is reciprocal to guarantee to have used " calibration back " channel of correcting the factor.Like this, can be based upon estimation acquisition second link estimation more accurately that first link is derived.

During normal running, the channel response that (for example down link) sends MIMO pilot tone (for example by access point) and be used to derive first link on first link is estimated.Channel response is estimated then can be through decomposing (for example using singular value decomposition by user terminal) to obtain the diagonal singular value matrix and to comprise the left eigenvector of first link and the right eigenvector of second link (for example up link).Therefore first unitary matrix can be used in transfer of data that receives on first link and the transfer of data implementation space that sends on second link and handle.

Can on second link, use the eigenvector in first unitary matrix to send the manipulation benchmark.Handle the pilot tone that benchmark (or handling pilot tone) is to use the eigenvector that is used for transfer of data to send on specific eigenmodes.This manipulation benchmark then can treated (for example by access point) to obtain diagonal matrix and to comprise the left eigenvector of second link and second unitary matrix of the right eigenvector of first link.Therefore second unitary matrix can be used in the transfer of data that receives on second link and the transfer of data implementation space that will send on first link and handle.

Various aspects of the present invention and embodiment are described further below.

Brief description of the drawings

Various aspects of the present invention and characteristic are described together with the following drawings following, wherein:

Fig. 1 is according to one embodiment of the invention, access point and user terminal block diagram in the TDD mimo system.

Fig. 2 A is illustrated in the block diagram that access point and user terminal place transmit and receive chain according to one embodiment of the invention;

Fig. 2 B illustrates the correction matrix application according to one embodiment of the invention, and described matrix is used to consider the poor of access point and user terminal place transmit.

Fig. 3 represents the down link of spatial multiplexing modes and the spatial manipulation of up link according to an embodiment of the invention.

Fig. 4 illustrates the down link of beam steering pattern and the spatial manipulation of up link according to one embodiment of the invention; And

Fig. 5 is illustrated in the process of access point and user terminal place realization channel estimating and spatial manipulation according to one embodiment of the invention.

Describe in detail

Fig. 1 is the access point 110 in the TDD mimo system 100 and the embodiment block diagram of user terminal 150.Access point 110 has the N that is used for transfer of data _ApTransmit/receive antenna is used for transfer of data/reception, and user terminal 150 has N _UtIndividual transmit/receive antenna.

On down link, at access point 110 places, emission (TX) data processor 114 receives traffic data (being information bit) and slave controller 130 reception signaling and other data from data source 112.114 pairs of data of TX data processor format, encode, interlock and modulate (being symbol mapped) so that modulated symbol to be provided.TX spatial processor 120 is handled so that N to be provided from TX data processor 114 reception modulated symbols and implementation space _ApIndividual transmitter code flow filament, stream of each antenna.The also suitably multiplexed pilot frequency code element of TX spatial processor 120.

Each modulator (MOD) 122 (this comprises the emission chain) receives and handles corresponding transmitter code flow filament so that corresponding down link modulated signal to be provided.From the N of modulator 122a to 122ap _ApIndividual down link modulated signal is then correspondingly from N _ApIndividual antenna 124a sends to 124ap.

At user terminal 150 places, N _UtThe down link modulated signal that individual antenna 152a receive to send to 152ut, and the corresponding demodulator of each day alignment (DEMOD) 154 provides and receives signal.Each demodulator 154 (this comprises the reception chain) is realized with the complementary processing that realizes at modulator 122 places and receiving symbol is provided.Receive (RX) spatial processor 160 then to handling so that the recovery code element to be provided to the code element implementation space that receives of 154ut from all demodulator 154a, this is the estimation of the modulated symbol of access point transmission.RX data processor 170 also further (for example code element is gone mapping, deinterleave and decoding) recovers code element so that decoding back data to be provided.Decoding back data can comprise the traffic data, signaling of recovery etc., they can be provided for data sink 172 with storage and/or controller 180 with further processing.

The processing of up link can be identical or different with the processing of up link.Data and signaling are handled (for example encode, interlock and modulation) by TX data processor 188, and are further handled by TX spatial processor 190, described processor also suitably multiplexed in pilot frequency code element (for example for calibration and normal running).Pilot tone and transmit symbol from TX spatial processor 190 are further handled to generate N to 154ut by modulator 154a _UtIndividual up link modulated signal, these signals are sent to access point by antenna 152a to 152ut then.

At access point 110 places, the up link modulated signal is received to 124ap by antenna 124a, by demodulator 122a to the 122ap demodulation, and by RX spatial processor 140 and RX data processor 142 with handle in the mode of user terminal place complementation.Data can be provided for data sink 144 to store and/or controller 130 is further handled after the decoding of up link.

Controller 130 and 180 is controlled at the operation of the corresponding various processing units of access point and user terminal place.Memory 132 and 182 is storage control 130 and 180 data and the program codes that use correspondingly.

1. calibration

For the TDD system,, between down link and uplink channel responses, there is high correlation because down link is shared identical frequency band with up link.Therefore down link and uplink channel responses matrix can be supposed (being transposition) reciprocal each other.Yet the response of the transmit at access point place generally is not equal to the response in user terminal place transmit.In order to improve performance, its difference can be determined and can be considered by calibration.

Fig. 2 A transmits and receives the chain block diagram according to what one embodiment of the invention illustrated access point 110 and user terminal 150 places.For down link, at access point 110 places, code element is (with sending vector x _DnExpression) by 214 processing of emission chain and from N _Ap Individual antenna 124 is sent out on mimo channel.At user terminal 150 places, down link signal is by N _Ut Individual antenna 152 receives and handles to provide the code element that receives (with " reception " vector by receiving chain 254 r _DnExpression).For up link, at user terminal 150 places, code element is (by the emission vector x _UpExpression) handle by emission chain 264 and on mimo channel from N _UtIndividual antenna is sent out.At access point 110 places, uplink signal is by N _Ap Individual antenna 124 receives and receives code element (by receiving vector by receiving chain 224 processing to provide r _UpExpression).

For down link, the reception vector at user terminal place r _Dn(under noise-free case) can be represented as:

r _Dn= R _Ut HT _Ap x _Dn, equation (1)

Wherein x _DnBe the N that has of down link _ApThe emission vector of item;

r _DnBe to have N _UtThe reception vector of item;

T _ApBe N _Ap* N _ApDiagonal matrix has the N with the access point place _ApThe complex gain item that the emission chain of individual antenna is associated;

R _UtBe N _Ut* N _UtDiagonal matrix has the N with the user terminal place _UtThe complex gain item that the reception chain of individual antenna is associated; And

HBe the N of down link _Ut* N _ApChannel response.

The response of transmit and mimo channel generally is the function of frequency.In order to simplify, for mild fading channel (promptly having mild frequency response) is supposed in following derivation.

For up link, the reception vector at access point place r ^Up(under noise-free case) can be expressed as:

r _Up= R _Ap H ^T T _Ut x _Up, equation (2)

Wherein x _UpBe the N that has of down link _UtThe emission vector of item;

r _UpBe to have N _ApThe reception vector of item;

T _UtBe N _Ut* N _UtDiagonal matrix has the N with the user terminal place _UtThe complex gain item that the emission chain of individual antenna is associated;

R _ApBe N _Ap* N _ApDiagonal matrix has the N at the access point place _ApThe complex gain item that the reception chain of individual antenna is associated; And

H ^TBe the N of up link _Ap* N _UtChannel response matrix.

Draw " effectively " down link and uplink channel responses from equation (1) and (2) H _DnWith H _Up(expression can be used the response that transmits and receives antenna) can be expressed as:

${\underset{\‾}{H}}_{\mathrm{dn}}{\underset{\‾}{=R}}_{\mathrm{ut}}\underset{\‾}{{\mathrm{HT}}_{\mathrm{ap}}}$ And ${\underset{\‾}{H}}_{\mathrm{up}}{\underset{\‾}{=R}}_{\mathrm{ap}}\underset{\‾}{{H^{T}T}_{\mathrm{ut}}}$

Equation (3)

Illustrate as equation (3), if the response of the transmit at access point place is not equal to the transmit response at user terminal place, then active downlink and uplink channel responses are not that inverse is promptly each other: ${\underset{\‾}{R}}_{\mathrm{ut}}\underset{\‾}{{\mathrm{HT}}_{\mathrm{ap}}}\≠{\underset{\‾}{R}}_{\mathrm{ap}}\underset{\‾}{H^T{T}_{\mathrm{ut}}}$

With the combination in equation set (3) of two equatioies, can obtain following relation:

$\underset{\‾}{H}={{\underset{\‾}{R}}_{\mathrm{ut}}}^{-1}{\underset{\‾}{H}}_{\mathrm{dn}}{T_{\underset{\‾}{\mathrm{ap}}}}^{-1}={\left({{\underset{\‾}{R}}_{\mathrm{ap}}}^{-1}{\underset{\‾}{H}}_{\mathrm{up}}{T_{\underset{\‾}{\mathrm{ut}}}}^{-1}\right)}^T={{\underset{\‾}{T}}_{\mathrm{ut}}}^{-1}{\underset{\‾}{H^T}}_{\mathrm{up}}{T_{\underset{\‾}{\mathrm{ap}}}}^{-1}$

Equation (4)

Rearrange equation (4), below can obtaining:

${{\underset{\‾}{H}}_{\mathrm{up}}}^T={\underset{\‾}{T}}_{\mathrm{ut}}{{\underset{\‾}{R}}^{-1}}_{\mathrm{ut}}{\underset{\‾}{H}}_{\mathrm{dn}}{T_{\underset{\‾}{\mathrm{ap}}}}^{-1}{\underset{\‾}{R}}_{\mathrm{ap}}={K_{\mathrm{ut}}}^{-1}{\underset{\‾}{H}}_{\mathrm{dn}}{K}_{\underset{\‾}{\mathrm{ap}}}$

Or

${\underset{\‾}{H}}_{\mathrm{up}}={\left({K_{\mathrm{ut}}}^{-1}{\underset{\‾}{H}}_{\mathrm{dn}}{K}_{\underset{\‾}{\mathrm{ap}}}\right)}^T$

Equation (5)

K wherein _Ut= T _Ut ^-1 R _Ut, and $K\underset{\‾}{{}_{\mathrm{ap}}=}{T_{\underset{\‾}{\mathrm{ap}}}}^{-1}{\underset{\‾}{R}}_{\mathrm{ap}}.$ Because T _Ut, R _Ut, T ApWith R _ApBe diagonal matrix, K ApAnd K _UtIt also is diagonal matrix.Equation (5) can be expressed as:

${\underset{\‾}{H}}_{\mathrm{up}}{K}_{\underset{\‾}{\mathrm{ut}}}={\left({\underset{\‾}{H}}_{\mathrm{dn}}{K}_{\underset{\‾}{\mathrm{ap}}}\right)}^T$

Equation (6)

Matrix K ApAnd K _UtCan be considered and comprise " the correction factor ", the described factor is considered poor in the transmit at access point and user terminal place.This allows the channel response of a link to be represented by the channel response of another link, as illustrating in the equation (5).

Can realize that calibration is to determine matrix K ApAnd K _UtGenerally, real channel response HResponse is not known with transmit, and they also can not accurately or simply be determined.On the contrary, active downlink and uplink channel responses H _DnWith H _UpCan be based on the MIMO pilot tone of corresponding transmission on down link and up link and estimated.MIMO pilot tone and generation and use are described in detail in aforesaid No. 60/421309 U.S. Patent application sequence.

Matrix K ApAnd K _UtEstimation be called as the correction matrix, With

Can estimate based on down link and uplink channel responses With

And derive, this can derive in every way, comprises matrix than calculating and least mean-square error (MMSE) calculating.Calculate (N for the matrix ratio _Ut* N _Ap) matrix CAt first estimate to compare and calculate as up link and uplink channel responses, as follows:

$\underset{\‾}{C}=\frac{{\underset{\‾}{\overset{^}{H}}}_{\mathrm{up}}^T}{{\underset{\‾}{\overset{^}{H}}}_{\mathrm{dn}}}$

Equation (7)

Wherein than being that each element is divided by. CEach element therefore can be calculated as:

${\underset{\‾}{c}}_{i,j}=\frac{{\hat{h}}_{\mathrm{upi},j}}{{\hat{h}}_{\mathrm{dni},j}}$ Wherein, i={1...N _UtAnd j={1...N _Ap,

Wherein

With Correspondingly be

With

The (i, j) (OK, row), and c _{I, j}Be C(i, j) individual element.

The correction vector of access point Include only

N _ApIndividual diagonal element can be defined as equaling CThe average of standardization row. CEvery row c _iAt first by each element in will going divided by first element of row through standardization to obtain corresponding standardization row If therefore ${\underset{\‾}{c}}_i\left(k\right)=[c_{i,1}...c_{{i,N}_{\mathrm{ap}}}]$ Be CI capable, standardization row then Can be expressed as:

${\underset{\‾}{\overset{~}{c}}}_i\left(k\right)=[c_{i,1}\left(k\right)/c_{i,1}\left(k\right)...c_{i,j}\left(k\right)/c_{i,1}\left(k\right)...c_{{i,N}_{\mathrm{ap}}}\left(k\right)/c_{i,1}\left(k\right)]$

Correct vector

Be set equal to then CN _UtThe average of individual standardization row, it can be expressed as:

${\underset{\‾}{\overset{^}{k}}}_{\mathrm{ap}}=\frac{1}{N_{\mathrm{ut}}}\underset{i=1}{\overset{N_{\mathrm{ut}}}{\mathrm{\Σ}}}\underset{\‾}{\overset{~}{c_i}}$

Equation (8)

Because standardization,

First element be unit one (unity).

User terminal has the vector of correction

Include only

N _UtIndividual diagonal element can be defined as equaling CStandardization row contrary. CEvery row c _jAt first by using vector J element each element in being listed as is carried out proportional zoom, to obtain corresponding standardization row

, described j element K _{Apj, j}Expression.Therefore, if ${\underset{\‾}{c}}_j\left(k\right)={[c_{1,j}...c_{N_{\mathrm{ut},j}}]}^T$ Be CJ capable, standardization row then

Can be expressed as:

Correct vector Be set equal to then CN _ApThe contrary average of individual standardization row, and can be represented as:

Equation (9)

Wherein standardization is listed as Contrary be that each element is realized.

Calibration is correspondingly corrected vector for access point and user terminal provide With Or corresponding correction matrix With

Correct matrix

With MMSE calculate and at length in aforesaid No. 60/421462 U.S. Patent application sequence, to describe in detail.

Fig. 2 B has illustrated the application of correcting matrix according to one embodiment of the invention, poor with the transmit of considering access point and user terminal place.On down link, the emission vector x _DnAt first multiply by matrix by unit 212 The emission chain 214 of down link with receive chain 254 handle in succession with Fig. 2 A in illustrate identical.Similarly, on up link, the emission vector x _UpAt first multiply by matrix by unit 262

What illustrate in emission chain 264 and the processing in succession that receives chain 224 and Fig. 2 A equally, is identical.

Observed " calibration back " down link of user terminal and access point and uplink channel responses correspondingly can be expressed as:

$H_{\mathrm{cdn}}={\underset{\‾}{H}}_{\mathrm{dn}}{\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}$ and $H_{\mathrm{cup}}={\underset{\‾}{H}}_{\mathrm{up}}{\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}$ Equation (10)

Wherein H _Cdn ^TWith H _CupIt is the estimation that " truly " calibration back channel response is expressed in the equation (6).From equation (6) and (10) as seen ${\underset{\‾}{H}}_{\mathrm{cup}}\≈{\underset{\‾}{H}}_{\mathrm{cdn}}^T.$ ${\underset{\‾}{H}}_{\mathrm{cup}}\≈{\underset{\‾}{H}}_{\mathrm{cdn}}^T$ The accuracy of relation depends on estimation

With

Accuracy, this depends on that then down link and uplink channel responses estimate

With Quality.As above illustrating, in case calibrated transmit, is that the calibration channel response that a link obtains (is for example estimated ) can (for example estimate as channel response after the calibration of another link

).

Being aligned in aforesaid No. 60/421309 U.S. Patent application sequence and No. 60/421462 U.S. Patent application of TDD mimo system described in detail.

2. spatial manipulation

For mimo system, data can send on one or more eigenmodes of mimo channel.Can the definition space multiplexing modes covering the transfer of data on a plurality of eigenmodes, and can define the beam steering pattern to cover the transfer of data on the single eigenmodes.Operator scheme requires the spatial manipulation at transmitter and receiver place.

Channel estimating described here and spatial processing technique can be used to have or do not have the mimo system of OFDM.OFDM is divided into a plurality of (N with the total system bandwidth effectively _F) orthogonal subbands, they also are called as frequency zone or subchannel.Under the OFDM situation, each subband is associated with corresponding subcarrier, thereon modulating data.For the mimo system (being the MIMO-OFDM system) that uses OFDM, each eigenmodes of each subband can be considered the independent transmission channel.For clear, channel estimating and spatial processing technique following be TDD MIMO-OFDM system description.For this system, each subband of wireless channel can be assumed to be inverse.

Correlation between down link and uplink channel responses can be used to simplify the channel estimating and the spatial manipulation of the access point and the user terminal of TDD system.This is reduced at has realized that calibration is effective afterwards to consider the difference in the transmit.The channel response of calibration can be expressed as the function of frequency, and is as follows:

${\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{dn}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right)$ Wherein k ∈ K, and equation (11)

${\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right)\_={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right)={\left({\underset{\‾}{H}}_{\mathrm{dn}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}\left(k\right)\right)}^T,$ K ∈ K wherein,

Wherein K represents to be used for all sets of subbands (" being data subband ") of transfer of data.Can realize calibrating and think that each data subband obtains matrix

With Perhaps, can realize calibration for a subclass of all data subbands, wherein the matrix of " without calibration " subband With

Can describe in No. 60/421462 U.S. Patent application sequence as described above by for " calibration back " subband interpolation matrix obtains.

The channel response matrix of each subband H(k) can " by diagonalization " to obtain the N of this subband _SIndividual eigenmodes.This can pass through channel response matrix H(k) realize singular value decomposition or right H(k) correlation matrix is realized eigen value decomposition and obtains that correlation matrix is R(k)= H ^H(k) H(k).For clear, singular value decomposition is used for following description.

Calibration back downlink channel response matrix H _Cup(k) singular value decomposition can be represented as:

${\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right)={\underset{\‾}{U}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\mathrm{\Σ}}\left(k\right){\underset{\‾}{V}}_{\mathrm{ut}}^H\left(k\right),$ K ∈ K equation (12)

Wherein U _Ap(k) be H _Cup(k) (N that makes eigenvector _Ap* N _Ap) unitary matrix;

∑(k) be H _Cup(the N of singular value (k) _Ap* N _Ut) diagonal matrix; And

V _Ut(k) be H _Cup(the N of right eigenvector (k) _Ut* N _Ut) unitary matrix.

The unitary matrix characteristic is M ^H M= I, wherein IIt is unit matrix.

Correspondingly, calibration downlink channel response matrix H _Cdn(k) singular value decomposition can be represented as:

${\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right)={\underset{\‾}{V}}_{\mathrm{ut}}^{*}\left(k\right)\underset{\‾}{\mathrm{\Σ}}\left(k\right){\underset{\‾}{U}}_{\mathrm{ap}}^T\left(k\right),$ K ∈ K equation (13)

Matrix wherein V _Ut ^*(k) and U _Ap ^*(k) correspondingly be H _CdnThe unitary matrix of left and right sides eigenvector (k).Illustrate as equation (12) and (13), and based on following description, the left and right sides eigenvector matrix of a link is corresponding to be the complex conjugate of the left and right sides eigenvector matrix of another link.Matrix V _Ut(k), V _Ut ^*(k), V _Ut ^T(k) and V _Ut ^H(k) be matrix V _Ut(k) multi-form, and matrix U _Ap(k), U _Ap ^*(k), U _Ap ^T(k) and U _Ap ^H(k) be U _Ap(k) multi-form.In order to simplify, the matrix in below describing U _Ap(k) and V _Ut(k) with reference to can also refer to its various other forms.Matrix U _Ap(k) and V _Ut(k) correspondingly be used for spatial manipulation and represent by its subscript by access point and user terminal.Eigenvector also often is called as " manipulation " vector.

Singular value decomposition is further described in detail in the book of Gilbert Strang, is entitled as " Linear Algebraand Its Applications ", second edition, Academic Press, 1980.

User terminal can be estimated calibration back downlink channel response based on the MIMO pilot tone that access point sends.User terminal can be estimated for the calibration downlink channel response Realize singular value decomposition, k ∈ K wherein is to obtain The diagonal matrix of left eigenvector And matrix V _Ut ^*(k).This singular value decomposition can be given ${\widehat{\underset{\‾}{H}}}_{\mathrm{cdn}}\left(k\right)={\underset{\‾}{V}}_{\mathrm{ut}}^{*}\left(k\right)\underset{\‾}{\widehat{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{U}}_{\mathrm{ap}}^T\left(k\right),$ Wherein the expression of (" ^ ") on each matrix is the estimation of actual matrix.

Similarly, access point can be estimated calibration back uplink channel responses based on the mimo channel that user terminal sends.Access point can be estimated for calibrating the back uplink channel responses Realize singular value decomposition, k ∈ K wherein is to obtain

The diagonal matrix of left eigenvector

And matrix This singular value decomposition can be given ${\widehat{\underset{\‾}{H}}}_{\mathrm{cup}}\left(k\right)={\underset{\‾}{\hat{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\widehat{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{\hat{V}}}_{\mathrm{ut}}^H\left(k\right).$

Yet because channel reciprocal and calibration, singular value decomposition only need be realized by user terminal or access point.If realize by user terminal, then matrix (wherein k ∈ K) is used for the spatial manipulation at the user terminal place, and matrix (wherein k ∈ K) can (promptly pass through sending metrix with direct form Item) or not direct form (for example by handling benchmark, following description) be provided for access point.

Each matrix

Interior singular value (wherein k ∈ K) can be sorted, and making the row of winning comprise maximum singular value, and secondary series comprises second maximum singular value, so analogizes (promptly ${\mathrm{\σ}}_1\≥{\mathrm{\σ}}_2\≥...\≥{\mathrm{\σ}}_{N_S}$ , σ wherein _iBe after ordering

I row in eigenvalue).When each matrix

Singular value be sorted after, the relevant unitary matrix of this subband With Eigenvector (or row) also correspondingly be sorted." broadband " eigenmodes can be defined in the same order eigenmodes set (promptly m broadband eigenmodes comprises m eigenmodes of all subbands) of all subbands of ordering back.Each broadband eigenmodes is relevant with the corresponding eigenvector set of all subbands.Main broadband eigenmodes be after ordering with each matrix The eigenmodes that interior maximum singular value is relevant.

A. handle in the up link space

Ul transmissions can be expressed as by the spatial manipulation that user terminal carries out:

${\underset{\‾}{x}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right),$ K ∈ K equation (14) wherein

Wherein x _Up(k) be the emission vector of the up link of k subband; And

s _Up(k) be the N that is used at k subband _SThe nearly N of the modulated symbol that sends on the individual eigenmodes _S" data " vector of individual nonzero term.

The receiving uplink transmission at access point place can be represented as:

${\underset{\‾}{r}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{x}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$ K ∈ K. equation (15)

$={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$\≈{\underset{\‾}{\overset{^}{H}}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^H\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

Wherein, r _Up(k) be k subband up link receive vector; And

n _Up(k) be the additive white Gaussian noise (AWGN) of k subband.

Equation (15) uses following relation: ${\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right)\≈{\underset{\‾}{\overset{~}{H}}}_{\mathrm{cup}}\left(k\right)$ And ${\underset{\‾}{\overset{^}{H}}}_{\mathrm{cup}}\left(k\right)={\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^H\left(k\right).$

Weighted registration electric-wave filter matrix from the ul transmissions of user terminal M _Ap(k) can be expressed as:

${\underset{\‾}{M}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{=}{\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^H\left(k\right),$ K ∈ K equation (16)

The spatial manipulation (or matched filtering) that receiving uplink is transmitted in the access point place can be expressed as:

${\underset{\‾}{\overset{^}{s}}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^H\left(k\right){\underset{\‾}{r}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^H\left(k\right)({\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)),$ K ∈ K equation (17) wherein

$={\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{\overset{^}{n}}}_{\mathrm{up}}\left(k\right)$

Wherein It is the data vector that user terminal sends on up link s _Up(k) estimation, and

Be to handle the back noise.

B. down link spatial manipulation

The access point place spatial manipulation of downlink transmission can be expressed as:

${\underset{\‾}{x}}_{\mathrm{dn}}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^{*}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right),$ K ∈ K equation (18) wherein

Wherein x _Dn(k) be to send vector, and s _Dn(k) be the data vector of down link.

The receiving downlink transmission at user terminal place can be represented as:

${\underset{\‾}{r}}_{\mathrm{dn}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{dn}}\left(k\right){\underset{\‾}{x}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right)$

$={\underset{\‾}{H}}_{\mathrm{dn}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^{*}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right)$

$={\underset{\‾}{\overset{^}{H}}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^{*}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right)$

$={\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^{*}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^T\left(k\right){\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^{*}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right))$

$={\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^{*}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right)$ K ∈ K equation (19) wherein

Weighted registration electric-wave filter matrix from the downlink transmission of access point M _Ut(k) can be expressed as:

${\underset{\‾}{M}}_{\mathrm{ut}}\left(k\right)={\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^T\left(k\right),$ K ∈ K equation (20) wherein

The spatial manipulation (or matched filtering) that receiving downlink is transmitted in the user terminal place can be expressed as:

$\underset{\‾}{{\hat{s}}_{\mathrm{dn}}\left(k\right)}={\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^T\left(k\right){\underset{\‾}{r}}_{\mathrm{dn}}\left(k\right)$

$={\underset{\‾}{\overset{^}{\mathrm{\Σ}}}}^{-1}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^T\left(k\right)({\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^{*}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right))$ Equation (21)

$={\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{\overset{~}{n}}}_{\mathrm{dn}}\left(k\right)$

Form 1 has been summarized the spatial manipulation at access point and user terminal place that is used for transfer of data and reception.

Form 1

In foregoing description and form 1, correct matrix With

The corresponding transmitting terminal that is applied to access point place and user terminal place.Correct matrix With Can also with other diagonal matrix (for example such as the weighting matrix that is used to obtain the channel counter-rotating W _Dn(k) and W _Up(k)).Yet, corrects matrix and can also be applied to receiving terminal rather than transmitting terminal, and this is within the scope of the present invention.

Fig. 3 is used for the down link of spatial multiplexing modes and the spatial manipulation block diagram of up link according to one embodiment of the invention.

For down link, in the TX of access point 110x spatial processor 120x, data vector s _Dn(k) (wherein k ∈ K) at first multiply by matrix by unit 310

, and further multiply by the correction matrix by unit 312 To obtain the emission vector x _Dn(k).Vector x _Dn(k) (wherein k ∈ K) handles and sends to user terminal 150x by the emission chain in the modulator 122x 314 then on mimo channel.Unit 310 is handled for the downlink transmission implementation space.

At user terminal 150x place, down link signal is handled to obtain to receive vector by the reception chain in the demodulator 154x 354 r _Dn(k) (wherein k ∈ K).In RX spatial processor 160x, receive vector r _Dn(k) (wherein k ∈ K) at first multiply by matrix by unit 356 , and further by the contrary diagonal matrix of unit 358 usefulness

Carry out proportional zoom to obtain

, this vector is a data vector s _Dn(k) estimation.Unit 356 and 358 is that handle down link matched filtering implementation space.

For up link, in the TX of user terminal 150x spatial processor 190x, data vector s _Up(k) (wherein k ∈ K) at first multiply by matrix by unit 360

, multiply by the correction matrix by unit 362 then

To obtain the emission vector x _Up(k).Vector x _Up(k) (wherein k ∈ K) handles and sends to access point 110x by the emission chain in the modulator 154x 364 then on mimo channel.Unit 360 is handled for the uplink data transmission implementation space.

At access point 110x place, uplink signal is handled to obtain to receive vector by the reception chain in the demodulator 122x 324 r _Up(k), k ∈ K wherein.In RX spatial processor 140x, receive vector r _Up(k) (wherein k ∈ K) at first multiply by matrix by unit 326

, and further by the contrary diagonal matrix of unit 328 usefulness

Through proportional zoom to obtain vector , it is a data vector s _Up(k) estimation.Unit 326 and 328 is that handle up link matched filtering implementation space.

3. beam steering

For certain channel condition, preferably only on a broadband eigenmodes, send data-generally be best or main broadband eigenmodes.If to by using on the eigenmodes of main broadband all available launch power when realizing improving performance, this situation may occur to the signal to noise ratio (snr) difference that receives of every other broadband eigenmodes.

Transfer of data on broadband eigenmodes can use beam shaping or beam steering to obtain.For beam shaping, for main broadband eigenmodes (promptly after ordering

Or

First row), the modulated symbol eigenvector

Or Through spatial manipulation, k ∈ K wherein.For beam steering, for main broadband eigenmodes, modulated symbol is generally used " standardization " (or " saturated ") eigenvector

Or Spatial manipulation is carried out in (wherein k ∈ K) set.For clear, beam steering is that following up link describes.

For up link, each eigenvector

Element different amplitudes can be arranged, k ∈ K wherein.Therefore, each subband through the pre-adjustment code element different amplitudes can be arranged, wherein each subband multiply by the eigenvector of subband k through the pre-adjustment code element by the modulated symbol with subband k

Element and obtain.Therefore, line emission every day vector has different amplitudes, and each comprises the pre-adjustment code element of all data subbands of given transmitting antenna.If the transmitting power of each transmitting antenna limited restriction of power amplifier (for example because), then the beam shaping gross power that may not use each antenna to use fully.

For main broadband eigenmodes, beam steering only uses from eigenvector

Phase information, each eigenvector of wherein k ∈ K, and standardization makes that all elements in the eigenvector has equal amplitude.The standardization eigenvector of k subband

Can be expressed as:

${\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right)=[{\mathrm{Ae}}^{{\mathrm{j\θ}}_i\left(k\right)}{\mathrm{Ae}}^{{\mathrm{j\θ}}_2\left(k\right)}...{\mathrm{Ae}}^{{\mathrm{j\θ}}_{N_{\mathrm{ut}}}\left(k\right)}{]}^T$ Equation (22)

Wherein A is constant (for example A=1); And

θ _i(k) be the phase place of k subband of i transmitting antenna, this can be given:

${\mathrm{\θ}}_i\left(k\right)=\∠{\hat{v}}_{\mathrm{ut},1}\left(k\right)={\tan }^{-1}\left(\frac{\mathrm{Im}\left\{{\hat{v}}_{\mathrm{ut},1,i}\left(k\right)\right\}}{\mathrm{Re}\left\{{\hat{v}}_{\mathrm{ut},1}\left(k\right)\right\}}\right)$ Equation (23)

As in the equation (23) vector being shown The phase place of each interior element is from eigenvector

Obtaining (is θ _i(k) from Wherein $\begin{array}{cc}{\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},1}\left(k\right)=[{\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},\mathrm{1,1}}\left(k\right)& {\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},\mathrm{1,2}}\left(k\right)...{\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},1,N_t}\left(k\right){]}^T\end{array}).$

A. uplink beam is handled

User terminal is that the spatial manipulation that the beam steering on the up link carries out can be expressed as:

${\underset{\‾}{\overset{~}{x}}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}{\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)$ Equation (24)

Wherein s _Up(k) be the modulated symbol that on k subband, sends; And

It is the emission vector of k subband of beam steering.

As illustrating in the equation (22), vector is handled in the standardization of each subband

N _UtIndividual unit have identical amplitude, but may the phase place difference.Therefore beam steering generates an emission vector for each subband

N _UtIndividual unit have identical amplitude, but may the phase place difference.

The access point place can be expressed as for the ul transmissions that beam steering receives:

${\underset{\‾}{\tilde{r}}}_{\mathrm{up}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\tilde{x}}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right),$ K ∈ K equation (25) wherein

$={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

Wherein Be beam steering k subband up link receive vector.

Use the capable vector of coupling filter of the ul transmissions of beam steering to be expressed as:

${\underset{\‾}{\overset{~}{m}}}_{\mathrm{ap}}\left(k\right)={\left({\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right)\right)}^H,$ K ∈ K equation (26) wherein

The matched filter vector

Can obtain on following description ground.The spatial manipulation that receives ul transmissions (or matched filtering) that has beam steering at the access point place can be expressed as:

${\underset{\‾}{\overset{^}{s}}}_{\mathrm{up}}\left(k\right)={\widetilde{\mathrm{\λ}}}_{\mathrm{up}}^{-1}\left(k\right){\underset{\‾}{\overset{~}{m}}}_{\mathrm{ap}}\left(k\right){\underset{\‾}{\overset{~}{r}}}_{\mathrm{up}}\left(k\right)$

$={\widetilde{\mathrm{\λ}}}_{\mathrm{up}}^{-1}\left(k\right){\left({\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right)\right)}^H({\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)),$ K ∈ K wherein

$={\underset{\‾}{s}}_{\mathrm{up}}\left(k\right)+{\underset{\‾}{\overset{~}{n}}}_{\mathrm{up}}\left(k\right)$ Equation (27)

Wherein ${\widetilde{\mathrm{\λ}}}_{\mathrm{up}}\left(k\right){=\left({\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right)\right)}^H\left({\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right)\right)$ (promptly

Be

Inner product with and conjugate transpose),

It is the modulated symbol that user terminal sends on up link s _Up(k) estimation, and

Be to handle the back noise.

B. downlink beamforming is handled

On down link, can be expressed as by the spatial manipulation that access point carries out for beam steering:

${\underset{\‾}{\overset{~}{x}}}_{\mathrm{dn}}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}{\underset{\‾}{s}}_{\mathrm{dn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right),$ K ∈ K equation (28) wherein

Wherein Be the standardization eigenvector of k subband, for main broadband eigenmodes, can be based on eigenvector

Generate, as mentioned above.

Use the capable vector of matched filter of the downlink transmission of beam steering

Can be expressed as:

${\underset{\‾}{\overset{~}{m}}}_{\mathrm{ut}}\left(k\right)={\left({\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right)\right)}^H,$ K ∈ K equation (29) wherein

The user terminal place can be expressed as the spatial manipulation (or matched filtering) that receives downlink transmission and carry out:

${\hat{s}}_{\mathrm{dn}}\left(k\right)={\widetilde{\mathrm{\λ}}}_{\mathrm{dn}}^{-1}\left(k\right){\underset{\‾}{\overset{~}{m}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{~}{r}}}_{\mathrm{dn}}\left(k\right)$

$={\widetilde{\mathrm{\λ}}}_{\mathrm{dn}}^{-1}\left(k\right){\left({\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right)\right)}^H({\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right)s_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{dn}}\left(k\right)),$ K ∈ K wherein

$=s_{\mathrm{dn}}\left(k\right)+{\underset{\‾}{\overset{~}{n}}}_{\mathrm{dn}}\left(k\right)$ Equation (30)

Wherein ${\widetilde{\mathrm{\λ}}}_{\mathrm{dn}}\left(k\right)={\left({\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right)\right)}^H\left({\underset{\‾}{H}}_{\mathrm{cdn}}\left(k\right){\underset{\‾}{\overset{~}{u}}}_{\mathrm{ap}}\left(k\right)\right)$ (promptly Be

Inner product with its conjugate transpose).

Beam steering can be considered the special case of spatial manipulation, wherein has only an eigenvector of an eigenmodes to be used for transfer of data, and this eigenvector can be by standardization there to be equal amplitude.

Fig. 4 is according to the down link of the beam steering pattern of one embodiment of the invention and the spatial manipulation block diagram of up link.

For down link, in the TX spatial processor 120y at access point 110y place, modulated symbol s _Dn(k) (wherein k ∈ K) at first multiply by the standardization eigenvector by unit 410

, further multiply by the correction matrix then by unit 412 To obtain the emission vector

Vector (wherein k ∈ K) handles and sends to user terminal 150y by the emission chain in the modulator 122y 414 then on mimo channel.The spatial manipulation of downlink transmission is realized in unit 410 for the beam steering pattern.

At user terminal 150y place, down link signal is handled to obtain to receive vector by the reception chain in the demodulator 154y 454 , k ∈ K wherein.In RX spatial processor 160y, vector is realized receiving in unit 456 With the matched filter vector Inner product, k ∈ K wherein.Inner product result is then by unit 458 usefulness Through proportional zoom to obtain code element s _Dn(k), this is a modulated symbol s _Dn(k) estimation.Unit 456 and 458 spatial manipulation for the matched filtering of beam steering pattern realization down link.

For up link, in the TX spatial processor 190y at user terminal 150y place, modulated symbol s _Up(k) (wherein k ∈ K) at first multiply by the standardization eigenvector by unit 460 , multiply by the correction matrix by unit 462 then To obtain the emission vector

Vector (wherein k ∈ K) handled by the emission chain in the modulator 154y 464 then, and sends to access point 110y on mimo channel.The spatial manipulation of uplink data transmission is realized in unit 460 for the beam steering pattern.

At access point 110y place, uplink signal is handled to obtain to receive vector by the reception chain in demodulator 124y 424

, k ∈ K wherein.In RX spatial processor 140y, vector is realized receiving in unit 426

With the matched filter vector

Inner product, k ∈ K wherein.Inner product result is then by unit 428 usefulness Through proportional zoom to obtain code element , this is a modulated symbol s _Up(k) estimation.Unit 426 and 428 spatial manipulation for the matched filtering of beam steering pattern realization up link.

4. manipulation benchmark

As illustrating in the equation (15),, receive the up link vector at the access point place r _Up(k) (wherein k ∈ K) under noise-free case, equal through The data vector of conversion s _Up(k),

Be Left eigenvector matrix Diagonal matrix through singular value Proportional zoom.As illustrating in equation (17) and (18), owing to channel reciprocal and calibration, matrix The spatial manipulation (matched filtering) that is respectively applied for the spatial manipulation of downlink transmission and receives ul transmissions with its transposition.

Handling benchmark (or handling pilot tone) can be used for obtaining by the user terminal transmission and by access point With Estimation, k ∈ K wherein, and do not need to estimate mimo channel or realize singular value decomposition.Similarly, handling benchmark can be used for obtaining by the access point transmission and by user terminal

With

Estimation.

Handle benchmark and comprise specific OFDM code element (this is called as pilot tone or " P " OFDM code element), it is from all N of user terminal (for up link) _UtThe N at individual antenna or access point place _ApIndividual antenna (for down link) sends.P OFDM code element is handled and is only sent on a broadband eigenmodes by the eigenvector set implementation space with this broadband eigenmodes.

A. up link is handled benchmark

The up link that user terminal sends is handled benchmark and can be expressed as:

${\underset{\‾}{x}}_{\mathrm{up},m}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},m}\left(k\right)p\left(k\right),$ K ∈ K equation (31) wherein

Wherein x _Up, m (k) is the emission vector of k subband of m broadband eigenmodes;

It is the eigenvector of k subband of m broadband eigenmodes; And

P (k) is the pilot modulated code element that will send on k subband.

Eigenvector It is matrix M row, wherein $\begin{array}{cc}{\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}\left(k\right)=[{\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},1}\left(k\right)& {\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},2}\left(k\right)...{\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},N_{\mathrm{ut}}}\left(k\right)].\end{array}$

The up link that the access point place receives is handled benchmark and can be expressed as:

${\underset{\‾}{r}}_{\mathrm{up},m}\left(k\right)={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{x}}_{\mathrm{up},m}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$ K ∈ K equation (32) wherein

$={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},m}\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{\overset{^}{H}}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},m}\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}\left(k\right)\underset{\‾}{\overset{^}{\mathrm{\Σ}}}\left(k\right){\underset{\‾}{\overset{^}{V}}}_{\mathrm{ut}}^H\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},m}\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right))$

$={\underset{\‾}{\overset{^}{u}}}_{\mathrm{ap},m}\left(k\right){\mathrm{\σ}}_m\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

Wherein r _{Up, m}(k) be the vector that receives that the up link of k subband of m broadband eigenmodes is handled benchmark; And

σ _m(k) be the singular value of k subband of m broadband eigenmodes.

Based on the technology of handling the base estimation channel response at following detailed description.

B. down link is handled benchmark

The down link that access point sends is handled benchmark and can be expressed as:

${\underset{\‾}{x}}_{\mathrm{dn},m}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ap}}\left(k\right){\underset{\‾}{\overset{^}{u}}}_{\mathrm{ap},m}^{*}\left(k\right)p\left(k\right),$ K ∈ K equation (33) wherein

Wherein x _{Dn, m}(k) be the emission vector of k subband of m broadband eigenmodes;

It is the eigenvector of k subband of m broadband eigenmodes.

Handle vector It is matrix M row, wherein $\begin{array}{cc}{\underset{\‾}{\overset{^}{U}}}_{\mathrm{ap}}^{*}\left(k\right)=[{\underset{\‾}{\overset{^}{u}}}_{\mathrm{ap},1}^{*}\left(k\right)& {\underset{\‾}{\overset{^}{u}}}_{\mathrm{ap},2}^{*}\left(k\right)...{\underset{\‾}{\overset{^}{u}}}_{\mathrm{ap},N_{\mathrm{ap}}}^{*}\left(k\right)].\end{array}$

Down link is handled benchmark can be used for various purposes by user terminal.For example down link manipulation benchmark allows user terminal to determine which kind of access point has estimate (because access point has the estimation of channel estimating) for mimo channel.Down link is handled the SNR that receives that benchmark can also be used for the estimating down-ward link transmission by user terminal.

C. the manipulation benchmark of beam steering

For the beam steering pattern, the spatial manipulation of transmitting terminal is to use " standardization " eigenvector set of main broadband eigenmodes to realize.The total transfer function that has the standardization eigenvector be different from have the not standardized eigenvector total transfer function (promptly ${\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut},1}^{*}\left(k\right)\≠{\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{~}{v}}}_{\mathrm{ut}}\left(k\right))$ 。The matched filter vector that uses manipulation benchmark that the standardization eigenvector set of all subbands generates can send and be used to derive the beam steering pattern by receiver by transmitter then.

For up link, the manipulation benchmark of beam steering pattern can be expressed as:

${\underset{\‾}{\overset{~}{x}}}_{\mathrm{up},\mathrm{sr}}\left(k\right)={\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut}}\left(k\right)p\left(k\right),$ K ∈ K equation (34) wherein

At access point, the receiving uplink of beam steering pattern is handled benchmark and can be expressed as:

${\underset{\‾}{r}}_{\mathrm{up},\mathrm{sr}}\left(k\right)={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{~}{x}}}_{\mathrm{up},\mathrm{sr}}\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right),$ K ∈ K equation (35) wherein

$={\underset{\‾}{H}}_{\mathrm{up}}\left(k\right){\underset{\‾}{\overset{^}{K}}}_{\mathrm{ut}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut}}\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

$={\underset{\‾}{H}}_{\mathrm{cup}}\left(k\right){\underset{\‾}{\overset{^}{v}}}_{\mathrm{ut}}\left(k\right)p\left(k\right)+{\underset{\‾}{n}}_{\mathrm{up}}\left(k\right)$

In order to obtain to carry out the capable vector of matched filter of ul transmissions with beam steering

, handle the vector that receives of benchmark Rup, sr(k) at first multiply by p ^*(k).The result handles the reference symbol upper integral to form a plurality of receiving then

Estimation.Vector

It then is the conjugate transpose of this estimation.

When operating under the beam steering pattern, user terminal can send a plurality of manipulation reference symbols, for example uses the standardization eigenvector

One or more code elements, use eigenvector for main eigenmodes One or more code elements, and the possible one or more code elements of eigenvector of using other eigenmodes.With The manipulation reference symbol that generates can be used to derive the matched filter vector by access point

With The manipulation reference symbol that generates can be used for obtaining

, this can be used to derive the standardization eigenvector that is used for beam steering on the down link Eigenvector with other eigenmodes Arrive The manipulation reference symbol that generates can be used for obtaining by access point

Arrive

And the singular value of these other eigenmodes.This information can be used to determine whether to use the spatial multiplexing modes or the beam steering pattern of transfer of data then by access point.

For down link, user terminal can be estimated based on calibration back downlink channel response For the beam steering pattern derives the matched filter vector

Particularly, user terminal have from Singular value decomposition

, and can derive the standardization eigenvector User terminal can be used

Multiply by

To obtain

, and based on

Derive

Perhaps, handle benchmark and can use the standardization eigenvector by access point

Send, and this manipulation benchmark can be handled to obtain in the above described manner by user terminal

D. based on the channel estimating of handling benchmark

As illustrating in the equation (32),, receive up link manipulation benchmark (under noise-free case) and be approximately at the access point place Therefore access point can obtain the estimation of uplink channel responses based on the manipulation benchmark that user terminal sends.Various estimation techniques can be used to obtain channel response and estimate.

In one embodiment, in order to obtain Estimation, the manipulation benchmark of m broadband eigenmodes receive vector r _{Up, m}(k) at first multiply by the pilot modulated code element p that is used to handle benchmark ^*(k) complex conjugate.The result handles the reference symbol upper integral to obtain a plurality of the receiving of each broadband eigenmodes then Estimation, this is a m broadband eigenmodes

Through proportional zoom left side eigenvector. N _ApEach based on r _{Up, m}(k) N _ApIn corresponding one and obtain, wherein r _{Up, m}(k) N _ApItem is from access point N _ApIndividual antenna obtains receives code element.Because eigenvector has unit power, singular value σ _m(k) can estimate that this can measure for each subband of each broadband eigenmodes based on handling the receiving power of benchmark.

In another embodiment, least mean-square error (MMSE) technology is used for based on the vector that receives of handling benchmark r _{Up, m}(k) obtain Estimation.Because known pilot modulated symbol p (k), access point can be derived Estimation, make to receive pilot frequency code element (to receiving vector r _{Up, m}(k) realized obtaining after the matched filtering) and the pilot frequency code element that sends between mean square error minimize.Using the MMSE technology for the spatial manipulation at receiver place describes in detail in the U.S. Patent Application Serial Number 09/993087 of public distribution, be entitled as " Multiple-Access Multiple-Input Multiple-Output (MIMO) Communication System ", be filed in November 6 calendar year 2001.

Handling benchmark is that a broadband eigenmodes sends in the period in any given code element, and then can be used to the estimation of an eigenvector of each subband acquisition of this broadband eigenmodes.Therefore, receiving function is that an eigenvector that obtains in the unitary matrix any given code element period is estimated.Because a plurality of eigenvectors of unitary matrix are estimated to obtain on the period in different code element, and because the noise in the transmission path and other deterioration sources, the estimation eigenvector of unitary matrix can not be a quadrature.If the eigenvector of estimating is in the transfer of data spatial manipulation that after this is used on other links, then the interior orthogonality error of the eigenvector of these estimations can cause the cross interference between eigenvector, and this can degrade performance.

The estimation eigenvector of each unitary matrix is forced to mutually orthogonal in one embodiment.The orthogonalization of eigenvector can use the Gram-Schmidt method to obtain, and this has a detailed description in aforesaid Gilbert Strang reference, perhaps can also use additive method to realize.

Can also use based on the other technologies of handling the base estimation channel response, and this within the scope of the present invention.The manipulation base estimation that therefore access point sends based on user terminal With , and do not need estimating uplink channel response or right

Realize singular value decomposition.Owing to have only N _UtIndividual broadband eigenmodes has power,

The matrix of left eigenvector

In fact dimension is N _Ap* N _Ut, and matrix

Dimension is considered to N _Ut* N _Ut

The user terminal place handles the base estimation matrix based on down link With Processing (wherein k ∈ K) can be similar to the above realization of handling benchmark description for up link.

5. channel estimating and spatial manipulation

Fig. 5 is the process 500 specific embodiment flow graphs in access point and user terminal place realization channel estimating and spatial manipulation according to one embodiment of the invention.Process 500 comprises two parts one calibration (frame 510) and normal running (frame 520).

During beginning, access point and user terminal transmit and receive chain according to it and realize calibration with definite difference, and obtain to correct matrix With

, k ∈ K (at frame 512 places) wherein.Calibration only needs to realize once (for example when communication session begins, or when user terminal is opened for the first time).Correct matrix

With After this use respectively by access point and user terminal at transmitting terminal as mentioned above.

During normal running, access point sends MIMO pilot tone (at frame 522 places) on the calibration downlink channel.User terminal receives and handles the MIMO pilot tone, and keeps the calibration downlink channel response and estimate (at frame 524 places).Can illustrate when channel response and estimate performance better (having lacked deterioration) when accurate.Channel response is estimated and can be obtained by the estimation of on average deriving from a plurality of MIMO of receiving pilot transmission accurately.

User terminal decomposes the calibration downlink channel response then and estimates K ∈ K wherein is to obtain diagonal matrix And unitary matrix (at frame 526 places).Matrix Comprise

Left eigenvector, and

Comprise Right eigenvector.Matrix Therefore can be used in transfer of data that receives on the down link and the transfer of data implementation space that on up link, sends by user terminal and handle.

User terminal uses matrix Interior eigenvector will be handled benchmark and send to access point on up link, as equation (31) (at frame 530 places) are shown.Access point receives and handles up link and handles benchmark to obtain diagonal matrix And unitary matrix K ∈ K (at frame 532 places) wherein.Matrix

Comprise

Left eigenvector, and

Comprise

Right eigenvector.Matrix Therefore can be used in transfer of data that receives on the up link and the transfer of data implementation space that on down link, sends by access point and handle.

Matrix

(wherein k ∈ K) handles the estimation of benchmark based on up link and obtains, and this benchmark is to use based on the calibration downlink channel response to estimate and the eigenvector generation of acquisition.Therefore, matrix

Be actually the estimation of estimation.Access point can be handled the benchmark transmission to up link and ask on average to obtain actual matrix Estimate more accurately.

In case user terminal and access point obtain corresponding matrix With

, transfer of data can begin on down link and/or up link.For downlink transmission, access point is used Right eigenvector matrix The code element implementation space is handled, and send to user terminal (frame 540).User terminal receives and uses matrix then

The downlink transmission implementation space is handled described matrix Be The matrix of left eigenvector

Conjugate transpose (at frame 542 places).For uplink data transmission, user terminal is used

The matrix of right eigenvector

The code element implementation space is handled, and send to access point (at frame 550 places).Access point receives and uses matrix then The uplink data transmission implementation space is handled described matrix

Be Left eigenvector matrix

Conjugate transpose (at frame 552 places).

Down link and/or uplink data transmission can continue up to being ended by access point or user terminal.When (not having data to send or reception) when user terminal is idle, MIMO pilot tone and/or manipulation benchmark still can be sent out to allow access point and terminal to keep the latest estimated of corresponding downstream link and uplink channel responses.This can so that transfer of data to begin ground when continuing faster.

For clear, for specific embodiment channel estimating and spatial processing technique have been described here, the downlink channel response after wherein user terminal is estimated to calibrate based on the downlink mimo pilot tone, and realize singular value decomposition.Channel estimating and singular value decomposition can also be realized by access point, and this within the scope of the invention.Generally, because the channel reciprocal of TDD system, channel estimating only need realize at an end of link.

Technology described here can have or not have calibration and use.The realization of calibration can improve channel estimating, and this can improve systematic function.

Technology described here can also be used together with other spatial processing techniques, such as the channel counter-rotating of pouring water and being used for carrying out at the intersubband of each broadband eigenmodes transmit power allocations that is used for carrying out transmit power allocations between the eigenmodes of broadband.The channel counter-rotating and the description in aforesaid United States Patent (USP) sequence number 60/421309 of pouring water.

Channel estimating described here and spatial processing technique can be realized by various means.For example, these technology can realize with hardware, software or their combination.For hardware is realized, be used for realizing that the processing unit of data processing, spatial manipulation and scheduling can realize in following equipment: one or more application-specific integrated circuit (ASIC)s (ASIC), digital signal processor (DSP), digital signal processing appts (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, be designed to carry out other electronic unit of function described here or their combination at the access point place.

For software was realized, channel estimating and spatial processing technique can be realized with the module (for example process, function or the like) of carrying out function described here.Software code can be stored in the memory cell (for example memory cell among Fig. 1 132 or 182), and can be carried out by processor (for example controller 130 or 180).Memory cell can realize in processor or realize that under the externally realization situation, it can be communicatively coupled to processor by variety of way known in the field outside processor.

Here the title that comprises supplies to quote, and helps the specific chapters and sections in location.These titles do not limit its scope of described notion down, and these notions can be applicable to other chapters and sections in the entire description.

The description of above preferred embodiment makes those skilled in the art can make or use the present invention.The various modifications of these embodiment are conspicuous for a person skilled in the art, and Ding Yi General Principle can be applied among other embodiment and not use creativity here.Therefore, the embodiment that the present invention is not limited to illustrate here, and will meet and the principle and the novel feature the most wide in range consistent scope that disclose here.

Claims (56)

1. the method that handle the implementation space in wireless TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

First transmission that processing receives by first link to be to obtain at least one eigenvector, can be used for the transfer of data that receives by first link and the spatial manipulation of the transfer of data that sends by second link; And

Before transmission on second link, be the second transmission implementation space processing with described at least one eigenvector.

2. the method for claim 1 is characterized in that, also comprises:

With described at least one eigenvector the 3rd transmission implementation space that receives by first link is handled to recover the data symbols of the 3rd transmission.

3. the method for claim 1 is characterized in that, described first transmission is the manipulation pilot tone that receives at least one eigenmodes of mimo channel for first link.

4. the method for claim 1 is characterized in that, described first transmission is the MIMO pilot tone that comprises from a plurality of pilot transmission of a plurality of transmitting antennas transmissions, and wherein can be by the receiver sign of MIMO pilot tone from the pilot transmission of each transmitting antenna.

5. method as claimed in claim 4 is characterized in that, described processing first transmission comprises:

Based on the MIMO pilot tone is that first link obtains the channel response estimation; And

Decompose channel response and estimate to can be used for a plurality of eigenvectors of the spatial manipulation of first and second links with acquisition.

6. method as claimed in claim 5 is characterized in that, the channel response estimated service life singular value decomposition of described first link and decomposing.

7. method as claimed in claim 4 is characterized in that, also comprises:

The pilot frequency code element implementation space is handled the manipulation pilot tone of transmitting at least one eigenmodes with the mimo channel that is created on second link with described at least one eigenvector.

8. the method for claim 1 is characterized in that, the spatial manipulation of the eigenvector that the described second transmission process is transmitted on an eigenmodes of the mimo channel of second link.

9. the method for claim 1, it is characterized in that, the spatial manipulation of the standardization eigenvector that transmits on the eigenmodes of described second transmission through the mimo channel of second link, described standardization eigenvector comprises a plurality of elements that have same magnitude.

10. the method for claim 1, it is characterized in that, described first transmission is the manipulation pilot tone for the standardization eigenvector generation of an eigenmodes of the mimo channel of first link, described standardization eigenvector comprises a plurality of elements that have same magnitude, and wherein obtains to can be used for an eigenvector of the spatial manipulation of first and second links.

11. the method for claim 1 is characterized in that, also comprises:

Calibrate first and second links, the channel response estimation that makes the link of winning is the inverse that the channel response of second link is estimated.

12. method as claimed in claim 11 is characterized in that, described calibration comprises

Estimate to obtain the correction factor of first link based on the channel response of first and second links; And

Estimate to obtain the correction factor of second link based on the channel response of first and second links.

13. the equipment in wireless TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

Be used to handle first transmission that receives by first link to obtain the device of at least one eigenvector, described eigenvector can be used for carrying out spatial manipulation to the transfer of data that receives by first link and by the transfer of data that second link sends; And

Be used on second link, transmitting the device that handle the implementation space with described at least one eigenvector to second before the transmission.

14. equipment as claimed in claim 13 is characterized in that, also comprises:

With described at least one eigenvector the 3rd transmission implementation space that receives by first link is handled to recover the device of the 3rd data symbols of transmitting.

15. equipment as claimed in claim 13 is characterized in that, the manipulation pilot tone that receives at least one eigenmodes of the mimo channel that described first transmission is first link.

16. equipment as claimed in claim 13 is characterized in that, described first transmission is the MIMO pilot tone that comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone.

17. equipment as claimed in claim 16 is characterized in that, also comprises:

Based on the MIMO pilot tone is that first link obtains the device that channel response is estimated; And

Decompose channel response and estimate device with a plurality of eigenvectors of the spatial manipulation that obtains to can be used for first and second links.

18. equipment as claimed in claim 13 is characterized in that:

Being used to handle first device that transmits is controller; And

The device that is used for the second transmission implementation space is handled is the emission space processor.

19. equipment as claimed in claim 14 is characterized in that, the device that is used for the 3rd transmission implementation space is handled is for receiving spatial processor.

20. equipment as claimed in claim 18 is characterized in that, the manipulation pilot tone that receives at least one eigenmodes of the mimo channel that described first transmission is first link.

21. equipment as claimed in claim 18 is characterized in that, described first transmission is the MIMO pilot tone that comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone.

22. equipment as claimed in claim 21, it is characterized in that, described controller also is used for obtaining based on the MIMO pilot tone channel response estimation of first link, and decomposes a plurality of eigenvectors that described channel response estimates to can be used for acquisition the spatial manipulation of first and second links.

23. the method that handle the implementation space in wireless TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

The MIMO pilot tone that processing receives by first link is to obtain a plurality of eigenvectors, the spatial manipulation that can be used for the transfer of data that receives by first link and pass through the transfer of data of second link transmission, wherein said MIMO pilot tone comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone;

With described a plurality of eigenvectors the first transfer of data implementation space that receives by first link is handled to recover the data symbols of first transfer of data; And

Before transmission on second link, realize the spatial manipulation of second transfer of data with described a plurality of eigenvectors.

24. method as claimed in claim 23 is characterized in that, also comprises:

With at least one eigenvector the pilot frequency code element implementation space is handled to generate and to handle pilot tone, at least one eigenmodes of the mimo channel of second link, to send.

25. method as claimed in claim 23 is characterized in that, also comprises:

Realize that calibration is to obtain to correct the factor; And

Carrying out proportional zoom with described correction factor pair second transfer of data before the transmission on second link.

26. method as claimed in claim 23 is characterized in that, described TDD MIMO communication system is used orthogonal frequency division multiplex OFDM, and wherein is each implementation space processing of a plurality of subbands.

27. the equipment in wireless time division multiplexing TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

The MIMO pilot tone that processing receives by first link is to obtain the device of a plurality of eigenvectors, the spatial manipulation that described eigenvector can be used for the transfer of data that receives by first link and passes through the transfer of data of second link transmission, wherein said MIMO pilot tone comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone;

With described a plurality of eigenvectors device with the data symbols of recovering first transfer of data is handled in the first transfer of data implementation space that receives by first link; And

Before transmission on second link, realize the device of the spatial manipulation of second transfer of data with described a plurality of eigenvectors.

28. equipment as claimed in claim 27 is characterized in that, also comprises:

The pilot frequency code element implementation space is handled the device of the manipulation pilot tone of transmitting at least one eigenmodes with the mimo channel that is created on second link with at least one eigenvector.

29. equipment as claimed in claim 27 is characterized in that, also comprises:

Realize that calibration is to obtain to correct the device of the factor; And

Before transmission on second link, carry out the device of proportional zoom with described correction factor pair second transfer of data.

30. equipment as claimed in claim 27 is characterized in that:

The device that is used to handle the MIMO pilot tone is a controller;

The device that is used for the first transfer of data implementation space is handled is for receiving spatial processor; And

The device that is used for the second transfer of data implementation space is handled is the emission space processor.

31. equipment as claimed in claim 30, it is characterized in that, described emission space processor also is used for at least one eigenvector the pilot frequency code element implementation space being handled, the manipulation pilot tone of transmitting at least one eigenmodes with the mimo channel that is created on second link.

32. equipment as claimed in claim 30, it is characterized in that, described controller also is used for realizing calibration with the acquisition correction factor, and wherein said emission space processor also is used for carrying out proportional zoom with described correction factor pair second transfer of data before the transmission on second link.

33. the method that handle the implementation space in wireless TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

The manipulation pilot tone that processing at least one eigenmodes by the mimo channel of first link receives, to obtain at least one eigenvector, the spatial manipulation that can be used for the transfer of data that receives by first link and pass through the transfer of data of second link transmission;

With described at least one eigenvector the first transfer of data implementation space that receives by first link is handled; And

With described at least one eigenvector the second transfer of data implementation space is being handled before the transmission on second link.

34. method as claimed in claim 33 is characterized in that, also comprises:

Be created on the MIMO pilot tone of transmitting on second link, wherein said MIMO pilot tone comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein can be by the receiver sign of MIMO pilot tone from the pilot transmission of each transmitting antenna.

35. the equipment in wireless TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

The manipulation pilot tone that processing at least one eigenmodes by the mimo channel of first link receives to be obtaining the device of at least one eigenvector, and described eigenvector can be used for the transfer of data that receives by first link and the spatial manipulation of the transfer of data that sends by second link;

The device of the first transfer of data implementation space that receives by first link being handled with described at least one eigenvector; And

The device of before transmission on second link, the second transfer of data implementation space being handled with described at least one eigenvector.

36. equipment as claimed in claim 35 is characterized in that, also comprises:

Be created on the device of the MIMO pilot tone of transmitting on second link, wherein said MIMO pilot tone comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone.

37. equipment as claimed in claim 35 is characterized in that, comprising:

Being used to handle the device of handling pilot tone is controller;

The device that is used for the first transfer of data implementation space is handled is for receiving spatial processor; And

The device that is used for the second transfer of data implementation space is handled is the emission space processor.

38. equipment as claimed in claim 37, it is characterized in that, described emission space processor is further used for generating the MIMO pilot tone into the transmission on second link, wherein the MIMO pilot tone comprises a plurality of pilot transmission that send from a plurality of transmitting antennas, and wherein from the pilot transmission of each transmitting antenna receiver sign by the MIMO pilot tone.

39. the method that handle the implementation space in wireless TDD multiple-input and multiple-output MIMO OFDM OFMD communication system is characterized in that, comprising:

Each acquisition eigenvector matrix of a plurality of subbands is thought in first transmission that processing receives by first link, wherein obtain a plurality of eigenvector matrixes, the spatial manipulation that can be used for the transfer of data that receives by first link and pass through the transfer of data of second link transmission for described a plurality of subbands; And

On second link, realize the spatial manipulation of second transmission before the transmission with described a plurality of eigenvector matrixes.

40. method as claimed in claim 39 is characterized in that, also comprises:

Based on the channel gain that is associated with eigenvector the eigenvector in each matrix is sorted.

41. method as claimed in claim 40 is characterized in that, described second is transmitted at least one broadband eigenmodes and sends, and with each broadband eigenmodes that eigenvector set in a plurality of matrixes is associated identical order is arranged after ordering.

42. the estimation wireless channel method in TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

Processing is estimated with the channel response that obtains first link by the pilot transmission that first link receives; And

Decompose channel response and estimate to obtain the eigenvector matrix spatial manipulation that can be used for the transfer of data that receives by first link and pass through the transfer of data of second link transmission.

43. the interior method of estimating wireless channel of TDD multiple-input and multiple-output MIMO communication system is characterized in that, comprising:

On at least one eigenmodes of the mimo channel of first link, receive and handle pilot tone; And

The manipulation pilot tone that processing receives to be to obtain at least one eigenvector, can be used for the transfer of data that receives by first link and the spatial manipulation of the transfer of data that sends by second link.

44. method as claimed in claim 43 is characterized in that, described processing comprises:

To the manipulation pilot demodulated that receives to remove owing to be used to generate the modulation that the pilot frequency code element of handling pilot tone causes, and

Manipulation pilot tone after the processing demodulation is to obtain described at least one eigenvector.

45. method as claimed in claim 43 is characterized in that, described at least one eigenvector is based on the least mean-square error technology and obtains.

46. method as claimed in claim 43 is characterized in that, obtains a plurality of eigenvectors and described eigenvector is forced to mutually orthogonal.

47. a method that realizes data processing in the wireless communication system that comprises access point and user terminal is characterized in that, comprising:

Calibrate a plurality of communication links, described link comprises first link and second link between access point and the user terminal, with first link of formation calibration and second link of calibration;

The channel response that obtains first link of calibration on first link of calibration based on one or more pilot tones that send is estimated; And

Decompose described channel response and estimate to can be used for one or more eigenvectors of the spatial manipulation of a plurality of communication links with acquisition.

48. method as claimed in claim 47, described calibration comprises:

Estimate to determine one or more correction factor set based on the channel response of a plurality of communication links; And

First and second links are used described one or more correction factor set to form first and second links of described calibration.

49. method as claimed in claim 47 is characterized in that, also comprises:

Use is handled the transfer of data implementation space on first and second links from described one or more eigenvectors that the channel response estimation of decomposing first link of calibrating obtains.

50. method as claimed in claim 49 is characterized in that, described implementation space is handled and is comprised:

Use described one or more eigenvector on second link, to send and handle benchmark.

51. method as claimed in claim 50 is characterized in that, also comprises:

With described one or more eigenvectors one or more pilot frequency code elements implementation space is handled to generate described manipulation benchmark.

52. an equipment of realizing data processing in the wireless communication system that comprises access point and user terminal is characterized in that, comprising:

Calibrate the device of a plurality of communication links with second link of first link that forms calibration and calibration, described link comprises first link and second link between access point and the user terminal;

Based on one or more on first link of calibration the pilot tones that send obtain the device that the channel response of first link of calibration is estimated; And

Decompose described channel response and estimate device with one or more eigenvectors of the spatial manipulation that obtains to can be used for a plurality of communication links.

53. equipment as claimed in claim 52 is characterized in that, described calibration comprises:

Estimate to determine the device of one or more correction factor set based on the channel response of a plurality of communication links; And

Described one or more correction factors set is applied to the device of first and second links with first and second links that form described calibration.

54. equipment as claimed in claim 52 is characterized in that, also comprises:

Use is handled the transfer of data implementation space on first and second links from described one or more eigenvectors that the channel response estimation of decomposing first link of calibrating obtains.

55. equipment as claimed in claim 54 is characterized in that, described implementation space is handled and is comprised:

Use described one or more eigenvector on second link, to send and handle benchmark.

56. equipment as claimed in claim 55 is characterized in that, also comprises:

With described one or more eigenvectors one or more pilot frequency code elements implementation space is handled to generate described manipulation benchmark.

Images