CN101444054B - Method for obtaining channel reciprocity, transceiver and MIMO communication system - Google Patents

Abstract

The invention provides a method for obtaining channel reciprocity, a transceiver and a Multiple Input Multiple Output (MIMO) communication system, wherein the system includes at least one transceiver A having N antennas and at least one transceiver B having N antennas, the system uses at least a first communication channel H AB from A to B and a second communication channel H BA from B to A. The main idea of the present invention is to utilize a precoding to convert the communication channel H AB and H BA to have identical characteristics. The invention also provides the transceiver and the MIMO communication system which implements the method of the invention.

Description

Obtain method, transceiver and the MIMO communication system of channel reciprocity

Technical field

The present invention relates to radio communication, relate in particular to a kind of method of definite communication channel, a kind of transceiver and a kind of MIMO communication system of implementing this method.

Background technology

All adopt in the wireless communication system of many antennas at transmitter and receiver both sides, promptly in multiple-input and multiple-output (MIMO) system of common name, will use which channel if transmitter is known communication in advance, performance will improve a lot.In this case, can use the precoding of many types of suggestion once, as preequalization [1], the singular value decomposition of utilizing water-filling algorithm (water filling) (singular valuedecomposition) [2] under the Single User MIMO situation, and dirty paper (dirty paper) precoding [3] also suitable to the multiuser MIMO situation.(the prior art reference in [] is listed in the ending of specification)

Originally the MIMO Communication Studies is devoted to improve performance unidirectional in the Radio Link, is different because it is believed that the demand of throughput on the link both direction.But now people more and more pay close attention to respectively on the both direction of Radio Link, and promptly up link and down link carry out MIMO communication.This two-way MIMO that has high robust (robust) and/or high-throughput on both direction communicates by letter very valuable to all high situation of the two-way demand of Radio Link.

Flat fading, no precoding, have a N _TIndividual transmitting antenna and N _RIn the mimo system of individual reception antenna, input/output relation can be expressed as:

y＝Hx+n (1)

In the formula (1), x is the N of emission symbol _T* 1 vector, y, n are respectively the N of receiving symbol _R* 1 vector and noise, H are the N of channel coefficients _R* N _TMatrix.Thereby transmitter symbol N on mimo channel H _THeavy spatial reuse (N _T-fold spatiallymultiplexed), in other words, N _TIndividual stream is parallel the transmission, can make spectrum efficiency increase N in theory _TDoubly.

Linear predictive coding means introduces N in (1) _T* N _SPre-coding matrix W comes the symbol in the vector x is carried out precoding.The dimension N of row _SCan select less than N _T, at this moment x is modified to N _S* 1 dimension.Therefore, N _SIndividual stream is parallel the transmission.The input/output relation of linear precoding MIMO system is expressed as:

y＝HWx+n (2)

Equally, nonlinear precoding can use for example THP precoding (Tomlinson-Harashima precoding) [7] to realize, has wherein defined feedforward and feedback filter matrix.The general introduction of linear/non-linear precoding is referring to [8].

When carrying out nonlinear precoding, in order to select to be applied in pre-coding matrix W optimum in the formula (2), or feedforward and feedback pre-coding matrix, need know the statistical information and the channel H of receiver noise.Also need forward channel quality information so that in the system that uses adaptive coding and modulating (AMC), correct modulation and code rate are set.

In Frequency Division Duplexing (FDD) (FDD) wireless communication system, forward direction is more much bigger than the coherence bandwidth of wireless channel usually with the frequency interval of reverse link.The channel that this means the tranmitting frequency place is unknown, and simultaneously, in order to adopt AMC and/or advanced MIMO precoding technique, the feedback of forward channel information needs receiver to be transmitted into transmitter by backward channel.Thereby this feedback has increased the spectrum efficiency that the signaling consumption of backward channel has reduced system.In addition, when adopting feedback, feedback may be made mistakes, and makes the performance severe exacerbation of MIMO precoding technique.Equally, on the link in the opposite direction (backward channel becomes forward channel at this moment, and vice versa), the MIMO emission faces identical expense and feedback error issues.

Prior art can be divided into two classes.In the first kind, thereby transmitter can be known the mimo channel coefficient and can use the precoding that any kind needs.In second class, receiver has been owing to grasped the information of mimo channel coefficient, can select one and feed back the index (index) of selected precoder from one group of known available precoder.

Prior art in the first kind [4] has been discussed the one way link emission of base station to the user, and this system is made up of a base station and K the single antenna user by M transmitting antenna.Each user estimates M * 1 forward channel, in a cycle of training forward channel coefficient estimation value is modulated to reverse carrier wave and sends to transmitter.Transmitter will obtain all K user's forward channel information.

Belong to [5] and [6] of the first kind, in the MIMO-OFDM system, introduce direct channels feedback (Direct Channel Feed Back, DCFB) method is fed back the forward channel coefficient with the forward channel index modulation to reverse pilot sequence (pilot sequence).Other reverse pilot sequence is used for the backward channel coefficient is estimated.Transmitter is optimized forward emitted according to the forward channel coefficient that obtains again.

Another example PCT/CN2006/001403 of prior art belongs to the second above-mentioned class, and it adopts the code table of available linear predictive coding matrix W.Receiver is estimated mimo channel, select a desirable pre-coding matrix, and the index of this matrix in the code table fed back to transmitter, this desirable pre-coding matrix is in some sense near mimo channel being carried out the desirable pre-coding matrix that characteristic value decomposition obtains.

All use in the system of many antenna emissions at uplink downlink, the problems referred to above are more outstanding.If the up-downgoing channel separates, optimum transmit and receiving filter need all calculate the up-downgoing channel.Therefore, amount of calculation need double.This may make receiver need be equipped with more expensive hardware and/or produce bigger power consumption.

Summary of the invention

The invention provides the solution or the mitigation scheme of prior art problem, design a kind of method of in the MIMO communication system, determining the uplink downlink communication channel characteristics, so that computation complexity is minimum, and minimum to the needs of communication channel characteristics feedback.

A kind of respective receiver and corresponding M IMO communication system are provided simultaneously.

The present invention utilizes linear MIMO precoding to generate equivalent mimo system, even wherein original up-downgoing mimo channel is different fully because of Frequency Division Duplexing (FDD), the up-downgoing mimo channel that present embodiment obtains also identical (multiplying each other with a scalar constant after Hermetian (Hermitian) transposition) is seen Fig. 1.This equivalent system can be used by known senior linearity or nonlinear precoding algorithm in second step then.

The present invention compares with prior art reference [4] [5] [6], and difference is that MIMO communication is two-way and does not need to obtain the mimo channel coefficient of " truly ".So the embodiment of the invention makes up-downgoing mimo channel equivalence (equal) by adopting linear predictive coding, so that the estimated value of uplink channel can be directly used in the optimization of down link.

Optimum transmit filter and optimum reception filter can be as [1] described linear predictive coding matrix, also can be as the combination of [8] described pre-coding matrix and feedback filter, and nonlinear precoding is provided.The calculating of optimum transmit and receiving filter based on channel matrix must be for known to the Design of Filter algorithm, and the optimum filter of transmitting-receiving both sides obtains simultaneously.

If the mimo channel matrix is identical or equivalent in transceiver mode to uplink downlink, can reduce amount of calculation this moment greatly, because can directly obtain the optimum reception filter calculating optimum emission filter time.Therefore, this optimization is only calculated and need be carried out once.

In addition, compare, need not pre-coding matrix is quantified as near a pre-coding matrix in one group of available pre-coding matrix in the code table with the second class prior art.Simultaneously, need not to feed back this precoding index, and this feedback may be made mistakes when sending on wireless channel.

As shown in Figure 2, equate mimo channel, all utilize expression formula (2) to carry out linear predictive coding at uplink downlink in order to generate uplink downlink.

In order to realize the arrangement of Fig. 1, two transceivers all will carry out channel characteristic estimation with the data that receive, and these characteristics of channel preferably multiply by after a scalar number carries out the amplitude adjustment, at the emission pusher side as linear predictive coding.

Provide an example among Fig. 3.Up channel is estimated, obtained scalar as the normalization constant.Then this up channel is carried out the Hermitian transposition, as the linear predictive coding of downlink.Down channel is adopted similar method.Equivalent system is seen Fig. 4, wherein $c_{\mathrm{UL}}{H}_{\mathrm{DL}}{\hat{H}}_{\mathrm{UL}}^{*}={(k\·c_{\mathrm{DL}}{H}_{\mathrm{UL}}{\hat{H}}_{\mathrm{DL}}^{*})}^{*},$ Wherein k is a scalar constant.

After obtaining equivalent mimo system as shown in Figure 4, can estimate this up-downgoing channel that equates.Because special construction of the present invention, at the two ends of link, transmitting channel can be directly by the Hermitian transposition acquisition of receive channel estimated value.Therefore, this equivalent system can adopt senior linearity or the non-linear emission precoding technique that needs channel information.And this equivalent system also can adopt other technology that needs channel information, comprises Adaptive Modulation, adaptive coding and adaptive power control etc.

Description of drawings

Below in conjunction with accompanying drawing the embodiment of the invention is described.

Fig. 1 is the schematic diagram that two different communication channels is converted to the characteristic of two channels much at one of the present invention;

Fig. 2 is the schematic diagram of precoding notion of the present invention;

Fig. 3 is the schematic diagram that the uplink downlink channel is carried out precoding of the present invention;

Fig. 4 is the schematic diagram that carries out the equivalent mimo system that precoding obtains;

The schematic diagram of Fig. 5 for adopting orthogonal reference signal that communication channel is estimated;

Fig. 6 is the use schematic diagram of interior precoder of the present invention.

Embodiment

Suppose that a mimo system is made up of two the transceiver A and the B of a wireless links, each transceiver has N antenna and is used to receive and send messages.The carrier frequency that communication is used on the both direction between A and the B is different, so the communication on the both direction can the phase mutual interference.

To communicate, define 2N orthogonal reference signal for ease of channel estimating, be called pilot tone.This 2N pilot tone is divided into two groups, every group N, with " pilot tone 1 " and " pilot tone 2 " expression.Should be noted that reference signal not necessarily wants quadrature, but the reference signal of quadrature can be improved performance.Preferred reference signal will have low cross correlation at least.During the reference signal quadrature, they can be time quadratures, and/or frequency orthogonal, and/or coded orthogonal.

In order to obtain the characteristic estimating value of mimo channel, used one group of N orthogonal pilot frequency sequence (each transmitting antenna uses) in the pilot tone 2, referring to Fig. 5.These pilot tones are used for estimating H respectively _ABAnd H _BA" truly " mimo channel of expression.

When the ends A of communication link and B utilize one group of orthogonal reference signal in the pilot tone 2 characteristic of these channels is estimated as respectively

With

The time, obtain the normalization constant and come the compensating for path loss.For example, can be calculated as follows:

$c_A=\frac{1}{{\left|\right|{\hat{H}}_{\mathrm{BA}}\left|\right|}_F}$ With

$c_B=\frac{1}{{\left|\right|{\hat{H}}_{\mathrm{AB}}\left|\right|}_F}$

‖ ‖ wherein _FFrobenius norm for this channel.Normalization also is optionally, can significantly improve systematic function.

Respectively press following formula selection wire pre-coding matrix at A and B place this moment:

$W_A=c_A{\hat{H}}_{\mathrm{BA}}^{*}$

$W_B=c_B{\hat{H}}_{\mathrm{AB}}^{*}$

Use these pre-coding matrixes by following formula respectively at A and B place then and carry out precoding, to obtain the channel of equivalence:

$G_{\mathrm{AB}}=H_{\mathrm{AB}}{W}_A=c_A{H}_{\mathrm{AB}}{\hat{H}}_{\mathrm{BA}}^{*}$

$G_{\mathrm{BA}}=H_{\mathrm{BA}}{W}_B=c_B{H}_{\mathrm{BA}}{\hat{H}}_{\mathrm{AB}}^{*}$

Should be noted that, except top used Hermitian transposed matrix channel estimation value

With

In addition, common transposed matrix

With

Also be available.In fact, to conventional channel estimated value transposed matrix

With

, the Hermitian transposed matrix gets only that the complex conjugate of each element in these channel estimation values obtains.

If adopt the channel estimating of " pilot tone 2 " to have good performance, this moment these two mimo channels just each other the complex-conjugate transpose matrix multiply by a scalar constant, perhaps, if use be

With

Then common each other transposed matrix multiply by a scalar constant.

$G_{\mathrm{AB}}=k\·G_{\mathrm{BA}}^{*}$

Second group of orthogonal pilot frequency sequence of " pilot tone 1 " expression is all with that organizes member's quadrature in pilot frequency sequence " pilot tone 2 " among Fig. 5." pilot tone 1 " is used for respectively each side at Radio Link to G _ABAnd G _BAEstimate.Estimated value is used

With

Expression.The same with the front, quadrature is not necessary, but can improve performance.

Pre-coding matrix W _AAnd W _BRenewal must follow channel dynamics.Each pilot tone may be separated (orthogonallyseparated in time and/or frequency) by quadrature in the ofdm system on time and/or frequency.If this method is applied to ofdm system, above-mentioned matrix must be pressed channel under the situation of frequency separation coherence bandwidth calculates, and must press the coherence time of channel and calculate under the situation of time-division.

Now, transmitter can use senior precoding technique, as adopt nonlinear precodings such as the characteristic value decomposition of power water filling on the eigenmodes or THP (Tomlinson-Harashima precoding) precoding, because at the A place, carrying out the required A of this precoding is according to the Hermitian transposition of B to A channel characteristic estimation value, promptly to the channel of B ${\hat{G}}_{\mathrm{AB}}=k\·{\hat{G}}_{\mathrm{BA}}^{*}$ Obtain, wherein k is a constant.

Equally, at the B place, B is according to the Hermitian transposition of A to the B channel estimation value, promptly to the channel of A ${\hat{G}}_{\mathrm{BA}}=k^{\′}\·{\hat{G}}_{\mathrm{AB}}^{*}$ Obtain, wherein k ' is a constant.

The input/output relation of linear predictive coding mimo system is (being got by (2))

y＝HWx+n

When also using linear filter R to estimate the emission vector x, receiver has

$\hat{x}=\mathrm{RHWx}+n$

In provided the optimization method different with W to R.For example, use in the example, singular value decomposition is used for linear predictive coding, decompose square mimo channel matrix H by following formula at one

H＝UΛV ^*

Wherein the row among U and the V comprise quadrature and normalized left and right sides singular vector (left and right singular vectors) respectively, and singular value is positioned on the diagonal of diagonal matrix Λ.If the emission pre-coding matrix selects for use W=V and receiving filter to select R=U for use ^*, equivalent system can be

$\hat{x}=U^{*}\mathrm{U\Λ}{V}^{*}\mathrm{Vx}+U^{*}n=\mathrm{\Λx}+U^{*}n$

As can be seen, because diagonal matrix Λ, the emission in the vector x is intersymbol to influence each other at the estimated value vector

In be completely removed (anti-interference).This needs transmitter and receiver to know that the situation of this channel is so that can calculate linear predictive coding matrix R and W in using.In addition, if up-downgoing channel difference then must be carried out channel singular value decomposition H=U Λ V respectively to the up-downgoing channel ^*, and this computational costs costliness.In the present invention, this calculating only need be carried out once, because up-downgoing mimo channel matrix equates.

On the other hand, in the present invention, the linear predictive coding of emission symbol must be undertaken so that generate identical channel at uplink downlink by Fig. 2.As shown in Figure 6, carry out " outside " after the precoding, for example adopt above-mentioned singular value decomposition, the symbol after the precoding must through " interior " precoder once more precoding to generate identical up-downgoing channel.Therefore shortcoming compared with prior art of the present invention is to need " interior " precoding structure.But the multiplicative complexity of matrix that interior precoding is relevant and vector sees Table the comparison in 1 far below singular value decomposition.

The complexity of singular value decomposition increases with cube value of antenna amount N, and matrix-vector multiplication only increases with its square value.Therefore, replace singular value decomposition with matrix-vector multiplication and greatly reduce computation complexity.

Table 1 carries out singular value decomposition and the required amount of calculation of matrix-vector multiplication, and wherein N is the size of square formation.

Computing	Operand singular value decomposition [9]	Operand matrix-vector multiplication
			The multiplication number of times	18N 3	N 2
The addition number of times	9N 3	N 2-1

Table 1

List of references

A.Scaglione，P.Stoica，S.Barbarossa，G.B.Giannakis，and?H.Sampath，“Optimal?designs?for?space-time?linear?precoders?and?decoders”，IEEETrans.Signal?Processing，vol.50，pp.1987-2006，July?1999.

I.E.Telatar，“Capacity?of?Multi-Antenna?Gaussian?channels”，EuropeanTransactions?on?Telecommunications，vol.10，no.6，pp.585-595，Nov./Dec.1999.

U.Erez?and?S.ten?Brink，“A?Close-to-Capacity?Dirty?Paper?PrecodingScheme”，IEEE?transactions?on?information?theory，vol.51，No.10，October2005，p.3417-3432.

T.Marzetta?and?B.Hochwald，“Fast?transfer?of?channel?state?information?inwireless?systems”，submitted?to?IEEE?Transactions?on?Signal?Processing，2004，http://mars.bell-labs.com/papers/channel_estimation/FDDvsTDD.pdf

IEEE802.16e-04/422，“Improvements?to?the?uplink?channel?sounding?forOFDMA”，IEEE?BWA?WG，2004-04-11，http://www.ieee802.org/16/tge/contrib/C80216e-04_422.pdf

3GPP?R1-050516，“Additional?details?on?DCFB?for?obtaining?MIMOchannel?information?at?Node?B”，Motorola，Seoul，Korea，November?2005.

R.F.H.Fischer，C.Windpassinger，A.Lampe?and?J.B.Huber，“Space-time?transmission?using?Tomlinson-Harashima?precoding”，Proc.of?4?ITGConference?on?Source?and?Channel?Coding，pp.139-147，January?2002.

O.Simeone，Y-Bar-Ness?and?U.Spagnolini，“Linear?and?nonlinearpreequalization/equalization?for?MIMO?Systems?with?long?term?channel?stateinformation?at?the?transmitter”，IEEE?Transactions?on?wireless?communications，p.373-378，Vol3.No.2，March?2004.

P.Vandenameele，L.van?Der?Perre，B.Gyselinckx，“An?SDMA?algorithmfor?High-Speed?WLAN?Performance?and?complexity”，Global?TelecommunicationsConference，pp.189-194，November?1998.

Claims (16)

1. A method for obtaining communication channel reciprocity in a MIMO communication system, wherein the system includes at least one multi-antenna transceiver A and at least one multi-antenna transceiver B, and the system uses at least one multi-antenna transceiver A first communication channel H _AB from antenna transceiver A to multi-antenna transceiver B and a second communication channel H _BA from multi-antenna transceiver B to multi-antenna transceiver A communicate between A and B, characterized in that The methods described include: The N antennas of the multi-antenna transceiver A transmit a first group of N reference signals to the multi-antenna transceiver B through the first communication channel _HAB , wherein each antenna transmits a reference signal; The N antennas of the multi-antenna transceiver B receive the first group of N reference signals transmitted by the multi-antenna transceiver A, wherein each antenna receives a reference signal; The multi-antenna transceiver B obtains the characteristic estimation value of the first communication channel H _AB according to the received first group of N reference signals

The N antennas of the multi-antenna transceiver B transmit a first group of N reference signals to the multi-antenna transceiver A through the second communication channel _HBA , wherein each antenna transmits a reference signal; The N antennas of the multi-antenna transceiver A receive the first group of N reference signals transmitted by the multi-antenna transceiver B, where each antenna receives one reference signal; The multi-antenna transceiver A obtains the characteristic estimation value of the second communication channel H _BA according to the received first group of N reference signals

use

The transposition matrix is used as the precoding matrix of the first communication channel H _AB , the first communication channel H _AB is precoded, and the channel G _AB is generated, and the transposition matrix is a hermitian transposition matrix or a common transposition matrix; use

The transposition matrix of is used as the precoding matrix of the second communication channel H _BA to precode the second communication channel H _BA to generate the channel G _BA , and the transposition matrix is a hermitian transposition matrix or an ordinary transposition matrix. 2. The method according to claim 1, further comprising: The multi-antenna transceiver A sends a second group of N reference signals to the multi-antenna transceiver B through the channel _GAB , wherein each antenna transmits a reference signal; The N antennas of the multi-antenna transceiver B receive the second group of N reference signals transmitted by the multi-antenna transceiver A, wherein each antenna receives a reference signal; The multi-antenna transceiver B obtains the first characteristic estimation value of the channel G _AB according to the received second group of N reference signals

The multi-antenna transceiver B sends a second group of N reference signals to the multi-antenna transceiver A through the channel _GBA , wherein each antenna transmits a reference signal; The N antennas of the multi-antenna transceiver A receive the second group of N reference signals transmitted by the multi-antenna transceiver B, where each antenna receives one reference signal; The multi-antenna transceiver A obtains the second characteristic estimation value of the channel G _BA according to the received second group of N reference signals

The second characteristic estimate of the channel G _BA will be obtained The transposed matrix of is multiplied by a constant as the third characteristic estimate of the channel G _AB

The first characteristic estimate of the received channel G _AB will be obtained

The transposed matrix of is multiplied by a constant as the estimated value of the fourth characteristic of the transmit channel G _BA

3. The method according to claim 2, wherein the first property estimate

second characteristic estimate

Estimated value of the third characteristic

and the fourth property estimate It is used in any one of transceivers A and B to improve communication characteristics. 4. The method according to claim 3, wherein the first characteristic estimate

second characteristic estimate

Estimated value of the third characteristic

and the fourth property estimate

for at least one of the following: adaptive modulation, adaptive encoding, Adaptive Power Control. 5. The method of claim 4, wherein the adaptive power control is performed using a water injection method. 6. The method according to claim 1, wherein the channel characteristic estimation value

and

Calculated by normalization to compensate for path loss. 7. The method according to claim 2, wherein the first set of reference signals and the second set of reference signals satisfy at least one of the following: Orthogonal in time; Orthogonal in frequency; Orthogonal in encoding. 8. The method according to claim 2, wherein the first property estimate

second characteristic estimate Estimated value of the third characteristic

and the fourth property estimate

for precoding. 9. The method according to any one of claims 1 to 8, wherein the transpose matrix is a Hermitian transpose matrix. 10. The method according to claim 9, wherein the channel G _AB and the channel G _BA are complex conjugate transpose matrices multiplied by a scalar constant. 11. A method according to any one of claims 1 to 8, wherein The adoption The transposed matrix of as the precoding matrix of the first communication channel H _AB includes: calculate The complex conjugate matrix of the transpose matrix, the calculated matrix is used as the precoding matrix of the first communication channel H _AB ; The adoption The transposed matrix of as the precoding matrix of the second communication channel H _BA includes: calculate

The complex conjugate matrix of the transpose matrix of is used as the precoding matrix of the second communication channel H _BA . 12. The method according to any one of claims 1 to 8, wherein the first set of reference signals are orthogonal to each other. 13. A method according to any one of claims 2 to 8, wherein The second set of reference signals are orthogonal to each other. 14. A multiple-input multiple-output MIMO communication system, characterized in that it comprises at least one multi-antenna transceiver A and at least one multi-antenna transceiver B, said system at least uses the first communication from transceiver A to transceiver B channel H _AB and a second communication channel H _BA from transceiver B to transceiver A, wherein, The multi-antenna transceiver A is used to transmit a first group of N reference signals to the transceiver B through the first communication channel _HAB using N antennas, wherein each antenna transmits a reference signal; using the N antennas Receive the first group of N reference signals transmitted by the multi-antenna transceiver B, wherein each antenna receives a reference signal; obtain the characteristic estimation value of the second communication channel H _BA according to the received N reference signals

use

The transposition matrix is used as the precoding matrix of the first communication channel H _AB , the first communication channel H _AB is precoded, and the channel G _AB is generated, and the transposition matrix is a hermitian transposition matrix or a common transposition matrix; The multi-antenna transceiver B is used to transmit a first set of N reference signals to the transceiver A through the second communication channel H _BA using N antennas, wherein each antenna transmits a reference signal; using the N antennas Receive the first group of N reference signals transmitted by the multi-antenna transceiver A, wherein each antenna receives a reference signal; obtain the characteristic estimation value of the first communication channel H _AB according to the received N reference signals

use

The transposition matrix of is used as the precoding matrix of the second communication channel H _BA to precode the second communication channel H _BA to generate the channel G _BA , and the transposition matrix is a hermitian transposition matrix or an ordinary transposition matrix. 15. The MIMO communication system according to claim 14, wherein, The multi-antenna transceiver A is further configured to send a second group of N reference signals to the multi-antenna transceiver B through the channel G _AB , wherein each antenna transmits a reference signal, and uses the N antennas to receive all The second group of N reference signals transmitted by the multi-antenna transceiver B, wherein each antenna receives a reference signal; obtain the second characteristic estimation value of the channel G _BA according to the received N reference signals

A second estimate of the channel G _BA will be obtained The transposed matrix of is multiplied by a constant as the third characteristic estimate of the channel G _AB

The multi-antenna transceiver B is further configured to send a second group of reference signals to the multi-antenna transceiver A through the channel G _BA , where each antenna transmits a reference signal, and uses the N antennas to receive the multiple The second group of N reference signals transmitted by the antenna transceiver A, wherein each antenna receives a reference signal; obtain the first characteristic estimation value of the channel G _AB according to the received N reference signals

The first characteristic estimate of the channel G _AB will be obtained

The transposed matrix of is multiplied by a constant as the estimated value of the fourth characteristic of the channel G _BA

16. The MIMO communication system according to claim 15, wherein the system is used to use the first characteristic estimation value

second characteristic estimate

Estimated value of the third characteristic and the fourth property estimate

For any one of the multi-antenna transceivers A and B to improve communication characteristics.

Images