CN102710365A - Channel statistical information-based precoding method for multi-cell cooperation system - Google Patents

Abstract

A channel statistical information-based precoding method for a multi-cell cooperation system comprises that (1) multi-cell base stations participating in cooperation obtain second-order statistical information of all user channels of the system; (2) base stations cooperate with each other, and precoding vectors of all users of the system are calculated by using the obtained second-order statistical information of channels according to the channel statistical information-based precoding method; and (3) base stations multiply data which are required to be sent by each user by the corresponding precoding vector of the user and send precoded data. The method is characterized in that real-time information is replaced with the statistical information of channels and precoding vectors are calculated by using the channel statistical information-based precoding method. In the system, when Doppler spread of channels is small, and when precoding vectors obtained on the basis of the second-order statistical information of channels are compared with precoding vectors obtained on the basis of the real-time information of channels, performance loss at operating points is small.

Description

Be applied to the method for precoding based on channel statistical information of multi-cell cooperating system

Technical field

The present invention relates to a kind of precoding algorithm that is applied to the multi-cell cooperating system.

Background technology

In Long Term Evolution plan (LTE), many input and output (MIMO) techniques make use many receptions, transmitting antenna is that system provides branch collection and spatial multiplexing gain through do precoding at transmitting terminal.But in many cell environments, owing to there be interference and the interference between the dispatched users between the sub-district, the capacity gain of system can reduce greatly, especially is in the user of cell edge, and its performance loss is serious.To this problem, 3GPP has proposed the scheme of multi-plot joint transmission (CoMP), promptly sends data through a plurality of base station collaborations to the user, comes the effectively interference between the control sub-district, the performance of elevator system.The downlink transfer of CoMP system considers that mainly two kinds of cooperation modes are joint transmission (JP) and cooperative scheduling/beam forming (CS/CB).Under joint transmission (JP) pattern, same user's data is shared by a plurality of base stations of participating in cooperation and is united transmission, and its equivalence is a single base stations and multiuser system; In united beam moulding (CB) pattern, each base station corresponding with service is in a user, and data are not shared between the base station.More than two kinds of pattern common ground be that transmitting terminal need adopt appropriate pre-coding scheme to reduce the interference between the user, the Hoisting System performance based on the channel information of receiving terminal feedback.

At present, the linear predictive coding scheme that is directed to multi-user system mainly contains two kinds, and a kind of maximization letter of being based on leaks and makes an uproar than the pre-coding scheme of (SLNR) thought, and a kind of is block diagonalization (BD) pre-coding scheme.Maximization SLNR pre-coding scheme is taken noise and targeted customer's leakage power into account, and its purpose is to make the noise that targeted customer's available power receives with it and the ratio of interference power to reach maximum.Because SLNR is directly connected to the throughput of Block Error Rate performance and system, the precoding that therefore maximizes SLNR usually is the effective scheme that obtains ideal performance.And the main thought of BD pre-coding scheme is to reduce the interference that the targeted customer causes other users, does not consider the Signal to Interference plus Noise Ratio that receiving terminal is actual, is a kind of " sharp he " scheme.Its effect receives the characteristic of channel, the influence of antenna number and number of users.

Summary of the invention

Technical problem: the method for precoding based on channel statistical information of the multi-cell cooperating system that is applied to that the present invention proposes; When being mainly solution pre existing encoding scheme and in actual multi-cell cooperating system, using, need the problem of a large amount of feedback informations, provided a kind of near ideal performance; Be easy to realize; Low to the feedback information accuracy requirement, the method for precoding of strong robustness has reduced the complexity of feedback of channel information in the multi-cell cooperating system effectively.

Technical scheme: the method for precoding based on channel statistical information that is applied to the multi-cell cooperating system of the present invention may further comprise the steps:

1) multi-cell base station of participation cooperation obtains each user's of system channel second-order statistics information;

2) cooperation each other between the base station, the channel second-order statistics information that utilize to obtain, according to following formula:

${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_i+N_{0}I)}^{-1}{R}_k$ Or ${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_{k,i}+N_{0}I)}^{-1}{R}_{k,k}$ K=1,2,3, K K

Calculate the precoding vector w of user k in the system _k

Wherein N representes to participate in the system total number of base of cooperation, and K representes intrasystem total number of users, R _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix; () ^-1Expression is to matrix inversion;

3) base station data that user k need be sent and this user's precoding vector w _kCorrespondence multiplies each other, and sends the data after the precoding.

Wherein, the multi-cell base station of the said participation cooperation of step 1) obtains each user's of system channel second-order statistics information, and its step is following:

11) user k sends detectable signal SRS to base station n, and the place obtains channel matrix H in the base station _{N, k}, its size is n _r* n _t, k=1,2,3 ... K, n=1,2,3...N, wherein, K representes intrasystem total number of users, N representes the total number of base of participating in cooperating, H _{N, k}Expression base station n is to the space channel of user k, n _rExpression reception antenna number, n _tThe expression number of transmit antennas;

12) if the cooperation mode of joint transmission (JP) is adopted in the minizone, a plurality of base stations equivalence of then participating in cooperation is the big base station of antenna more than, and multi-cell cooperating system equivalence this moment is a single base stations and multiuser system, and the equivalent space channel of user k is expressed as:

H _k＝[H _1k H _2k?L?H _Nk]

If the cooperation mode of united beam moulding (CB) is adopted in cooperation between many sub-districts, then this step is omitted;

13) design time length L through statistical average, calculates channel second-order statistics metric matrix,

If the cooperation mode of minizone employing JP then

$R_k=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_k}^H\left(i\right)H_k\left(i\right)$

If the cooperation mode of minizone employing CB then

$R_{k,k}=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_{k,k}}^H\left(i\right)H_{k,k}\left(i\right)$

Wherein, R _kThe space channel second-order statistics metric matrix of expression user k, R _{K, k}Expression base station k is to the space channel second-order statistics metric matrix of user k, H _k(i), H _{K, k}(i) expression i moment spatial channel matrix, i representes some observing time.

Wherein, step 2) cooperation each other between the said base station utilizes the channel second-order statistics information that obtains to calculate each user's of system precoding vector according to the precoding vector computing formula, and its step is following:

21) if the cooperation mode of JP, then compute matrix A=R are adopted in the minizone _kWith

If the cooperation mode of CB, then compute matrix A=R are adopted in the minizone _{K, k}And matrix

R wherein _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix;

22) to matrix B ^-1A carries out characteristic root and decomposes, and extracts eigenvalue of maximum characteristic of correspondence vector, promptly obtains the precoding vector of user k.B wherein ^-1The inverse matrix of representing matrix B.

Beneficial effect: the method for precoding based on channel statistical information that is applied to the multi-cell cooperating system provided by the present invention is compared based on the precoding technique of Real-time Channel information with existing, and its beneficial effect is:

1, method for precoding provided by the present invention the time become slowly and can obtain the performance similar in the channel with prior art.Because the service object of multi-cell cooperating system is the user who is in cell edge, its characteristic of channel time changes slowly, and therefore method for precoding provided by the invention can be advantageously applied in the multi-cell cooperating system.

2, method for precoding provided by the present invention only needs the relevant information of feedback space channel, wants easy to be many compared with feedback Real-time Channel information.In the multi-cell cooperating system, existing technological feedback overhead is very big, uses method for precoding provided by the present invention can reduce the Feedback Design requirement to the multi-cell cooperating system greatly.

3, method for precoding provided by the present invention has provided the closed solutions of calculating precoding vector, with existing to compare computation complexity based on the precoding technique of Real-time Channel information suitable and performance is approaching.

Description of drawings

Fig. 1 is a practical implementation flow chart of the present invention;

Fig. 2,3 are the performance of the present invention in actual emulation and the performance comparison of desirable pre-coding scheme;

Explanation of symbols:

JP: joint transmission multi-cell cooperating pattern

CB: united beam moulding multi-cell cooperating pattern

K: the total number of users in the system

N: the total number of base in the system

N ₀: the power of receiving terminal additive white Gaussian noise

L: timing statistics length when calculating the second-order statistics metric matrix

H _{N, k}: base station n is to the space channel of user k

H _k: in the JP pattern, the equivalent space channel of user k

k base station other users' beyond k user in the system space channel

I: unit matrix

R _k: in the JP pattern, the space channel second-order statistics metric matrix of user k

R _{N, k}: in the CB pattern, base station n is to the space channel second-order statistics metric matrix of user k

n _r: the reception antenna number

n _t: number of transmit antennas

s _k: k the data that the user need send

n _k: k user's additive white Gaussian noise vector

w _{I, k}: the precoding vector of k user on i base station

w _k: k user's equivalent precoding vector

|| ||: the Frobenius norm

E{}: the expectation of element

Embodiment

The present invention is a kind of method for precoding based on channel statistical information that is applied to the multi-cell cooperating system, respectively JP and two kinds of collaboration modes of CB is provided principle and computational methods based on the precoding vector of channel second-order statistics information at present.

(1) the precoding vector computational methods under the JP collaboration mode

For the system of a JP collaboration mode, suppose that N base station communicating by letter with K user simultaneously, there is n each base station _tThe root transmitting antenna, each user has n _rThe root reception antenna, the user carries out the single layer data flow transmission.Then k user's reception data can be expressed as

$y_k=\left(\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{H}_{\mathrm{ik}}{w}_{\mathrm{ik}}\right)s_k+\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{H}_{\mathrm{ik}}\left(\underset{j=1,j\≠k}{\overset{K}{\mathrm{\Σ}}}{w}_{\mathrm{ij}}{s}_j\right)+n_k$

Wherein, s _kThe transmission data of representing k user, w _Ik∈ (n _t* 1) precoding vector of k user of expression on i base station, H _IkRepresent the spatial channel matrix of k user, n to i base station _kThe additive white Gaussian noise vector of representing k user.

Note H _k=[H _1kH _2kL H _Nk], $w_k=\begin{array}{c}\left[\begin{array}{c}{w}_{1k}\\ w_{2k}\\ M\\ w_{\mathrm{Nk}}\end{array}\right]\end{array},$

Then the signal power that receives of k user does || H _kw _k|| ², wherein || || the F norm of representing matrix.Therefore the signal power of k user's leakage can obtain k user for

SLNR expression formula does

${\mathrm{SLNR}}_k=\frac{{w_k}^H{H_k}^H{H}_k{w}_k}{{w_k}^H\left(\underset{i=1,i\≠k}{\overset{K}{\mathrm{\Σ}}}{H_i}^H{H}_i\right)w_k+N_0}$

Have by Jensen's inequality

$E\left\{{\mathrm{SLNR}}_k\right\}=E\left\{\frac{{w_k}^H{H_k}^H{H}_k{w}_k}{{w_k}^H\left(\underset{i=1,i\≠k}{\overset{K}{\mathrm{\Σ}}}{H_i}^H{H}_i\right)w_k+N_0}\right\}$

$\≥\frac{{w_k}^{H}E\left\{{H_k}^H{H}_k\right\}{w}_k}{{w_k}^H\left(\underset{i=1,i\≠k}{\overset{K}{\mathrm{\Σ}}}E\right\{{H_i}^H{H}_i\left\}\right)w_k+N_0}$

Make inequality the right maximum, only need w _kGet ${\left(\underset{i=1,i\≠k}{\overset{K}{\mathrm{\Σ}}}E\right\{{H_i}^H{H}_i\}+N_{0}I)}^{-1}\left(E\right\{{H_k}^H{H}_k\left\}\right)$ Eigenvalue of maximum characteristic of correspondence vector, wherein E{} representes the expectation of element.

(2) the precoding vector computational methods under the CB collaboration mode

Different with the JP collaboration mode, under the CB collaboration mode, each base station corresponding with service is in a user.For the system of a CB collaboration mode, suppose that N base station communicating by letter with N user simultaneously, there is n each base station _tThe root transmitting antenna, each user has n _rThe root reception antenna, the user carries out the single layer data flow transmission.Then k user power of receiving signal does || H _Kkw _k|| ², the power of k user's leakage signal does

Wherein || || the F norm of representing matrix.

Wherein ${\stackrel{\‾}{H}}_{\mathrm{Kk}}={\left[\begin{array}{cccccc}{H^H}_{k1}& L& {H^H}_{k(k-1)}& {H^H}_{k(k+1)}& L& {H^H}_{\mathrm{KN}}\end{array}\right]}^H,$ w _k∈ (n _t* 1) k user's of expression precoding vector, H _IkRepresent the spatial channel matrix of k user to i base station.

The SLNR expression formula that therefore can obtain k user does

${\mathrm{SLNR}}_k=\frac{{w^H}_k{H^H}_{\mathrm{kk}}{H}_{\mathrm{kk}}{w}_k}{{w^H}_k{{\stackrel{\‾}{H}}^H}_{\mathrm{kk}}{\stackrel{\‾}{H}}_{\mathrm{kk}}{w}_k+N_0}$

Have by Jensen's inequality

$E\left\{{\mathrm{SLNR}}_k\right\}=E\left\{\frac{{w^H}_k{H^H}_{\mathrm{kk}}{H}_{\mathrm{kk}}{w}_k}{{w^H}_k{{\stackrel{\‾}{H}}^H}_{\mathrm{kk}}{\stackrel{\‾}{H}}_{\mathrm{kk}}{w}_k+N_0}\right\}$

$\≥\frac{{w_k}^{H}E\left\{{H^H}_{\mathrm{kk}}{H}_{\mathrm{kk}}\right\}{w}_k}{{w_k}^{H}E\left\{{{\stackrel{\‾}{H}}^H}_{\mathrm{kk}}{\stackrel{\‾}{H}}_{\mathrm{kk}}\right\}{w}_k+N_0}$

Make inequality the right maximum, only need w _kGet ${\left(E\right\{{{\stackrel{\‾}{H}}^H}_{\mathrm{Kk}}{\stackrel{\‾}{H}}_{\mathrm{Kk}}\}+N_{0}I)}^{-1}\left(E\right\{{H^H}_{\mathrm{Kk}}{H}_{\mathrm{Kk}}\left\}\right)$ Eigenvalue of maximum characteristic of correspondence vector.

Below in conjunction with accompanying drawing, embodiments of the invention are elaborated.

Fig. 1 is that the present invention is at multi-cell cooperating system implementation flow chart.Present embodiment is a kind of method for precoding based on channel statistical information that is applied to the multi-cell cooperating system, and its complete step is following:

(1) multi-cell base station of participation cooperation obtains each user's of system channel second-order statistics information;

(2) cooperation each other between the base station, the channel second-order statistics information that utilize to obtain, according to following formula:

${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_i+N_{0}I)}^{-1}{R}_k$ Or ${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_{k,i}+N_{0}I)}^{-1}{R}_{k,k}$ K=1,2,3, K K

Calculate the precoding vector w of user k in the system _k

Wherein N representes to participate in the system total number of base of cooperation, and K representes intrasystem total number of users, R _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix; () ^-1Expression is to matrix inversion.

(3) base station data that user k need be sent and this user's precoding vector w _kCorrespondence multiplies each other, and sends the data after the precoding.

Said method, wherein the multi-cell base station of the said participation cooperation of step (1) obtains each user's of system channel second-order statistics information, and its step is following:

11) user k (k=1,2 ..., K) to base station n (n=1,2,3 ..., N) send detectable signal (SRS), the place, base station obtains channel matrix H _{N, k}, its size is n _r* n _t

Wherein, K representes intrasystem total number of users, and N representes the total number of base of participating in cooperating, H _{N, k}Expression base station n is to the space channel of user k, n _rExpression reception antenna number, n _tThe expression number of transmit antennas.

12) if the cooperation mode of JP is adopted in the minizone, a plurality of base stations equivalence of then participating in cooperation is the big base station of antenna more than, and multi-cell cooperating system equivalence this moment is a single base stations and multiuser system.The equivalent space channel of user k is expressed as:

H _k＝[H _1k?H _2k?L?H _Nk]，k＝1，2，K?K

If what the cooperation between the sub-district was adopted is the CB cooperation mode, then this step is omitted.

13) design time length L through statistical average, calculates channel second-order statistics metric matrix.

If the cooperation mode of minizone employing JP then

$R_k=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_k}^H\left(i\right)H_k\left(i\right)$

If the cooperation mode of minizone employing CB then

$R_{k,k}=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_{k,k}}^H\left(i\right)H_{k,k}\left(i\right)$

Wherein, R _kThe space channel second-order statistics metric matrix of expression user k, R _{K, k}Expression base station k is to the space channel second-order statistics metric matrix of user k, H _k(i), H _{K, k}(i) expression i moment spatial channel matrix, i representes some observing time.

Said method, cooperation each other between the said base station of step (2), the channel second-order statistics information that utilize to obtain calculates each user's of system precoding vector according to the precoding vector computing formula, and its step is following:

21) if the cooperation mode of JP, then compute matrix A=R are adopted in the minizone _kAnd matrix

If the cooperation mode of CB, then compute matrix A=R are adopted in the minizone _{K, k}And matrix

R wherein _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix;

22) to matrix B ^-1A carries out characteristic root and decomposes, and extracts eigenvalue of maximum characteristic of correspondence vector, promptly obtains the precoding vector of user k.B wherein ^-1The inverse matrix of representing matrix B.

Data that said method, the said base station of step (3) need send user k and this user's precoding vector w _kCorrespondence multiplies each other, and sends the data after the precoding, and its step is following:

31) if the cooperation mode of JP is adopted in the minizone, then the transmission data on the n of base station are:

$x_n=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{w}_{n,k}{s}_k$

And w _{N, k}And s _kSatisfy following normalizing condition:

E{|s _k|2}＝1， ${\left|\right|w_k\left|\right|}^2=\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{\left|\right|w_{k,n}\left|\right|}^2=1$ k＝1，2，K?K

Wherein, w _{N, k}Represent the precoding vector of k user on the n of base station, s _kThe transmission data of representing k user, these data are shared on all base stations.

32) if the cooperation mode of CB is adopted in the minizone, then the transmission data on the n of base station are:

x _n＝w _ns _n

And w _nAnd s _nSatisfy following normalizing condition:

E{|s _n| ²}＝1，||w _n|| ²＝1 n＝1，2，KN

Wherein, w _nPrecoding vector on the expression base station n, s _nThe transmission data of expression user n, these data are only transmitted on the n of base station.

Table 1 is provided with for the simulated environment parameter.

Table 1

System parameters	LTE multi-cell cooperating system
		The cell cooperative pattern	JP，CB
Carrier frequency	2.0GHz
		System bandwidth	1.4MHz
The base station number	2
		Number of users	2
Number of transmit antennas	4
		The reception antenna number	2
The data fluxion	Single current
		Channel model	EPA5
The PMI time delay	1ms
		Correlation	Low relevant
Channel estimating	Perfect estimation
		Chnnel coding	Turbo code
Modulation system	QPSK，16QAM
		Precoding	SLNR，ESLNR
The statistics duration.	20ms
		MIMO detects	MRC

Claims (3)

1. method for precoding based on channel statistical information that is applied to the multi-cell cooperating system is characterized in that this method may further comprise the steps:

1) multi-cell base station of participation cooperation obtains each user's of system channel second-order statistics information;

2) cooperation each other between the base station, the channel second-order statistics information that utilize to obtain, according to following formula:

${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_i+N_{0}I)}^{-1}{R}_k$ Or ${(\underset{i\≠k}{\overset{N}{\mathrm{\Σ}}}{R}_{k,i}+N_{0}I)}^{-1}{R}_{k,k}$ K=1,2,3, K K

Calculate the precoding vector w of user k in the system _k

Wherein N representes to participate in the system total number of base of cooperation, and K representes intrasystem total number of users, R _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix; () ^-1Expression is to matrix inversion;

3) base station data that user k need be sent and this user's precoding vector w _kCorrespondence multiplies each other, and sends the data after the precoding.

2. the method for precoding that is applied to the multi-cell cooperating system according to claim 1 based on channel statistical information; It is characterized in that; The multi-cell base station of the said participation cooperation of step 1) obtains each user's of system channel second-order statistics information, and its step is following:

11) user k sends detectable signal SRS to base station n, and the place obtains channel matrix H in the base station _{N, k}, its size is n _r* n _t, k=1,2,3 ... K, n=1,2,3...N, wherein, K representes intrasystem total number of users, N representes the total number of base of participating in cooperating, H _{N, k}Expression base station n is to the space channel of user k, n _rExpression reception antenna number, n _tThe expression number of transmit antennas;

12) if the cooperation mode of joint transmission (JP) is adopted in the minizone, a plurality of base stations equivalence of then participating in cooperation is the big base station of antenna more than, and multi-cell cooperating system equivalence this moment is a single base stations and multiuser system, and the equivalent space channel of user k is expressed as:

H _k＝[H _1k?H _2k?L?H _Nk]

If the cooperation mode of united beam moulding (CB) is adopted in cooperation between many sub-districts, then this step is omitted;

13) design time length L through statistical average, calculates channel second-order statistics metric matrix,

If the cooperation mode of minizone employing JP then

$R_k=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_k}^H\left(i\right)H_k\left(i\right)$

If the cooperation mode of minizone employing CB then

$R_{k,k}=\frac{1}{L}\underset{i=0}{\overset{L-1}{\mathrm{\Σ}}}{H_{k,k}}^H\left(i\right)H_{k,k}\left(i\right)$

Wherein, R _kThe space channel second-order statistics metric matrix of expression user k, R _{K, k}Expression base station k is to the space channel second-order statistics metric matrix of user k, H _k(i), H _{K, k}(i) expression i moment spatial channel matrix, i representes some observing time.

3. the method for precoding that is applied to the multi-cell cooperating system according to claim 1 based on channel statistical information; It is characterized in that; Step 2) cooperation each other between the said base station; The channel second-order statistics information that utilize to obtain calculates each user's of system precoding vector according to the precoding vector computing formula, and its step is following:

21) if the cooperation mode of JP, then compute matrix A=R are adopted in the minizone _kWith

If the cooperation mode of CB, then compute matrix A=R are adopted in the minizone _{K, k}And matrix

R wherein _kAnd R _iThe space channel second-order statistics metric matrix of representing user k and user i respectively, R _{K, k}And R _{K, i}Represent the space channel second-order statistics metric matrix of base station k respectively, N to the space channel second-order statistics metric matrix of user k and base station k to user i ₀The power of expression receiving terminal white Gaussian noise; I representation unit matrix;

22) to matrix B ^-1A carries out characteristic root and decomposes, and extracts eigenvalue of maximum characteristic of correspondence vector, promptly obtains the precoding vector of user k.B wherein ^-1The inverse matrix of representing matrix B.

Images