CN103595455B - LTE A non-code book beam form-endowing method based on user satisfaction - Google Patents

Abstract

Based on the LTE A non-code book beam form-endowing method of user satisfaction, belong to the beamforming technique field in mobile communication system.In order to solve current beam form-endowing method while improving systematic function, also deposit fairness among different users and the low problem of the corresponding satisfaction of each user.It includes：Step one：Corresponding wave beam formed matrix is obtained using existing SLNR beam form-endowing method；Step 2：Under the wave beam formed matrix that step one obtains, obtain the SINR value of each user and be satisfied with angle value；Step 3：Using the SINR value obtaining and be satisfied with angle value, the wave beam formed matrix that step one is obtained is modified weighting, and obtains the wave beam formed matrix after revising weighting；Step 4：For GBR traffic and non GBR traffic, the wave beam formed matrix after weighting of revising that step 3 is obtained is revised further using different weights, obtains optimal beam figuration matrix.It is used for forming multiuser wave beam.

Description

LTE-A based on user satisfaction non-code book beam form-endowing method

Technical field

The invention belongs to the beamforming technique field in mobile communication system, it is based on user satisfaction particularly to a kind of LTE-A non-code book beam form-endowing method.

Background technology

In LTE-A standard, MIMO technology has obtained further reinforcement, its highest can support descending 8 layers, up 4 The transmission mode of layer, additionally, the technology such as multiuser MIMO and beam shaping has also obtained further development and has applied.In LTE- The up-link of A, can reduce the interference between user by the joint-detection at eNodeB end；And consider the equipment of descending UE Ability, then mainly realize the suppression of inter-user interference by precoding or beam shaping technology, wherein non-code book beam shaping Method is subject to more in recent years due to can be reduced the signaling consumption of channel estimation using the reciprocity of tdd mode lower channel Many concerns.

Release 8 version of 3GPP starts support the beam shaping of single user, open after Release 10 version Begin to support multi-user's multi-flow beam excipient.For single user wave beam forming, beamforming algorithm that feature based value is decomposed due to Realize simple and be widely used, and for forming multiuser wave beam, it can be found that dirty paper code algorithm can reach highest Channel capacity, but the complexity of this algorithm is too high, is difficult to realize in actual hardware.Letter leakage noise ratio algorithm is Mirette A kind of beamforming algorithm that Sadek proposes, this algorithm utilizes SLNR Lai approximate antithesis SINR (Signal to Interference and Noise Ratio), can realize while reducing algorithm complex higher power system capacity and Relatively low average error bit rate is preferably compromised such that it is able to obtain between complexity and systematic function.However, multi-user SLNR Algorithm is only to turn to optimization aim with the SLNR value maximum of transmitting terminal to obtain corresponding wave beam forming vector, does not consider user The impact to systematic function for the different channel conditions, therefore has scholar to propose weighting SLNR beam shaping method, using letter again Road coefficient is weighted to each disturbed user as weighter factor, thus the situation of change of channel gain is considered wave beam During excipient, substantially increase the performance of SLNR beam shaping method.On here basis, there is scholar to improve further again The alternative manner of weight coefficient, and introduce iterative process, thus the system that further increases overall and speed.

But, said method is all to improve systematic function as target, but while improving systematic function, also exists different Fairness between user and the low situation of the corresponding satisfaction of each user.

Content of the invention

The invention aims to solving current beam form-endowing method while improving systematic function, also exist not With the low problem of the fairness between user and the corresponding satisfaction of each user, the present invention provides one kind based on user's satisfaction The LTE-A non-code book beam form-endowing method of degree.

LTE-A based on the user satisfaction non-code book beam form-endowing method of the present invention,

It comprises the steps：

Step one：Corresponding wave beam formed matrix is obtained using existing SLNR beam form-endowing method；

Step 2：Under the wave beam formed matrix that step one obtains, obtain the SINR value of each user and be satisfied with angle value；

Step 3：Using the SINR value obtaining and be satisfied with angle value, the wave beam formed matrix that step one is obtained is modified Weighting, obtains the wave beam formed matrix after revising weighting；

Step 4：For GBR traffic and non-GBR business, the wave beam forming square revised after weighting that step 3 is obtained Battle array is revised further using different weights, obtains optimum wave beam formed matrix.

It is an advantage of the current invention that on the basis of existing SLNR algorithm, being modified by the other variables of introducing can To realize different targets, therefore, in order to improve the user fairness of weighting SLNR algorithm, user satisfaction is incorporated into optimum In the calculating process of beam shaping matrix, traditional SLNR method is weighted using different channels condition and satisfaction simultaneously Adjustment, had both improved systematic function, can ensure that satisfaction and the fairness of user again.

Brief description

Fig. 1 is the former of the non-code book beam form-endowing method of the LTE-A based on user satisfaction described in specific embodiment one Reason schematic diagram；

Fig. 2 is beam form-endowing method of the present invention and weights illustrating with capacity comparison curve of existing SLNR algorithm Figure.

Fig. 3 is beam form-endowing method of the present invention and the satisfaction correlation curve weighting existing SLNR algorithm is illustrated Figure.

Fig. 4 is beam form-endowing method of the present invention and the satisfaction contrast weighting each user under existing SLNR algorithm Curve synoptic diagram.Wherein, user 1-4 is GBR traffic user, and user 5-8 is non-GBR service-user.

Fig. 5 is that GBR traffic priority weighting is asynchronous and capacity comparison curve synoptic diagram.

Fig. 6 is GBR traffic priority weighting asynchronous satisfaction correlation curve schematic diagram.

Specific embodiment

Specific embodiment one：In conjunction with Fig. 1, present embodiment is described, described in present embodiment based on user satisfaction LTE-A non-code book beam form-endowing method, it comprises the steps：

Step one：Corresponding wave beam formed matrix is obtained using existing SLNR beam form-endowing method；

Step 2：Under the wave beam formed matrix that step one obtains, obtain the SINR value of each user and be satisfied with angle value；

Step 3：Using the SINR value obtaining and be satisfied with angle value, the wave beam formed matrix that step one is obtained is modified Weighting, obtains the wave beam formed matrix after revising weighting；

Step 4：For GBR traffic and non-GBR business, the wave beam forming square revised after weighting that step 3 is obtained Battle array is revised further using different weights, obtains optimum wave beam formed matrix.

Specific embodiment two：Present embodiment is to the LTE- based on user satisfaction described in specific embodiment one The restriction further of A non-code book beam form-endowing method,

In described step one using the method that existing SLNR beam form-endowing method obtains corresponding wave beam formed matrix it is：

When the transmission power of user side configurations match filtering receiver and user has been normalized, the SLNR of user k It is expressed as：

In formula, W_kIt is the wave beam formed matrix of user k,Wherein H_iIt is user The channel matrix of i,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side,For user k Noise power,For N_TThe unit matrix of dimension, tr is the mark of matrix；

When SLNR maximizes, the wave beam formed matrix of the user k that iteration obtainsFor：

Specific embodiment three：Present embodiment is to the LTE- based on user satisfaction described in specific embodiment one The restriction further of A non-code book beam form-endowing method,

In step 2 under the wave beam formed matrix that step one obtains, the method obtaining the SINR value of each user is：

The SINR value of user k is：

In formula, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel matrix of user i,Representing matrix A's Frobenius norm, N_RIt is the reception antenna number of user side,Noise power for user k.

Specific embodiment four：Present embodiment be to described in specific embodiment one or three based on user satisfaction The restriction further of LTE-A non-code book beam form-endowing method, for GBR traffic, the described wave beam forming square obtaining in step one Under battle array, the method being satisfied with angle value obtaining each user is：

In formula,For the Mean Speed of GBR traffic actual transmissions, r_k,tarr_k,MBRAllow the high speed of transmission for GBR traffic Rate, r_k,minr_k,GBRFor GBR traffic minimum transmission rate；For GBR traffic actual average propagation delay time, d_k,tard_k,PDBFor GBR The maximum transmitted time delay that business allows.

Specific embodiment five：Present embodiment be to described in specific embodiment one or three based on user satisfaction The restriction further of LTE-A non-code book beam form-endowing method, for non-GBR business, the described wave beam obtaining in step one is assigned Under shape matrix, the method being satisfied with angle value obtaining each user is：

In formula,For the Mean Speed of non-GBR business actual transmissions, r_k,MBRAllow transmission for non-GBR business Two-forty, r_k,GBRFor non-GBR business minimum transmission rate；For non-GBR business actual average propagation delay time, d_k,PDBFor The maximum transmitted time delay that non-GBR business allows.

Specific embodiment six：Present embodiment be to described in specific embodiment four or five based on user satisfaction The restriction further of LTE-A non-code book beam form-endowing method, the Mean Speed of actual transmissionsPreparation method be：

Channel matrix information that base station obtains by using channel estimation and the wave beam formed matrix that step one obtains, simulation The Mean Speed of the actual transmissions that actual transmissions estimation obtains

Specific embodiment seven：Present embodiment is to the LTE- based on user satisfaction described in specific embodiment one The restriction further of A non-code book beam form-endowing method, using the SINR value obtaining and be satisfied with angle value in described step 3, to step The wave beam formed matrix of one acquisition is modified weighting, and the method obtaining the wave beam formed matrix after revising weighting is：

The described wave beam formed matrix revised after weighting

In formula, J (W_k, α) be weighting after SLNR, W_kIt is the wave beam formed matrix of user k, α is weight coefficient, H_iIt is to use The channel matrix of family i,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side, α_kFor user k Weight coefficient,For the noise power of user k, k is positive integer；

The weight coefficient α of user i_iFor：

Wherein, SINR_iFor the SINR value of user i, SINR_jFor the SINR value of user j, sat_iIt is satisfied with angle value for user i, sat_i/sat^(max)For normalized user satisfaction value, sat^(max)It is that in all users, highest is satisfied with angle value.

Specific embodiment eight：Present embodiment is to the LTE- based on user satisfaction described in specific embodiment one The restriction further of A non-code book beam form-endowing method, is directed to GBR traffic and non-GBR business, step 3 is obtained in step 4 The wave beam formed matrix revised after weighting revised further using different weights, obtain optimum wave beam formed matrix Method be：

In formula,For optimum wave beam formed matrix, γ is the weight coefficient of GBR traffic, and 1- γ is non-GBR business Weight coefficient, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel matrix of user i,Representing matrix A's Frobenius norm, N_RIt is the reception antenna number of user side,For the noise power of user k, α_kWeighting system for user k Number, K₁It is the number of users of transmission GBR traffic.

The priority of GBR traffic is bigger, and its shared weight coefficient is also bigger, so can improve this kind of service-user Satisfaction；The weight of contrary non-GBR business is less, and such way mainly can be used for different service types The adjustment of family satisfaction.

Specific embodiment nine：Present embodiment is to the LTE- based on user satisfaction described in specific embodiment seven The restriction further of A non-code book beam form-endowing method, for the different service types under GBR traffic or non-GBR business, accordingly The wave beam formed matrix of optimum be：

In formula, β_iIt is the weight coefficient of type of service i, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel square of user i Battle array,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side,Noise work(for user k Rate, α_kWeight coefficient for user k.

For making the purpose of the present invention, technological means and advantage become more apparent, below in conjunction with accompanying drawing, the present invention is done into One step describes in detail.

Embodiment：

Fig. 1 is the principle schematic of multiuser downstream wave beam forming, and total number of users of system is set to K, and eNodeB sends out Penetrating antenna number is N_T, the reception antenna number of UE is N_R.At eNodeB end, the data of all users passes through serioparallel exchange module, via Beam shaping module is sent to UE, and for user k, its wave beam formed matrix can be expressed as Consider beam shaping algorithm, it is therefore assumed that the transmission power of each user and data flow is identical, now ripple due to main herein The mark of bundle excipient matrix needs to meet

Receiving terminal, the receipt signal of user k is represented by：

Wherein,It is the sending signal of user k after beam shaping module, m_k For user k corresponding data fluxion, and need to meet m_k≤N_R；N is the information symbol number sending；For To the channel matrix of k-th user, when channel is for the accidental channel of flat fading, its element obeys zero-mean, list to eNodeB The multiple Gauss distribution of position variance；n_kIt is the zero-mean additive Gaussian noise corresponding to different channels.

As can be seen that in receipt signal r_kIn in addition to the desired signal of user k, also there are other users to user The interference of k, if can not eliminating or reducing this interference, can greatly reduce the Signal to Interference plus Noise Ratio of receiving terminal, so lead to be The decline of system performance.First, apply traditional corresponding wave beam formed matrix of SLNR beam form-endowing method primary Calculation.When UE end When the transmission power of configurations match filtering receiver and user has been normalized, the SLNR of user k is represented by：

W in formula_kIt is the wave beam formed matrix of user k,Wherein H_iIt is user i Channel matrix,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side,For user k's Noise power,For N_TThe unit matrix of dimension.Beamforming algorithm based on SLNR criterion to maximize above formula as target, i.e. ripple Bundle figuration matrixCan be obtained by equation below：

By above formula it is found that when using SLNR algorithm, the selection of the optimal beam figuration matrix of user k is only with it The wave beam formed matrix of itself is relevant, and has no direct relation with the wave beam formed matrix of other users.Existing document channel syndrome Bright, whenFor pencil of matrixGeneralized eigenvalue, and X_kFor each broad sense During the corresponding generalized eigenvector of eigenvalue, the optimal solution of user's k beam shaping matrix is X_kFront m_kRow.Therefore, user k Wave beam formed matrix can be expressed as：In formula, β_kIt is to make wave beam formed matrixMeet transmitting Power constraintWeights coefficient, X_k(:,1:m_k) it is by X_kFront m_kThe matrix of row composition.

According to this initial wave beam formed matrix, calculate the SINR of each user, specific formula for calculation is as follows：

Then, base station obtains by using channel estimation channel matrix information and being obtained using SLNR beam form-endowing method The initial beam figuration matrix arriving, simulation actual transmissions estimate the Mean Speed of user's transmission under initial beam figuration matrix, Thus calculating the satisfaction of each user under initial beam figuration matrix, for expiring of GBR traffic and non-GBR service definition Meaning degree computational methods are as follows：

In formula,r_k,tarAnd r_k,minRespectively represent business actual transmissions Mean Speed, allow transmission flank speed with And minimum transmission rate requires；And d_k,tarRepresent the actual average propagation delay time of business and the maximum delay of permission respectively.

Weighted using the correction that calculated SINR value and satisfaction sat value are carried out to initial beam figuration matrix, this Wave beam formed matrix after Shi Jiaquan can be obtained by following formula：

In formula, weight coefficient α_iCircular as follows：

Wherein sat_i/sat^(max)It is normalized user satisfaction, sat^(max)It is that in all users, highest is satisfied with angle value.

Priority in view of different business is also different, for GBR traffic and non-GBR business, using different weights Revised further, revised wave beam formed matrix can be by formula：

It is calculated.In formula, γ is the weight coefficient of GBR traffic, and 1- γ is the weight coefficient of non-GBR business, specifically The setting of root determines to the priority of each business according to operator, K₁It is the number of users of transmission GBR traffic.As to delineation of activities more Careful, as shown in table 1, the different service types in GBR traffic (non-GBR business) classification can select to be set to different power Weight coefficient, corresponding wave beam forming can be by formula：

It is calculated.β in formula_iIt is the weight coefficient of type of service i.

The priority of table 1 different kinds of business divides

According to the parameter request of LTE system, simulation parameter setting is as follows：The transmitting antenna sum of eNB is set to 8, user End configuration single antenna, the total number of users K=8 of the work in system, the number of users of wherein GBR traffic and non-GBR business is 4 Individual；System bandwidth is 1.4MHz；System adopts extended cyclic prefix, i.e. CP=32；Channel is the random of flat multiple Gauss distribution Channel, the design parameter with regard to GBR traffic and non-GBR business satisfaction configures as shown in table 2.Use for making simulation process project The change of family satisfaction, does not consider the processes such as resource allocation herein, and assumes that channel estimation is completely preferable, supposes each simultaneously The noise variance of receiver user is consistent.

Table 2 is the parameter configuration of GBR traffic and non-GBR business satisfaction

Type of service	rk,MBR/Mbps	rk,GBR/Mbps	dk,PDB/ms
				GBR	1.5	1	100
Non-GBR	2	-	100

Fig. 2 and Fig. 3 sets forth the present invention offer beam form-endowing method and weighting SLNR algorithm and capacity and The comparison diagram of satisfaction is it can be seen that both is very nearly the same with capacity, but employs the wave beam forming side of present invention offer Method, the average satisfaction of system has larger lifting.Fig. 4 is when signal to noise ratio is for 10dB, the satisfaction ratio of each user Relatively it can be seen that employing the beam form-endowing method of present invention offer, the satisfaction difference between each user is less than 3%, Therefore, the fairness between user is improved.

Fig. 5 and Fig. 6 sets forth GBR traffic and non-GBR business under different weights and capacity and satisfaction Contrast, in figure gives the priority weighting γ of GBR traffic.From fig. 5, it can be seen that when γ takes different value, the sum of system Capacity has almost no change, but the satisfaction of the user of transmission GBR traffic, with the increase of γ, has and significantly lifted. Γ increases to 0.7 from 0.3, and the average satisfaction of GBR traffic rises about 5%.

By embodiments above as can be seen that the non-code book wave beam of the LTE-A based on user satisfaction that the present invention provides is assigned Shape method, can effectively improve the fairness between the total satisfactory grade of system and user, and operator can be according to need Will, the weight of different business is adjusted, to reach more preferable laser propagation effect.

The present invention also can have other various embodiments, in the case of without departing substantially from present invention spirit and its essence, this area Technical staff when can according to the present invention make various corresponding change and deform, but these corresponding change and deformation all should belong to The protection domain of appended claims of the invention.

Claims (8)

1. the non-code book of the LTE-A based on user satisfaction beam form-endowing method, it comprises the steps：

Step one：Corresponding wave beam formed matrix is obtained using existing SLNR beam form-endowing method；

Step 2：Under the wave beam formed matrix that step one obtains, obtain the SINR value of each user and be satisfied with angle value；

Step 3：Using the SINR value obtaining and be satisfied with angle value, the wave beam formed matrix that step one is obtained is modified weighting, Obtain the wave beam formed matrix after revising weighting；

Step 4：For GBR traffic and non-GBR business, the wave beam formed matrix revised after weighting that step 3 obtains is adopted Revised further with different weights, obtained optimum wave beam formed matrix；

It is characterized in that,

Using the SINR value obtaining and be satisfied with angle value in described step 3, the wave beam formed matrix that step one is obtained is modified Weight, the method obtaining the wave beam formed matrix after revising weighting is：

The described wave beam formed matrix revised after weighting

In formula, J (W_k, α) be weighting after SLNR, W_kIt is the wave beam formed matrix of user k, α is weight coefficient, H_iIt is user i Channel matrix,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side, α_kAdding for user k Weight coefficient,For the noise power of user k, k is positive integer；

The weight coefficient α of user i_iFor：

$\begin{array}{c}{\mathrm{\α}}_i^{US}=\frac{{\mathrm{SINR}}_i/({\mathrm{sat}}_i/{\mathrm{sat}}^{\left(\max \right)})}{{\Σ}_{j=0,j\≠i}^K{\mathrm{SINR}}_j/({\mathrm{sat}}_j/{\mathrm{sat}}^{\left(\max \right)})}\\ =\frac{{\mathrm{SINR}}_i/{\mathrm{sat}}_i}{{\Σ}_{j=0,j\≠i}^K{\mathrm{SINR}}_j/{\mathrm{sat}}_j}\end{array}$

Wherein, SINR_iFor the SINR value of user i, SINR_jFor the SINR value of user j, sat_iThe angle value that is satisfied with for user i, sat_i/ sat^(max)For normalized user satisfaction value, sat^(max)It is that in all users, highest is satisfied with angle value.

2. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 1 it is characterised in that In described step one using the method that existing SLNR beam form-endowing method obtains corresponding wave beam formed matrix it is：

When the transmission power of user side configurations match filtering receiver and user has been normalized, the SLNR of user k represents For：

$\begin{array}{c}J\left(W_k\right)=\frac{\left|\right|H_k{W}_k|{|}_F^2}{N_R{\mathrm{\σ}}_k^2+{\Σ}_{i=1,i\≠k}^K\left|\right|H_i{W}_k|{|}_F^2}\\ =\frac{tr\left(W_k^H{H}_k^H{H}_k{W}_k\right)}{tr\left(W_k^H\right({\stackrel{\‾}{H}}_k^H{\stackrel{\‾}{H}}_k+N_R{\mathrm{\σ}}_k^2{I}_{N_T}\left)W_k\right)}\end{array}$

In formula, W_kIt is the wave beam formed matrix of user k,Wherein H_iIt is the letter of user i Road matrix,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side,Noise for user k Power,For N_TThe unit matrix of dimension, tr is the mark of matrix；

When SLNR maximizes, the wave beam formed matrix of the user k that iteration obtainsFor：

$\begin{array}{c}{\stackrel{\‾}{W}}_k=\underset{W_k}{\arg \max }J\left(W_k\right)\\ =\underset{W_k}{\arg \max }\frac{tr\left(W_k^H{H}_k^H{H}_k{W}_k\right)}{tr\left(W_k^H\right({\stackrel{\‾}{H}}_k^H{\stackrel{\‾}{H}}_k+N_R{\mathrm{\σ}}_k^2{I}_{N_t}\left)W_k\right)}\end{array}.$

3. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 1 it is characterised in that In step 2 under the wave beam formed matrix that step one obtains, the method obtaining the SINR value of each user is：

The SINR value of user k is：

${\mathrm{SINR}}_k=\frac{\left|\right|H_k{W}_k|{|}_F^2}{N_R{\mathrm{\σ}}_k^2+{\Σ}_{i=1,i\≠k}^K\left|\right|H_k{W}_i|{|}_F^2}$

In formula, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel matrix of user i,The Frobenius of representing matrix A Norm, N_RIt is the reception antenna number of user side,Noise power for user k.

4. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 1 it is characterised in that For GBR traffic, described under the wave beam formed matrix that step one obtains, the method being satisfied with angle value obtaining each user is：

${\mathrm{sat}}_k^{GBR}=\left\{\begin{array}{cc}1& \stackrel{\&OverBar;}{r_k}\&GreaterEqual;r_{k,MBR}and\stackrel{\&OverBar;}{d_k}\&le;d_{k,PDB}\\ \frac{\stackrel{\&OverBar;}{r_k}-r_{k,GBR}}{r_{k,MBR}-r_{k,GBR}}& else\\ 0& \stackrel{\&OverBar;}{r_k}<r_{k,GBR}or\stackrel{\&OverBar;}{d_k}>d_{k,PDB}\end{array}\right.$

In formula,For the Mean Speed of GBR traffic actual transmissions, r_k,MBRAllow the flank speed of transmission, r for GBR traffic_k,GBRFor GBR traffic minimum transmission rate；For GBR traffic actual average propagation delay time, d_k,PDBFor GBR traffic allow maximum transmitted when Prolong.

5. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 1 it is characterised in that

For non-GBR business, described under the wave beam formed matrix that step one obtains, obtain the angle value that is satisfied with of each user Method is：

${\mathrm{sat}}_k^{Non-GBR}=\left\{\begin{array}{cc}1& \stackrel{\&OverBar;}{r_k}\&GreaterEqual;r_{k,MBR}and\stackrel{\&OverBar;}{d_k}\&le;d_{k,PDB}\\ {\left(\frac{\stackrel{\&OverBar;}{r_k}}{r_{k,MBR}}\right)}^2& else\\ 0& \stackrel{\&OverBar;}{d_k}>d_{k,PDB}\end{array}\right.$

In formula,For the Mean Speed of non-GBR business actual transmissions, r_k,MBRAllow the high speed of transmission for non-GBR business Rate, r_k,GBRFor non-GBR business minimum transmission rate；For non-GBR business actual average propagation delay time, d_k,PDBFor non- The maximum transmitted time delay that GBR traffic allows.

6. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 4 or 5, its feature exists In,

The Mean Speed of actual transmissionsPreparation method be：

Channel matrix information that base station obtains by using channel estimation and the wave beam formed matrix that step one obtains, simulation is actual The Mean Speed of the actual transmissions that transmission estimation obtains

7. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 1 it is characterised in that

It is directed to GBR traffic and non-GBR business in step 4, the wave beam formed matrix revised after weighting that step 3 obtains is adopted Revised further with different weights, the method obtaining optimum wave beam formed matrix is：

$\begin{array}{c}{W}_k^o=\underset{W_k}{\arg \max }\left(J\right(W_k,\mathrm{\&alpha;}\left)\right)\\ =\arg \underset{W_k}{\max }\frac{\left|\right|H_k{W}_k|{|}_F^2}{{\mathrm{\&alpha;}}_k{N}_R{\mathrm{\&sigma;}}_k^2+\mathrm{\&gamma;}{\&Sigma;}_{i=1,\&NotEqual;k}^{K_1}{\mathrm{\&alpha;}}_i\left|\right|H_i{W}_k|{|}_F^2+(1-\mathrm{\&gamma;}){\&Sigma;}_{i=K_1+1,\&NotEqual;k}^{K_1}{\mathrm{\&alpha;}}_i|\left|H_i{W}_k\right|{|}_F^2}\end{array}$

In formula,For optimum wave beam formed matrix, γ is the weight coefficient of GBR traffic, and 1- γ is the weight of non-GBR business Coefficient, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel matrix of user i,The Frobenius model of representing matrix A Number, N_RIt is the reception antenna number of user side,For the noise power of user k, α_kFor the weight coefficient of user k, K₁It is transmission The number of users of GBR traffic.

8. the non-code book beam form-endowing method of the LTE-A based on user satisfaction according to claim 6 it is characterised in that For the different service types under GBR traffic or non-GBR business, optimum wave beam formed matrix is accordingly：

$\begin{array}{c}{W}_k^o=\underset{W_k}{\arg \max }\left(J\right(W_k,\mathrm{\&alpha;}\left)\right)\\ =\arg \underset{W_k}{\max }\frac{\left|\right|H_k{W}_k|{|}_F^2}{{\mathrm{\&alpha;}}_k{N}_R{\mathrm{\&sigma;}}_k^2+{\&Sigma;}_{i=1,i\&NotEqual;k}^K{\mathrm{\&beta;}}_i{\mathrm{\&alpha;}}_i\left|\right|H_i{W}_k|{|}_F^2}\end{array}$

In formula, β_iIt is the weight coefficient of type of service i, W_kIt is the wave beam formed matrix of user k, H_iIt is the channel matrix of user i,The Frobenius norm of representing matrix A, N_RIt is the reception antenna number of user side,For the noise power of user k, α_k Weight coefficient for user k.