CN101553026B - Method and device for controlling power of downlink closed-loop - Google Patents

Abstract

The present invention provides a method and a device for controlling power of downlink closed-loop, wherein the method comprises the following steps: A. pre-establishing a predictive matrix used for calculating the coefficient of fitting curve; B. recording the measured value of signal-to-noise ratio of current subframe and n-1 subframes before the current subframe; C. calculating the coefficientof fitting curve according to the measured value of predictive matrix and the recorded n signal-to-noise ratios for further obtaining the assured fitting curve; D. calculating the predicted value of signal-to-noise ratio of next subframe time according to the assured fitting value; and E. comparing the predicted value of signal-to-noise ratio with the target value of signal-to-noise ratio, and generating power control command word according to the comparing result. The method and the device of the invention realizes an accurate control to the downlink transmitting power thereby reducing the power consumption of whole network.

Description

A kind of power control method for downward closed loop and device

Technical field

The invention belongs to wireless communication field, particularly relate to a kind of power control method for downward closed loop and device.

Background technology

In communication system, transmitting power is a kind of resource of preciousness, and the base station sends data by the transmitting power of setting to terminal.In communicating system descending link, each subscriber terminal equipment is different with the distance of base station, like this, in the base station when the transmitting power of each terminal equipment is identical, the signal that will cause the terminal equipment nearer apart from the base station to receive is strong, a little less than the signal that terminal equipment far away receives, thereby make that the terminal equipment that is positioned at base station far-end can't operate as normal; And, also running into the problem of " shadow effect " in the radio wave propagation through regular meeting, the position of user terminal in the sub-district is often change at random, so path loss can significantly change.Therefore, transmitting power that should the real time altering base station can guarantee communication quality, and this process is exactly the control of communication system down link power.

As shown in Figure 1, in the descending control of base station transmitting power, can adopt the power control mode of closed loop, under this control model, power control is divided into inner-loop power control and external circule power control.Described external circule power control its objective is by the actual signal quality (being Block Error Rate) to monitoring to compare with professional required target signal qualities, obtains the signal-to-noise target value of a real-time update, is configured to the inner-loop power control process again and uses; And described inner-loop power control, signal to noise ratio according to current measurement, and several signal-to-noise ratio measurements of past, predict the signal to noise ratio of next subframe, and by comparing the signal-to-noise target value that signal-to-noise ratio to predict value and external circule power control provide, draw the power control command word of adjusting about power, this power control command word is sent to the base station by upward signal, the base station is regulated transmitting power according to this power control command word.

Portable terminal dopes the signal to noise ratio of next subframe according to current and historical snr of received signal, and compare with target signal to noise ratio with this signal to noise ratio of predicting and to draw power control command word, next is controlled constantly to the transmitting power of portable terminal according to this power control command word in the base station.On the one hand, under the situation of slow fading, prior art estimated performance to signal to noise ratio in control loop is not good.On the other hand, because the prediction of signal demodulation and signal to noise ratio be because the restriction of firmware can not realize real-time processing, power control command word may uplink be given the base station behind a fixed response time.As shown in Figure 3, the base station is measured subframe n and is predicted, at subframe n+2 constantly, the base station is given in the power control command word uplink; Also there is a fixed response time in processing controls command word in base station, in Fig. 3, the base station is controlled the transmitting power of portable terminal correspondence at subframe n+3 constantly, therefore, originally wish the transmitting power of the control subframe n+1 moment to portable terminal, but be deferred to subframe n+3 and control constantly, this obviously can not realize the accurate control to transmitting power.

Summary of the invention

Technical problem to be solved by this invention provides a kind of power control method for downward closed loop and device, to realize the accurate control to down transmitting power, reduces power consumption of whole network.

For solving the problems of the technologies described above, it is as follows to the invention provides technical scheme:

A kind of power control method for downward closed loop comprises the steps:

A, set up the prediction matrix of the coefficient be used to calculate matched curve in advance;

The signal-to-noise ratio measurements of preceding n-1 subframe of B, the current subframe of record and current subframe, n is the integer greater than 1;

C, calculate the coefficient of matched curve according to n signal-to-noise ratio measurements of described prediction matrix and described record, and then the matched curve that obtains determining;

D, calculate next subframe signal-to-noise ratio to predict value constantly according to described definite matched curve;

E, described signal-to-noise ratio to predict value and signal-to-noise target value are compared, produce power control command word according to comparative result.

Above-mentioned method, steps A comprises: the number n of determining the signal-to-noise ratio measurements of needs record; Definite exponent number m that the signal-to-noise ratio measurements of n record is carried out the employed matched curve of match, m is the integer greater than 0; According to described definite n, m, set up described prediction matrix according to criterion of least squares.

Above-mentioned method, among the step C, the column vector that n signal-to-noise ratio measurements of described prediction matrix and described record constituted multiplies each other, and obtains: [a _m..., a ₀] ^T, wherein, a _iBe the i time coefficient of matched curve, i=0 ..., m.

Above-mentioned method among the step D, adopts following formula to calculate next subframe signal-to-noise ratio to predict value constantly

$\widehat{y_{n+1}}=a_0+a_1\·(n+\mathrm{step})+\·\·\·+a_m\·{(n+\mathrm{step})}^m$

Wherein, step adjusts the prediction step that step-length is determined according to power.

Above-mentioned method wherein, when the base station is a n+N subframe constantly the time to the response of n subframe power control command word constantly, also comprises between step D and the step e, adopts following formula right

Compensate:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\α}\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein,

Be the signal-to-noise ratio to predict value after the compensation, N is the integer greater than 1, and α is a compensating factor, 0＜α≤1, TPC_step_db _jIt is the power adjustment of the power control command word correspondence that produces constantly of n-N+j subframe.

Above-mentioned method in the steps A, determines that n is 5,6 or 7.

Above-mentioned method in the steps A, determines that m is 2 or 3.

Above-mentioned method, among the step D, when power adjustment step-length was 1dB, the span of step was (0.2,0.6), when power adjustment step-length was 2dB, the span of step was (0.4,1.2), when power adjustment step-length was 3dB, the span of step was (0.6,1.8).

Above-mentioned method, in the step e, when described comparative result is greater than or equal to signal-to-noise target value for the signal-to-noise ratio to predict value, produce the power control command word that reduces power, when described comparative result is a signal-to-noise ratio to predict value during less than signal-to-noise target value, produce the power control command word of rising power.

A kind of downlink closed-loop output control device comprises:

Prediction matrix is set up module, is used to set up prediction matrix, and described prediction matrix is used to calculate the coefficient of matched curve;

Logging modle is used to write down the signal-to-noise ratio measurements of preceding n-1 subframe of current subframe and current subframe, and n is the integer greater than 1;

Fitting module is used for the coefficient according to n signal-to-noise ratio measurements calculating matched curve of described prediction matrix and described record, and then the matched curve that obtains determining;

Prediction module is used for calculating next subframe signal-to-noise ratio to predict value constantly according to described definite matched curve;

The power control command word generation module is used for described signal-to-noise ratio to predict value and signal-to-noise target value are compared, and produces power control command word according to comparative result.

Compared with prior art, embodiments of the invention are predicted signal to noise ratio based on criterion of least squares, and are comparatively accurate to predicting the outcome of signal to noise ratio; Can also the response delay of power control command word be compensated, further improved signal-to-noise ratio to predict result's accuracy.So, realized accurate control, reduced power consumption of whole network down transmitting power.

Description of drawings

The schematic diagram of Fig. 1 for down transmitting power is controlled;

Fig. 2 is the flow chart of the power control method for downward closed loop of the embodiment of the invention;

Fig. 3 is a power control command word response delay schematic diagram;

Fig. 4 is the structure chart of the downlink closed-loop output control device of the embodiment of the invention.

Embodiment

Embodiments of the invention are at first selected the historical record value set: (x ₁, y ₁), (x ₂, y ₂) ..., (x _n, y _n), wherein, x _iBe sample point (being the subframe sequence number), y _iBe sample point x _iCorresponding signal-to-noise ratio measurements uses a curve f (x) to carry out match then, makes it have minimum error:

Embodiments of the invention adopt criterion of least squares to carry out match, suppose that promptly this matched curve is the multinomial on m rank, and expression formula is as follows:

y＝a ₀+a ₁x+…+a _mx ^m

With above-mentioned formula to historical record value set (x ₁, y ₁), (x ₂, y ₂) ..., (x _n, y _n) be similar to, make it have minimum square error:

Here, a ₀..., a _mBe the coefficient of one group of the unknown, in order to obtain least squares error, to a _iThe single order inverse have null value, that is:

$\left\{\begin{array}{c}\frac{\∂\mathrm{\Π}}{\∂a_0}=2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}[y_i-(a_0+a_1{x}_i+a_2{x}_i^2+\·\·\·+a_m{x}_i^m)]=0\\ \frac{\∂\mathrm{\Π}}{\∂a_1}=2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i\·[y_i-(a_0+a_1{x}_i+a_2{x}_i^2+\·\·\·+a_m{x}_i^m)]=0\\ \·\\ \·\\ \·\\ \frac{\∂\mathrm{\Π}}{\∂a_m}=2\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i^m\·[y_i-(a_0+a_1{x}_i+a_2{x}_i^2+\·\·\·+a_m{x}_i^m)]=0\end{array}\right.$

Above-mentioned equation is simplified, is expressed, can obtain with matrix equation:

I ^T·Y ^T＝(I ^T·I)·A ^T

Here, Y=(y ₁, y ₂..., y _n), A=(a _m, a _M-1..., a ₀), and:

$I=\left[\begin{array}{cccc}{x}_1^m& \·\·\·& x_1^1& x_1^0\\ x_2^m& \·\·\·& x_2^1& x_2^0\\ \·& \·& \·& \·\\ \·& \·& \·& \·\\ \·& \·& \·& \·\\ x_n^m& \·\·\·& x_n^1& x_n^0\end{array}\right]$

Because the Y vector all is a known matrix with the I matrix, can find the solution the unknowm coefficient vector by matrix operation:

A ^T＝(I ^T·I) ^-1·I ^T·Y ^T

Like this, just can utilize known historical record value, use the polynomial coefficient a that finds the solution acquisition ₀, a ₁..., a _m, next unknown sample point is predicted estimation:

$\widehat{y_{n+1}}=a_0+a_1\·(n+\mathrm{step})+\·\·\·+a_m\·{(n+\mathrm{step})}^m$

Wherein

Expression n+1 subframe signal-to-noise ratio to predict value constantly, step is a prediction step.

The situation of the above-mentioned power control command word that is applicable to that the base station sends in the n subframe constantly in n+1 subframe time of day response terminal of predicting the outcome, if the base station is the n+N subframe moment to the response of this power control command word, need so power control command word is compensated, the true signal-to-noise ratio to predict value that is suitable for should be:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\α}\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein

Be the signal-to-noise ratio to predict value after the compensation, TPC_step_db _jBe the power control command word of n-N+j subframe terminal generation constantly, α is a compensating factor, 0＜α≤1.

With reference to Fig. 2, the power control method for downward closed loop of the embodiment of the invention, carry out following steps in subscriber terminal side:

Step 201: determine the number n of the signal-to-noise ratio measurements of needs record, and the record register is carried out initialization;

According to the theoretical analysis result of least square method, n={5,6,7} is more reasonable.

Step 202: definite exponent number m that the signal-to-noise ratio measurements of n record is carried out the employed matched curve of match;

According to the link level simulation result, use m={2, it is more reasonable that the curve on 3} rank carries out match to the signal-to-noise ratio measurements of record.

Step 203: calculate prediction matrix PM=(I ^TI) ^-1I ^T, and to preserve this prediction matrix be known matrix, is used for the coefficient calculations of matched curve, wherein,

$I=\left[\begin{array}{cccc}{x}_1^m& \·\·\·& x_1^1& x_1^0\\ x_2^m& \·\·\·& x_2^1& x_2^0\\ \·& \·& \·& \·\\ \·& \·& \·& \·\\ \·& \·& \·& \·\\ x_n^m& \·\·\·& x_n^1& x_n^0\end{array}\right]$

Step 204: the coefficient that calculates matched curve

[a _m，…，a ₀] ^T＝PM×[y ₁，y ₂，…，y _n] ^T

Step 205: calculate next signal-to-noise ratio to predict value of (being the n+1 subframe moment) constantly

$\widehat{y_{n+1}}=a_0+a_1\·(n+\mathrm{step})+\·\·\·+a_m\·{(n+\mathrm{step})}^m$

Here, step is for adjusting the definite prediction step of step-length (be after the base station receives power control command word, whenever carry out a transmitting power and adjust corresponding power recruitment or power reduction) according to power, according to simulation result, when power adjustment step-length was 1dB, the span of step was (0.2,0.6), when power adjustment step-length is 2dB, the span of step is (0.4,1.2), when power adjustment step-length is 3dB, the span of step is (0.6,1.8).

Step 206:, adopt following formula right when the base station is a n+N subframe constantly the time to the response of n subframe power control command word constantly

Compensate:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\α}\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein,

Be the signal-to-noise ratio to predict value after the compensation, α is a compensating factor, 0＜α≤1, TPC_step_db _jIt is the power adjustment of the power control command word correspondence that produces constantly of n-N+j subframe.

Step 207: generate power control command word (TPC), and send to the base station;

Described signal-to-noise ratio to predict value and signal-to-noise target value are compared, produce power control command word according to comparative result, if

(when not compensating, use

) value be not less than the signal-to-noise target value (SNR_target) that exterior ring power control obtains, then produce the power control command word (00) that reduces power, otherwise, produce the power control command word (11) of rising power:

$\mathrm{TPC}=\left\{\begin{array}{cc}00& {\hat{y}}_{i+N}\&GreaterEqual;\mathrm{SNR}\_t\arg \mathrm{et}\\ 11& {\hat{y}}_{i+N}<\mathrm{SNR}\_t\arg \mathrm{et}\end{array}\right.$

When next sub-frame arrives constantly, upgrade the historical record of signal-to-noise ratio measurements one by one, update method is as follows:

$\left\{\begin{array}{c}{y}_1=y_2\\ y_2=y_3\\ \&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\\ y_n=y_{n+1}\end{array}\right.$

Repeating step 205 is to the process of step 207 then.

The base station is adjusted the power that transmits according to power control command word after receiving the power control command word of user terminal transmission.

Below provide an application example.

The historical record value of supposing the least square method of use is 6, i.e. (x ₁, y ₁), (x ₂, y ₂) ..., (x ₆, y ₆); The exponent number of the matched curve of using is 2 rank, i.e. y=a ₀+ a ₁X+a ₂x ²Memory cell to 6 history value is carried out initialization, promptly

$\left\{\begin{array}{c}{y}_1=C\\ y_2=C\\ \&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\\ y_6=C\end{array}\right.$

Wherein C is a fixed constant, if the unit of the signal to noise ratio of input is dB, then C can be made as 9.

Calculate prediction matrix PM=(I ^TI) ^-1I ^T, and preserve this matrix, because the matched curve of choosing is a parabola, and the historical record value of using is 6, its unit matrix is

$\left[\begin{array}{ccc}1& 1& 1\\ 2^2& 2& 1\\ 3^2& 3& 1\\ 4^2& 4& 1\\ 5^2& 5& 1\\ 6^2& 6& 1\end{array}\right]$

Prediction matrix is

$\mathrm{PM}=\left[\begin{array}{cccccc}0.089286& -0.01786& -0.07143& -0.07143& -0.01786& 0.089286\\ -0.76786& 0.039286& 0.47123& 0.52857& 0.21071& -0.48214\\ 1.5& 0.3& -0.4& -0.6& -0.3& 0.5\end{array}\right]$

According to the prediction matrix PM that preserves, calculate the coefficient (the n subframe constantly) of matched curve

[a ₂，a ₁，a ₀] ^T＝PM×[y ₁，y ₂，y ₃，y ₄，y ₅，y ₆] ^T

$\left\{\begin{array}{c}{y}_1=y_{n-5}\\ y_2=y_{n-4}\\ \&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\\ y_6=y_n\end{array}\right.$

Calculate then next subframe predicted value constantly

${\hat{y}}_{n+1}=a_0+a_1\&CenterDot;\left(6.4\right)+a_2\&CenterDot;{\left(6.4\right)}^2$

According to specific implementation, judge the transmitting time of power control command word, with reference to Fig. 3, if constantly could send to the base station to power control command word in the n+2 subframe, and the base station can only constantly could respond the power control command word that receives in the n+3 subframe the soonest, so, and need be to predicted value

Compensate

${\hat{y}}_{n+3}={\hat{y}}_{n+1}+\mathrm{\&alpha;}\underset{j=1}{\overset{2}{\mathrm{\&Sigma;}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein, TPC_step_db ₁The transmitting power adjusted value of the control command word correspondence that produces constantly for n-2, TPC_step_db ₂The transmitting power adjusted value of the power control command word correspondence that produces constantly for n-1.

With

Be benchmark, the signal-to-noise target value of importing with exterior ring power control compares, and generates power control command word, and sends to the base station:

$\mathrm{TPC}=\left\{\begin{array}{cc}00& {\hat{y}}_{i+3}\&GreaterEqual;\mathrm{SNR}\_t\arg \mathrm{et}\\ 11& {\hat{y}}_{i+3}<\mathrm{SNR}\_t\arg \mathrm{et}\end{array}\right.$

When the n+1 arrival constantly of next sub-frame, upgrade the historical record of signal-to-noise ratio measurements one by one, update method is as follows:

$\left\{\begin{array}{c}{y}_1=y_{n-4}\\ y_2=y_{n-3}\\ \&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\\ y_6=y_{n+1}\end{array}\right.$

Repeat above-mentioned steps then, obtain new power control command word, and send to the base station by ascending time slot.The base station is adjusted the power that transmits according to power control command word after receiving the power control command word of user terminal transmission.

With reference to Fig. 4, the embodiment of the invention also provides a kind of downlink closed-loop output control device, comprises, prediction matrix is set up module, logging modle, fitting module, prediction module and power control command word generation module, wherein:

Prediction matrix is set up module, is used to set up prediction matrix, and described prediction matrix is used to calculate the coefficient of matched curve, can set up prediction matrix in the following manner:

Definite number n that needs the signal-to-noise ratio measurements of record; Definite exponent number m that the signal-to-noise ratio measurements of n record is carried out the employed matched curve of match, m is the integer greater than 0; According to described definite n, m, set up described prediction matrix according to criterion of least squares.

Logging modle is used to write down the signal-to-noise ratio measurements of preceding n-1 subframe of current subframe and current subframe, and n is the integer greater than 1.

Fitting module, be used for calculating the coefficient of matched curve according to n signal-to-noise ratio measurements of described prediction matrix and described record, and then the matched curve that obtains determining, particularly, the column vector that n signal-to-noise ratio measurements of described prediction matrix and described record constituted multiplies each other, and obtains: [a _m..., a ₀] ^T, wherein, a _iBe the i time coefficient of matched curve, i=0 ..., m.

Prediction module is used for calculating next subframe signal-to-noise ratio to predict value constantly according to described definite matched curve, particularly, adopts following formula to calculate next subframe signal-to-noise ratio to predict value constantly

$\widehat{y_{n+1}}=a_0+a_1\&CenterDot;(n+\mathrm{step})+\&CenterDot;\&CenterDot;\&CenterDot;+a_m\&CenterDot;{(n+\mathrm{step})}^m$

Wherein, step adjusts the prediction step that step-length is determined according to power.

When the base station is a n+N subframe constantly the time to the response of n subframe power control command word constantly, described prediction module also adopts following formula right

Compensate:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\&alpha;}\underset{j=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein,

Be the signal-to-noise ratio to predict value after the compensation, N is the integer greater than 1, and α is a compensating factor, 0＜α≤1, TPC_step_db _jIt is the power adjustment of the power control command word correspondence that produces constantly of n-N+j subframe.

The power control command word generation module, be used for described signal-to-noise ratio to predict value and signal-to-noise target value are compared, produce power control command word according to comparative result, particularly, when described comparative result is greater than or equal to signal-to-noise target value for the signal-to-noise ratio to predict value, produce to reduce the power control command word of power,, produce the power control command word of rising power when described comparative result is a signal-to-noise ratio to predict value during less than signal-to-noise target value.

In sum, embodiments of the invention are predicted signal to noise ratio based on criterion of least squares, and are comparatively accurate to predicting the outcome of signal to noise ratio; Can also the response delay of power control command word be compensated, further improved signal-to-noise ratio to predict result's accuracy.So, realized accurate control, reduced power consumption of whole network down transmitting power.

Should be noted that at last, above embodiment is only unrestricted in order to technical scheme of the present invention to be described, those of ordinary skill in the art is to be understood that, can make amendment or be equal to replacement technical scheme of the present invention, and not breaking away from the spiritual scope of technical solution of the present invention, it all should be encompassed in the middle of the claim scope of the present invention.

Claims (13)

1. a power control method for downward closed loop is characterized in that, comprises the steps:

A, set up the prediction matrix of the coefficient be used to calculate matched curve in advance;

The signal-to-noise ratio measurements of preceding n-1 subframe of B, the current subframe of record and current subframe, n is the integer greater than 1;

C, calculate the coefficient of matched curve according to n signal-to-noise ratio measurements of described prediction matrix and described record, and then the matched curve that obtains determining;

D, calculate next subframe signal-to-noise ratio to predict value constantly according to described definite matched curve;

E, described signal-to-noise ratio to predict value and signal-to-noise target value are compared, produce power control command word according to comparative result;

Steps A comprises:

Definite number n that needs the signal-to-noise ratio measurements of record;

Definite exponent number m that the signal-to-noise ratio measurements of n record is carried out the employed matched curve of match, m is the integer greater than 0;

According to described definite n, m, set up described prediction matrix according to criterion of least squares.

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

Among the step C, the column vector that n signal-to-noise ratio measurements of described prediction matrix and described record constituted multiplies each other, and obtains: [α _m..., α ₀] ^T, wherein, α _iBe the i time coefficient of matched curve, i=0 ..., m.

3. method as claimed in claim 2 is characterized in that, among the step D, adopts following formula to calculate next subframe signal-to-noise ratio to predict value constantly

$\widehat{y_{n+1}}=a_0+a_1\&CenterDot;(n+\mathrm{step})+...+a_m\&CenterDot;{(n+\mathrm{step})}^m$

Wherein, step adjusts the prediction step that step-length is determined according to power.

4. method as claimed in claim 3 is characterized in that, when the base station is a n+N subframe constantly the time to the response of n subframe power control command word constantly, also comprises between step D and the step e, adopts following formula right

Compensate:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\&alpha;}\underset{j=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein,

Be the signal-to-noise ratio to predict value after the compensation, N is the integer greater than 1, and α is a compensating factor, 0＜α≤1, TPC_step_db _jIt is the power adjustment of the power control command word correspondence that produces constantly of n-N+j subframe.

5. the method for claim 1 is characterized in that:

In the steps A, determine that n is 5,6 or 7.

6. the method for claim 1 is characterized in that:

In the steps A, determine that m is 2 or 3.

7. method as claimed in claim 3 is characterized in that:

Among the step D, when power adjustment step-length was 1dB, the span of step was (0.2,0.6), and when power adjustment step-length was 2dB, the span of step was (0.4,1.2), and when power adjustment step-length was 3dB, the span of step was (0.6,1.8).

8. the method for claim 1 is characterized in that:

In the step e, when described comparative result is greater than or equal to signal-to-noise target value for the signal-to-noise ratio to predict value, produce to reduce the power control command word of power,, produce the power control command word of rising power when described comparative result is a signal-to-noise ratio to predict value during less than signal-to-noise target value.

9. a downlink closed-loop output control device is characterized in that, comprising:

Prediction matrix is set up module, is used to set up prediction matrix, and described prediction matrix is used to calculate the coefficient of matched curve;

Logging modle is used to write down the signal-to-noise ratio measurements of preceding n-1 subframe of current subframe and current subframe, and n is the integer greater than 1;

Fitting module is used for the coefficient according to n signal-to-noise ratio measurements calculating matched curve of described prediction matrix and described record, and then the matched curve that obtains determining;

Prediction module is used for calculating next subframe signal-to-noise ratio to predict value constantly according to described definite matched curve;

The power control command word generation module is used for described signal-to-noise ratio to predict value and signal-to-noise target value are compared, and produces power control command word according to comparative result;

Described prediction matrix is set up module and is set up prediction matrix in the following manner:

Definite number n that needs the signal-to-noise ratio measurements of record;

Definite exponent number m that the signal-to-noise ratio measurements of n record is carried out the employed matched curve of match, m is the integer greater than 0;

According to described definite n, m, set up described prediction matrix according to criterion of least squares.

10. device as claimed in claim 9 is characterized in that:

Described fitting module, the column vector that n signal-to-noise ratio measurements of described prediction matrix and described record constituted multiplies each other, and obtains: [α _m..., α ₀] ^T, wherein, α _iBe the i time coefficient of matched curve, i=0 ..., m.

11. device as claimed in claim 10 is characterized in that, described prediction module adopts following formula to calculate next subframe signal-to-noise ratio to predict value constantly

$\widehat{y_{n+1}}=a_0+a_1\&CenterDot;(n+\mathrm{step})+...+a_m\&CenterDot;{(n+\mathrm{step})}^m$

Wherein, step adjusts the prediction step that step-length is determined according to power.

12. device as claimed in claim 11 is characterized in that, when the base station is a n+N subframe constantly the time to the response of n subframe power control command word constantly, described prediction module also adopts following formula right

Compensate:

${\hat{y}}_{n+N}={\hat{y}}_{n+1}+\mathrm{\&alpha;}\underset{j=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\mathrm{TPC}\_\mathrm{step}\_{\mathrm{db}}_j$

Wherein,

Be the signal-to-noise ratio to predict value after the compensation, N is the integer greater than 1, and α is a compensating factor, 0＜α≤1, TPC_step_db _jIt is the power adjustment of the power control command word correspondence that produces constantly of n-N+j subframe.

13. device as claimed in claim 9 is characterized in that:

Described power control command word generation module, when described comparative result is greater than or equal to signal-to-noise target value for the signal-to-noise ratio to predict value, produce the power control command word that reduces power, when described comparative result is a signal-to-noise ratio to predict value during less than signal-to-noise target value, produce the power control command word of rising power.

Images