CN104869091A - Method and system for training digital predistortion coefficient - Google Patents

Abstract

The embodiment of the invention relates to the technical field of communication, specifically to a method and a system for training a digital predistortion coefficient, which are used for training a DPD coefficient meeting performance requirements. In the embodiment of the invention, the adjacent channel power ratio (ACPR) of a feedback signal of a training sequence is judged after each DPD processing, so that the effect of each DPD processing can be accurately judged; and when the ACPR meet the requirements, i.e. the effect of DPD processing meets the requirements, the process is ended, and therefore a DPD coefficient meeting the performance requirements is obtained.

Description

A kind of digital pre-distortion coefficient training method and system

Technical field

The embodiment of the present invention relates to communication technical field, particularly relates to a kind of digital pre-distortion coefficient training method and system.

Background technology

Height with domestic and international mobile operator is approved, large scale deployment in recent years commercial, baseband processing unit (Base Band Unit, being called for short BBU) net construction pattern of+Remote Radio Unit (Radio Remote Unit, be called for short RRU) presents the trend fundamentally changing traditional network architecture.Due to various countries and regional frequency spectrum Policy Difference, the frequency spectrum resource relative distribution that each mobile operator obtains, such as comprises 900MHz/1800MHz/2100MHz/2300MHz/2600MHz etc., is generally faced with the challenge of multi-modulation scheme, multiband.Although frequency spectrum resource enriches, frequency range interval is comparatively large, and the broadband of radio frequency becomes developing direction.

Wideband digital predistortion (Digital PreDistortion, being called for short DPD) technology is one of the core technology solving broadband RRU, it carrys out non-linear the caused distortion of compensating power amplifier effectively at numeric field by existing powerful signal processing technology.

Provide a kind of digital pre-distortion method in prior art, the method comprises: gather digital feedback signal; Signal before described digital feedback signal and predistortion is done difference and obtain error signal, the signal before described error signal and described predistortion is obtained the correction value of predistortion correction coefficient by adaptive algorithm iteration convergence; Based on described correction value, predistortion correction coefficient is revised, obtain new predistortion correction coefficient, carry out digital pre-distortion process with the predistortion correction coefficient that this is new.Can find out, the method is revised DPD coefficient, but can not ensure that revised DPD coefficient is better, and in some cases, revised DPD coefficient may be poorer.

In sum, need a kind of DPD coefficient training method and DPD system at present badly, for training the DPD coefficient being met performance requirement.

Summary of the invention

The embodiment of the present invention provides a kind of DPD coefficient training method and DPD system, is met the DPD coefficient of performance requirement in order to training.

The embodiment of the present invention provides a kind of DPD coefficient training method, comprising:

Steps A: sent by radio frequency unit by original training sequence, receives the feedback signal of this original training sequence, and the feedback signal according to this original training sequence and this original training sequence generates DPD coefficient, according to the DPD coefficient update look-up table generated;

Step B: carry out DPD process according to the DPD coefficient in this look-up table to this original training signal, sent by the training sequence after DPD process by radio frequency unit, receives the feedback signal of the training sequence after this DPD process;

Step C: judge whether the Adjacent Channel Power Ratio ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and if so, then terminates the training process of current cycle of training, otherwise proceeds to step D;

Step D: the feedback signal according to the training signal after the training sequence after this DPD process and this DPD process generates DPD coefficient, according to the DPD coefficient update look-up table generated, and proceeds to step B.

Preferably, before proceeding to step B, also comprise:

Judge whether the frequency of training in current cycle of training reaches maximum times;

This proceeds to step B, is specially: if the frequency of training in current cycle of training does not reach maximum times, then proceed to step B.

Preferably, also comprise:

If the frequency of training in current cycle of training reaches maximum times, then terminate the training process of current cycle of training.

Preferably, the ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after this DPD process is greater than threshold value.

Preferably, adopt adaptive algorithm to generate DPD coefficient, this adaptive algorithm is any one in following algorithm:

Least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.

The embodiment of the present invention provides a kind of digital pre-distortion DPD system, comprising:

Training sequence generation unit, for generating original training sequence;

DPD processing unit, for carrying out DPD process according to look-up table to this original training sequence, and exports to radio frequency unit by the sequence after DPD process and sends;

Switch element, for connecting this training sequence generation unit selectivity between this DPD processing unit and radio frequency unit;

DPD coefficient generation unit, for generating DPD coefficient according to the feedback signal received and the training sequence corresponding with this feedback signal;

Look-up table updating block, for the DPD coefficient update look-up table generated according to this DPD coefficient generation unit;

Adjacent Channel Power Ratio ACPR determining unit, for determining the ACPR of this feedback signal according to the feedback signal received;

Control unit, for after cycle of training starts:

Control this switch element the connection between this training sequence generation unit and DPD processing unit to be disconnected, the connection between this training sequence generation unit and radio frequency unit is closed, this original training sequence is sent by this radio frequency unit, the feedback signal of this original series is by after the reception of this DPD coefficient generation unit, this DPD coefficient generation unit generates DPD coefficient according to the feedback signal of this original training sequence and this original series sequence, and this look-up table updating block is according to the DPD coefficient update look-up table generated;

Control this switch element the connection between this training sequence generation unit and this DPD processing unit to be closed, the connection between this training sequence generation unit and radio frequency unit is disconnected, training sequence after this DPD process is sent by this radio frequency unit, the feedback signal of the training sequence after this DPD process is by after the reception of this ACPR determining unit, and this ACPR determining unit determines the ACPR of the feedback signal received; This control unit judges whether this ACPR meets the demands, and if so, then terminates the training process of current cycle of training, otherwise indicates this training sequence generation unit again to generate original training sequence.

Preferably, this control unit, also for

Frequency of training in current cycle of training is counted;

If determine that the frequency of training in current cycle of training reaches maximum times according to count value, then terminate the training process of current cycle of training.

Preferably, the ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after this DPD process is greater than threshold value.

Preferably, this DPD coefficient generation unit specifically for: adopt adaptive algorithm to generate DPD coefficient, this adaptive algorithm is any one in following algorithm: least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.

In the embodiment of the present invention, first perform steps A, the feedback signal according to training sequence and training sequence generates DPD coefficient, according to the DPD coefficient update look-up table generated; Then perform step B, according to the DPD coefficient in look-up table, DPD process is carried out to training signal, receive the feedback signal of the training sequence after DPD process; Then Adjacent Channel Power Ratio (the Adjacent Channel Power Ratio of the feedback signal of the training sequence after DPD process is judged, be called for short ACPR) whether meet the demands, if, then terminate training process, otherwise then generate DPD coefficient according to the feedback signal of the training signal after the training sequence after DPD process and DPD process, according to the DPD coefficient update look-up table generated, and again perform step B.Owing to all judging the ACPR of the feedback signal of the training sequence after DPD process at every turn, so, then the effect of each DPD process accurately can be judged, and meet the requirements at ACPR, namely when DPD treatment effect meets the requirements, process ends, thus the DPD coefficient being met performance requirement.

Accompanying drawing explanation

In order to be illustrated more clearly in the technical scheme in the embodiment of the present invention, below the accompanying drawing used required in describing embodiment is briefly introduced, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.

A kind of DPD coefficient training method schematic flow sheet that Fig. 1 provides for the embodiment of the present invention;

The another kind of DPD coefficient training method schematic flow sheet that Fig. 2 provides for the embodiment of the present invention;

The structural representation of a kind of DPD system that Fig. 3 provides for the embodiment of the present invention.

Embodiment

In order to make object of the present invention, technical scheme and beneficial effect clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.

Fig. 1 illustrates the flow chart that the embodiment of the present invention provides a kind of DPD coefficient training method.

The embodiment of the present invention provides a kind of DPD coefficient training method, comprises the following steps as shown in Figure 1:

Steps A: original training sequence is sent by radio frequency unit, receive the feedback signal of this original training sequence, feedback signal according to this original training sequence and this original training sequence generates DPD coefficient, according to the DPD coefficient update look-up table generated, proceeds to step B afterwards.Specifically, this original training sequence can stochastic generation, for convenience of description, by stochastic generation in the embodiment of the present invention, and become original training sequence without the training sequence of DPD process.

Step B: carry out DPD process according to the DPD coefficient in this look-up table to this original training signal, sent by the training sequence after DPD process by radio frequency unit, is received the feedback signal of the training sequence after this DPD process, proceeds to step C afterwards.

Step C: judge whether the ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and if so, then proceeds to step e, otherwise proceeds to step D.

Step e: the training process terminating current cycle of training.

Step D: the feedback signal according to the training signal after the training sequence after this DPD process and this DPD process generates DPD coefficient, according to the DPD coefficient update look-up table generated, and proceeds to step B.

For convenience of description, in the embodiment of the present invention, original training sequence will be sent, and be called once directly training process based on the process that the feedback signal of original training sequence obtains DPD coefficient, be called an indirect discipline process by the training sequence after sending DPD process and based on the process that the feedback signal of the training sequence after this DPD process obtains DPD coefficient.

The method that the embodiment of the present invention provides is and first uses original training sequence to perform once directly training process, the DPD coefficient update look-up table obtained with direct training process, then performs indirect discipline process.Indirect discipline process may perform repeatedly, in each indirect discipline process, all ACPR judgement is carried out to the feedback signal received, namely judge whether the ACPR of feedback signal meets the demands, if judge, ACPR does not meet the demands, then again perform indirect discipline process, till the ACPR of feedback signal meets the demands.

The DPD normally cycle of training carries out.The embodiment of the present invention can in each DPD cycle of training all according to above-mentioned flow performing.

In above-mentioned steps A and step B, preferably, radio frequency unit specifically comprises the digital to analog converter, the first radio-frequency channel device, the power amplifier that connect successively.Original training sequence is sent to successively digital to analog converter, the first radio-frequency channel device, power amplifier, receive the feedback signal of this original training sequence exported from this power amplifier afterwards.Further, preferably by the feedback signal of the original training sequence exported from this power amplifier that receives successively through the second radio-frequency channel device, analog to digital converter.

Preferably, look-up table is stored in random asccess memory (random access memory is called for short RAM).

In above-mentioned steps B, preferably, by the computing of off-line floating number, in conjunction with the DPD coefficient in look-up table, DPD process is carried out to this original training sequence.

ACPR is used to weigh the standard of interference volume in adjacent frequency channels or quantity of power.ACPR is often defined as the ratio of the average power of adjacent frequency channels or side-play amount and the average power of tranmitting frequency channel.ACPR describes the distortion value because the non-linear factor in radio frequency unit composition causes.Therefore, by judging that the value of ACPR determines whether generated DPD coefficient is the value meeting performance requirement.

Preferably, the ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after this DPD process is greater than threshold value.Choosing of this place's threshold value is relevant with concrete network environment at that time, automatically configures by network, also by manually arranging an empirical value.Preferably, usually the threshold value of ACPR is set to 47dBc, absolute value is greater than this threshold value and then illustrates that Nonlinear perturbations is little, is less than this threshold value and then illustrates that Nonlinear perturbations is large.

Preferred mode is generate DPD coefficient according to the feedback signal of this training sequence and this original training sequence or generate DPD coefficient mode for an employing adaptive algorithm generation DPD coefficient according to the feedback signal of the training signal after the original training sequence after this DPD process and this DPD process, and this adaptive algorithm is any one in following algorithm: least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.

In the embodiment of the present invention, owing to all judging the ACPR of the feedback signal of the training sequence after DPD process at every turn, so, then accurately can judge the effect of each DPD process, and meet the requirements at ACPR, namely when DPD treatment effect meets the requirements, process ends, thus the DPD coefficient being met performance requirement.

The embodiment of the present invention additionally provides another kind of DPD coefficient training method, and as shown in Figure 2, Fig. 2 is the flow chart that the embodiment of the present invention additionally provides another kind of DPD coefficient training method.

Step 201: original training sequence is sent by radio frequency unit, receive the feedback signal of this original training sequence, feedback signal according to this original training sequence and this original training sequence generates DPD coefficient, according to the DPD coefficient update look-up table generated, proceeds to step 202 afterwards.The specific implementation process of steps A is stated described in this step specific implementation process is the same.

Step 202: carry out DPD process according to the DPD coefficient in this look-up table to this original training signal, sent by the training sequence after DPD process by radio frequency unit, is received the feedback signal of the training sequence after this DPD process, proceeds to step 203 afterwards.The specific implementation process of step B is stated described in this step specific implementation process is the same.

Step 203: judge whether the ACPR of the feedback signal of the training sequence after this DPD process meets the demands, and if so, then proceeds to step 206, otherwise proceeds to step 204.The specific implementation process of step C is stated described in this step specific implementation process is the same.

Step 204: the feedback signal according to the training signal after the training sequence after this DPD process and this DPD process generates DPD coefficient, according to the DPD coefficient update look-up table generated, performs step 205 afterwards.The specific implementation process of step D is stated described in this step specific implementation process is the same.

Step 205, judges whether the frequency of training in current cycle of training reaches maximum frequency of training, if the frequency of training in current cycle of training does not reach maximum frequency of training, then performs step 202.

Preferably, if the frequency of training in step 205 in current cycle of training reaches maximum frequency of training, then perform step 206.

Step 206, terminates the training process of current cycle of training.

In above-mentioned steps 205, preferably, the frequency of training in current cycle of training is the number of times of indirect discipline.In concrete enforcement, also can add up all number of times carrying out training, i.e. the number of times of directly training adds the number of times of indirect discipline.The definition of direct training and indirect discipline as previously mentioned.

Specifically, the judgement to indirect discipline number of times is added in above-mentioned steps 205, now can avoid in some cases, when ACPR does not meet the requirements always, perform indirect discipline, so always, then took multi-system resource, therefore, by the judgement to the frequency of training in current cycle of training in the embodiment of the present invention, reduce the consumption of system resource.

In various embodiments of the present invention, carry out to original training sequence the DPD model that DPD process uses, can select as required, the embodiment of the present invention does not limit used DPD model.

Along with the increase of bandwidth, DPD model can become more complicated, due to DPD model analysis and simplify all based on the circuit characteristic of power amplifier, make DPD model maintain the ability describing power amplifier various actual dynamic characteristic.

Such as, the embodiment of the present invention can use the Volterra progression DPD model of discrete form, and this DPD model can be expressed as formula (1):

$\begin{array}{c}z\left(n\right)=\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{h}_1\left(k\right)\·(n-k)\\ +\underset{k_1=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_2=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_3=0}{\overset{K-1}{\mathrm{\Σ}}}{h}_3(k_1,k_2,k_3)\·x(n-k_1)\·x(n-k_2)\·x^{*}(n-k_3)\\ +\underset{k_1=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_2=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_3=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_4=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{k_5=0}{\overset{K-1}{\mathrm{\Σ}}}{h}_5(k_1,k_2,k_3,k_4,k_5)\·x(n-k_1)\·x(n-k_2)\\ \·x(n-k_3)\·x^{*}(n-k_4)\·x^{*}(n-k_5)\end{array}\·\·\·\·\·\·\left(1\right)$

In formula (1):

The input signal that x (n) is system;

X ^*(n) for ask for conjugation to complex signal, x ^*(n)=[xi (n)+jxq (n)] ^*=xi (n)-jxq (n);

Z (n) is system output signal;

H ₁(k) and h ₃(k ₁, k ₂, k ₃) and h ₅(k ₁, k ₂, k ₃, k ₄, k ₅) be power amplifier model parameter;

K is the memory span of system.

The embodiment of the present invention additionally provides the MP model that expression formula is formula (2), expression formula is the time interleaving memory polynomial model of formula (3) and formula (4), and expression formula is the time interleaving conjugation memory polynomial model of formula (5) and formula (6):

$z_{\mathrm{mp}}\left(n\right)=\underset{m=0}{\overset{M}{\mathrm{\Σ}}}y(n-m)\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q}{\left|y(n-m)\right|}^{(q-1)}\·\·\·\·\·\·\left(2\right)$

$z_{\mathrm{mp}-\mathrm{cl}}\left(n\right)=\underset{l=1}{\overset{L_c}{\mathrm{\Σ}}}\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}y(n-m)\underset{q=2}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q,-l}{\left|y(n-m-l)\right|}^{(q-1)}\·\·\·\·\·\·\left(3\right)$

$z_{\mathrm{mp}+\mathrm{cl}}\left(n\right)=\underset{l=1}{\overset{L_c}{\mathrm{\Σ}}}\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}y(n-m)\underset{q=2}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q,l}{\left|y(n-m+l)\right|}^{(q-1)}\·\·\·\·\·\·\left(4\right)$

$z_{\mathrm{mp}-\mathrm{tl}}\left(n\right)=\underset{l=1}{\overset{L_c}{\mathrm{\Σ}}}\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{y}^{*}(n-m)\·y^2(n-m-l)\underset{q=3}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q,-\mathrm{xl}}^1{\left|y(n-m-l)\right|}^{(q-3)}\·\·\·\·\·\·\left(5\right)$

$z_{\mathrm{mp}+\mathrm{tl}}\left(n\right)=\underset{l=1}{\overset{L_c}{\mathrm{\Σ}}}\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{y}^{*}(n-m)\·y^2(n-m+l)\underset{q=3}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q,\mathrm{xl}}^1{\left|y(n-m+l)\right|}^{(q-3)}\·\·\·\·\·\·\left(6\right)$

Above-mentioned formula (2) is in formula (6):

M is memory depth;

Q is intermodulation exponent number;

L _cfor the intermodulation degree of depth, it is namely the interlaced sampling degree of depth;

The input signal that y (n-m), y (n-m+l) are system;

Y ^*n () is for ask for conjugation to complex signal;

Z (n) is system output signal;

Z _mpn () is memory polynomial MP model;

Z _mp-cl(n) and z _mp+cln () is time interleaving memory polynomial;

Z _mp-tl(n) and z _mp+tl(n) time interleaving conjugation memory polynomial.

Based on above-mentioned several DPD model, preferably, also provide one relatively simple DPD model in the embodiment of the present invention, expression formula is as shown in formula (7):

$\begin{array}{c}z\left(n\right)=\underset{m=0}{\overset{M}{\mathrm{\Σ}}}y(n-m)\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q}{\left|y(n-m)\right|}^{(q-1)}+\\ +\underset{m=0}{\overset{M-r}{\mathrm{\Σ}}}y(n-m-r)\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q}^r{\left|y(n-m)\right|}^{(q-1)}\\ +\underset{m=0}{\overset{M-r}{\mathrm{\Σ}}}y(n-m)\underset{q=1}{\overset{Q}{\mathrm{\Σ}}}{w}_{m,q}^{-r}{\left|y(n-m-r)\right|}^{(q-1)}\end{array}\·\·\·\·\·\·\left(7\right)$

In formula (7):

r＝1,2,3…R；

M is memory depth;

Q is intermodulation exponent number;

W _m,qfor target DPD coefficient; M is memory index, and q is non-linear directory;

The input signal that y (n-m), y (n-m-r) are system;

Z (n) is system output signal.

After determining the DPD model used in the embodiment of the present invention, in order to keep power-balance, feedback signal need eliminate the gain of power amplifier rated linear, obtains signal, specifically see formula (8):

$u_{m,q,r_1,r_2}\left(n\right)=\frac{y(n-m-r_1)}{G}{\left|\frac{y(n-m-r_2)}{G}\right|}^{q-1}\·\·\·\·\·\·\left(8\right)$

In above-mentioned formula (8):

0≤r ₁≤R,0≤r ₂≤R；

Y (n-m-r ₁), y (n-m-r ₂) be feedback signal;

G is the gain of power amplifier rated linear;

for the signal obtained after the gain of feedback signal elimination power amplifier rated linear.

In formula (8), the matrix notation of reference signal is as follows:

z＝Uw……(9)

The least square solution of target DPD coefficient is as shown in formula (10):

$\hat{w}={\left(U^{H}U\right)}^{-1}{U}^{H}z\·\·\·\·\·\·\left(10\right)$

In formula (9) and formula (10):

Z is reference signal; Z=[z (0) ..., z (N-1)] ^t;

U is feedback signal in formula (8) matrix form, $U=[u_{10},\·\·\·u_{M000},\·\·\·u_{1Q},\·\·\·u_{\mathrm{MQ}{R}_1{R}_2}],u_{\mathrm{mq}{r}_1{r}_2}={[u_{\mathrm{mq}{r}_1{r}_2}\left(0\right),u_{\mathrm{mq}{r}_1{r}_2}\left(1\right),\·\·\·u_{\mathrm{mq}{r}_1{r}_2}(N-1)]}^T$

W is target DPD coefficient, $w={[w_{1000},\·\·\·w_{M000},\·\·\·w_{1Q00},\·\·\·w_{\mathrm{MQ}{R}_1{R}_2}]}^T.$

Because the number of sampled point is more than the number of the coefficient of model, therefore formula (9) is over-determined systems.For this reason, the solution of principle of least square determination linear equation in the embodiment of the present invention, can be applied, in Practical Calculation process, adopt the QR decomposition method of matrix or quick Cholesky decomposition method solution matrix coefficient.For PVS model, by configuration memory depth, intermodulation exponent number, interlaced sampling point, the combination of autocorrelation matrix just can be completed.

Power amplifier model in the embodiment of the present invention and the relation between DPD model as follows:

First power amplifier model is through linear time-varying characteristic, is memoryless nonlinear characteristic afterwards; First DPD model is memoryless nonlinear characteristic, then be linear time-varying characteristics, after power amplifier model and DPD models coupling, the signal processed out is exactly the signal after linear model process, therefore, feedback signal after predistortion is judged, accurately can determine the effect of predistortion.

It can be seen from the above, owing to all judging the ACPR of the feedback signal of the original training sequence after DPD process at every turn in the embodiment of the present invention, so, then accurately can judge the effect of each DPD process, and meet the requirements at ACPR, namely when DPD treatment effect meets the requirements, process ends, thus the DPD coefficient being met performance requirement.

Fig. 3 illustrates a kind of structural representation of DPD system.

Based on same idea, the embodiment of the present invention provides a kind of DPD system, as shown in Figure 3, comprise training sequence generation unit 301, the switch element 304 be connected with this training sequence generation unit 301, the DPD processing unit 302 be connected with switch element 304 and radio frequency unit 303, switch element 304 can realize training sequence generation unit 301 and be connected with DPD processing unit, also can realize training sequence generation unit 301 to be directly connected with radio frequency unit 303, the radio frequency unit 303 be connected with DPD processing unit 302, ACPR determining unit 307, DPD coefficient generation unit 305, ACPR determining unit 307 and DPD coefficient generation unit 305 all can get feedback signal 310, the look-up table updating block 306 be connected with DPD coefficient generation unit 305, look-up table updating block 306 can be used for upgrading the look-up table stored, control unit 308, with training sequence generation unit 301 while of control unit 308, switch element 304, ACPR determining unit 307, DPD coefficient generation unit 305 connects.

Preferably, radio frequency unit 303 comprises digital to analog converter 313, the first radio-frequency channel device 314, the power amplifier 315 that connect successively.Preferably, feedback signal 310 enters ACPR determining unit 307 and DPD coefficient generation unit 305 through the second radio-frequency channel device 316, analog to digital converter 317.

Training sequence generation unit 301, for generating original training sequence;

DPD processing unit 302, for carrying out DPD process according to look-up table to original training sequence, and exports to radio frequency unit 303 by the sequence after DPD process and sends;

Switch element 304, for connecting training sequence generation unit 301 selectivity between DPD processing unit 302 and radio frequency unit 303;

DPD coefficient generation unit 305, for generating DPD coefficient according to the feedback signal received and the training sequence corresponding with feedback signal;

Look-up table updating block 306, for the DPD coefficient update look-up table generated according to DPD coefficient generation unit 305;

ACPR determining unit 307, for the ACPR according to the feedback signal determination feedback signal received;

Control unit 308, for after cycle of training starts:

Connection between training sequence generation unit 301 and DPD processing unit 302 disconnects by control switch unit 304, the connection between training sequence generation unit 301 and radio frequency unit 303 closed, original training sequence is sent by radio frequency unit 303, after the feedback signal of original series is received by DPD coefficient generation unit 305, DPD coefficient generation unit 305 generates DPD coefficient according to the feedback signal of original training sequence and original series sequence, and look-up table updating block 306 is according to the DPD coefficient update look-up table generated;

Connection between training sequence generation unit 301 and DPD processing unit 302 closes by control switch unit 304, the connection between training sequence generation unit 301 and radio frequency unit 303 disconnected, training sequence after DPD process is sent by radio frequency unit 303, after the feedback signal of the training sequence after DPD process is received by ACPR determining unit 307, ACPR determining unit 307 determines the ACPR of the feedback signal received; Control unit 308 judges whether ACPR meets the demands, and if so, then terminates the training process of current cycle of training, otherwise instruction training sequence generation unit 301 generates original training sequence again.

Specifically, training sequence generation unit 301 can all regenerate identical original training sequence in each training, also different original training sequence can all be regenerated in each training, also an original training sequence can be generated in the direct training process of a cycle of training, stored afterwards, and all used the original training sequence of this storage in indirect discipline afterwards.It is that example is introduced that embodiment of the present invention following proposal all regenerates identical original training sequence based on training sequence generation unit 301 in each training.

Preferably, control unit 308 is also for counting the frequency of training in current cycle of training; If determine that the frequency of training in current cycle of training reaches maximum times according to count value, then terminate the training process of current cycle of training.Frequency of training in current cycle of training is the number of times of indirect discipline, and the total degree that the number of times of indirect discipline equals to train deducts the number of times of once directly training.The definition of indirect discipline and directly training as previously mentioned.

Preferably, control unit 308 counts by the number of times generating DPD coefficient according to the feedback signal received and the training sequence corresponding with feedback signal to DPD coefficient generation unit 305.Other implementation is, control unit 308 also can count the number of times generating original training sequence in training sequence generation unit 301, counts etc. the update times of look-up table.

Preferably, the ACPR of the feedback signal of the training sequence after DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after DPD process is greater than threshold value.

Preferably, DPD coefficient generation unit 305 specifically for: adopt adaptive algorithm to generate DPD coefficient, adaptive algorithm is any one in following algorithm: least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.

It can be seen from the above: owing to all judging the ACPR of the feedback signal of the original training sequence after DPD process at every turn in the embodiment of the present invention, so, then accurately can judge the effect of each DPD process, and meet the requirements at ACPR, namely when DPD treatment effect meets the requirements, process ends, thus the DPD coefficient being met performance requirement.

Those skilled in the art should understand, embodiments of the invention can be provided as method or computer program.Therefore, the present invention can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the present invention can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disc store, CD-ROM, optical memory etc.) of computer usable program code.

The present invention describes with reference to according to the flow chart of the method for the embodiment of the present invention, equipment (system) and computer program and/or block diagram.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block diagram and/or square frame and flow chart and/or block diagram and/or square frame.These computer program instructions can being provided to the processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing device to produce a machine, making the instruction performed by the processor of computer or other programmable data processing device produce system for realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command system, and this command system realizes the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

These computer program instructions also can be loaded in computer or other programmable data processing device, make on computer or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computer or other programmable devices is provided for the step realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.

Although describe the preferred embodiments of the present invention, those skilled in the art once obtain the basic creative concept of cicada, then can make other change and amendment to these embodiments.So claims are intended to be interpreted as comprising preferred embodiment and falling into all changes and the amendment of the scope of the invention.

Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.

Claims (9)

1. a digital pre-distortion DPD coefficient training method, is characterized in that, comprising:

Steps A: sent by radio frequency unit by original training sequence, receives the feedback signal of described original training sequence, and the feedback signal according to described original training sequence and described original training sequence generates DPD coefficient, according to the DPD coefficient update look-up table generated;

Step B: carry out DPD process according to the DPD coefficient in described look-up table to described original training signal, sent by the training sequence after DPD process by radio frequency unit, receives the feedback signal of the training sequence after described DPD process;

Step C: judge whether the Adjacent Channel Power Ratio ACPR of the feedback signal of the training sequence after described DPD process meets the demands, and if so, then terminates the training process of current cycle of training, otherwise proceeds to step D;

Step D: the feedback signal according to the training signal after the training sequence after described DPD process and described DPD process generates DPD coefficient, according to the DPD coefficient update look-up table generated, and proceeds to step B.

2. the method for claim 1, is characterized in that, before proceeding to step B, also comprises:

Judge whether the frequency of training in current cycle of training reaches maximum times;

Describedly proceed to step B, be specially: if the frequency of training in current cycle of training does not reach maximum times, then proceed to step B.

3. method as claimed in claim 2, is characterized in that, also comprise:

If the frequency of training in current cycle of training reaches maximum times, then terminate the training process of current cycle of training.

4. method as claimed any one in claims 1 to 3, it is characterized in that, the ACPR of the feedback signal of the training sequence after described DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after described DPD process is greater than threshold value.

5. as claimed any one in claims 1 to 3 method, is characterized in that, adopt adaptive algorithm to generate DPD coefficient, and described adaptive algorithm is any one in following algorithm:

Least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.

6. a digital pre-distortion DPD system, is characterized in that, comprising:

Training sequence generation unit, for generating original training sequence;

DPD processing unit, for carrying out DPD process according to look-up table to described original training sequence, and exports to radio frequency unit by the sequence after DPD process and sends;

Switch element, for connecting described training sequence generation unit selectivity between described DPD processing unit and radio frequency unit;

DPD coefficient generation unit, for generating DPD coefficient according to the feedback signal received and the training sequence corresponding with described feedback signal;

Look-up table updating block, for the DPD coefficient update look-up table generated according to described DPD coefficient generation unit;

Adjacent Channel Power Ratio ACPR determining unit, for determining the ACPR of described feedback signal according to the feedback signal received;

Control unit, for after cycle of training starts:

Control described switch element the connection between described training sequence generation unit and DPD processing unit to be disconnected, the connection between described training sequence generation unit and radio frequency unit is closed, described original training sequence is sent by described radio frequency unit, the feedback signal of described original series is by after described DPD coefficient generation unit reception, described DPD coefficient generation unit generates DPD coefficient according to the feedback signal of described original training sequence and described original series sequence, and described look-up table updating block is according to the DPD coefficient update look-up table generated;

Control described switch element the connection between described training sequence generation unit and described DPD processing unit to be closed, the connection between described training sequence generation unit and radio frequency unit is disconnected, training sequence after described DPD process is sent by described radio frequency unit, the feedback signal of the training sequence after described DPD process is by after described ACPR determining unit reception, and described ACPR determining unit determines the ACPR of the feedback signal received; Described control unit judges whether described ACPR meets the demands, and if so, then terminates the training process of current cycle of training, otherwise indicates described training sequence generation unit again to generate original training sequence.

7. DPD system as claimed in claim 1, is characterized in that, described control unit, also for

Frequency of training in current cycle of training is counted;

If determine that the frequency of training in current cycle of training reaches maximum times according to count value, then terminate the training process of current cycle of training.

8. DPD system as claimed in claims 6 or 7, it is characterized in that, the ACPR of the feedback signal of the training sequence after described DPD process meets the demands, and refers to that the ACPR absolute value of the feedback signal of the training sequence after described DPD process is greater than threshold value.

9. DPD system as claimed in claims 6 or 7, it is characterized in that, described DPD coefficient generation unit specifically for: adopt adaptive algorithm to generate DPD coefficient, described adaptive algorithm is any one in following algorithm: least-squares algorithm, least mean square algorithm, recursive least squares, least-mean-square error algorithm.