CN107612856A  A kind of digital predistortion processing method and device  Google Patents
A kind of digital predistortion processing method and device Download PDFInfo
 Publication number
 CN107612856A CN107612856A CN201710934026.7A CN201710934026A CN107612856A CN 107612856 A CN107612856 A CN 107612856A CN 201710934026 A CN201710934026 A CN 201710934026A CN 107612856 A CN107612856 A CN 107612856A
 Authority
 CN
 China
 Prior art keywords
 matrix
 structural matrix
 default
 structural
 3rd
 Prior art date
Links
 238000003672 processing method Methods 0 abstract title 2
 239000011159 matrix materials Substances 0 abstract 11
 238000010276 construction Methods 0 abstract 6
 239000000562 conjugates Substances 0 abstract 1
 230000000875 corresponding Effects 0 abstract 1
 230000015654 memory Effects 0 abstract 1
 239000000047 products Substances 0 abstract 1
Classifications

 G—PHYSICS
 G06—COMPUTING; CALCULATING; COUNTING
 G06F—ELECTRIC DIGITAL DATA PROCESSING
 G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
 G06F17/10—Complex mathematical operations
 G06F17/16—Matrix or vector computation, e.g. matrixmatrix or matrixvector multiplication, matrix factorization

 H—ELECTRICITY
 H04—ELECTRIC COMMUNICATION TECHNIQUE
 H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
 H04L25/00—Baseband systems
 H04L25/38—Synchronous or startstop systems, e.g. for Baudot code
 H04L25/40—Transmitting circuits; Receiving circuits
 H04L25/49—Transmitting circuits; Receiving circuits using code conversion at the transmitter; using predistortion; using insertion of idle bits for obtaining a desired frequency spectrum; using three or more amplitude levels ; Baseband coding techniques specific to data transmission systems
Abstract
Description
Technical field
The application is related to communication technical field, more particularly to a kind of digital predistortion processing method and device.
Background technology
With the development of radio communication, the fluctuating of its envelope is increasing with peaktoaverage force ratio, and RF system circuit produces substantial amounts of Nonlinear distortion, reduce communication quality.The nonlinear distortion of signal be mainly derived from power amplifier (Power Amplifier, PA nonlinear, electric memory effect and dynamic nonlinear effect especially caused by wideband radio frequency power amplifier).It is right at present The linearization technique of power amplifier mainly includes, and manufacture linearizes higher power amplifier apparatus, method for suppressing peak to average ratio With digital predistortion (Digital PreDistortion, DPD) method etc..Wherein, high, the applicable band of DPD methods stability Wide scope is wide, precision is high, realizes that difficulty is low.It is main that a distortion with input signal is produced using predistorter in DPD methods The opposite signal of characteristic, so as to offset distortion component, is corrected to distorted signal, is then sent to the signal after correction Power amplifier is amplified output, so as to obtain undistorted signal output.
But in actual applications, as bandwidth is increasing, the digital predistortion model of power amplifier is also increasingly Complexity, in some application scenarios, it may appear that DPD solution has unstable phenomenon, can be caused when serious to power amplifier forever Property damage.
In summary, how to improve the stability of DPD solutions is a urgent problem to be solved.
The content of the invention
A kind of digital predistortion processing method of the application offer and device, the memory polynomial model based on DPD, according to The conditional number of the predistortion coefficient matrix of DPD memory polynomial, judge the stability of DPD solutions, realize the memory according to DPD The conditional number of polynomial predistortion coefficient matrix, determine that conditional number is less than the predistortion coefficient matrix of threshold value, reduce DPD's The unstable possibility of solution, ensure that the stability of power amplifier output signal, improves the overall performance of communication equipment.
The embodiment of the present application provides a kind of digital predistortion processing method, including：
Obtain input signal, the input signal of power amplifier and the feedback signal of digital predistortion DPD module；
First structural matrix F is determined according to the feedback signal and default predistortion memory polynomial_{1}；
According to the first structural matrix F_{1}Determine the first matrix U_{1}, however, it is determined that first matrix U_{1}Conditional number be more than Default first threshold, then to the first structural matrix F_{1}It is adjusted, until according to the first construction square after adjustment Battle array F_{1}The conditional number of determination is less than default Second Threshold；First matrix U_{1}For the first structural matrix F_{1}Conjugate torque Battle array and the first structural matrix F_{1}Product；
According to the first structural matrix F after the adjustment_{1}Corresponding digital predistortion coefficient matrices A is to the input signal Carry out predistortion.
A kind of possible implementation, it is described if it is determined that first matrix U_{1}Conditional number be more than default first threshold Value, then to the first structural matrix F_{1}It is adjusted, until according to the first structural matrix F after adjustment_{1}The condition of determination Number is less than default Second Threshold, including：
Maximum order K and maximal memory depth L corresponding to the default predistortion memory polynomial are adjusted to respectively Default first maximum order K_{2}With default first maximal memory depth L_{2}, wherein the K_{2}<K, L_{2}<L；
According to the K_{2}And L_{2}Obtain the second structural matrix F_{2}, and the second structural matrix F_{2}Corresponding second matrix U_{2}'s Conditional number is less than the default Second Threshold, second matrix U_{2}For the second structural matrix F_{2}Conjugate matrices and institute State the second structural matrix F_{2}Product；
Element in matrix Q is sequentially added into the second structural matrix F_{2}The 3rd structural matrix F of middle acquisition_{3}, and calculate every Add the 3rd structural matrix F obtained during an element_{3}Corresponding 3rd matrix U_{3}Conditional number；3rd matrix U_{3}For institute State the 3rd structural matrix F_{3}Conjugate matrices and the 3rd structural matrix F_{3}Product；The matrix Q is by appearing in described first Structural matrix F_{1}And the second structural matrix F is not appeared in_{2}Element composition；
By all 3rd matrix Us_{3}Conditional number be ranked up, obtain minimum conditional number and the minimum condition 3rd structural matrix F corresponding to number_{3}；
Judge whether the minimum conditional number is less than the default Second Threshold, if so, then with the minimum bar 3rd structural matrix F corresponding to number of packages_{3}As the second structural matrix F_{2}, and recalculate the 3rd structural matrix F_{3}With the 3rd Structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number the step of；If it is not, then by the second structural matrix F_{2}As described One structural matrix F_{1}The matrix obtained after adjustment.
Optionally, the default Second Threshold is smaller 0.3dB than the default first threshold.
A kind of possible implementation, it is described if it is determined that first matrix U_{1}Conditional number be more than default first threshold Value, then to the first structural matrix F_{1}It is adjusted, until according to the first structural matrix F after adjustment_{1}The condition of determination Number is less than default Second Threshold, including：
Maximum order K and maximal memory depth L corresponding to the default predistortion memory polynomial are adjusted to respectively Second default maximum order K_{3}With the second default maximal memory depth L_{3}, wherein the K_{3}<K, the L_{3}<L；
According to the K_{3}And L_{3}Obtain the 4th structural matrix F_{4}, the 4th structural matrix F_{4}For the first structural matrix F_{1} The matrix obtained after adjustment, and the 4th structural matrix F_{4}Corresponding 4th matrix U_{4}Conditional number be less than described default the Two threshold values；4th matrix U_{4}For the 4th structural matrix F_{4}Conjugate matrices and the 4th structural matrix F_{4}Product.
Optionally, the first structural matrix F_{1}By element F_{10}To element F_{KL}Composition, wherein for either element F_{kl}By with Lower formula determines：
F_{kl}=y (nl)  y (nl) ^{k1}
Wherein, y (nl) is the sample sequence of the feedback signal of the power amplifier, and n span is [0, M1], M is the sampling sum of feedback signal, and K is maximum order, and k is polynomial order, and span, between [1, K], l is memory Depth, and span is [0, L], L is maximal memory depth.
First matrix U_{1}Determined by below equation：
First matrix U_{1}Conditional number C_{1}Determined by below equation：
C_{1}=cond (U_{1})=  U_{1}U_{1} ^{1}。
In specific implementation process, the predistortion, including：
According to the first structural matrix F and PA input signal z (n), predistortion coefficient matrix A is determined：A= (F^{H}F)^{1}F^{H}Z；
According to predistortion coefficient matrix A and the PA input signal, new PA input signal is generated.
The embodiment of the present application provides a kind of digital predistortion process apparatus, and described device includes：
Data acquisition unit, for obtaining input signal, the input signal of power amplifier of digital predistortion DPD module And feedback signal；
DPD coefficient matrix determining units, are used for：
First structural matrix F is determined according to the feedback signal and default predistortion memory polynomial_{1}；
According to the first structural matrix F_{1}Determine the first matrix U_{1}, however, it is determined that first matrix U_{1}Conditional number be more than Default first threshold, then to the first structural matrix F_{1}It is adjusted, until according to the first construction square after adjustment Battle array F_{1}The conditional number of determination is less than default Second Threshold；First matrix U_{1}For the first structural matrix F_{1}Conjugate torque Battle array and the product of the first structural matrix F；
DPD processing units, for according to the first structural matrix F after the adjustment_{1}Corresponding digital predistortion coefficient square Battle array A carries out predistortion to the input signal.
A kind of possible implementation, the DPD coefficient matrixes determining unit, is specifically used for：
Maximum order K and maximal memory depth L corresponding to the default predistortion memory polynomial are adjusted to respectively Default first maximum order K_{2}With default first maximal memory depth L_{2}, wherein the K_{2}<K, L_{2}<L；
According to the K_{2}And L_{2}Obtain the second structural matrix F_{2}, and the second structural matrix F_{2}Corresponding second matrix U_{2}'s Conditional number is less than the default Second Threshold；Second matrix U_{2}For the second structural matrix F_{2}Conjugate matrices and institute State the second structural matrix F_{2}Product；
Element in matrix Q is sequentially added into the second structural matrix F_{2}The 3rd structural matrix F of middle acquisition_{3}, and calculate every Individual 3rd structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number；3rd matrix U_{3}For the 3rd structural matrix F_{3}'s Conjugate matrices and the 3rd structural matrix F_{3}Product；The matrix Q is by appearing in the first structural matrix F_{1}And do not occur In the second structural matrix F_{2}Element composition；
By all 3rd matrix Us_{3}Conditional number be ranked up, obtain minimum conditional number and the minimum condition 3rd structural matrix F corresponding to number_{3}；
Judge whether the minimum conditional number is less than the default Second Threshold, if so, then with the minimum bar 3rd structural matrix F corresponding to number of packages_{3}As the second structural matrix F_{2}, and recalculate the 3rd structural matrix F_{3}With the 3rd Structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number the step of；If it is not, then by the second structural matrix F_{2}As described One structural matrix F_{1}The matrix obtained after adjustment.
Optionally, the DPD coefficient matrixes determining unit, is specifically used for：The default Second Threshold is more default than described The small 0.3dB of first threshold.
A kind of possible implementation, the DPD coefficient matrixes determining unit, is specifically used for：
The first structural matrix F_{1}By element F_{10}To element F_{KL}Composition, wherein for either element F_{kl}By below equation It is determined that：
F_{kl}=y (nl)  y (nl) ^{k1}
Wherein, y (nl) is the sample sequence of the feedback signal of the power amplifier, and n span is [0, M1], M is the sampling sum of feedback signal, and K is maximum order, and the integer of k value between [1, K], l value is between [0, L] Integer, L is maximal memory depth.
First matrix U_{1}Determined by below equation：
First matrix U_{1}Conditional number C_{1}Determined by below equation：
C_{1}=cond (U_{1})=  U_{1}U_{1} ^{1}。
The DPD processing units, are specifically used for：
According to the first structural matrix F and PA input signal z (n), predistortion coefficient matrix A is determined：A= (F^{H}F)^{1}F^{H}Z；
In specific implementation process, according to predistortion coefficient matrix A and the input signal, the input for generating new PA is believed Number.
Using such scheme, due to the conditional number of the predistortion coefficient matrix of the memory polynomial according to DPD, DPD is judged The stability of solution, the conditional number of the predistortion coefficient matrix of memory polynomial according to DPD is realized, determines that conditional number is less than threshold The predistortion coefficient matrix of value, the unstable possibility of DPD solution is reduced, ensure that the stabilization of power amplifier output signal Property, improve the overall performance of communication equipment.
Brief description of the drawings
Fig. 1 is a kind of schematic architectural diagram for digital predistortion processing method that the embodiment of the present application provides；
Fig. 2 is a kind of digital predistortion processing method schematic flow sheet that the embodiment of the present application provides；
Fig. 3 is a kind of digital predistortion process apparatus structural representation that the embodiment of the present application provides.
Embodiment
The application is described in further detail below in conjunction with accompanying drawing.
Fig. 1 is a kind of structural representation for digital predistortion processing method that the embodiment of the present application provides.As shown in Figure 1 In the structural representation of digital predistortion method, mainly include DPD processing units 101, PA102, data acquisition unit 103, DPD Coefficient matrix determining unit 104.Input signal generates PA input signal, the input signal of the PA by DPD processing units Amplify generation PA output signal by PA, and generate feedback signal, data acquisition unit 103 is defeated by the PA's of acquisition Enter signal and feedback signal, be sent to DPD coefficient matrixes determining unit 104, DPD coefficient matrixes determining unit 104 is according to described PA input signal and feedback signal determines required DPD coefficient matrixes, and generation DPD coefficient matrixes are sent to DPD processing Unit 101.Fig. 1 is the rough schematic view illustrated, and can also include other equipment in system, not drawn in Fig. 1.
In order to more fully understand the application, the application is illustrated below with reference to accompanying drawing.
With reference to foregoing description, referring to Fig. 2, a kind of digital predistortion processing method flow provided for the embodiment of the present application is shown It is intended to.This method comprises the following steps：
Step 201：Obtain input signal, the input signal of power amplifier and the feedback letter of digital predistortion DPD module Number；
Step 202：First structural matrix F is determined according to the feedback signal and default predistortion memory polynomial；
Step 203：According to the first structural matrix F_{1}Determine the first matrix U_{1}, however, it is determined that first matrix U_{1}Bar Number of packages is more than default first threshold, then to the first structural matrix F_{1}It is adjusted, until according to described the after adjustment One structural matrix F_{1}The conditional number of determination is less than default Second Threshold；First matrix U_{1}For the first structural matrix F_{1} Conjugate matrices and the first structural matrix F_{1}Product；
Step 204：According to the first structural matrix F after the adjustment_{1}Corresponding digital predistortion coefficient matrices A is to described Input signal carries out predistortion.
With reference to Fig. 1, in step 201, the input signal of the PA is by input of the DPD processing units 101 to DPD module Signal obtains after carrying out predistortion, and the feedback signal is putting according to the PA102 PA exported input signal What big signal obtained.
Specifically, assume it is current there is input signal to input to DPD processing units 101, then DPD processing units 101, to input Input signal carry out digital predistortion processing after obtain PA input signal, and export to PA102.By PA102 to the defeated of PA Enter signal be amplified processing after obtain PA input signal amplified signal, by amplified signal divided by PA102 multiplication factor It can obtain feedback signal.
In step 202, it is assumed that the PA currently got input signal is z (n), n=0,1,2 ... M1, and feedback signal is Y (n), n=0,1,2 ... M1, wherein, M is PA input signal and the sampling total length of feedback signal, then with memory polynomial Exemplified by model, below equation is obtained：
Wherein, a_{kl}For predistortion coefficients to be determined, z (n) is the sample sequence of PA input signal, and y (n) is feedback signal Signal sample sequence, k value is the integer in [1, K] section, and l value is the integer in [0, L] section.Maximum order K, maximal memory depth L is the predistortion parameters in the predistortion memory polynomial, can be set according to actual conditions.
The form of formula [1] matrix of being write as is：
Z=F_{1}A [2]
Wherein：
Z=[z (0), z (1), z (2), z (3) ..., z (M1)]^{T} [3]
Wherein, the first structural matrix F_{1}For：
F_{1}=[F_{10},F_{20},...,F_{K0},F_{11},...,F_{1L},...,F_{KL}] [4]
First structural matrix F_{1}In by element F_{10}To element F_{KL}Composition, wherein for either element F_{kl}For：
F_{kl}=[F_{kl}(0),F_{kl}(1),...,F_{kl}(M1)]^{T} [5]
Wherein, F_{kl}(n) it is the first structural matrix F_{kl}In n coefficient value.
For example, the first structural matrix F_{1}Element F_{10}For：
F_{10}=[F_{10}(0),F_{10}(1),...,F_{10}(M1)]^{T}
Wherein, F_{10}(n) it is the first structural matrix F_{10}In n coefficient value.
F_{kl}(n)=y (nl)  y (nl) ^{k1} [6]
In step 203, first matrix U_{1}Determined by below equation：
Wherein, F_{1} ^{H}For F_{1}Conjugate matrices.
First matrix U_{1}Conditional number C_{1}Determined by below equation：
C_{1}=cond (U_{1})=  U_{1}U_{1} ^{1} [8]
Wherein, U_{1} ^{1}For U_{1}Inverse matrix,     to take norm operator.
If the first matrix F_{1}Conditional number C_{1}It is more than default first threshold, then the default predistortion memory is multinomial Maximum order K corresponding to formula and maximal memory depth L is adjusted to default first maximum order K respectively_{2}With default first most Big memory depth L_{2}, wherein the K_{2}<K, L_{2}<L；
According to the K_{2}And L_{2}Generate the second structural matrix F_{2}；The second structural matrix F_{2}Determination method and described the One structural matrix F_{1}Determination method it is identical, difference is the second structural matrix F_{2}In maximum order K value adjustment For default first maximum order K_{2}, maximal memory depth L value is adjusted to default first maximal memory depth L_{2}, and institute State the second structural matrix F_{2}Corresponding second matrix U_{2}Conditional number be less than the default Second Threshold.
In a kind of possible implementation, the default first threshold is obtained by many experiments, when digital pre When unusual fluctuations occurs in anamorphic system, PA input signal and predistortion feedback signal, and conditional number corresponding to determination are recorded, in advance If first threshold be according to multiple abnormal data determine conditional number average value.The default Second Threshold can be equal to Or less than the default first threshold.
In step 203, the second structural matrix F_{2}For：
Wherein, the second structural matrix F_{2}Second matrix U determined_{2}Conditional number C_{2}Less than described default Two threshold values.
A kind of possible implementation, to improve the precision of predistortion model, if second matrix U_{2}Conditional number it is small In the default Second Threshold, the conditional number of second matrix after ensureing to adjust still meets to be less than default second threshold Under conditions of value, the second structural matrix F can be updated in the following manner_{2}：
Element in matrix Q is sequentially added into the second structural matrix F_{2}The 3rd structural matrix F of middle acquisition_{3}, and calculate every Add the 3rd structural matrix F obtained during an element_{3}Corresponding 3rd matrix U_{3}Conditional number；3rd matrix U_{3}For institute State the 3rd structural matrix F_{3}Conjugate matrices and the 3rd structural matrix F_{3}Product；The matrix Q is by appearing in described first Structural matrix F_{1}And the second structural matrix F is not appeared in_{2}Element composition；
By all 3rd matrix Us_{3}Conditional number be ranked up, obtain minimum conditional number and the minimum condition 3rd structural matrix F corresponding to number_{3}；
Judge whether the minimum conditional number is less than the default Second Threshold, if so, then with the minimum bar 3rd structural matrix F corresponding to number of packages_{3}To the second structural matrix F_{2}It is replaced, the second structural matrix F after being replaced_{2}, and The 3rd structural matrix F will be recalculated_{3}With the 3rd structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number the step of；If it is not, then By the second structural matrix F_{2}The matrix obtained after being adjusted as first structural matrix.
For example, however, it is determined that second matrix U_{2}Conditional number C_{2}Less than default Second Threshold, then the matrix Q For：
According to the second structural matrix F_{2}With the matrix Q, N number of 3rd structural matrix F is obtained_{3}, specifically, described Three structural matrix F_{3}For：
Wherein, F_{i}For a coefficient in matrix Q；N is matrix Q number of coefficients.Each 3rd structural matrix F_{3}From matrix The coefficient chosen in Q is different, therefore, the 3rd structural matrix F_{3}There are N kinds may.
Each 3rd structural matrix F_{3}The 3rd matrix U can be determined according to formula [7]_{3}, each 3rd matrix U_{3}Root again Conditional number C is determined according to formula [8]_{3}, N number of conditional number C can be determined altogether_{3}, determine N number of conditional number C_{3}In minimum bar Number of packages.
If the minimal condition number is less than the default Second Threshold, by the second structural matrix F_{2}It is updated to institute State the 3rd structural matrix F corresponding to minimal condition number_{3}, will the second structural matrix F_{2}Corresponding to the minimal condition number 3rd structural matrix F_{3}Replace, obtain and realize to the second structural matrix F_{2}Renewal.The second structural matrix F is updated_{2} Afterwards, return to and calculate the 3rd structural matrix F_{3}With the 3rd structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number the step of, after The second structural matrix F of continuous adjustment_{2}。
Accordingly, if the minimal condition number is more than or equal to the default Second Threshold, the institute can be determined State the second structural matrix F_{2}For required structural matrix, so as to according to the second structural matrix F_{2}Corresponding predistortion Coefficient carries out predistortion to the input signal.
The first structural matrix F after the adjustment obtained according to this method_{1}In element, after the adjustment first construction Matrix F_{1}Corresponding first matrix meets that the conditional number is less than under conditions of the Second Threshold, first structure after adjustment Make matrix F_{1}The element of corresponding digital predistortion coefficient is enough, greatly improves the precision of digital predistortion model, and And it ensure that the stability of digital predistortion model.
In step 203, a kind of possible implementation, will be maximum corresponding to the default predistortion memory polynomial Exponent number K and maximal memory depth L is adjusted to the second default maximum order K respectively_{3}With the second default maximal memory depth L_{3}, wherein The K_{3}<K, the L_{3}<L；
According to the K_{3}And L_{3}Obtain the 4th structural matrix F_{4}, the 4th structural matrix F_{4}For the first structural matrix F_{1} The matrix obtained after adjustment, and the 4th structural matrix F_{4}Corresponding 4th matrix U_{4}Conditional number be less than described default the Two threshold values；4th matrix U_{4}For the 4th structural matrix F_{4}Conjugate matrices and the 4th structural matrix F_{4}Product.
According to the first structural matrix F after the adjustment_{1}Corresponding digital predistortion coefficient matrices A is to the input signal Carry out predistortion.
The embodiment of the present application provides a kind of digital predistortion processing method, comprises the following steps：
Step 1：Assuming that it is 4 that K values, which are 7, L values, then the first structural matrix F is determined according to formula [4]_{1}For：
F_{1}=[F_{10},F_{20},…,F_{70},F_{11},…,F_{71},…F_{14},…,F_{74}]
Step 2：If the first structural matrix F_{1}Corresponding first matrix U_{1}Conditional number C_{1}More than default first threshold, then Generate the second structural matrix F_{2}。
Assuming that K_{2}For 6, L_{2}For 3, then the second structural matrix F_{2}For：
F_{2}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63}]
Step 3：Second structural matrix F_{2}Second matrix U determined_{2}Conditional number C_{2}Less than default Second Threshold, Then according to the second structural matrix F_{2}Predistortion coefficient matrix is determined with the input signal of the PA, according to predistortion coefficients square Battle array carries out predistortion to the input signal.
The embodiment of the present application provides a kind of digital predistortion processing method, comprises the following steps：
Step 1：Assuming that it is 4 that K values, which are 7, L values, then the first structural matrix F is determined according to formula [4]_{1}For：
F_{1}=[F_{10},F_{20},…,F_{70},F_{11},…,F_{71},…F_{14},…,F_{74}]
Step 2：If the first structural matrix F_{1}Corresponding first matrix U_{1}Conditional number C_{1}More than default first threshold, then Generate the second structural matrix F_{2}。
Assuming that K_{2}For 6, L_{2}For 3, then the second structural matrix F_{2}For：
F_{2}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63}]
Step 3：If it is determined that second matrix U_{2}Conditional number C_{2}Less than default Second Threshold, then matrix Q is：
Q=[F_{70},F_{71},F_{72},F_{73},F_{74},F_{14},F_{24},F_{34},F_{44},F_{54},F_{64}]
According to the second structural matrix F_{2}With matrix Q, 11 the 3rd structural matrix F are obtained_{3}, specifically, the 3rd structure Make matrix F_{3}For：
F_{301}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63},F_{70}]
…
F_{311}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63},F_{74}]
Each 3rd structural matrix F_{3}The 3rd matrix U can be determined according to formula [7]_{3}, each 3rd matrix U_{3}Root again Conditional number C is determined according to formula [8]_{3}, 11 conditional number C can be determined altogether_{3}, determine 11 conditional number C_{3}In minimum Conditional number number C_{311}。
Step 4：If the minimal condition number C_{311}Less than the default Second Threshold, then with the 3rd structural matrix F_{311} To the second structural matrix F_{2}It is replaced, the second structural matrix F after being replaced_{2}, and step 3 is back to, if minimal condition Number C_{311}More than or equal to the default Second Threshold, then step 5 is gone to.Detailed process is as follows：
Now, the second structural matrix F_{2}For：
F_{2}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63},F_{74}]
Matrix Q is：
Q=[F_{70},F_{71},F_{72},F_{73},F_{14},F_{24},F_{34},F_{44},F_{54},F_{64}]
According to the second structural matrix F_{2}With the 3rd structural matrix F_{3}, obtain 10 the 4th structural matrix F_{3}, specifically It is as follows：
F_{301}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63},F_{74},F_{70}]
…
F_{310}=[F_{10},F_{20},...,F_{60},F_{11},...,F_{13},...,F_{63},F_{74},F_{64}]
Each 3rd structural matrix F_{3}The 3rd matrix U can be determined according to formula [7]_{3}, each 3rd matrix U_{3}Root again Conditional number C is determined according to formula [8]_{3}, 10 conditional number C can be determined altogether_{3}, determine 10 conditional number C_{3}In minimum Conditional number.Assuming that F_{310}Corresponding conditional number is minimal condition number C_{310}。
If minimal condition number C_{310}More than or equal to the default Second Threshold, then according to the second structural matrix F_{2} (now, second structural matrix has been updated to F_{311}), predistortion is carried out to the input signal；
If the minimal condition number C_{310}Less than the default Second Threshold, then by the second structural matrix F_{2}Renewal For the 3rd structural matrix F_{310}, and return to step 3.Detailed process is same as described above, will not be repeated here.
Circulation step three and step 4, until the 3rd structural matrix F of generation_{3}Corresponding 3rd matrix U_{3}Minimal condition Number is more than or equal to the default Second Threshold, stops updating second structural matrix.
Step 5：According to the second structural matrix F_{2}Predistortion is carried out to the input signal.
Optionally, to avoid frequently carrying out step 203, default Second Threshold could be arranged to be less than default first threshold Value, during practical application, default Second Threshold value can be smaller 0.3dB than default first threshold.
In step 204, predistortion coefficient matrix A is：
A=[a_{10},a_{20},...,a_{K0},a_{11},...,a_{1L},...,a_{KL}]^{T}[12]
The solution of formula [2] is equivalent to the solution of below equation group：
F_{1} ^{H}F_{1}A=F_{1} ^{H}Z[13]
Wherein, F_{1} ^{H}For F_{1}Conjugate matrices, work as F_{1} ^{H}F_{1}Can the inverse time, then can obtain：
A=(F_{1} ^{H}F_{1})^{1}F_{1} ^{H}Z[14]
Wherein, (F_{1} ^{H}F_{1})^{1}For F_{1} ^{H}F_{1}Inverse matrix.
New PA input signal is generated according to predistortion coefficient matrix A and the input signal.Formula loses in advance in [14] The specific determination method of true coefficient matrix A, the embodiment of the present application are not limited this, will not be repeated here.
As shown in figure 3, the embodiment of the present application provides a kind of digital predistortion process apparatus, described device includes：
Data acquisition unit 301, the input of input signal, power amplifier for obtaining digital predistortion DPD module Signal and feedback signal；
DPD coefficient matrixes determining unit 302, is used for：
First structural matrix F is determined according to the feedback signal and default predistortion memory polynomial_{1}；
According to the first structural matrix F_{1}Determine the first matrix U_{1}, however, it is determined that first matrix U_{1}Conditional number be more than Default first threshold, then to the first structural matrix F_{1}It is adjusted, until according to the first construction square after adjustment Battle array F_{1}The conditional number of determination is less than default Second Threshold；First matrix U_{1}For the first structural matrix F_{1}Conjugate torque Battle array and the product of the first structural matrix F；
DPD processing units 303, for according to the first structural matrix F after the adjustment_{1}Corresponding digital predistortion coefficient Matrix A carries out predistortion to the input signal.
A kind of possible implementation, DPD coefficient matrixes determining unit 302, is specifically used for：
Maximum order K and maximal memory depth L corresponding to the default predistortion memory polynomial are adjusted to respectively Default first maximum order K_{2}With default first maximal memory depth L_{2}, wherein the K_{2}<K, L_{2}<L；
According to the K_{2}And L_{2}Obtain the second structural matrix F_{2}, and the second structural matrix F_{2}Corresponding second matrix U_{2}'s Conditional number is less than the default Second Threshold；Second matrix U_{2}For the second structural matrix F_{2}Conjugate matrices and institute State the second structural matrix F_{2}Product；
Element in matrix Q is sequentially added into the second structural matrix F_{2}The 3rd structural matrix F of middle acquisition_{3}, and calculate every Individual 3rd structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number；3rd matrix U_{3}For the 3rd structural matrix F_{3}'s Conjugate matrices and the 3rd structural matrix F_{3}Product；The matrix Q is by appearing in the first structural matrix F_{1}And do not occur In the second structural matrix F_{2}Element composition；
By all 3rd matrix Us_{3}Conditional number be ranked up, obtain minimum conditional number and the minimum condition 3rd structural matrix F corresponding to number_{3}；
Judge whether the minimum conditional number is less than the default Second Threshold, if so, then with the minimum bar 3rd structural matrix F corresponding to number of packages_{3}To the second structural matrix F_{2}It is replaced, the second structural matrix F after being replaced_{2}, and Recalculate the 3rd structural matrix F_{3}With the 3rd structural matrix F_{3}Corresponding 3rd matrix U_{3}Conditional number the step of；If it is not, then will The second structural matrix F_{2}The matrix obtained after being adjusted as first structural matrix.
Optionally, DPD coefficient matrixes determining unit 302, is specifically used for：The default Second Threshold is more default than described The small 0.3dB of first threshold.
A kind of possible implementation, DPD coefficient matrixes determining unit 302, is specifically used for：
The first structural matrix F_{1}By element F_{10}To element F_{KL}Composition, wherein for either element F_{kl}By below equation It is determined that：
F_{kl}=y (nl)  y (nl) ^{k1}
Wherein, y (nl) is the sample sequence of the feedback signal of the power amplifier, and n span is [0, M1], M is the sampling sum of feedback signal, and K is maximum order, and the integer of k value between [1, K], l value is between [0, L] Integer, L is maximal memory depth.
First matrix U_{1}Determined by below equation：
First matrix U_{1}Conditional number C_{1}Determined by below equation：
C_{1}=cond (U_{1})=  U_{1}U_{1} ^{1}。
DPD processing units 303, are specifically used for：
According to the first structural matrix F_{1}With the input signal z (n) of the PA, predistortion coefficient matrix A is determined：A= (F_{1} ^{H}F_{1})^{1}F_{1} ^{H}Z；
In specific implementation process, according to predistortion coefficient matrix A and the input signal, the input for generating new PA is believed Number.
In the embodiment of the present application, by judging whether the conditional number of the first matrix is more than default first threshold, to judge The stability of DPD solutions, if the conditional number of the first matrix is more than default first threshold, generates the second structural matrix.If second The conditional number of second matrix corresponding to structural matrix is more than default Second Threshold, then determines predistortion according to the second structural matrix Coefficient matrix；If the conditional number of the second matrix corresponding to the second structural matrix is less than default Second Threshold, according to the first structure Matrix is made to select the coefficient of the second structural matrix.In prior art, according to the different maximum order of selection and most The value of big memory depth, it is determined that optimal DPD solution.The method that the embodiment of the present application uses is more flexible, avoids repetition meter Predistortion coefficient matrix is calculated, greatly reduces amount of calculation, more stable DPD solution can be obtained faster, reduce DPD Solution unstable possibility, ensure that the stability of power amplifier output signal, improve the overall performance of communication equipment.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or determine machine program Product.Therefore, the application can use the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware Apply the form of example.Moreover, the application can use the determination machine for wherein including determination machine usable program code in one or more The shape for the determination machine program product that usable storage medium is implemented on (including but is not limited to magnetic disk storage, optical memory etc.) Formula.
The application be with reference to according to the present processes, equipment (system) and determine machine program product flow chart and/or Block diagram describes.It should be understood that can by each flow in determination machine programmed instruction implementation process figure and/or block diagram and/or Square frame and the flow in flow chart and/or block diagram and/or the combination of square frame.These determination machine programmed instruction can be provided to arrive General determination machine, special determination machine, the processor of Embedded Processor or other programmable data processing devices are to produce one Machine so that produced by the instruction of determination machine or the computing device of other programmable data processing devices and flowed for realizing The device for the function of being specified in one flow of journey figure or multiple flows and/or one square frame of block diagram or multiple square frames.
These determination machine programmed instruction, which may be alternatively stored in, can guide determination machine or other programmable data processing devices with spy Determine in the determination machine readable memory that mode works so that the instruction being stored in the determination machine readable memory, which produces, to be included referring to Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one square frame of block diagram or The function of being specified in multiple square frames.
These determination machine programmed instruction can be also loaded into determination machine or other programmable data processing devices so that true Determine on machine or other programmable devices perform series of operation steps with produce determination machine realization processing, so as to it is determined that machine or The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one The step of function of being specified in individual square frame or multiple square frames.
Obviously, those skilled in the art can carry out the essence of various changes and modification without departing from the application to the application God and scope.So, if these modifications and variations of the application belong to the scope of the application claim and its equivalent technologies Within, then the application is also intended to comprising including these changes and modification.
Claims (10)
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

CN201710934026.7A CN107612856B (en)  20171010  20171010  Digital predistortion processing method and device 
Applications Claiming Priority (2)
Application Number  Priority Date  Filing Date  Title 

CN201710934026.7A CN107612856B (en)  20171010  20171010  Digital predistortion processing method and device 
PCT/CN2018/096530 WO2019071992A1 (en)  20171010  20180720  Digital predistortion processing method and apparatus 
Publications (2)
Publication Number  Publication Date 

CN107612856A true CN107612856A (en)  20180119 
CN107612856B CN107612856B (en)  20200114 
Family
ID=61068296
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

CN201710934026.7A CN107612856B (en)  20171010  20171010  Digital predistortion processing method and device 
Country Status (2)
Country  Link 

CN (1)  CN107612856B (en) 
WO (1)  WO2019071992A1 (en) 
Cited By (1)
Publication number  Priority date  Publication date  Assignee  Title 

WO2019071992A1 (en) *  20171010  20190418  京信通信系统（中国）有限公司  Digital predistortion processing method and apparatus 
Citations (6)
Publication number  Priority date  Publication date  Assignee  Title 

CN1547801A (en) *  20020514  20041117  松下电器产业株式会社  Method and device for compensating hybrid distortion 
CN101459636A (en) *  20071212  20090617  中兴通讯股份有限公司  Adaptive predistortion method 
CN102651634A (en) *  20110228  20120829  英特尔移动通信有限公司  Joint adaptive bias point adjustment and digital predistortion for power amplifier 
CN104168238A (en) *  20130517  20141126  上海无线通信研究中心  Method for lowering selfadaptive digital predistortion algorithm computation complexity 
US20140355714A1 (en) *  20130604  20141204  Qualcomm Incorporated  Configurable preemphasis component for transmission circuitry 
CN105763503A (en) *  20160425  20160713  京信通信技术(广州)有限公司  Digital predistortion processing method and apparatus 
Family Cites Families (7)
Publication number  Priority date  Publication date  Assignee  Title 

CN101610093B (en) *  20080617  20120905  富士通株式会社  Predistortion device, predistortion system and predistortion method 
AU2010364182B2 (en) *  20101116  20160707  Telefonaktiebolaget L M Ericsson (Publ)  Configurable basisfunction generation for nonlinear modeling 
CN102739586B (en) *  20120614  20150325  京信通信系统（中国）有限公司  Method, equipment and system for adjusting linear performance of predistortion 
CN103618684B (en) *  20131125  20170929  京信通信系统（中国）有限公司  A kind of method and device of estimation DPD coefficients 
CN103701738B (en) *  20131209  20170215  京信通信系统（中国）有限公司  Predistortion parameter determination method, equipment and digital predistortion processing system 
CN103957179B (en) *  20140404  20180420  京信通信系统（中国）有限公司  DPD method and systems 
CN107612856B (en) *  20171010  20200114  京信通信系统（中国）有限公司  Digital predistortion processing method and device 

2017
 20171010 CN CN201710934026.7A patent/CN107612856B/en active IP Right Grant

2018
 20180720 WO PCT/CN2018/096530 patent/WO2019071992A1/en unknown
Patent Citations (6)
Publication number  Priority date  Publication date  Assignee  Title 

CN1547801A (en) *  20020514  20041117  松下电器产业株式会社  Method and device for compensating hybrid distortion 
CN101459636A (en) *  20071212  20090617  中兴通讯股份有限公司  Adaptive predistortion method 
CN102651634A (en) *  20110228  20120829  英特尔移动通信有限公司  Joint adaptive bias point adjustment and digital predistortion for power amplifier 
CN104168238A (en) *  20130517  20141126  上海无线通信研究中心  Method for lowering selfadaptive digital predistortion algorithm computation complexity 
US20140355714A1 (en) *  20130604  20141204  Qualcomm Incorporated  Configurable preemphasis component for transmission circuitry 
CN105763503A (en) *  20160425  20160713  京信通信技术(广州)有限公司  Digital predistortion processing method and apparatus 
Cited By (1)
Publication number  Priority date  Publication date  Assignee  Title 

WO2019071992A1 (en) *  20171010  20190418  京信通信系统（中国）有限公司  Digital predistortion processing method and apparatus 
Also Published As
Publication number  Publication date 

WO2019071992A1 (en)  20190418 
CN107612856B (en)  20200114 
Similar Documents
Publication  Publication Date  Title 

Bianco et al.  Highorder central schemes for hyperbolic systems of conservation laws  
Ding et al.  Memory polynomial predistorter based on the indirect learning architecture  
KR100864196B1 (en)  Digital predistortion system and method for linearizing an RF power amplifier with nonlinear gain characteristics and memory effects  
CN100511974C (en)  Selfadapting digital predistortion circuti, corresponding method and wireless communication system  
JP5231218B2 (en)  Inplace transform with application to encoding and decoding of various code classes  
US20050195919A1 (en)  Digital predistortion system and method for high efficiency transmitters  
US20090088093A1 (en)  Signal predistortion in radio transmitter  
Cavers  Optimum table spacing in predistorting amplifier linearizers  
CN100544203C (en)  Nonlinear filter  
Litsyn  Peak power control in multicarrier communications  
JP2008504721A (en)  Method and apparatus for equalizing speakers  
JP2011527860A (en)  Method and system for reducing the PAPR of an OFDM signal  
JP4685860B2 (en)  Improvements in multicarrier modulation systems  
US9225501B2 (en)  Nonlinear modeling of a physical system using lookup table with polynomial interpolation  
US8831133B2 (en)  Recursive digital predistortion (DPD)  
JP2008295089A (en)  Distortion compensating amplifier apparatus  
US7091779B2 (en)  Nonlinear modeling method  
US6737918B2 (en)  Distortion compensation apparatus  
US8271859B2 (en)  Segmented CRC design in high speed networks  
JP4077449B2 (en)  Table reference predistorter  
CN1790920A (en)  A transmitter with an envelope tracking power amplifier utilizing digital predistortion of the signal envelope  
Wiechman et al.  An improved spherepacking bound for finitelength codes over symmetric memoryless channels  
US8037399B2 (en)  Techniques for segmented CRC design in high speed networks  
US7729446B2 (en)  System and method for digitally correcting a nonlinear element using a multiply partitioned architecture for predistortion  
US7822146B2 (en)  System and method for digitally correcting a nonlinear element 
Legal Events
Date  Code  Title  Description 

PB01  Publication  
PB01  Publication  
SE01  Entry into force of request for substantive examination  
SE01  Entry into force of request for substantive examination  
GR01  Patent grant 