CN106685868B - Adjacent multiband digital predistortion system and method - Google Patents

Adjacent multiband digital predistortion system and method Download PDF

Info

Publication number
CN106685868B
CN106685868B CN201710001793.2A CN201710001793A CN106685868B CN 106685868 B CN106685868 B CN 106685868B CN 201710001793 A CN201710001793 A CN 201710001793A CN 106685868 B CN106685868 B CN 106685868B
Authority
CN
China
Prior art keywords
digital
signal
nonlinear
power amplifier
predistortion
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710001793.2A
Other languages
Chinese (zh)
Other versions
CN106685868A (en
Inventor
全欣
刘颖
潘文生
唐友喜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201710001793.2A priority Critical patent/CN106685868B/en
Publication of CN106685868A publication Critical patent/CN106685868A/en
Application granted granted Critical
Publication of CN106685868B publication Critical patent/CN106685868B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/38—Synchronous or start-stop 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
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03F—AMPLIFIERS
    • H03F1/00—Details of amplifiers with only discharge tubes, only semiconductor devices or only unspecified devices as amplifying elements
    • H03F1/32—Modifications of amplifiers to reduce non-linear distortion
    • H03F1/3241—Modifications of amplifiers to reduce non-linear distortion using predistortion circuits
    • H03F1/3247—Modifications of amplifiers to reduce non-linear distortion using predistortion circuits using feedback acting on predistortion circuits

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Nonlinear Science (AREA)
  • Power Engineering (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Amplifiers (AREA)

Abstract

The invention discloses a system and a method for adjacent multiband digital predistortion, which comprises a multiband digital predistortion module, an adder, a power amplifier and a multiband predistortion parameter calculation module; the multi-band digital predistortion module comprises a plurality of digital predistorters which run in parallel; each digital predistorter processes signals from different signal sources at the same time, and the output ends of the digital predistorters are connected with the adder through respective radio frequency channels; the output end of the adder is connected with a power amplifier, and the output end of the power amplifier is connected with an antenna; the output end of the power amplifier is also connected with a multi-band predistortion parameter calculation module through a feedback channel, the output end of the multi-band predistortion parameter calculation module is respectively connected with each digital predistorter, a predistortion parameter vector is calculated by the multi-band predistortion parameter calculation module, and parameter adjustment is carried out on each digital predistorter. The invention can respectively adjust the digital predistorters corresponding to different frequency bands, thereby effectively inhibiting the nonlinearity of the close-range multiband power amplifier from being distortion.

Description

Adjacent multiband digital predistortion system and method
Technical Field
The invention relates to a system and a method for adjacent multiband digital predistortion.
Background
The multi-band power amplifier is generated, so that signals of a plurality of different frequency bands can be multiplexed by the same power amplifier, and cost is saved. However, it is a key issue how to perform predistortion processing on these multiple signals to eliminate the nonlinear distortion inherent in the power amplifier.
If the conventional single-band digital predistortion method is used to perform predistortion processing on signals of multiple frequency bands, it is necessary to integrate all the frequency band signals into a signal of one frequency channel, so that the signal bandwidth is very wide. This has the consequence that high sampling rates of DAC and ADC are required, which leads to a drastic increase in costs.
In the traditional digital predictive true method of non-adjacent multiband, the signal of each frequency band is respectively subjected to independent predistortion processing operation. Due to the large separation of the frequency bands, the cross modulation effect between the frequency bands is ignored. Thereby affecting the adjacent multiband digital predistortion effect.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a system and a method for adjacent multiband digital predistortion, which can respectively correct nonlinear distortion of each frequency band of a power amplifier.
The purpose of the invention is realized by the following technical scheme: an adjacent multiband digital predistortion system comprises a multiband digital predistortion module, an adder, a power amplifier and a multiband predistortion parameter calculation module;
the multi-band digital predistortion module comprises a plurality of digital predistorters which run in parallel and correspond to different frequency bands of the power amplifier one by one; each digital predistorter processes signals from different signal sources at the same time, and the output ends of the digital predistorters are connected with the adder through respective radio frequency channels; the output end of the adder is connected with a power amplifier, and the output end of the power amplifier is connected with an antenna; the output end of the power amplifier is also connected with a multi-band predistortion parameter calculation module through a feedback channel, the output end of the multi-band predistortion parameter calculation module is respectively connected with each digital predistorter, a predistortion parameter vector is calculated by the multi-band predistortion parameter calculation module, and parameter adjustment is carried out on each digital predistorter.
The radio frequency channel comprises a digital-to-analog conversion module and an up-conversion module, wherein the input end of the digital-to-analog conversion module is connected with the digital predistorter, the output end of the digital-to-analog conversion module is connected with the up-conversion module, and the output end of the up-conversion module is connected with the adder.
The feedback channel comprises a down-conversion module and an analog-to-digital conversion module, wherein the input end of the down-conversion module is connected with the power amplifier, and the output end of the down-conversion module is connected with the analog-to-digital conversion module.
The feedback channel further comprises a filter arranged between the down-conversion module and the analog-to-digital conversion module.
The multi-band predistortion parameter calculation module comprises:
the multi-band nonlinear model extraction unit is used for extracting a multi-band nonlinear power amplifier model according to the signals output by each digital predistorter and the feedback signals from the feedback channels;
and the predistortion parameter extraction unit is used for extracting the predistortion parameter vector of each digital predistorter according to the multi-band nonlinear power amplifier model and the signals output by each digital predistorter.
A method of adjacent multi-band digital predistortion, comprising the steps of:
s1, simultaneously processing signals u from different signal sources by using each digital predistorter in a multi-band digital predistortion module1(n),u2(n),...,uN(n) obtaining a plurality of pre-distorted signals x1(n),x2(n),...,xN(N), the number of signal sources is equal to the number of predistorters, and the number of signal sources is N, so that a predistortion signal x is obtained1(n),x2(n),...,xN(n) the inverse characteristic of the nonlinear distortion characteristic of the power amplifier is included, so that the nonlinear distortion of the power amplifier can be reduced or counteracted;
s2, pre-distortion signals x obtained by each pre-distorter1(n),x2(n),...,xN(n) respectively passing through respective radio frequency channels to synthesize the same path of signal
Figure BDA0001201715450000021
And sending the signal to a power amplifier for amplification to obtain a signal
Figure BDA0001201715450000022
And transmitting through an antenna;
s3, collecting signals output by the power amplifier
Figure BDA0001201715450000023
Obtaining a digital signal y (n) after processing through a feedback channel;
s4, utilizing digital signal y (n) and each input signal x of power amplifier1(n),x2(n),...,xNAnd (n) extracting the predistortion parameter vector and adjusting each digital predistorter.
The step S1 includes the following sub-steps:
s11, signals u of each signal source1(n),u2(n),...,uN(n), carrying out nonlinear processing to obtain nonlinear vectors corresponding to the digital predistorters;
s12, multiplying the predistortion parameter vector of each digital predistorter with the transpose of the corresponding nonlinear vector to obtain an output signal of each digital predistorter:
the output signal of the 1 st predistorter is:
Figure BDA0001201715450000024
the output signal of the 2 nd predistorter is:
Figure BDA0001201715450000025
……
the output signal of the nth predistorter is:
Figure BDA0001201715450000026
wherein, W1A predistortion parameter vector for the 1 st predistorter;
Figure BDA0001201715450000027
transpose of the nonlinear vector corresponding to the 1 st predistorter;
W2a predistortion parameter vector for the 2 nd predistorter;
Figure BDA0001201715450000028
transpose of the nonlinear vector corresponding to the 2 nd predistorter;
WNa predistortion parameter vector of an Nth predistorter;
Figure BDA0001201715450000029
is the transpose of the nonlinear vector corresponding to the nth predistorter.
The step S2 includes the following sub-steps:
s21, pre-distortion signals obtained by each pre-distorterx1(n),x2(n),...,xN(n) sending the signals into respective radio frequency channels; performing digital-to-analog conversion and up-converting to radio frequency;
s22, synthesizing all pre-distortion signals output after radio frequency channel processing into the same path of signal through an adder
Figure BDA0001201715450000031
S23, utilizing a power amplifier to transmit signals
Figure BDA0001201715450000032
Amplified and transmitted through an antenna.
The signal output to the power amplifier by the feedback channel in the step S3
Figure BDA0001201715450000033
The processing is performed to obtain the digital signal y (n) in the following ways:
first, the
Figure BDA0001201715450000034
Performing down-conversion to a low-frequency band to obtain an analog signal y (t), and directly performing analog-to-digital conversion on the analog signal y (t) to obtain a digital signal y (n);
second, to the signal
Figure BDA0001201715450000035
Each frequency band is subjected to down-conversion, filtering and analog-to-digital conversion respectively, and finally digital signals obtained by conversion of each frequency band are synthesized into one path to be output to obtain digital signals y (n);
thirdly, the signals are transmitted
Figure BDA0001201715450000036
Performing down-conversion by selecting the clock signal s via the frequency selector1(t),s2(t),…,sNAnd (t) taking one of the signals as a frequency conversion frequency to obtain a low-frequency signal, and performing filtering and analog-to-digital conversion to obtain a digital signal y (n).
The step S4 includes the following sub-steps:
s41, utilizing the signal x1(n),x2(n),...,xN(n) and y (n) estimating a nonlinear model of the original power amplifier:
(1) establishing a nonlinear power amplifier model of each frequency band:
Figure BDA0001201715450000037
Figure BDA0001201715450000038
……
Figure BDA0001201715450000039
in the formula, R1,R2,...,RNThe power amplifier nonlinear parameters of each frequency band are represented, N represents the number of y (N) intermediate frequency bands, and the number of the frequency bands is the same as that of the digital predistorters and corresponds to that of the digital predistorters one by one; y is1(n) an output signal of the nonlinear power amplifier model representing the 1 st frequency band; y is2(n) represents the output signal of the nonlinear power amplifier model of the 2 nd frequency band, yN(N) an output signal of the nonlinear power amplifier model representing the nth frequency band;
Figure BDA0001201715450000041
transpose of the nonlinear vector corresponding to the 1 st frequency band;
Figure BDA0001201715450000042
transpose of the nonlinear vector corresponding to the 2 nd frequency band;
Figure BDA0001201715450000043
transpose of the nonlinear vector corresponding to the nth frequency band;
(2) due to the signal y1(n)、y2(n) and yN(n) are all included in y (n), so that the nonlinear model parameter R of the power amplifier is obtained from y (n) in a model identification mode1,R2,...,RN：
Simplifying the nonlinear power amplifier model of each frequency band into:
Y1＝R1Ψ1；
Y2＝R2Ψ2；
......
YN＝RNΨN；
Y1representing the signal y1(n) a vector of components; y is2Representing the signal y2(n) a vector of components; y isNRepresenting the signal yN(n) are formed into vectors, and the lengths of the vectors are A and psi1Vector phi representing N sets of consecutive times1(x1,x2,...,xN) Arranged in a non-linear matrix; Ψ2Vector phi representing N sets of consecutive times2(x1,x2,...,xN) Arranged in a non-linear matrix; ΨNVector Ψ representing N sets of consecutive timesN(x1,x2,...,xN) Arranged in a non-linear matrix;
using LS algorithm to solve parameters and replace Y1And Y2Is Y; where Y ═ Y (1), Y (2), …, Y (n), represents the vector formed by signals Y (n); the estimated values of the parameters were obtained as:
Figure BDA0001201715450000044
Figure BDA0001201715450000045
……
Figure BDA0001201715450000046
Figure BDA0001201715450000047
are each R1,R2,...,RNAn estimated value of (d);
(3) from non-linear parametersEstimated value
Figure BDA0001201715450000048
Respectively obtaining an estimated value of the nonlinear model output of each frequency band:
Figure BDA0001201715450000049
Figure BDA00012017154500000410
……
Figure BDA0001201715450000051
s42, utilizing the signal x1(n),x2(n),...,xN(n) and an estimate of the nonlinear model output for each frequency band
Figure BDA0001201715450000052
Extracting predistortion parameters:
Figure BDA0001201715450000053
Figure BDA0001201715450000054
……
Figure BDA0001201715450000055
matrix E1Vector phi representing N sets of consecutive times1(y1,y2,...,yN) Arranged in a non-linear matrix of phi1(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000056
Carrying out nonlinear processing to obtain a nonlinear vector; matrix E2Vector phi representing N sets of consecutive times2(y1,y2,...,yN) Arranged in a non-linear matrix of phi2(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000057
Carrying out nonlinear processing to obtain a nonlinear vector; matrix ENVector phi representing N sets of consecutive timesN(y1,y2,...,yN) Arranged in a non-linear matrix of phiN(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000058
Carrying out nonlinear processing to obtain a nonlinear vector;
matrix X1,X2,...,XNRespectively as follows:
X1＝[x1(1),x1(2),…x1(n)]；
X2＝[x2(1),x2(2),…x2(n)]；
……
XN＝[xN(1),xN(2),…xN(n)]；
s43, pre-distortion vector parameters obtained through calculation
Figure BDA0001201715450000059
The digital predistorter transmitted to the corresponding frequency band carries out predistortion adjustment on the frequency band to replace the original predistortion parameter W1,W2，…，WN。
The invention has the beneficial effects that: the invention can extract the nonlinear predistortion parameters including the nonlinear distortion information among frequency bands from the output signal of the power amplifier and the signal output by the multi-band digital predistortion module, and respectively adjust the digital predistorters corresponding to different frequency bands, thereby effectively inhibiting the nonlinear distortion of the close-range multi-band power amplifier.
Drawings
FIG. 1 is a schematic block diagram of the system of the present invention;
FIG. 2 is a flow chart of a method of the present invention;
FIG. 3 is a diagram illustrating a first signal processing method of the feedback path;
FIG. 4 is a diagram illustrating a second signal processing method of the feedback channel;
fig. 5 is a schematic diagram of a third signal processing method of the feedback channel.
Detailed Description
The technical solutions of the present invention are further described in detail below with reference to the accompanying drawings, but the scope of the present invention is not limited to the following.
As shown in fig. 1, an adjacent multiband digital predistortion system includes a multiband digital predistortion module, an adder, a power amplifier and a multiband predistortion parameter calculation module;
the multi-band digital predistortion module comprises a plurality of digital predistorters (DPD _ 1-DPD _ N) which run in parallel and correspond to different frequency bands of the power amplifier one by one; each digital predistorter processes signals from different signal sources at the same time, and the output ends of the digital predistorters are connected with the adder through respective radio frequency channels; the output end of the adder is connected with a power amplifier, and the output end of the power amplifier is connected with an antenna; the output end of the power amplifier is also connected with a multi-band predistortion parameter calculation module through a feedback channel, the output end of the multi-band predistortion parameter calculation module is respectively connected with each digital predistorter, a predistortion parameter vector is calculated by the multi-band predistortion parameter calculation module, and parameter adjustment is carried out on each digital predistorter.
The radio frequency channel comprises a digital-to-analog conversion module and an up-conversion module, wherein the input end of the digital-to-analog conversion module is connected with the digital predistorter, the output end of the digital-to-analog conversion module is connected with the up-conversion module, and the output end of the up-conversion module is connected with the adder.
The feedback channel comprises a down-conversion module and an analog-to-digital conversion module, wherein the input end of the down-conversion module is connected with the power amplifier, and the output end of the down-conversion module is connected with the analog-to-digital conversion module.
The feedback channel further comprises a filter arranged between the down-conversion module and the analog-to-digital conversion module.
The multi-band predistortion parameter calculation module comprises:
the multi-band nonlinear model extraction unit is used for extracting a multi-band nonlinear power amplifier model according to the signals output by each digital predistorter and the feedback signals from the feedback channels;
and the predistortion parameter extraction unit is used for extracting the predistortion parameter vector of each digital predistorter according to the multi-band nonlinear power amplifier model and the signals output by each digital predistorter.
As shown in fig. 2, a method of adjacent multiband digital predistortion, comprising the steps of:
s1, simultaneously processing signals u from different signal sources by using each digital predistorter in a multi-band digital predistortion module1(n),u2(n),...,uN(n) obtaining a plurality of pre-distorted signals x1(n),x2(n),...,xN(N), the number of signal sources is equal to the number of predistorters, and the number of signal sources is N, so that a predistortion signal x is obtained1(n),x2(n),...,xN(n) the inverse characteristic of the nonlinear distortion characteristic of the power amplifier is included, so that the nonlinear distortion of the power amplifier can be reduced or counteracted;
s2, pre-distortion signals x obtained by each pre-distorter1(n),x2(n),...,xN(n) respectively passing through respective radio frequency channels to synthesize the same path of signal
Figure BDA0001201715450000071
And sending the signal to a power amplifier for amplification to obtain a signal
Figure BDA0001201715450000072
And transmitting through an antenna;
s3, collecting signals output by the power amplifier
Figure BDA0001201715450000073
Obtaining a digital signal y (n) after processing through a feedback channel;
s4. benefitUsing digital signals y (n) and input signals x of power amplifiers1(n),x2(n),...,xNAnd (n) extracting the predistortion parameter vector and adjusting each digital predistorter.
The step S1 includes the following sub-steps:
s11, signals u of each signal source1(n),u2(n),...,uN(n), carrying out nonlinear processing to obtain nonlinear vectors corresponding to the digital predistorters;
s12, multiplying the predistortion parameter vector of each digital predistorter with the transpose of the corresponding nonlinear vector to obtain an output signal of each digital predistorter:
the output signal of the 1 st predistorter is:
Figure BDA0001201715450000074
the output signal of the 2 nd predistorter is:
Figure BDA0001201715450000075
……
the output signal of the nth predistorter is:
Figure BDA0001201715450000076
wherein, W1A predistortion parameter vector for the 1 st predistorter;
Figure BDA0001201715450000077
transpose of the nonlinear vector corresponding to the 1 st predistorter;
W2a predistortion parameter vector for the 2 nd predistorter;
Figure BDA0001201715450000078
transpose of the nonlinear vector corresponding to the 2 nd predistorter;
WNa predistortion parameter vector of an Nth predistorter;
Figure BDA0001201715450000079
is the transpose of the nonlinear vector corresponding to the nth predistorter.
The step S2 includes the following sub-steps:
s21, pre-distortion signals x obtained by each pre-distorter1(n),x2(n),...,xN(n) sending the signals into respective radio frequency channels; performing digital-to-analog conversion and up-converting to radio frequency;
s22, synthesizing all pre-distortion signals output after radio frequency channel processing into the same path of signal through an adder
Figure BDA00012017154500000710
S23, utilizing a power amplifier to transmit signals
Figure BDA00012017154500000711
Amplified and transmitted through an antenna.
In the embodiment of the present application, in step S3, the feedback channel is used to output the signal to the power amplifier
Figure BDA0001201715450000081
The processing is performed to obtain the digital signal y (n) in the following ways:
first, as shown in FIG. 3, will
Figure BDA0001201715450000082
Performing down-conversion to a low-frequency band to obtain an analog signal y (t), and directly performing analog-to-digital conversion (ADC) on the analog signal y (t) to obtain a digital signal y (n); this approach requires a high enough analog-to-digital conversion rate to ensure that the digital signal y (n) contains all the frequency information of the analog signal y (t);
second, as shown in FIG. 4, for signals
Figure BDA0001201715450000083
Each frequency band is respectively subjected to down conversion, filtering and analog-to-digital conversion (ADC), and finally, digital signals obtained by conversion of each frequency band are synthesized into one path to be output to obtain digital signals y (n);
third, as shown in FIG. 5, the signals are combined
Figure BDA0001201715450000084
Performing down-conversion by selecting the clock signal s via the frequency selector1(t),s2(t),…,sNAnd (t) taking one of the signals as a frequency conversion frequency to obtain a low-frequency signal, and performing filtering and analog-to-digital conversion (ADC) to obtain a digital signal y (n).
The step S4 includes the following sub-steps:
s41, utilizing the signal x1(n),x2(n),...,xN(n) and y (n) estimating a nonlinear model of the original power amplifier:
(1) establishing a nonlinear power amplifier model of each frequency band:
Figure BDA0001201715450000085
Figure BDA0001201715450000086
……
Figure BDA0001201715450000087
in the formula, R1,R2,...,RNThe power amplifier nonlinear parameters of each frequency band are represented, N represents the number of y (N) intermediate frequency bands, and the number of the frequency bands is the same as that of the digital predistorters and corresponds to that of the digital predistorters one by one; y is1(n) an output signal of the nonlinear power amplifier model representing the 1 st frequency band; y is2(n) represents the output signal of the nonlinear power amplifier model of the 2 nd frequency band, yN(N) an output signal of the nonlinear power amplifier model representing the nth frequency band;
Figure BDA0001201715450000088
transpose of the nonlinear vector corresponding to the 1 st frequency band;
Figure BDA0001201715450000089
transpose of the nonlinear vector corresponding to the 2 nd frequency band;
Figure BDA00012017154500000810
transpose of the nonlinear vector corresponding to the nth frequency band;
(2) due to the signal y1(n)、y2(n) and yN(n) are all included in y (n), so that the nonlinear model parameter R of the power amplifier is obtained from y (n) in a model identification mode1,R2,...,RN：
Simplifying the nonlinear power amplifier model of each frequency band into:
Y1＝R1Ψ1；
Y2＝R2Ψ2；
......
YN＝RNΨN；
Y1representing the signal y1(n) a vector of components; y is2Representing the signal y2(n) a vector of components; y isNRepresenting the signal yN(n) are formed into vectors, and the lengths of the vectors are A and psi1Vector phi representing N sets of consecutive times1(x1,x2,...,xN) Arranged in a non-linear matrix; Ψ2Vector phi representing N sets of consecutive times2(x1,x2,...,xN) Arranged in a non-linear matrix; ΨNVector phi representing N sets of consecutive timesN(x1,x2,...,xN) Arranged in a non-linear matrix;
using LS algorithm to solve parameters and replace Y1And Y2Is Y; where Y ═ Y (1), Y (2), …, Y (n), represents the vector formed by signals Y (n); the estimated values of the parameters were obtained as:
Figure BDA0001201715450000091
Figure BDA0001201715450000092
……
Figure BDA0001201715450000093
Figure BDA0001201715450000094
are each R1,R2,...,RNAn estimated value of (d);
(3) from non-linear parameter estimates
Figure BDA0001201715450000095
Respectively obtaining an estimated value of the nonlinear model output of each frequency band:
Figure BDA0001201715450000096
Figure BDA0001201715450000097
……
Figure BDA0001201715450000098
s42, utilizing the signal x1(n),x2(n),...,xN(n) and an estimate of the nonlinear model output for each frequency band
Figure BDA0001201715450000099
Extracting predistortion parameters:
Figure BDA0001201715450000101
Figure BDA0001201715450000102
……
Figure BDA0001201715450000103
matrix E1Vector phi representing N sets of consecutive times1(y1,y2,...,yN) The non-linear matrix is arranged in a non-linear matrix,1(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000104
Carrying out nonlinear processing to obtain a nonlinear vector; matrix E2Vector phi representing N sets of consecutive times2(y1,y2,...,yN) Arranged in a non-linear matrix of phi2(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000105
Carrying out nonlinear processing to obtain a nonlinear vector; matrix ENVector phi representing N sets of consecutive timesN(y1,y2,...,yN) Arranged in a non-linear matrix of phiN(y1,y2,...,yN) Representing the N signals
Figure BDA0001201715450000106
Carrying out nonlinear processing to obtain a nonlinear vector;
matrix X1,X2,...,XNRespectively as follows:
X1＝[x1(1),x1(2),…x1(n)]；
X2＝[x2(1),x2(2),…x2(n)]；
……
XN＝[[xN(1),xN(2),…xN(n)]；
s43, pre-distortion vector parameters obtained through calculation
Figure BDA0001201715450000107
Transmitted to a corresponding frequency band digital predistorter to be processedAdjusting line predistortion, replacing original predistortion parameter W1,W2，…，WN。
In one embodiment of the present application, N ═ 2, step S1 is as follows:
Figure BDA0001201715450000108
Figure BDA0001201715450000109
representing 2 signal sources u1(n),u2(n) a nonlinear vector obtained by performing predistortion nonlinear processing, the center frequencies of all elements in the vector and the signal source u1(n) have the same center frequency, wherein the vectors
Figure BDA00012017154500001010
The k × j × L + L elements of (1) are formed in a manner that:
Figure BDA0001201715450000111
Figure BDA0001201715450000112
is a pre-distortion parameter vector, vector length and
Figure BDA0001201715450000113
the same;
Figure BDA0001201715450000114
(i>1, i is odd) represents the number of signal sources u for 2 signal sources1(n),u2(n) a nonlinear vector obtained by performing predistortion nonlinear processing, wherein the signal center frequencies of all elements in the vector are equal to the signal source u1(n),u2I order intermodulation component of (n)
Figure BDA0001201715450000115
Has the same center frequency, wherein the vector
Figure BDA0001201715450000116
The k × j × L + L elements of (1) are formed in a manner that:
Figure BDA0001201715450000117
Figure BDA0001201715450000118
is a pre-distortion parameter vector, vector length and
Figure BDA0001201715450000119
the same;
when all the vectors are processed
Figure BDA00012017154500001110
Arranged in sequence as a vector
Figure BDA00012017154500001111
Arranging all non-linear vectors into one vector at a time
Figure BDA00012017154500001112
The output of the first digital predistorter can be written as:
Figure BDA00012017154500001113
where T denotes a matrix transpose operation.
Similarly, the output of the second digital predistorter can be written as:
Figure BDA00012017154500001114
wherein phi2(u1,u2) Is formed of1(u1,u2) Similarly; only in the construction of specific vectors
Figure BDA00012017154500001115
Exchange u1(n) and u2(n) position.
In the embodiment where N is 2, step S4 is specifically as follows:
firstly, establishing a nonlinear power amplifier model of 2 frequency bands as follows:
Figure BDA00012017154500001116
wherein, y1(n) and y2(n) are the first band and second band power amplifier model outputs, respectively; phi1(x1,x2) Representing 2 signals x1(n),x2(n) performing nonlinear processing to obtain a nonlinear vector; internal constitution and phi1(u1,u2) Similarly, except that the signal u is replaced separately1(n),u2(n) is x1(n),x2(n)；Φ2(x1,x2) Representing 2 signals x1(n),x2(n) performing nonlinear processing to obtain a nonlinear vector; internal constitution and phi2(u1,u2) Similarly, except that the signal u is replaced separately1(n),u2(n) is x1(n),x2(n)；
Due to the signal y1(n) and y2And (n) comprises y (n), and the nonlinear model parameters R1 and R2 of the power amplifier can be respectively obtained from y (n) in a model identification mode. The specific identification method can be Least Square (LS) algorithm or Recursive Least Square (RLS) algorithm. Taking the LS algorithm as an example:
the above equation is written in matrix form:
Figure BDA0001201715450000121
wherein Y is1And Y2Respectively represent signals y1(n) and y2(N) a vector of length N; r1 and R2 represent nonlinear parameter vectors corresponding to band 1 and band 2, respectively, and are constructed in a manner similar to W1 and W2; Ψ1To representCombining N sets of continuous-time vectors phi1(x1,x2) Arranged in a non-linear matrix; Ψ2Vector phi representing N sets of consecutive times2(x1,x2) Arranged in a non-linear matrix;
solving parameters with LS algorithm and replacing Y1And Y2For Y, the estimated value of the parameter is obtained as:
Figure BDA0001201715450000122
Figure BDA0001201715450000123
and
Figure BDA0001201715450000124
estimated values for R1 and R2, respectively.
From non-linear parameter estimates
Figure BDA0001201715450000125
And
Figure BDA0001201715450000126
respectively obtaining an estimated value of the nonlinear model output of each frequency band:
Figure BDA0001201715450000127
and further:
Figure BDA0001201715450000128
Figure BDA0001201715450000129
extracting predistortion parameters: using x1(n),x2(n) and output of power amplifier model estimation module
Figure BDA00012017154500001210
And
Figure BDA00012017154500001211
to extract predistortion parameters, one of the extraction methods is as follows:
one of the calculation methods may use a conventional indirect learning structure to obtain W:
Figure BDA00012017154500001212
Figure BDA00012017154500001213
wherein the matrix E1Vector phi representing N sets of consecutive times1(y1,y2) Arranged in a non-linear matrix; e2Vector phi representing N sets of consecutive times2(y1,y2) Arranged in a non-linear matrix; phi1(y1,y2) Represents 2 signals
Figure BDA00012017154500001214
And
Figure BDA00012017154500001215
carrying out nonlinear processing to obtain a nonlinear vector; internal constitution and phi1(u1,u2) Similarly, except that the signal u is replaced separately1(n) and u2(n) is
Figure BDA0001201715450000136
And
Figure BDA0001201715450000131
2(y1,y2) Represents 2 signals
Figure BDA0001201715450000132
And
Figure BDA0001201715450000133
to perform non-linearCarrying out linear processing on the obtained nonlinear vector; internal constitution and phi2(u1,u2) Similarly, except that the signal u is replaced separately1(n) and u2(n) is
Figure BDA0001201715450000137
And
Figure BDA0001201715450000138
and:
Figure BDA0001201715450000134
derived from the estimation
Figure BDA0001201715450000135
The digital predistorter transmitted to the corresponding frequency band carries out predistortion adjustment on the frequency band to replace the original predistortion parameter W1,W2。

Claims (6)

1. A contiguous multi-band digital predistortion system, characterized by: the system comprises a multi-band digital predistortion module, an adder, a power amplifier and a multi-band predistortion parameter calculation module;
the multi-band digital predistortion module comprises a plurality of digital predistorters which run in parallel and correspond to different frequency bands of the power amplifier one by one; each digital predistorter processes signals from different signal sources at the same time, and the output ends of the digital predistorters are connected with the adder through respective radio frequency channels; the output end of the adder is connected with a power amplifier, and the output end of the power amplifier is connected with an antenna; the output end of the power amplifier is also connected with a multi-band predistortion parameter calculation module through a feedback channel, the output end of the multi-band predistortion parameter calculation module is respectively connected with each digital predistorter, a predistortion parameter vector is calculated by the multi-band predistortion parameter calculation module, and parameter adjustment is carried out on each digital predistorter;
a method of adjacent multi-band digital predistortion for an adjacent multi-band digital predistortion system, comprising the steps of:
s1, simultaneously processing signals u from different signal sources by using each digital predistorter in a multi-band digital predistortion module1(n)，u2(n)，...，uN(n) obtaining a plurality of pre-distorted signals x1(n)，x2(n)，...，xN(N), the number of signal sources is equal to the number of predistorters, and the number of signal sources is N, so that a predistortion signal x is obtained1(n)，x2(n)，....xN(n) the inverse characteristic of the nonlinear distortion characteristic of the power amplifier is included, so that the nonlinear distortion of the power amplifier can be reduced or counteracted;
s2, pre-distortion signals x obtained by each pre-distorter1(n)，x2(n)，...，xN(n) respectively passing through respective radio frequency channels to synthesize the same path of signal
Figure FDA0002575890670000011
And sending the signal to a power amplifier for amplification to obtain a signal
Figure FDA0002575890670000012
And transmitting through an antenna;
s3, collecting signals output by the power amplifier
Figure FDA0002575890670000013
Obtaining a digital signal y (n) after being processed by a feedback channel
S4, utilizing digital signal y (n) and each input signal x of power amplifier1(n)，x2(n)，...，xN(n), extracting a predistortion parameter vector, and adjusting each digital predistorter;
the step S1 includes the following sub-steps:
s11, signals u of each signal source1(n)，u2(n)，...，uN(n), carrying out nonlinear processing to obtain nonlinear vectors corresponding to the digital predistorters;
s12, multiplying the predistortion parameter vector of each digital predistorter with the transpose of the corresponding nonlinear vector to obtain an output signal of each digital predistorter:
the output signal of the 1 st predistorter is
Figure FDA0002575890670000014
The output signal of the 2 nd predistorter is
Figure FDA0002575890670000015
……
The output signal of the Nth predistorter is
Figure FDA0002575890670000016
Wherein, W1 is a predistortion parameter vector of the 1 st predistorter;
Figure FDA0002575890670000017
transpose of the nonlinear vector corresponding to the 1 st predistorter;
w2 is the predistortion parameter vector of the 2 nd predistorter;
Figure FDA0002575890670000021
transpose of the nonlinear vector corresponding to the 2 nd predistorter;
WN is a predistortion parameter vector of the Nth predistorter;
Figure FDA0002575890670000022
transpose the nonlinear vector corresponding to the nth predistorter;
the signal output to the power amplifier by the feedback channel in the step S3
Figure FDA0002575890670000023
The processing is performed to obtain the digital signal y (n) in the following ways:
first, the
Figure FDA0002575890670000024
Performing down-conversion to a low-frequency band to obtain an analog signal y (t), and directly performing analog-to-digital conversion on the analog signal y (t) to obtain a digital signal y (n);
second, to the signal
Figure FDA0002575890670000025
Each frequency band is subjected to down-conversion, filtering and analog-to-digital conversion respectively, and finally digital signals obtained by conversion of each frequency band are synthesized into one path to be output to obtain digital signals y (n);
thirdly, the signals are transmitted
Figure FDA0002575890670000026
Performing down-conversion by selecting the clock signal s via the frequency selector1(t)，s2(t)，…，sNOne of (t) is used as a frequency conversion frequency to obtain a low-frequency signal, and a digital signal y (n) is obtained after filtering and analog-to-digital conversion; the step S4 includes the following sub-steps:
s41, utilizing the signal x1(n)，x2(n)，...，xN(n) and y (n) estimating a nonlinear model of the original power amplifier:
(1) establishing a nonlinear power amplifier model of each frequency band:
Figure FDA0002575890670000027
Figure FDA0002575890670000028
……
Figure FDA0002575890670000029
in the formula, R1，R2，...，RNThe power amplifier nonlinear parameters of each frequency band are represented, N represents the number of y (N) intermediate frequency bands, and the number of the frequency bands is the same as that of the digital predistorters and corresponds to that of the digital predistorters one by one; y is1(n) output of the 1 st band nonlinear power amplifier modelA signal; y is2(n) represents the output signal of the nonlinear power amplifier model of the 2 nd frequency band, yN(N) an output signal of the nonlinear power amplifier model representing the nth frequency band;
Figure FDA00025758906700000210
transpose of the nonlinear vector corresponding to the 1 st frequency band;
Figure FDA00025758906700000211
transpose of the nonlinear vector corresponding to the 2 nd frequency band;
Figure FDA00025758906700000212
transpose of the nonlinear vector corresponding to the nth frequency band;
(2) due to the signal y1(n)、y2(n) and yN(n) are all included in y (n), so that the nonlinear model parameter R of the power amplifier is obtained from y (n) in a model identification mode1，R2，...，RN：
Simplifying the nonlinear power amplifier model of each frequency band into:
Y1＝R1Ψ1；
Y2＝R2Ψ2；
……
YN＝RNΨN；
y1 denotes the signal Y1(n) a vector of components; y2 denotes the signal Y2(n) a vector of components; YN represents signal yN(N) are all A, psi 1 represents N groups of continuous time vectors phi 1 (x)1，x2，...，xN) Arranged in a non-linear matrix; Ψ 2 represents a vector Φ 2 (x) that divides N sets of consecutive times1，x2，...，xN) Arranged in a non-linear matrix; Ψ N represents a vector Φ N (x) that represents N groups of consecutive times1，x2，...，xN) Arranged in a non-linear matrix;
carrying out parameter solution by using an LS algorithm, and replacing Y1 and Y2 with Y; where Y ═ Y (1), Y (2), …, Y (n) ], represents a vector of signals Y (n); the estimated values of the parameters were obtained as:
Figure FDA0002575890670000031
Figure FDA0002575890670000032
Figure FDA0002575890670000033
Figure FDA0002575890670000034
are each R1，R2，...，RNAn estimated value of (d);
(3) from non-linear parameter estimates
Figure FDA0002575890670000035
Respectively obtaining an estimated value of the nonlinear model output of each frequency band:
Figure FDA0002575890670000036
Figure FDA0002575890670000037
Figure FDA0002575890670000038
s42, utilizing the signal x1(n)，x2(n)，...，xN(n) and an estimate of the nonlinear model output for each frequency band
Figure FDA0002575890670000039
Extracting predistortion parameters:
Figure FDA00025758906700000310
Figure FDA00025758906700000311
……
Figure FDA00025758906700000312
the matrix E1 represents a vector Φ 1 (y) of N sets of consecutive times1，y2，...，yN) Arranged in a non-linear matrix, Φ 1 (y)1y2，...，yN) Representing the N signals
Figure FDA00025758906700000313
Carrying out nonlinear processing to obtain a nonlinear vector; the matrix E2 represents a vector Φ 2 (y) of N sets of consecutive times1，y2，...，yN) Arranged in a non-linear matrix, Φ 2 (y)1，y2，...，yN) Representing the N signals
Figure FDA0002575890670000041
Carrying out nonlinear processing to obtain a nonlinear vector; the matrix EN represents a vector of N successive time groups φ N (y)1，y2，...，yN) Arranged in a non-linear matrix, Φ N (y)1，y2，...，yN) Representing the N signals
Figure FDA0002575890670000042
Carrying out nonlinear processing to obtain a nonlinear vector;
matrix X1，X2，...，XNRespectively as follows:
X1＝[x1(1)，x1(2)，…x1(n)]；
X2＝[x2(1)，x2(2)，…x2(n)]；
……
XN＝[xN(1)，xN(2)，…xN(n)]；
s43, pre-distortion vector parameters obtained through calculation
Figure FDA0002575890670000043
The digital predistorter transmitted to the corresponding frequency band carries out predistortion adjustment on the frequency band to replace the original predistortion parameter W1，W2，…，WN。
2. The adjacent multiband digital predistortion system of claim 1, wherein the radio frequency channel comprises a digital-to-analog conversion module and an up-conversion module, an input end of the digital-to-analog conversion module is connected with the digital predistorter, an output end of the digital-to-analog conversion module is connected with the up-conversion module, and an output end of the up-conversion module is connected with the adder.
3. The adjacent multiband digital predistortion system of claim 1, wherein the feedback channel comprises a down-conversion module and an analog-to-digital conversion module, an input end of the down-conversion module is connected with the power amplifier, and an output end of the down-conversion module is connected with the analog-to-digital conversion module.
4. A contiguous multiband digital predistortion system according to claim 1 or 3, wherein the feedback path further comprises a filter disposed between the down-conversion module and the analog-to-digital conversion module.
5. The system of claim 1, wherein the multi-band predistortion parameter calculation module comprises:
the multi-band nonlinear model extraction unit is used for extracting a multi-band nonlinear power amplifier model according to the signals output by each digital predistorter and the feedback signals from the feedback channels;
and the predistortion parameter extraction unit is used for extracting the predistortion parameter vector of each digital predistorter according to the multi-band nonlinear power amplifier model and the signals output by each digital predistorter.
6. The adjacent multiband digital predistortion system of claim 1, wherein the step S2 includes the sub-steps of:
s21, pre-distortion signals x obtained by each pre-distorter1(n)，x2(n)，...，xN(n) sending the signals into respective radio frequency channels; performing digital-to-analog conversion and up-converting to radio frequency;
s22, synthesizing all pre-distortion signals output after radio frequency channel processing into the same path of signal through an adder
Figure FDA0002575890670000044
And S23, amplifying the signals by using a power amplifier and then transmitting the signals through an antenna.
CN201710001793.2A 2017-01-03 2017-01-03 Adjacent multiband digital predistortion system and method Active CN106685868B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710001793.2A CN106685868B (en) 2017-01-03 2017-01-03 Adjacent multiband digital predistortion system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710001793.2A CN106685868B (en) 2017-01-03 2017-01-03 Adjacent multiband digital predistortion system and method

Publications (2)

Publication Number Publication Date
CN106685868A CN106685868A (en) 2017-05-17
CN106685868B true CN106685868B (en) 2020-09-08

Family

ID=58848767

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710001793.2A Active CN106685868B (en) 2017-01-03 2017-01-03 Adjacent multiband digital predistortion system and method

Country Status (1)

Country Link
CN (1) CN106685868B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11476809B2 (en) * 2017-08-11 2022-10-18 Nokia Solutions And Networks Oy Polyphase digital signal predistortion in radio transmitter
CN107994923B (en) * 2017-11-03 2019-11-12 京信通信系统（中国）有限公司 Ultra-broadband digital pre-distortion method, device and system
CN112543156B (en) * 2019-09-20 2024-04-12 中兴通讯股份有限公司 Digital predistortion method for multiband signal, electronic device and readable storage medium
KR20210108196A (en) * 2020-02-25 2021-09-02 주식회사 케이엠더블유 Multi band transmitter
FI20205889A1 (en) 2020-09-15 2022-03-16 Nokia Technologies Oy Interference mitigation with Multi-Band Digital Pre-Distortion
CN113612452B (en) * 2021-08-11 2023-09-26 电子科技大学 Digital predistortion correction method and device with frequency selection characteristic

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105765861A (en) * 2013-11-22 2016-07-13 赛灵思公司 Multi-path digital pre-distortion
CN106034096A (en) * 2015-03-20 2016-10-19 瑞昱半导体股份有限公司 Transmitter and method for reducing distortion of input signals

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8817859B2 (en) * 2011-10-14 2014-08-26 Fadhel Ghannouchi Digital multi-band predistortion linearizer with nonlinear subsampling algorithm in the feedback loop

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105765861A (en) * 2013-11-22 2016-07-13 赛灵思公司 Multi-path digital pre-distortion
CN106034096A (en) * 2015-03-20 2016-10-19 瑞昱半导体股份有限公司 Transmitter and method for reducing distortion of input signals

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Novel Multiband Linearization Technique for Closely-Spaced Dual-Band Signals of Wide Bandwidth;Ying Liu, et al.;《2015 IEEE MTT-S International Microwave Symposium》;20150630;第1-4页 *

Also Published As

Publication number Publication date
CN106685868A (en) 2017-05-17

Similar Documents

Publication Publication Date Title
CN106685868B (en) Adjacent multiband digital predistortion system and method
CN106506417B (en) Narrow-band feedback digital predistortion system and method
US10103755B2 (en) Multi-band radio-frequency digital predistortion
US11476809B2 (en) Polyphase digital signal predistortion in radio transmitter
CN103314523B (en) Systems and methods for digital predistortion in a dual band transmitter
CN104604126B (en) Low sampling rate adaptation scheme for double frequency-band linearisation
CN102017553B (en) Method and system for baseband predistortion linearization in multi-channel wideband communication systems
CN102014090B (en) Digital predistortion method and device
US9166840B2 (en) Method of transmitting data samples with reduced bandwidth
CN101645862B (en) Method and device for reducing signal peak-to-average ratio
CN103493367A (en) A method for pre-distorting and a pre-distorter
CN105763495B (en) digital predistortion method and device
CN106685368A (en) Digital predistortion system and method achieving undersampling feedback
EP1833214B1 (en) A method and system for out of band predistortion linearization
CN105471784A (en) Digital predistortion method of jointly compensating for IQ imbalance and PA non-linearity
CN103888396A (en) Predistortion method based on self-adaptive variable step size iteration
CN113659937B (en) Pre-distortion processing method and device, communication equipment and storage medium
WO2015188578A1 (en) Nonlinear system distortion correction device and method
CN201409180Y (en) Self-adapting baseband linearization device of digital television transmitter
CN106921600A (en) The method of adjustment and device of a kind of predistortion model
CN110808746B (en) DPD model parameter extraction method for MIMO transmitter
CN113612453A (en) Low-sampling-rate feedback digital predistortion correction method and device
CN108134584A (en) Combine digital pre-distortion system and method with outer in the band of wide-band radio frequency power amplifier
CN113630134B (en) Digital feedforward-assisted broadband digital predistortion method and device
CN118137987B (en) Simplified undersampled digital predistortion method and system

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
GR01 Patent grant