CN109683115A - A kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system - Google Patents
A kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system Download PDFInfo
- Publication number
- CN109683115A CN109683115A CN201910111648.9A CN201910111648A CN109683115A CN 109683115 A CN109683115 A CN 109683115A CN 201910111648 A CN201910111648 A CN 201910111648A CN 109683115 A CN109683115 A CN 109683115A
- Authority
- CN
- China
- Prior art keywords
- power amplifier
- frequency power
- signal
- radio
- magnetic resonance
- 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.)
- Granted
Links
- 238000013528 artificial neural network Methods 0.000 claims abstract description 59
- 230000006870 function Effects 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 24
- 230000008569 process Effects 0.000 claims abstract description 15
- 238000006243 chemical reaction Methods 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 13
- 238000004088 simulation Methods 0.000 claims abstract description 9
- 238000001228 spectrum Methods 0.000 claims description 15
- 230000011664 signaling Effects 0.000 claims description 7
- 230000003321 amplification Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 5
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 5
- 230000008878 coupling Effects 0.000 claims description 4
- 238000010168 coupling process Methods 0.000 claims description 4
- 238000005859 coupling reaction Methods 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 3
- 210000001367 artery Anatomy 0.000 claims 1
- 238000001914 filtration Methods 0.000 claims 1
- 210000003462 vein Anatomy 0.000 claims 1
- 230000003446 memory effect Effects 0.000 description 11
- 210000002569 neuron Anatomy 0.000 description 11
- 230000008447 perception Effects 0.000 description 9
- 230000008901 benefit Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000012804 iterative process Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000009022 nonlinear effect Effects 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 238000012897 Levenberg–Marquardt algorithm Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000002939 conjugate gradient method Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 210000004218 nerve net Anatomy 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000011017 operating method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/28—Details of apparatus provided for in groups G01R33/44 - G01R33/64
- G01R33/32—Excitation or detection systems, e.g. using radio frequency signals
- G01R33/36—Electrical details, e.g. matching or coupling of the coil to the receiver
- G01R33/3614—RF power amplifiers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/28—Details of apparatus provided for in groups G01R33/44 - G01R33/64
- G01R33/32—Excitation or detection systems, e.g. using radio frequency signals
- G01R33/36—Electrical details, e.g. matching or coupling of the coil to the receiver
- G01R33/3621—NMR receivers or demodulators, e.g. preamplifiers, means for frequency modulation of the MR signal using a digital down converter, means for analog to digital conversion [ADC] or for filtering or processing of the MR signal such as bandpass filtering, resampling, decimation or interpolation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/565—Correction of image distortions, e.g. due to magnetic field inhomogeneities
- G01R33/5659—Correction of image distortions, e.g. due to magnetic field inhomogeneities caused by a distortion of the RF magnetic field, e.g. spatial inhomogeneities of the RF magnetic field
Abstract
Present disclose provides magnetic resonance radio frequency power amplifier apparatus and magnetic resonance systems.Wherein, magnetic resonance radio frequency power amplifier apparatus, including the first balun are configured as receiving feed-forward signal and are sent to processor through the first analog-digital converter;Second balun is configured as receiving feedback signal and is sent to processor through the second analog-digital converter;Processor, it is configured as received feed-forward signal and feedback signal carrying out digital frequency conversion and signal synchronization process, by neural network indirect learning structured training radio-frequency power amplifier inverse function estimator, and the weight coefficient of radio-frequency power amplifier inverse function estimator is copied to predistorter;The output signal of predistorter successively enters radio-frequency power amplifier module through third balun, digital analog converter and simulation up-converter module;Radio-frequency power amplifier module is divided into two-way after amplifying received signal.
Description
Technical field
The disclosure belongs to magnetic resonance arts more particularly to a kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system
System.
Background technique
Only there is provided background technical informations relevant to the disclosure for the statement of this part, it is not necessary to so constitute first skill
Art.
In magnetic resonance system, radio-frequency power amplifier is needed to be amplified to the radio-frequency pulse that resonance spectrometer emits several
Ten arrive several hectowatts, transmitting coil are output to, with excitation experiment sample.Therefore radio-frequency power amplifier be can not in magnetic resonance system
The a part lacked.But such as general power amplifier, radio-frequency power amplifier has nonlinear characteristic, works as radio-frequency power
Amplifier operation can cause pulse signal non-linear distortion at saturation region or cut-off region.The power of input signal is bigger, radio frequency
The distortion of power amplifier is more serious.This distortion includes amplitude and phase distortion and the extension of frequency band external spectrum in frequency band, is done
The signal transmission for disturbing adjacent channel, finally makes magnetic resonance system imaging artifact, distortion occur, and image quality reduces, and interferes disease
Disease diagnosis and scientific research.The line of power amplifier is realized while therefore, it is necessary to guarantee radio-frequency power amplifier working efficiency
Property.
Solving radio-frequency power amplifier non-linear method includes following several method:
(1) most straightforward approach is manufacture high-precision High Linear radio-frequency power amplifier.But inventor has found its manufacture
It is complex process, expensive, it generally is not suitable for use in high power amplifier.
(2) back-off is to improve a kind of simplest method of power amplifier linearity, by reducing power amplifier
Operating point so that power amplifier work from saturation point retract 10 arrive 15dB place.Power back realization is simple, but
Inventor's discovery is the disadvantage is that will be greatly reduced the working efficiency of power amplifier, and then increase the maintenance cost of system.
(3) direct feedback is directly to inhibit input signal using output signal, and inventor has found the disadvantage is that being difficult to estimate
Delay of the output signal for input signal is counted, so that system stability is poor.Therefore in engineering more through frequently with indirectly negative
Feedback transmitter, indirect negative-feedback method refer to that output and input signal are compared by a kind of indirect connection type.Therefore, indirectly
The advantages of negative-feedback method, is precision height, technology maturation, cheap etc..Inventor's discovery is its shortcoming is that feedback loop is delayed
It is difficult to control, system is not sufficiently stable, is not suitable for frequency band compared with wide field conjunction etc..
(4) feed forward approach basic principle is as follows: isolating interference signal to the power amplifier through being delayed using offset loop
Output signal is overlapped counteracting, and then realizes the linearisation of power amplifier.The method fast speed, linearisation is preferable, makes
With broader bandwidth, but its structure is complicated, at high cost, efficiency is lower and adaptivity is poor for inventor's discovery.
(5) basic thought of pre-distortion technology is that a curve characteristic and power amplifier song are inserted into before power amplifier
The reciprocal predistorter of line characteristic uses realization power amplifier linearity by the connection grade of predistorter and power amplifier
Change.The mode that the predistorter of pre-distortion technology is mostly used analog circuit realizes, its advantage is that it is at low cost, circuit structure is simple, suitable
Should be with wide, however inventor's discovery is especially distorted high order component and carries out predistortion the disadvantage is that predistortion effect is very limited
It is more difficult.
In conclusion inventors have found that the working efficiency of radio-frequency power amplifier is lower and non-linear processed at present
Predistortion effect in journey is poor.
Summary of the invention
To solve the above-mentioned problems, the first aspect of the disclosure provides a kind of magnetic resonance radio frequency power amplifier apparatus,
It has the advantages that work efficiency is high and predistortion effect in Nonlinear Processing process is good.
A kind of technical solution of magnetic resonance radio frequency power amplifier apparatus of the first aspect of the disclosure are as follows:
A kind of magnetic resonance radio frequency power amplifier apparatus of the disclosure, comprising:
First balun is configured as receiving feed-forward signal and is sent to processor through the first analog-digital converter;
Second balun is configured as receiving feedback signal and is sent to processor through the second analog-digital converter;
Processor is configured as carrying out received feed-forward signal into digital frequency conversion place synchronous with signal with feedback signal
Reason then by neural network indirect learning structured training radio-frequency power amplifier inverse function estimator, and radio-frequency power is put
The weight coefficient of big device inverse function estimator is copied to predistorter;The radio-frequency power amplifier inverse function estimator and pre- mistake
True device structure is identical;
The output signal of the predistorter, which successively enters through third balun, digital analog converter and simulation up-converter module, penetrates
Frequency power amplifier module;
The radio-frequency power amplifier module is configured as being divided into two-way after amplifying received signal, wherein emitting all the way
It goes out, successively coupled module and analog down resume module obtain feedback signal to another way.
Further, received feed-forward signal and feedback signal are carried out the process of digital frequency conversion by the processor are as follows:
By feed-forward signal Digital Down Convert at baseband signal, then by feed-forward signal Digital Up Convert be center frequency meet place
Manage the signal of device centre frequency;
By feedback signal Digital Down Convert at baseband signal, then by feedback signal Digital Up Convert be center frequency meet place
Manage the signal of device centre frequency.
The advantages of above scheme, is, by the way that received feed-forward signal and feedback signal are carried out digital frequency conversion, so that
The centre frequency of received feed-forward signal and feedback signal meets the center frequency signal of processor, can be improved processing in this way
The working efficiency of device, and then improve the stability of entire magnetic resonance radio frequency power amplifier apparatus work.
Further, the predistorter is made of linear portion and non-linear.
Further, the linear portion is FIR filter structure;The non-linear is neural network structure.
Further, neural network structure is shallow-layer learning neural network or the neural network based on deep learning.
Wherein, predistorter also needs to be fitted radio-frequency power in addition to needing to be fitted the non-linear distortion of radio-frequency power amplifier
Amplifier memory effect, function neural network predistortion device also must band memory effect at this time.
It is worth noting that, neural network here, can be traditional shallow-layer learning neural network, such as backpropagation
(BP) neural network, multi-layered perception neural networks etc. are also possible to the neural network based on deep learning, such as depth nerve net
Network, recurrent neural network, convolutional neural networks.
The training algorithm that neural network uses can be gradient descent method, additional guide vanes, conjugate gradient method, newton and calculate
Method, Levenberg-Marquardt algorithm etc..
Further, the processor is estimated by neural network indirect learning structured training radio-frequency power amplifier inverse function
The process of gauge are as follows:
Radio frequency function is updated according to predistorter and radio-frequency power amplifier inverse function estimator output difference adaptive polo placement
The weight coefficient of rate amplifier inverse function estimator;
When predistorter and radio-frequency power amplifier inverse function estimator output difference are less than certain value, by radio-frequency power
Amplifier inverse function estimator weight coefficient is copied to predistorter.
The advantages of above scheme, is that such predistorter weighs radio-frequency power amplifier inverse function estimator in convergence
Weight coefficient is copied to predistorter, can be improved the stability of entire magnetic resonance radio frequency power amplifier apparatus work.
Further, the processor, is also configured to
According to the variation of radio-frequency power amplifier module, periodically update the weight coefficient of predistorter, real-time monitoring with
Track radio-frequency power amplifier module performance change, realizes the self-adapted pre-distortion of radio-frequency power amplifier module.
Further, the simulation up-converter module includes: that the first numerically-controlled oscillator, bandpass filter and first are mixed
Frequency device;First numerically-controlled oscillator completes frequency for generating sine and cosine two-way orthogonal signalling, and by the first frequency mixer jointly
The resettlement of spectrum;Bandpass filter is used to filter out the redundant frequency spectrum generated when up-conversion mixing.
Further, the analog down module includes: the second numerically-controlled oscillator, digital Finite filter and
Two frequency mixers;Second numerically-controlled oscillator is for generating sine and cosine two-way orthogonal signalling, and it is jointly complete to pass through the second frequency mixer
At the resettlement of frequency spectrum;Digital Finite filter is used to filter out the redundant frequency spectrum generated when down coversion mixing.
Further, the coupling module includes: power divider and signal attenuator, the power divider be used for from
A part of radiofrequency signal is extracted in radio-frequency power amplifier, this portion radio frequency signal successively passes through signal attenuator and simulation again
Down conversion module processing, obtains feedback signal.
To solve the above-mentioned problems, the second aspect of the disclosure provides a kind of magnetic resonance system, with working efficiency
The good advantage of predistortion effect in high and Nonlinear Processing process.
A kind of technical solution of magnetic resonance system of the second aspect of the disclosure are as follows:
A kind of magnetic resonance system of the disclosure, comprising:
Magnetic resonance signal generating device is configured as generating magnetic resonance pulse signal;
Magnetic resonance radio frequency power amplifier apparatus described above is configured as in advance being lost magnetic resonance pulse signal
Really amplifies and export to magnetic resonance signal emitter;
The magnetic resonance signal emitter is configured as the magnetic resonance signal of transmitting predistortion amplification.
The beneficial effect of the disclosure is:
(1) disclosure is using the non-linear of neural network fitting magnetic resonance system radio-frequency power amplifier, to radio-frequency power
Amplifier carries out predistortion, improves the working efficiency of magnetic resonance system radio-frequency power amplifier device and non-linear processed
Predistortion effect in journey.
(2) disclosure is fitted radio-frequency power amplifier memory effect using filter segment, quasi- using part of neural network
The non-linear distortion of radio-frequency power amplifier is closed, and then realizes the non-linear and memory effect of separate type fitting radio-frequency power amplifier
It answers.It is compared with traditional band Memory Neural Networks predistorter model, not only greatly reduces legacy network parameter and network rule
Mould, and greatly reduce the calculation amount that weight coefficient updates iterative process.
Detailed description of the invention
The Figure of description for constituting a part of this disclosure is used to provide further understanding of the disclosure, and the disclosure is shown
Meaning property embodiment and its explanation do not constitute the improper restriction to the disclosure for explaining the disclosure.
Fig. 1 is a kind of magnetic resonance radio frequency power amplifier apparatus structural schematic diagram that the embodiment of the present disclosure provides.
Fig. 2 is a kind of processor structure signal for magnetic resonance radio frequency power amplifier apparatus that the embodiment of the present disclosure provides
Figure.
Fig. 3 is a kind of magnetic resonance radio frequency power amplifier apparatus schematic diagram that the embodiment of the present disclosure provides.
Fig. 4 is the indirect learning structure principle chart based on multi-layered perception neural networks that the embodiment of the present disclosure provides.
Fig. 5 is the multi-layered perception neural networks schematic diagram that the embodiment of the present disclosure provides.
Fig. 6 is the AM/AM characteristic curve before the predistortion that the embodiment of the present disclosure provides.
Fig. 7 is the AM/AM characteristic curve after the predistortion that the embodiment of the present disclosure provides.
Fig. 8 is the AM/PM characteristic curve before the predistortion that the embodiment of the present disclosure provides.
Fig. 9 is the AM/PM characteristic curve after the predistortion that the embodiment of the present disclosure provides.
Figure 10 is the power spectrum of signal before and after the original signal and predistortion that the embodiment of the present disclosure provides.
Figure 11 is a kind of magnetic resonance system structural schematic diagram that the embodiment of the present disclosure provides.
Specific embodiment
It is noted that following detailed description is all illustrative, it is intended to be further explained to disclosure offer.Unless another
It indicates, all technical and scientific terms used herein has usual with disclosure person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the disclosure.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
As shown in Figure 1, a kind of magnetic resonance radio frequency power amplifier apparatus of the present embodiment, comprising:
First balun is configured as receiving feed-forward signal and is sent to processor through the first analog-digital converter;
Second balun is configured as receiving feedback signal and is sent to processor through the second analog-digital converter;
Processor is configured as carrying out received feed-forward signal into digital frequency conversion place synchronous with signal with feedback signal
Reason then by neural network indirect learning structured training radio-frequency power amplifier inverse function estimator, and radio-frequency power is put
The weight coefficient of big device inverse function estimator is copied to predistorter;The radio-frequency power amplifier inverse function estimator and pre- mistake
True device structure is identical;
The output signal of the predistorter, which successively enters through third balun, digital analog converter and simulation up-converter module, penetrates
Frequency power amplifier module;
The radio-frequency power amplifier module is configured as being divided into two-way after amplifying received signal, wherein emitting all the way
It goes out, successively coupled module and analog down resume module obtain feedback signal to another way.
Wherein, balun: balun is responsible for mutually converting between high frequency single-ended signal and differential signal.
Analog-digital converter: having Differential Input mouth, is digital signal by analog-signal transitions.
Processor: DSP or fpga chip can be used to realize.
Processor mainly realizes that Digital up and down convert pretreatment, signal are synchronous and magnetic resonance radio frequency power amplifier number is pre-
Three functions are distorted, as shown in Figure 2.
Digital analog converter: converting digital signals into analog signal, exports analog signal.
Simulation up-converter module: radio frequency band will be adjusted in the intermediate-freuqncy signal of digital analog converter output.
Radio-frequency power amplifier module: the signal of up-conversion is after radio-frequency power amplifier amplifies, using radio frequency line
Circle is launched.
Coupling module: it is equivalent to signal attenuator.
Analog down module: by radiofrequency signal frequency reducing.
Specifically, the simulation up-converter module includes: the first numerically-controlled oscillator, bandpass filter and the first mixing
Device;First numerically-controlled oscillator completes frequency spectrum for generating sine and cosine two-way orthogonal signalling, and by the first frequency mixer jointly
Resettlement;Bandpass filter is used to filter out the redundant frequency spectrum generated when up-conversion mixing.
The analog down module includes: the second numerically-controlled oscillator, digital Finite filter and the second frequency mixer;
Second numerically-controlled oscillator completes removing for frequency spectrum for generating sine and cosine two-way orthogonal signalling, and by the second frequency mixer jointly
It moves;Digital Finite filter is used to filter out the redundant frequency spectrum generated when down coversion mixing.
The coupling module includes: power divider and signal attenuator, and the power divider is used for from radio-frequency power
A part of radiofrequency signal is extracted in amplifier, this portion radio frequency signal successively passes through signal attenuator and analog down mould again
Block processing, obtains feedback signal.
In specific implementation, received feed-forward signal and feedback signal are carried out the process of digital frequency conversion by the processor
Are as follows:
By feed-forward signal Digital Down Convert at baseband signal, then by feed-forward signal Digital Up Convert be center frequency meet place
Manage the signal of device centre frequency;
By feedback signal Digital Down Convert at baseband signal, then by feedback signal Digital Up Convert be center frequency meet place
Manage the signal of device centre frequency.
The advantages of above scheme, is, by the way that received feed-forward signal and feedback signal are carried out digital frequency conversion, so that
The centre frequency of received feed-forward signal and feedback signal meets the center frequency signal of processor, can be improved processing in this way
The working efficiency of device, and then improve the stability of entire magnetic resonance radio frequency power amplifier apparatus work.
Wherein, neural network indirect learning structure is also to realize on DSP or fpga chip, as shown in figure 3, in the knot
In structure, predistorter is identical with the structure of radio-frequency power amplifier inverse function estimator.The predistorter is by linear portion
It is constituted with non-linear.
The linear portion is FIR filter structure;The non-linear is neural network structure.
Neural network structure is shallow-layer learning neural network or the neural network based on deep learning.
Wherein, predistorter also needs to be fitted radio-frequency power in addition to needing to be fitted the non-linear distortion of radio-frequency power amplifier
Amplifier memory effect, function neural network predistortion device also must band memory effect at this time.
It is worth noting that, neural network here, can be traditional shallow-layer learning neural network, such as backpropagation
(BP) neural network, multi-layered perception neural networks etc. are also possible to the neural network based on deep learning, such as depth nerve net
Network, recurrent neural network, convolutional neural networks.
The training algorithm that neural network uses can be gradient descent method, additional guide vanes, conjugate gradient method, newton and calculate
Method, Levenberg-Marquardt algorithm etc..
Further, the processor is estimated by neural network indirect learning structured training radio-frequency power amplifier inverse function
The process of gauge are as follows:
Radio frequency function is updated according to predistorter and radio-frequency power amplifier inverse function estimator output difference adaptive polo placement
The weight coefficient of rate amplifier inverse function estimator;
When predistorter and radio-frequency power amplifier inverse function estimator output difference are less than certain value, by radio-frequency power
Amplifier inverse function estimator weight coefficient is copied to predistorter.
The advantages of above scheme, is that such predistorter weighs radio-frequency power amplifier inverse function estimator in convergence
Weight coefficient is copied to predistorter, can be improved the stability of entire magnetic resonance radio frequency power amplifier apparatus work.
In specific implementation, the processor, is also configured to
According to the variation of radio-frequency power amplifier module, periodically update the weight coefficient of predistorter, real-time monitoring with
Track radio-frequency power amplifier module performance change, realizes the self-adapted pre-distortion of radio-frequency power amplifier module.
Below with the indirect learning structure based on multi-layered perception neural networks, as being described in detail for Fig. 4:
Predistorter is exactly the pre-distortion system being made of two parts of linear and nonlinear.Wherein linear portion is
FIR filter, non-linear partial are 3 layers of common multi-layered perception neural networks of a double-input double-output.It is quasi- using filter segment
Radio-frequency power amplifier memory effect is closed, using the non-linear distortion of part of neural network fitting radio-frequency power amplifier, in turn
Realize the non-linear and memory effect of separate type fitting radio-frequency power amplifier.
It is compared with traditional band Memory Neural Networks predistorter model, it should indirect based on multi-layered perception neural networks
It practises structure and not only greatly reduces legacy network parameter and network size, but also greatly reduce weight coefficient and update iterative process
Calculation amount.
Multi-layered perception neural networks described in Fig. 5 embodiment, feedforward calculate as follows:
In formula (1),For the output of the 1st layer of j-th of neuron of multi-layered perception neural networks, M is memory depth,For
The connection weight coefficient of the 1st layer of j-th of neuron and i-th of input layer input, x-iIt is inputted for i-th of neural network;
In formula (2),For the output of the 2nd layer of j-th of neuron, hidden layer activation primitiveFor tanh letter
Number,For the connection weight coefficient of the 2nd layer of j-th of neuron and the 1st layer of i-th of neuron,For the 1st layer of j-th of neuron
Output,For the biasing coefficient of the 2nd layer of j-th of neuron;
In formula (3),For the output of the 3rd layer of k-th of neuron, l1For hidden layer number of nodes,For the 3rd layer of k-th of mind
Connection weight coefficient through member with the 2nd layer of j-th of neuron,For the biasing coefficient of the 3rd layer of k-th of neuron.It selects f (x)
=x is as output layer neuron activation primitive.
Training algorithm using Bayesian-Levenberg-Marquardt optimization algorithm as neural network, target letter
Number expression formula are as follows:
F (X)=α EW+βED(4)
In formulaWherein N is the update of neural network weight
The rounds of iteration;S2 is neural network output layer number of nodes;Q is the training sample number of every bout,Training sample number
Neural network indicates the error amount square of true device Yu inverse function estimator when for q;vj(x) true to indicate during iteration j
The error amount square of device and inverse function estimator;wjFor network weight coefficient;α, β are coefficient;M is the total of weight coefficient in network
Number, for the positive integer more than or equal to 1.
As it can be seen that by new objective function, while network can guarantee that network output error is small as far as possible in the training process
Guarantee that network has lesser network weight coefficient.Bayesian-Levenberg-Marquardt updates neural network coefficient
Iterative formula are as follows:
Xk+1=Xk-[αJTJ-(μ+β)I]-1JTe(5)
Wherein, μ and e is constant coefficient.
In above formula, J is Jacobi matrix, and expression formula is
Wherein, X is the weight coefficient vector of network.The optimum value of factor alpha and β value is respectively αMPAnd βMP, and given by following formula
Out:
Wherein γ=m-2 αMP·tr(HMP)-1Indicate effective weight coefficient number, range is 0 between m.M is to weigh in network
The sum of weight coefficient, HMPIt is objective function F (X) in its smallest point XMPThe Hessian matrix at place.It needs to count in calculating process
Calculate the Hessian matrix.Simplify Hessian matrix using Gauss-Newton approximatioss, then: ▽2F(XMP)≈βJTJ+αIm, wherein J
It is EDIn point XMPJacobi matrix.
The operating procedure of Bayesian-LM algorithm is as follows:
(1) initialization network parameter, and initialize factor alpha=0, β=1.
(2) Levenberg-Marquardt algorithmic minimizing network performance objective function F (X)=α E is utilizedW+βED。
(3) H ≈ β J is solved using Gauss-Newton approximatiossTJ+αIm, and solve actual parameter number γ.
(4) the new estimated value of design factor
(5) step is repeated (2) to arrive (4), until algorithmic statement.
The AM/AM performance plot of radio-frequency power amplifier before predistortion, as shown in Figure 6.AM/AM performance plot after prerun very,
As shown in Figure 7.From Fig. 6 and Fig. 7: the AM/AM characteristic of radio-frequency power amplifier is the non-thread of band sluggishness before predistortion
Linearity curve.The AM/AM characteristic of radio-frequency power amplifier with predistorter be almost a slope be 1 straight line and without slow
It is stagnant, illustrate that amplitude amplification substantially achieves linearisation.
The AM/PM performance plot of radio-frequency power amplifier is as shown in Figure 8 before predistortion.AM/PM performance plot after prerun very is such as
Shown in Fig. 9.From Fig. 8 and Fig. 9: before predistortion, when input changes between 0 to 1, the AM/PM of radio-frequency power amplifier
There are different degrees of offsets for phase on performance plot, and the offset of phase is bigger when inputting smaller.AM/PM characteristic exists after predistortion
When input range changes, the offset of output phase is essentially 0, and the purpose for illustrating phase predistortion substantially achieves.
Figure 10 is the power spectrum of radio frequency amplifier before and after predistortion, smooth, the predistorter that inband signaling becomes after predistortion
Structure can effectively reduce Adjacent Channel Power Ratio 30dB or so.
The magnetic resonance radio frequency power amplifier apparatus of the present embodiment is fitted magnetic resonance system radio-frequency power using neural network
Amplifier nonlinearity carries out predistortion to radio-frequency power amplifier, improves magnetic resonance system radio-frequency power amplifier device
Working efficiency and Nonlinear Processing process in predistortion effect.
The magnetic resonance radio frequency power amplifier apparatus of the present embodiment is remembered using filter segment fitting radio-frequency power amplifier
Recall effect, using the non-linear distortion of part of neural network fitting radio-frequency power amplifier, and then realizes that separate type is fitted radio frequency
The non-linear and memory effect of power amplifier.It is compared with traditional band Memory Neural Networks predistorter model, not only significantly
Reduce legacy network parameter and network size, and greatly reduces the calculation amount that weight coefficient updates iterative process.
Figure 11 is a kind of magnetic resonance system structural schematic diagram that the embodiment of the present disclosure provides.
As shown in figure 11, a kind of magnetic resonance system of the present embodiment, comprising:
Magnetic resonance signal generating device is configured as generating magnetic resonance pulse signal;
Magnetic resonance radio frequency power amplifier apparatus as shown in Figure 1 is configured as carrying out magnetic resonance pulse signal pre-
Distortion is amplified and is exported to magnetic resonance signal emitter;
The magnetic resonance signal emitter is configured as the magnetic resonance signal of transmitting predistortion amplification.
Specifically, resonance spectrometer can be used to realize in magnetic resonance signal generating device.
The radio-frequency coil of magnetic resonance system can be used to realize in magnetic resonance signal emitter.
The magnetic resonance system of the present embodiment is non-linear using neural network fitting magnetic resonance system radio-frequency power amplifier,
Predistortion is carried out to radio-frequency power amplifier, improves the working efficiency of magnetic resonance system radio-frequency power amplifier device and non-
Predistortion effect during linear process.
The magnetic resonance system of the present embodiment is fitted radio-frequency power amplifier memory effect using filter segment, utilizes nerve
Network portion is fitted the non-linear distortion of radio-frequency power amplifier, and then realizes the non-thread of separate type fitting radio-frequency power amplifier
Property and memory effect.It is compared with traditional band Memory Neural Networks predistorter model, not only greatly reduces legacy network ginseng
Several and network size, and greatly reduce the calculation amount that weight coefficient updates iterative process.
Although above-mentioned be described in conjunction with specific embodiment of the attached drawing to the disclosure, model not is protected to the disclosure
The limitation enclosed, those skilled in the art should understand that, on the basis of the technical solution of the disclosure, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within the protection scope of the disclosure.
Claims (10)
1. a kind of magnetic resonance radio frequency power amplifier apparatus characterized by comprising
First balun is configured as receiving feed-forward signal and is sent to processor through the first analog-digital converter;
Second balun is configured as receiving feedback signal and is sent to processor through the second analog-digital converter;
Processor is configured as received feed-forward signal and feedback signal carrying out digital frequency conversion and signal synchronization process,
Then by neural network indirect learning structured training radio-frequency power amplifier inverse function estimator, and by radio-frequency power amplifier
The weight coefficient of inverse function estimator is copied to predistorter;The radio-frequency power amplifier inverse function estimator and predistorter
Structure is identical;
The output signal of the predistorter successively enters radio frequency function through third balun, digital analog converter and simulation up-converter module
Rate amplifier module;
The radio-frequency power amplifier module is configured as being divided into two-way after amplifying received signal, wherein launching all the way
It goes, successively coupled module and analog down resume module obtain feedback signal to another way.
2. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the processor will connect
The feed-forward signal and feedback signal of receipts carry out the process of digital frequency conversion are as follows:
By feed-forward signal Digital Down Convert at baseband signal, then by feed-forward signal Digital Up Convert be center frequency meet processor
The signal of centre frequency;
By feedback signal Digital Down Convert at baseband signal, then by feedback signal Digital Up Convert be center frequency meet processor
The signal of centre frequency.
3. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the predistorter by
Linear portion and non-linear are constituted.
4. a kind of magnetic resonance radio frequency power amplifier apparatus as claimed in claim 3, which is characterized in that the linear portion is
FIR filter structure;The non-linear is neural network structure.
5. a kind of magnetic resonance radio frequency power amplifier apparatus as claimed in claim 4, which is characterized in that neural network structure is
Shallow-layer learning neural network or neural network based on deep learning.
6. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the processor passes through
The process of neural network indirect learning structured training radio-frequency power amplifier inverse function estimator are as follows:
It updates radio-frequency power according to predistorter and radio-frequency power amplifier inverse function estimator output difference adaptive polo placement and puts
The weight coefficient of big device inverse function estimator;
When predistorter and radio-frequency power amplifier inverse function estimator output difference are less than certain value, radio-frequency power is amplified
Device inverse function estimator weight coefficient is copied to predistorter.
7. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the processor, also
It is configured as:
According to the variation of radio-frequency power amplifier module, the weight coefficient of predistorter is periodically updated, real-time monitoring tracking is penetrated
Frequency power amplifier module performance change, realizes the self-adapted pre-distortion of radio-frequency power amplifier module.
8. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the simulation up-conversion
Module includes: the first numerically-controlled oscillator, bandpass filter and the first frequency mixer;First numerically-controlled oscillator is for generating
Sine and cosine two-way orthogonal signalling, and pass through the resettlement that the first frequency mixer completes frequency spectrum jointly;Bandpass filter is for filtering out change
The redundant frequency spectrum that frequency generates when being mixed;
Or the analog down module includes: the second numerically-controlled oscillator, digital Finite filter and the second frequency mixer;The
Two numerically-controlled oscillators complete removing for frequency spectrum for generating sine and cosine two-way orthogonal signalling, and by the second frequency mixer jointly
It moves;Digital Finite filter is used to filter out the redundant frequency spectrum generated when down coversion mixing.
9. a kind of magnetic resonance radio frequency power amplifier apparatus as described in claim 1, which is characterized in that the coupling module packet
Include: power divider and signal attenuator, the power divider are penetrated for extracting a part from radio-frequency power amplifier
Frequency signal, this portion radio frequency signal successively pass through signal attenuator and analog down resume module again, obtain feedback signal.
10. a kind of magnetic resonance system characterized by comprising
Magnetic resonance signal generating device is configured as generating magnetic resonance pulse signal;
Magnetic resonance radio frequency power amplifier apparatus as claimed in any one of claims 1-9 wherein, is configured as magnetic resonance arteries and veins
Signal is rushed to carry out predistortion amplification and export to magnetic resonance signal emitter;
The magnetic resonance signal emitter is configured as the magnetic resonance signal of transmitting predistortion amplification.
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910111648.9A CN109683115B (en) | 2019-02-12 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system | |
PCT/CN2020/074422 WO2020164435A1 (en) | 2019-02-12 | 2020-02-06 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
GB2112953.1A GB2595828B (en) | 2019-02-12 | 2020-02-06 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
ZA2021/06725A ZA202106725B (en) | 2019-02-12 | 2021-09-10 | Magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910111648.9A CN109683115B (en) | 2019-02-12 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109683115A true CN109683115A (en) | 2019-04-26 |
CN109683115B CN109683115B (en) | 2024-05-03 |
Family
ID=
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111324861A (en) * | 2020-02-28 | 2020-06-23 | 厦门大学 | Deep learning magnetic resonance spectrum reconstruction method based on matrix decomposition |
WO2020164435A1 (en) * | 2019-02-12 | 2020-08-20 | 山东第一医科大学(山东省医学科学院) | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
CN112763952A (en) * | 2019-11-06 | 2021-05-07 | 西门子(深圳)磁共振有限公司 | Method and apparatus for monitoring radio frequency transmission circuit of magnetic resonance imaging equipment |
Citations (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5140268A (en) * | 1990-06-15 | 1992-08-18 | The Board Of Trustees Of The Leland Stanford Junior University | Method and means for correcting RF amplifier distortion in magnetic resonance imaging |
US6239593B1 (en) * | 1998-09-21 | 2001-05-29 | Southwest Research Institute | Method and system for detecting and characterizing mechanical damage in pipelines using nonlinear harmonics techniques |
CN1299532A (en) * | 1998-04-30 | 2001-06-13 | 诺基亚网络有限公司 | Linearization method for amplifier, and amplifier arrangement |
CN1516493A (en) * | 2003-01-10 | 2004-07-28 | 深圳市中兴通讯股份有限公司上海第二 | Power amplifier linearization method and linear power amplifier device |
CN1870614A (en) * | 2005-05-27 | 2006-11-29 | 华为技术有限公司 | Universal device for predistortion processing base band digital signal |
CN101072220A (en) * | 2006-05-08 | 2007-11-14 | 中兴通讯股份有限公司 | Radical basis function neural network predistortion method for adaptive power amplifier |
CN101110571A (en) * | 2007-08-30 | 2008-01-23 | 京信通信***(中国)有限公司 | Digital pre-distorting power amplifier and implementing method thereof |
CN101320960A (en) * | 2008-07-18 | 2008-12-10 | 东南大学 | Power amplifier predistortion method of Hammerstein model based on fuzzy neural network |
US20100141257A1 (en) * | 2007-05-04 | 2010-06-10 | Koninklijke Philips Electronics N.V. | Rf transmitter with digital feedback for mri |
CN102082751A (en) * | 2009-11-27 | 2011-06-01 | 电子科技大学 | Neural network pre-distortion method based on improved MLBP (Levenberg-Marquardt back propagation) algorithm |
US8179133B1 (en) * | 2008-08-18 | 2012-05-15 | Hypres, Inc. | High linearity superconducting radio frequency magnetic field detector |
CN102804596A (en) * | 2009-06-22 | 2012-11-28 | 高通股份有限公司 | Method and apparatus for using pre-distortion and feedback to mitigate nonlinearity of circuits |
CN103326973A (en) * | 2013-07-03 | 2013-09-25 | 苏州工业园区昶辰通信科技有限公司 | Distortion characteristic estimation method predistortion device and method thereof |
US20140118066A1 (en) * | 2012-10-31 | 2014-05-01 | Korea Advanced Institute Of Science And Technology | Low-cost digital predistortion apparatus and method using envelope detection feedback |
US20140254716A1 (en) * | 2013-03-05 | 2014-09-11 | Qualcomm Incorporated | Power Amplifier System Including A Composite Digital Predistorter |
CN104811149A (en) * | 2014-01-28 | 2015-07-29 | 马克西姆综合产品公司 | Adaptively controlled digital pre-distortion in an rf power amplifier using an integrated signal analyzer with enhanced analog-to-digital conversion |
CN105262447A (en) * | 2015-11-26 | 2016-01-20 | 中国电子科技集团公司第三十研究所 | Pre-distortion method and device for power amplifier and radio frequency system |
CN105450184A (en) * | 2014-08-08 | 2016-03-30 | 瑞昱半导体股份有限公司 | Digit pre-distortion circuit and method and digit pre-distortion training circuit |
CN105897188A (en) * | 2016-03-29 | 2016-08-24 | 上海联影医疗科技有限公司 | Method and system for nonlinear correction of power amplifier |
CN205958733U (en) * | 2016-08-26 | 2017-02-15 | 西门子(深圳)磁共振有限公司 | Radio frequency emission device and magnetic resonance imaging system |
CN106603022A (en) * | 2016-11-10 | 2017-04-26 | 电子科技大学 | Short-wave high-linearity balance structure power amplifier |
CN209570689U (en) * | 2019-02-12 | 2019-11-01 | 泰山医学院 | A kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system |
Patent Citations (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5140268A (en) * | 1990-06-15 | 1992-08-18 | The Board Of Trustees Of The Leland Stanford Junior University | Method and means for correcting RF amplifier distortion in magnetic resonance imaging |
CN1299532A (en) * | 1998-04-30 | 2001-06-13 | 诺基亚网络有限公司 | Linearization method for amplifier, and amplifier arrangement |
US6239593B1 (en) * | 1998-09-21 | 2001-05-29 | Southwest Research Institute | Method and system for detecting and characterizing mechanical damage in pipelines using nonlinear harmonics techniques |
CN1516493A (en) * | 2003-01-10 | 2004-07-28 | 深圳市中兴通讯股份有限公司上海第二 | Power amplifier linearization method and linear power amplifier device |
CN1870614A (en) * | 2005-05-27 | 2006-11-29 | 华为技术有限公司 | Universal device for predistortion processing base band digital signal |
CN101072220A (en) * | 2006-05-08 | 2007-11-14 | 中兴通讯股份有限公司 | Radical basis function neural network predistortion method for adaptive power amplifier |
US20100141257A1 (en) * | 2007-05-04 | 2010-06-10 | Koninklijke Philips Electronics N.V. | Rf transmitter with digital feedback for mri |
CN101110571A (en) * | 2007-08-30 | 2008-01-23 | 京信通信***(中国)有限公司 | Digital pre-distorting power amplifier and implementing method thereof |
CN101320960A (en) * | 2008-07-18 | 2008-12-10 | 东南大学 | Power amplifier predistortion method of Hammerstein model based on fuzzy neural network |
US8179133B1 (en) * | 2008-08-18 | 2012-05-15 | Hypres, Inc. | High linearity superconducting radio frequency magnetic field detector |
CN102804596A (en) * | 2009-06-22 | 2012-11-28 | 高通股份有限公司 | Method and apparatus for using pre-distortion and feedback to mitigate nonlinearity of circuits |
CN102082751A (en) * | 2009-11-27 | 2011-06-01 | 电子科技大学 | Neural network pre-distortion method based on improved MLBP (Levenberg-Marquardt back propagation) algorithm |
US20140118066A1 (en) * | 2012-10-31 | 2014-05-01 | Korea Advanced Institute Of Science And Technology | Low-cost digital predistortion apparatus and method using envelope detection feedback |
US20140254716A1 (en) * | 2013-03-05 | 2014-09-11 | Qualcomm Incorporated | Power Amplifier System Including A Composite Digital Predistorter |
CN103326973A (en) * | 2013-07-03 | 2013-09-25 | 苏州工业园区昶辰通信科技有限公司 | Distortion characteristic estimation method predistortion device and method thereof |
CN104811149A (en) * | 2014-01-28 | 2015-07-29 | 马克西姆综合产品公司 | Adaptively controlled digital pre-distortion in an rf power amplifier using an integrated signal analyzer with enhanced analog-to-digital conversion |
CN105450184A (en) * | 2014-08-08 | 2016-03-30 | 瑞昱半导体股份有限公司 | Digit pre-distortion circuit and method and digit pre-distortion training circuit |
CN105262447A (en) * | 2015-11-26 | 2016-01-20 | 中国电子科技集团公司第三十研究所 | Pre-distortion method and device for power amplifier and radio frequency system |
CN105897188A (en) * | 2016-03-29 | 2016-08-24 | 上海联影医疗科技有限公司 | Method and system for nonlinear correction of power amplifier |
CN205958733U (en) * | 2016-08-26 | 2017-02-15 | 西门子(深圳)磁共振有限公司 | Radio frequency emission device and magnetic resonance imaging system |
CN106603022A (en) * | 2016-11-10 | 2017-04-26 | 电子科技大学 | Short-wave high-linearity balance structure power amplifier |
CN209570689U (en) * | 2019-02-12 | 2019-11-01 | 泰山医学院 | A kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system |
Non-Patent Citations (4)
Title |
---|
李环;张羽;祝乐芳;: "一种改进的数字预失真迭代算法", 沈阳理工大学学报, no. 03 * |
杨文考, 周尚波, 朱维乐: "OFDM***中的一种自适应预失真器", 电波科学学报, no. 02 * |
袁子龙;张照喜;邱建峰;陈迢;刘涛;刘玉林;: "基于ACR体模的磁共振射频线圈性能的对比", 中国医疗设备, no. 01 * |
邓洪敏, 何松柏, 虞厥邦: "基于BP神经网络的功放自适应预失真", 通信学报, no. 11 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020164435A1 (en) * | 2019-02-12 | 2020-08-20 | 山东第一医科大学(山东省医学科学院) | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
GB2595828A (en) * | 2019-02-12 | 2021-12-08 | Shandong First Medical Univ & Shandong Academy Of Medical Sciences | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
GB2595828B (en) * | 2019-02-12 | 2023-02-15 | Shandong First Medical Univ & Shandong Academy Of Medical Sciences | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
CN112763952A (en) * | 2019-11-06 | 2021-05-07 | 西门子(深圳)磁共振有限公司 | Method and apparatus for monitoring radio frequency transmission circuit of magnetic resonance imaging equipment |
CN112763952B (en) * | 2019-11-06 | 2023-08-25 | 西门子(深圳)磁共振有限公司 | Method and apparatus for monitoring radio frequency transmit circuit of magnetic resonance imaging device |
CN111324861A (en) * | 2020-02-28 | 2020-06-23 | 厦门大学 | Deep learning magnetic resonance spectrum reconstruction method based on matrix decomposition |
Also Published As
Publication number | Publication date |
---|---|
WO2020164435A1 (en) | 2020-08-20 |
GB202112953D0 (en) | 2021-10-27 |
GB2595828B (en) | 2023-02-15 |
ZA202106725B (en) | 2021-10-27 |
GB2595828A (en) | 2021-12-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
FI107212B (en) | I / Q modulator DC offset correction | |
CN109061581B (en) | Radar target distance accurate simulation device and method for linear frequency modulation signals | |
CN1076903C (en) | Efficient linear power amplification | |
CN106877825A (en) | The digital predistortion apparatus and method of the simplified nonlinear filter based on band limit | |
CN206389365U (en) | A kind of multi-tap is non-to wait work(point Full-Duplex Analog self-interference cancellation element | |
CN102143108A (en) | Improved self-adaption predistortion technology | |
CN107395538B (en) | A kind of digital pre-distortion system and method for frequency selectivity harmonics restraint | |
GB2587066A (en) | Method for compensating gain flatness of transceiver | |
CN207939511U (en) | A kind of RF transceiver chip | |
CN105656434A (en) | Power amplifier digital pre-distortion device and method based on modified piecewise linear function | |
CN103685109A (en) | Broadband digital pre-distortion multicarrier power amplification system, and power-amplification gain equalization device and method thereof | |
CN201947295U (en) | Base band linearization device for digital television transmitter based on multi-channel feedback | |
Tripathi et al. | Swish activation based deep neural network predistorter for RF-PA | |
CN111064439B (en) | System and method for improving short wave digital predistortion performance | |
CN101217522A (en) | A method and device to determine the predistortion parameter of open loop, commutator and commutating method | |
CN114400979A (en) | System and method for improving digital predistortion of short wave interaction and harmonic distortion | |
CN114448454A (en) | Short-wave transmitter capable of suppressing harmonic wave and intermodulation distortion | |
CN109683115A (en) | A kind of magnetic resonance radio frequency power amplifier apparatus and magnetic resonance system | |
CN111010095B (en) | Broadband power amplifier linearization processing method and system of wireless communication system | |
CN109831231B (en) | A kind of bluetooth baseband receives system and its implementation | |
CN104301269B (en) | Equivalent sampling pre-distortion system and its method | |
CN109120285A (en) | A kind of solution of Tetra terminal power amplifier linearization | |
CN102647378B (en) | Pre-distortion device, pre-distortion method and transmitter/ receiver system | |
CN107294546B (en) | Radio frequency link and its working method in a kind of digital pre-distortion system | |
CN109683115B (en) | Magnetic resonance radio frequency power amplifier device and magnetic resonance 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 |