CN109683115B - Magnetic resonance radio frequency power amplifier device and magnetic resonance system - Google Patents
Magnetic resonance radio frequency power amplifier device and magnetic resonance system Download PDFInfo
- Publication number
- CN109683115B CN109683115B CN201910111648.9A CN201910111648A CN109683115B CN 109683115 B CN109683115 B CN 109683115B CN 201910111648 A CN201910111648 A CN 201910111648A CN 109683115 B CN109683115 B CN 109683115B
- Authority
- CN
- China
- Prior art keywords
- radio frequency
- power amplifier
- frequency power
- signal
- 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.)
- Active
Links
- 238000013528 artificial neural network Methods 0.000 claims abstract description 62
- 230000006870 function Effects 0.000 claims abstract description 39
- 238000006243 chemical reaction Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000000034 method Methods 0.000 claims description 29
- 238000001228 spectrum Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 11
- 230000008878 coupling Effects 0.000 claims description 7
- 238000010168 coupling process Methods 0.000 claims description 7
- 238000005859 coupling reaction Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 3
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims description 3
- 230000003446 memory effect Effects 0.000 description 12
- 210000002569 neuron Anatomy 0.000 description 12
- 230000008901 benefit Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000012549 training Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 5
- 238000012804 iterative process Methods 0.000 description 4
- 230000009022 nonlinear effect Effects 0.000 description 4
- 230000008447 perception Effects 0.000 description 4
- 230000001953 sensory effect Effects 0.000 description 4
- 238000012897 Levenberg–Marquardt algorithm Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 125000003275 alpha amino acid group Chemical group 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 150000001875 compounds Chemical class 0.000 description 2
- 238000002939 conjugate gradient method Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 238000003892 spreading Methods 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
Landscapes
- Physics & Mathematics (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- High Energy & Nuclear Physics (AREA)
- Amplifiers (AREA)
Abstract
The present disclosure provides a magnetic resonance radio frequency power amplifier device and a magnetic resonance system. Wherein the magnetic resonance radio frequency power amplifier arrangement comprises a first balun configured to receive the feed forward signal and to transmit it to the processor via a first analog-to-digital converter; a second balun configured to receive the feedback signal and transmit it to the processor via a second analog-to-digital converter; the processor is configured to perform digital frequency conversion and signal synchronization processing on the received feedforward signal and the feedback signal, train the radio frequency power amplifier inverse function estimator through the neural network indirect learning structure, and copy the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter; the output signal of the predistorter sequentially enters a radio frequency power amplifier module through a third balun, a digital-to-analog converter and an analog up-conversion module; the radio frequency power amplifier module amplifies the received signal and divides the amplified signal into two paths.
Description
Technical Field
The disclosure belongs to the field of magnetic resonance, and in particular relates to a magnetic resonance radio frequency power amplifier device and a magnetic resonance system.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
In a magnetic resonance system, a radio frequency power amplifier is required to amplify radio frequency pulses emitted by a magnetic resonance spectrometer to several tens to several hundreds watts and output the amplified radio frequency pulses to a transmitting coil to excite an experimental sample. The radio frequency power amplifier is thus an integral part of the magnetic resonance system. However, as with a typical power amplifier, the rf power amplifier has a nonlinear characteristic, which causes nonlinear distortion of the pulse signal when the rf power amplifier operates in a saturation region or a cut-off region. The greater the power of the input signal, the more severe the distortion of the radio frequency power amplifier. The distortion comprises in-band amplitude and phase distortion and out-of-band spectrum spreading, and interferes with signal transmission of adjacent channels, so that imaging of a magnetic resonance system finally causes artifacts and distortion, and the picture quality is reduced, and the disease diagnosis and scientific research are interfered. Therefore, it is necessary to achieve linearization of the power amplifier while ensuring the operating efficiency of the radio frequency power amplifier.
The method for solving the nonlinearity of the radio frequency power amplifier comprises the following steps:
(1) The most straightforward method is to fabricate high precision high linearity radio frequency power amplifiers. The inventors have found that it is complex and expensive to manufacture and is generally unsuitable for use with high power amplifiers.
(2) Power backoff is the simplest way to increase the linearity of a power amplifier by lowering the operating point of the power amplifier so that the power amplifier operates 10 to 15dB back-off from the saturation point. The power backoff method is simple to implement, but the inventor finds that the disadvantage is that the working efficiency of the power amplifier is greatly reduced, thereby increasing the maintenance cost of the system.
(3) The direct feedback method directly suppresses an input signal by using an output signal, and the inventor finds that the disadvantage is that it is difficult to estimate the delay of the output signal to the input signal, so that the system stability is poor. Therefore, an indirect negative feedback method is more often adopted in engineering, and the indirect negative feedback method refers to that output signals and input signals are compared through an indirect connection mode. Therefore, the indirect negative feedback method has the advantages of high precision, mature technology, low price and the like. The inventor finds that the disadvantage is that the feedback loop delay is difficult to control, the system is not stable enough, and the feedback loop is not suitable for occasions with wider frequency bands.
(4) The basic principle of the feedforward method is as follows: and the offset loop is used for separating out interference signals to carry out superposition offset on the delayed power amplifier output signals, so that linearization of the power amplifier is realized. The method has the advantages of high speed, good linearization and wide use bandwidth, but the inventor discovers that the method has the advantages of complex structure, high cost, low efficiency and poor self-adaptability.
(5) The basic idea of the predistortion technology is to insert a predistorter with the curve characteristic being reciprocal to the curve characteristic of the power amplifier before the power amplifier, and the linearization of the power amplifier is realized by using the predistorter and the cascade of the power amplifier. The predistorter of the predistortion technology is realized in a multi-purpose analog circuit mode, and has the advantages of low cost, simple circuit structure and wide adaptation bandwidth, however, the inventor finds that the predistortion effect is limited, and particularly, the predistortion on high-order component distortion is difficult.
In summary, the inventors have found that the current rf power amplifiers have low operating efficiency and poor predistortion effects during nonlinear processing.
Disclosure of Invention
In order to solve the above-mentioned problems, a first aspect of the present disclosure provides a magnetic resonance radio frequency power amplifier device, which has advantages of high working efficiency and good predistortion effect in nonlinear processing.
The technical scheme of the magnetic resonance radio frequency power amplifier device of the first aspect of the disclosure is that:
A magnetic resonance radio frequency power amplifier device of the present disclosure, comprising:
A first balun configured to receive the feedforward signal and to communicate to the processor via a first analog-to-digital converter;
a second balun configured to receive the feedback signal and transmit it to the processor via a second analog-to-digital converter;
The processor is configured to perform digital frequency conversion and signal synchronization processing on the received feedforward signal and the feedback signal, train the radio frequency power amplifier inverse function estimator through the neural network indirect learning structure, and copy the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter; the radio frequency power amplifier inverse function estimator and the predistorter have the same structure;
The output signal of the predistorter sequentially enters a radio frequency power amplifier module through a third balun, a digital-to-analog converter and an analog up-conversion module;
the radio frequency power amplifier module is configured to amplify a received signal and then divide the amplified signal into two paths, wherein one path of the amplified signal is transmitted, and the other path of the amplified signal is processed by the coupling module and the analog down-conversion module in sequence to obtain a feedback signal.
Further, the process of digital frequency conversion of the received feedforward signal and the feedback signal by the processor is as follows:
digitally down-converting the feedforward signal into a baseband signal, and then digitally up-converting the feedforward signal into a signal having a center frequency that corresponds to the center frequency of the processor;
the feedback signal is digitally down-converted to a baseband signal and then the feedback signal is digitally up-converted to a signal having a center frequency that matches the center frequency of the processor.
The advantage of the above scheme is that the received feedforward signal and the feedback signal are subjected to digital frequency conversion, so that the center frequencies of the received feedforward signal and the received feedback signal are both in accordance with the center frequency signal of the processor, thus the working efficiency of the processor can be improved, and the working stability of the whole magnetic resonance radio frequency power amplifier device is further improved.
Further, the predistorter is composed of a linear portion and a nonlinear portion.
Further, the linear part is an FIR filter structure; the nonlinear part is of a neural network structure.
Further, the neural network structure is a shallow learning neural network or a neural network based on deep learning.
The predistorter needs to fit the memory effect of the radio frequency power amplifier besides the nonlinear distortion of the radio frequency power amplifier, and the neural network predistorter also needs to have the memory effect.
It is noted that the neural network may be a conventional shallow learning neural network, such as a Back Propagation (BP) neural network, a multi-layer perception neural network, or the like, or may be a deep learning-based neural network, such as a deep neural network, a recurrent neural network, or a convolutional neural network.
The training algorithm adopted by the neural network can be a gradient descent method, an additional momentum method, a conjugate gradient method, a Newton algorithm, a Levenberg-Marquardt algorithm and the like.
Further, the process of training the radio frequency power amplifier inverse function estimator by the processor through the neural network indirect learning structure is as follows:
Adaptively calculating and updating the weight coefficient of the radio frequency power amplifier inverse function estimator according to the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator;
and when the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator is smaller than a certain value, copying the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter.
The scheme has the advantages that the predistorter copies the weight coefficient of the inverse function estimator of the radio frequency power amplifier to the predistorter when the predistorter converges, so that the working stability of the whole magnetic resonance radio frequency power amplifier device can be improved.
Further, the processor is further configured to:
According to the change of the radio frequency power amplifier module, the weight coefficient of the predistorter is updated periodically, the performance change of the radio frequency power amplifier module is monitored and tracked in real time, and the self-adaptive predistortion of the radio frequency power amplifier module is realized.
Further, the analog up-conversion module includes: a first digitally controlled oscillator, a band pass filter and a first mixer; the first digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the first mixer; the band-pass filter is used for filtering redundant frequency spectrums generated during up-conversion mixing.
Further, the analog down-conversion module includes: a second digitally controlled oscillator, a digital finite filter and a second mixer; the second digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the second mixer; the digital finite filter is used to filter out unwanted spectrum generated during the down-conversion mixing.
Further, the coupling module includes: the power divider is used for extracting a part of radio frequency signals from the radio frequency power amplifier, and the part of radio frequency signals are sequentially processed by the signal attenuator and the analog down-conversion module to obtain feedback signals.
In order to solve the above-mentioned problems, a second aspect of the present disclosure provides a magnetic resonance system having advantages of high working efficiency and good predistortion effect in a nonlinear processing.
The technical scheme of the magnetic resonance system in the second aspect of the disclosure is that:
a magnetic resonance system of the present disclosure, comprising:
A magnetic resonance signal generation device configured to generate a magnetic resonance pulse signal;
The magnetic resonance radio frequency power amplifier device is configured to perform predistortion amplification on a magnetic resonance pulse signal and output the magnetic resonance pulse signal to the magnetic resonance signal transmitting device;
The magnetic resonance signal transmitting device is configured to transmit a predistortion amplified magnetic resonance signal.
The beneficial effects of the present disclosure are:
(1) The method adopts the neural network to fit the nonlinearity of the radio frequency power amplifier of the magnetic resonance system, and predistorts the radio frequency power amplifier, thereby improving the working efficiency of the radio frequency power amplifier of the magnetic resonance system and the predistortion effect in the nonlinear processing process.
(2) The method utilizes the filter part to fit the memory effect of the radio frequency power amplifier, utilizes the neural network part to fit the nonlinear distortion of the radio frequency power amplifier, and further realizes the nonlinear and memory effects of the separated fitting radio frequency power amplifier. Compared with the traditional predistorter model with the memory neural network, the method not only greatly reduces the original network parameters and network scale, but also greatly reduces the calculated amount of the weight coefficient updating iterative process.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure, illustrate and explain the exemplary embodiments of the disclosure and together with the description serve to explain the disclosure, and do not constitute an undue limitation on the disclosure.
Fig. 1 is a schematic structural diagram of a magnetic resonance radio frequency power amplifier device according to an embodiment of the present disclosure.
Fig. 2 is a schematic diagram of a processor structure of a magnetic resonance radio frequency power amplifier device according to an embodiment of the disclosure.
Figure 3 is a schematic diagram of a magnetic resonance radio frequency power amplifier device provided by an embodiment of the present disclosure.
Fig. 4 is a schematic diagram of an indirect learning structure based on a multi-layer sensory neural network provided in an embodiment of the present disclosure.
Fig. 5 is a schematic diagram of a multi-layer sensory neural network provided by an embodiment of the present disclosure.
Fig. 6 is an AM/AM characteristic curve before predistortion provided by an embodiment of the present disclosure.
Fig. 7 is an AM/AM characteristic curve after predistortion provided by an embodiment of the present disclosure.
Fig. 8 is an AM/PM characteristic curve before predistortion provided by an embodiment of the present disclosure.
Fig. 9 is an AM/PM characteristic after predistortion provided by an embodiment of the present disclosure.
Fig. 10 is a power spectrum of an original signal and pre-and post-predistortion signals provided by an embodiment of the present disclosure.
Fig. 11 is a schematic structural diagram of a magnetic resonance system according to an embodiment of the present disclosure.
Detailed Description
It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments in accordance with the present disclosure. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
As shown in fig. 1, a magnetic resonance radio frequency power amplifier device of the present embodiment includes:
A first balun configured to receive the feedforward signal and to communicate to the processor via a first analog-to-digital converter;
a second balun configured to receive the feedback signal and transmit it to the processor via a second analog-to-digital converter;
The processor is configured to perform digital frequency conversion and signal synchronization processing on the received feedforward signal and the feedback signal, train the radio frequency power amplifier inverse function estimator through the neural network indirect learning structure, and copy the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter; the radio frequency power amplifier inverse function estimator and the predistorter have the same structure;
The output signal of the predistorter sequentially enters a radio frequency power amplifier module through a third balun, a digital-to-analog converter and an analog up-conversion module;
the radio frequency power amplifier module is configured to amplify a received signal and then divide the amplified signal into two paths, wherein one path of the amplified signal is transmitted, and the other path of the amplified signal is processed by the coupling module and the analog down-conversion module in sequence to obtain a feedback signal.
Wherein, balun: the balun is responsible for interconverting between high frequency single ended signals and differential signals.
An analog-to-digital converter: there is a differential input port that converts the analog signal to a digital signal.
A processor: may be implemented using DSP or FPGA chips.
The processor mainly realizes three functions of digital up-down conversion preprocessing, signal synchronization and digital predistortion of the magnetic resonance radio frequency power amplifier, as shown in fig. 2.
A digital-to-analog converter: the digital signal is converted into an analog signal, and the analog signal is output.
Analog up-conversion module: and up-regulating the intermediate frequency signal output by the digital-to-analog converter to a radio frequency band.
A radio frequency power amplifier module: the up-converted signal is amplified by a radio frequency power amplifier and then transmitted by a radio frequency coil.
And a coupling module: corresponding to a signal attenuator.
Analog down-conversion module: the radio frequency signal is down-converted.
Specifically, the analog up-conversion module includes: a first digitally controlled oscillator, a band pass filter and a first mixer; the first digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the first mixer; the band-pass filter is used for filtering redundant frequency spectrums generated during up-conversion mixing.
The analog down-conversion module includes: a second digitally controlled oscillator, a digital finite filter and a second mixer; the second digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the second mixer; the digital finite filter is used to filter out unwanted spectrum generated during the down-conversion mixing.
The coupling module includes: the power divider is used for extracting a part of radio frequency signals from the radio frequency power amplifier, and the part of radio frequency signals are sequentially processed by the signal attenuator and the analog down-conversion module to obtain feedback signals.
In a specific implementation, the process of digital frequency conversion of the received feedforward signal and the feedback signal by the processor is as follows:
digitally down-converting the feedforward signal into a baseband signal, and then digitally up-converting the feedforward signal into a signal having a center frequency that corresponds to the center frequency of the processor;
the feedback signal is digitally down-converted to a baseband signal and then the feedback signal is digitally up-converted to a signal having a center frequency that matches the center frequency of the processor.
The advantage of the above scheme is that the received feedforward signal and the feedback signal are subjected to digital frequency conversion, so that the center frequencies of the received feedforward signal and the received feedback signal are both in accordance with the center frequency signal of the processor, thus the working efficiency of the processor can be improved, and the working stability of the whole magnetic resonance radio frequency power amplifier device is further improved.
The neural network indirect learning structure is also implemented on a DSP or FPGA chip, as shown in fig. 3, in which the predistorter and the rf power amplifier inverse function estimator are identical in structure. The predistorter is composed of a linear portion and a nonlinear portion.
The linear part is of an FIR filter structure; the nonlinear part is of a neural network structure.
The neural network structure is a shallow learning neural network or a neural network based on deep learning.
The predistorter needs to fit the memory effect of the radio frequency power amplifier besides the nonlinear distortion of the radio frequency power amplifier, and the neural network predistorter also needs to have the memory effect.
It is noted that the neural network may be a conventional shallow learning neural network, such as a Back Propagation (BP) neural network, a multi-layer perception neural network, or the like, or may be a deep learning-based neural network, such as a deep neural network, a recurrent neural network, or a convolutional neural network.
The training algorithm adopted by the neural network can be a gradient descent method, an additional momentum method, a conjugate gradient method, a Newton algorithm, a Levenberg-Marquardt algorithm and the like.
Further, the process of training the radio frequency power amplifier inverse function estimator by the processor through the neural network indirect learning structure is as follows:
Adaptively calculating and updating the weight coefficient of the radio frequency power amplifier inverse function estimator according to the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator;
and when the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator is smaller than a certain value, copying the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter.
The scheme has the advantages that the predistorter copies the weight coefficient of the inverse function estimator of the radio frequency power amplifier to the predistorter when the predistorter converges, so that the working stability of the whole magnetic resonance radio frequency power amplifier device can be improved.
In an implementation, the processor is further configured to:
according to the change of the radio frequency power amplifier module, the weight coefficient of the predistorter is updated periodically, the change of the module performance of the radio frequency power amplifier is monitored and tracked in real time, and the self-adaptive predistortion of the radio frequency power amplifier module is realized.
The following describes in detail an indirect learning structure based on a multi-layer sensory neural network, as an example in fig. 4:
A predistorter is a predistortion system that consists of two parts, linear and nonlinear. The linear part is an FIR filter, and the nonlinear part is a double-input double-output 3-layer common multi-layer perception neural network. And fitting the memory effect of the radio frequency power amplifier by using the filter part, and fitting the nonlinear distortion of the radio frequency power amplifier by using the neural network part so as to realize the nonlinear and memory effects of the separated fitting radio frequency power amplifier.
Compared with the traditional predistorter model with the memory neural network, the indirect learning structure based on the multi-layer perception neural network not only greatly reduces the original network parameters and network scale, but also greatly reduces the calculated amount of the weight coefficient updating iterative process.
The multi-layer sensory neural network described in the embodiment of fig. 5, the feedforward is calculated as follows:
in the formula (1), the components are as follows, Output of the jth neuron of the 1 st layer of the multi-layer sensing neural network, M is memory depth,/>The connection weight coefficient of the jth neuron of the 1 st layer and the ith input of the input layer is x -i, and the ith input of the neural network;
in the formula (2), the amino acid sequence of the compound, For the output of layer 2, jth neuron, the hidden layer activates the function/>As tanh function,/>For the connection weight coefficient of the jth neuron of the 2 nd layer and the ith neuron of the 1 st layer,/>For the output of layer 1, jth neuron,/>Bias coefficients for layer 2, jth neuron;
In the formula (3), the amino acid sequence of the compound, For the output of layer 3, the kth neuron, l 1 is the number of hidden layer nodes,/>For the connection weight coefficient of the layer 3 kth neuron and the layer 2 jth neuron,/>Is the bias coefficient of the layer 3 kth neuron. F (x) =x is chosen as the output layer neuron activation function.
The Bayesian-Levenberg-Marquardt optimization algorithm is adopted as a training algorithm of the neural network, and the objective function expression is as follows:
F(X)＝αEW+βED(4)
In the middle of Wherein N is the number of rounds of neural network weight update iterations; s2 is the number of nodes of the output layer of the neural network; q is the number of training samples per round,/>When the number of training samples is q, the neural network predicts the square of error values of the true ware and the inverse function estimator; v j (x) is the square of the error value of the predictor and the inverse function estimator in the j-th iteration process; w j is a network weight coefficient; alpha and beta are coefficients; m is the total number of weight coefficients in the network, and is a positive integer greater than or equal to 1.
Therefore, through the new objective function, the network can ensure that the network output error is as small as possible in the training process and simultaneously ensure that the network has smaller network weight coefficient. The iterative formula for updating the neural network coefficients by Bayesian-Levenberg-Marquardt is as follows:
Xk+1＝Xk-[αJTJ-(μ+β)I]-1JTe(5)
wherein μ and e are both constant coefficients.
In the above formula, J is Jacobi matrix, and its expression is
Wherein X is a weight coefficient vector of the network. The optimized values for the coefficients α and β are α MP and β MP, respectively, and are given by:
Wherein γ=m-2α MP·tr(HMP)-1 represents the number of significant weight coefficients, ranging from 0 to m. m is the total number of weight coefficients in the network, and H MP is the Hessian matrix of the objective function F (X) at its minimum point X MP. The Hessian matrix needs to be calculated during the calculation process. Simplifying the Hessian matrix by using Gaussian-Newton approximation, then: is 2F(XMP)≈βJTJ+αIm, where J is the Jacobi matrix of E D at point X MP.
The Bayesian-LM algorithm operates as follows:
⑴ Network parameters are initialized and the initialization coefficients α=0, β=1.
⑵ The network performance objective function F (X) =αe W+βED was minimized using the Levenberg-Marquardt algorithm.
⑶ And solving H (approximately equal to beta) J TJ+αIm by using a Gaussian-Newton approximation method, and solving the number gamma of the effective parameters.
⑷ Calculating new estimated value of coefficient
⑸ Steps ⑵ to ⑷ are repeated until the algorithm converges.
The AM/AM characteristic diagram of the radio frequency power amplifier before predistortion is shown in fig. 6. The AM/AM profile after pre-test is shown in figure 7. As can be seen from fig. 6 and 7: the AM/AM characteristic of the radio frequency power amplifier before predistortion is a nonlinear curve of the banding hysteresis. The AM/AM characteristic of the radio frequency power amplifier with predistorter is almost a straight line with a slope of 1 and no hysteresis, indicating that the amplitude amplification is substantially linearized.
An AM/PM characteristic diagram of the radio frequency power amplifier before predistortion is shown in fig. 8. The AM/PM characteristic after pre-test is shown in FIG. 9. As can be seen from fig. 8 and 9: before predistortion, when the input is changed from 0 to 1, the phase on the AM/PM characteristic diagram of the radio frequency power amplifier is shifted to different degrees, and the phase shift amount is larger when the input is smaller. The AM/PM characteristic after predistortion has an output phase offset of substantially 0 when the input amplitude varies, indicating that the objective of phase predistortion is substantially achieved.
Fig. 10 shows power spectra of the rf amplifier before and after predistortion, and the in-band signal becomes flat after predistortion, and the predistorter structure can effectively reduce the adjacent channel power ratio by about 30 dB.
The magnetic resonance radio frequency power amplifier device of the embodiment adopts a neural network to fit the nonlinearity of the radio frequency power amplifier of the magnetic resonance system, and predistortion is carried out on the radio frequency power amplifier, so that the working efficiency of the radio frequency power amplifier device of the magnetic resonance system and the predistortion effect in the nonlinear processing process are improved.
The magnetic resonance radio frequency power amplifier device of the embodiment utilizes the filter part to fit the memory effect of the radio frequency power amplifier, and utilizes the neural network part to fit the nonlinear distortion of the radio frequency power amplifier, thereby realizing the nonlinear and memory effects of the separated fit radio frequency power amplifier. Compared with the traditional predistorter model with the memory neural network, the method not only greatly reduces the original network parameters and network scale, but also greatly reduces the calculated amount of the weight coefficient updating iterative process.
Fig. 11 is a schematic structural diagram of a magnetic resonance system according to an embodiment of the present disclosure.
As shown in fig. 11, a magnetic resonance system of the present embodiment includes:
A magnetic resonance signal generation device configured to generate a magnetic resonance pulse signal;
A magnetic resonance radio frequency power amplifier device as shown in fig. 1 configured to predistort and amplify a magnetic resonance pulse signal and output to a magnetic resonance signal transmitting device;
The magnetic resonance signal transmitting device is configured to transmit a predistortion amplified magnetic resonance signal.
In particular, the magnetic resonance signal generating means may be realized by means of a magnetic resonance spectrometer.
The magnetic resonance signal transmitting means may be implemented by radio frequency coils of a magnetic resonance system.
The magnetic resonance system of the embodiment adopts a neural network to fit the nonlinearity of the radio frequency power amplifier of the magnetic resonance system, and predistortion is carried out on the radio frequency power amplifier, so that the working efficiency of the radio frequency power amplifier device of the magnetic resonance system and the predistortion effect in the nonlinear processing process are improved.
The magnetic resonance system of the embodiment utilizes the filter part to fit the memory effect of the radio frequency power amplifier, utilizes the neural network part to fit the nonlinear distortion of the radio frequency power amplifier, and further realizes the nonlinear and memory effects of the separated fitting radio frequency power amplifier. Compared with the traditional predistorter model with the memory neural network, the method not only greatly reduces the original network parameters and network scale, but also greatly reduces the calculated amount of the weight coefficient updating iterative process.
While the specific embodiments of the present disclosure have been described above with reference to the drawings, it should be understood that the present disclosure is not limited to the embodiments, and that various modifications and changes can be made by one skilled in the art without inventive effort on the basis of the technical solutions of the present disclosure while remaining within the scope of the present disclosure.
Claims (8)
1. A magnetic resonance radio frequency power amplifier apparatus, comprising:
A first balun configured to receive the feedforward signal and to communicate to the processor via a first analog-to-digital converter;
a second balun configured to receive the feedback signal and transmit it to the processor via a second analog-to-digital converter;
The processor is configured to perform digital frequency conversion and signal synchronization processing on the received feedforward signal and the feedback signal, train the radio frequency power amplifier inverse function estimator through the neural network indirect learning structure, and copy the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter; the radio frequency power amplifier inverse function estimator and the predistorter have the same structure;
the process of digital frequency conversion of the received feedforward signal and the feedback signal by the processor is as follows:
digitally down-converting the feedforward signal into a baseband signal, and then digitally up-converting the feedforward signal into a signal having a center frequency that corresponds to the center frequency of the processor;
digitally down-converting the feedback signal to a baseband signal, and then digitally up-converting the feedback signal to a signal having a center frequency that corresponds to the center frequency of the processor;
The processor is further configured to:
according to the change of the radio frequency power amplifier module, the weight coefficient of the predistorter is updated periodically, the performance change of the radio frequency power amplifier module is monitored and tracked in real time, and the self-adaptive predistortion of the radio frequency power amplifier module is realized;
The output signal of the predistorter sequentially enters a radio frequency power amplifier module through a third balun, a digital-to-analog converter and an analog up-conversion module;
the radio frequency power amplifier module is configured to amplify a received signal and then divide the amplified signal into two paths, wherein one path of the amplified signal is transmitted, and the other path of the amplified signal is processed by the coupling module and the analog down-conversion module in sequence to obtain a feedback signal.
2. A magnetic resonance radio frequency power amplifier arrangement as claimed in claim 1, characterized in that the predistorter is formed by a linear part and a non-linear part.
3. A magnetic resonance radio frequency power amplifier arrangement as claimed in claim 2, characterized in that the linear part is a FIR filter structure; the nonlinear part is of a neural network structure.
4. A magnetic resonance radio frequency power amplifier device as claimed in claim 3, characterized in that the neural network structure is a shallow learning neural network or a deep learning based neural network.
5. The magnetic resonance radio frequency power amplifier device as set forth in claim 1, wherein the processor trains the radio frequency power amplifier inverse function estimator through a neural network indirect learning structure by:
Adaptively calculating and updating the weight coefficient of the radio frequency power amplifier inverse function estimator according to the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator;
and when the output difference value of the predistorter and the radio frequency power amplifier inverse function estimator is smaller than a certain value, copying the weight coefficient of the radio frequency power amplifier inverse function estimator to the predistorter.
6. A magnetic resonance radio frequency power amplifier arrangement as claimed in claim 1, characterized in that the analog up-conversion module comprises: a first digitally controlled oscillator, a band pass filter and a first mixer; the first digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the first mixer; the band-pass filter is used for filtering redundant frequency spectrums generated during up-conversion mixing;
or the analog down-conversion module comprises: a second digitally controlled oscillator, a digital finite filter and a second mixer; the second digital control oscillator is used for generating sine and cosine two-way quadrature signals and completing frequency spectrum relocation through the second mixer; the digital finite filter is used to filter out unwanted spectrum generated during the down-conversion mixing.
7. A magnetic resonance radio frequency power amplifier device as claimed in claim 1, characterized in that the coupling module comprises: the power divider is used for extracting a part of radio frequency signals from the radio frequency power amplifier, and the part of radio frequency signals are sequentially processed by the signal attenuator and the analog down-conversion module to obtain feedback signals.
8. A magnetic resonance system, comprising:
A magnetic resonance signal generation device configured to generate a magnetic resonance pulse signal;
The magnetic resonance radio frequency power amplifier apparatus as claimed in any one of claims 1-7, configured to predistort and amplify a magnetic resonance pulse signal and output to a magnetic resonance signal transmitting apparatus;
The magnetic resonance signal transmitting device is configured to transmit a predistortion amplified magnetic resonance signal.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910111648.9A CN109683115B (en) | 2019-02-12 | 2019-02-12 | 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 |
| PCT/CN2020/074422 WO2020164435A1 (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 | 2019-02-12 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109683115A CN109683115A (en) | 2019-04-26 |
| CN109683115B true CN109683115B (en) | 2024-05-03 |
Family
ID=66195803
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910111648.9A Active CN109683115B (en) | 2019-02-12 | 2019-02-12 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
Country Status (4)
| Country | Link |
|---|---|
| CN (1) | CN109683115B (en) |
| GB (1) | GB2595828B (en) |
| WO (1) | WO2020164435A1 (en) |
| ZA (1) | ZA202106725B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109683115B (en) * | 2019-02-12 | 2024-05-03 | 泰山医学院 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
| 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 |
| CN111324861B (en) * | 2020-02-28 | 2022-05-03 | 厦门大学 | Deep learning magnetic resonance spectrum reconstruction method based on matrix decomposition |
Citations (19)
| 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 | 京信通信系统（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 |
| 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 |
| 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 |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2376581B (en) * | 2001-06-15 | 2005-03-09 | Wireless Systems Int Ltd | Control scheme for signal processing arrangement |
| CN101675354A (en) * | 2007-05-04 | 2010-03-17 | 皇家飞利浦电子股份有限公司 | The RF transmitter that is used for MRI with digital feedback |
| US8023588B1 (en) * | 2008-04-08 | 2011-09-20 | Pmc-Sierra, Inc. | Adaptive predistortion of non-linear amplifiers with burst data |
| CN102613974B (en) * | 2011-01-26 | 2014-08-20 | 成都芯通科技股份有限公司 | Digital nuclear magnetic resonance radio-frequency amplifier and realization method thereof |
| CN102281221B (en) * | 2011-06-23 | 2017-03-15 | 中兴通讯股份有限公司 | Nonlinear systematic distortion correction device and method |
| CN102427336B (en) * | 2011-11-30 | 2015-07-08 | 钱骅 | Radio frequency power amplification system with function of adaptive digital predistortion linearization |
| CN103378391B (en) * | 2012-04-27 | 2016-08-31 | Ge医疗系统环球技术有限公司 | A kind of balun |
| KR101389880B1 (en) * | 2012-10-31 | 2014-05-07 | 한국과학기술원 | Apparatus and method for low cost implementation of adaptive digital predistortion algorithm using envelope detection feedback |
| US8989307B2 (en) * | 2013-03-05 | 2015-03-24 | Qualcomm Incorporated | Power amplifier system including a composite digital predistorter |
| CN109683115B (en) * | 2019-02-12 | 2024-05-03 | 泰山医学院 | Magnetic resonance radio frequency power amplifier device and magnetic resonance system |
-
2019
- 2019-02-12 CN CN201910111648.9A patent/CN109683115B/en active Active
-
2020
- 2020-02-06 GB GB2112953.1A patent/GB2595828B/en active Active
- 2020-02-06 WO PCT/CN2020/074422 patent/WO2020164435A1/en active Application Filing
-
2021
- 2021-09-10 ZA ZA2021/06725A patent/ZA202106725B/en unknown
Patent Citations (19)
| 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 |
| CN101110571A (en) * | 2007-08-30 | 2008-01-23 | 京信通信系统（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 |
| 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 (5)
| Title |
|---|
| OFDM系统中的一种自适应预失真器;杨文考, 周尚波, 朱维乐;电波科学学报(第02期);全文 * |
| 一种改进的数字预失真迭代算法;李环;张羽;祝乐芳;;沈阳理工大学学报(第03期);全文 * |
| 基于ACR体模的磁共振射频线圈性能的对比;袁子龙;张照喜;邱建峰;陈迢;刘涛;刘玉林;;中国医疗设备(第01期);全文 * |
| 基于BP神经网络的功放自适应预失真;邓洪敏, 何松柏, 虞厥邦;通信学报(第11期);全文 * |
| 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.2010,(第03期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| GB2595828A (en) | 2021-12-08 |
| WO2020164435A1 (en) | 2020-08-20 |
| GB202112953D0 (en) | 2021-10-27 |
| ZA202106725B (en) | 2021-10-27 |
| CN109683115A (en) | 2019-04-26 |
| GB2595828B (en) | 2023-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11245429B2 (en) | System and method for increasing bandwidth for digital predistortion in multi-channel wideband communication systems | |
| US11095326B2 (en) | Wide bandwidth digital predistortion system with reduced sampling rate | |
| CN109683115B (en) | Magnetic resonance radio frequency power amplifier device and magnetic resonance system | |
| EP1402700B1 (en) | Adaptive predistorter based on the probability distribution function of the output amplitude | |
| US9461698B2 (en) | Communications device with simultaneous transmit and receive and related methods | |
| CN100556015C (en) | A kind of pre-distortion device and method | |
| CN100594669C (en) | Power amplifier predistortion method of Hammerstein model based on fuzzy neural network | |
| DE102015101197A1 (en) | Adaptively controlled digital predistortion in an integrated signal analyzer with improved analog-to-digital conversion and RF power amplifiers using the same | |
| US8649745B2 (en) | Adaptive predistortion for a non-linear subsystem based on a model as a concatenation of a non-linear model followed by a linear model | |
| CN111064439A (en) | System and method for improving short-wave digital predistortion performance | |
| CN114400979A (en) | System and method for improving digital predistortion of short wave interaction and harmonic distortion | |
| EP2795802A2 (en) | Architecture of a low bandwidth predistortion system for non-linear rf components | |
| CN1531195B (en) | Enlarger controlling system for counting improvement of digital controlling signal resolution ratio | |
| CN115118562A (en) | Dual-target optimization self-adaptive radio frequency predistortion system and method | |
| WO2023028795A1 (en) | Frequency predistortion apparatus and frequency predistortion method | |
| CN113346930A (en) | Beam-oriented feed-forward architecture system and method for use in a multi-antenna system | |
| CN115940833A (en) | Digital predistortion method, storage medium and apparatus for integrated correction of frequency response of shortwave power amplifier | |
| Wei et al. | The research of adaptive digital predistortion based on SISO-Neural network | |
| Hammi et al. | Subband Adaptive Digital Predistorter For 3G Applications “ | |
| CN107154804A (en) | The binary channels TIADC Parameter Estimation of Nonlinear Systems methods of low-pass signal |
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 |