CN102487367B - Adaptive amplifying digital baseband pre-distortion method - Google Patents

Adaptive amplifying digital baseband pre-distortion method Download PDF

Info

Publication number
CN102487367B
CN102487367B CN201010571253.6A CN201010571253A CN102487367B CN 102487367 B CN102487367 B CN 102487367B CN 201010571253 A CN201010571253 A CN 201010571253A CN 102487367 B CN102487367 B CN 102487367B
Authority
CN
China
Prior art keywords
distortion
digital baseband
coefficient
covariance matrix
described step
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201010571253.6A
Other languages
Chinese (zh)
Other versions
CN102487367A (en
Inventor
张�浩
阎跃鹏
陈立平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Zhongke Micro Investment Management Co ltd
Original Assignee
Institute of Microelectronics of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Institute of Microelectronics of CAS filed Critical Institute of Microelectronics of CAS
Priority to CN201010571253.6A priority Critical patent/CN102487367B/en
Publication of CN102487367A publication Critical patent/CN102487367A/en
Application granted granted Critical
Publication of CN102487367B publication Critical patent/CN102487367B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Amplifiers (AREA)

Abstract

The invention relates to an adaptive amplifying digital baseband pre-distortion method. The method comprises the following detailed steps of: step A, judging whether an input signal and an output signal of a pre-distortion system are equal or not, if so, disconnecting with a pre-distortion training module, otherwise, performing step B; and step B, using a training method of a memory multinomial coefficient of Kalman filter to set the to-be-estimated memory multinomial coefficient of the pre-distortion training module to be a state variable, establishing a state space model, and using an estimation value of the prior time and a sampling value of the present time to update the state variable to acquire the estimation value of the present time. Compared with the traditional pre-distortion scheme, the technical scheme provided by the invention has the advantages of high robustness and convergence rate; and the requirement of updating parameters in real time can be met through the technical scheme.

Description

A kind of adaptive amplifying digital baseband pre-distortion method
Technical field
The present invention relates to digital baseband signal process field, relate in particular to a kind of adaptive amplifying digital baseband pre-distortion method.
Background technology
Power amplifier (abbreviation power amplifier) is most important device in communication system, between its input and output, has unintentional nonlinearity, makes signal occur distortion.In modern communications standard, often adopt some novel modulation techniques (for example 16QAM, OFDM etc.), to improve the availability of frequency spectrum.But these modulation techniques are non-linear very responsive to power amplifier.Non-linearly cause spectral re-growth or the expansion outside signal bandwidth, adjacent channel has been caused to interference.Meanwhile, thisly non-linearly in signal bandwidth, also cause distortion, thereby worsened the bit error rate of system.In addition, the signal with higher peak-to-average power ratio is especially easily subject to the impact of nonlinear distortion, causes the generation of power amplifier memory effect, and the characteristic of power amplifier is changed.Therefore, ensure the service behaviour of power amplifier and the normal transmission of signal, must be to the non-linear linearization process of carrying out of power amplifier.
The linearization of power amplifier mainly can be divided into feed-forward technique, negative-feedback technology and pre-distortion technology.In these technology, pre-distortion technology efficiency is high, cost is low and applicable broadband connections, has good prospect.Pre-distortion technology is divided into analog radio frequency predistortion and digital baseband predistortion (Digital Predistortion, DPD) technology.Analog radio frequency pre-distortion technology compensates the non-linear of power amplifier in radio frequency part, need to use the non-linear active device of radio frequency, has that difficulty is high, the shortcoming of wayward processing.On the contrary, digital baseband pre-distortion technology does not relate to radiofrequency signal processing, only in base band, signal is compensated, and is convenient to adopt Digital Signal Processing, thereby is widely applied.
For the nonlinear memory characteristic of power amplifier, in order to reduce the number of parameters of required estimation, reduce computation complexity, conventionally adopt memory multinomial model to describe.The non-linear of power amplifier be can compensate well based on the polynomial predistortion model of memory, direct study and two kinds of structures of non-direct study comprised.Indirect learning architecture has advantages of that the function of need not negating obtains finally remembering polynomial parameter, thereby has obtained application in increasing occasion.
The self adaptation of the memory multinomial coefficient of predistortion model is estimated conventionally to adopt based on least mean-square error (LMS) criterion or the algorithm for estimating based on iterative least square (RLS) criterion, and convergence of algorithm speed and robustness are the challenges of real-time update always.For this problem, in existing paper and patent, some improved algorithm for estimating are proposed, for example the method based on neural net, method based on wavelet transformation etc.But these methods have increased very large amount of calculation and memory space with respect to traditional algorithm based on LMS criterion (or RLS criterion), thereby have limited to a great extent their practical application.
Summary of the invention
In order to solve the deficiency that in prior art, amount of calculation is large and memory space is large, the present invention has proposed a kind of adaptive amplifying digital baseband pre-distortion method especially.
The invention provides a kind of adaptive amplifying digital baseband pre-distortion method, specifically comprise the steps:
Whether steps A, the input signal that judges pre-distortion system and output signal equate, are to disconnect with predistortion module; Otherwise, execution step B;
The training method of the memory multinomial coefficient of step B, employing Kalman filtering, the memory multinomial coefficient to be estimated of setting predistortion module is state variable, set up state-space model, utilize the sampled value of estimated value, current time input signal and the output signal of previous moment pre-distortion coefficients to upgrade state variable, obtain the estimated value of current time pre-distortion coefficients.
Described step B specifically comprises the steps:
Step B1, setting pre-distortion coefficients vector are time variable, and the coefficient vector w (n) of initialization predistortion training module, and setting w (n) estimation error covariance matrix is P (n), given constant σ 1..., σ k, r;
Step B2, respectively described input signal and output signal are sampled, form digital list entries x (n) and output sequence z (n);
Step B3: utilize constant σ given in described step B1 1..., σ kupgrade covariance matrix P (n);
Step B4: according to given constant r in covariance matrix P (n), described output sequence z (n) after upgrading in described step B3 and described step B1, calculated gains matrix K (n);
Step B5: utilize described gain matrix K (n), described output sequence z (n), again upgrade covariance matrix P (n), and utilize P (n) and P (n) after again upgrading taverage replacement upgrade P (n+1);
Step B6: according to described coefficient vector w (n), described list entries x (n) and described output sequence z (n), calculate residual error e (n);
Step B7: utilize described gain matrix K (n) and described residual error e (n) to upgrade coefficient vector, obtain w (n+1);
Step B8: judge whether to meet power amplifier linearization requirement, if meet, renewal process finishes; If do not meet, put n=n+1, return to step B2.
Coefficient vector w (n)=[w in described step B1 10(n) ..., w 1Q(n) ..., w k0(n) ..., w kQ(n)] t
Wherein, in coefficient vector w (n), comprise the individual estimated parameter that needs of K (Q+1), the individual coefficient of the polynomial K of corresponding predistortion (Q+1);
And described covariance matrix P (n) is the symmetry square matrix of a K (Q+1) × K (Q+1).
List entries x (n) in described step B2=[x (n) ..., x (n-Q) ..., x (n) | x (n) | ..., x (n-Q) | x (n-Q) | k-1] t;
Described output sequence z (n)=[z (n) ..., z (n-Q) ..., z (n) | z (n) | ..., z (n-Q) | z (n-Q) | k-1] t.
In described step B3, covariance matrix P (n) only carries out diagonal entry renewal, and concrete steps are as follows:
The diagonal entry of setting P (n) is P ii(n), wherein, 1≤i≤K (Q+1);
The individual order of elements of K (Q+1) is divided into K group, and every group comprises Q+1 element;
For the element P in L group ii(n), more new formula is as follows: P ii(n)=P ii(n)+σ l, wherein, 1≤L≤K.
Gain matrix in described step B4 K ( n ) = P ( n ) z ( n ) z ( n ) T P ( n ) z ( n ) + r .
In described step B5, again upgrade covariance matrix P (n)=(I-K (n) z (n) t) P (n);
Average replacement upgraded P ( n + 1 ) = 1 2 ( P ( n ) + P ( n ) T ) .
Residual error e (n)=w (n) in described step B6 t(x (n)-z (n)), wherein, w (n) is current coefficient estimate vector.
W (n+1)=w (n)+K (n) e (n) in described step 7.
The present invention can compensate power amplifier non-linear that has memory effect adaptively, there is strong robustness, fast convergence rate with respect to traditional predistortion scheme, can meet parameter real-time update, and computation complexity and memory space lower, be convenient to software or hard-wired occasion.
Brief description of the drawings
Fig. 1 is the structural representation of power amplifier digital baseband pre-distortion system of the present invention;
Fig. 2 is the flow chart of amplifying digital baseband pre-distortion method of the present invention.
Embodiment
With reference to Fig. 1, a kind of structural representation of self adaptation power amplifier digital baseband pre-distortion system, wherein this system nature is the predistortion module of connecting before power amplifier, described predistortion module at least comprises predistortion module and predistortion training module, the structure of the two is identical, and actual predistortion module is to be applied in the predistortion module of series connection after the parameter of predistortion training module is copied.Wherein comparison module, for according to coefficient vector w (n), described list entries x (n) and described output sequence z (n), calculates residual error e (n); And export residual error e (n) to predistortion training module, to upgrade coefficient vector, obtain w (n+1).
Predistortion adopts the polynomial construction that has memory, and its input signal and output signal relation are respectively:
y 1 ( n ) = Σ k = 1 K Σ q = 0 Q w kq x ( n - q ) | x ( n - q ) | k - 1 - - - ( 1 )
y 2 ( n ) = Σ k = 1 K Σ q = 0 Q w kq z ( n - q ) | z ( n - q ) | k - 1 - - - ( 2 )
Wherein, K is polynomial exponent number, and Q is memory span, the individual coefficient W of K (Q+1) kq(1≤k≤K, 0≤q≤Q) is parameter to be estimated.Can be regarded as the process of a parameter training for the estimation of predistortion multinomial coefficient.
Reach at pre-distortion system under the condition of stable (being that power amplifier non-linear obtains fine compensation), input signal x (n) should equate with output signal z (n), the parameter W now obtaining kq(1≤k≤K, 0≤q≤Q) is called best estimate parameter, is equivalent to the disconnection that is connected of training module and system.In the time that the characteristic of power amplifier changes, training network and system connect again, recover parameter W kqthe training process of (1≤k≤K, 0≤q≤Q).
The present invention proposes a kind of training method of the memory multinomial coefficient based on Kalman filtering.Suppose that memory multinomial coefficient to be estimated in predistortion model is state variable, set up state-space model, utilize the sampled value of estimated value, current time input signal and the output signal of previous moment pre-distortion coefficients to upgrade state variable, obtain the estimated value of current time pre-distortion coefficients.For the general principle of Kalman filtering, in this area, know personnel and all should be appreciated that, do not repeat them here.
With reference to Fig. 2, the adaptive amplifying digital baseband pre-distortion method of one that the embodiment of the present invention proposes, specifically comprises the steps:
Whether steps A, the input signal that judges pre-distortion system and output signal equate, are to disconnect with predistortion training module; Otherwise, execution step B;
The training method of the memory multinomial coefficient of step B, employing Kalman filtering, the memory multinomial coefficient to be estimated of setting predistortion training module is state variable, set up state-space model, utilize the sampled value of estimated value, current time input signal and the output signal of previous moment pre-distortion coefficients to upgrade state variable, obtain the estimated value of current time pre-distortion coefficients.
Described step B specifically comprises the steps:
The coefficient vector w (n) of step 101, initialization predistortion training module, it comprises the individual estimated parameter that needs of K (Q+1), the corresponding individual coefficient of the polynomial K of predistortion (Q+1) to set w (n) estimation error covariance matrix be P (n), given constant σ 1..., σ k, r; Wherein, the renewal of coefficient vector w (n) is that iteration is carried out, and sets n moment coefficient vector to be estimated to be: w (n)=[w 10(n) ..., w 1Q(n) ..., w k0(n) ..., w kQ(n)] t.
In coefficient vector w (n), comprise the individual estimated parameter that needs of K (Q+1), the individual coefficient of the polynomial K of its corresponding predistortion (Q+1), the estimation error covariance matrix of setting coefficient vector w (n) is P (n).It should be noted that, P (n) is the symmetric positive definite square formation of a K (Q+1) × K (Q+1), and when initial, P (n) is set to K (Q+1) I, and wherein I is unit matrix.
Step 102, respectively described input signal and output signal are sampled, form digital list entries x (n) and output sequence z (n); Wherein, list entries x (n)=[x (n) ..., x (n-Q) ..., x (n) | x (n) | ..., x (n-Q) | x (n-Q) | k-1] t;
Described output sequence z (n)=[z (n) ..., z (n-Q) ..., z (n) | z (n) | ..., z (n-Q) | z (n-Q) | k-1] t.
Step 103: utilize constant σ given in described step B1 1..., σ kupgrade covariance matrix P (n); Wherein, covariance matrix P (n) only carries out diagonal entry renewal, and concrete steps are as follows:
The diagonal entry of setting P (n) is P ii(n), wherein, 1≤i≤K (Q+1);
The individual order of elements of K (Q+1) is divided into K group, and every group comprises Q+1 element;
For the element P in L group ii(n), more new formula is as follows: P ii(n)=P ii(n)+σ l, wherein, 1≤L≤K.
Step 104: according to given constant r in covariance matrix P (n), described output sequence z (n) after upgrading in described step B3 and described step B1, calculated gains matrix K (n); Wherein, gain matrix K ( n ) = P ( n ) z ( n ) z ( n ) T P ( n ) z ( n ) + r .
Step 105: utilize described gain matrix K (n), described output sequence z (n), again upgrade covariance matrix P (n), again upgrade covariance matrix P (n)=(I-K (n) z (n) t) P (n);
And utilize again upgrade after P (n) and P (n) taverage replacement upgrade P ( n + 1 ) = 1 2 ( P ( n ) + P ( n ) T ) .
Step 106: according to described coefficient vector w (n), described list entries x (n) and described output sequence z (n), calculate residual error e (n), wherein, residual error e (n)=w (n) t(x (n)-z (n)), w (n) is current coefficient estimate vector.
Step 107: utilize described gain matrix K (n) and described residual error e (n) to upgrade coefficient vector, obtain w (n+1); Wherein, w (n+1)=w (n)+K (n) e (n).
Step 108: judge whether to meet power amplifier linearization requirement, if meet, renewal process finishes; If do not meet, put n=n+1, return to step B2.
The present invention can compensate power amplifier non-linear that has memory effect adaptively, there is strong robustness, fast convergence rate with respect to traditional predistortion scheme, can meet parameter real-time update, and computation complexity and memory space lower, be convenient to software or hard-wired occasion.
The foregoing is only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.All any amendments of doing within the spirit and principles in the present invention, be equal to replacement, improvement etc., within protection scope of the present invention all should be included in.

Claims (8)

1. an adaptive amplifying digital baseband pre-distortion method, is characterized in that comprising the steps:
Whether steps A, the input signal that judges pre-distortion system and output signal equate, are to disconnect with predistortion training module; Otherwise, execution step B;
The training method of the memory multinomial coefficient of step B, employing Kalman filtering, the memory multinomial coefficient to be estimated of setting predistortion module is state variable, set up state-space model, utilize the sampled value of estimated value, current time input signal and the output signal of previous moment pre-distortion coefficients to upgrade state variable, obtain the estimated value of current time pre-distortion coefficients;
Described step B specifically comprises the steps:
Step B1, setting pre-distortion coefficients vector are time variable, and the coefficient vector w (n) of initialization predistortion module, and setting w (n) estimation error covariance matrix is P (n), given constant σ 1..., σ k, r;
Step B2, respectively described input signal and output signal are sampled, form digital list entries x (n) and output sequence z (n);
Step B3: utilize constant σ given in described step B1 1..., σ kupgrade covariance matrix P (n);
Step B4: according to given constant r in covariance matrix P (n), described output sequence z (n) after upgrading in described step B3 and described step B1, calculated gains matrix K (n);
Step B5: utilize described gain matrix K (n), described output sequence z (n), again upgrade covariance matrix P (n), and utilize P (n) and P (n) after again upgrading taverage replacement upgrade P (n+1);
Step B6: according to described coefficient vector w (n), described list entries x (n) and described output sequence z (n), calculate residual error e (n);
Step B7: utilize described gain matrix K (n) and described residual error e (n) to upgrade coefficient vector, obtain w (n+1);
Step B8: judge whether to meet power amplifier linearization requirement, if meet, renewal process finishes; If do not meet, put n=n+1, return to step B2.
2. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, coefficient vector w (n)=[w in described step B1 10(n) ..., w 1Q(n) ..., w k0(n) ..., w kQ(n)] t
Wherein, in coefficient vector w (n), comprise the individual estimated parameter that needs of K (Q+1), the individual coefficient of the polynomial K of corresponding predistortion (Q+1);
And described covariance matrix P (n) is the symmetry square matrix of a K (Q+1) × K (Q+1).
3. amplifying digital baseband pre-distortion method according to claim 1, it is characterized in that, list entries x (n)=[x (n) in described step B2, x (n-Q) ..., x (n) | x (n) |,, x (n-Q) | x (n-Q) | k-1] t;
Described output sequence z (n)=[z (n) ..., z (n-Q) ..., z (n) | z (n) | ..., z (n-Q) | z (n-Q) | k-1] t.
4. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, in described step B3, covariance matrix P (n) only carries out diagonal entry renewal, and concrete steps are as follows:
The diagonal entry of setting P (n) is P ii(n), wherein, 1≤i≤K (Q+1);
The individual order of elements of K (Q+1) is divided into K group, and every group comprises Q+1 element; For the element P in L group ii(n), more new formula is as follows: P ii(n)=P ii(n)+σ l, wherein, 1≤L≤K.
5. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, the gain matrix in described step B4 K ( n ) = P ( n ) z ( n ) z ( n ) T P ( n ) z ( n ) + r .
6. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, again upgrades covariance matrix P (n)=(I-K (n) z (n) in described step B5 t) P (n);
Average replacement upgraded P ( n + 1 ) = 1 2 ( P ( n ) + P ( n ) T ) .
7. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, residual error e (the n)=w (n) in described step B6 t(x (n)-z (n)), wherein, w (n) is current coefficient estimate vector.
8. amplifying digital baseband pre-distortion method according to claim 1, is characterized in that, w (n+1)=w (n)+K (n) e (n) in described step 7.
CN201010571253.6A 2010-12-02 2010-12-02 Adaptive amplifying digital baseband pre-distortion method Active CN102487367B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201010571253.6A CN102487367B (en) 2010-12-02 2010-12-02 Adaptive amplifying digital baseband pre-distortion method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201010571253.6A CN102487367B (en) 2010-12-02 2010-12-02 Adaptive amplifying digital baseband pre-distortion method

Publications (2)

Publication Number Publication Date
CN102487367A CN102487367A (en) 2012-06-06
CN102487367B true CN102487367B (en) 2014-09-10

Family

ID=46152822

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201010571253.6A Active CN102487367B (en) 2010-12-02 2010-12-02 Adaptive amplifying digital baseband pre-distortion method

Country Status (1)

Country Link
CN (1) CN102487367B (en)

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102970262B (en) * 2012-11-16 2015-07-01 华南理工大学 Method for improving digital pre-distortion stability
CN103499812A (en) * 2013-09-23 2014-01-08 中国科学院电子学研究所 Baseband signal predistortion method of broadband multi-channel coherent radar imaging system
CN104883330A (en) * 2014-02-27 2015-09-02 清华大学 Blind equalization method and blind equalization system
CN103888396A (en) * 2014-03-26 2014-06-25 西安电子科技大学 Predistortion method based on self-adaptive variable step size iteration
CN103929136B (en) * 2014-04-22 2017-01-11 西安易晓通讯科技有限公司 Broadband mobile communication LTE system transmitter power amplifier predistortion processing method
CN105450184B (en) * 2014-08-08 2019-01-11 瑞昱半导体股份有限公司 Digital predistortion circuit and method and digital pre-distortion drill circuit
CN104320095B (en) * 2014-10-23 2017-09-22 清华大学 The method and device of ADC sampling rates in a kind of reduction digital pre-distortion
CN104363191B (en) * 2014-11-03 2019-03-29 西安烽火电子科技有限责任公司 A kind of digital pre-distortion method of frequency-hopping communication system
WO2016078038A1 (en) * 2014-11-19 2016-05-26 华为技术有限公司 Pre-distortion processing device and method
CN107132511B (en) * 2017-05-11 2020-01-21 中国科学院电子学研究所 Accurate radar linear frequency modulation source predistortion method
CN109039969B (en) * 2017-06-09 2020-11-27 中国工程物理研究院电子工程研究所 Method for implementing broadband digital predistorter
CN108063739A (en) * 2017-12-15 2018-05-22 北京卫星信息工程研究所 Broadband digital communication system transmitting terminal power amplifier adaptive digital pre-distortion method
CN109067369B (en) * 2018-07-06 2022-11-01 京信网络系统股份有限公司 Predistortion optimization method, device and system
CN111245377B (en) * 2020-03-20 2021-11-05 厦门大学 Digital predistortion system and method based on direct learning structure
CN113468814B (en) * 2021-07-09 2024-02-27 成都德芯数字科技股份有限公司 Neural network-based digital predistortion training data screening method and device
US11770145B2 (en) 2021-08-27 2023-09-26 Samsung Electronics Co., Ltd System and method for efficient initialization of memory digital pre-distorter coefficients to reduce calibration time

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
DISTORTION COMPENSATION OF NONLINEAR SYSTEMS BASED ON INDIRECT LEARNING ARCHITECTURE;Emad Abd-Elrady, Li Gan and Gernot Kubin;《ISCCSP》;20080314;正文第2页第2节、第4节及图1 *
Emad Abd-Elrady, Li Gan and Gernot Kubin.DISTORTION COMPENSATION OF NONLINEAR SYSTEMS BASED ON INDIRECT LEARNING ARCHITECTURE.《ISCCSP》.2008,正文第2页第2节、第4节及图1.

Also Published As

Publication number Publication date
CN102487367A (en) 2012-06-06

Similar Documents

Publication Publication Date Title
CN102487367B (en) Adaptive amplifying digital baseband pre-distortion method
CN101175061B (en) Self-adapting digital predistortion method and apparatus for OFDM transmitter
CN103873407B (en) Digital pre-distortion system for having the radiofrequency launcher of the sample rate reduced in observation ring
WO2015096735A1 (en) Digital pre-distortion parameter obtaining method and pre-distortion system
CN102652397A (en) Method and apparatus for digital pre-distortion of non-linear systems with reduced bandwidth feedback
US20140118066A1 (en) Low-cost digital predistortion apparatus and method using envelope detection feedback
CN110765720B (en) Power amplifier predistortion method of complex-valued pipeline recurrent neural network model
CN111245375B (en) Power amplifier digital predistortion method of complex-valued fully-connected recurrent neural network model
CN104363191A (en) Digital pre-distortion method of frequency hopping communication system
CN105356854A (en) Method for applying combined algorithm to combined learning structure digital predistortion system
KR101386239B1 (en) Predistorter for compensating of nonlinear distortion and method for the same
CN103929136B (en) Broadband mobile communication LTE system transmitter power amplifier predistortion processing method
CN103888396A (en) Predistortion method based on self-adaptive variable step size iteration
Tripathi et al. Swish activation based deep neural network predistorter for RF-PA
CN101350597A (en) Method for modeling wideband radio-frequency power amplifier
CN111988254B (en) Low-complexity peak-to-average ratio compression and predistortion joint optimization method
CN201409180Y (en) Self-adapting baseband linearization device of digital television transmitter
CN113221308A (en) Transfer learning rapid low-complexity modeling method facing power amplifier
CN102611661A (en) Predistortion device and method based on precise inverse solution memory polynomial model equation
CN102983820A (en) Nonlinear injection type linearization system and digital predistortion method
Nair et al. Impact of total harmonic distortion in SWIPT enabled wireless communication networks
CN108063739A (en) Broadband digital communication system transmitting terminal power amplifier adaptive digital pre-distortion method
US8633769B2 (en) Dual loop adaptation digital predistortion architecture for power amplifiers
Zhu et al. A blind AM/PM estimation method for power amplifier linearization
Huang et al. RLS-DPD algorithm for hybrid precoding architecture in MIMO-OFDM systems

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20220823

Address after: Room 108, floor 1, building 4, No. 2 dacuodeng Hutong, Dongcheng District, Beijing 100010

Patentee after: Beijing Zhongke micro Investment Management Co.,Ltd.

Address before: 100029 Beijing city Chaoyang District Beitucheng West Road No. 3 Institute of Microelectronics

Patentee before: Institute of Microelectronics, Chinese Academy of Sciences

TR01 Transfer of patent right