CN105958952A - Single-bit DPD (Digital Pre-Distortion) method for power amplifier - Google Patents

Abstract

The invention relates to a single-bit DPD (Digital Pre-Distortion) method for a power amplifier. The single-bit DPD method comprises the steps of: carrying out time delay estimation by a frequency domain method, and carrying out time delay compensation in a digital domain; taking difference values of input and output signals; replacing a conventional feedback branch circuit ADC (Analog-to-Digital Converter) with a 1bit ADC or a comparator, sequentially carrying out sampling on the difference values of the input and output signals, i.e., x still represents an input signal sequence, and carrying out difference value taking processing on output y signals; estimating a step length required in iteration; and carrying out iteration by employing an iteration format suitable for 1bit after carrying out the processing to complete parameter extraction of DPD. According to the method, a 1bit technology developed in the field of digital signal processing in recent years is used for a DPD technology of the power amplifier, the single-bit DPD method is formed, power consumption and cost of a feedback loop can be greatly reduced, a bandwidth range of the DPD technical application also can be greatly improved and is expected to reach hundreds of MHz, and the broadband DPD technical requirements of 5G (5th-generation) communication are possibly solved.

Description

A kind of monobit digital pre-distortion method of power amplifier

Technical field

The present invention relates to the monobit digital pre-distortion method of a kind of power amplifier, belong to electronic information and communication technology neck Territory.

Background technology

In a wireless communication system, power amplifier (Power Amplifier, PA) is main nonlinear device.Merit Put non-linear meeting serious near saturation point and cause the inband distortion of signal, also there will be spectral leakage simultaneously thus produce neighbour Road disturbs, and on the contrary, allows the online property district of high peak-to-average power ratio signal work then can reduce power efficiency.Therefore, the linearisation of power amplifier Technology becomes study hotspot in recent years.Wherein, digital pre-distortion (Digital PreDistortion, DPD) becomes with it The advantages such as this low, flexible in programming become current widely used a kind of technology [1,2].DPD technology is in power amplifier prime Join the predistorter (PreDistorter, PD) of a nonlinear characteristic contrary therewith, so that cascade system linearly [2], It is illustrated in figure 1 a typical conventional DPD block diagram.

DPD technology is because being operated in numeric field, and analog-digital converter (Analog-to-Digital Converter, ADC) is required Device.Non-linear owing to existing, frequency spectrum will broadening after power amplifier for signal.Enter to analyze its nonlinear characteristic One step tries to achieve the characteristic of predistorter, if it is considered that rate 5 rank crosstalk, at least needs to analyze 5 times of input signal bandwidth Frequency spectrum.In other words, people to sample to be not less than 5 times of input signal nyquist sampling rates.Along with the 3rd Generation, the 4th generation wireless communication technology universal, input signal bandwidth becomes more and more wider, reaches 100MHz even broader, This just proposes high requirement to ADC, and cost also can be significantly increased [3].Therefore, reduce ADC rate requirement and Power consumption is the problem that current DPD research is badly in need of solving.

(1) digital pre-distortion is the research emphasis of radio-frequency front-end

Digital pre-distortion is the research emphasis of radio-frequency front-end at present.Digital pre-distortion is suggested in eighties of last century the nineties, And this technology starts to obtain the attention of people and having had and is widely applied after entering basic annals.Existing numerous researchs in the world Mechanism is working on the research work being correlated with, including Canada's iRadio laboratory, Dublin, Ireland university, Yi Jide State, many laboratorys of Japan and other countries.And domestic had Tsing-Hua University, Chengdu University of Electronic Science and Technology, Southeast China University, Numerous seminar such as University Of Ningbo, Beijing University of Post & Telecommunication, Chinese University of Science and Technology have carried out the research work that DPD is relevant.

The study hotspot in DPD field focuses primarily upon the selection of DPD model, improves and optimizates and reduce complexity at present [4-9], relatively conventional model include memory polynomial (MP) model, Hammerstein model, Wiener model, Volterra model, LUT model and some enhancing models etc. on this basis.In recent years, DPD application scenarios is also Gradually be transitioned into double frequency even multifrequency [10-12] from single-frequency, therefore for double frequency, multifrequency DPD scale-model investigation the most gradually Becoming hot topic, process bandwidth is also had higher requirement by this.

Along with developing rapidly of current Digital Signal Processing software and hardware, and next generation wireless communication technology is popularized, and signal passes What defeated speed, the availability of frequency spectrum and power amplifier linearization required improves constantly, and digital pre-distortion technology will be made to continue to become nothing One of primary study direction of line radio-frequency front-end.

(2) lack sampling DPD research gradually heats up

Reducing systematic sampling rate is one of following main direction of studying in DPD field, has had research both at home and abroad Person begins to focus on this field, and proposes some feasible solutions.

One of which method is by direct lack sampling, and do so is very simple and convenient, application effect in some arrowband scenes The best, document [3,13] has turned out the feasibility that direct lack sampling method is applied in digital pre-distortion, but to lose band External information is cost, and predistortion effect in the scene of broadband is the most bad.Nonetheless, document [14] demonstrates use directly Connect the digital pre-distortion method of lack sampling to the frequency content suppression near master tape substantially, and successfully applied in wideband field Jing Zhong, cost is to add one-level band filter after PA.Document [15] proposes a kind of band limit DPD, is only suitable for low Speed ADC gathers signal in-band information, and band external information is obtained by frequency spectrum Extrapolation method.Corresponding is a kind of mark Amount predistortion (scalar DPD) method [16], the method only need to gather the band external information of signal, and minimize it thus Extracting DPD parameter, its shortcoming is that convergence rate is slow cannot tackle memory effect simultaneously.Another kind of solution is parallel Lack sampling, it can reduce single-channel sampling speed, in turn ensure that band external information will not be lost, and has given up increase extra simultaneously The trouble brought of band filter.Applicant has been proposed that a kind of parallel lack sampling DPD knot in document [17] Structure, and its feasibility of preliminary identification, as shown in Figure 2.

Along with the development of next generation wireless communication technology is with universal, two-forty broadband wireless communications signal is to lack sampling DPD Demand more urgent.The present invention is aiming at what this demand proposed.

(3) 1bit technology is in the application of radar imagery

In radar imaging system, owing to echo data amount is the biggest, general the highest (every to the quantified precision of echo General 4 to 8 bit quantizations of individual sample) [18].Synthetic aperture radar image-forming based on single-bit quantification is in last century 90 Age is reviewed and analysed [18-22] by G.Schirinzi et al..Single-bit quantification encodes [18,19,22] (one also known as single-bit Bit coded), symbolic coding [21] (signum coded) or phase quantization [20] (phase quantization), it is adopted Sample loading mode is the I to echo-signal, and Q two-way carries out single-bit encoding operation respectively, the most as shown in Figure 3.

In single-bit synthetic aperture radar image-forming system, traditional ADC is replaced by comparator, after down coversion Signal, through single-bit measure, the measurement data obtained is { the form of ± 1 ± j}.Then, originally value is real number value I, Q two paths of signals be mapped to two discrete values, the most also referred to as symbolic codings.And single-bit quantification be equivalent to right Phase place has carried out the quantization of four quadrants, therefore also becomes four points of phase quantizations (four level phase quantization). After single-bit quantification, the waveform of signal receives the biggest impact, the frequency spectrum of signal also corresponding broadening.Pass through Substantial amounts of the researching and analysing of G.Schirinzi et al. is pointed out, under conditions of low signal-to-noise ratio and the sampling of Nyquist excessively, and digital ratio Special quantization is less on the impact of signal spectrum, therefore the method that tradition matched filtering can be used, come by space-time two-dimension convolution Scene is carried out imaging.

For sparse scene, it is possible to use the method for sparse optimization, single-bit compressed sensing and single-bit synthetic aperture Radar imagery combines, and breaks through the restriction of tradition single-bit matched filtering method and improves image quality.Based on compression sense The research of the sparse imaging of single-bit synthetic aperture radar known is still at an early stage.

The technology its development relevant based on above three and trend, it can be seen that for tackling the ever-increasing demand of bandwidth, On the one hand it is can to use suitable down-sampled scheme under conditions of keeping sampling resolution constant so that overall bandwidth requirement Reducing, the total bit number being equivalent in the unit interval reduces, such that it is able to complete broadband DPD.Another brand-new think of Lu Shi, the present invention proposes the monobit digital pre-distortion method of power amplifier, 1bit thought is directly incorporated into the DPD of power amplifier In, use figure place to be only the ADC technology of 1 bit (alternatively referred to as single-bit), it is achieved the reduction of total bit number (though So may need sampling rate faster, but total bit number is the most less), and the speed of single-bit ADC is exceedingly fast, Even can substitute ADC with comparator, speed is faster.Therefore it is expected to break through current predicament, overall lifting DPD Technology bandwidth level applicatory.

Monobit digital pre-distortion method, for power amplifier digital pre-distortion technology, is a brand-new method.Pass to be solved Key problem be how just with 1bit to reflect the amplitude information of signal.There is no pertinent literature report at present.

The present invention proposes the thinking of key: although due to input/output information because power amplifier non-linear has distortion, but There is preferable dependency, it is proposed that after taking the difference between the two, difference is carried out 1bit sampling.

List of references

[1]D.R.Morgan,Z.Ma,J.Kim,M.G.Zierdt,J.Pastalan.A Generalized Memory Polynomial Model for Digital Predistortion of RF Power Amplifiers[J].IEEE Transactions on Signal Processing,2006,54(10):3852-3860.

[2]F.M.Ghannouchi,O.Hammi.Behavioral Modeling and Predistortion[J].IEEE Microwave Magazine,2009,10(12):52-64.

[3] Chen Changwei, Qin Kaiyu. lack sampling is in the application [J] of base band predistortion power amplifier linearization. telecommunication skill Art, 2010,50 (5): 47-50.

[4]J.C.Pedro and S.A.Maas,A Comparative Overview of Microwave and Wireless Power-Amplifier Behavioral Modeling Approaches[J],IEEE Transactions on Microwave Theory and Techniques,vol.53,pp.1150-1163,April 2005.

[5]Z.Anding,J.C.Pedro,and T.J.Brazil,Dynamic Deviation Reduction-Based Volterra Behavioral Modeling of RF Power Amplifiers[J],IEEE Transactions on Microwave Theory and Techniques,vol.54,pp.4323-4332,2006.

[6]T.Liu,S.Boumaiza,and F.M.Ghannouchi,Augmented hammerstein predistorter for linearization of broad-band wireless transmitters[J],IEEE Transactions on Microwave Theory and Techniques,vol.54,pp.1340-1349,June 2006.

[7]L.Ding,G.T.Zhou,D.R.Morgan,Z.Ma,J.S.Kenney,J.Kim,et al.,A robust digital baseband predistorter constructed using memory polynomials[J],IEEE Transactions on Communications,vol.52,pp.159-165,Jan 2004.

[8]Y.J.Liu,J.Zhou,W.Chen,and B.H.Zhou,A Robust Augmented Complexity-Reduced Generalized Memory Polynomial for Wideband RF Power Amplifiers[J],IEEE Transactions on Industrial Electronics,vol.61,pp.2389-2401, 2014.

[9]L.Guan and A.Zhu,Optimized Low-Complexity Implementation of Least Squares Based Model Extraction for Digital Predistortion of RF Power Amplifiers[J], IEEE Transactions on Microwave Theory and Techniques,vol.60,pp.594-603,March 2012.

[10]F.M.Ghannouchi,M.Younes,and M.Rawat,Distortion and impairments mitigation and compensation of single-and multi-band wireless transmitters(invited) [J],IET Microwaves,Antennas&Propagation,vol.7,pp.518-534,2013.

[11]L.Ding,Z.Yang,and H.Gandhi,Concurrent Dual-band Digital Predistortion [C],2012 IEEE MTT-S International Microwave Symposium Digest,Montreal,QC, Canada,2012.

[12]G.Yang,F.Liu,L.Li,H.Wang,C.Zhao,and Z.Wang,2D Orthogonal Polynomials for Concurrent Dual-Band Digital Predistortion[C],2013 International Microwave Symposium,Seattle,WA,2013.

[13]H.Koeppl and P.Singerl,An efficient scheme for nonlinear modeling and predistortion in mixed-signal systems[J].IEEE Trans.Circuits Syst.II,vol.53,no.12, pp.1368–1372,Dec.2006.

[14]C.Yu,L.Guan,E.Zhu and A.Zhu.Band-Limited Volterra Series-Based Digital Predistortion for Wideband RF Power Amplifiers[J].IEEE Transactions on Microwave Theory and Techniques,2012,60(12):4198-4208.

[15]Y.Ma,Y.Yamao,Y.Akaiwa,and K.Ishibashi,Wideband Digital Predistortion Using Spectral Extrapolation of Band-Limited Feedback Signal[J].IEEE Transactions on Circuits and Systems I:Regular Papers,vol.PP,pp.1-10,2014.

[16]I.Takanori,J.Zhou,H.Li,Z.Shi,P.Chen,B.Zhao,and K.Kakinuma,A simple predistorter infixed microwave radio system with the transmission power control [C].IEEE Topical Conf.Power Amplifiers Wireless Radio Appl.,Jan.2012,vol.65,no. 68,pp.15–18.

[17] Liu Falin, Wang Haoyu, Yang Guang, Li Gang. lack sampling digital pre-distortion [J]. microwave journal, 2013Vol. 29(5-6):81-85

[18]G.Fornaro,V.Pascazio,G.Schirinzi,Synthetic aperture radar interferometry using one bit coded raw and reference signals[J],Ieee Transactions on Geoscience and Remote Sensing,1997,35:1245-1253.

[19]V.Pascazio,G.Schirinzi,Synthetic aperture radar imaging by one bit coded signals[J],Electronics&Communication Engineering Journal,1998,10:17-28.

[20]G.Franceschetti,S.Merolla,M.Tesauro,Phase quantized SAR signal processing:Theory and experiments[J],IEEE Transactions on Aerospace and Electronic Systems,1999,35:201-214.

[21]G.Franceschetti,V.Pascazio,G.Schirinzi,Processing of Signum Coded Sar Signal-Theory and Experiments[J],IEE Proceedings-F Radar and Signal Processing, 1991,138:192-198.

[22]G.Alberti,G.Franceschetti,V.Pascazio,et al,Time-Domain Convolution of One-Bit Coded Radar Signals[J],IEE Proceedings-F Radar and Signal Processing,1991, 138:438-444.

Summary of the invention

The technology of the present invention solves problem: overcoming the deficiencies in the prior art, the monobit digital proposing a kind of power amplifier is pre- Distortion methods, is used for digital processing field developed recently 1bit technology out in the digital pre-distortion technology of power amplifier, Form monobit digital pre-distortion method.On the one hand power consumption and the cost of feedback circuit can be greatly reduced, on the other hand can Increase substantially the applicable bandwidth range of pre-distortion technology, it is expected to reach hundreds of MHz, it is possible to solve 5G communication Broadband DPD technical need.This technology can also be used for the linearization technique of wideband radar transmitter.

The technology of the present invention solution is described as follows: the monobit digital pre-distortion method of a kind of power amplifier, it is achieved step Suddenly it is:

(1) difference of input/output signal is taken；

(2) conventional feedback branch ADC is replaced with ADC or the comparator of 1bit, according to this input and output is believed Number difference sample, i.e. x is still input signal sequence, has carried out taking difference for the y signal of output and has processed；

(3) after above-mentioned process, then use the Iteration being adapted to 1bit to be iterated, complete the whole of DPD Individual process.

The Iteration of the described 1bit of being adapted to is accomplished by

(31) difference for input/output signal carry out 1bit quantify sampling obtain:

$s=sign(\tilde{y}-\tilde{x})$

This formula represents that the difference for input/output signal carries out 1bit and quantifies sampling, and the s on the formula left side is that difference is taken symbol Vector after number, wherein sign () expression takes 1 bit quantization, and the data being i.e. sampled are fixed more than or equal to during certain thresholding It is 1, is set to 0 or-1 less than this thresholding, represents then have with+1 and-1:

S=[± 1 ... ± 1]^T

$\tilde{y}\≈\tilde{x}+c_{k}s$

Wherein, c_kFor iteration step length, effect is control disturbance amplitude；

(32) Iteration is as follows:

$\begin{array}{c}{\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-\mathrm{\μ}{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^T(\tilde{y}-\tilde{x})\\ \≈{\widetilde{\mathrm{\α}}}_k-c_k{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^{T}s\end{array}$

WhereinIt is the expression matrix form of the basic function obtained by list entries, typically using Volterra progression as base letter Number describes the non-linear of power amplifier and memory effect.α_kRepresent the predistorter parameter after kth time iteration

(33), after such formula iteration, the parameter extraction of monobit digital predistortion is just completed.

Use time-delay alignment algorithm to need to estimate two respectively when the described difference to input/output signal is sampled to return The time delay on road.Represent that former transmitting branch starts the analog-digital converter up to feedback circuit from digital to analog converter with path 1, Path 2 represents that the input signal branch road being transferred through time delay processing starts the modulus up to feedback circuit from digital to analog converter Transducer, particularly as follows: first gather the signal in path 1 with the ADC of 1bit, now path 2 ground connection, then this Signal carries out FFT to frequency domain, compared with the band inner portion of primary signal, estimates time delay t in path 1₁；Use again The ADC of 1bit gathers the signal in path 2, now path 1 ground connection, then this signal is carried out FFT to frequency domain, Compared with the band inner portion of primary signal, estimate time delay t in path 2₂；After the time delay of two paths all estimates, The t that the signal in path 2 is fixed at numeric field₁-t₂Time delay, thus ensure that when carrying out DPD parameter extraction, Input and output difference signal is mutually aligned.

Iteration step length employing below equation during described iteration:

$c_k=\frac{c_0}{k^{\mathrm{\γ}}}$

Wherein c₀Being initial step length, k is iterations, and γ is zoom factor.

Described initial step length c₀Can be estimated by the priori of power amplifier, i.e. the product manual of power amplifier or power amplifier is carried out list Sound is tested, as follows:

$c_0=\mathrm{\λ}\·rms\left(\frac{x}{\max \left|x\right|}\right)\·(\sqrt{\frac{P_{out}^{peak,sg}}{P_{out}^{peak}}}-1)$

WhereinIt is true peak output,Being the expectation maximum output under small-signal gain, λ is little In the constant of 1, x is then input signal.

Through substantial amounts of experiment test, according to different signal bandwidths and power amplifier, the span suggestion of γ is 1.1 to 1.8.

Present invention advantage compared with prior art is: the present invention uses the ADC of 1bit precision at feedback circuit, greatly Reduce greatly power consumption and cost.Because ADC is one of device the most expensive in DPD system, and ADC's is basic Composed component is comparator, and the number of comparator is with the exponential growth of sampling precision.When sampling precision is only 1bit, Power consumption and cost are obtained for and significantly reduce, and this is applied to mini-plant for DPD and provides convenience (current DPD Be mainly used in large-scale base station, mini-plant many with back-off replace DPD thus efficiency is the lowest)；The opposing party Face, can the most easily reach the sampling rate of several GSPS compared to high-precision ADC, the ADC of 1bit, this It is very easy to DPD and broadband system is carried out linearisation.It is expected to solve next generation communication system for broadband DPD's Needs of problems.Additionally, this method is compared with traditional DPD method of employing long number ADC, result has suitable essence Degree, solving speed is close.

Accompanying drawing explanation

Fig. 1 is typical case's DPD block diagram；

Fig. 2 is a kind of lack sampling DPD structured flowchart；

Fig. 3 is that single-bit complex signal measures block diagram；

Fig. 4 is the simplified block diagram that typical case directly learns DPD system；

The single-bit DPD theory diagram of Fig. 5 present invention

A kind of single-bit DPD implementation schematic diagram in Fig. 6 present invention；

Fig. 7 is power spectral density comparison diagram；

Fig. 8 is convergence rate comparison diagram.

Detailed description of the invention

The solution principle of the 1bit digital pre-distortion method that the present invention proposes is as it is shown in figure 5, process is: (1) takes input The difference of output signal；(2) conventional feedback branch ADC is replaced with ADC or the comparator of 1bit, according to this The difference of input/output signal is sampled.I.e. x is still input signal sequence, and the y signal for output is taken Difference processes.After such process, then use the new Iteration being adapted to 1bit to be iterated, complete DPD Whole process.

Therefore, the main points of the inventive method are also embodied in and give the DPD that new iterative equation replacement is traditional theoretically Iterative equation.Time-delay alignment algorithm therein and step-length selection strategy have been steps necessary to iteration, fully demonstrate Feature under 1bit treatment conditions.This part of theory content is discussed in detail below:

(1) theoretical derivation of tradition DPD

The theory diagram of tradition DPD as it is shown in figure 1, at its feedback branch, due to will to output signal (rather than this The difference of the signal that invention is used in single-bit method) directly sample, the ADC of long number typically should be utilized to force The dynamic range of this signal nearly, dynamic range is the biggest, and the figure place of the ADC of needs is the most.

Fig. 1 is simplified, the conventional feedback branch schematic diagram directly learning structure as shown in Figure 4 can be obtained.Wherein, X is input signal sequence, and y is the feedback signal sequence after conventional ADC sampling.

Typically take optimization object function as follows, it is therefore an objective to make the error between input and output minimum, non-to eliminate power amplifier The impact of linear effect.

Optimization object function is:

The gradient of note object function and extra large gloomy matrix:

Then Iteration is as follows:

α_k+1=α_k-μH^-1G=α_k-μ(X^HX)^-1X^H(y-x)

Wherein μ is iteration step length, and X is the expression matrix form of the basic function obtained by list entries, typically with Volterra progression describes the non-linear of power amplifier and memory effect as basic function.α_kRepresent the pre-mistake after kth time iteration True device parameter.This is plural form.If note:

$\widetilde{\mathrm{\α}}=\left[\begin{array}{c}\mathrm{Re}\left(\mathrm{\α}\right)\\ \mathrm{Im}\left(\mathrm{\α}\right)\end{array}\right],\tilde{x}=\left[\begin{array}{c}\mathrm{Re}\left(x\right)\\ \mathrm{Im}\left(x\right)\end{array}\right],\tilde{y}=\left[\begin{array}{c}\mathrm{Re}\left(y\right)\\ \mathrm{Im}\left(y\right)\end{array}\right],\tilde{X}=\left[\begin{array}{cc}\mathrm{Re}\left(X\right)& -\mathrm{Im}\left(X\right)\\ \mathrm{Im}\left(X\right)& \mathrm{Re}\left(X\right)\end{array}\right]$

Then can turn to following real number iteration form:

${\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-\mathrm{\μ}{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^T(\tilde{y}-\tilde{x})$

After so iteration completes, just complete digital pre-distortion.

(2) theoretical derivation of the inventive method

The central principle block diagram of the single-bit DPD method that the present invention proposes is as shown in Figure 5.For from hardware system, At traditional band limit feedback channel, having carried out two crucial variations, one is the difference taking input/output signal, reduces The amplitude dynamic range of signal to be sampled；Two is the ADC replacing long number with single-bit ADC.

For realizing single-bit DPD algorithm, it is necessary to study its theoretical frame and method for solving.It is given below and changes with this hardware The dynamic theoretical derivation wanting to adapt to, the two combines the core innovation constituting the application.

As a example by directly study structure (can also be used for other structure such as indirect learning structures), the 1bit number that the present invention proposes The schematic diagram of a kind of implementation of word pre-distortion method as shown in Figure 6, is characterized by: (1) takes input/output signal Difference；(2) conventional feedback branch ADC is replaced with ADC or the comparator of 1bit, 1bit ADC pair according to this The difference of input/output signal is sampled.I.e. x is still input signal sequence, and the y signal for output has carried out taking difference Value processes.

In the present invention, optimization object function is still so that the error between input and output is minimum, and this is power amplifier DPD Common demands.But Iteration there occurs important change.Note:

$s=sign(\tilde{y}-\tilde{x})$

This formula represents that the difference for input/output signal carries out 1bit and quantifies sampling, and the s on the formula left side is that difference is taken symbol Vector after number, wherein, sign () expression takes 1 bit quantization, when the data being i.e. sampled are more than or equal to certain thresholding It is set to 1, is set to 0 or-1 less than this thresholding.Typically represent then have with+1 and-1:

S=[± 1 ... ± 1]^T

$\tilde{y}\≈\tilde{x}+c_{k}s$

Wherein, c_kFor iteration step length, effect is control disturbance amplitude, and s is interpreted as random disturbance vector.

Then can derive Iteration as follows:

$\begin{array}{c}{\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-\mathrm{\μ}{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^T(\tilde{y}-\tilde{x})\\ \≈{\widetilde{\mathrm{\α}}}_k-c_k{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^{T}s\end{array}$

Wherein μ has been merged into c_kIn, the latter is as total iteration step length.With traditional method same meaning.So far, Theoretical derivation completes.After so iteration terminates, just complete monobit digital predistortion.

More than derive and can also understand from the process minimizing a norm.If object function becomes:

$\underset{\widetilde{\mathrm{\α}}}{\mathrm{argmin}}f\left(\widetilde{\mathrm{\α}}\right)={||\tilde{y}-\tilde{x}||}_1$

Can try to achieve its subgradient is:

Then the iterative formula of single order Newton iteration method is:

${\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-c_k{\tilde{X}}^{T}s$

In order to accelerate convergence of algorithm speed, addThis, then above formula becomes:

${\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-c_k{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^{T}s$

Visible, this iterative formula derived with the concept of random disturbance used above is just the same.This algorithm relates to Time-delay alignment is summarized as follows with step-size estimation the two important algorithm:

(3) the time-delay alignment algorithm of the inventive method

1bit is only only had, it is impossible to use common cross correlation algorithm to estimate experiment due to feedback circuit precision.Use frequency domain Algorithm carrys out estimation time delay, because the band inner portion of signal is quantified to be affected not quite by 1bit, the most just says that in-band information retains phase To completely, this contributes to using in-band information to carry out time delay estimation.Need to estimate respectively the time delay in two paths, as Shown in Fig. 6, " path 1 " represents that former transmitting branch starts the analog-digital converter up to feedback circuit, " road from digital to analog converter Footpath 2 " represent that the input signal branch road being transferred through time delay processing starts the modulus up to feedback circuit from digital to analog converter Transducer.First gather the signal (now path 2 ground connection) in path 1 with the ADC of 1bit, then this signal is entered Row FFT is to frequency domain, compared with the band inner portion of primary signal, estimates time delay t in path 1₁.Path 2 time Prolong t₂Estimate similar.After the time delay of two paths all estimates, the t that the signal in path 2 is fixed at numeric field₁- t₂Time delay, thus ensure that, when carrying out DPD parameter extraction, input and output difference signal is mutually aligned.

(4) step-length of the inventive method selects

Iteration step length needs strictly to control to reach Fast Convergent and good precision.If step-length is too big, may The convergence error causing algorithm excessive is not even restrained；If step-length is the least, then restrain the slowest, it is impossible to meet a lot of system The demand of real-time.Propose to use equation below to determine iteration step length:

$c_k=\frac{c_0}{k^{\mathrm{\γ}}}$

Wherein c₀Being initial step length, k is iterations, and γ is zoom factor.c₀(merit can be estimated by the priori of power amplifier The product manual put or power amplifier is carried out singletone test)

$c_0=\mathrm{\λ}\·rms\left(\frac{x}{\max \left|x\right|}\right)\·(\sqrt{\frac{P_{out}^{peak,sg}}{P_{out}^{peak}}}-1)$

WhereinIt is true peak output,It it is the expectation maximum output under small-signal gain.λ is little In the constant of 1.X is then input signal.Institute is through substantial amounts of experiment test, according to different signal bandwidths and power amplifier, γ Span suggestion be 1.1 to 1.8.

Blue arrow as shown in Figure 5 is constituted with the circular subtractor represented, and its effect is to carry out for input/output signal Taking difference to process, owing to the two has preferable relatedness, therefore the amplitude of difference has obtained effective reduction.This is to be transported Key with 1bit ADC.(2) 1bit ADC, as shown in Figure 5.This module carries out 1bit for above-mentioned difference ADC samples, it is also possible to realize this sampling process with comparator.(3) new theoretical method for solving.Including (a) Iterative core methed, (b) time-delay alignment algorithm and (c) step-length selection strategy etc..Above-mentioned hard owing to have employed Part tupe, traditional DPD Solution model is the most directly suitable for.The present invention proposes one and is applicable to 1bit process DPD method for solving, its detailed process has been described above in above-mentioned theory part.

Matlab emulation experiment is utilized to carry out preliminary identification.Condition is as follows:

Power amplifier: Wiener Model (Wiener model)

DPD: memory polynomial (memory depth P=5, non-linear exponent number M=4)

Step-length:

Input signal: LTE 20MHz bandwidth, peak-to-average force ratio PAPR 6.6dB

One group towards the result after the 1bit DPD process of 20MHz bandwidth as shown in Figure 7, Figure 8.From in figure, Its digital pre-distortion effect (with 1bit DPD) and DPD technology (the with DL through the directly routine of study structure DPD) effect is suitable, and convergence rate is close with conventional DPD technology.Demonstrate the effectiveness of institute of the present invention extracting method.

After another set monobit digital pre-distortion, AM-AM (AM-AM) noise characteristic (is adjusted with AM-PM Width-phase modulation) noise characteristic, pass through and existing DPD Technical comparing, also obtain the most same performance.

There is provided above example to be only used to describe the purpose of the present invention, and be not intended to limit the scope of the present invention.This Bright scope is defined by the following claims.Various equivalents made without departing from spirit and principles of the present invention and repairing Change, all should contain within the scope of the present invention.

Claims (6)

1. the monobit digital pre-distortion method of a power amplifier, it is characterised in that realize step as follows:

(1) carry out time delay estimation by frequency domain method, and carry out delay compensation at numeric field；

(2) difference of input/output signal is taken；

(3) conventional feedback branch ADC is replaced with ADC or the comparator of 1bit, according to this input and output is believed Number difference sample, i.e. x is still input signal sequence, has carried out taking difference for the y signal of output and has processed；

(4) step-length required in iteration estimated by technical manual or priori by PA；

(5) after above-mentioned process, then use the Iteration being adapted to 1bit to be iterated, complete the ginseng of DPD Number extracts.

The monobit digital pre-distortion method of power amplifier the most according to claim 1, it is characterised in that: suitable Should be accomplished by the Iteration of 1bit

(1) difference for input/output signal carry out 1bit quantify sampling obtain:

$s=sign(\tilde{y}-\tilde{x})$

This formula represents that the difference for input/output signal carries out 1bit and quantifies sampling, and the s on the formula left side is that difference is taken symbol Vector after number, wherein sign () expression takes 1 bit quantization, and the data being i.e. sampled are fixed more than or equal to during certain thresholding It is 1, is set to 0 or-1 less than this thresholding, represents then have with+1 and-1:

S=[± 1 ... ± 1]^T

$\tilde{y}\≈\tilde{x}+c_{k}s$

Wherein, c_kFor iteration step length, effect is control disturbance amplitude；

(2) Iteration is as follows:

$\begin{array}{c}{\widetilde{\mathrm{\α}}}_{k+1}={\widetilde{\mathrm{\α}}}_k-\mathrm{\μ}{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^T\left(\tilde{y}-\tilde{x}\right)\\ \≈{\widetilde{\mathrm{\α}}}_k-c_k{\left({\tilde{X}}^T\tilde{X}\right)}^{-1}{\tilde{X}}^{T}s\end{array}$

WhereinIt is the expression matrix form of the basic function obtained by list entries, typically using Volterra progression as base letter Number describes the non-linear of power amplifier and memory effect,Represent the predistorter parameter after kth time iteration；

(3), after such formula iteration, the parameter extraction of monobit digital predistortion is just completed.

The monobit digital pre-distortion method of power amplifier the most according to claim 1, it is characterised in that: institute State when the difference to input/output signal is sampled use time-delay alignment algorithm be: estimate respectively two paths time Prolong, represent that former transmitting branch starts the analog-digital converter up to feedback circuit, path 2 from digital to analog converter with path 1 Represent that the input signal branch road being transferred through time delay processing starts the analog digital conversion up to feedback circuit from digital to analog converter Device, particularly as follows: first gather the signal in path 1 with the ADC of 1bit, now path 2 ground connection, then this signal Carry out FFT to frequency domain, compared with the band inner portion of primary signal, estimate time delay t in path 1₁；Use 1bit again ADC gather the signal in path 2, now path 1 ground connection, then this signal is carried out FFT to frequency domain, with The band inner portion of primary signal is compared, and estimates time delay t in path 2₂；After the time delay of two paths all estimates, The t that the signal in path 2 is fixed at numeric field₁-t₂Time delay, thus ensure that when carrying out DPD parameter extraction, Input and output difference signal is mutually aligned.

The monobit digital pre-distortion method of power amplifier the most according to claim 1, it is characterised in that: institute Iteration step length when stating iteration uses below equation:

$c_k=\frac{c_0}{k^{\mathrm{\γ}}}$

Wherein c₀Being initial step length, k is iterations, and γ is zoom factor.

The monobit digital pre-distortion method of power amplifier the most according to claim 4, it is characterised in that: institute State initial step length c₀Estimated by the priori of power amplifier, i.e. the product manual of power amplifier or power amplifier is carried out singletone test, As follows:

$c_0=\mathrm{\λ}\·rms\left(\frac{x}{max\left|x\right|}\right)\·(\sqrt{\frac{P_{out}^{peak,sg}}{P_{out}^{peak}}}-1)$

WhereinIt is true peak output,Being the expectation maximum output under small-signal gain, λ is little In the constant of 1, x is then input signal.

The monobit digital pre-distortion method of power amplifier the most according to claim 4, it is characterised in that: institute Through substantial amounts of experiment test, according to different signal bandwidths and power amplifier, the span suggestion of γ is 1.1 to 1.8.