CN101212206A - Method for handling nonlinear pre-distortion of amplifier - Google Patents

Abstract

The invention discloses a method for treating non-linear predistortion of an amplifier, which includes the following processes: according to an amplitude An of a current signal and an amplitude Ahistory of a historic signal sample, an index value is generated; according to the index value, a predistortion factor corresponding to the index value is researched in a look-up table; the predistortion factor multiplies the current signal, and after digital analog conversion and frequency conversion treatment, the signal after predistortion treatment is output to an input end of the amplifier. During the process of predistortion treatment, output signals of the amplifier and feedback signals of a baseband can be obtained. A new predistortion factor is obtained by measuring the feedback signals of the baseband and the delayed current signal with the adaptive adjust algorithm, and then the new predistortion factor is written into the look-up table according to the index value for replacing the former predistortion factor. The invention adopts the signal amplitude to measure the index value of the look-up table and is easy in measurement. Moreover, better system performance can be achieved, and the complexity for realizing the system is reduced.

Description

A kind of method of amplifier nonlinearity pre-distortion

Technical field

The present invention relates to communication technical field, relate in particular to the method for a kind of amplifier nonlinearity pre-distortion in the communication equipment.

Background technology

Amplifier is a kind of electronic device commonly used in the electronic equipment.Any amplifier all exists the phenomenon of linear and nonlinear distortion, especially high power amplifier (HPA, High Power Amplifier) then distortion phenomenon is more obvious, and, As time goes on, owing to factors such as wearing out of device, feature linear and that nonlinear distortion shows also will change.

Linear distortion (being also referred to as frequency distortion) is caused the response of unlike signal frequency is different by the linear reactance element in the amplifier circuit; Linear distortion only can make amplitude proportional relation and the time delay relation of each frequency component signal change, and has perhaps filtered the signal of some frequency component, and can't increase the signal of new frequency component.

And nonlinear distortion is to be caused by the non-linear element in the amplifier circuit; Nonlinear distortion can become non-sinusoidal waveform with sine wave, and it not only comprises the frequency content (first-harmonic) of input signal, but also has produced many new harmonic componentss.

Therefore, equally all make the output signal of amplifier produce distortion although we can say linear distortion and nonlinear distortion, two kinds of distortion phenomenons substantially are diverse.Wherein, nonlinear distortion is compared the more difficult elimination of linear distortion because of its regular difference.

Multicarrier system such as OFDM (OFDM, Orthogonal Frequency DivisionMultiplexing) system, owing to have higher peak-to-average power ratio value (PAPR, Peak-to-AveragePower Ratio), compare with single-carrier system, the nonlinear distortion of pair amplifier is more responsive.The amplifier nonlinearity distorted characteristic also has " memory " feature, and promptly the current output signal of amplifier is not only relevant with the current input signal of amplifier, and is also relevant with the input signal in its past.Therefore, in order to obtain higher amplifier operating efficiency, must try every possible means to eliminate the negative effect of nonlinear distortion.

A kind of method of elimination nonlinear distortion negative effect commonly used was not carried out anti-distortion processing to signal exactly in advance before signal enters amplifier as yet, to send into amplifier through the signal after the anti-distortion processing again, distortion through amplifier is handled, the effect that anti-distortion processing and distortion are handled will be cancelled out each other, thereby the output signal that can guarantee amplifier is not subjected to the influence of amplifier nonlinearity distorted characteristic, but shows linear characteristic.This method is commonly referred to as the non-linear predistortion technology.

A kind of common implementation method of non-linear predistortion technology is to adopt look-up table (LUT, Look-UpTable) method, in advance the predistortion factor being stored in LUT is in the look-up table, power according to input signal carries out index to LUT, search out the suitable predistortion factor, again input signal is carried out pre-distortion.

In above-mentioned non-linear predistortion technology, because the implementation complexity of non-linear predistortion technology and the size of LUT (look-up table) and the highway width of signal variable have direct relation, in the method for traditional generation X and Y index, adopted signal power as the tolerance means.Compare with signal amplitude for example, signal power is the square value of signal amplitude.Suppose to adopt 16 bits to come the amplitude of complete representation signal, have only so and adopt the power that 32 bits can the complete representation signal.That is to say that adopt signal power many greatly as the metric signal bus width that metric needs more than the employing signal amplitude as metric, this has directly influenced the implementation complexity of system.

On the other hand, there is directly relation in the size of the performance of system and LUT.In theory, LUT is big more, and the fineness of the signal that LUT can represent is also just strong more, just can reflect the variations in detail of signal more, thereby the performance of system is just good more; Otherwise the performance of system is just poor more.But the size of LUT can't be infinitely great; Perhaps requiring under certain systematic function, just requiring under the certain situation of the signal fineness of LUT reflection, because the span of power measurement value is much larger than the span of amplitude metric, so the complexity that power method requires will be higher.

As from the foregoing, traditional non-linear predistortion technology has poor-performing, the higher shortcoming of complexity.Especially when hardware is realized, for example adopt FPGA (Field Programmable Gate Array, field programmable gate array) and IC (Integrated Circuit, integrated circuit) when realizing, the data width of signal also will directly influence final system complexity.

Summary of the invention

The technical problem to be solved in the present invention provides a kind of method of amplifier nonlinearity pre-distortion, overcomes existing non-linear predistortion processing method poor-performing and the high shortcoming of complexity, realizes the non-linear predistortion processing simply efficiently.

At above-mentioned technical problem, the invention provides a kind of method of amplifier nonlinearity pre-distortion, comprise the steps:

A, according to the amplitude A of current demand signal _nAmplitude A with the historical signal sample _HistoryGenerate index value;

B, in the look-up table that has the predistortion factor in advance, find the predistortion factor of described index value correspondence according to described index value;

C, the described predistortion factor and the current demand signal that finds multiplied each other, carry out exporting the signal after the pre-distortion to amplifier in after digital-to-analogue conversion and the upconversion process.

Described method further comprises:

D, catch the output signal of amplifier and it is carried out obtaining baseband feedback signal after down-conversion and the analog-to-digital conversion, utilize Adaptive adjusting algorithm to calculate the new predistortion factor current demand signal after this baseband feedback signal and the delay again, and according to index value the new predistortion factor is write in the look-up table, replace original predistortion factor.

Further, in described method, described look-up table is a two-dimensional look-up table, and correspondingly index value comprises X index and Y index, in the steps A according to the amplitude A of current demand signal sample _nGenerate the X index of look-up table, according to the amplitude A of current demand signal sample _nAmplitude A with the historical signal sample _HistoryGenerate the Y index.

Further, in described method, in the described steps A according to the amplitude A of current demand signal sample _nThe concrete steps that generate the X index of look-up table are:

A101, with the amplitude A of current demand signal sample _nWhole span be divided into L _XThe maximum A of part and definite amplitude _Max

A102, according to described A _n, L _X, A _MaxGenerating the X index according to the following equation is:

Work as A _n＜A _MaxThe time;

X=L _X-1, work as A _n〉=A _MaxThe time;

Wherein,

Expression is the operation of round numbers down; The span of X index be 0,1 ..., L _X-2, L _X-1}.

Further, in described method, in the described steps A according to the amplitude A of current demand signal sample _nAmplitude A with the historical signal sample _HistoryThe concrete steps that generate the Y index are:

A201, determine the amplitude A of historical signal sample according to some historical signal _History

A202, with the amplitude A of described historical signal sample _HistoryAmplitude A with the current demand signal sample _nThe operation of being divided by obtains ratio $B_n=\frac{A_{\mathrm{history}}}{A_n},$ Determine described ratio B _nMaximum B _MaxAnd with B _nWhole span be divided into L _YPart;

A203, according to described B _n, L _Y, B _MaxGenerating the Y index according to the following equation is:

Work as B _n＜B _MaxThe time;

Y=L _Y-1, work as B _n〉=B _MaxThe time;

Wherein,

Expression is the operation of round numbers down; The span of Y index be 0,1 ..., L _Y-2, L _Y-1}.

Further, in described method, the amplitude A of described current demand signal sample _nBe defined as: in the mould of the imaginary part of current demand signal and the mould of real part, the maximum among two moulds adds 1/2nd gained sums of minimum value, promptly

$A_n=\max \left(\right|S_{n,I}|,|S_{n,Q}\left|\right)+\frac{1}{2}\min \left(\right|S_{n,I}|,|S_{n,Q}\left|\right),$

Wherein, S _{N, I}And S _{N, Q}Represent sample of signal S respectively _nReal part and imaginary part.

The Adaptive adjusting algorithm that adopts among the described step D is the minimum mean square self-adaption adjustment algorithm.Described current demand signal or described historical signal are that OFDM is an orthogonal frequency-division multiplex singal.

Because the implementation complexity of non-linear predistortion technology and the size of LUT (look-up table) and the highway width of signal variable have direct relation.Use non-linear predistortion processing method of the present invention, adopt signal amplitude to calculate needed index value as metric, needed metric signal bus width compare with traditional non-linear predistortion processing method in adopt signal power as the tolerance means, implementation complexity obviously reduces.On the other hand, under certain systematic function, just require under the certain situation of the signal fineness of LUT reflection, because the span of power measurement value is much larger than the span of amplitude metric, therefore adopt signal amplitude to carry out pre-distortion, then desired complexity will be lower.

Description of drawings

Fig. 1 is the schematic flow sheet of non-linear predistortion processing method in the embodiment of the invention;

Fig. 2 is the system principle diagram that realizes the non-linear predistortion processing method in the embodiment of the invention.

Embodiment

Below in conjunction with drawings and Examples technical scheme of the present invention is described in detail.

The present invention proposes the method that a kind of non-linear predistortion that can be applicable to high power amplifier is handled, this method be a kind of adaptive, based on the method for 2D (two-dimensional, two dimension) look-up table.Main innovate point is to have proposed in the method a kind of method that comparatively simply generates look-up table index, promptly in the method that generates X and Y index, do not adopt signal power as the tolerance means, but the employing signal amplitude generates required X of pre-distortion and Y index according to signal amplitude.The present invention is particularly useful for systematic function and requires can reduce the complexity that non-linear predistortion is handled under certain situation.

As shown in Figure 1, provided the schematic flow sheet that non-linear predistortion of the present invention is handled, comprised the steps:

Step 101 is according to the amplitude A of current demand signal _nAmplitude A with the historical signal sample _HistoryGenerate index value;

Step 102 is found the predistortion factor of described index value correspondence in the look-up table that has the predistortion factor in advance according to described index value;

Step 103 multiplies each other the described predistortion factor and the current demand signal that finds, and carries out exporting the signal after the pre-distortion to amplifier in after digital-to-analogue conversion and the upconversion process.

Step 104, catch the output signal of amplifier and it is carried out obtaining baseband feedback signal after down-conversion and the analog-to-digital conversion, utilize Adaptive adjusting algorithm to calculate the new predistortion factor current demand signal after this baseband feedback signal and the delay again, and according to index value the new predistortion factor is write in the look-up table, replace original predistortion factor.

As shown in Figure 2, the system principle diagram that realizes the non-linear predistortion processing method of the embodiment of the invention has been described.This system comprises OFDM source module, symbol delay module, two-dimensional seek Table X index generation module, two-dimensional look-up table Y index generation module, look-up table means, multiplying module, digital-to-analogue conversion and up-conversion module, high power amplifier, delay feedback module, analog-to-digital conversion and down conversion module, self adaptation adjusting module and antenna.

As shown in Figure 2, from the signal of OFDM source module output be the predistortion factor exported of look-up table means from LUT after the multiplying module is carried out mutually multiplication, the signal that obtains is after digital-to-analogue conversion and up-conversion (Up-Converter) are handled through DAC again, entering HPA is that high power amplifier carries out the signal amplification, by antenna signal is launched at last again.

The described predistortion factor is stored in the look-up table means, for the predistortion factor that need from the look-up table of look-up table means, to obtain, then need to come joint index LUT by the X index and the Y index of two-dimensional look-up table, search out the suitable predistortion factor, the signal that outputs to the output of multiplying module and OFDM source module then multiplies each other and afterwards exports;

Two-dimensional seek Table X index, the signal that is OFDM source module output are sent into and are generated in the X index generation module after the symbol delay module postpones to handle, and are indexed to look-up table means by this X index generation module output X;

Two-dimensional look-up table Y index, the signal that is OFDM source module output are sent into and are generated in the Y index generation module after the symbol delay module postpones to handle, and are indexed to look-up table means by this Y index generation module output Y;

It is in the look-up table that the predistortion factor is stored in LUT, and the particular content (that is: the different predistortion factors of different index correspondence) of LUT storage can pass through LMS (Least Mean Square, lowest mean square) Adaptive adjusting algorithm is upgraded, wherein, a predistortion factor is exactly a complex values, distinguish all corresponding such complex values with all index that the predistortion factor pair is answered, because of LUT in the present embodiment is a bivariate table, so index herein is meant the joint index of X and Y;

Upgrading when adjusting, from the output signal of transmitting antenna end by coupler acquisition amplifier, is analog-to-digital conversion process through down-conversion and ADC again, obtains baseband feedback signal and sends into the self adaptation adjusting module; Simultaneously, utilize the delay feedback module that the ofdm signal of OFDM source module is postponed to handle, the ofdm signal after delay is handled is sent into the self adaptation adjusting module;

The self adaptation adjusting module then utilizes the adaptive algorithm of self, and for example the minimum mean square self-adaption adjustment algorithm is handled the predistortion factor after obtaining to upgrade to the baseband feedback signal of input and the ofdm signal after the delay processing; At last the new predistortion factor being write LUT is in the look-up table, replaces the predistortion factor original among the LUT.

Further each module in the system shown in Figure 2 is elaborated below.

(1) OFDM source module, the signal of this module output are the baseband signals after handling through up-sampling.Concerning Chinese mobile multimedia broadcast system CMMB (China Mobile MultimediaBroadcasting), the OFDM source signal can be the burst after 3 times of up-samplings again behind the base band framing.

(2) X index generation module is used for the amplitude A according to the current demand signal sample _nGenerate the X index.

Suppose that the current demand signal sample is S _n=S _{N, I}+ jS _{N, Q}, S wherein _{N, I}And S _{N, Q}Represent sample of signal S respectively _nReal part and imaginary part, then the amplitude of current demand signal sample is $A_n=\left|S_n\right|=\sqrt{S_{n,I}^2+S_{n,Q}^2}.$ Because this formula contains out the radical sign computing, implementation is comparatively complicated during processing, so can adopt a kind of comparatively simple approximation method, the amplitude of current demand signal sample can be defined as follows:

$A_n=\max \left(\right|S_{n,I}|,|S_{n,Q}\left|\right)+\frac{1}{2}\min \left(\right|S_{n,I}|,|S_{n,Q}\left|\right)$

When generating the X index, at first with the amplitude A of current demand signal sample _nWhole span be divided into L _XAs 256 parts, and the maximum of hypothesis amplitude is A _MaxAs 32, then every part amplitude is spaced apart A _Max/ L _XAs 1/8, then the generation formula of X index is:

Work as A _n＜A _MaxThe time;

X=L _X-1, work as A _n〉=A _MaxThe time;

Wherein,

Expression is the operation of round numbers down; Thereby, the span of X index be 0,1 ..., L _X-2, L _X-1}.Through above-mentioned conversion, just can utilize the amplitude A of current demand signal sample _nObtain the X index of look-up table.

(3) Y index generation module is used for the amplitude A according to the current demand signal sample _nAmplitude A with the historical signal sample _HistoryGenerate the Y index.

The generation of Y index not only with the amplitude A of current demand signal sample _nRelevant, also with the amplitude A of historical signal sample _HistoryRelevant.The amplitude A of definition historical signal sample _HistoryFor:

A _history＝f(A _n-M，A _n-M+1，Λ，A _n-1)，

It is the amplitude A of historical signal sample _HistoryBe a function of M sample of signal amplitude in history, wherein M is a positive integer, and f () represents a function.In a specific embodiment of the present invention, can adopt A _HistoryA kind of comparatively simple functional form, that is:

$A_{\mathrm{history}}=\frac{A_{n-M}+A_{n-M+1}+\mathrm{\Λ}+A_{n-1}}{M},$ A kind of M=2 that specifically is exemplified as of M wherein.

When generating the Y index, at first with the amplitude A of historical signal sample _History, with the amplitude A of current demand signal sample _nThe operation of being divided by, gained is the result be designated as $B_n=\frac{A_{\mathrm{history}}}{A_n};$

Afterwards, again with ratio B _nWhole span be divided into L _YAs 16 parts, and hypothesis ratio B _nMaximum be B _MaxAs 4, then every part is spaced apart B _Max/ L _YAs 1/4, then the generation formula of Y index is:

Work as B _n＜B _MaxThe time;

Y=L _Y-1, work as B _n〉=B _MaxThe time;

Wherein,

Expression is the operation of round numbers down; Thereby, the span of Y index be 0,1 ..., L _Y-2, L _Y-1}.Through above-mentioned conversion, just can utilize the amplitude A of current demand signal sample _nAmplitude A with the historical signal sample _HistoryObtain the Y index of look-up table.

(4) LUT is a look-up table means, is a block storage during described LUT module specific implementation, and it stores size and is L _X* L _YAs 256*16 element, it is 1 entirely that the initial value of each element can dispose.After handling through the adjustment of self adaptation adjusting module, upgrade initial value with the predistortion factor of upgrading according to index value, as the new preliminary treatment factor, after for example the self adaptation adjusting module adopted the LMS iterative processing to upgrade, the storage complex values was predistortion factor-beta=α exp (j σ), wherein, j represents the phase place prefix in the complex representation, and Sigma represents phase value, and the described predistortion factor itself is exactly a plural form.

(5) self adaptation adjusting module is used for utilizing adaptive algorithm to handle the predistortion factor that generates renewal according to the output signal and the OFDM source signal of high power amplifier, replaces original predistortion factor among the LUT with this new predistortion factor.

Because the LUT table is for bivariate table, promptly with X, Y joint index, wherein, X, Y can be regarded as the coordinate figure of a predistortion factor in the LUT table, directly can find corresponding points by X, Y in the LUT table, can obtain the corresponding predistortion factor, when initial, the value of the predistortion factor can all be changed to 1.

During the adaptive updates predistortion factor, can find the predistortion factor of correspondence position in the LUT table according to X, Y joint index, replace the predistortion factor before with the new predistortion factor, then can realize adaptive updates, make that the predistortion factor of finding is more accurate from the LUT table.

In the present embodiment, the self adaptation adjusting module is a LMS self adaptation adjusting module, and employing LMS Adaptive adjusting algorithm is carried out self adaptation adjustment processing to the output signal and the OFDM source signal of high power amplifier.In the self adaptation adjustment process, sample of signal of every input, following LMS iterative equation group with regard to computing once, and the predistortion factor after will upgrading writes among the LUT.LMS iterative equation group is described below:

e _ρ＝ρ _in-ρ _out

e _θ＝θ _in-θ _out

α _i+1＝α _i+e _ρ*μ _ρ

σ _i+1＝σ _i+e _θ*μ _θ

Wherein:

e _ρAnd e _θRepresent OFDM source signal sample and corresponding amplitude difference and phase difference between the amplifier output signal sample that coupler feeds back after the delayed processing respectively;

α and σ represent the amplitude and the phase place of predistortion factor-beta respectively, and subscript i is corresponding to the original predistortion factor, and subscript i+1 is corresponding to the predistortion factor after upgrading;

μ _ρAnd μ _θRepresent the amplitude LMS stepping factor and the phase place LMS stepping factor respectively.

The stepping factor can influence LMS convergence of algorithm speed, and the performance of pre-distortion algorithms.Set suitable parameters, and after the abundant data of OFDM source output, pre-distortion algorithms just can restrain.As a kind of application example, easy for what calculate, the stepping factor can be set to 2 respectively ^-k, wherein k is a positive integer, as the stepping factor is set is 1/8.In other embodiments of the invention, two stepping factors can be the same, also can be different; Can all be set to 2 negative power power form, concrete power is inferior can be different.

Non-linear predistortion processing method of the present invention can tackle the class ab ammplifier of memory-type; Further, can be applied to ofdm system better.A kind of model of using for emulation of HPA amplifier is the AB class, and memory-type is arranged, and its mathematic(al) representation is as follows:

$x_{\mathrm{out}}\left(n\right)=\underset{p=1,\mathrm{odd}}{\overset{P}{\mathrm{\Σ}}}\underset{q=0}{\overset{Q}{\mathrm{\Σ}}}{a}_{\mathrm{pq}}{x}_{\mathrm{in}}(n-q){\left|x_{\mathrm{in}}(n-q)\right|}^{p-1}$

The parameter that obtains from the class ab ammplifier of the memory-type of a reality is as follows:

a ₁₀＝1.0513+0.0904j a ₃₀＝-0.0542-0.2900j

a ₅₀＝-0.9657-0.7028j a ₁₁＝-0.0680-0.0023j

a ₃₁＝0.2234+0.2317j a ₅₁＝-0.2451-0.3735j

a ₁₂＝0.0289-0.0054j a ₃₂＝-0.0621-0.0932j

a ₅₂＝0.1229+0.1508j

Emulation shows, use the non-linear predistortion processing method of the high power amplifier of the present invention's proposition, no longer adopt power to generate X, the Y index that LUT is a look-up table, but adopt the amplitude of sample of signal to generate X, Y index, make gained X, Y index calculation simple, can obtain systematic function preferably, can realize with lower complexity again.In the planisphere of assessed for performance quality, can obtain planisphere more clearly, show and adopt non-linear predistortion treatment technology of the present invention can obtain systematic function preferably.

Certainly; the present invention also can have other various embodiments; under the situation that does not deviate from spirit of the present invention and essence thereof; those of ordinary skill in the art work as can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection range of the appended claim of the present invention.

Claims (10)

1. the method for an amplifier nonlinearity pre-distortion is characterized in that, comprises the steps:

A, according to the amplitude A of current demand signal _nAmplitude A with the historical signal sample _HistoryGenerate index value;

B, in the look-up table that has the predistortion factor in advance, find the predistortion factor of described index value correspondence according to described index value;

C, the described predistortion factor and the current demand signal that finds multiplied each other, carry out exporting the signal after the pre-distortion to amplifier in after digital-to-analogue conversion and the upconversion process.

2. the method that non-linear predistortion as claimed in claim 1 is handled is characterized in that described method further comprises:

D, catch the output signal of amplifier and it is carried out obtaining baseband feedback signal after down-conversion and the analog-to-digital conversion, utilize Adaptive adjusting algorithm to calculate the new predistortion factor current demand signal after this baseband feedback signal and the delay again, and according to index value the new predistortion factor is write in the look-up table, replace original predistortion factor.

3. the method that non-linear predistortion as claimed in claim 2 is handled is characterized in that,

Described look-up table is a two-dimensional look-up table, and correspondingly index value comprises X index and Y index, in the steps A according to the amplitude A of current demand signal sample _nGenerate the X index of look-up table, according to the amplitude A of current demand signal sample _nAmplitude A with the historical signal sample _HistoryGenerate the Y index.

4. the method that non-linear predistortion as claimed in claim 3 is handled is characterized in that, in the described steps A according to the amplitude A of current demand signal sample _nThe concrete steps that generate the X index of look-up table are:

A101, with the amplitude A of current demand signal sample _nWhole span be divided into L _XThe maximum A of part and definite amplitude _Max

A102, according to described A _n, L _X, A _MaxGenerating the X index according to the following equation is:

Work as A _n＜A _MaxThe time;

X=L _X-1, work as A _n〉=A _MaxThe time;

Wherein,

Expression is the operation of round numbers down; The span of X index be 0,1 ..., L _X-2, L _X-1}.

5. the method that non-linear predistortion as claimed in claim 3 is handled is characterized in that, in the described steps A according to the amplitude A of current demand signal sample _nAmplitude A with the historical signal sample _HistoryThe concrete steps that generate the Y index are:

A201, determine the amplitude A of historical signal sample according to some historical signal _History

A202, with the amplitude A of described historical signal sample _HistoryAmplitude A with the current demand signal sample _nThe operation of being divided by obtains ratio $B_n=\frac{A_{\mathrm{history}}}{A_n},$ Determine described ratio B _nMaximum B _MaxAnd with B _nWhole span be divided into L _YPart;

A203, according to described B _n, L _Y, B _MaxGenerating the Y index according to the following equation is:

Work as B _n＜B _MaxThe time;

Y=L _Y-1, work as B _n〉=B _MaxThe time;

Wherein,

Expression is the operation of round numbers down; The span of Y index be 0,1 ..., L _Y-2, L _Y-1}.

6. the method that non-linear predistortion as claimed in claim 6 is handled is characterized in that,

In the described steps A 201, determine the amplitude A of historical signal sample according to some historical signal _HistoryThe time, determine by functional form, i.e. the amplitude A of historical signal sample _HistoryBe a function of M sample of signal amplitude in history:

A _history＝f(A _n-M，A _n-M+1，Λ，A _n-1)

Wherein, M is a positive integer, and f () represents a function.

7. the method that non-linear predistortion as claimed in claim 7 is handled is characterized in that,

Described function is for asking average function, then the amplitude A of historical signal sample _HistoryFor:

$A_{\mathrm{history}}=\frac{A_{n-M}+A_{n-M+1}+\mathrm{\Λ}+A_{n-1}}{M}.$

8. the method for handling as each described non-linear predistortion in the claim 4 to 7 is characterized in that the amplitude A of described current demand signal sample _nBe defined as: in the mould of the imaginary part of current demand signal and the mould of real part, the maximum among two moulds adds 1/2nd gained sums of minimum value, promptly

$A_n=\max \left(\right|S_{n,I}|,|S_{n,Q}\left|\right)+\frac{1}{2}\min \left(\right|S_{n,I}|,|S_{n,Q}\left|\right),$

Wherein, S _{N, I}And S _{N, Q}Represent sample of signal S respectively _nReal part and imaginary part.

9. the method for handling as each described non-linear predistortion in the claim 2 to 7 is characterized in that the Adaptive adjusting algorithm that adopts among the described step D is the minimum mean square self-adaption adjustment algorithm.

10. the method for handling as each described non-linear predistortion in the claim 1 to 7 is characterized in that described current demand signal or described historical signal are that OFDM is an orthogonal frequency-division multiplex singal.