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

Method for nonlinear predistortion processing of amplifier

Technical Field

The present invention relates to the field of communications technologies, and in particular, to a method for processing nonlinear predistortion of an amplifier in a communications device.

Background

An amplifier is a commonly used type of electronic device in electronic equipment. Linear and nonlinear distortion phenomena occur in any Amplifier, especially in High Power Amplifiers (HPAs), and the distortion phenomena become more significant, and the characteristics of linear and nonlinear distortion phenomena change over time due to aging of the device and other factors.

Linear distortion (also referred to as frequency distortion) which is caused by the different responses of linearly reactive components in an amplifier circuit to different signal frequencies; the linear distortion only changes the amplitude proportional relation and the time delay relation of each frequency component signal, or filters out signals of certain frequency components, but does not increase signals of new frequency components.

And the non-linear distortion is caused by non-linear elements in the amplifier circuit; the nonlinear distortion changes a sine wave into a non-sine wave, which not only contains frequency components (fundamental waves) of an input signal but also generates many new harmonic components.

Thus, it can be said that linear distortion and non-linear distortion, although both also distort the output signal of the amplifier, are essentially completely different. Among them, nonlinear distortion is less easily eliminated than linear distortion due to its poor regularity.

Multicarrier systems, such as OFDM (Orthogonal Frequency division multiplexing) systems, are more sensitive to nonlinear distortion of the amplifier than single carrier systems due to their higher Peak-to-average power Ratio (PAPR). The nonlinear distortion characteristics of an amplifier also have a "memory" feature, i.e., the current output signal of the amplifier is not only related to the current input signal of the amplifier, but also to its past input signals. Therefore, in order to achieve higher amplifier operating efficiency, one must think of ways to eliminate the negative effects of nonlinear distortion.

A common method for eliminating the negative effect of non-linear distortion is to perform inverse distortion treatment on the signal before the signal enters the amplifier, and then send the signal after inverse distortion treatment to the amplifier, and after distortion treatment of the amplifier, the effects of inverse distortion treatment and distortion treatment will cancel each other, thereby ensuring that the output signal of the amplifier is not affected by the non-linear distortion characteristic of the amplifier, but shows linear characteristic. This method is generally referred to as a non-linear predistortion technique.

A common implementation method of the nonlinear predistortion technique is to use a Look-up table (LUT) method, store the predistortion factors in the LUT in advance, index the LUT according to the power of the input signal, find the appropriate predistortion factor, and then perform predistortion processing on the input signal.

In the above-mentioned nonlinear predistortion technique, since the implementation complexity of the nonlinear predistortion technique has a direct relationship with the size of the LUT (look-up table) and the bus width of the signal variable, in the conventional method of generating X and Y indices, the signal power is adopted as a measurement means. The signal power is the square of the signal amplitude, compared to e.g. the signal amplitude. Assuming that 16 bits can be used to fully represent the amplitude of the signal, only 32 bits can be used to fully represent the power of the signal. That is, the signal bus width is much larger by using the signal power as the metric than by using the signal amplitude as the metric, which directly affects the implementation complexity of the system.

On the other hand, there is a direct relationship between the performance of the system and the size of the LUT. Theoretically, the larger the LUT is, the more fineness the LUT can represent the signal is, the more detail changes of the signal can be reflected, and the better the performance of the system is; conversely, the worse the performance of the system. However, the size of the LUT cannot be infinite; or under the condition of requiring certain system performance, namely requiring certain signal fineness reflected by the LUT, the power method requires higher complexity because the value range of the power metric value is far larger than that of the amplitude metric value.

As can be seen from the above, the conventional non-linear predistortion technique has the disadvantages of poor performance and high complexity. Especially when implemented in hardware, such as in FPGA (Field Programmable Gate Array) and IC (Integrated Circuit), the data width of the signal will also directly affect the final system complexity.

Disclosure of Invention

The technical problem to be solved by the invention is to provide a method for nonlinear predistortion processing of an amplifier, which overcomes the defects of poor performance and high complexity of the existing nonlinear predistortion processing method and simply and efficiently realizes nonlinear predistortion processing.

In view of the above technical problem, the present invention provides a method for nonlinear predistortion processing of an amplifier, comprising the following steps:

A. according to the amplitude A of the current signal_nAnd amplitude A of the historical signal samples_historyGenerating an index value;

B. searching a predistortion factor corresponding to the index value in a lookup table pre-storing predistortion factors according to the index value;

C. and multiplying the found predistortion factor by the current signal, performing digital-to-analog conversion and up-conversion processing, and outputting the signal after predistortion processing to the input end of the amplifier.

The method further comprises the following steps:

D. capturing an output signal of the amplifier, performing down-conversion and analog-to-digital conversion on the output signal to obtain a baseband feedback signal, calculating the baseband feedback signal and a delayed current signal by using a self-adaptive adjustment algorithm to obtain a new predistortion factor, and writing the new predistortion factor into a lookup table according to an index value to replace the original predistortion factor.

Further, in the method, the lookup table is a two-dimensional lookup table, and the corresponding index value includes an X index and a Y index, and the step a is based on the amplitude a of the current signal sample_nGenerating an X index of a look-up table based on the amplitude A of the current signal sample_nAnd amplitude A of the historical signal samples_historyA Y index is generated.

Further, in the method, the step A is based on the amplitude A of the current signal sample_nThe specific steps of generating the X index of the lookup table are as follows:

a101, the amplitude A of the current signal sample_nIs equally divided into L_XDividing and determining the maximum value A of the amplitude_max；

A102, according to A_n、L_X、A_maxThe X index is generated according to the following formula:

when A is_n＜A_maxWhen the current is over;

X＝L_X-1, when A_n≥A_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; the value range of the X index is {0, 1,. eta.. and L }_X-2、L_X-1}。

Further, in the method, the step A is based on the amplitude A of the current signal sample_nAnd amplitude A of the historical signal samples_historyThe specific steps for generating the Y index are as follows:

a201, determining the amplitude A of a history signal sample according to a plurality of history signals_history；

A202, the amplitude A of the historical signal sample_historyWith the amplitude a of the current signal sample_nPerforming a phase division operation to obtain a ratio $B_n=\frac{A_{\mathrm{history}}}{A_n},$ Determining the ratio B_nMaximum value of (B)_maxAnd B is_nIs equally divided into L_YPreparing;

a203 according to B_n、L_Y、B_maxThe Y index is generated according to the following formula:

when B is present_n＜B_maxWhen the current is over;

Y＝L_Y1, when B is_n≥B_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; the value range of the Y index is {0, 1, ·_Y-2、L_Y-1}。

Further, in the method, the amplitude a of the current signal sample_nIs defined as: of the modes of the imaginary part and the real part of the current signal, the sum of the maximum of the two modes plus one half of the minimum, i.e.

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

Wherein S is_n，IAnd S_n，QRespectively representing signal samples S_nReal and imaginary parts of (c).

And D, the adaptive adjustment algorithm adopted in the step D is a least mean square adaptive adjustment algorithm. The current signal or the historical signal is an OFDM, orthogonal frequency division multiplexing, signal.

The implementation complexity due to the non-linear predistortion technique has a direct relationship with the size of the LUT (look-up table) and the bus width of the signal variables. By applying the nonlinear predistortion processing method, the required index value is calculated by using the signal amplitude as the metric value, and compared with the traditional nonlinear predistortion processing method in which the signal power is used as the measurement means, the required metric value signal bus width obviously reduces the complexity. On the other hand, under certain system performance, that is, under the condition that the signal fineness reflected by the LUT is required to be certain, since the value range of the power metric is much larger than that of the amplitude metric, the signal amplitude is adopted for predistortion processing, and the required complexity is lower.

Drawings

FIG. 1 is a flow chart of a nonlinear predistortion processing method in an embodiment of the invention;

fig. 2 is a schematic block diagram of a system for implementing a nonlinear predistortion processing method in an embodiment of the present invention.

Detailed Description

The technical solution of the present invention will be described in more detail with reference to the accompanying drawings and examples.

The invention provides a method for nonlinear predistortion processing applicable to a high-power amplifier, which is a self-adaptive method based on a 2D (two-dimensional) lookup table. The main innovation point is that a simpler method for generating the lookup table index is provided in the method, namely, the method for generating the X and Y indexes does not adopt the signal power as a measurement means, but adopts the signal amplitude, and generates the X and Y indexes required by the predistortion processing according to the signal amplitude. The invention is especially suitable for the condition that the system performance requirement is certain, and can reduce the complexity of nonlinear predistortion processing.

As shown in fig. 1, a flow chart of the nonlinear predistortion processing of the present invention is given, which comprises the following steps:

step 101, according to the amplitude A of the current signal_nAnd amplitude A of the historical signal samples_historyGenerating an index value;

step 102, finding a predistortion factor corresponding to the index value in a lookup table in which the predistortion factor is pre-stored according to the index value;

and 103, multiplying the found predistortion factor by the current signal, performing digital-to-analog conversion and up-conversion processing, and outputting the signal after predistortion processing to the input end of the amplifier.

And 104, capturing the output signal of the amplifier, performing down-conversion and analog-to-digital conversion on the output signal to obtain a baseband feedback signal, calculating the baseband feedback signal and the delayed current signal by using a self-adaptive adjustment algorithm to obtain a new predistortion factor, and writing the new predistortion factor into a lookup table according to an index value to replace the original predistortion factor.

Referring to fig. 2, a system block diagram of a nonlinear predistortion processing method embodying an embodiment of the present invention is depicted. The system comprises an OFDM source module, a symbol delay module, a two-dimensional lookup table X index generation module, a two-dimensional lookup table Y index generation module, a lookup table module, a multiplication operation module, a digital-to-analog conversion and up-conversion module, a high-power amplifier, a delay feedback module, an analog-to-digital conversion and down-conversion module, a self-adaptive adjustment module and an antenna.

As can be seen from fig. 2, after the signal output from the OFDM source module and the predistortion factor output from the LUT, i.e., the lookup table module, are multiplied by the multiplication module, the obtained signal is processed by DAC, i.e., digital-to-analog conversion and Up-conversion (Up-Converter), and then enters the HPA, i.e., the high-power amplifier, to amplify the signal, and finally, the signal is transmitted through the antenna.

The predistortion factor is stored in a lookup table module, in order to obtain the required predistortion factor from the lookup table of the lookup table module, an LUT (look-up table) needs to be jointly indexed through an X index and a Y index of a two-dimensional lookup table, the appropriate predistortion factor is found, and then the predistortion factor is output to a multiplication operation module to be multiplied by a signal output by an OFDM (orthogonal frequency division multiplexing) source module and then output;

the two-dimensional lookup table X index is generated by sending a signal output by the OFDM source module into an X index generating module after the signal is delayed by the symbol delay module, and outputting the X index to the lookup table module by the X index generating module;

the two-dimensional lookup table Y index is generated by sending a signal output by the OFDM source module into a Y index generating module after the signal is delayed by the symbol delay module, and outputting the Y index to the lookup table module by the Y index generating module;

the predistortion factors are stored in an LUT (look-up table), and the specific content stored in the LUT (namely, different predistortion factors corresponding to different indexes) can be updated by an LMS (Least Mean Square) adaptive adjustment algorithm, wherein one predistortion factor is a complex value, and all indexes corresponding to the predistortion factors respectively correspond to the complex value;

when updating and adjusting, obtaining an output signal of the amplifier from a transmitting antenna end through the coupler, and obtaining a baseband feedback signal and sending the baseband feedback signal to the self-adaptive adjusting module through down-conversion and ADC (analog-to-digital conversion); meanwhile, the OFDM signal of the OFDM source module is delayed by using a delay feedback module, and the delayed OFDM signal is sent to a self-adaptive adjusting module;

the adaptive adjusting module processes the input baseband feedback signal and the delayed OFDM signal by using a self adaptive algorithm, such as a least mean square adaptive adjusting algorithm, to obtain an updated predistortion factor; and finally, writing the new predistortion factor into an LUT (look-up table), namely a lookup table, and replacing the original predistortion factor in the LUT.

The various modules in the system of FIG. 2 are described in further detail below.

(1) And the OFDM source module outputs a baseband signal after up-sampling processing. For the china Mobile multimedia broadcasting system cmmb (china Mobile multimedia broadcasting), the OFDM source signal may be a signal sequence after being framed by a baseband and then up-sampled by 3 times.

(2) An X index generation module for generating an X index according to the amplitude A of the current signal sample_nAn X index is generated.

Assume that the current signal sample is S_n＝S_n，I+j·S_n，QIn which S is_n，IAnd S_n，QRespectively representing signal samples S_nThe real and imaginary parts of, then the amplitude of the current signal sample is $A_n=\left|S_n\right|=\sqrt{S_{n,I}^2+S_{n,Q}^2}.$ Because the formula contains the root-opening operation and the implementation mode is complex during processing, a simpler approximation method can be adopted, and the amplitude of the current signal sample can be defined as follows:

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

in generating the X index, the amplitude A of the current signal sample is first determined_nIs equally divided into L_XE.g. 256, and assume that the maximum value of the amplitude is a_maxIf 32, the amplitude interval of each portion is A_max/L_XIf 1/8, the generation formula of the X index is:

when A is_n＜A_maxWhen the current is over;

X＝L_X-1, when A_n≥A_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; thus, the value range of the X index is {0, 1, ·_X-2、L_X-1}. By the conversion, the amplitude A of the current signal sample can be used_nAn X index of the lookup table is obtained.

(3) A Y index generation module for generating an amplitude A according to the current signal sample_nAnd amplitude A of the historical signal samples_historyA Y index is generated.

The generation of the Y index is not only related to the amplitude A of the current signal sample_nRelated also to the amplitude A of the historical signal samples_historyIt is related. Defining the amplitude A of the historical signal samples_historyComprises the following steps:

A_history＝f(A_n-M，A_n-M+1，Λ，A_n-1)，

i.e. the amplitude a of the historical signal samples_historyIs a function of the amplitude of the historical M signal samples, where M is a positive integer and f (·) represents a function. In one embodiment of the present invention, A may be employed_historyA simpler functional form of (a), namely:

<math><mrow> <msub> <mi>A</mi> <mi>history</mi> </msub> <mo>=</mo> <mfrac> <mrow> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mi>M</mi> </mrow> </msub> <mo>+</mo> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mi>M</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <mi>&Lambda;</mi> <mo>+</mo> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mrow> <mi>M</mi> </mfrac> <mo>,</mo> </mrow></math> one specific example of M is M ═ 2.

In generating the Y index, the amplitude A of the historical signal sample is first determined_historyWith the amplitude A of the current signal sample_nThe result of the division is recorded $B_n=\frac{A_{\mathrm{history}}}{A_n};$

Then, the ratio B is calculated_nIs equally divided into L_YE.g., 16, and assume the ratio B_nMaximum value of (A) is B_maxIf 4, each interval is B_max/L_YIf 1/4, the generation formula of the Y index is:

when B is present_n＜B_maxWhen the current is over;

Y＝L_Y1, when B is_n≥B_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; thus, the value range of the Y index is {0, 1, ·_Y-2、L_Y-1}. By the conversion, the amplitude A of the current signal sample can be used_nAnd amplitude A of the historical signal samples_historyA Y index of the lookup table is obtained.

(4) The LUT is a look-up table module, which is implemented as a block memory having a size L stored therein_X*L_YFor example, 256 × 16 elements, the initial value of each element may be configured to be 1. After the adjustment processing of the adaptive adjustment module, the initial value is updated by the updated predistortion factor according to the index value, and the updated predistortion factor is used as a new preprocessing factor, for example, after the adaptive adjustment module is updated by the LMS iteration processing, the complex value is stored as a predistortion factor β ═ α exp (j σ), wherein,j denotes the phase prefix in a complex representation and sigma denotes the phase value, the predistortion factor itself being in complex form.

(5) And the self-adaptive adjusting module is used for generating an updated predistortion factor by utilizing self-adaptive algorithm processing according to the output signal of the high-power amplifier and the OFDM source signal, and replacing the original predistortion factor in the LUT with the new predistortion factor.

Since the LUT table is a two-dimensional table, i.e., the LUT table is jointly indexed by X, Y, X, Y can be understood as a coordinate value of a predistortion factor in the LUT table, and the corresponding point can be directly found in the LUT table by X, Y, i.e., the corresponding predistortion factor can be obtained, and initially, the values of the predistortion factors can all be set to 1.

When the predistortion factors are updated in a self-adaptive mode, the predistortion factors at the corresponding positions in the LUT can be found according to the X, Y combined index, and the previous predistortion factors are replaced by the new predistortion factors, so that the self-adaptive updating can be realized, and the predistortion factors searched from the LUT are more accurate.

In this embodiment, the adaptive adjustment module is an LMS adaptive adjustment module, and an LMS adaptive adjustment algorithm is used to perform adaptive adjustment processing on the output signal of the high-power amplifier and the OFDM source signal. In the adaptive adjustment process, the LMS iterative equation system described below operates once every input of one signal sample, and the updated predistortion factor is written into the LUT. The LMS iteration equation set is described as follows:

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_θRespectively representing the amplitude difference and the phase difference between the OFDM source signal sample after the delay processing and the corresponding amplifier output signal sample fed back by the coupler;

alpha and sigma respectively represent the amplitude and phase of the predistortion factor beta, subscript i corresponds to the original predistortion factor, and subscript i +1 corresponds to the updated predistortion factor;

μ_ρand mu_θThe amplitude LMS step factor and the phase LMS step factor are represented separately.

The step factor may affect the convergence speed of the LMS algorithm, as well as the performance of the predistortion algorithm. After appropriate parameters are set and enough data are output by the OFDM source, the predistortion algorithm can be converged. As an application example, for simplicity of calculation, the step factors may be set to 2 respectively^-kWhere k is a positive integer, such as setting the step factor to 1/8. In other embodiments of the present invention, the two step factors may or may not be the same; may be set to the negative power form of 2, and the specific powers may be different.

The nonlinear predistortion processing method of the invention can deal with the AB class amplifier with memory; furthermore, the method can be better applied to an OFDM system. One model for HPA amplifiers for simulation is class AB, memory type, whose mathematical expression is as follows:

<math><mrow> <msub> <mi>x</mi> <mi>out</mi> </msub> <mrow> <mo>(</mo> <mi>n</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>p</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mi>odd</mi> </mrow> <mi>P</mi> </munderover> <munderover> <mi>&Sigma;</mi> <mrow> <mi>q</mi> <mo>=</mo> <mn>0</mn> </mrow> <mi>Q</mi> </munderover> <msub> <mi>a</mi> <mi>pq</mi> </msub> <msub> <mi>x</mi> <mi>in</mi> </msub> <mrow> <mo>(</mo> <mi>n</mi> <mo>-</mo> <mi>q</mi> <mo>)</mo> </mrow> <msup> <mrow> <mo>|</mo> <msub> <mi>x</mi> <mi>in</mi> </msub> <mrow> <mo>(</mo> <mi>n</mi> <mo>-</mo> <mi>q</mi> <mo>)</mo> </mrow> <mo>|</mo> </mrow> <mrow> <mi>p</mi> <mo>-</mo> <mn>1</mn> </mrow> </msup> </mrow></math>

the parameters obtained from a practical class AB amplifier of the memory type are as follows:

a₁₀＝1.0513+0.0904j a₃₀＝-0.0542-0.2900j

a₅₀＝-0.9657-0.7028j a₁₁＝-0.0680-0.0023j

a₃₁＝0.2234+0.2317j a₅₁＝-0.2451-0.3735j

a₁₂＝0.0289-0.0054j a₃₂＝-0.0621-0.0932j

a₅₂＝0.1229+0.1508j

simulation shows that by applying the nonlinear predistortion processing method of the high-power amplifier provided by the invention, power is not used for generating X, Y indexes of LUT (look-up table), but amplitude of signal samples is used for generating X, Y indexes, so that the obtained X, Y index is simple to calculate, and not only can better system performance be obtained, but also the nonlinear predistortion processing method can be realized with lower complexity. In the constellation diagram with good and bad performance, a clearer constellation diagram can be obtained, which shows that better system performance can be obtained by adopting the nonlinear predistortion processing technology of the invention.

The present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof, and it should be understood that various changes and modifications can be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (10)

1. A method of nonlinear predistortion processing of an amplifier, comprising the steps of:

A. according to the amplitude A of the current signal_nAnd amplitude A of the historical signal samples_historyGenerating an index value;

B. searching a predistortion factor corresponding to the index value in a lookup table pre-storing predistortion factors according to the index value;

C. and multiplying the found predistortion factor by the current signal, performing digital-to-analog conversion and up-conversion processing, and outputting the signal after predistortion processing to the input end of the amplifier.

2. The method of nonlinear predistortion processing as set out in claim 1, wherein said method further comprises:

D. capturing an output signal of the amplifier, performing down-conversion and analog-to-digital conversion on the output signal to obtain a baseband feedback signal, calculating the baseband feedback signal and a delayed current signal by using a self-adaptive adjustment algorithm to obtain a new predistortion factor, and writing the new predistortion factor into a lookup table according to an index value to replace the original predistortion factor.

3. The method of nonlinear predistortion processing as set out in claim 2,

the lookup table is a two-dimensional lookup table, the index values correspondingly comprise an X index and a Y index, and the step A is carried out according to the amplitude A of the current signal sample_nGenerating an X index of a look-up table based on the amplitude A of the current signal sample_nAnd amplitude A of the historical signal samples_historyA Y index is generated.

4. A method for non-linear pre-distortion processing as claimed in claim 3, wherein the step a is based on the amplitude a of the current signal sample_nThe specific steps of generating the X index of the lookup table are as follows:

a101, the amplitude A of the current signal sample_nIs equally divided into L_XDividing and determining the maximum value A of the amplitude_max；

A102, according to A_n、L_X、A_maxThe X index is generated according to the following formula:

when A is_n＜A_maxWhen the current is over;

X＝L_X-1, when A_n≥A_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; the value range of the X index is {0, 1,. eta.. and L }_X-2、L_X-1}。

5. A method for non-linear pre-distortion processing as claimed in claim 3, wherein the step a is based on the amplitude a of the current signal sample_nAnd amplitude A of the historical signal samples_historyThe specific steps for generating the Y index are as follows:

a201, determining the amplitude A of a history signal sample according to a plurality of history signals_history；

A202, the amplitude A of the historical signal sample_historyWith the amplitude a of the current signal sample_nPerforming a phase division operation to obtain a ratio $B_n=\frac{A_{\mathrm{history}}}{A_n},$ Determining the ratio B_nMaximum value of (B)_maxAnd B is_nIs equally divided into L_YPreparing;

a203 according to B_n、L_Y、B_maxThe Y index is generated according to the following formula:

when B is present_n＜B_maxWhen the current is over;

Y＝L_Y1, when B is_n≥B_maxWhen the current is over;

wherein,

represents the operation of taking down an integer; the value range of the Y index is 0、1、......、L_Y-2、L_Y-1}。

6. The method of nonlinear predistortion processing as set out in claim 6,

in step A201, the amplitude A of the historical signal sample is determined according to a plurality of historical signals_historyIs determined by a functional form, i.e. the amplitude A of the historical signal sample_historyIs a function of the amplitude of the historical M signal samples:

A_history＝f(A_n-M，A_n-M+1，Λ，A_n-1)

where M is a positive integer and f (·) represents a function.

7. The method of nonlinear predistortion processing as set out in claim 7,

the function is an averaging function, the amplitude A of the historical signal sample_historyComprises the following steps:

<math><mrow> <msub> <mi>A</mi> <mi>history</mi> </msub> <mo>=</mo> <mfrac> <mrow> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mi>M</mi> </mrow> </msub> <mo>+</mo> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mi>M</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <mi>&Lambda;</mi> <mo>+</mo> <msub> <mi>A</mi> <mrow> <mi>n</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mrow> <mi>M</mi> </mfrac> <mo>.</mo> </mrow></math>

8. a method for non-linear pre-distortion processing as claimed in any of claims 4 to 7, characterized in that the amplitude A of the current signal sample_nIs defined as: of the modes of the imaginary part and the real part of the current signal, the sum of the maximum of the two modes plus one half of the minimum, i.e.

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

Wherein S is_n，IAnd S_n，QRespectively representing signal samples S_nReal and imaginary parts of (c).

9. The method according to any of claims 2 to 7, wherein the adaptive adjustment algorithm used in step D is a least mean square adaptive adjustment algorithm.

10. The method of nonlinear predistortion processing as claimed in any of claims 1 to 7, characterized in that the current signal or the historical signal is an OFDM (orthogonal frequency division multiplexing) signal.