CN117880045A

CN117880045A - Predistortion model construction method, predistortion model construction device, electronic equipment and medium

Info

Publication number: CN117880045A
Application number: CN202311841914.6A
Authority: CN
Inventors: 姜宗琳
Original assignee: Suzhou Huatai Electronics Co Ltd
Current assignee: Suzhou Huatai Electronics Co Ltd
Priority date: 2023-12-28
Filing date: 2023-12-28
Publication date: 2024-04-12

Abstract

The application provides a predistortion model construction method, a predistortion model construction device, electronic equipment and a predistortion model construction medium, and relates to the technical field of digital communication, wherein the specific scheme comprises the following steps: acquiring target mapping relation information of a first signal and a second signal, wherein the first signal is an output signal or a baseband signal, the second signal is a predistortion signal, and the predistortion signal is obtained based on the baseband signal; vectorizing information in the target mapping relation information, and converting the vectorized information to obtain relation information of a second signal and a first signal, wherein the relation information of the second signal and the first signal comprises coefficient vectors, baseband signals containing phase information and output signals containing envelope information; and constructing a predistortion model according to the relation information of the second signal and the first signal, wherein the predistortion model comprises the relation information of the baseband signal and the predistortion signal. Thus, the accuracy of the output result of the power amplifier can be improved.

Description

Predistortion model construction method, predistortion model construction device, electronic equipment and medium

Technical Field

The application belongs to the technical field of digital communication, and particularly relates to a method, a device, electronic equipment and a medium for constructing a predistortion model.

Background

Modern communication systems use Power Amplifiers (PA) to amplify signals. However, the power amplifier is not perfectly linear, i.e. the power amplifier is not capable of linearly amplifying the input signal.

On the basis, in order to improve the linearity of the power amplifier, the information in the output signal and the input signal is ensured to be consistent as much as possible. A predistortion model may be established by a digital predistortion technique, the predistortion model being used to represent a mapping of an input signal x (t) to a predistortion signal u (t), and then u (t) being used as an input signal to the power amplifier, i.e. the nonlinear distortion of the power amplifier is compensated by the predistortion model. In this way, the output signal y (t) of the power amplifier can be more accurate.

In an ideal case, the output signal of the power amplifier may be expressed as y (t) =g·x (t), where G represents the gain, typically set to a constant. According to the above formula, the output signals y (t) and x (t) are generally considered to include the same information, and thus the prior art builds a predistortion model using either the input signal or the output signal alone when building the predistortion model. In practice, however, the output signal x (t) of the power amplifier and the input signal y (t) contain different information, since the power amplifier is not capable of linearly amplifying the input signal. Therefore, the accuracy of the output signal y (t) of the power amplifier is poor on the basis of constructing a predistortion model using the input signal or the output signal alone.

Disclosure of Invention

The embodiment of the application provides a method, a device, electronic equipment and a medium for constructing a predistortion model, which can improve the accuracy of an output result of a power amplifier.

In a first aspect, an embodiment of the present application provides a method for constructing a predistortion model, including:

acquiring target mapping relation information of a first signal and a second signal, wherein the first signal is an output signal or a baseband signal, the second signal is a predistortion signal, and the predistortion signal is obtained based on the baseband signal;

vectorizing information in the target mapping relation information, and converting the vectorized information to obtain relation information of the second signal and the first signal, wherein the relation information of the second signal and the first signal comprises coefficient vectors, baseband signals containing phase information and output signals containing envelope information;

and constructing a predistortion model according to the relation information of the second signal and the first signal, wherein the predistortion model comprises the relation information of the baseband signal and the predistortion signal.

In one possible implementation manner, the vectorizing the information in the target mapping relationship information, and converting the vectorized information to obtain relationship information between the second signal and the first signal includes:

Converting the target mapping relation information into a vector form to obtain a vector matrix of the target mapping relation;

performing conversion processing on the vector matrix to obtain an inverse model vector matrix corresponding to the relation information of the second signal and the first signal;

replacing vectors in the inverse model vector matrix to obtain a target vector matrix comprising the coefficient vector, the phase information and the envelope information;

calculating a target coefficient vector in the case of a minimum of an error function based on the target vector matrix and the output signal;

and determining the relation information of the second signal and the first signal according to the target coefficient vector and the target vector matrix.

In one possible implementation, each vector of the inverse model vector matrix composed of nonlinear terms includes an absolute value portion and a non-absolute value portion; the replacing the vector in the inverse model vector matrix to obtain a target vector matrix including the coefficient vector, the phase information and the envelope information includes:

the inverse model vector matrix is used for representing relation information of the predistortion signal and the output signal, and for each vector formed by nonlinear terms in the inverse model vector matrix, a non-absolute value part in the vector is replaced by a baseband signal comprising the phase information, so that the target vector matrix is obtained; or,

The inverse model vector matrix is used for representing relation information of the predistortion signal and the baseband signal, and for each vector formed by nonlinear terms in the first vector matrix, the absolute value part in the vector is replaced by an output signal comprising the envelope information, so that the target vector matrix is obtained.

In one possible implementation manner, after the pre-distortion model is constructed according to the relation information of the second signal and the first signal, the method further includes:

acquiring a baseband signal to be processed at intervals of a preset time length;

inputting the baseband signal to be processed into the predistortion model to obtain a target predistortion signal;

calculating an error value of the error function based on the target predistortion signal and an output signal at the current moment;

judging whether the error value is larger than a preset threshold value or not;

if yes, executing the target mapping relation information of the first signal and the second signal, carrying out vectorization processing on the information in the target mapping relation information, converting to obtain relation information of the second signal and the first signal, and constructing a predistortion model according to the relation information of the second signal and the first signal;

If not, adopting a gradient descent method to carry out iterative computation on the error function, and determining an update coefficient vector corresponding to the error function after iterative computation;

and updating the coefficient vector of the predistortion model into the updated coefficient vector.

In one possible implementation, the expression of the relationship information of the second signal and the first signal is as follows:

wherein K is _a 、K _b 、K _c An index set for representing an order of the relationship information of the second signal and the first signal, L _a 、L _b 、L _c Is the index set of memory depth, M _b 、M _c Is an index set expressing the dissynchrony of the signal and the envelope, l represents the envelope orAnd a time difference between the phase and the first signal, m representing the time difference between the envelope and the first signal.

In a second aspect, an embodiment of the present application provides an apparatus for predistortion model construction, where the apparatus includes:

the device comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring target mapping relation information of a first signal and a second signal, the first signal is an output signal or a baseband signal, the second signal is a predistortion signal, and the predistortion signal is obtained based on the baseband signal;

the processing module is used for vectorizing the information in the target mapping relation information and converting the vectorized information to obtain relation information of the second signal and the first signal, wherein the relation information of the second signal and the first signal comprises coefficient vectors, baseband signals containing phase information and output signals containing envelope information;

And the construction module is used for constructing a predistortion model according to the relation information of the second signal and the first signal, wherein the predistortion model comprises the relation information of the baseband signal and the predistortion signal.

In a third aspect, an embodiment of the present application provides an electronic device, including: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, implements the method of predistortion model construction as described in the first aspect.

In a fourth aspect, embodiments of the present application provide a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement a method of predistortion model construction as set forth in the first aspect.

According to the method, the device, the electronic equipment and the medium for constructing the digital predistortion model, after target mapping relation information of the first signal and the second signal is obtained, vectorization processing is carried out on the target mapping relation information, and conversion is carried out to obtain relation information of the second signal and the first signal, wherein the relation information of the second signal and the first signal comprises relation information of a predistortion signal and an output signal or relation information of the predistortion signal and a baseband signal, the relation information of the second signal and the first signal comprises coefficient vectors, the output signal comprising envelope information and the baseband signal comprising phase information, so that the constructed relation information of the second signal and the first signal refers to the envelope information and the phase information at the same time, and the relation information of the second signal and the first signal comprises relation information capable of more accurately reflecting the predistortion signal, the baseband signal and the output signal. When the relation information of the second signal and the first signal comprises the predistortion model construction, the accuracy of the predistortion model is improved due to the fact that the output signal comprising envelope information and the baseband signal comprising phase information, namely the input signal, are considered at the same time, and therefore the accuracy of the power amplifier is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described, and it is possible for a person skilled in the art to obtain other drawings according to these drawings without inventive effort.

FIG. 1 is an exemplary schematic diagram of a prior art predistortion technique;

FIG. 2 is a schematic diagram of a system architecture of a predistortion model construction method in the prior art;

FIG. 3 is a flow chart of a method for predistortion model construction provided in an embodiment of the present application;

fig. 4 is a schematic system architecture diagram of a predistortion model construction method according to an embodiment of the present application;

FIG. 5 is a flowchart of a method for updating a predistortion model according to an embodiment of the present application;

FIG. 6 is an exemplary schematic diagram of a method of predistortion model construction provided by an embodiment of the present application;

fig. 7 is a schematic structural diagram of an apparatus for predistortion model construction according to an embodiment of the present application;

fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Features and exemplary embodiments of various aspects of the present application are described in detail below to make the objects, technical solutions and advantages of the present application more apparent, and to further describe the present application in conjunction with the accompanying drawings and the detailed embodiments. It should be understood that the specific embodiments described herein are intended to be illustrative of the application and are not intended to be limiting. It will be apparent to one skilled in the art that the present application may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present application by showing examples of the present application.

It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.

The following describes the terminology involved in this application:

a power amplifier: for power amplifying an input signal;

digital Pre-Distortion (DPD): by establishing a digital predistortion model, the input signal is compensated in advance, and the quality of the PA output signal is improved.

As shown in fig. 1, assuming that the input signal is x (t) and the output signal is y (t), since PA is not an amplified input signal that can be perfectly linear, a DPD model is constructed by DPD technique, and during the mapping, the predistortion signal u (t) is mapped to x (t), and in consideration of the nonlinear distortion of PA, the x (t) is compensated by DPD model in advance, so as to obtain u (t), so that after u (t) is input to PA, the output signal of PA is a result of linearly amplifying x (t), i.e., y (t) =g×x (t), where G represents a gain, and is generally set to be constant. As an example, G is set to 1.

Currently, the DPD model can be constructed in two ways:

mode 1, directly constructing a mapping relationship between u (t) and x (t).

Mode 2 constructs a mapping relationship between u (t) and x (t) from an output signal y (t) of PA.

Taking the example of mode 2, how to construct the DPD model is described with reference to fig. 2:

let x (t) be input, y (t) be output, where x (n) is obtained by discretizing the sample of x (t), y (n) is obtained by discretizing the sample of y (t), DPD signal u (n), error function e (n) =x (n) -y (n)/G.

As shown in fig. 2, x (n) is input to the DPD model to obtain u (n), and u (n) is input to the PA to obtain y (n).

In the DPD model construction stage, the generalized memory polynomial expression from u (n) to y (n) is:

wherein K is _a 、K _b 、K _c Index set for representing order, L _a 、L _b 、L _c Is the index set of memory depth, M _b 、M _c Is an index set that expresses signal and envelope dyssynchrony. The current time is n, (n-l) indicates that one time is pushed forward from the current time n, and n-l+m indicates that one time is pushed forward from the current time n and then m times are pushed backward.

The nonlinear term in the generalized memory polynomial expression may be written in the form of a vector, and the generalized memory polynomial may be represented by a vector matrix as follows:

y＝Uw

wherein U is represented by U (n-l) |u (n-l) | ^k Vector of non-linear terms, w being the coefficient to be solved for a _kl 、b _klm 、c _klm And the like.

On the basis, an inverse model can be constructed by the vector matrix, wherein the inverse model is u=yw, and Y is Y (n-l) |y (n-l) | ^k And the vector of the coefficients to be calculated, and w is a coefficient vector of the coefficients to be calculated. The inverse model is used to represent the relationship information of the predistortion signal and the output signal.

The value of the coefficient vector is then determined by taking the above inverse model u=yw as x (n) into the error function e (n) =x (n) -y (n)/G, and solving the minimum value of the absolute value or the square of the absolute value of the error function.

And (5) bringing the values of the coefficient vector into a vector matrix of the inverse model to determine the target inverse model.

With the target inverse model, a DPD model from x (t) to u (t) can be constructed by taking y (n) =gx (n) into the target inverse model vector matrix.

As shown in fig. 2, 1/G in fig. 2 is nonlinear compensation of an output signal, and a DPD Training module (DPD Training) determines coefficient vectors in a DPD model by calculating an absolute value of an error function e (n) or a minimum value of squares of the absolute values, and then waits for next DPD model update, replaces the DPD model in use with the latest DPD model, and discards the DPD model in use.

On this basis, when constructing the DPD model, the DPD training module models only using x (t) or y (t), and because PA is not perfectly linear, the information in x (t) or y (t) is different, so the information used when constructing the DPD model is incomplete, resulting in inaccurate DPD model constructed, further resulting in inaccurate PA output signal,

in order to solve the problems in the prior art, the embodiment of the application provides a method, a device, electronic equipment and a medium for constructing a predistortion model. The method for constructing the predistortion model provided by the embodiment of the application is first described below.

Fig. 3 shows a flowchart of a method for constructing a predistortion model according to an embodiment of the present application. As shown in fig. 3, the method is applied to an electronic device, and specifically includes:

s301, acquiring target mapping relation information of a first signal and a second signal.

The first signal is an output signal or a baseband signal, the second signal is a predistortion signal, and the predistortion signal is obtained based on the baseband signal.

It will be appreciated that if the first signal is an output signal and the second signal is a predistortion signal, the target mapping relationship information is a mapping relationship between the output signal and the predistortion signal, that is, the predistortion signal is used to represent the output signal, and as an example, the target mapping relationship may be represented as

If the first signal is a baseband signal and the second signal is a predistortion signal, the target mapping relation information is the mapping relation between the baseband signal and the predistortion signal, namely, the predistortion signal is used for representing the baseband signal.

S302, vectorizing the information in the target mapping relation information, and converting the information to obtain relation information of the second signal and the first signal.

The relation information of the second signal and the first signal comprises coefficient vectors, baseband signals containing phase information and output signals containing envelope information.

The vectorization processing is to represent the information in the target mapping relation information by vectors and the mathematical relation between the information by matrixes.

It is understood that in S301, the first signal is represented by the second signal in the obtained target mapping relationship information. After the target mapping relation information of the vectorization processing is converted, the first signal can be used for representing the second signal, namely, an inverse model is constructed.

The expression of the relationship information of the second signal and the first signal is as follows:

wherein K is _a 、K _b 、K _c Index set for representing the order of the relation information of the second signal and the first signal, L _a 、L _b 、L _c Is the index set of memory depth, M _b 、M _c Is an index set expressing the dissynchronability of the signal from the envelope, i represents the time difference between the envelope or the phase and the first signal, and m represents the time difference between the envelope and the first signal.

Continuing the above example, the first signal is the output signal y _GMP (n), the second signal is a predistortion signal U (n), and vectorization processing is performed on the target mapping relation information to obtain y=uw, where U is U (n-l) |u (n-l) | ^k And the nonlinear term is formed by converting the y=uw and performing inverse model construction to obtain u=yw, wherein Y is x (n-l) |y (n-l) | ^k And the like. The information of the relationship between the second signal and the first signal corresponding to the vector matrix u=yw is:

i.e. u=yw denotes the second signal by the first signal.

S303, constructing a predistortion model according to the relation information of the second signal and the first signal.

The predistortion model is the digital predistortion technical model. The predistortion model includes relation information of the baseband signal and the predistortion signal. The predistortion model is used to represent the mapping relationship of the predistortion signal and the baseband signal.

After acquiring the relationship information between the second signal and the first signal, a predistortion model may be constructed using y (n) =g×x (n).

Continuing the above example, replacing y in the u (n) expression with G x (n) to obtain mapping relation information of u (n) and x (n), and obtaining mapping relation information of x (n) and u (n) by a matrix transformation mode to obtain a predistortion model.

After the target mapping relation information of the first signal and the second signal is obtained, vectorizing the target mapping relation information and converting the target mapping relation information to obtain the relation information of the second signal and the first signal, wherein the relation information of the second signal and the first signal comprises the relation information of a predistortion signal and an output signal or the relation information of the predistortion signal and a baseband signal, the relation information of the second signal and the first signal comprises a coefficient vector, the output signal comprising envelope information and the output signal comprising phase information, and thus, the constructed relation information of the second signal and the first signal simultaneously refers to the envelope information and the phase information, and the relation information of the second signal and the first signal comprises the relation capable of more accurately reflecting the predistortion signal and the baseband signal and the output signal. When the relation information of the second signal and the first signal comprises the predistortion model construction, the output signal comprising the envelope information and the baseband signal comprising the phase information are considered at the same time, namely the input signal and the output signal are considered at the same time, so that the accuracy of the predistortion model is improved, and the accuracy of the power amplifier is improved.

Aiming at the S302, the information in the target mapping relation information is vectorized and converted to obtain the relation information of the second signal and the first signal, the implementation mode comprises the following steps of specifically comprising the steps of 1 to 5:

and step 1, converting the target mapping relation information into a vector form to obtain a vector matrix of the target mapping relation.

Specifically, the nonlinear item and the unknown coefficient in the target mapping relation information are written in the form of vectors. The unknown coefficients are undetermined coefficients in the target mapping relation, and the unknown coefficients can be model parameters for adjusting the predistortion model. The method of calculating the unknown coefficients is elaborated in the following embodiments.

And 2, performing conversion processing on the vector matrix to obtain an inverse model vector matrix corresponding to the relation information of the second signal and the first signal.

It can be understood that the vector matrix in step 1 is transformed, and the transformed result is also a vector matrix, which is distinguished in that the inverse model vector matrix is used for representing the relationship information of the second signal and the first signal, and the vector matrix of the target mapping relationship is used for representing the relationship information of the first signal and the second signal.

And 3, replacing vectors in the inverse model vector matrix to obtain a target vector matrix comprising coefficient vectors, phase information and envelope information.

The specific method of replacing vectors refers to the following embodiments.

And 4, calculating a target coefficient vector under the condition of the minimum value of the error function based on the target vector matrix and the output signal.

The target coefficient vector is the unknown coefficient.

The minimum value of the error function means the minimum value of the absolute value of the error function or the minimum value of the square of the absolute value. The errors of the baseband signal and the output signal can be reflected more clearly by the absolute value of the error function or the square of the absolute value.

As an example, the error function is e (n) =x (n) -y (n)/G, and the target coefficient vector is obtained by calculating the minimum value of the error function by taking the target vector matrix as x (n) in the error function, and determining the value of the coefficient vector when the value of the error function is the minimum value.

And 5, determining the relation information of the second signal and the first signal according to the target coefficient vector and the target vector matrix.

It can be understood that the above calculated target coefficient vector is replaced by the coefficient vector in the inverse model vector matrix to obtain the target vector matrix, and then the target vector matrix is rewritten into a polynomial expression composed of nonlinear terms, where the polynomial expression is used to represent the relationship information of the second signal and the first signal. The information of the relationship between the second signal and the first signal is that the first signal is used to represent the second signal.

In this way, the target vector matrix obtained by means of matrix transformation can be used to characterize the mapping relationship between the second signal and the first signal. And then calculating error values of the baseband signal and the output signal by using an error function, so that an accurate target coefficient vector can be obtained, and the relation information of the second signal and the first signal obtained by considering the phase information of the baseband signal and the envelope information of the output signal can be more accurate, thereby reducing the nonlinear distortion of the power amplifier.

Specifically, each vector formed by nonlinear terms in the inverse model vector matrix comprises an absolute value part and a non-absolute value part, and on the basis, the vector in the inverse model vector matrix is replaced in the step 3 to obtain a target vector matrix comprising coefficient vectors, phase information and envelope information, and the following two implementation modes specifically exist:

in a specific implementation manner, if an inverse model vector matrix is used to represent the relation information of the predistortion signal and the output signal, for each vector in the inverse model vector matrix, which is formed by nonlinear terms, a non-absolute value part in the vector is replaced by a baseband signal comprising phase information, so as to obtain a target vector matrix.

As an example, the inverse model vector matrix expression is u=yw, where Y is Y (n-l) |y (n-l) | ^k And (3) replacing the non-absolute value part y (n-l) in the nonlinear term with x (n-l) to obtain a target vector matrix, wherein w is a coefficient vector formed by the nonlinear term and w is an unknown coefficient.

In a specific implementation, if an inverse model vector matrix is used to represent the relationship information between the predistortion signal and the baseband signal, for each vector in the first vector matrix that is formed by nonlinear terms, the absolute value part in the vector is replaced by an output signal that includes envelope information, so as to obtain a target vector matrix.

As an example, the inverse model vector matrix expression is u=xw, X is the expression X (n-l) |x (n-l) | ^k The vector of nonlinear terms, where w is the coefficient vector of unknown coefficients, will be the absolute value of the component |x (n-l) |in the nonlinear term ^k Replaced by |y (n-l))| ^k And obtaining a target vector matrix.

With the above two specific implementations, if an inverse model vector matrix is used to represent the relationship information between the predistortion signal and the output signal, each vector formed by the nonlinear term in the inverse model vector matrix is formed by the output signal, i.e. the inverse model vector matrix refers only to the output signal. For each vector of the inverse model vector matrix that is composed of nonlinear terms, the non-absolute value portion of the vector is replaced with a baseband signal that includes phase information. And compared with envelope information, the influence of the phase information on the accuracy of the predistortion model is larger, so that the accuracy of the predistortion model is improved by replacing the non-absolute value part with a baseband signal comprising the phase information. Similarly, if the inverse model vector matrix is used to represent the relationship information between the predistortion signal and the baseband signal, for each vector in the first vector matrix, which is formed by nonlinear terms, the absolute value part in the vector is replaced by an output signal comprising envelope information, so that the accuracy of the predistortion model is improved.

Based on the foregoing embodiments, an embodiment of the present application provides a system architecture schematic diagram of a predistortion model, as shown in fig. 4, where the system includes the predistortion model, a PA, and a DPD training module.

Wherein 1/G is used to represent the nonlinear compensation of the output signal.

The DPD training module comprehensively considers nonlinear compensation 1/G of the baseband signal x (n) and the output signal y (n), and an accurate predistortion model is constructed by calculating an error function e (n), adjusting coefficient vectors and continuously reducing errors of the baseband signal and the output signal. The specific method for constructing the predistortion model refers to the related description in the above embodiment, and will not be repeated here.

In some embodiments of the present application, after the predistortion model is constructed, the coefficient vector of the predistortion model may also be updated during the operation of the predistortion model. As shown in fig. 5, after constructing a predistortion model according to the relationship information between the second signal and the first signal in S303, the method further includes:

s304, acquiring a baseband signal to be processed at intervals of a preset time.

The preset time period may be preset empirically.

S305, inputting the baseband signal to be processed into a predistortion model to obtain a target predistortion signal.

Specifically, after the baseband signal to be processed is obtained, the target predistortion signal can be obtained by calculation according to the mapping relation between the predistortion signal and the baseband signal in the predistortion model.

S306, calculating an error value of an error function based on the target predistortion signal and the output signal of the current moment.

The target predistortion signal is taken as x (n) in the error function, and the error value of the error function is calculated, and the error function is referred to the related description in the above embodiment, which is not repeated here.

S307, judging whether the error value is larger than a preset threshold value.

If yes, the step S302 is executed again.

It will be appreciated that if the error value is greater than the predetermined threshold, this indicates that the predistortion model has low accuracy, which results in a reduced accuracy of the output of the power amplifier, and therefore requires reconstruction of the predistortion model.

And S308, if not, carrying out iterative computation on the error function by adopting a gradient descent method, and determining an update coefficient vector corresponding to the error function after iterative computation.

If the error value is smaller than or equal to the preset threshold, the accuracy of the predistortion model is within the error allowable range, that is, the predistortion model can be continuously used. However, in order to further improve the accuracy of the output result of the power amplifier, the gradient descent method may be used to iteratively calculate the error value of the error function, optimize the error function, reduce the error value of the error function, determine the coefficient vector after iterative calculation, and use the coefficient vector after iterative calculation as an updated coefficient vector, thereby achieving the purpose of fine tuning the coefficient vector.

Specifically, the iteration number of performing iterative computation on the error function may be preset according to the user requirement.

If the user has lower precision requirements on the predistortion model, the iteration times can be set to be smaller correspondingly, so that the calculation amount of iterative calculation can be reduced. For example, the number of iterations may be set to 1. If the user has higher precision requirement on the predistortion model, the iteration times can be set to be more times correspondingly, so that the precision of the predistortion model can be improved and the service time of the predistortion model can be prolonged.

It should be noted that, the error function is calculated by using the least square method, the calculated amount is large, but the granularity of the calculated result is small, and the accurate solution can be performed, so that the calculated coefficient vector is accurate, and compared with the least square method, the calculated amount of the gradient descent method is small, and correspondingly, the granularity of the calculated result calculated by using the gradient descent method is coarse. In addition, in the operation process of the predistortion model, due to errors in the construction of the predistortion model, the errors are amplified along with the increase of operation times, so that the errors of the predistortion model finally exceed a preset error allowable range and cannot meet the requirement of accuracy. Therefore, during the operation of the predistortion model, a gradient descent method with small calculation amount can be adopted to finely tune the predistortion model.

By adopting the method, in the operation process of the predistortion model, the error value of the error function can be calculated according to the target predistortion signal and the output signal at the current moment every preset time, and the error value can be used for indicating the accuracy of the output result of the power amplifier so as to reflect the accuracy of the predistortion model. When the error value is greater than the preset threshold, it may be determined that the predistortion model is low in accuracy, so that the steps of S302-S303 described above may be performed to reconstruct the predistortion model. When the error value is smaller than the preset threshold, the predistortion model is higher in accuracy, and in order to further improve the accuracy of the model, coefficient vectors in the predistortion model can be updated.

For ease of understanding, embodiments of the present application provide an exemplary schematic diagram of a method for predistortion model construction, as shown in fig. 6, the method includes:

s601, initializing an inverse model.

Wherein the inverse model is used to represent the relation information of the predistortion signal and the output signal or the baseband signal. The inverse model is the inverse model vector matrix in the above embodiment, and the initialization of the inverse model is the process of constructing the inverse model vector matrix, and the specific content refers to the related description in the above embodiment and is not repeated here.

S602, calculating an inverse model coefficient.

Specifically, the inverse model is brought into the error function in the above embodiment, and the coefficient vector in the inverse model, that is, the above inverse model coefficient, is calculated in the case that the error value of the error function is the minimum value.

After the inverse model coefficient is obtained by calculation, updating the inverse model according to the inverse model coefficient to obtain a target inverse model, and then constructing a predistortion model according to the target inverse model.

S603, calculating a predistortion signal.

After the predistortion model is constructed, parameters of the predistortion model can be periodically adjusted to meet the requirement on the accuracy of the predistortion model.

And acquiring a baseband signal to be processed, and calculating to obtain a predistortion signal through the predistortion model and the baseband signal to be processed.

S604, calculating a predistortion error.

Specifically, the method for calculating the predistortion error is a method for calculating an error value of an error function, and specific content refers to the related description in the above embodiment, which is not repeated here.

S605, whether the error is larger than a preset threshold value.

If yes, returning to the step S601, only retaining the history information of the baseband signal and the output signal, and discarding the inverse model constructed in the step S.

If not, S606 is performed.

S606, fine tuning the coefficient vector by adopting a gradient descent method.

After the method is adopted, after the inverse model is initialized, the inverse model simultaneously refers to the baseband signal and the output signal, so that the inverse model can fully apply the information of the baseband signal and the output signal, the predistortion model generated by subsequent construction is more accurate, and the accuracy of the output result of the power amplifier is improved. And then in the running process of the generated predistortion model, the error function can be periodically utilized to monitor the errors of the predistortion signal and the output signal, when the error value is larger than a preset threshold value, the accuracy of the predistortion model is lower, and the normal running of the system is ensured by reconstructing the predistortion model. When the error value is smaller than or equal to a preset threshold value, the accuracy of the predistortion model is indicated to be within the error allowable range, so that the coefficient vector can be finely adjusted through a gradient descent method, and the accuracy of the predistortion model is further improved.

Based on the same conception, the embodiment of the application also provides a device for constructing a predistortion model, as shown in fig. 7, the device comprises:

an obtaining module 701, configured to obtain target mapping relationship information of a first signal and a second signal, where the first signal is an output signal or a baseband signal, and the second signal is a predistortion signal, and the predistortion signal is obtained based on the baseband signal;

The processing module 702 is configured to vectorize information in the target mapping relationship information, and convert the vectorized information to obtain relationship information between the second signal and the first signal, where the relationship information between the second signal and the first signal includes a coefficient vector, a baseband signal including phase information, and an output signal including envelope information;

a construction module 703, configured to construct a predistortion model according to the relationship information of the second signal and the first signal, where the predistortion model includes the relationship information of the baseband signal and the predistortion signal.

In one possible implementation, the processing module 702 is configured to:

In one possible implementation, each vector of nonlinear terms in the inverse model vector matrix includes an absolute value portion and a non-absolute value portion; the processing module 702 is configured to:

In one possible implementation, the apparatus further includes:

the acquiring module 701 is configured to acquire a baseband signal to be processed at intervals of a preset duration;

the calculation module is used for inputting the baseband signal to be processed into the predistortion model to obtain a target predistortion signal;

the calculating module is further used for calculating an error value of the error function based on the target predistortion signal and an output signal at the current moment;

the judging module is used for judging whether the error value is larger than a preset threshold value or not;

If not, carrying out iterative computation on the error function by adopting a gradient descent method through an adjustment module, and determining an update coefficient vector corresponding to the error function after iterative computation;

and the updating module is used for updating the coefficient vector of the predistortion model into the updated coefficient vector.

wherein K is _a 、K _b 、K _c An index set for representing an order of the relationship information of the second signal and the first signal, L _a 、L _b 、L _c Is the index set of memory depth, M _b 、M _c Is an index set expressing the dissynchrony of a signal with an envelope, i represents the time difference of the envelope or phase with the signal, and m represents the time difference of the envelope and the signal.

Fig. 8 shows a schematic hardware structure of an electronic device according to an embodiment of the present application.

A processor 801 and a memory 802 storing computer program instructions may be included in an electronic device.

In particular, the processor 801 may include a Central Processing Unit (CPU), or an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or may be configured to implement one or more integrated circuits of embodiments of the present application.

Memory 802 may include mass storage for data or instructions. By way of example, and not limitation, memory 802 may include a Hard Disk Drive (HDD), floppy Disk Drive, flash memory, optical Disk, magneto-optical Disk, magnetic tape, or universal serial bus (Universal Serial Bus, USB) Drive, or a combination of two or more of the above. Memory 802 may include removable or non-removable (or fixed) media, where appropriate. Memory 802 may be internal or external to the integrated gateway disaster recovery device, where appropriate. In a particular embodiment, the memory 802 is a non-volatile solid-state memory.

The memory may include Read Only Memory (ROM), random Access Memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical/tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software comprising computer-executable instructions and when the software is executed (e.g., by one or more processors) it is operable to perform the operations described with reference to methods in accordance with aspects of the present disclosure.

The processor 801 implements any of the predistortion model construction methods of the above embodiments by reading and executing computer program instructions stored in the memory 802.

In one example, the electronic device may also include a communication interface 803 and a bus 804. As shown in fig. 8, the processor 801, the memory 802, and the communication interface 803 are connected to each other via a bus 804 and perform communication with each other.

The communication interface 803 is mainly used to implement communication between each module, apparatus, unit and/or device in the embodiments of the present application.

Bus 804 includes hardware, software, or both, coupling components of the online data flow billing device to each other. By way of example, and not limitation, the buses may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a micro channel architecture (MCa) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable bus, or a combination of two or more of the above. Bus 804 may include one or more buses, where appropriate. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

It should be clear that the present application is not limited to the particular arrangements and processes described above and illustrated in the drawings. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications, and additions, or change the order between steps, after appreciating the spirit of the present application.

The functional blocks shown in the above-described structural block diagrams may be implemented in hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, the elements of the present application are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine readable medium or transmitted over transmission media or communication links by a data signal carried in a carrier wave. A "machine-readable medium" may include any medium that can store or transfer information. Examples of machine-readable media include electronic circuitry, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio Frequency (RF) links, and the like. The code segments may be downloaded via computer networks such as the internet, intranets, etc.

It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be different from the order in the embodiments, or several steps may be performed simultaneously.

Aspects of the present disclosure are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to being, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware which performs the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In the foregoing, only the specific embodiments of the present application are described, and it will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the systems, modules and units described above may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein. It should be understood that the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, which are intended to be included in the scope of the present application.

Claims

1. A method of predistortion model construction, comprising:

2. The method according to claim 1, wherein the vectorizing and converting information in the target mapping relationship information to obtain relationship information between the second signal and the first signal includes:

3. The method of claim 2, wherein each vector of nonlinear terms in the inverse model vector matrix comprises an absolute value portion and a non-absolute value portion; the replacing the vector in the inverse model vector matrix to obtain a target vector matrix including the coefficient vector, the phase information and the envelope information includes:

4. The method of claim 1, wherein after said constructing a predistortion model from relationship information of said second signal and said first signal, said method further comprises:

Judging whether the error value is larger than a preset threshold value or not;

if not, carrying out iterative computation on the error function by adopting a gradient descent method, and determining an update coefficient vector corresponding to the error function after iterative computation;

5. The method of claim 1, wherein the expression of the relationship information of the second signal and the first signal is as follows:

wherein K is _a 、K _b 、K _c An index set for representing an order of the relationship information of the second signal and the first signal, L _a 、L _b 、L _c Is the index set of memory depth, M _b 、M _c Is a reference to express that the first signal is unsynchronized with the envelopeThe set of labels, i represents the time difference of the envelope or phase and the first signal, and m represents the time difference of the envelope and the first signal.

6. An apparatus for predistortion model construction, the apparatus comprising:

7. The apparatus of claim 6, wherein the processing module is specifically configured to:

8. The apparatus of claim 7, wherein each vector of nonlinear terms in the inverse model vector matrix comprises an absolute value portion and a non-absolute value portion; the processing module is specifically configured to:

9. An electronic device, the device comprising: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, implements a method of predistortion model construction as claimed in any one of claims 1-5.

10. A computer readable storage medium, having stored thereon computer program instructions, which when executed by a processor, implement a method of predistortion model construction according to any of the claims 1-5.