CN118971810A - Predistorter lookup table updating method, predistorter lookup table updating device and predistortion processing system - Google Patents

Abstract

The invention provides a predistorter lookup table updating method, a predistorter lookup table updating device and a predistortion processing system, wherein the method comprises the steps that a processor constructs a linear equation set to obtain a data matrix of the linear equation set, and a solution of the linear equation set is represented through a first intermediate matrix and a second intermediate matrix; collecting an observation value at the beginning of a new preset period; updating the first intermediate matrix and the second intermediate matrix after each acquisition in the current preset period; when the preset acquisition times are reached in the current preset period, calculating a solution of a linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period; and updating the predistorter lookup table according to the model coefficient of the current preset period. The invention can save the calculation resources, disperse intensive matrix calculation, reduce the occurrence probability of task blocking and improve the system performance.

Description

Predistorter lookup table updating method, predistorter lookup table updating device and predistortion processing system

Technical Field

The present invention relates to the field of signal processing technologies, and in particular, to a method and an apparatus for updating a lookup table of a predistorter, and a predistortion processing system.

Background

Digital predistortion is used to improve linearity and efficiency index of a radio frequency power amplifier, and is commonly used in radio frequency transceiver chips. As shown in fig. 1, a transmitting channel of a radio frequency transceiver chip is inserted into a predistorter, an input signal u (n) input to the radio frequency transceiver chip is predistorted, and an output signal x (n) of the predistorter is output to a power amplifier PA through digital-to-analog conversion and the like, and the power amplifier PA outputs a signal y (n). The lookup table is a pre-distortion processing method commonly used in a pre-distorter, and the pre-distorter lookup table corrects the non-linear behavior of the power amplifier by storing a series of compensation coefficients, so that the purpose of compensating the non-linearity of the power amplifier is achieved, and the distortion from the signal x (n) to the signal y (n) is reduced.

The compensation coefficient stored in the predistorter lookup table is closely related to the model coefficient of the predistortion model used by the predistorter, and in order to ensure the stability and the precision of digital predistortion, the predistortion model needs to be updated in real time according to the change of environment and working state, the model coefficient is changed continuously, and the compensation coefficient stored in the predistorter lookup table also needs to be updated.

Currently, for nonlinear predistortion models, adaptive algorithms are generally used to determine and optimize model coefficients, including least squares algorithms, recursive least squares algorithms, least mean squares algorithms, and the like. The least square algorithm has high solving precision and is widely applied to the field of digital predistortion. The key steps of the least square method solving include matrix inversion and matrix multiplication operation, and the size of the data matrix is determined by the product of the number of the characteristic values and the length of the observed value. However, if the data matrix is large in size, a large amount of CPU memory will be consumed, and on the other hand, dense matrix operation is a challenge for the embedded system of the radio frequency transceiver chip, which is easy to cause the situation of system task blocking.

Disclosure of Invention

Aiming at the defects existing in the prior art, the invention provides a predistorter lookup table updating method, which solves the problems that the model coefficient of a predistortion model is determined and optimized through a least square method in the prior art, namely, the calculation resource consumption required for updating the predistorter lookup table is large, and the performance of an embedded system using a radio frequency transceiver chip is influenced.

According to a first aspect of embodiments of the present invention, there is provided a predistorter lookup table updating method applied to a predistortion processing system including a processor, a predistorter, a power amplifier, the predistorter lookup table updating method including:

the method comprises the steps that when each preset period starts, the processor acquires a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the previous preset period, builds a linear equation set, acquires a data matrix of the linear equation set, and represents a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

collecting an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

After each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector;

calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observed value;

Updating the first intermediate matrix and the second intermediate matrix by collecting the calculated first sub-matrix and second sub-matrix at the present time;

When the preset acquisition times are reached in the current preset period, calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period;

And updating the predistorter lookup table according to the model coefficient of the current preset period, wherein the updated predistorter lookup table obtained in the current preset period is used for indicating the predistorter to perform predistortion processing on an input signal input to the predistortion processing system in a future preset period.

Optionally, calculating the first sub-matrix and the second sub-matrix according to the intermediate feature vector of the current acquisition and the observed value of the current acquisition includes:

Extracting a calculation part comprising the intermediate feature vector from a first intermediate matrix to obtain a first submatrix; and extracting a calculation part comprising the intermediate feature vector and the observation value acquired at the current time from a second intermediate matrix to obtain a second submatrix.

Optionally, updating the first intermediate matrix and the second intermediate matrix by the first sub-matrix and the second sub-matrix calculated by the intermediate eigenvector acquired at the time, where the formula is:

Where Q is a first intermediate matrix, Q _i is a first sub-matrix calculated for the current acquisition, U is a second intermediate matrix, and U _i is a second sub-matrix calculated for the current acquisition.

Optionally, before calculating the feature vector corresponding to the observation value of the current acquisition in the data matrix according to the observation value of the current acquisition, the method includes:

and (5) performing delay, phase and gain alignment processing on the observation value acquired at the present time.

Optionally, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value, and after obtaining the current acquired intermediate feature vector, including:

the intermediate feature vector acquired at the time is written into the hardware accelerator in a DMA mode.

Optionally, when the preset acquisition times are reached, solving the linear equation set with the last updated first intermediate matrix and second intermediate matrix to obtain the model coefficient includes:

Invoking the first intermediate matrix and the second intermediate matrix which are calculated and updated according to the last written intermediate feature vector in the hardware accelerator to obtain the last updated first intermediate matrix and the last updated second intermediate matrix;

And solving the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated finally to obtain the model coefficient of the current signal acquisition period.

Optionally, updating the predistorter lookup table according to the model coefficient of the current signal acquisition period includes:

decomposing the model coefficient of the current signal acquisition period by using Cholesky to obtain a compensation coefficient;

updating a predistorter lookup table by the compensation coefficient.

A second aspect provides a predistorter look-up table updating apparatus for use in a predistortion processing system comprising a processor, a predistorter, a power amplifier, the predistorter look-up table updating apparatus comprising:

The model splitting module is used for acquiring a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the last preset period when each preset period starts, constructing a linear equation set, acquiring a data matrix of the linear equation set, and expressing a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

the observation value acquisition module is used for acquiring an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

The intermediate feature vector calculation module is used for calculating the feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value after each acquisition in the current preset period to obtain the current acquired intermediate feature vector;

the submatrix calculation module is used for calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observation value;

The intermediate matrix updating module is used for updating the first intermediate matrix and the second intermediate matrix through the first sub-matrix and the second sub-matrix which are calculated by current acquisition;

the model coefficient calculation module is used for calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period when the preset acquisition times are reached in the current preset period;

And the predistorter lookup table updating module is used for updating the predistorter lookup table according to the model coefficient of the current preset period, wherein the updated predistorter lookup table obtained in the current preset period is used for indicating the predistorter to perform predistortion processing on an input signal input into the predistortion processing system in the future preset period.

A third aspect provides a predistortion processing system, comprising:

The processor is used for acquiring a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the last preset period when each preset period starts, constructing a linear equation set, acquiring a data matrix of the linear equation set, and expressing the solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

collecting an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

After each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector;

calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observed value;

Updating the first intermediate matrix and the second intermediate matrix by collecting the calculated first sub-matrix and second sub-matrix at the present time;

When the preset acquisition times are reached in the current preset period, calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period;

updating a predistorter lookup table according to the model coefficient of the current preset period;

The predistorter instructs the predistorter to perform predistortion processing on a target input signal input into the predistortion processing system in a future preset period according to an updated predistorter lookup table obtained in the current preset period, and outputs a target predistortion signal;

a power amplifier for receiving the target predistortion signal and amplifying the target predistortion signal

The technical principle of the invention is as follows: firstly, constructing a linear equation set aiming at a predistortion model used by a predistorter, wherein the linear equation set comprises a predistorter output signal, a power amplifier output signal and model coefficients of the predistortion model, wherein the predistorter output signal and the power amplifier output signal are observation values acquired in each preset period, and the model coefficients of the predistortion model are solutions of the linear equation set. The present invention also calculates a solution of a system of linear equations using a least squares method, and therefore the solution of the system of linear equations is represented by a first intermediate matrix and a second intermediate matrix. However, in the solving process, each time an observed value is acquired, the first intermediate matrix and the second intermediate matrix are updated once, which is equivalent to updating part of the feature matrix, so as to obtain part of the calculation result of the first intermediate matrix and part of the calculation result of the second intermediate matrix, and when the preset acquisition times are reached in the current preset period, the finally updated first intermediate matrix and second intermediate matrix are the same as the first intermediate matrix and second intermediate matrix used in the traditional least square method calculation. And finally, the updated first intermediate matrix and second intermediate matrix can solve the linear equation set to obtain the model coefficient of the current signal acquisition period, and then the predistorter lookup table can be updated according to the model coefficient of the current preset period.

Compared with the prior art, the invention has the following beneficial effects: according to the predistorter lookup table updating method provided by the embodiment of the invention, after the observation value is collected once, the feature vector is calculated, the first intermediate matrix and the second intermediate matrix which are used for solving the model coefficient by the original least square method are split according to the observation value collected at present and the calculated intermediate feature vector, namely the first submatrix and the second submatrix are calculated, the first intermediate matrix and the second intermediate matrix are gradually updated through the first submatrix and the second submatrix, and finally the updated first intermediate matrix and the updated second intermediate matrix are equivalent to the first intermediate matrix and the second intermediate matrix which are used for solving the model coefficient by the original least square method.

Drawings

Fig. 1 is a schematic diagram of a component structure of a radio frequency transceiver chip;

FIG. 2 is a schematic diagram of an implementation flow of a predistorter lookup table update method according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of the composition structure of a predistortion processing system according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of the algorithm logic for solving model coefficients by the conventional least squares method;

FIG. 5 is a schematic diagram of algorithm logic for calculating model coefficients in a predistorter lookup table update method according to an embodiment of the present invention;

Fig. 6 is a schematic diagram of a composition structure of a predistorter lookup table updating method device according to an embodiment of the present invention.

Detailed Description

The technical scheme of the invention is further described below with reference to the accompanying drawings and examples.

Referring to fig. 2 and 3, as shown in fig. 2, an embodiment of the present invention provides a predistorter lookup table updating method applied to a predistortion processing system 30 shown in fig. 3, where the predistortion processing system 30 includes a processor 31, a predistorter 32 and a power amplifier 33. According to the embodiment of the invention, the step of solving the model coefficient by the processor optimizing least square method algorithm is adopted, so that the consumption of computing resources is reduced, intensive matrix operation is dispersed, and the risk of blocking of system tasks can be avoided.

It will be appreciated that the predistortion processing system in the embodiment of the present invention is not limited to the above-described processor, predistorter and power amplifier, but also includes other functional devices such as a digital-to-analog converter and a low pass filter in the rf transceiver chip shown in fig. 1.

Referring to fig. 2, the method for updating a predistorter lookup table according to an embodiment of the present invention includes, but is not limited to, the following steps:

s1, when each preset period starts, the processor acquires a predistorter output signal, a power amplifier output signal and model coefficients of a predistortion model in the previous preset period, constructs a linear equation set, acquires a data matrix of the linear equation set, and expresses a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix.

In the embodiment of the invention, the preset period is set according to the input signal of the input predistortion processing system.

It can be understood that if the preset period in the step S1 is the first preset period, the predistorter output signal, the power amplifier output signal, and the model coefficients of the predistortion model are directly recorded.

Assuming that the predistorter uses the GMP DPD model, the GMP DPD model is expressed as:

Where x (n) represents the predistorter output signal, y (n) represents the power amplifier output signal, i represents the memory term, j represents the cross memory term, k represents the nonlinear order, and c _ijk represents the model coefficients.

Based on GMPDPD model, and step S1 above, the following set of linear equations can be constructed:

Wherein F _ij is a feature vector satisfying F _ij＝f(x_i,c_j), and F _X is used to represent the feature matrix of the data matrix of the above formula (2), then there are:

Y＝F_Xc(3)

wherein,

In the above equations (1) to (4), the predistorter output signal and the power amplifier output signal are obtained by observation, and thus, considering x _i and y _i as known numbers, the solution of the above linear equation set (2) represents the model coefficient, i.e., c in the above equation (3).

The key steps of solving model coefficients by the traditional least square method comprise matrix inversion and matrix multiplication operation, and the embodiment of the invention also calculates the solution of the linear equation set by the matrix inversion and the matrix multiplication operation, namely, the solution of the linear equation set is represented by a first intermediate matrix and a second intermediate matrix. Let the first intermediate matrix be Q, the second intermediate matrix be U, the model coefficient be c, the mathematical expressions of the first intermediate matrix and the second intermediate matrix be:

Q＝F_X ^HF_X(5)；

U＝F_X ^HY(6)；

the mathematical expression of the least square method solving model coefficient is:

c＝Q^-1U(7)。

Wherein F _X represents the feature matrix of the data matrix of the above formula (2), and Y represents Y ₁,...,y_i,...,y_n of the above formula (2).

According to the predistorter lookup table updating method provided by the embodiment of the invention, the process of solving the model coefficient by using the least square method is optimized through the following steps S21 to S26.

S21, collecting an observation value when a new preset period starts.

In the step S21, the observed values are collected for multiple times in each preset period until the preset collection times are reached. In the embodiment of the invention, the observation value acquired at the time comprises the predistorter output signal acquired at the time and the power amplifier output signal acquired at the time.

And S22, after each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector.

In the above, F _ij represents the eigenvector of the data matrix based on the linear equation set (2), F _ij＝f(x_i,c_j. It is assumed that in step S22 described above, when the acquired observations comprise a number a of predistorter input signals x _i (i=1, 2,3, a.), a feature vector F _ij is obtained (i=1, 2, 3), a) when the intermediate eigenvectors of the sub-acquisition include a eigenvectors F _ij, (i=1, 2,3,., a), a part of the eigenvector F _X of the data matrix belonging to the above formula (2).

S23, calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observation value.

In the step S23, the first sub-matrix is a part of the first intermediate matrix, the second sub-matrix is a part of the second intermediate matrix, and the implementation steps of the step S23 are as follows:

Extracting a calculation part comprising the intermediate feature vector from a first intermediate matrix to obtain a first submatrix; and extracting a calculation part comprising the intermediate feature vector and the observation value acquired at the current time from a second intermediate matrix to obtain a second submatrix.

Preferably, the second intermediate matrix includes the intermediate eigenvectors and the calculation part of the output signal of the power amplifier acquired at the present time, so as to obtain a second sub-matrix.

S24, updating the first intermediate matrix and the second intermediate matrix through the first sub-matrix and the second sub-matrix calculated by current acquisition.

The above step S24 is expressed as:

Where Q is a first intermediate matrix, Q _i is a first sub-matrix calculated for the current acquisition, U is a second intermediate matrix, and U _i is a second sub-matrix calculated for the current acquisition.

The embodiment of the invention also describes the relationship between the first sub-matrix and the first intermediate matrix and the relationship between the second sub-matrix and the second intermediate matrix through detailed data.

Illustratively, developing equation (5) yields

Expanding the formula (6) to obtain

Order the

Obtaining

Order the

Obtaining

In practical application, the conventional square method solves the model coefficient and needs to calculate the eigenvector through all the observed values obtained in the period to obtain the complete eigenvector, and through the step S23 and the step S24, the embodiment of the invention solves the first submatrix and the second submatrix for each group of observed values and updates the first intermediate matrix and the second intermediate matrix, thereby obviously reducing the consumption of the system memory in the solving process, and dispersing the intensive matrix operation on the time axis, so as to avoid the risk of system task blocking.

It will be appreciated that, if it is the first acquisition, the first intermediate matrix updated by the current acquired observation and the current acquired intermediate eigenvector is equal to the first sub-matrix and the second intermediate matrix updated by the current acquired observation and the current acquired intermediate eigenvector is equal to the second sub-matrix.

In one embodiment, the steps S23 and S24 may be implemented by a hardware accelerator, and the step S22 calculates, according to the currently acquired observation value, a feature vector corresponding to the currently acquired observation value in the data matrix, to obtain a currently acquired intermediate feature vector, and then includes:

The intermediate feature vector of the current acquisition is written to the hardware accelerator by way of DMA (Direct Memory Access ).

And S25, when the preset acquisition times are reached in the current preset period, calculating the solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain the model coefficient of the current preset period.

In the above steps S21 to S24, each time the observed value is acquired, the first intermediate matrix and the second intermediate matrix are updated once, which is equivalent to updating part of the feature matrix, and obtaining part of the calculation result of the first intermediate matrix and part of the calculation result of the second intermediate matrix, so when the preset acquisition times are reached in the current preset period, the finally updated first intermediate matrix and second intermediate matrix are the same as the first intermediate matrix and second intermediate matrix used in the traditional least square method calculation.

After the step S22, the updating of the first intermediate matrix and the updating of the second intermediate matrix in the step S23 and the step S24 are implemented in the hardware accelerator, when the preset acquisition times are reached, the hardware accelerator outputs the last updated first intermediate matrix and the last updated second intermediate matrix, and when the preset acquisition times are reached, the linear equation set is solved by the last updated first intermediate matrix and the last updated second intermediate matrix to obtain the model coefficient, including:

Invoking the first intermediate matrix and the second intermediate matrix which are calculated and updated according to the last written intermediate feature vector in the hardware accelerator to obtain the last updated first intermediate matrix and the last updated second intermediate matrix;

And solving the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated finally to obtain the model coefficient of the current signal acquisition period.

S26, updating a predistorter lookup table according to the model coefficient of the current preset period.

The updated predistorter lookup table obtained in the current preset period is used for indicating the predistorter to perform predistortion processing on an input signal input to the predistortion processing system in the future preset period.

It should be noted that, the model coefficient is not equal to the compensation coefficient of the predistorter lookup table, and the embodiment of the invention uses Cholesky to decompose the model coefficient of the current signal acquisition period to obtain the compensation coefficient; updating a predistorter lookup table by the compensation coefficient.

In one embodiment, before updating the predistorter lookup table by the compensation coefficient, it comprises:

verifying a model coefficient of a current signal acquisition period;

the validated model coefficients are stored to a predistorter lookup table.

Exemplary, validating model coefficients includes:

calculating decision coefficients of the predistortion model using the model coefficients;

and when the determined coefficient is in a preset range, the model coefficient passes verification.

The embodiment of the invention also compares the provided predistorter lookup table updating method with the traditional method for solving the model coefficient by the least square method. As shown in fig. 4, a logic diagram of a predistorter lookup table updating method according to an embodiment of the present invention shows steps of solving model coefficients in a preset period, i.e. steps S21 to S26 described above. Fig. 5 is a logic diagram of solving model coefficients by using a conventional least square method, and also shows a step of solving model coefficients in one period. In fig. 5, after each time of collecting the observed values, the feature matrix is updated, and if the observed values are collected N times in the period, the feature matrix is updated N times, and finally a feature matrix with a larger magnitude is obtained, and then the calculation of the first intermediate matrix and the second intermediate matrix is performed according to the feature matrix with the larger magnitude, so as to calculate the model coefficient. As shown in fig. 4, in the updating method of the predistorter lookup table provided by the embodiment of the present invention, after the observation value is collected once, the eigenvector is calculated, and the first intermediate matrix and the second intermediate matrix used for solving the model coefficient by the original least square method are split according to the observation value collected at the present time and the calculated intermediate eigenvector, that is, the first sub-matrix and the second sub-matrix are calculated, the first intermediate matrix and the second intermediate matrix are gradually updated through the first sub-matrix and the second sub-matrix, and finally the updated first intermediate matrix and the second intermediate matrix are equivalent to the first intermediate matrix and the second intermediate matrix used for solving the model coefficient by the original least square method.

As shown in table 1 below, the computational resources required for the conventional method of solving model coefficients by the least squares method shown in fig. 5.

It can be seen that dense matrix operations place high demands on the computational power of the CPU. Meanwhile, other real-time tasks are required to be considered simultaneously by the embedded system, so that the task blockage of the system can be caused by intensive matrix operation required by the traditional least square method solution. According to fig. 2 and fig. 4, in the method for updating the predistorter lookup table provided by the embodiment of the present invention, when solving the model coefficient, if the predistortion model is fixed (the number of feature quantities is not changed), the required memory will not increase with the increase of the length of the observed data, and the feature vector is calculated only by using the currently collected observed value, and only the line of intermediate feature vector corresponding to the current observed value needs to be saved in the feature matrix all the time, for example, in the method shown in fig. 2 and fig. 4, the memory allocation of the feature matrix changes as shown in table 2 below:

Memory requirements	Before optimization	After optimization
			Feature matrix F X	M×N	M×1

Taking m=96 and n=32k as an example, the memory required for the feature matrix F after optimization will be reduced by 3040k (98.9%).

As shown in fig. 6, the embodiment of the present invention further provides a predistorter lookup table updating device 60, which can be applied to the predistortion processing system 30 including the processor 31, the predistorter 32 and the power amplifier 33 as shown in fig. 3. In a preferred implementation, predistorter lookup table update means 60 is a virtual device of processor 31 comprising:

The model splitting module 61 is configured to obtain, at the beginning of each preset period, a predistorter output signal, a power amplifier output signal, and a model coefficient of a predistortion model in a previous preset period, and construct a linear equation set, obtain a data matrix of the linear equation set, and represent a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

An observation value acquisition module 62, configured to acquire an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

the intermediate feature vector calculation module 63 is configured to calculate, after each acquisition in a current preset period, a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value, so as to obtain a current acquired intermediate feature vector;

a sub-matrix calculation module 64, configured to calculate a first sub-matrix and a second sub-matrix according to the currently acquired intermediate feature vector and the currently acquired observation value;

An intermediate matrix updating module 65, configured to update the first intermediate matrix and the second intermediate matrix by acquiring the calculated first sub-matrix and second sub-matrix at the present time;

the model coefficient calculation module 66 is configured to calculate a solution of the linear equation set with the last updated first intermediate matrix and the second intermediate matrix to obtain a model coefficient of the current preset period when the preset acquisition times reach in the current preset period;

the predistorter lookup table updating module 67 is configured to update the predistorter lookup table according to a model coefficient of a current preset period, where the updated predistorter lookup table obtained in the current preset period is used to instruct the predistorter to perform predistortion processing on an input signal input to the predistortion processing system in a future preset period.

The embodiment of the present invention further describes in detail the predistortion processing system 30 shown in fig. 3, the predistortion processing system 30 comprises a processor 31, a predistorter 32 and a power amplifier 33, wherein:

A processor 31, configured to obtain, at the beginning of each preset period, a predistorter output signal x (n), a power amplifier output signal y (n), and model coefficients of a predistortion model in a previous preset period, and construct a linear equation set, obtain a data matrix of the linear equation set, and represent a solution of the linear equation set by using a first intermediate matrix and a second intermediate matrix;

collecting an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

After each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector;

calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observed value;

Updating the first intermediate matrix and the second intermediate matrix by collecting the calculated first sub-matrix and second sub-matrix at the present time;

When the preset acquisition times are reached in the current preset period, calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period;

updating a predistorter lookup table according to the model coefficient of the current preset period;

A predistorter 32, which instructs the predistorter to perform predistortion processing on a target input signal u (n) input to the predistortion processing system in a future preset period according to an updated predistorter lookup table obtained in the current preset period, and outputs a target predistortion signal x (n);

And a power amplifier 33 for receiving the target predistortion signal x (n), amplifying the target predistortion signal, and outputting a signal y (n).

It should be noted that, the predistortion processing system provided in the embodiment of the present invention is an embedded system, and may be applied to different signal processing requirements, such as signal transmission in an intelligent communication system, audio signal processing in an audio device, etc., where the application of the embodiment of the present invention is not limited.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The foregoing examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.

Finally, it is noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered by the scope of the claims of the present invention.

Claims (9)

1. A predistorter look-up table updating method for a predistortion processing system comprising a processor, a predistorter, a power amplifier, the predistorter look-up table updating method comprising:

the method comprises the steps that when each preset period starts, the processor acquires a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the previous preset period, builds a linear equation set, acquires a data matrix of the linear equation set, and represents a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

collecting an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

After each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector;

calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observed value;

Updating the first intermediate matrix and the second intermediate matrix by collecting the calculated first sub-matrix and second sub-matrix at the present time;

When the preset acquisition times are reached in the current preset period, calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period;

And updating the predistorter lookup table according to the model coefficient of the current preset period, wherein the updated predistorter lookup table obtained in the current preset period is used for indicating the predistorter to perform predistortion processing on an input signal input to the predistortion processing system in a future preset period.

2. The predistorter lookup table updating method of claim 1, wherein calculating a first sub-matrix and a second sub-matrix from the currently acquired intermediate feature vector and the currently acquired observation comprises:

Extracting a calculation part comprising the intermediate feature vector from a first intermediate matrix to obtain a first submatrix; and extracting a calculation part comprising the intermediate feature vector and the observation value acquired at the current time from a second intermediate matrix to obtain a second submatrix.

3. The predistorter lookup table updating method of claim 2, wherein the first intermediate matrix and the second intermediate matrix are updated by a first sub-matrix and a second sub-matrix calculated from the intermediate feature vector acquired at the time, as:

Where Q is a first intermediate matrix, Q _i is a first sub-matrix calculated for the current acquisition, U is a second intermediate matrix, and U _i is a second sub-matrix calculated for the current acquisition.

4. A predistorter lookup table updating method as claimed in any one of claims 1 to 3, comprising, before calculating a feature vector in the data matrix corresponding to a current acquired observation from the current acquired observation:

and (5) performing delay, phase and gain alignment processing on the observation value acquired at the present time.

5. A predistorter lookup table updating method as claimed in any one of claims 1 to 3, comprising, after computing feature vectors in the data matrix corresponding to the current acquired observations from the current acquired observations, obtaining the current acquired intermediate feature vectors:

the intermediate feature vector acquired at the time is written into the hardware accelerator in a DMA mode.

6. The method of updating a predistorter lookup table as claimed in claim 5, wherein solving said set of linear equations with a last updated first intermediate matrix and second intermediate matrix to obtain said model coefficients when a preset number of acquisitions is reached, comprising:

Invoking the first intermediate matrix and the second intermediate matrix which are calculated and updated according to the last written intermediate feature vector in the hardware accelerator to obtain the last updated first intermediate matrix and the last updated second intermediate matrix;

And solving the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated finally to obtain the model coefficient of the current signal acquisition period.

7. The predistorter lookup table updating method of claim 1, wherein updating the predistorter lookup table based on model coefficients of a current signal acquisition cycle comprises:

decomposing the model coefficient of the current signal acquisition period by using Cholesky to obtain a compensation coefficient;

updating a predistorter lookup table by the compensation coefficient.

8. A predistorter look-up table updating apparatus for use in a predistortion processing system comprising a processor, a predistorter, a power amplifier, said predistorter look-up table updating apparatus comprising:

The model splitting module is used for acquiring a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the last preset period when each preset period starts, constructing a linear equation set, acquiring a data matrix of the linear equation set, and expressing a solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

the observation value acquisition module is used for acquiring an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

The intermediate feature vector calculation module is used for calculating the feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value after each acquisition in the current preset period to obtain the current acquired intermediate feature vector;

the submatrix calculation module is used for calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observation value;

The intermediate matrix updating module is used for updating the first intermediate matrix and the second intermediate matrix through the first sub-matrix and the second sub-matrix which are calculated by current acquisition;

the model coefficient calculation module is used for calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period when the preset acquisition times are reached in the current preset period;

And the predistorter lookup table updating module is used for updating the predistorter lookup table according to the model coefficient of the current preset period, wherein the updated predistorter lookup table obtained in the current preset period is used for indicating the predistorter to perform predistortion processing on an input signal input into the predistortion processing system in the future preset period.

9. A predistortion processing system, comprising:

The processor is used for acquiring a predistorter output signal, a power amplifier output signal and a model coefficient of a predistortion model in the last preset period when each preset period starts, constructing a linear equation set, acquiring a data matrix of the linear equation set, and expressing the solution of the linear equation set through a first intermediate matrix and a second intermediate matrix;

collecting an observation value when a new preset period starts; wherein the observation value of the current acquisition comprises a predistorter output signal of the current acquisition and a power amplifier output signal of the current acquisition;

After each acquisition in the current preset period, calculating a feature vector corresponding to the current acquired observation value in the data matrix according to the current acquired observation value to obtain a current acquired intermediate feature vector;

calculating a first submatrix and a second submatrix according to the currently acquired intermediate feature vector and the currently acquired observed value;

Updating the first intermediate matrix and the second intermediate matrix by collecting the calculated first sub-matrix and second sub-matrix at the present time;

When the preset acquisition times are reached in the current preset period, calculating a solution of the linear equation set by using the first intermediate matrix and the second intermediate matrix which are updated last to obtain a model coefficient of the current preset period;

updating a predistorter lookup table according to the model coefficient of the current preset period;

The predistorter instructs the predistorter to perform predistortion processing on a target input signal input into the predistortion processing system in a future preset period according to an updated predistorter lookup table obtained in the current preset period, and outputs a target predistortion signal;

And the power amplifier is used for receiving the target predistortion signal and performing power amplification on the target predistortion signal.