Hybrid learning structure digital pre-distortion method, device, equipment, medium and product

Abstract

The invention discloses a mixed learning structure digital pre-distortion method, device, equipment, medium and product, and the method comprises the steps: carrying out the cascading of pre-distortion signals obtained through the calculation of a power amplifier and a forward pre-distorter, and obtaining a power amplification signal; monitoring the power amplification signal, and calculating the signal-to-noise ratio of the power amplification signal; determining a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal; and pre-distortion parameter estimation is carried out according to the determined parameter learning link, and pre-distortion parameter updating is carried out according to an estimation result. According to the scheme, flexible loading and application of the parameter learning structure can be realized, and optimal convergence of digital pre-distortion is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital communication, in particular to a hybrid learning structure digital predistortion method, device, equipment, medium and product. BACKGROUND

[0002] The power amplifier is an important component of the transmitter system, which amplifies the input signal to ensure that the transmitted signal can reach the required transmission power. Due to the non-linear effect of the power amplifier, the signal passing through the power amplifier produces in-band and out-of-band distortion, which increases the bit error rate and causes interference to the adjacent channel, thereby reducing the performance of the communication system. Therefore, it is necessary to use a power amplifier with high linearity to reduce the impact on the performance of the communication system. Among all linearization technologies, digital predistortion technology is gradually replacing various analog linearization technologies and has become a key technology for commercial application of wireless communication.

[0003] At present, the digital predistortion technology generally adopts an indirect learning structure or a direct learning structure, but both the indirect learning structure and the direct learning structure have certain defects, for example: the direct learning structure has complex calculation and slow convergence speed, and the convergence of the indirect learning structure is easily affected by noise and has low convergence accuracy. The existing technical solutions cannot realize flexible loading and use of the parameter learning structure, and the digital predistortion cannot achieve optimal convergence. SUMMARY

[0004] In order to solve the above problems, the present application provides a hybrid learning structure digital predistortion method, device, equipment, medium and product, which can realize flexible loading and use of the parameter learning structure and achieve optimal convergence of the digital predistortion.

[0005] The present application provides a hybrid learning structure digital predistortion method, which comprises the following steps:

[0006] The predistortion signal calculated by the power amplifier and the forward predistorter is cascaded to obtain a power amplification signal;

[0007] The power amplification signal is monitored, and the signal-to-noise ratio of the power amplification signal is calculated;

[0008] According to the signal-to-noise ratio and the power amplification signal, a parameter learning link is determined in a preset parameter learning structure;

[0009] According to the determined parameter learning link, the predistortion parameter estimation is performed, and the predistortion parameter is updated according to the estimation result.

[0010] Preferably, according to the signal-to-noise ratio and the power amplification signal, a parameter learning link is determined in a preset parameter learning structure, which comprises the following steps:

[0011] According to the signal-to-noise ratio and the power amplification signal, a stability value is calculated;

[0012] According to the size of the calculated stability value, a corresponding parameter learning link in the parameter learning structure is matched.

[0013] Preferably, the parameter learning structure includes a direct learning structure and an indirect learning structure.

[0014] Preferably, the calculation of the stability value according to the signal-to-noise ratio and the power amplification signal includes:

[0015] The difference between the ideal output signal of the power amplifier and the power amplification signal is calculated;

[0016] According to the difference and the signal-to-noise ratio, the stability value is calculated;

[0017] Wherein, the stability value t = w1t1 + w2t2, t1 is the difference, t2 is the reciprocal of the signal-to-noise ratio, w1 is a preset difference proportion coefficient, and w2 is a preset signal-to-noise ratio proportion coefficient.

[0018] Preferably, according to the size of the calculated stability value, a corresponding parameter learning link in the parameter learning structure is matched, including:

[0019] When the stability value is greater than a preset threshold, the direct learning structure in the parameter learning structure is determined as the parameter learning link;

[0020] When the stability value is not greater than the threshold, the indirect learning structure in the parameter learning structure is determined as the parameter learning link.

[0021] Preferably, according to the determined parameter learning link, pre-distortion parameter estimation is performed, and pre-distortion parameter updating is performed according to the estimation result, including:

[0022] When the parameter learning link is a direct learning structure, the power amplification signal and the baseband signal input to the forward pre-distortion device are collected for parameter estimation;

[0023] The parameters of the forward pre-distortion device are updated according to the solved parameters.

[0024] Preferably, according to the determined parameter learning link, pre-distortion parameter estimation is performed, and pre-distortion parameter updating is performed according to the estimation result, including:

[0025] When the parameter learning link is an indirect learning structure, the power amplification signal is collected and input to the backward pre-distortion device for pre-distortion calculation to obtain a backward output signal;

[0026] performing parameter estimation on the backward output signal and the pre-distortion signal;

[0027] updating parameters of the backward pre-distortor and the forward pre-distortor according to the solved parameters.

[0028] Preferably, the pre-distortion parameter estimation according to the determined parameter learning link comprises:

[0029] performing minimum adaptive estimation on the two signals calculated in the parameter estimation by using an estimation function to obtain a parameter estimation result;

[0030] The estimation function is e(n) is an error signal, and Z(n) are the two signals calculated in the parameter estimation.

[0031] The embodiment of the present application also provides a hybrid learning structure digital pre-distortion device, which comprises:

[0032] a pre-distortion module, configured to cascade a pre-distortion signal calculated by a power amplifier and a forward pre-distortor to obtain a power amplification signal;

[0033] a signal-to-noise ratio calculation module, configured to monitor the power amplification signal and calculate a signal-to-noise ratio of the power amplification signal;

[0034] a link determination module, configured to determine a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal;

[0035] a parameter updating module, configured to perform pre-distortion parameter estimation according to the determined parameter learning link and to perform pre-distortion parameter updating according to an estimation result.

[0036] Preferably, the link determination module is configured to:

[0037] calculate a stability value according to the signal-to-noise ratio and the power amplification signal;

[0038] match a corresponding parameter learning link in the parameter learning structure according to the size of the calculated stability value.

[0039] As a preferred solution, the parameter learning structure comprises a direct learning structure and an indirect learning structure.

[0040] Further, the link determination module is configured to:

[0041] calculate a difference between an ideal output signal of the power amplifier and the power amplification signal;

[0042] calculate the stability value according to the difference and the signal-to-noise ratio.

[0043] The stability value t is t1*w1+t2*w2, t1 is the difference value, t2 is the inverse of the signal-to-noise ratio, w1 is a preset difference value proportion coefficient, and w2 is a preset signal-to-noise ratio proportion coefficient.

[0044] Further, the link determination module is configured to:

[0045] When the stability value is greater than a preset threshold value, a direct learning structure in the parameter learning structure is determined as the parameter learning link.

[0046] When the stability value is not greater than the threshold value, an indirect learning structure in the parameter learning structure is determined as the parameter learning link.

[0047] Preferably, the parameter updating module is configured to:

[0048] When the parameter learning link is the direct learning structure, the power amplification signal and the baseband signal input to the forward pre-distorter are collected for parameter estimation.

[0049] The parameters of the forward pre-distorter are updated according to the solved parameters.

[0050] Preferably, the parameter updating module is configured to:

[0051] When the parameter learning link is the indirect learning structure, the power amplification signal is collected and input to the backward pre-distorter for pre-distortion calculation to obtain a backward output signal.

[0052] The backward output signal and the pre-distortion signal are subjected to parameter estimation.

[0053] The parameters of the backward pre-distorter and the forward pre-distorter are updated according to the solved parameters.

[0054] Preferably, the parameter updating module is configured to:

[0055] The two signals subjected to parameter estimation calculation are subjected to minimum adaptive estimation by using an estimation function to obtain a parameter estimation result.

[0056] The estimation function is e(n) is an error signal, and Z(n) are the two signals subjected to parameter estimation calculation.

[0057] The embodiment of the application also provides a terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the hybrid learning structure digital pre-distortion method according to any one of the above embodiments when executing the computer program.

[0058] The embodiment of the present application also provides a computer readable storage medium, which comprises a stored computer program, wherein the computer readable storage medium controls a device where the computer readable storage medium is located to execute the hybrid learning structure digital pre-distortion method according to any one of the above embodiments when the computer program is executed.

[0059] The embodiment of the present application also provides a computer program product, which comprises computer programs / instructions, and the computer programs / instructions are executed by a processor to implement the steps of any one of the above methods.

[0060] The present application provides a hybrid learning structure digital pre-distortion method, device, equipment, medium and product, a power amplifier and a forward pre-distortion device are cascaded to obtain a power amplification signal; the power amplification signal is monitored, and a signal-to-noise ratio of the power amplification signal is calculated; according to the signal-to-noise ratio and the power amplification signal, a parameter learning link is determined in a preset parameter learning structure; pre-distortion parameter estimation is performed according to the determined parameter learning link, and pre-distortion parameter updating is performed according to the estimation result. The present application can realize flexible loading and use of the parameter learning structure, and realize optimal convergence of digital pre-distortion. BRIEF DESCRIPTION OF DRAWINGS

[0061] Figure 1 is a flow diagram of a hybrid learning structure digital pre-distortion method provided by the embodiment of the present application;

[0062] Figure 2 is a principle diagram of a hybrid learning structure digital pre-distortion method provided by the embodiment of the present application;

[0063] Figure 3 is another principle diagram of a hybrid learning structure digital pre-distortion method provided by the embodiment of the present application;

[0064] Figure 4 is a structure diagram of a hybrid learning structure digital pre-distortion device provided by the embodiment of the present application;

[0065] Figure 5 is a structure diagram of a terminal device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0067] In the process of mixed learning structure digital pre-distortion, the existing digital pre-distortion parameter calculation generally adopts an indirect learning structure or a direct learning structure. The working principle is to make the baseband signal pre-produce an anti-distortion inversely proportional to the power amplifier characteristics, so that the cascade of digital pre-distortion and feedback response can achieve the expected linear response. The digital pre-distortion technology corrects the pre-distortion of the transmission signal by collecting the in-band and out-of-band output data of the feedback loop with nonlinear distortion.

[0068] Since the direct learning architecture (DLA) is not affected by noise and has good performance, but the calculation is complex and the convergence speed is slow, the convergence of the indirect learning architecture (ILA) is easily affected by noise and has low convergence accuracy, but the calculation is simple and the convergence speed is fast. Therefore, the existing technical solutions cannot balance the convergence speed and convergence accuracy, which may cause the pre-distortion system to have slow convergence speed or low convergence accuracy.

[0069] In view of the above defects, the embodiment of the present application provides a mixed learning structure digital pre-distortion method, as shown in Figure 1 , which is a flowchart of a mixed learning structure digital pre-distortion method provided by the embodiment of the present application. The method comprises the following steps:

[0070] Step S1, cascade the pre-distortion signal calculated by the power amplifier and the forward pre-distortion device to obtain a power amplification signal;

[0071] Step S2, monitor the power amplification signal and calculate the signal-to-noise ratio of the power amplification signal;

[0072] Step S3, according to the signal-to-noise ratio and the power amplification signal, determine the parameter learning link in the preset parameter learning structure;

[0073] Step S4, perform pre-distortion parameter estimation according to the determined parameter learning link, and perform pre-distortion parameter update according to the estimation result.

[0074] In the specific execution of the embodiment, the pre-distortion device is inserted before the power amplifier, the pre-distortion signal opposite to the nonlinearity of the power amplifier is calculated, the pre-distortion device is cascaded with the power amplifier to obtain a linear power amplification signal.

[0075] In the process of mixed parameter learning, the power amplification signal is input into the signal-to-noise ratio detection submodule for signal-to-noise ratio detection, and the signal-to-noise ratio (SNR) of the power amplification signal at the output end of the power amplifier is monitored and calculated according to the change of the circuit environment.

[0076] Then, the signal-to-noise ratio and the power amplification signal are input into a judgment switching submodule for judgment, and a parameter learning link is determined in a preset parameter learning structure.

[0077] By selecting the parameter learning link according to the quantization noise, performing the pre-distortion parameter estimation according to the determined parameter learning link, the optimal convergence of the parameter learning can be ensured, the balance of the convergence speed and the convergence accuracy is realized, and the pre-distortion parameter is updated according to the estimation result.

[0078] Although the prior art also discloses a scheme of quantization noise of a hybrid digital pre-distortion system, it can only solve the problems of the influence of quantization noise on the non-linear pre-distortion of the power amplifier and limited computing resources, but the method does not realize flexible loading and use of the parameter learning method, the saving of hardware resources is limited, and the universality is poor. The digital pre-distortion method does not meet the demand of flexible use of computing power and saving of hardware resources in the future.

[0079] The scheme can select the parameter learning link according to the quantization noise, determine the optimal learning link for parameter learning, and ensure the optimal convergence of the parameter learning. The computing resources are saved while the correction accuracy is ensured, the computing efficiency is improved, and the power consumption of the equipment is reduced.

[0080] In another embodiment provided by the application, referring to Figure 2 is a principle schematic diagram of the hybrid learning structure digital pre-distortion method provided by the embodiment of the application. When performing pre-distortion calculation, the baseband signal x(n) is input into a forward pre-distortion calculator to perform forward pre-distortion calculation, and a pre-distortion signal u(n) is obtained, wherein n represents a sampling point.

[0081] The pre-distortion signal u(n) is cascaded with a power amplifier PA to obtain a power amplification signal y(n).

[0082] When performing hybrid parameter learning, the power amplification signal y(n) is input into a signal-to-noise ratio detection submodule for signal-to-noise ratio detection, and the signal-to-noise ratio s(n) of the power amplification signal y(n) at the output end of the power amplifier is monitored and calculated.

[0083] Then, the signal-to-noise ratio s(n) and the power amplification signal y(n) are input into a judgment switching submodule for judgment. According to the signal-to-noise ratio s(n) and the power amplification signal y(n), the stability value of the system is calculated. The stability value is used to measure the difference between the actual output and the ideal output of the power amplifier, can evaluate the influence of the signal-to-noise ratio on the system, judge whether the system state is stable, and then determine the corresponding parameter learning link for parameter learning calculation. The learning architecture of the parameter estimation can be determined according to the working state of the system, and the computing power is flexibly loaded.

[0084] In another embodiment provided by the application, the parameter learning structure comprises a direct learning structure and an indirect learning structure.

[0085] Referring to Figure 2 , the application selects different learning structures as the parameter learning link by judgment, the convergence effect of the direct learning structure is not affected by noise, and the performance is good, but the calculation is complex, and the convergence of the indirect learning structure is easily affected by noise, but the calculation is simple and the convergence speed is fast. Therefore, the application simultaneously designs two learning architectures, selects the parameter learning method according to the quantization noise, and improves the digital pre-distortion effect.

[0086] In another embodiment provided by the application, the following steps are specifically performed when the stability is calculated:

[0087] The ideal output signal i(n) of the power amplifier needs to be obtained, and the difference t1 between the ideal output signal i(n) of the power amplifier and the power amplification signal y(n) is calculated; t1=i(n)-y(n), t1 represents the difference between the actual output and the ideal output, and in actual application, the smaller the difference between the actual output and the ideal output is, the better.

[0088] The reciprocal t2 of the signal-to-noise ratio is calculated, t2 represents the reciprocal of the signal-to-noise ratio value of the power amplifier output end calculated by the signal-to-noise ratio monitoring module, and in actual application, the larger the value of SNR is, the better, that is, the smaller the value of t2 is.

[0089] According to the difference and the reciprocal of the signal-to-noise ratio, the stability value t is calculated, t=w1t1+w2t2, w1 is a preset difference ratio coefficient, representing the proportion of the difference between the actual output and the ideal output in judging the stability of the system; w2 is a preset signal-to-noise ratio ratio coefficient, representing the proportion of the signal-to-noise ratio in judging the stability of the system w1, w2 can be set according to the actual ability of the circuit and the specific application scene.

[0090] In another embodiment provided by the application, when the corresponding parameter learning link is matched in the parameter learning structure, the matching is performed according to the stability value.

[0091] When the stability value t>a, it is considered that the difference between the actual output and the ideal output of the power amplifier is large, the system is greatly affected by the signal-to-noise ratio, and the system state is unstable. The direct learning structure is selected as the parameter learning link for parameter learning, that is, Figure 2 link ① in the direct learning structure is used, the convergence effect of the direct learning structure is not affected by noise, and the performance is good, so that the digital pre-distortion converges to a better linearization result, the convergence precision is ensured, and the parameter estimation is avoided due to the influence of the signal-to-noise ratio. The power amplification signal y(n) is transmitted to the direct learning structure for parameter learning.

[0092] When the stability value t is less than or equal to a, it is considered that the difference between the actual output and the ideal output of the power amplifier is small, the system is less affected by the signal-to-noise ratio, and the system state is stable. The indirect learning structure is selected as the parameter learning link to learn the parameters, that is, link ②. The indirect learning structure has the characteristics of simple calculation and fast convergence speed, so that the digital pre-distortion quickly converges and saves the calculation resources. The power amplifier signal y(n) is transmitted to link ② for parameter learning.

[0093] It should be noted that a is a set threshold value for switching the parameter learning link, which can be set or adjusted according to the actual ability of the circuit and the specific application scenario.

[0094] In another embodiment provided by the present application, referring to Figure 3 , another principle schematic diagram of the hybrid learning structure digital pre-distortion method provided by the embodiment of the present application is shown. When step S4 is specifically executed, the following steps are specifically executed.

[0095] When the parameter learning link is the direct learning structure, that is, the result of the judgment of whether the stability value is greater than the preset threshold a is yes, the direct learning structure is used for parameter learning, that is, link ① in the formula. Figure 3

[0096] At this time, the power amplifier signal y(n) output by the power amplifier and the baseband signal x(n) input into the forward pre-distorter are acquired for parameter estimation, and then solving is performed.

[0097] The parameters of the forward pre-distorter are updated according to the solving result, the correction of the forward pre-distorter is completed, and the pre-distortion result is output to the local module.

[0098] In another embodiment provided by the present application, when step S4 is specifically executed, referring to Figure 3 , the following steps are specifically executed.

[0099] When the parameter learning link is the indirect learning structure, that is, the result of the judgment of whether the stability value is greater than the preset threshold a is no, the indirect learning structure is used for parameter learning, that is, link ② in the formula. Figure 3

[0100] At this time, the power amplifier signal y(n) output by the power amplifier is acquired and input into the backward pre-distorter for backward pre-distortion and parameter estimation, and the backward output signal is calculated.

[0101] The pre-distortion signal u(n) output by the forward pre-distortor and the backward output signal are acquired for parameter estimation, and then solving is performed.

[0102] ​​The parameter updating backward pre-distortion device is copied to the forward pre-distortion device, the parameters of the forward pre-distortion device are updated, and the pre-distortion result is output to the local module.

[0103] In another embodiment provided by the application, the pre-distortion parameter estimation calculation process specifically adopts an estimation function, and the estimation function is e(n) is an error signal, and Z(n) are two signals for parameter estimation calculation, respectively;

[0104] When learning by using the direct learning structure, the error signal e(n) = y(n)-x(n), y(n) is a power amplification signal, and x(n) is a baseband signal input into the forward pre-distortion device.

[0105] When learning by using the indirect learning structure, the error signal e(n) = z(n)-u(n), z(n) is a backward output signal calculated by the backward pre-distortion device, and u(n) is a pre-distortion signal calculated by the forward pre-distortion device.

[0106] The estimation function is used to minimize and adaptively estimate the two signals for parameter estimation calculation, the Newton method is used to minimize the error, the polynomial coefficients corresponding to the minimum error signal e(n) are obtained, and the polynomial coefficients corresponding to the minimum error signal e(n) are taken as the parameter estimation results obtained by solving.

[0107] The scheme provided in the application is aimed at the problem of computing power demand caused by large bandwidth and high data rate in linearization pre-correction processing of a next-generation wireless base station, two learning structures are designed, the selection of the parameter learning method is performed according to quantization noise, and optimal convergence of parameter learning can be ensured. While ensuring correction accuracy, computing resources are saved, computing efficiency is improved, and device power consumption is reduced.

[0108] The scheme provided in the application can be combined with subsequent chip products or solutions to solve the problem of computing power demand caused by large bandwidth and high data rate in linearization pre-correction processing of a next-generation wireless base station. The present application can further improve the execution efficiency and execution effect of digital pre-distortion, improve the application efficiency of hardware computing power, and realize flexible loading of computing power modules. Computing resources are saved, and device power consumption is reduced.

[0109] Referring to Figure 4 is a structural schematic diagram of a hybrid learning structure digital pre-distortion device provided by an embodiment of the application, and the device comprises:

[0110] The pre-distortion module is configured to cascade the pre-distortion signal calculated by the power amplifier and the forward pre-distortion device to obtain a power amplification signal.

[0111] a signal-to-noise ratio calculation module, configured to monitor the power amplification signal and calculate a signal-to-noise ratio of the power amplification signal;

[0112] a link determination module, configured to determine a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal;

[0113] a parameter updating module, configured to perform pre-distortion parameter estimation according to the determined parameter learning link and perform pre-distortion parameter updating according to the estimation result.

[0114] The hybrid learning structure digital pre-distortion apparatus provided in the embodiment can perform all steps and functions of the hybrid learning structure digital pre-distortion method provided in any of the above embodiments, and the specific functions of the apparatus will not be repeated here.

[0115] Referring to Figure 5 is a structural schematic diagram of a terminal device provided in an embodiment of the present application. The terminal device comprises a processor, a memory, and a computer program stored in the memory and capable of running on the processor, for example, a hybrid learning structure digital pre-distortion program. The processor implements the steps in any of the above hybrid learning structure digital pre-distortion method embodiments when executing the computer program, for example, steps S1-S4 shown in the figure. Figure 1 Alternatively, the processor implements the functions of the modules in the above device embodiments when executing the computer program.

[0116] For example, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present application. The one or more modules can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the hybrid learning structure digital pre-distortion apparatus. For example, the computer program can be divided into several modules, and the specific functions of the modules have been described in detail in any of the above hybrid learning structure digital pre-distortion method embodiments, and the specific functions of the apparatus will not be repeated here.

[0117] The terminal device can be a desktop computer, a notebook computer, a palm computer, a cloud server, and the like. The terminal device can include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the schematic diagram is only an example of the terminal device, and does not limit the hybrid learning structure digital pre-distortion apparatus, which can include more or fewer components than the diagram, or combine certain components, or different components, for example, the terminal device can also include an input / output device, a network access device, a bus, and the like.

[0118] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The processor is a control center of the hybrid learning structure digital pre-distortion device, and is connected with various parts of the hybrid learning structure digital pre-distortion device through various interfaces and lines.

[0119] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the hybrid learning structure digital pre-distortion device by running or executing the computer program and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0120] If the modules of the mixed learning structure digital pre-distortion device set are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. When the processor executes the computer program, the steps of the above-mentioned various method embodiments can be realized. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0121] The embodiment of the present application also provides a computer program product, which includes computer program / instruction, and the computer program / instruction realizes the steps of the method when executed by the processor.

[0122] The computer program product provided by the embodiment can execute all steps and functions of the mixed learning structure digital pre-distortion method provided by any of the above-mentioned embodiments, and the specific functions of the product will not be described here.

[0123] It should be noted that those skilled in the art can make some improvements and refinements without departing from the principles of the present application, and these improvements and refinements are also considered within the protection scope of the present application.

Claims

1. A hybrid learning structure digital pre-distortion method, characterized in that, The method comprises: obtaining a power amplification signal by cascading a pre-distortion signal calculated by a power amplifier and a forward pre-distorter; monitoring the power amplification signal and calculating a signal-to-noise ratio of the power amplification signal; determining a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal; performing pre-distortion parameter estimation according to the determined parameter learning link and updating pre-distortion parameters according to the estimation result.

2. The hybrid learning structure digital pre-distortion method according to claim 1, wherein, The method comprises: calculating a stability value according to the signal-to-noise ratio and the power amplification signal; matching a corresponding parameter learning link in the parameter learning structure according to the size of the calculated stability value.

3. The hybrid learning structure digital pre-distortion method of claim 1, wherein, The parameter learning structure comprises a direct learning structure and an indirect learning structure.

4. The hybrid learning structure digital pre-distortion method of claim 2, wherein, The method comprises: calculating a difference between an ideal output signal of the power amplifier and the power amplification signal; calculating the stability value according to the difference and the signal-to-noise ratio; wherein the stability value t = w1t1 + w2t2, t1 is the difference, t2 is the reciprocal of the signal-to-noise ratio, w1 is a preset difference proportion coefficient, and w2 is a preset signal-to-noise ratio proportion coefficient.

5. The hybrid learning structure digital pre-distortion method of claim 2, wherein, The method comprises: when the stability value is greater than a preset threshold, determining the direct learning structure in the parameter learning structure as the parameter learning link; when the stability value is not greater than the threshold, determining the indirect learning structure in the parameter learning structure as the parameter learning link.

6. The hybrid learning structure digital pre-distortion method of claim 1, wherein, The method comprises: when the parameter learning link is the direct learning structure, collecting the power amplification signal and a baseband signal input into the forward pre-distorter for parameter estimation; updating parameters of the forward pre-distorter according to the solved parameters.

7. The hybrid learning structure digital pre-distortion method of claim 1, wherein, The method comprises: when the parameter learning link is the indirect learning structure, collecting the power amplification signal and inputting the power amplification signal into a backward pre-distorter for pre-distortion calculation to obtain a backward output signal; performing parameter estimation on the backward output signal and the pre-distortion signal; updating parameters of the backward pre-distorter and the forward pre-distorter according to the solved parameters.

8. The hybrid learning structure digital pre-distortion method of claim 1, wherein, The method comprises: performing minimum adaptive estimation on two signals calculated by parameter estimation calculation by using an estimation function to obtain a parameter estimation result. wherein the estimation function is e(n) is an error signal, and Z(n) are two signals of the parameter estimation calculation, respectively.

9. A hybrid learning structure digital pre-distortion apparatus, characterized by, The device comprises: a pre-distortion module configured to obtain a power amplification signal by cascading a pre-distortion signal calculated by a power amplifier and a forward pre-distorter; a signal-to-noise ratio calculation module configured to monitor the power amplification signal and calculate a signal-to-noise ratio of the power amplification signal; and a parameter learning link determination module configured to determine a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal. The link determination module is configured to determine a parameter learning link in a preset parameter learning structure according to the signal-to-noise ratio and the power amplification signal. The parameter updating module is configured to perform pre-distortion parameter estimation according to the determined parameter learning link, and perform pre-distortion parameter updating according to the estimation result.

10. A terminal device, comprising: The computer readable storage medium comprises a computer program stored therein, wherein the computer program, when executed, controls a device in which the computer readable storage medium is located to perform the hybrid learning structure digital pre-distortion method according to any one of claims 1 to 8.

11. A computer readable storage medium, characterized in that, The computer readable storage medium comprises a computer program stored therein, wherein the computer program, when executed, controls a device in which the computer readable storage medium is located to perform the hybrid learning structure digital pre-distortion method according to any one of claims 1 to 8.

12. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instruction, when executed by the processor, implements the steps of the method according to any one of claims 1 to 8.

Images