Signal nonlinear distortion compensation method, model and communication system

Abstract

The invention discloses a signal nonlinear distortion compensation method, a model and a communication system, and relates to the technical field of wireless communication. In the present disclosure, based on a preset non-linear compensation fusion model, DPD processing, CFR processing and error compensation processing are performed on an initial OFDM signal, thereby obtaining an initial OFDM signal after non-linear distortion compensation; therefore, through a mode of connecting a BL-DPD module, a BL-CFR module and an error compensation module in parallel in a nonlinear compensation fusion model, the technical defects that in the prior art, after a DPD module is arranged behind a CFR module, the requirement for the sampling rate of ADC / DAC is high, the requirement for the convergence speed of hardware and an algorithm is improved, and the implementation difficulty and cost of a system are increased are overcome; therefore, not only is the requirement of ADC / DAC sampling rate reduced, but also the compensation performance of the DPD for the nonlinearity of the power amplifier is improved, and the communication performance and the sensing performance of the OFDM system are also improved.

Description

A signal nonlinear distortion compensation method, model and communication system Technical Field

[0001] The present disclosure relates to the field of wireless communication technology, and in particular to a compensation method, model and communication system for signal nonlinear distortion. Background Art

[0002] Orthogonal Frequency Division Multiplexing (OFDM) is a multi-carrier modulation technology that divides a carrier into several mutually orthogonal sub-carriers to overcome frequency selective fading and narrowband interference.

[0003] Therefore, signals modulated using OFDM technology, also known as OFDM signals, typically have characteristics such as a non-constant envelope, wide bandwidth, and a high peak-to-average power ratio (PAPR). However, when OFDM signals pass through a power amplifier (PA), nonlinear distortion is inevitably introduced.

[0004] Initially, to ensure good linearity of the signal output by the power amplifier, a simple power back-off method was generally used. However, this method reduced the efficiency of the power amplifier and resulted in a waste of resources. Then, to improve the efficiency of the power amplifier, the power amplifier was usually operated near the saturation point, but this also caused severe in-band distortion, increased the bit error rate of the communication system, and also generated out-of-band spectrum expansion, interfering with adjacent channels.

[0005] In view of this, in order to improve the impact of nonlinear distortion of power amplifiers, crest factor reduction (CFR) technology and digital pre-distortion (DPD) technology are usually used. Among them, CFR technology aims to reduce the PAPR of the signal by reducing the signal peak value. In this way, after the signal PAPR is reduced, the value of the output peak power back-off at the average power operating point of the power amplifier can be reduced, thereby improving the power amplifier efficiency. DPD technology is an effective method to compensate for the nonlinearity and memory effects of the power amplifier in the high-efficiency area. It can be seen that the combination of CFR technology and DPD technology can simultaneously meet the application needs of improving power amplifier efficiency and improving linearity indicators.

[0006] Therefore, in related technologies, the CFR module and the DPD module are usually cascaded to achieve the purpose of simultaneously improving the power amplifier efficiency and improving the linearity. Moreover, in the solution combining the classic CFR technology with the DPD technology, the DPD module is usually applied after the CFR module.

[0007] However, when the CFR module and the DPD module are cascaded, the DPD module is applied after the CFR module. This will cause the PAPR of the signal, which has been reduced after peak clipping by the CFR module, to increase again when passing through the DPD module. In addition, due to the nonlinearity of the power amplifier, the output signal will produce spectrum spread, thereby requiring a higher sampling rate for the analog-to-digital converter (ADC) / digital-to-analog converter (DAC) in the OFDM system, thereby increasing the requirements for hardware and algorithm convergence speed, and increasing the difficulty and cost of system implementation.

[0008] It can be seen that how to effectively reduce the requirements for ADC / DAC sampling rate and improve the DPD compensation performance for the nonlinearity of the power amplifier are technical problems that need to be solved at present.

[0009] Summary of the Invention

[0010] The embodiments of the present disclosure provide a method, model, and communication system for compensating for signal nonlinear distortion, which are used to reduce the requirements for ADC / DAC sampling rates and improve the DPD compensation performance for power amplifier nonlinearity, thereby further improving the communication performance and perception performance of the OFDM system.

[0011] In a first aspect, an embodiment of the present disclosure provides a method for compensating for nonlinear distortion of a signal, the method comprising:

[0012] Inputting the initial orthogonal frequency division multiplexing (OFDM) signal into a preset nonlinear compensation fusion model; wherein the nonlinear compensation fusion model includes: a band-limited digital predistortion (BL-DPD) module, a band-limited crest factor reduction (BL-CFR) module, and an error compensation module, wherein the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module;

[0013] Obtaining, for the initial OFDM signal, a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module;

[0014] An initial OFDM signal after nonlinear distortion compensation is obtained based on the first OFDM signal, the second OFDM signal, and the third OFDM signal.

[0015] In an optional embodiment, the BL-DPD module and the BL-CFR module each use the same basis function.

[0016] In an optional embodiment, obtaining, for an initial OFDM signal, a first OFDM signal processed by a BL-DPD module, a second OFDM signal processed by a BL-CFR module, and a third OFDM signal processed by an error compensation module, respectively, includes:

[0017] Based on the model parameter sets converged by the BL-DPD module, the BL-CFR module and the error compensation module in the offline mode, the initial OFDM signal is modulated to obtain the first OFDM signal, the second OFDM signal and the third OFDM signal.

[0018] In an optional embodiment, the model parameter set includes any one of the following parameter combinations:

[0019] The kernel coefficients, nonlinear order, memory depth and order of the low-order low-pass filter LFP of the BL-DPD module;

[0020] The kernel coefficients, nonlinear order, memory depth and order of low-order LFP of the BL-CFR module;

[0021] Kernel coefficients, nonlinearity order, and memory depth of the error compensation module.

[0022] In an optional embodiment, if the parameter combination of the model parameter set is: the kernel coefficient of the BL-DPD module, the nonlinear order, the memory depth and the order of the low-order LFP, the model parameter set is obtained in the following manner:

[0023] Input the sample OFDM signal in offline mode into the BL-DPD module to obtain the sample OFDM signal after DPD processing;

[0024] Based on the sample OFDM signal after DPD processing and the conjugate parameter set corresponding to the initial parameter set of the BL-DPD module, obtaining the sample OFDM signal after inverse DPD processing;

[0025] Obtaining a target DPD error signal based on the sample OFDM signal in the offline mode and the sample OFDM signal after inverse DPD processing;

[0026] Based on the target DPD error signal and a preset low-noise variable step-size least mean square algorithm, the initial parameter set is iteratively modified until the absolute value of the target DPD error signal is less than the set DPD error signal threshold;

[0027] The iteratively modified initial parameter set is used as the model parameter set of the BL-DPD module.

[0028] In an optional embodiment, obtaining a sample OFDM signal after inverse DPD processing based on the sample OFDM signal after DPD processing and a conjugate parameter set corresponding to the initial parameter set of the BL-DPD module includes:

[0029] Performing digital-to-analog conversion, up-conversion, and power amplification on the sample OFDM signal after DPD processing in sequence to obtain a power-amplified sample OFDM signal;

[0030] The power-amplified sample OFDM signal is subjected to power attenuation, down-conversion, and analog-to-digital conversion in sequence to obtain an analog-to-digital converted sample OFDM signal;

[0031] Based on the analog-to-digital converted sample OFDM signal and the conjugate parameter set corresponding to the initial parameter set, the sample OFDM signal after inverse DPD processing is obtained.

[0032] In an optional embodiment, the initial parameter set is iteratively modified based on the target DPD error signal and a preset low-noise variable step-size least mean square algorithm, including:

[0033] During each modification of the initial parameter set, the following operations are performed:

[0034] Obtain sample OFDM signals corresponding to multiple historical moments adjacent to the current sample OFDM signal in the offline mode;

[0035] Based on the historical DPD error signals corresponding to the multiple sample OFDM signals, the average DPD error signal of the sample OFDM signal in the offline mode at the current moment is obtained;

[0036] Obtaining a first target step size factor based on a mean value of the DPD error signal, a step size factor at a previous historical moment adjacent to the current moment, a target DPD error signal at the current moment, and a historical DPD error signal at the previous historical moment;

[0037] Based on the first target step size factor, the conjugate DPD error signal corresponding to the target DPD error signal at the current moment, and the sample OFDM signal after analog-to-digital conversion, the initial parameter set at the current moment is modified to obtain a modified initial parameter set.

[0038] In an optional embodiment, before iteratively modifying the initial parameter set based on the target DPD error signal and a preset low-noise variable step-size least mean square algorithm, the method further includes:

[0039] If the absolute value of the target DPD error signal is not less than the DPD error signal threshold, the initial parameter set is directly used as the model parameter set of the BL-DPD module.

[0040] In an optional embodiment, if the parameter combination of the model parameter set is: the kernel coefficient of the BL-CFR module, the nonlinear order, the memory depth and the order of the low-order LFP, the model parameter set is obtained as follows:

[0041] The sample OFDM signal in the offline mode is sequentially input into the BL-DPD module and the preset CFR module to obtain the sample OFDM signal after DPD-CFR processing;

[0042] Based on the sample OFDM signal in the offline mode and the conjugate parameter set corresponding to the initial parameter set of the BL-CFR module, the sample OFDM signal after CFR processing is obtained;

[0043] Obtaining a target CFR error signal based on the sample OFDM signal after DPD-CFR processing and the sample OFDM signal after CFR processing;

[0044] Based on the target CFR error signal and a preset low-noise variable step-size least mean square algorithm, the initial parameter set is iteratively modified until the absolute value of the target CFR error signal is less than the set CFR error signal threshold;

[0045] The iteratively modified initial parameter set is used as the model parameter set of the BL-CFR module.

[0046] In an optional embodiment, before iteratively modifying the initial parameter set based on the target CFR error signal and a preset low-noise variable step-size least mean square algorithm, the method further includes:

[0047] If the absolute value of the target CFR error signal is not less than the CFR error signal threshold, the initial parameter set is directly used as the model parameter set of the BL-CFR module.

[0048] In an optional embodiment, if the parameter combination of the model parameter set is: kernel coefficient, nonlinear order and memory depth of the error compensation module, the model parameter set is obtained in the following manner:

[0049] The sample OFDM signal in the offline mode is input into the nonlinear compensation fusion model and the preset high-order LPF respectively to obtain the sample OFDM signal after nonlinear distortion compensation and the sample OFDM signal after filtering;

[0050] Obtaining a target compensation error signal based on the sample OFDM signal after nonlinear distortion compensation and the sample OFDM signal after filtering;

[0051] Based on the target compensation error signal and a preset low-noise variable step-size least mean square algorithm, the initial parameter set of the error compensation module is iteratively modified until the absolute value of the target compensation error signal is less than the set compensation error signal threshold;

[0052] The iteratively modified initial parameter set is used as the model parameter set of the error compensation module.

[0053] In an optional embodiment, before iteratively modifying the initial parameter set of the error compensation module based on the target compensation error signal and a preset low-noise variable step-size least mean square algorithm, the method further includes:

[0054] If the absolute value of the target compensation error signal is not less than the compensation error signal threshold, the initial parameter set is directly used as the model parameter set of the error compensation module.

[0055] In an optional embodiment, after obtaining the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal, the second OFDM signal, and the third OFDM signal, the method further includes:

[0056] Based on the initial OFDM signal after nonlinear distortion compensation and the conjugate parameter set corresponding to the model parameter set of the BL-DPD module, the initial OFDM signal after inverse DPD processing is obtained;

[0057] Obtaining a target distortion compensation error signal based on the initial OFDM signal after nonlinear distortion compensation and the initial OFDM signal after inverse DPD processing;

[0058] Based on the target distortion compensation error signal and the preset sine sum error variable step-least mean square algorithm, the model parameter set is iteratively modified until the absolute value of the target distortion compensation error signal is less than the set distortion compensation error threshold.

[0059] In an optional embodiment, obtaining the initial OFDM signal after inverse DPD processing based on the initial OFDM signal after nonlinear distortion compensation and the conjugate parameter set corresponding to the model parameter set of the BL-DPD module includes:

[0060] The initial OFDM signal after nonlinear distortion compensation is subjected to digital-to-analog conversion, up-conversion and power amplification in sequence to obtain the initial OFDM signal after power amplification;

[0061] The initial OFDM signal after power amplification is subjected to power attenuation, down-conversion and analog-to-digital conversion in sequence to obtain the initial OFDM signal after analog-to-digital conversion;

[0062] Based on the initial OFDM signal after analog-to-digital conversion and the conjugate parameter set corresponding to the model parameter set, the initial OFDM signal after inverse DPD processing is obtained.

[0063] In an optional embodiment, based on the target distortion compensation error signal and a preset sine sum error variable step-size least mean square algorithm, the model parameter set is iteratively modified, including:

[0064] Each time a model parameter set is modified, the following operations are performed:

[0065] Obtaining sample OFDM signals after nonlinear distortion compensation corresponding to multiple historical moments adjacent to the initial OFDM signal after nonlinear distortion compensation;

[0066] Based on multiple nonlinear distortion compensated sample OFDM signals and their corresponding historical distortion compensation error signals, obtaining a distortion compensation error signal mean value of the initial OFDM signal after nonlinear distortion compensation at a current moment;

[0067] Obtaining a second target step size factor based on an error term corresponding to a mean value of the distortion compensation error signal, a target distortion compensation error signal at a current moment, and a historical distortion compensation error signal at a previous historical moment adjacent to the current moment;

[0068] Based on the second target step size factor, the conjugate distortion compensation error signal corresponding to the target distortion compensation error signal at the current moment, and the initial OFDM signal after analog-to-digital conversion, the model parameter set at the current moment is modified to obtain a modified model parameter set.

[0069] In an optional embodiment, the method further includes:

[0070] According to the set cycle time, the model parameter set of the BL-DPD module is iteratively modified.

[0071] In a second aspect, the embodiments of the present disclosure further provide a nonlinear compensation fusion model, the nonlinear compensation fusion model comprising: a BL-DPD module, a BL-CFR module, and an error compensation module;

[0072] The BL-DPD module, the BL-CFR module and the error compensation module are connected in parallel; wherein, the BL-DPD module and the BL-CFR module each use the same basis function, and the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module.

[0073] In an optional embodiment, the BL-DPD module and the BL-CFR module each use the same basis function.

[0074] In an optional embodiment, the nonlinear compensation fusion model is used to:

[0075] Based on the initial OFDM signal, obtaining a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module;

[0076] An initial OFDM signal after nonlinear distortion compensation is obtained based on the first OFDM signal, the second OFDM signal, and the third OFDM signal.

[0077] In an optional embodiment, the nonlinear compensation fusion model is further used to:

[0078] In offline mode, the preset low-noise variable step-size least mean square algorithm is used to extract the model parameter set of the BL-DPD module, BL-CFR module and error compensation module.

[0079] In an optional embodiment, the nonlinear compensation fusion model is further used to: in online mode, use a preset sine sum error variable step-least mean square algorithm and a set cycle time to refresh the model parameter set of the nonlinear compensation fusion model.

[0080] In an optional embodiment, the nonlinear compensation fusion model is specifically used to:

[0081] In online mode, the preset sine sum error variable step-least mean square algorithm and the set cycle time are used to refresh the model parameter set only for the BL-DPD module in the nonlinear compensation fusion model.

[0082] In a third aspect, an embodiment of the present disclosure further provides an OFDM communication system, the OFDM communication system comprising: the nonlinear compensation fusion model as described in the second aspect, a first branch, a second branch, a third branch, and a fourth branch;

[0083] The BL-DPD module and the first branch in the nonlinear compensation fusion model are used in an offline mode to iteratively modify the initial parameter set of the BL-DPD module to obtain a model parameter set of the BL-DPD module;

[0084] The BL-CFR module and the second branch in the nonlinear compensation fusion model are used in offline mode to iteratively modify the initial parameter set of the BL-CFR module to obtain the model parameter set of the BL-CFR module;

[0085] The nonlinear compensation fusion model and the third branch are used in an offline mode to iteratively modify the initial parameter set of the error compensation module to obtain the model parameter set of the error compensation module;

[0086] The nonlinear compensation fusion model and the fourth branch are used to iteratively modify the model parameter set of the BL-DPD module in online mode.

[0087] In an optional embodiment, the first branch includes: a digital-to-analog converter DAC, an up-converter, a power amplifier PA, an attenuator, a bandpass filter BPF, a down-converter, an analog-to-digital converter ADC, a training network POST-BL-DPD module, and a preset low-noise variable step size-least mean square algorithm module arranged in sequence.

[0088] In an optional embodiment, the second branch includes: a CFR module, a training network POST-BL-DPD module, and a preset low-noise variable step size-least mean square algorithm module arranged in sequence.

[0089] In an optional embodiment, the third branch includes: a low-pass filter LPF and a preset low-noise variable step-size-least mean square algorithm module arranged in sequence.

[0090] In an optional embodiment, the fourth branch includes: a DAC, an up-converter, a PA, an attenuator, a BPF, a down-converter, an ADC, a training network POST-BL-DPD module, and a preset sine and error variable step-least mean square algorithm module arranged in sequence.

[0091] In a fourth aspect, an electronic device is proposed, comprising a processor and a memory, wherein the memory stores a program code, and when the program code is executed by the processor, the processor executes the steps of the method for compensating for nonlinear distortion of signals described in the first aspect.

[0092] In a fifth aspect, a computer-readable storage medium is proposed, which includes a program code. When the program code is run on an electronic device, the program code is used to enable the electronic device to execute the steps of the signal nonlinear distortion compensation method described in the first aspect.

[0093] In a sixth aspect, a computer program product is provided. When the computer program product is called by a computer, the computer is caused to execute the steps of the method for compensating for nonlinear distortion of a signal as described in the first aspect.

[0094] The beneficial effects of the present disclosure are as follows:

[0095] In the signal nonlinear distortion compensation method provided in the embodiments of the present disclosure, based on a preset nonlinear compensation fusion model, DPD processing, CFR processing, and error compensation processing are performed on the initial OFDM signal, respectively, so as to obtain the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal processed by the BL-DPD module, the second OFDM signal processed by the BL-CFR module, and the third OFDM signal processed by the error compensation module. In this way, by connecting the BL-DPD module, the BL-CFR module, and the error compensation module in parallel in the preset nonlinear compensation fusion model, the technical disadvantages of the prior art, in which the DPD module is applied after the CFR module, which has a high requirement on the sampling rate of the ADC / DAC in the OFDM system, increases the requirements on the hardware and algorithm convergence speed, and increases the difficulty and cost of system implementation, are avoided. Therefore, the requirement on the ADC / DAC sampling rate is effectively reduced, and the DPD compensation performance for the nonlinearity of the power amplifier is improved, thereby further improving the communication performance and perception performance of the OFDM system.

[0096] In addition, other features and advantages of the present disclosure will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present disclosure. The purposes and other advantages of the present disclosure can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without inventive work. In the drawings:

[0098] FIG1 is a schematic diagram of the structure of an OFDM communication system provided by an embodiment of the present disclosure;

[0099] FIG2 is a schematic diagram of an implementation flow of a method for compensating for nonlinear distortion of a signal provided by an embodiment of the present disclosure;

[0100] FIG3 is a logic diagram of processing an initial OFDM signal provided by an embodiment of the present disclosure;

[0101] FIG4 is a flow chart of a method for obtaining a model parameter set of a BL-DPD module according to an embodiment of the present disclosure;

[0102] FIG5 is a logic diagram of obtaining a sample OFDM signal after inverse DPD processing according to an embodiment of the present disclosure;

[0103] FIG6 is a schematic diagram of an implementation flow of a method for modifying an initial parameter set in an offline mode provided by an embodiment of the present disclosure;

[0104] FIG7 is a flow chart of a method for obtaining a model parameter set of a BL-CFR module provided by an embodiment of the present disclosure;

[0105] FIG8 is a flow chart of a method for obtaining a model parameter set of an error compensation module provided by an embodiment of the present disclosure;

[0106] FIG9 is a schematic diagram of an implementation flow of a method for modifying a model parameter set in an online mode provided by an embodiment of the present disclosure;

[0107] FIG10 is a logic diagram of obtaining an initial OFDM signal after inverse DPD processing provided by an embodiment of the present disclosure;

[0108] FIG11 is a schematic diagram of a specific implementation process based on FIG9 provided in an embodiment of the present disclosure;

[0109] FIG12 is a schematic structural diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0110] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the technical solutions of the present disclosure, but not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments described in this disclosure without making any creative efforts shall fall within the scope of protection of the technical solutions of the present disclosure.

[0111] It should be noted that in the description of this disclosure, "multiple" is understood to mean "at least two." "And / or" describes the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. A and B are connected, which can mean: A and B are directly connected, and A and B are connected through C. In addition, in the description of this disclosure, words such as "first" and "second" are used only for the purpose of distinguishing descriptions, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying an order.

[0112] First, the design concept of the embodiment of the present disclosure is briefly introduced below:

[0113] OFDM technology is a multi-carrier modulation technique that divides a carrier into several mutually orthogonal subcarriers to overcome frequency-selective fading and narrowband interference. Therefore, OFDM signals, the communication signals of the global 5G standard (5G NR, 5th Generation Mobile Networks New Radio) based on a new air interface design based on OFDM, offer a wider range of options for time slots and subcarriers within each subframe, effectively supporting not only different communication scenarios but also different perception scenarios.

[0114] Therefore, signals modulated using OFDM technology, also known as OFDM signals, typically have characteristics such as non-constant envelope, wide bandwidth, and high PAPR. However, when OFDM signals pass through the PA, nonlinear distortion is inevitably introduced.

[0115] Initially, to ensure good linearity of the signal output by the power amplifier, a simple power back-off method was generally used. However, this method reduced the efficiency of the power amplifier and resulted in a waste of resources. Then, to improve the efficiency of the power amplifier, the power amplifier was usually operated near the saturation point, but this also caused severe in-band distortion, increased the bit error rate of the communication system, and also generated out-of-band spectrum expansion, interfering with adjacent channels.

[0116] At present, in order to improve the impact of PA nonlinear distortion, CFR technology and DPD technology are combined with their respective advantages to achieve the application needs or purposes of simultaneously improving power amplifier efficiency and improving linearity indicators.

[0117] Specifically, in the classic solution that combines CFR technology with DPD technology, the DPD module is usually applied after the CFR module. However, when the CFR module and DPD module are cascaded, since the DPD module is applied after the CFR module, the PAPR of the signal, which has been reduced after peak clipping by the CFR module, will increase again when passing through the DPD module. On the other hand, due to the nonlinearity of the power amplifier, the output signal will produce spectral expansion. When DPD is implemented, the bandwidth of the feedback receiving channel is 3 to 5 times the bandwidth of the input signal. For bandwidths greater than or equal to 400 MHz, the sampling rate of the ADC of the feedback receiving channel must be at least 4 GSPS. Such a high-speed sampling ADC not only increases the requirements for hardware and algorithm convergence speed, but also increases the difficulty and cost of system implementation.

[0118] In view of this, in order to effectively reduce the ADC / DAC sampling rate and at the same time minimize the inhibitory effect of peak clipping (CFR) on the DPD effect, so as to improve the communication performance and perception performance of 5G NR (such as OFDM system), a nonlinear distortion compensation method is proposed in an embodiment of the present disclosure, specifically including: inputting the OFDM signal into a preset nonlinear compensation fusion model; wherein the nonlinear compensation fusion model includes: a BL-DPD module, a BL-CFR module and an error compensation module, and the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module; further, for the initial OFDM signal, a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module are respectively obtained; finally, based on the first OFDM signal, the second OFDM signal and the third OFDM signal, the initial OFDM signal after nonlinear distortion compensation is obtained.

[0119] In particular, the preferred embodiments of the present disclosure are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure and are not used to limit the present disclosure. In addition, the embodiments of the present disclosure and the features in the embodiments may be combined with each other if there is no conflict.

[0120] Referring to FIG1 , which is a schematic diagram of the structure of an OFDM communication system proposed in an embodiment of the present disclosure, the OFDM communication system includes: a nonlinear compensation fusion model, a first branch, a second branch, a third branch, and a fourth branch; wherein:

[0121] The nonlinear compensation fusion model includes: a BL-DPD module, a BL-CFR module and an error compensation module ECM. The BL-DPD module, the BL-CFR module and the error compensation module ECM are connected in parallel. The error compensation module ECM is used to compensate for the errors of the OFDM signals output by the BL-DPD module and the BL-CFR module.

[0122] Optionally, the BL-DPD module and the BL-CFR module each use the same basis function.

[0123] The first branch includes: a DAC, an up-converter, a PA, an attenuator 1 / G, a band-pass filter (BPF), a down-converter, an ADC, a training network POST-BL-DPD module, and a preset low-noise variable step size-least mean square (LNVSS-LMS) algorithm module, which are arranged in sequence.

[0124] The second branch includes: a CFR module, a training network POST-BL-DPD module and a preset low-noise variable step size-least mean square algorithm module which are arranged in sequence.

[0125] The third branch includes: a low-pass filter (LPF) and a preset low-noise variable step-size-least mean square algorithm module arranged in sequence.

[0126] The fourth branch includes: a DAC, an upconverter, a PA, an attenuator 1 / G, a BPF, a downconverter, an ADC, a training network POST-BL-DPD module, and a preset Sine and Error Variable Step Size-Least Mean Square (SEVSS-LMS) algorithm module, which are arranged in sequence.

[0127] It can be seen that the only difference between the first branch and the fourth branch is that the first branch adopts the LNVSS-LMS algorithm, while the fourth branch adopts the SEVSS-LMS algorithm; among them, the LNVSS-LMS algorithm adjusts the change of the step size factor function by adopting the error energy at the current moment and the absolute error energy of the error at the previous moment, as well as parameters such as the average value, which can not only effectively suppress the susceptibility of the step size factor function to noise and enhance the noise resistance of the LNVSS-LMS algorithm, but also adjust the parameter value to be smaller than the error value at this time, so as to obtain a smaller step size and a smaller steady-state error; the SEVSS-LMS algorithm controls the variable step size factor function by adjusting three parameters, so that the variable step size factor function automatically increases in the initial stage when the error signal is large, and the convergence speed is fast; after steady state, it can maintain a small step size, slow convergence speed, and small error.

[0128] Furthermore, the nonlinear compensation fusion model may also be referred to as a compensation band limited-crest factor reduction-digital predistortion (CBL-CFR-DPD) model.

[0129] In addition, the above-mentioned first branch, second branch and third branch can share the same preset low-noise variable step size-least mean square algorithm module to save circuit overhead; it should be noted that in the embodiment of the present disclosure, the number of non-preset low-noise variable step size-least mean square algorithm modules is limited.

[0130] It is worth noting that in the embodiment of the present disclosure, the above-mentioned nonlinear compensation fusion model is used to perform DPD processing, CFR processing and error compensation processing on the initial OFDM signal respectively, to obtain a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module, thereby obtaining the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal, the second OFDM signal and the third OFDM signal.

[0131] The BL-DPD module and the first branch in the nonlinear compensation fusion model are used in offline mode to iteratively modify the initial parameter set of the BL-DPD module to obtain the model parameter set of the BL-DPD module, that is, to extract the kernel parameters of the BL-DPD module in offline mode. The offline mode is characterized by the absence of the initial OFDM signal, i.e., the module / model (pre-) training mode.

[0132] The BL-CFR module and the second branch in the nonlinear compensation fusion model are used in offline mode to iteratively modify the initial parameter set of the BL-CFR module to obtain the model parameter set of the BL-CFR module, that is, to extract the kernel parameters of the BL-CFR module in offline mode.

[0133] The nonlinear compensation fusion model and the third branch are used in offline mode to iteratively modify the initial parameter set of the error compensation module ECM to obtain the model parameter set of the error compensation module ECM, that is, to extract the kernel parameters of the error compensation module ECM in offline mode.

[0134] The nonlinear compensation fusion model and the fourth branch are used in online mode to iteratively modify the model parameter set of the BL-DPD module, that is, the parameters of the nonlinear compensation fusion model (i.e., the CBL-CFR-DPD model) are effectively refreshed in online mode, ultimately ensuring that the pre-distortion of the entire OFDM communication system after cascading can effectively improve the distortion of the OFDM signal caused by the nonlinearity of the power amplifier, thereby improving the efficiency of the entire OFDM communication system; among them, the online mode is characterized by: receiving the initial OFDM signal, that is, the real-time working mode.

[0135] Optionally, the nonlinear compensation fusion model is also used for: in offline mode, using a preset low-noise variable step size-least mean square algorithm to extract a model parameter set for the BL-DPD module, BL-CFR module and error compensation module; and / or, in online mode, using a preset sine sum error variable step size-least mean square algorithm and a set cycle time to refresh the model parameter set of the nonlinear compensation fusion model; optionally, the nonlinear compensation fusion model is specifically used for: in online mode, using a preset sine sum error variable step size-least mean square algorithm and a set cycle time to refresh the model parameter set only for the BL-DPD module in the nonlinear compensation fusion model.

[0136] It should also be noted that the Mode in FIG1 can be understood as a state selection module, so the corresponding state can be: offline mode or online mode.

[0137] Obviously, based on the above-mentioned OFDM communication system, a new nonlinear compensation fusion model (i.e., a CBL-CFR-DPD integrated model) is constructed. The BL-DPD module and the BL-CFR module are integrated into one model through the error compensation module ECM. An "offline + online" mechanism is adopted to extract and update the parameters of each module (i.e., the model parameter set). In offline mode, the LNVSS-LMS adaptive algorithm is used to extract / modify the parameters of the BL-DPD module, BL-CFR module, and ECM module; in online mode, the SEVSS-LMS adaptive algorithm is used to effectively refresh / modify only the parameters of the BL-DPD module, thereby effectively reducing the order of the band-limited filter and the complexity of the entire nonlinear compensation fusion model, while ensuring the accuracy of the entire nonlinear compensation fusion model. In addition, the operational complexity of the nonlinear compensation fusion model is comparable to that of a system with only the DPD module, except for the additional addition and subtraction operations of some coefficients. Compared with the classic solution of independent application of the CFR module and the DPD module, the operational complexity is reduced, and it can effectively improve the nonlinear performance of the digital predistorter in compensating the wideband power amplifier.

[0138] The following describes the method for compensating for signal nonlinear distortion provided by an exemplary embodiment of the present disclosure in combination with the above-mentioned OFDM communication system and with reference to the accompanying drawings. It should be noted that the above-mentioned system architecture is only shown to facilitate understanding of the spirit and principles of the present disclosure, and the embodiments of the present disclosure are not limited in this respect.

[0139] 2 is a flowchart illustrating a method for compensating for nonlinear distortion of a signal provided by an embodiment of the present disclosure, which is applied to the above-mentioned OFDM communication system. The specific implementation process of the method is as follows:

[0140] S201: Inputting the initial OFDM signal into a preset nonlinear compensation fusion model.

[0141] The nonlinear compensation fusion model includes a BL-DPD module, a BL-CFR module and an error compensation module ECM. The error compensation module ECM is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module.

[0142] It should be noted that the BL-DPD module and the BL-CFR module each use the same basis function but different model parameter sets (ie, coefficients).

[0143] S202: Obtaining, for the initial OFDM signal, a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module.

[0144] In an optional implementation, referring to FIG3 , when executing step S202 , the nonlinear compensation fusion model modulates the initial OFDM signal based on the model parameter sets converged by the BL-DPD module, the BL-CFR module, and the error compensation module ECM in the offline mode, respectively, to obtain a first OFDM signal, a second OFDM signal, and a third OFDM signal, wherein the offline mode is characterized by: the initial OFDM signal is not received; in this way, the initial OFDM signal is processed by three parallel modules, effectively avoiding the technical disadvantages of the prior art that the DPD module is applied after the CFR module, which has a high sampling rate requirement for the ADC / DAC in the OFDM system, increases the requirements for hardware and algorithm convergence speed, and increases the difficulty and cost of system implementation.

[0145] Exemplarily, assuming that the initial OFDM signal is s(n), the first OFDM signal x(n) is obtained through core processing of the BL-DPD module. The calculation formula of the first OFDM signal x(n) is as follows:

[0146] Where a kml , K1 and M1 are the kernel coefficient, nonlinear order and memory depth of the BL-DPD module, L is the order of the low-order LFP in the BL-DPD module, that is, the set of model parameters that the BL-DPD module converges in offline mode; optionally, L is less than 16.

[0147] The initial OFDM signal s(n) is processed by the core of the BL-CFR module to obtain the second OFDM signal c(n). The calculation formula of the second OFDM signal c(n) is as follows:

[0148] Where b kml, K1 and M1 are the kernel coefficient, nonlinear order and memory depth of the BL-CFR module, L is the order of the low-order LFP in the BL-CFR module, that is, the set of model parameters that the BL-CFR module converges in offline mode; optionally, the order L is less than 16.

[0149] The initial OFDM signal s(n) is processed by the core of the error compensation module ECM to obtain a third OFDM signal δ(n). The calculation formula of the third OFDM signal δ(n) is as follows:

[0150] Where c km , K2 and M2 are the kernel coefficients, nonlinear order and memory depth of the error compensation module ECM, that is, the set of model parameters that the error compensation module ECM converges in offline mode.

[0151] Therefore, it can be seen that the above model parameter set includes any of the following parameter combinations:

[0152] 1. The kernel coefficient, nonlinear order, memory depth and low-order LFP order of the BL-DPD module;

[0153] 2. The kernel coefficient, nonlinear order, memory depth and order of low-order LFP of the BL-CFR module;

[0154] 3. Kernel coefficients, nonlinear order and memory depth of the error compensation module ECM.

[0155] S203: Obtain an initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal, the second OFDM signal, and the third OFDM signal.

[0156] Specifically, when executing step S203, after obtaining the first OFDM signal processed by the BL-DPD module, the second OFDM signal processed by the BL-CFR module, and the third OFDM signal processed by the error compensation module, the nonlinear compensation fusion model can obtain the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal, the second OFDM signal and the third OFDM signal.

[0157] For example, assuming that the initial OFDM signal is s(n), the kernel processing of the nonlinear compensation fusion model (i.e., the CBL-CFR-DPD model) obtains the initial OFDM signal z(n) after nonlinear distortion compensation. The calculation formula of the initial OFDM signal z(n) after nonlinear distortion compensation is as follows:

[0158] It should be noted that the model parameter sets converged by the BL-DPD module, BL-CFR module and error compensation module in offline mode are respectively obtained by the nonlinear compensation fusion model and the first branch, the nonlinear compensation fusion model and the second branch, and the nonlinear compensation fusion model and the third branch.

[0159] In an optional implementation, as shown in FIG4 , if the parameter combination of the model parameter set is: the kernel coefficient of the BL-DPD module, the nonlinear order, the memory depth, and the order of the low-order LFP, then the model parameter set is obtained as follows:

[0160] S401: Inputting a sample OFDM signal in an offline mode into a BL-DPD module to obtain a sample OFDM signal after DPD processing.

[0161] Exemplarily, when executing step S401 , assuming that the sample OFDM signal in the offline mode is s(n), the DPD-processed sample OFDM signal x(n) is obtained through the core processing of the BL-DPD module.

[0162] S402: Obtain a sample OFDM signal after inverse DPD processing based on the sample OFDM signal after DPD processing and a conjugate parameter set corresponding to the initial parameter set of the BL-DPD module.

[0163] In an optional implementation, referring to FIG5 , when executing step S402, the OFDM communication system obtains the sample OFDM signal after DPD processing, and then, through the first branch in the offline mode, sequentially performs digital-to-analog conversion, up-conversion, and power amplification processing on the sample OFDM signal after DPD processing to obtain the power-amplified sample OFDM signal; then, sequentially performs power attenuation, down-conversion, and analog-to-digital conversion processing on the power-amplified sample OFDM signal to obtain the analog-to-digital converted sample OFDM signal; finally, based on the conjugate parameter set corresponding to the analog-to-digital converted sample OFDM signal and the initial parameter set, obtains the sample OFDM signal after inverse DPD processing.

[0164] Exemplarily, first, the sample OFDM signal x(n) after DPD processing is processed by DAC and up-converted to the working frequency band of PA, and then the sample OFDM signal after frequency conversion is sent to PA, and part of the power-amplified sample OFDM signal is coupled back to the feedback channel; then, on the feedback channel of the first branch, the power-amplified sample OFDM signal first passes through an attenuator 1 / G to attenuate the power of the signal, and then down-converts and samples the ADC to obtain the feedback signal vector (i.e., the sample OFDM signal after analog-to-digital conversion) v(n); finally, the feedback signal vector v(n) is processed by the training network POST-BL-DPD model to obtain the sample OFDM signal after inverse DPD processing

[0165] Optional, sample OFDM signal after inverse DPD processing The calculation formula is as follows:

[0166] Where, It is the transpose of the (complex) conjugate parameter set corresponding to the initial parameter set w0 of the BL-DPD module; it should be noted that before executing step S401, the model parameter set of the training network POST-BL-DPD model needs to be initialized to the initial parameter set w0.

[0167] S403: Obtain a target DPD error signal based on the sample OFDM signal in the offline mode and the sample OFDM signal after inverse DPD processing.

[0168] Optionally, when executing step S403, the target DPD error signal is calculated as follows:

[0169] Where, e BL-DPD (n) represents the target DPD error signal, x(n) is the sample OFDM signal, is the sample OFDM signal after inverse DPD processing.

[0170] S404: Based on the target DPD error signal and a preset low-noise variable step-size least mean square algorithm, iteratively modify the initial parameter set until the absolute value of the target DPD error signal is less than a set DPD error signal threshold.

[0171] Exemplarily, the absolute value of the target DPD error signal is expressed as: |e BL-DPD (n)|, if the set DPD error signal threshold is e0, then when determining |e BL-DPDWhen (n)|<e0, the initial parameter set w0 is no longer iteratively modified based on the target DPD error signal and the preset LNVSS-LMS (adaptive) algorithm; among them, the DPD error signal threshold e0 is the convergence error judgment threshold, which can be 0.001.

[0172] In an optional implementation, referring to FIG6 , when executing step S405 , the OFDM communication system performs the following operations each time the initial parameter set w0 is modified:

[0173] S601: Acquire sample OFDM signals corresponding to a plurality of historical moments adjacent to a current sample OFDM signal in an offline mode.

[0174] Exemplarily, when executing step S601, the OFDM communication system obtains sample OFDM signals s(i) corresponding to multiple historical moments adjacent to the current sample OFDM signal s(n) in offline mode, for example, the first N sample OFDM signals s(i) are obtained, where i is an integer ∈(1, N).

[0175] S602: Based on the historical DPD error signals corresponding to the plurality of sample OFDM signals, obtain the DPD error signal mean value of the sample OFDM signals in the offline mode at the current moment.

[0176] Optionally, when executing step S602, the calculation formula of the DPD error signal mean is as follows:

[0177] Where, represents the mean value of the DPD error signal, e BL-DPD (i) represents the historical DPD error signal corresponding to the i-th sample OFDM signal.

[0178] S603: Obtain a first target step size factor based on the DPD error signal mean, the step size factor of the previous historical moment adjacent to the current moment, the target DPD error signal at the current moment, and the historical DPD error signal of the previous historical moment.

[0179] Optionally, when executing step S603, the calculation formula of the first target step size factor is expressed as follows:

[0180] In the formula, μ(n+1) represents the first target step size factor, μ(n) represents the step size factor of the previous historical moment adjacent to the current moment, is the mean value of the DPD error signal, e BL-DPD (n) represents the target DPD error signal at the current moment, e BL-DPD(n-1) represents the historical DPD error signal at the previous historical moment, abs(*) is the absolute value operation, and α is the forgetting factor of the step size. Usually, its value cannot exceed 1, otherwise the LNVSS-LMS algorithm will not converge. If the value is too small, when the LNVSS-LMS algorithm converges, the step size changes too quickly, resulting in a large steady-state error. Therefore, the value of α is less than 1 and close to 1; β determines the degree to which the step size factor is affected by the error. When the LNVSS-LMS algorithm converges, the error is small. In order to obtain a smaller μ value, the value of β is generally very small.

[0181] It should be noted that the above α and β are set in advance according to actual conditions; and the calculation formula of the first target step size factor can also be regarded as the update formula of the first target step size factor.

[0182] S604: Based on the first target step size factor, the conjugate DPD error signal corresponding to the target DPD error signal at the current moment, and the sample OFDM signal after analog-to-digital conversion, modify the initial parameter set at the current moment to obtain a modified initial parameter set.

[0183] Optionally, when executing step S604, the calculation formula of the modified initial parameter set is as follows:

[0184] Where w0(n+1) represents the modified initial parameter set, w0(n) represents the initial parameter set at the current moment, μ(n+1) is the first target step size factor, and v(n) represents the sample OFDM signal after analog-to-digital conversion. The target DPD error signal e at the current moment BL-DPD (n) The corresponding (complex) conjugate DPD error signal.

[0185] Obviously, based on the method described in steps S601 to S604 above, the iterative modification of the initial parameter set is achieved in offline mode, so that the BL-DPD module can complete the nonlinear distortion compensation of the initial OFDM signal based on a more accurate model parameter set.

[0186] S405: Using the iteratively modified initial parameter set as the model parameter set of the BL-DPD module.

[0187] Specifically, when executing step S405, when it is determined that the absolute value of the current target DPD error signal is less than the set DPD error signal threshold, the iteratively modified initial parameter set at this time can be used as the model parameter set of the BL-DPD module, and the model parameter set can be assigned to the BL-DPD module.

[0188] Optionally, before executing step S404 , if the absolute value of the target DPD error signal is not less than the DPD error signal threshold, the initial parameter set is directly used as the model parameter set of the BL-DPD module.

[0189] Therefore, based on the method described in steps S401 to S405 above, the offline mode is selected in the Mode of the OFDM communication system, the sample OFDM signal is processed by the nonlinear compensation fusion model and the first branch, and the POST-BL-DPD module is trained using the LNVSS-LMS adaptive algorithm, thereby realizing the modification and extraction of the model parameter set of the BL-DPD module.

[0190] In an optional implementation, as shown in FIG7 , if the parameter combination of the model parameter set is: the kernel coefficient of the BL-CFR module, the nonlinear order, the memory depth, and the order of the low-order LFP, then the model parameter set is obtained as follows:

[0191] S701: Inputting a sample OFDM signal in an offline mode into a BL-DPD module and a preset CFR module in sequence to obtain a sample OFDM signal after DPD-CFR processing.

[0192] For example, when executing step S701, assuming that the sample OFDM signal in the offline mode is s(n), the core processing of the BL-DPD module obtains the sample OFDM signal x(n) after DPD processing, and the sample OFDM signal x(n) after DPD processing is processed by the preset CFR module to obtain the sample OFDM signal x after DPD-CFR processing. c (n).

[0193] S702: Obtain a sample OFDM signal after CFR processing based on the sample OFDM signal in the offline mode and the conjugate parameter set corresponding to the initial parameter set of the BL-CFR module.

[0194] Optionally, when executing step S702, the calculation formula of the sample OFDM signal u(n) after CFR processing is specifically as follows:

[0195] Where, It is the transpose of the (complex) conjugate parameter set corresponding to the initial parameter set w1 of the BL-CFR module; it should be noted that before executing step S701, the model parameter set of the training network POST-BL-CFR model needs to be initialized to the initial parameter set w1.

[0196] It should be noted that when training the BL-CFR model, the improved CFR technology is used. When the input signal amplitude (i.e., the sample OFDM signal x(n) after DPD processing) is greater than the threshold P, the amplitude is assigned to the threshold amplitude, and its phase remains unchanged from the original signal phase. Therefore, the sample OFDM signal x after DPD-CFR processing is c (n) is calculated as follows:

[0197] Among them, abs(*) is the absolute value operation, and exp(*) is the exponential function.

[0198] S703: Obtain a target CFR error signal based on the sample OFDM signal after DPD-CFR processing and the sample OFDM signal after CFR processing.

[0199] Optionally, when executing step S703, the target CFR error signal is calculated as follows: BL-CFR (n) = u(n) - x c (n)

[0200] Where, e BL-CFR (n) represents the target CFR error signal, u(n) is the sample OFDM signal after CFR processing, and x c (n) is the sample OFDM signal after DPD-CFR processing.

[0201] S704: Based on the target CFR error signal and a preset low-noise variable step-size least mean square algorithm, iteratively modify the initial parameter set until the absolute value of the target CFR error signal is less than a set CFR error signal threshold.

[0202] Exemplarily, the absolute value of the target CFR error signal is expressed as: |e BL-CFR (n)|, if the set CFR error signal threshold is e1, then when determining |e BL-CFR When (n)|<e1, the initial parameter set w1 is no longer iteratively modified based on the target CFR error signal and the preset LNVSS-LMS (adaptive) algorithm; wherein the CFR error signal threshold e1 is the convergence error judgment threshold, which can be 0.001.

[0203] It should be noted that the specific process of each modification of the initial parameter set w1 by the OFDM communication system is consistent with the process of each modification of the initial parameter set w0 recorded in steps S601 to S604 above, so it will not be repeated in the embodiment of the present disclosure.

[0204] S705: Using the iteratively modified initial parameter set as the model parameter set of the BL-CFR module.

[0205] Specifically, when executing step S705, when it is determined that the absolute value of the current target CFR error signal is less than the set CFR error signal threshold, the iteratively modified initial parameter set can be used as the model parameter set of the BL-CFR module, and the model parameter set can be assigned to the BL-CFR module.

[0206] Optionally, before executing step 704 , if the absolute value of the target CFR error signal is not less than the CFR error signal threshold, the initial parameter set is directly used as the model parameter set of the BL-CFR module.

[0207] Therefore, based on the method described in steps S701 to S705 above, the offline mode is selected in Mode in the OFDM communication system, the sample OFDM signal is processed by the nonlinear compensation fusion model and the second branch, and the POST-BL-CFR module is trained using the LNVSS-LMS adaptive algorithm, thereby realizing the modification and extraction of the model parameter set of the BL-CFR module.

[0208] In an optional implementation, as shown in FIG8 , if the parameter combination of the model parameter set is: the kernel coefficient, the nonlinear order, and the memory depth of the error compensation module ECM, then the model parameter set is obtained in the following manner:

[0209] S801: Inputting the sample OFDM signal in the offline mode into the nonlinear compensation fusion model and the preset high-order LPF respectively to obtain the sample OFDM signal after nonlinear distortion compensation and the sample OFDM signal after filtering.

[0210] It should be noted that when obtaining the model parameter set of the error compensation module ECM, ie, when executing the method described in steps S801 to S804, the sample OFDM signal in the offline mode is small, ie, has a small amplitude, and is in the linear region of the PA.

[0211] In addition, to reduce the complexity of the model, the BL-DPD module and the BL-CFRM module are generated through low-order FIR processing. In order to compensate for the signal loss caused by the low-order FIR, the error compensation parameters are trained using high-order FIR in an offline state, which effectively reduces the order of the band-limited FIR, thereby reducing the complexity of the entire nonlinear distortion fusion model while ensuring the accuracy and performance of the entire model.

[0212] Optionally, when executing step S801, assuming that the sample OFDM signal in the offline mode is s(n), the calculation formula of the sample OFDM signal z(n) after nonlinear distortion compensation is specifically as follows:

[0213] Where, is the transpose of the (complex) conjugate parameter set corresponding to the initial parameter set w0 of the BL-DPD module, is the transpose of the (complex) conjugate parameter set corresponding to the initial parameter set w1 of the BL-CFR module, is the transpose of the (complex) conjugate parameter set corresponding to the initial parameter set w2 of the error compensation module ECM; it should be noted that before executing step S801, the model parameter set of the error compensation module ECM needs to be initialized to the initial parameter set w2.

[0214] Sample OFDM signal s after filtering L The calculation formula of (n) is as follows:

[0215] Wherein, h represents the coefficient of the above-mentioned preset high-order LPF, and L′ represents the order of the above-mentioned preset high-order LPF; optionally, the order L′ is not less than 91.

[0216] S802: Obtain a target compensation error signal based on the sample OFDM signal after nonlinear distortion compensation and the sample OFDM signal after filtering.

[0217] Optionally, when executing step S802, the target compensation error signal is calculated as follows: ECM (n) = z(n) - s L (n)

[0218] Where, e ECM (n) represents the target compensation error signal, z(n) is the sample OFDM signal after nonlinear distortion compensation, s L (n) is the sample OFDM signal after filtering.

[0219] S803: Based on the target compensation error signal and a preset low-noise variable step-size least mean square algorithm, iteratively modify the initial parameter set of the error compensation module until the absolute value of the target compensation error signal is less than a set compensation error signal threshold.

[0220] Exemplarily, the absolute value of the target compensation error signal is expressed as: |e ECM (n)|, if the compensation error signal threshold is set to e2, then when determining |e ECM When (n)|<e2, the initial parameter set w2 is no longer iteratively modified based on the target compensation error signal and the preset LNVSS-LMS (adaptive) algorithm; among which, the compensation error signal threshold e2 is the convergence error judgment threshold, which can be 0.001.

[0221] It should be noted that the specific process of each modification of the initial parameter set w2 by the OFDM communication system is consistent with the process of each modification of the initial parameter set w0 recorded in steps S601 to S604 above, so it will not be repeated in the embodiment of the present disclosure.

[0222] S804: Using the iteratively modified initial parameter set as the model parameter set of the error compensation module.

[0223] Specifically, when executing step S805, when it is determined that the absolute value of the current target compensation error signal is less than the set compensation error signal threshold, the iteratively modified initial parameter set at this time can be used as the model parameter set of the error compensation module ECM, and the model parameter set can be assigned to the error compensation module ECM.

[0224] Optionally, before executing step S804 , if the absolute value of the target compensation error signal is not less than the compensation error signal threshold, the initial parameter set is directly used as the model parameter set of the error compensation module ECM.

[0225] Therefore, based on the method described in steps 801 to S804 above, the offline mode is selected in the Mode of the OFDM communication system, the sample OFDM signal is processed by the nonlinear compensation fusion model and the third branch, and the CBL-CFR-DPD module is trained using the LNVSS-LMS adaptive algorithm, thereby realizing the modification and extraction of the model parameter set of the error compensation module ECM.

[0226] Furthermore, referring to FIG9 , after the OFDM communication system obtains the initial OFDM signal after the first nonlinear distortion compensation through the nonlinear compensation fusion model, it can further refresh / modify the model parameter set of the BL-DPD module. The specific operation process is as follows:

[0227] S901: Based on the initial OFDM signal after nonlinear distortion compensation and the conjugate parameter set corresponding to the model parameter set of the BL-DPD module, obtain the initial OFDM signal after inverse DPD processing.

[0228] In an optional implementation, referring to FIG10 , after obtaining the initial OFDM signal after nonlinear distortion compensation, the OFDM communication system can, through the fourth branch in online mode, sequentially perform digital-to-analog conversion, up-conversion, and power amplification processing on the initial OFDM signal after nonlinear distortion compensation to obtain the initial OFDM signal after power amplification; then, sequentially perform power attenuation, down-conversion, and analog-to-digital conversion processing on the initial OFDM signal after power amplification to obtain the initial OFDM signal after analog-to-digital conversion; finally, based on the initial OFDM signal after analog-to-digital conversion and the conjugate parameter set corresponding to the model parameter set, obtain the initial OFDM signal after inverse DPD processing.

[0229] For example, assuming that the initial OFDM signal after nonlinear distortion compensation is z′(n), the initial OFDM signal z′(n) after nonlinear distortion compensation is first processed by DAC and up-converted to the working frequency band of PA, and then the initial OFDM signal after frequency conversion is sent to PA, and part of the initial OFDM signal after power amplification is coupled back to the feedback channel; then, on the feedback channel of the fourth branch, the initial OFDM signal after power amplification first passes through an attenuator 1 / G to attenuate the power of the signal, and then down-converts and samples the ADC to obtain the feedback signal vector (i.e., the initial OFDM signal after analog-to-digital conversion) v′(n); finally, the feedback signal vector v′(n) is processed by the training network POST-BL-DPD model to obtain the initial OFDM signal after inverse DPD processing.

[0230] Optional, initial OFDM signal after inverse DPD processing The calculation formula is as follows:

[0231] Where, It is the transpose of the (complex) conjugate parameter set corresponding to the model parameter set w0 when the BL-DPD module converges in offline mode; it should be noted that before executing step S901, the model parameter set of the training network POST-BL-DPD model needs to be initialized to the model parameter set w0 when the offline mode converges.

[0232] S902: Obtain a target distortion compensation error signal based on the initial OFDM signal after nonlinear distortion compensation and the initial OFDM signal after inverse DPD processing.

[0233] Optionally, when executing step S902, the target distortion compensation error signal is calculated as follows:

[0234] Where eCBL-CFR-DPD(n) represents the target distortion compensation error signal, z′(n) is the initial OFDM signal after nonlinear distortion compensation, is the initial OFDM signal after inverse DPD processing.

[0235] S903: Based on the target distortion compensation error signal and the preset sine sum error variable step-least mean square algorithm, iteratively modify the model parameter set until the absolute value of the target distortion compensation error signal is less than the set distortion compensation error threshold.

[0236] Exemplarily, the absolute value of the target distortion compensation error signal is expressed as: |eCBL-CFR-DPD(n)|. If the set distortion compensation error signal threshold is e3, then when it is determined that |eCBL-CFR-DPD(n)|<e3, the model parameter set w0 during offline mode convergence is no longer iteratively modified based on the target distortion compensation error signal and the preset SEVSS-LMS (adaptive) algorithm; wherein the distortion compensation error signal threshold is e3, i.e., the refresh error judgment threshold, which can be 0.01.

[0237] It can be seen that based on the above method, in offline mode, there is no need to consider issues such as convergence time and complexity. Therefore, it is necessary to adopt an adaptive identification algorithm with better performance, that is, the LNVSS-LMS adaptive algorithm, and by setting smaller threshold convergence thresholds (i.e., e0, e1, and e2), the BL-DPD, BL-CFRM, and ECM module parameters (i.e., the model parameter set of each module) are extracted, thereby improving the model accuracy; in online mode, it is necessary to consider convergence time and complexity, and since the parameters have been converged offline, the parameters only need to be converged in a small range in the online state. Therefore, the SEVSS-LMS adaptive algorithm is used to set a larger threshold convergence threshold (i.e., e3) to effectively refresh the CBL-CFR-DPD model parameters (i.e., the model parameter set of the BL-DPD module), which can quickly converge.

[0238] In addition, a new CBL-CFR-DPD integrated solution is constructed, adopting an "offline + online" mechanism for extracting and updating the parameters of each module. That is, in offline mode, the parameters of each module are extracted, effectively improving work efficiency; in online mode, the CBL-CFR-DPD module parameters are effectively refreshed, which can quickly pre-distort the broadband signal, thereby effectively improving the nonlinear performance of the digital pre-distorter in compensating for the broadband PA in engineering. Moreover, the BL-DPD module and the BL-CFRM module are truly integrated into one module through the error compensation module ECM. Therefore, the operating complexity of this model is comparable to that of a system that only performs DPD, except that some coefficient addition and subtraction operations are added, which reduces the operating complexity compared with the classic solution where the CFR module and DPD module are applied independently.

[0239] In an optional implementation, referring to FIG. 11 , when executing step S903 , the OFDM communication system performs the following operations each time the model parameter set w0 is modified during offline mode convergence:

[0240] S1101: Obtain sample OFDM signals after nonlinear distortion compensation corresponding to a plurality of historical moments adjacent to the initial OFDM signal after nonlinear distortion compensation.

[0241] Exemplarily, when executing step S1101, the OFDM communication system obtains sample OFDM signals z′(i) after nonlinear distortion compensation corresponding to multiple historical moments adjacent to the initial OFDM signal z′(n) after nonlinear distortion compensation. For example, the first N sample OFDM signals z′(i) after nonlinear distortion compensation are obtained, where i is an integer ∈(1, N).

[0242] S1102: Based on a plurality of sample OFDM signals after nonlinear distortion compensation and their corresponding historical distortion compensation error signals, obtain a distortion compensation error signal mean value of the initial OFDM signal after nonlinear distortion compensation at a current moment.

[0243] Optionally, when executing step S1102, the calculation formula of the distortion compensation error signal mean value is specifically as follows:

[0244] Where, represents the mean value of the distortion compensation error signal, and eCBL-CFR-DPD(i) represents the historical distortion compensation error signal corresponding to the i-th sample OFDM signal.

[0245] S1103: Obtain a second target step size factor based on the error term corresponding to the mean value of the distortion compensation error signal, the target distortion compensation error signal at the current moment, and the historical distortion compensation error signal at the previous historical moment adjacent to the current moment.

[0246] Optionally, when executing step S1103, the calculation formula of the second target step size factor is expressed as follows:

[0247] Where μ(n) represents the second target step size factor, eCBL-CFR-DPD(n) represents the target distortion compensation error signal at the current moment, and eCBL-CFR-DPD(n-1) represents the historical distortion compensation error signal at the previous historical moment adjacent to the current moment. represents the error term corresponding to the mean value of the distortion compensation error signal, erf(*) is the error function; a, b and c are adjustment factors, which are pre-set according to actual conditions; and the above-mentioned calculation formula for the second target step size factor can also be regarded as the update formula for the second target step size factor.

[0248] S1104: Based on the second target step size factor, the conjugate distortion compensation error signal corresponding to the target distortion compensation error signal at the current moment, and the initial OFDM signal after analog-to-digital conversion, the model parameter set at the current moment is modified to obtain a modified model parameter set.

[0249] Optionally, when executing step S1104, the calculation formula of the modified model parameter set is as follows:

[0250] Wherein, w0(n+1) represents the modified model parameter set, w0(n) represents the model parameter set at the current moment, μ(n) is the second target step size factor, and v′(n) represents the initial OFDM signal after analog-to-digital conversion. It represents the (complex) conjugate distortion compensation error signal corresponding to the current target distortion compensation error signal eCBL-CFR-DPD(n).

[0251] Obviously, based on the method described in steps S1101 to S1104 above, the iterative modification of the model parameter set of the BL-DPD module is realized in the online mode, so that the BL-DPD module can complete the nonlinear distortion compensation of the subsequent OFDM signal based on a more accurate model parameter set.

[0252] Optionally, in online mode, the OFDM communication system can iteratively modify the model parameter set of the BL-DPD module according to a set period time, that is, the initial OFDM signal can be processed in the online mode branch (i.e., the nonlinear compensation fusion model and the fourth branch) according to the preset period T.

[0253] Furthermore, when it is determined that the absolute value of the current target distortion compensation error signal is less than the set distortion compensation error threshold, the iteratively modified model parameter set can be used as a new model parameter set of the BL-DPD module, and the model parameter set can be assigned to the BL-DPD module.

[0254] Therefore, based on the method described in steps S901 to S903 above, in the OFDM communication system, the Mode is switched to the online mode, and the initial OFDM signal is processed by the online branches (nonlinear compensation fusion model and the fourth branch) according to the preset period T, and the SEVSS-LMS adaptive algorithm is adopted. The parameters of the nonlinear compensation fusion model (i.e., the CBL-CFR-DPD model) (i.e., the model parameter set of the BL-DPD module) can be effectively refreshed / modified.

[0255] In summary, in the signal nonlinear distortion compensation method provided in the embodiment of the present disclosure, based on the preset nonlinear compensation fusion model, the initial OFDM signal is subjected to DPD processing, CFR processing and error compensation processing respectively, thereby obtaining the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal processed by the BL-DPD module, the second OFDM signal processed by the BL-CFR module, and the third OFDM signal processed by the error compensation module.

[0256] In this way, by connecting the BL-DPD module, BL-CFR module and error compensation module in parallel in the preset nonlinear compensation fusion model, the technical disadvantages of the prior art, in which the DPD module is applied after the CFR module, which places high requirements on the sampling rate of the ADC / DAC in the OFDM system, increases the requirements on the hardware and algorithm convergence speed, and increases the difficulty and cost of system implementation, are avoided; therefore, the requirements on the ADC / DAC sampling rate are effectively reduced, and the DPD compensation performance for the nonlinearity of the power amplifier is improved, thereby further improving the communication performance and perception performance of the OFDM system.

[0257] Based on the same inventive concept, the embodiment of the present application further provides a nonlinear compensation fusion model, including: a BL-DPD module, a BL-CFR module and an error compensation module;

[0258] The BL-DPD module, the BL-CFR module and the error compensation module are connected in parallel; wherein, the BL-DPD module and the BL-CFR module each use the same basis function, and the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module.

[0259] In an optional embodiment, the BL-DPD module and the BL-CFR module each use the same basis function.

[0260] In an optional embodiment, the nonlinear compensation fusion model is used to:

[0261] Based on the initial OFDM signal, obtaining a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module;

[0262] An initial OFDM signal after nonlinear distortion compensation is obtained based on the first OFDM signal, the second OFDM signal, and the third OFDM signal.

[0263] In an optional embodiment, the nonlinear compensation fusion model is further used to:

[0264] In offline mode, the preset low-noise variable step-size least mean square algorithm is used to extract the model parameter set of the BL-DPD module, BL-CFR module and error compensation module.

[0265] In an optional embodiment, the nonlinear compensation fusion model is further used to:

[0266] In online mode, the preset sine sum error variable step-least mean square algorithm and the set cycle time are used to refresh the model parameter set of the nonlinear compensation fusion model.

[0267] In an optional embodiment, the nonlinear compensation fusion model is specifically used to:

[0268] In online mode, the preset sine sum error variable step-least mean square algorithm and the set cycle time are used to refresh the model parameter set only for the BL-DPD module in the nonlinear compensation fusion model.

[0269] Based on the same inventive concept, an embodiment of the present application further provides an OFDM communication system, comprising: a nonlinear compensation fusion model, a first branch, a second branch, a third branch, and a fourth branch;

[0270] The BL-DPD module and the first branch in the nonlinear compensation fusion model are used in an offline mode to iteratively modify the initial parameter set of the BL-DPD module to obtain a model parameter set of the BL-DPD module;

[0271] The BL-CFR module and the second branch in the nonlinear compensation fusion model are used in offline mode to iteratively modify the initial parameter set of the BL-CFR module to obtain the model parameter set of the BL-CFR module;

[0272] The nonlinear compensation fusion model and the third branch are used in an offline mode to iteratively modify the initial parameter set of the error compensation module to obtain the model parameter set of the error compensation module;

[0273] The nonlinear compensation fusion model and the fourth branch are used to iteratively modify the model parameter set of the BL-DPD module in online mode.

[0274] In an optional embodiment, the first branch includes: a digital-to-analog converter DAC, an up-converter, a power amplifier PA, an attenuator, a bandpass filter BPF, a down-converter, an analog-to-digital converter ADC, a training network POST-BL-DPD module, and a preset low-noise variable step size-least mean square algorithm module arranged in sequence.

[0275] In an optional embodiment, the second branch includes: a CFR module, a training network POST-BL-DPD module, and a preset low-noise variable step size-least mean square algorithm module arranged in sequence.

[0276] In an optional embodiment, the third branch includes: a low-pass filter LPF and a preset low-noise variable step-size-least mean square algorithm module arranged in sequence.

[0277] In an optional embodiment, the fourth branch includes: a DAC, an up-converter, a PA, an attenuator, a BPF, a down-converter, an ADC, a training network POST-BL-DPD module, and a preset sine and error variable step-least mean square algorithm module arranged in sequence.

[0278] Based on the same technical concept, the present disclosure also provides an electronic device that can implement the signal nonlinear distortion compensation method provided in the above-mentioned embodiments of the present disclosure. In one embodiment, the electronic device can be a server, a terminal device, or other electronic device. Referring to FIG12 , the electronic device may include:

[0279] At least one processor 1201, and a memory 1202 connected to at least one processor 1201. The specific connection medium between the processor 1201 and the memory 1202 is not limited in the embodiments of the present disclosure. Figure 12 takes the connection between the processor 1201 and the memory 1202 via the bus 1200 as an example. The bus 1200 is represented by a bold line in Figure 12, and the connection between other components is only for schematic illustration and is not intended to be limiting. The bus 1200 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bold line is used in Figure 12, but this does not mean that there is only one bus or one type of bus. Alternatively, the processor 1201 can also be called a controller, and there is no limitation on the name.

[0280] In the embodiment of the present disclosure, memory 1202 stores instructions executable by at least one processor 1201. At least one processor 1201 can execute the aforementioned method for compensating for nonlinear signal distortion by executing the instructions stored in memory 1202. Processor 1201 can implement the functions of each module in the corresponding device.

[0281] Among them, the processor 1201 is the control center of the device, which can use various interfaces and lines to connect the various parts of the entire control device, and monitor the device as a whole by running or executing instructions stored in the memory 1202 and calling data stored in the memory 1202, the various functions of the device and processing data.

[0282] In one possible design, processor 1201 may include one or more processing units. Processor 1201 may integrate an application processor and a modem processor. The application processor primarily processes the operating system, user interface, and application programs, while the modem processor primarily processes wireless communications. It is understood that the modem processor may not be integrated into processor 1201. In some embodiments, processor 1201 and memory 1202 may be implemented on the same chip. In some embodiments, they may also be implemented on separate chips.

[0283] Processor 1201 can be a general-purpose processor, such as a CPU, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the signal nonlinear distortion compensation method disclosed in conjunction with the embodiments of this disclosure can be directly implemented and executed by a hardware processor, or by a combination of hardware and software modules in the processor.

[0284] The memory 1202 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 1202 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 1202 is any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 1202 in the embodiment of the present disclosure can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0285] By designing and programming the processor 1201, the code corresponding to the method for compensating for nonlinear signal distortion described in the aforementioned embodiment can be embedded in the chip, thereby enabling the chip to execute the steps of the method for compensating for nonlinear signal distortion described in the embodiment of FIG. 2 during operation. Designing and programming the processor 1201 is well known to those skilled in the art and will not be further described here.

[0286] Based on the same inventive concept, an embodiment of the present disclosure further provides a storage medium storing computer instructions. When the computer instructions are executed on a computer, the computer executes a method for compensating for nonlinear distortion of a signal discussed above.

[0287] In some optional embodiments, the present disclosure also provides various aspects of a method for compensating for nonlinear distortion of a signal, which can also be implemented in the form of a program product, which includes program code. When the program product is run on an apparatus, the program code is used to enable the control device to execute the steps of a method for compensating for nonlinear distortion of a signal according to various exemplary embodiments of the present disclosure described above in this specification.

[0288] It should be noted that although several units or subunits of the device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0289] Furthermore, although the operations of the disclosed method are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all illustrated operations must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0290] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0291] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a server, so that the instructions executed by the processor of the computer or other programmable data processing device generate a device for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0292] Program code for performing the operations of the present disclosure may be written using any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may execute entirely on the user's computing device, partially on the user's device, as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0293] Where a remote computing device is involved, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0294] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0295] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0296] Obviously, those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include these modifications and variations.

Claims

1. A method for compensating signal nonlinear distortion, It is characterized in that include: Inputting the initial orthogonal frequency division multiplexing OFDM signal into a preset nonlinear compensation fusion model; wherein the nonlinear compensation fusion model includes: a band-limited-digital predistortion BL-DPD module, a band-limited-peak factor reduction BL-CFR module and an error compensation module, wherein the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module; Respectively obtaining, for the initial OFDM signal, a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module; An initial OFDM signal after nonlinear distortion compensation is obtained based on the first OFDM signal, the second OFDM signal and the third OFDM signal.

2. The method according to claim 1, It is characterized in that The BL-DPD module and the BL-CFR module respectively use the same basis function.

3. The method according to claim 1, It is characterized in that The step of respectively obtaining, for the initial OFDM signal, a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module comprises: The initial OFDM signal is modulated based on the model parameter sets converged by the BL-DPD module, the BL-CFR module and the error compensation module in the offline mode to obtain the first OFDM signal, the second OFDM signal and the third OFDM signal.

4. The method according to claim 3, It is characterized in that The model parameter set includes any of the following parameter combinations: The kernel coefficients, nonlinear order, memory depth and order of the low-order low-pass filter LFP of the BL-DPD module; The kernel coefficients, nonlinear order, memory depth and order of low-order LFP of the BL-CFR module; The kernel coefficients, nonlinear order and memory depth of the error compensation module.

5. The method according to claim 4, It is characterized in that If the parameter combination of the model parameter set is: the kernel coefficient, the nonlinear order, the memory depth and the order of the low-order LFP of the BL-DPD module, then the model parameter set is obtained in the following manner: Inputting the sample OFDM signal in the offline mode into the BL-DPD module to obtain the sample OFDM signal after DPD processing; Based on the sample OFDM signal after the DPD processing and the conjugate parameter set corresponding to the initial parameter set of the BL-DPD module, obtaining the sample OFDM signal after the inverse DPD processing; Obtaining a target DPD error signal based on the sample OFDM signal in the offline mode and the sample OFDM signal after the inverse DPD processing; Iteratively modifying the initial parameter set based on the target DPD error signal and a preset low-noise variable step-size least mean square algorithm until the absolute value of the target DPD error signal is less than a set DPD error signal threshold; The iteratively modified initial parameter set is used as the model parameter set of the BL-DPD module.

6. The method according to claim 5, It is characterized in that The obtaining of the sample OFDM signal after inverse DPD processing based on the sample OFDM signal after DPD processing and the conjugate parameter set corresponding to the initial parameter set of the BL-DPD module includes: The sample OFDM signal after DPD processing is subjected to digital-to-analog conversion, up-conversion and power amplification processing in sequence to obtain a power amplification signal. The sample OFDM signal after The power-amplified sample OFDM signal is subjected to power attenuation, frequency down-conversion and analog-to-digital conversion in sequence to obtain an analog-to-digital converted sample OFDM signal; Based on the sample OFDM signal after analog-to-digital conversion and the conjugate parameter set corresponding to the initial parameter set, the sample OFDM signal after inverse DPD processing is obtained.

7. The method according to claim 5, It is characterized in that The iterative modification of the initial parameter set based on the target DPD error signal and a preset low-noise variable step-size-least mean square algorithm includes: In each process of modifying the initial parameter set, the following operations are performed: Acquire sample OFDM signals corresponding to a plurality of historical moments that are currently adjacent to the sample OFDM signal in the offline mode; Based on the historical DPD error signals corresponding to the multiple sample OFDM signals, respectively, obtaining the DPD error signal mean value of the sample OFDM signal in the offline mode at the current moment; Obtaining a first target step factor based on the DPD error signal mean, the step factor of the previous historical moment adjacent to the current moment, the target DPD error signal at the current moment, and the historical DPD error signal at the previous historical moment; Based on the first target step factor, the conjugate DPD error signal corresponding to the target DPD error signal at the current moment, and the sample OFDM signal after analog-to-digital conversion, the initial parameter set at the current moment is modified to obtain a modified initial parameter set.

8. The method according to claim 5, It is characterized in that Before iteratively modifying the initial parameter set based on the target DPD error signal and the preset low-noise variable step-size minimum mean square algorithm, the method further includes: If the absolute value of the target DPD error signal is not less than the DPD error signal threshold, the initial parameter set is directly used as the model parameter set of the BL-DPD module.

9. The method according to claim 4, It is characterized in that If the parameter combination of the model parameter set is: the kernel coefficient, the nonlinear order, the memory depth and the order of the low-order LFP of the BL-CFR module, then the model parameter set is obtained in the following manner: Inputting the sample OFDM signal in the offline mode into the BL-DPD module and the preset CFR module in sequence to obtain the sample OFDM signal after DPD-CFR processing; Based on the sample OFDM signal in the offline mode and the conjugate parameter set corresponding to the initial parameter set of the BL-CFR module, obtaining the sample OFDM signal after CFR processing; Obtaining a target CFR error signal based on the sample OFDM signal after the DPD-CFR processing and the sample OFDM signal after the CFR processing; Iteratively modifying the initial parameter set based on the target CFR error signal and a preset low-noise variable step-size-least mean square algorithm until the absolute value of the target CFR error signal is less than a set CFR error signal threshold; The iteratively modified initial parameter set is used as the model parameter set of the BL-CFR module.

10. The method according to claim 9, It is characterized in that Before iteratively modifying the initial parameter set based on the target CFR error signal and the preset low-noise variable step-size minimum mean square algorithm, the method further includes: If the absolute value of the target CFR error signal is not less than the CFR error signal threshold, the initial parameter set is directly used as is a set of model parameters of the BL-CFR module.

11. The method according to claim 4, It is characterized in that If the parameter combination of the model parameter set is: the kernel coefficient, the nonlinear order and the memory depth of the error compensation module, then the model parameter set is obtained in the following manner: Inputting the sample OFDM signal in the offline mode into the nonlinear compensation fusion model and the preset high-order LPF respectively, to obtain the sample OFDM signal after nonlinear distortion compensation and the sample OFDM signal after filtering; Obtaining a target compensation error signal based on the sample OFDM signal after the nonlinear distortion compensation and the sample OFDM signal after the filtering process; Based on the target compensation error signal and a preset low-noise variable step-size-least mean square algorithm, iteratively modifying the initial parameter set of the error compensation module until the absolute value of the target compensation error signal is less than a set compensation error signal threshold; The iteratively modified initial parameter set is used as the model parameter set of the error compensation module.

12. The method according to claim 11, It is characterized in that Before iteratively modifying the initial parameter set of the error compensation module based on the target compensation error signal and the preset low-noise variable step-size-least mean square algorithm, the method further includes: If the absolute value of the target compensation error signal is not less than the compensation error signal threshold, the initial parameter set is directly used as the model parameter set of the error compensation module.

13. The method according to any one of claims 1 to 12, It is characterized in that After obtaining the initial OFDM signal after nonlinear distortion compensation based on the first OFDM signal, the second OFDM signal and the third OFDM signal, the method further includes: Based on the initial OFDM signal after nonlinear distortion compensation and the conjugate parameter set corresponding to the model parameter set of the BL-DPD module, obtaining the initial OFDM signal after inverse DPD processing; Obtaining a target distortion compensation error signal based on the initial OFDM signal after the nonlinear distortion compensation and the initial OFDM signal after the inverse DPD processing; Based on the target distortion compensation error signal and a preset sine sum error variable step-least mean square algorithm, the model parameter set is iteratively modified until the absolute value of the target distortion compensation error signal is less than a set distortion compensation error threshold.

14. The method according to claim 13, It is characterized in that The method of obtaining the initial OFDM signal after inverse DPD processing based on the initial OFDM signal after nonlinear distortion compensation and the conjugate parameter set corresponding to the model parameter set of the BL-DPD module includes: The initial OFDM signal after nonlinear distortion compensation is subjected to digital-to-analog conversion, up-conversion and power amplification processing in sequence to obtain the initial OFDM signal after power amplification; The power-amplified initial OFDM signal is subjected to power attenuation, frequency down-conversion and analog-to-digital conversion in sequence to obtain an initial OFDM signal after analog-to-digital conversion; Based on the initial OFDM signal after analog-to-digital conversion and a conjugate parameter set corresponding to the model parameter set, the initial OFDM signal after inverse DPD processing is obtained.

15. The method of claim 13, It is characterized in that The iterative modification of the model parameter set based on the target distortion compensation error signal and a preset sine sum error variable step-minimum mean square algorithm includes: In each modification process of the model parameter set, the following operations are performed: Obtain the nonlinear distortion compensation of the initial OFDM signal after the nonlinear distortion compensation, and the corresponding nonlinear distortion compensation of the multiple historical moments. Sample OFDM signal after distortion compensation; Based on multiple nonlinear distortion compensated sample OFDM signals and their corresponding historical distortion compensation error signals, obtaining a distortion compensation error signal mean value of the initial OFDM signal after nonlinear distortion compensation at a current moment; Obtaining a second target step size factor based on an error term corresponding to the mean value of the distortion compensation error signal, the target distortion compensation error signal at the current moment, and a historical distortion compensation error signal at a previous historical moment adjacent to the current moment; Based on the second target step factor, the conjugate distortion compensation error signal corresponding to the target distortion compensation error signal at the current moment, and the initial OFDM signal after analog-to-digital conversion, the model parameter set at the current moment is modified to obtain a modified model parameter set.

16. The method of claim 13, It is characterized in that The method further comprises: The model parameter set of the BL-DPD module is iteratively modified according to the set cycle time.

17. A nonlinear compensation fusion model, It is characterized in that include: BL-DPD module, BL-CFR module and error compensation module; The BL-DPD module, the BL-CFR module and the error compensation module are connected in parallel; wherein the error compensation module is used to perform error compensation on the OFDM signals output by the BL-DPD module and the BL-CFR module.

18. The nonlinear compensation fusion model according to claim 17, It is characterized in that The BL-DPD module and the BL-CFR module respectively use the same basis function.

19. The nonlinear compensation fusion model according to claim 17, It is characterized in that The nonlinear compensation fusion model is used for: Based on the initial OFDM signal, obtaining a first OFDM signal processed by the BL-DPD module, a second OFDM signal processed by the BL-CFR module, and a third OFDM signal processed by the error compensation module; An initial OFDM signal after nonlinear distortion compensation is obtained based on the first OFDM signal, the second OFDM signal and the third OFDM signal.

20. The nonlinear compensation fusion model according to claim 17, It is characterized in that The nonlinear compensation fusion model is also used for: In the offline mode, a preset low-noise variable step-size-least mean square algorithm is used to extract a set of model parameters for the BL-DPD module, the BL-CFR module and the error compensation module.

21. The nonlinear compensation fusion model according to claim 17, It is characterized in that The nonlinear compensation fusion model is also used for: In the online mode, a preset sinusoidal and error variable step-minimum mean square algorithm and a set cycle time are used to refresh the model parameter set of the nonlinear compensation fusion model.

22. The nonlinear compensation fusion model according to claim 21, It is characterized in that The nonlinear compensation fusion model is specifically used for: In the online mode, the preset sine sum error variable step-minimum mean square algorithm and the set cycle time are used to refresh the model parameter set only for the BL-DPD module in the nonlinear compensation fusion model.

23. An OFDM communication system, It is characterized in that include: The nonlinear compensation fusion model, the first branch, the second branch, the third branch and the fourth branch as described in any one of claims 17 to 22; The BL-DPD module and the first branch in the nonlinear compensation fusion model are used in an offline mode to iteratively modify the initial parameter set of the BL-DPD module to obtain a model parameter set of the BL-DPD module; The BL-CFR module and the second branch in the nonlinear compensation fusion model are used in an offline mode to iteratively modify the initial parameter set of the BL-CFR module to obtain a model parameter set of the BL-CFR module; The nonlinear compensation fusion model and the third branch are used in an offline mode to iteratively modify the initial parameter set of the error compensation module to obtain a model parameter set of the error compensation module; The nonlinear compensation fusion model and the fourth branch are used to iteratively modify the model parameter set of the BL-DPD module in an online mode.

24. The OFDM communication system according to claim 23, It is characterized in that The first branch includes: a digital-to-analog converter DAC, an up-converter, a power amplifier PA, an attenuator, a bandpass filter BPF, a down-converter, an analog-to-digital converter ADC, a training network POST-BL-DPD module and a preset low-noise variable step-size-minimum mean square algorithm module arranged in sequence.

25. The OFDM communication system according to claim 23, It is characterized in that The second branch includes: a CFR module, a training network POST-BL-DPD module and a preset low-noise variable step-size-least mean square algorithm module which are arranged in sequence.

26. The OFDM communication system according to claim 23, It is characterized in that The third branch includes: a low-pass filter LPF and a preset low-noise variable step-length-minimum mean square algorithm module which are arranged in sequence.

27. The OFDM communication system according to claim 24, It is characterized in that The fourth branch includes: the DAC, the up-converter, the PA, the attenuator, the BPF, the down-converter, the ADC, the training network POST-BL-DPD module and a preset sine and error variable step-minimum mean square algorithm module arranged in sequence.

28. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the method according to any one of claims 1 to 16 is implemented.

29. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 16 are implemented.

Images