A pre-distortion-based EVM compensation method for a satellite-borne transmit front-end

Abstract

The application provides a pre-distortion-based EVM compensation method for a satellite-borne transmitting front end, wherein the variable frequency gain of the satellite-borne transmitting front end at different temperatures Tn and different output frequencies is calibrated on the ground to obtain an amplitude-frequency response database and store the database in a channel data storage of the satellite-borne transmitting front end; when in orbit, an FPGA queries the amplitude-frequency response database to obtain a variable frequency gain data set and form a gain-frequency curve, the gain-frequency curve is fitted to obtain a channel amplitude-frequency gain coefficient of a signal in a frequency spectrum shift process in a current state, and the channel amplitude-frequency gain coefficient is used to make the FPGA perform pre-distortion processing on an input intermediate frequency signal to compensate for the EVM caused by the in-band amplitude unevenness. The application designs a pre-distortion algorithm by reading actual measurement parameters, improves the correction accuracy compared with a traditional theoretical pre-distortion algorithm, establishes a frequency index mechanism between the FPGA and the storage, proposes a standardized data reading method, and creates conditions for the universality implementation of the pre-distortion algorithm.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data technology, and specifically to a pre-distortion-based spaceborne transmitter front-end EVM compensation method. Background Technology

[0002] In space communication, the Ka band offers advantages over other bands, including large available communication capacity, wide bandwidth, high transmission quality, and small equipment size, making it the primary band used in space communication. To fully utilize this spectrum resource, higher-order modulation methods are widely adopted. The parameter used to measure the quality of the RF signal output from the transmitting front-end is the error vector amplitude (EVM), which directly determines the final demodulation bit error rate performance.

[0003] In the Ka-band transmitter front-end, the RF link generation section includes a digital baseband module and a Ka-band up-conversion channel. The parameters of the Ka-band up-conversion channel directly determine the differences in the EVM (Electronic Performance Measurement) of the modulated signal. The main influencing factors include: 1) nonlinear distortion in the link (including nonlinearity in the mixer and final stage amplifier); and 2) amplitude-frequency response introduced by in-band non-flatness caused by various stages of components, including filters. Other patents have proposed methods to compensate for the EVM degradation caused by the former; however, there is currently no effective solution for the amplitude-frequency response introduced by in-band non-flatness, especially in spaceborne communication systems. Due to the generally customized design and significant differences in component characteristics of up-conversion links, the non-flatness of the channel response within the passband varies from product to product.

[0004] Therefore, there is an urgent need for a method that can efficiently and accurately compensate for the EVM degradation caused by in-band unevenness. Summary of the Invention

[0005] This invention aims to address the problem of EVM degradation caused by in-band amplitude-frequency response unevenness in upconversion channels. It provides a predistortion-based spaceborne transmitter front-end EVM compensation method, which effectively compensates for the EVM degradation caused by in-band unevenness at the transmitter front-end using predistortion, thereby reducing the EVM of the modulated signal.

[0006] This invention provides a predistortion-based spaceborne launch front-end EVM compensation method, comprising the following steps: S1. The frequency conversion gain of the spaceborne transmitter front-end at different temperatures Tn and different output frequencies is calibrated on the ground to obtain the amplitude-frequency response database, which is then stored in the channel data memory of the spaceborne transmitter front-end. The amplitude-frequency response data is G(Tn, F). out G(Tn, Fout) represents the temperature Tn and output frequency F of the spaceborne launch front-end. out Output RF signal gain; S2. When EVM deterioration occurs during on-orbit operation of the spaceborne launch front-end, the FPGA of the spaceborne launch front-end will adjust the current temperature and output signal frequency band accordingly. The amplitude-frequency response database is queried to obtain the frequency conversion gain dataset and form a gain-frequency curve. The gain-frequency curve is fitted to obtain the channel frequency amplitude gain coefficient of the signal in the current state during the spectrum shifting process. The channel frequency amplitude gain coefficient is used to enable the FPGA to perform pre-distortion processing on the input intermediate frequency signal to compensate for the EVM caused by in-band amplitude unevenness.

[0007] The predistortion-based spaceborne launch front-end EVM compensation method of the present invention, as a preferred embodiment, includes the following steps in step S2: S21. The FPGA reads the input intermediate frequency signal frequency and local oscillator signal frequency to obtain the theoretical radio frequency carrier range; S22. The FPGA reads the current temperature value of the onboard transmitter front end; S23. The FPGA reads the frequency-gain curve stored in the channel memory based on the current temperature value and the theoretical RF carrier range; S24. The FPGA obtains the channel amplitude gain coefficient by interpolating and fitting the frequency-gain curve. The FPGA performs inverse calculation based on the channel amplitude gain coefficient, the intermediate frequency signal modulation method and the theoretical RF carrier range, and performs pre-distortion processing on the intermediate frequency signal to compensate for the EVM caused by in-band amplitude unevenness.

[0008] In the preferred embodiment of the predistortion-based spaceborne transmitter front-end EVM compensation method described in this invention, in step S2, a frequency indexing mechanism is set between the channel data memory and the FPGA, so that the FPGA reads the data according to the frequency indexing mechanism when reading the amplitude-frequency response data in the channel data memory.

[0009] In the predistortion-based spaceborne transmitter front-end EVM compensation method described in this invention, as a preferred embodiment, the channel data memory and the FPGA use an SPI interface for data interaction. In the channel data memory, amplitude-frequency response data G(Tn, F) at different temperatures are stored. out (Using temperature Tn - frequency F) out The address index parameter is stored in the form of a two-dimensional array; After acquiring the temperature, operating frequency, and operating bandwidth of the onboard transmitter front end, the FPGA reads the amplitude-frequency response data through the SPI interface.

[0010] In the preferred embodiment of the predistortion-based spaceborne transmitter front-end EVM compensation method described in this invention, the channel data storage is a Flash memory, where the first address represents the temperature and the remaining addresses represent the frequency points.

[0011] The EVM compensation method for a predistortion-based spaceborne transmitter front-end described in this invention, as a preferred embodiment, includes the input intermediate frequency signal F in the amplitude-frequency response data table. in Frequency, input level, output RF signal F out Frequency, output level, and frequency conversion gain G; The output RF signal F in the amplitude-frequency response data table out Frequency setting: Step frequency.

[0012] The EVM compensation method for a predistortion-based spaceborne transmitter front-end described in this invention, as a preferred embodiment, uses a modulated signal from the spaceborne transmitter front-end, outputs a Ka-band radio frequency signal from the spaceborne transmitter front-end, and includes N amplitude-frequency response data tables in the amplitude-frequency response database, with N Tn values.

[0013] The EVM compensation method for a predistortion-based spaceborne transmitter front-end described in this invention, as a preferred method, uses QPSK modulation for the intermediate frequency signal, with Tn values ​​of -25℃, -15℃, 0℃, 15℃, 25℃, 45℃, and 60℃, and a total of 7 amplitude-frequency response data tables.

[0014] The present invention discloses a predistortion-based spaceborne transmitter front-end EVM compensation method. In a preferred embodiment, the spaceborne transmitter front-end includes a connected digital baseband module, an upconversion channel, and a power supply module connected to both the digital baseband module and the upconversion channel. The digital baseband module includes a connected FPGA and a DAC; The upconversion channel includes an intermediate frequency amplifier, an intermediate frequency filter, a mixer, an RF filter, an RF amplifier connected in sequence, a local oscillator connected to the mixer, and a channel data memory connected to the FPGA. The FPGA is connected to the channel data memory and the local oscillator, and the DAC outputs the intermediate frequency modulation signal to the input of the intermediate frequency amplifier.

[0015] The spaceborne transmitter front-end EVM compensation method based on predistortion described in this invention, as a preferred embodiment, includes an FPGA comprising a complex FIR filter, a coefficient storage ROM connected to the complex FIR filter, and a digital upconverter. The coefficient storage ROM stores the channel frequency amplitude gain coefficient calculated based on the read amplitude frequency response data. The complex FIR filter performs inverse operation based on the input I and Q signals and the channel frequency amplitude gain coefficient, and outputs predistorted I and Q output signals to the digital up-converter. The digital up-converter performs frequency conversion and outputs the signals to the DAC. The DAC performs digital-to-analog conversion and outputs the signals to the intermediate frequency amplifier. Complex FIR filters are predistortion filters.

[0016] EVM degradation caused by in-band amplitude unevenness exists in the linear region. Since the linear gain at the transmitting front end differs at different frequencies, the modulated signal is a spread spectrum signal in the frequency domain. Moreover, the higher the speed of the modulated signal, the larger the frequency domain span, and the more susceptible it is to the effects of in-band unevenness.

[0017] This invention proposes a predistortion-based EVM compensation spaceborne transmitter front-end for spaceborne applications. Based on measured channel response characteristics, it can achieve efficient and accurate compensation for EVM degradation caused by in-band unevenness, significantly improving signal quality.

[0018] This invention designs a hardware system for an EVM-compensated high-code-rate spaceborne Ka-band transmission front-end. The system includes: a digital baseband module, a Ka-band up-conversion channel, and a power supply module. The core of the digital baseband module is an FPGA and a DAC. The Ka-band up-conversion channel includes an up-conversion module (containing an IF amplifier, an IF filter, a mixer, an RF filter, and an RF amplifier), a local oscillator, and a channel data memory.

[0019] This invention takes a QPSK modulated signal with an intermediate frequency (IF) of 1.2 GHz and a transmission rate of 150 Mbps as an example. In the digital baseband module, an L-band carrier signal is output through an FPGA and a DAC, which manifests as a spread spectrum signal with a center frequency of 1.2 GHz and a bandwidth of 300 MHz in the frequency domain. This IF modulated signal is then up-converted to the Ka band (taking 26 GHz as an example). Since the up-conversion channel has amplitude-frequency characteristics that degrade the signal's EVM characteristics, pre-distortion is performed at the digital baseband end to ensure that the final Ka-band output modulated signal has good EVM characteristics.

[0020] This invention incorporates a channel data memory into the Ka upconversion module, which stores in-band amplitude-frequency response data at different temperatures that are fully compatible with the local unit, and interacts with the FPGA through a standardized interface.

[0021] FPGA pre-distortion processing of intermediate frequency (IF) signals needs to be based entirely on the actual amplitude-frequency characteristics of the up-conversion channel. This solution stores detailed measured amplitude-frequency response data of the up-conversion channel at different temperatures in the memory within the channel module. This memory can be considered an "amplitude-frequency response database" for that channel module. The digital baseband module reads relevant data from the "database" according to the actual operating frequency and mode. This data serves as the reference for the FPGA's pre-distortion processing of IF signals.

[0022] This invention designs a predistortion algorithm for FPGAs. By calculating the operating conditions and frequency under the current environment, reading the amplitude-frequency response data at the corresponding frequency, and then using interpolation to back-calculate the predistortion coefficients, the intermediate frequency signal is predistorted.

[0023] The core reference of the FPGA predistortion algorithm of this invention is the amplitude-frequency response of the channel module. After the FPGA reads the amplitude-frequency response data of the channel module, it can fit the amplitude-frequency response of the signal in the current state during the spectrum shift by interpolation. Then, the predistortion coefficients can be further calculated, and finally the predistortion intermediate frequency modulation signal adapted to the current hardware conditions can be obtained.

[0024] This invention establishes a frequency indexing mechanism between the channel memory and the FPGA, enabling the FPGA to read data from the memory according to this indexing mechanism when performing predistortion operations, thus creating conditions for the universal implementation of the predistortion algorithm.

[0025] Data exchange between the memory and the FPGA is achieved using a standard SPI interface. Within the memory, amplitude-frequency response data at different temperatures are indexed using a "temperature-frequency" model, with the gain of the frequency conversion channel at that temperature and operating frequency stored at the corresponding address, i.e., stored in the form of a two-dimensional array. At the FPGA level, only the local temperature, operating frequency, and bandwidth data need to be collected to read the amplitude-frequency response data via the SPI interface. Changes in the channel module's design, operating status, and operating mode will not affect the FPGA's reading method and interpolation calculation. This standardized data reading method also creates conditions for the universal implementation of predistortion algorithms.

[0026] The present invention has the following advantages: This invention addresses not only the nonlinear characteristics caused by traditional mixers and PAs, but also the EVM degradation problem caused by nonlinear characteristics in other stages of the frequency conversion channel. A predistortion algorithm is designed by reading measured parameters, which improves correction accuracy compared to traditional theoretical predistortion algorithms. A frequency indexing mechanism between the FPGA and memory is established, and a standardized data reading method is proposed, creating conditions for the universal implementation of the predistortion algorithm. Attached Figure Description

[0027] Figure 1 A flowchart of a predistortion-based spaceborne launch front-end EVM compensation method; Figure 2 This is a schematic diagram of the hardware system and frequency conversion for a predistortion-based spaceborne transmitter front-end EVM compensation method; Figure 3 This is a schematic diagram of data storage in the memory of a predistortion-based spaceborne launch front-end EVM compensation method; Figure 4 This is an example diagram of memory data indexing and data reading in a predistortion-based spaceborne launch front-end EVM compensation method; Figure 5This is a mid-frequency gain diagram of a predistortion-based spaceborne transmitter front-end EVM compensation method; Figure 6 This is a schematic diagram of mid-frequency gain fitting for a pre-distortion-based spaceborne launch front-end EVM compensation method; Figure 7 This is a schematic diagram of the internal structure of an FPGA for a pre-distortion-based spaceborne transmitter front-end EVM compensation method.

[0028] Figure label: 1. Digital baseband module; 11. FPGA; 111. Complex FIR filter; 112. Coefficient storage ROM; 113. Digital upconverter; 12. DAC; 2. Upconversion channel; 21. Intermediate frequency amplifier; 22. Intermediate frequency filter; 23. Mixer; 24. Radio frequency filter; 25. Radio frequency amplifier; 26. Local oscillator; 27. Channel data memory; 3. Power supply module. Detailed Implementation

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Example 1

[0030] like Figures 1-7 As shown, a predistortion-based spaceborne launch front-end EVM compensation method is as follows: Figure 2 The diagram shows the hardware system and frequency conversion of the present invention, representing its basic form. It includes a digital baseband module 1, a Ka upconversion channel 2, and a power supply module 3. The core of the digital baseband module 1 is an FPGA 11 and a DAC 12. The Ka upconversion channel 2 includes an upconversion module (containing an IF amplifier 21, an IF filter 22, a mixer 23, an RF filter 24, and an RF amplifier 25), a local oscillator 26, and a channel data memory 27. The power supply module 3 provides power to the other two modules. Taking a QPSK modulated signal as an example, in the digital baseband module 1, the FPGA 11 generates a QPSK signal, which is then modulated into an IF signal via the DAC 12. Simultaneously, the FPGA 11 communicates with the local oscillator 26 and the channel data memory 27 of the upconversion channel via an SPI interface. The upconversion channel upconverts the modulated IF signal to the RF band and ultimately outputs it to the phased array antenna. When the FPGA 11 performs predistortion processing, it can obtain relevant data from the channel data memory 27 via the SPI interface and use this data as a reference to perform predistortion calculations on the signal.

[0031] like Figure 1The diagram shows the predistortion algorithm block diagram of this invention. Taking a QPSK modulated signal with an intermediate frequency of 1.2 GHz and a transmission rate of 150 Mbps, and an FPGA11 controlling the local oscillator to output a 24.8 GHz signal as an example, in the digital baseband module 1, by reading the current intermediate frequency and local oscillator frequency, the theoretical output RF of the QPSK modulated signal with a carrier frequency of 26 GHz ± 150 MHz can be calculated. In the frequency domain, it is represented by a passband with a bandwidth of 300 MHz. The channel amplitude unevenness within the 300 MHz passband and the center frequency of 26 GHz will degrade the signal's EVM. At this time, the FPGA11 reads the gain of each frequency point in the corresponding frequency band at the corresponding temperature value from the channel data memory, fits the amplitude-frequency response of the current product in the form of interpolation, and then performs predistortion processing on the signal.

[0032] The frequency-gain discrete points are obtained by looking up the table as follows: Figure 5 As shown (discrete point response of amplitude and frequency, with frequency (GHz) on the horizontal axis and gain (dB) on the vertical axis), the fitted curve is obtained by fitting with a cubic polynomial, and the channel amplitude and frequency gain coefficient is obtained from the fitted curve (e.g., Figure 6 (as shown) Figure 6 The fitted curve is: R 2 The value is 0.9628, and the RMSE is 0.1514, where y is G(Tn, F out ), x is F out .

[0033] like Figure 7 As shown, the channel amplitude-frequency gain coefficient is stored in the coefficient storage ROM 112 of FPGA 11. The complex FIR filter 111 of FPGA 11 reads the channel amplitude-frequency gain coefficient through the AXI interface and performs digital predistortion. The inputs of the complex FIR filter 111 are I_in and Q_in and the read channel amplitude-frequency gain coefficient. The outputs I_out and Q_out are sent to the digital up-converter 113, and the digital up-converter 113 outputs a signal to DAC 12. The complex FIR filter 111 is a predistortion filter.

[0034] like Figure 3 The diagram shows a data storage diagram in the memory of the present invention. The memory stores amplitude-frequency response data in the form of a two-dimensional array. Each address represents the gain G(Tn, Fm) of the current frequency point Fm at the current temperature Tn.

[0035] like Figure 4The image shows an example of memory data indexing and data reading in this invention. Taking a 128Mbit Flash memory as an example, the address range of this memory includes 16,777,216 addresses from 0x000000 to 0xFFFFFF. The first address bit represents the temperature, and the last 5 address bits represent the frequency. This Flash memory can store 16 temperature points (assuming T1~T16 cover -55℃~85℃, with T10 being 25℃), and nearly 100,000 frequency response data points for each temperature point. When the starting frequency is 20GHz and the step frequency is 20kHz, the amplitude-frequency response of the channel module in the range of 20~40GHz can be stored at a certain temperature point. This data storage capacity can cover all types of Ka upconversion channels. Moreover, this data indexing method is only used for communication between the channel data memory and FPGA11. Changes in frequency, modulation method, and RF link of the frequency conversion channel will not affect the data reading between the two. That is, at the level of predistortion algorithm data reading, FPGA11 can form a standardized data reading language based on this data indexing method. When FPGA11 calculates that the data to be read in the current state—within a range of 25.85GHz to 26.15GHz, a center frequency of 26GHz, and a bandwidth of 300MHz—is required, the reading method is as follows: Figure 4 As shown, in this case, the address bits that need to be read are (0xA4AE14~0xA4EB85), where the first "A" represents the temperature point T10, i.e., 25℃; 0xA4CCCC is the address of the center frequency 26GHz.

[0036] The method for obtaining the amplitude and frequency gain of a spread spectrum signal is as follows: Figure 2 The Ka upconversion channel section in this invention upconverts the L-band signal to the Ka-band. Since this invention is aimed at the spaceborne transmission front-end, this method assumes that the center frequency of the L-band signal of each transmission front-end is known in actual applications. At this time, a database is established, and the frequency conversion gain of each frequency point from 20 to 40 GHz is calibrated by inputting the L-band signal. After the Ka-band upconversion channel was manufactured, it was calibrated on the ground using the following specific testing methods, and the results are shown in the table below:

[0037] These values ​​were measured at various typical temperature points, including -25℃, -15℃, 0℃, 15℃, 25℃, 45℃, and 60℃, forming 7 tables. Only the output frequency (Fout) and temperature are needed to obtain the frequency conversion gain value, which is then stored in the memory for table lookup, such as: G (25℃, 30GHz) = 33dB.

[0038] In practical applications, the usage process is as follows: Figure 1 As shown, based on the output intermediate frequency signal, modulation method, and local oscillator signal, the center frequency and spread spectrum range of the up-converted Ka signal can be calculated in advance. Knowing this information, and adding temperature data, the above table can be used for lookup. Assuming a project outputs a 1.2GHz QPSK modulated signal with a transmission rate of 150Mbps, and the transmitted SPI signal controls the local oscillator output at 24.8GHz, it can be calculated that the L-band signal needs to be up-converted to 26GHz, ultimately outputting a spread spectrum signal of 26GHz ± 150MHz. Knowing this information, by looking up the table, the amplitude response data can be found for every 20kHz frequency point between G(25℃, 25.85GHz) and G(25℃, 26.15GHz), thus forming a spread spectrum gain curve.

[0039] Because calibration is performed at each temperature point, Figure 3 In this context, Tn can be replaced with T7, where T1 to T7 represent seven temperature points. Figure 3 The data in the array is a 7*m two-dimensional array.

[0040] In practical applications, a thermistor is present in the unit, which indicates the current temperature of the unit. Based on the current temperature, an appropriate Tn is selected, and the data in the row containing Tn becomes the data reference for subsequent pre-distortion.

[0041] When the input signal is a 1.2 GHz QPSK modulated signal with a symbol rate of 150 Mbps, the EVM is caused by amplitude error and phase error, and the EVM of the input signal is 0.5%. After the QPSK modulated signal passes through up-conversion channel 2, gain non-flatness worsens the amplitude error, and group delay and local oscillator phase noise worsen the phase error, causing the EVM of the 26 GHz signal output from the up-conversion channel to deteriorate to about 5.6% (including the contributions of amplitude error and phase error).

[0042] After the pre-distortion processing in this embodiment, the EVM is reduced to 3.2%. That is, the pre-distortion method in this embodiment minimizes the contribution of amplitude error to EVM degradation, and most of the remaining EVM is caused by phase error.

[0043] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A pre-distortion-based spaceborne launch front-end EVM compensation method, characterized in that: Includes the following steps: S1. The amplitude-frequency response database is obtained by calibrating the frequency conversion gain of the satellite-borne transmitter front-end at different temperatures Tn and different output frequencies on the ground and storing it in the channel data storage (27) of the satellite-borne transmitter front-end. The amplitude-frequency response data is G(Tn, F out G(Tn, Fout) represents the temperature Tn and output frequency Fout of the spaceborne launch front-end. out Output RF signal gain; S2. When the EVM deteriorates during the on-orbit operation of the spaceborne launch front-end, the FPGA (11) of the spaceborne launch front-end will adjust the current temperature and output signal frequency band accordingly. The frequency conversion gain dataset is obtained by querying the amplitude-frequency response database and a gain-frequency curve is formed. The gain-frequency curve is fitted to obtain the channel frequency amplitude gain coefficient of the signal in the current state during the spectrum shifting process. The channel frequency amplitude gain coefficient is used to enable the FPGA (11) to perform pre-distortion processing on the input intermediate frequency signal to compensate for the EVM caused by the in-band amplitude non-flatness.

2. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 1, characterized in that: Step S2 includes the following steps: S21. The FPGA (11) reads the input intermediate frequency signal frequency and local oscillator signal frequency to obtain the theoretical radio frequency carrier range; S22, The FPGA (11) reads the current temperature value of the spaceborne launch front end; S23, the FPGA (11) reads the frequency-gain curve stored in the channel memory (27) according to the current temperature value and the theoretical radio frequency carrier range; S24. The FPGA (11) obtains the channel amplitude gain coefficient by interpolating and fitting the frequency-gain curve. The FPGA (11) performs inverse calculation based on the channel amplitude gain coefficient, the intermediate frequency signal modulation method and the theoretical radio frequency carrier range, and performs pre-distortion processing on the intermediate frequency signal to compensate for the EVM caused by the in-band amplitude non-flatness.

3. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 1, characterized in that: In step S2, a frequency indexing mechanism is set between the channel data memory (27) and the FPGA (11) so that the FPGA (11) reads the data according to the frequency indexing mechanism when reading the amplitude-frequency response data in the channel data memory (27).

4. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 3, characterized in that: The channel data memory (27) and the FPGA (11) use an SPI interface for data interaction; Within the channel data memory (27), amplitude-frequency response data G(Tn, F) at different temperatures are stored. out (Using temperature Tn - frequency F) out The address index parameter is stored in the form of a two-dimensional array; The FPGA (11) collects the temperature, operating frequency and operating bandwidth of the satellite-borne transmitter front end and then reads the amplitude-frequency response data through the SPI interface.

5. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 1, characterized in that: The channel data storage (27) is a Flash memory, where the first address represents the temperature and the remaining addresses represent the frequency.

6. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 1, characterized in that: The amplitude-frequency response data table includes the input intermediate frequency signal F. in Frequency, input level, output RF signal F out Frequency, output level, and frequency conversion gain G; The frequency response data table outputs the radio frequency signal F. out Frequency setting: Step frequency.

7. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 1, characterized in that: The signal from the spaceborne transmitting front end is a modulated signal, the radio frequency signal output by the spaceborne transmitting front end is a Ka band signal, and the amplitude-frequency response database includes N amplitude-frequency response data tables, with the number of Tn being N.

8. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 7, characterized in that: The modulation method of the intermediate frequency signal is QPSK, and the Tn values ​​are -25℃, -15℃, 0℃, 15℃, 25℃, 45℃, and 60℃. There are 7 amplitude-frequency response data tables.

9. A pre-distortion-based spaceborne launch front-end EVM compensation method according to any one of claims 1 to 8, characterized in that: The satellite-borne transmission front end includes a connected digital baseband module (1), an upconversion channel (2), and a power supply module (3) connected to both the digital baseband module (1) and the upconversion channel (2). The digital baseband module (1) includes a connected FPGA (11) and DAC (12). The upconversion channel (2) includes an intermediate frequency amplifier (21), an intermediate frequency filter (22), a mixer (23), a radio frequency filter (24), a radio frequency amplifier (25) connected in sequence, a local oscillator (26) connected to the mixer (23), and a channel data memory (27) connected to the FPGA (11). The FPGA (11) is connected to the channel data memory (27) and the local oscillator (26), and the DAC (12) outputs the intermediate frequency modulation signal to the input of the intermediate frequency amplifier (21).

10. The spaceborne launch front-end EVM compensation method based on predistortion according to claim 9, characterized in that: The FPGA (11) includes a complex FIR filter (111) and a coefficient storage ROM (112) and a digital up-converter (113) connected to the complex FIR filter (111). The coefficient storage ROM (112) stores the channel frequency amplitude gain coefficient calculated based on the read amplitude frequency response data. The complex FIR filter (13) performs inverse operation based on the input I and Q signals and the channel frequency amplitude gain coefficient to output pre-distorted I and Q output signals to the digital upconverter (113). The digital upconverter (113) performs frequency conversion and outputs the signal to the DAC (12). The DAC (12) performs digital-to-analog conversion and outputs the signal to the intermediate frequency amplifier (21). The complex FIR filter (13) is a predistortion filter.

Images