IQ imbalance calibration method, device, signal transmitter and zero intermediate frequency transceiver

Abstract

This application provides an IQ imbalance calibration method, apparatus, signal transmitter, and zero-IF transceiver. The method includes: acquiring a reference baseband signal and a radio frequency (RF) signal output by the IQ modulation channel based on the reference baseband signal, and converting the RF signal into a modulation baseband signal; reconstructing the modulation baseband signal to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel; obtaining pre-distortion parameters for compensating for IQ imbalance based on the reference baseband signal and the baseband feedback signal; performing pre-distortion processing on the reference baseband signal using the pre-distortion parameters to generate a baseband pre-distortion signal; and performing imbalance calibration on the IQ modulation channel based on the baseband pre-distortion signal. Because the IQ modulation channel is calibrated, the image rejection ratio and error vector amplitude performance of the output signal can be improved, thus increasing the calibration accuracy of the IQ modulation channel.

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, and in particular to a calibration method, apparatus, signal transmitter, zero-IF transceiver, and electronic device for IQ imbalance. Background Technology

[0002] Zero-IF architecture is widely used in wireless signal transmitters such as oscilloscopes, spectrum analyzers, radar transmitters, and base stations due to its unique advantages in size, power consumption, and integration.

[0003] In zero-IF architecture, the transmitter uses analog quadrature mixing technology to convert the radio frequency signal into baseband in-phase (I) and quadrature (Q) signals for transmission to the mixer. However, due to the non-ideal characteristics of analog devices, the IQ signals suffer from wideband amplitude and phase distortion and IQ imbalance. As a result, IQ modulation distortions such as image interference and local oscillator leakage spurious signals occur, leading to problems such as increased bit error rate and deterioration of error vector amplitude.

[0004] To overcome the IQ modulation imbalance problem, related techniques involve sending a known, preset baseband reference signal to the IQ modulation channel, then acquiring the modulated output RF signal and down-converting it back to baseband. By comparing the received baseband signal with the original reference signal, gain and phase imbalance parameters can be estimated, and pre-distortion compensation can be performed on the transmitted signal. However, frequency offset, symbol timing error, additional phase noise, and channel noise are introduced during signal transmission, which contaminate the acquired signal and cause deviations in the estimation of imbalance parameters, thereby reducing the accuracy of calibration. Summary of the Invention

[0005] This application discloses a calibration method, apparatus, signal transmitter, zero-IF transceiver, and electronic equipment for IQ imbalance, which is used to solve the problem of low calibration accuracy of IQ modulation channels.

[0006] In a first aspect, this application provides an IQ imbalance calibration method applied to an IQ modulation channel, comprising: acquiring a reference baseband signal and a modulation baseband signal output by the IQ modulation channel based on the reference baseband signal, and converting the radio frequency signal into a modulation baseband signal; reconstructing the modulation baseband signal to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel; obtaining predistortion parameters for compensating for IQ imbalance based on the reference baseband signal and the baseband feedback signal; performing predistortion processing on the reference baseband signal using the predistortion parameters to generate a baseband predistortion signal; and performing imbalance calibration on the IQ modulation channel based on the baseband predistortion signal.

[0007] The above method acquires a reference baseband signal and an RF signal output from the IQ modulation channel, converts the RF signal into a modulated baseband signal, and then performs signal recovery processing on the modulated baseband signal to obtain a baseband feedback signal. This effectively eliminates linear impairments such as frequency offset and phase offset introduced by the transmission channel, thereby ensuring that the baseband feedback signal can accurately reflect the gain and phase imbalance of the analog components (such as mixers, amplifiers, and local oscillators) of the IQ modulation channel. Based on this, predistortion parameters are calculated based on the reference baseband signal and the baseband feedback signal, which can comprehensively and accurately extract the gain and phase imbalance characteristics of the channel itself. Finally, the reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal, which is then used to calibrate the IQ modulation channel. This achieves deep compensation for IQ imbalance, significantly improves the image rejection ratio and error vector amplitude performance of the output signal, and thus improves calibration accuracy.

[0008] Furthermore, using a reference baseband signal can simultaneously compensate for distortions generated by signals of different frequencies in the transmission channel, which improves the efficiency of calibrating the IQ modulation channel compared to calibrating multiple frequencies separately.

[0009] In some embodiments, converting the radio frequency signal into a modulated baseband signal includes: downconverting the radio frequency signal and converting it from analog to digital to a modulated baseband signal.

[0010] By down-converting and digitizing, a digital baseband radio frequency signal is obtained. The analog radio frequency signal output from the IQ modulation channel, which carries channel distortion information, is down-converted to the baseband and then converted into a discrete reference baseband signal via analog-to-digital conversion. This process moves the signal to be analyzed from the continuously time-varying and noise-sensitive analog domain to the stable, accurate, and infinitely reproducible digital domain. This allows for more accurate feedback on the gain and phase imbalance of the reference baseband signal after passing through the IQ modulation channel, which is beneficial for improving the accuracy of calibration.

[0011] In some embodiments, reconstructing the radio frequency signal to generate a baseband feedback signal characterizing the IQ modulation channel distortion includes:

[0012] The modulated baseband signal is preprocessed to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel: wherein the preprocessing includes at least one of frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration.

[0013] Combining the solution provided in the first aspect, the aforementioned preprocessing can eliminate linear distortions such as frequency offset and phase shift introduced by the transmission channel. This preprocessing transforms the noisy, interference-laden RF signal into a high-fidelity, robust baseband feedback signal, thereby improving the accuracy, robustness, and refinement of the distortion characterization of the IQ modulation channel, providing data support for subsequent improvements to the communication system's performance.

[0014] In some embodiments, the preprocessing of the modulated baseband signal includes:

[0015] The modulation baseband signal is frequency offset compensated to obtain the first processed signal;

[0016] The first processed signal is symbol-synchronized and periodically truncated to obtain the second processed signal;

[0017] The second processed signal is subjected to joint phase error compensation for time delay, frequency offset, and sampling rate error to obtain the third processed signal;

[0018] The signal-to-noise ratio is optimized and the complex gain is calibrated on the third processed signal to generate the baseband feedback signal.

[0019] By combining the above schemes, and through sequentially performing frequency offset compensation, symbol synchronization and period truncation, joint time-frequency domain phase error compensation, signal-to-noise ratio optimization, and complex gain calibration on the modulated baseband signal, various complex impairments introduced during signal transmission and acquisition are systematically eliminated. This allows for the gradual recovery of the original RF signal, which contains carrier frequency offset, timing deviation, sampling rate error, and noise interference, into a high-quality baseband feedback signal that is strictly aligned with the reference baseband signal in time, frequency, amplitude, and phase. This overcomes the shortcomings of traditional calibration methods that suffer from performance degradation due to the coupling of various impairments in broadband systems. It lays a precise and reliable signal foundation for the subsequent high-precision and high-efficiency extraction of intrinsic distortion parameters of the IQ modulation channel, thereby ensuring the final performance and robustness of the entire unbalanced calibration system.

[0020] In some embodiments, the step of performing frequency offset compensation on the modulated baseband signal to obtain a first processed signal includes:

[0021] The modulated baseband signal is acquired in multiple cycles, and the phase difference of the same index symbol in different cycles is calculated.

[0022] The signal frequency offset of the modulated baseband signal is corrected based on the phase difference to obtain the first processed signal.

[0023] By comprehensively determining the frequency offset calibration parameters based on the phase difference of multi-cycle reference baseband signals, the error present in a single-cycle signal can be eliminated, thus improving the accuracy of frequency offset calibration.

[0024] In some embodiments, the symbol synchronization and periodic truncation of the first processed signal includes:

[0025] By utilizing the correlation between the reference baseband signal and the first processed signal, the starting symbol of each integer cycle signal in the first processed signal is identified;

[0026] Based on the starting symbol, the first processed signal is periodically truncated.

[0027] By utilizing the correlation between the reference baseband signal and the first processed signal for symbol synchronization, the starting point of each period of a periodic broadband signal can be accurately identified, and accurate period truncation can be performed accordingly. This effectively overcomes the performance degradation of traditional synchronization methods under low signal-to-noise ratio or complex channel impairments, significantly improving the accuracy and robustness of period positioning. It provides a well-structured and clearly defined data foundation for subsequent time-frequency alignment, multi-period averaging, and high-precision parameter estimation, thereby ensuring the reliability of the entire calibration process and the final compensation accuracy.

[0028] In some embodiments, the joint phase error compensation of the second processed signal for time delay, frequency offset, and sampling rate error includes:

[0029] Based on the reference baseband signal and the second processed signal, a time-frequency grid equation is constructed to perform the joint phase error compensation.

[0030] By employing time-frequency grid equations to jointly compensate for residual carrier frequency offset, sampling rate clock frequency deviation, and timing deviation, predistortion parameters and compensations for different frequencies can be estimated simultaneously (jointly) through a single fitting process, shortening calibration time and improving calibration efficiency. Furthermore, by utilizing a two-dimensional time-frequency signal structure, the coupling effects of various distortions are mathematically decoupled, preventing error propagation and resulting in more accurate parameter estimation. This lays a precise foundation for subsequent IQ modulation channel compensation.

[0031] In some embodiments, the signal-to-noise ratio optimization and complex gain calibration of the third processed signal include:

[0032] The third processed signal is averaged and denoised in units of signal period.

[0033] Complex gain alignment is performed based on the denoised third processed signal and the reference baseband signal.

[0034] By averaging the aligned signal period and combining it with the reference baseband signal for complex gain calibration, on the one hand, period averaging effectively suppresses random noise, significantly improves the signal-to-noise ratio, and enhances the stability and anti-interference capability of subsequent parameter estimation; on the other hand, complex gain alignment eliminates the overall amplitude scaling and constant phase shift introduced during signal transmission, ensuring that the feedback signal and the reference signal are accurately comparable in the complex domain.

[0035] In some embodiments, obtaining the predistortion parameters for compensating IQ imbalance based on the reference baseband signal and the baseband feedback signal includes:

[0036] Based on the reference baseband signal and the baseband feedback signal, calculate the predistortion parameters;

[0037] Using the predistortion parameters, the reference baseband signal and its conjugate signal are filtered to generate a baseband predistortion signal.

[0038] By using a reference baseband signal and a baseband feedback signal, predistortion parameters are calculated. These predistortion parameters may include the parameters of a first complex filter corresponding to the reference baseband signal and the parameters of a second complex filter corresponding to the conjugate signal of the reference baseband signal. This can provide high-precision and high-efficiency compensation for frequency-selective IQ imbalance in broadband systems, improve the image rejection ratio, and enhance signal quality.

[0039] In some embodiments, the step of filtering the reference baseband signal and its conjugate signal using the predistortion parameters to generate a baseband predistortion signal includes:

[0040] Using the predistortion parameters, the reference baseband signal and its conjugate signal are subjected to parallel filtering, and the superimposed signal of the two filtering results is combined with the local oscillator leakage suppression coefficient to generate the baseband predistortion signal.

[0041] Combining the solution provided in the first aspect, a baseband predistortion signal is generated by parallel filtering of the original reference baseband signal and superimposing it with the DC offset, thereby achieving joint calibration of IQ imbalance and local oscillator leakage.

[0042] In some embodiments, calculating the predistortion parameters based on the reference baseband signal and the baseband feedback signal includes:

[0043] Based on the reference baseband signal and the baseband feedback signal, the predistortion parameters are calculated using a nonlinear optimization algorithm.

[0044] By introducing a nonlinear optimization algorithm, predistortion parameters are calculated based on the reference baseband signal and the precisely preprocessed baseband feedback signal. This enables adaptive and accurate fitting of the composite model of frequency-selective distortion and nonlinear distortion in a broadband IQ modulation channel, overcoming the shortcomings of traditional linear or piecewise calibration methods in terms of accuracy under broadband conditions.

[0045] In some embodiments, it also includes:

[0046] The baseband predistortion signal is input to the IQ modulation channel, and after unbalance calibration of the IQ modulation channel, the following iterative steps are included:

[0047] The newly modulated baseband signal is reconstructed to obtain the updated baseband feedback signal;

[0048] Based on the reference baseband signal and the updated baseband feedback signal, the predistortion parameters are recalculated, and a new round of baseband predistortion signal is generated using the recalculated predistortion parameters. The iterative optimization steps are repeated until a preset iteration termination condition is met. The iteration termination condition includes at least one of the following: the output signal quality of the IQ modulation channel meets a preset threshold, the number of iterations reaches a preset maximum number, and the change in the predistortion parameters calculated in two adjacent iterations is less than a preset tolerance.

[0049] Through iterative optimization, multiple parameters used to compensate for IQ imbalance and local oscillator leakage are solved simultaneously, significantly improving the convergence speed and accuracy of parameter estimation.

[0050] Secondly, this application provides an unbalanced calibration device for an IQ modulation channel, comprising:

[0051] The signal acquisition module is used to acquire a reference baseband signal and a radio frequency signal output by the IQ modulation channel based on the reference baseband signal, and convert the radio frequency signal into a modulation baseband signal;

[0052] The signal recovery module is used to reconstruct the modulated baseband signal and generate a baseband feedback signal that characterizes the distortion of the IQ modulation channel;

[0053] The parameter estimation module is used to calculate the predistortion parameters for compensating for IQ imbalance based on the reference baseband signal and the baseband feedback signal.

[0054] The predistortion processing module is used to perform predistortion processing on the reference baseband signal using the predistortion parameters to generate a baseband predistortion signal;

[0055] The calibration execution module is used to perform unbalanced calibration on the IQ modulation channel based on the baseband predistortion signal.

[0056] The aforementioned system acquires a reference baseband signal and an RF signal output from the IQ modulation channel, converts the RF signal into a modulated baseband signal, and then performs signal recovery processing on the RF signal to obtain a baseband feedback signal. This effectively eliminates linear impairments such as frequency offset and phase offset introduced by the transmission channel, ensuring that the baseband feedback signal accurately reflects the gain and phase imbalance of the analog components (such as mixers, amplifiers, and local oscillators) in the IQ modulation channel. Based on this, predistortion parameters are calculated using the reference baseband signal and the baseband feedback signal, enabling comprehensive and accurate extraction of the channel's own gain and phase imbalance characteristics. Finally, the reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal, which is then used to calibrate the IQ modulation channel. This achieves deep compensation for IQ imbalance, significantly improving the image rejection ratio and error vector amplitude performance of the output signal, thereby enhancing calibration accuracy.

[0057] Thirdly, this application provides a signal transmitter, including: a signal generator, a signal acquisition unit, and a processor;

[0058] The signal generator, connected to the processor, is used to generate and send a reference baseband signal to the IQ modulation channel;

[0059] The signal acquisition unit is connected to the processor and is used to acquire the radio frequency signal output by the IQ modulation channel based on the reference baseband signal, and convert the radio frequency signal into a modulation baseband signal and send it to the processor.

[0060] The processor is used for:

[0061] Receive the modulated baseband signal from the signal acquisition unit;

[0062] The modulated baseband signal is reconstructed to generate a baseband feedback signal that characterizes the distortion of the IQ modulation channel;

[0063] Based on the reference baseband signal and the baseband feedback signal, calculate the predistortion parameters used to compensate for IQ imbalance;

[0064] The reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal;

[0065] The signal generator is controlled to input the baseband predistortion signal to the IQ modulation channel to perform unbalance calibration on the IQ modulation channel.

[0066] The processor in the above system acquires a reference baseband signal and an RF signal output from the IQ modulation channel, converts the RF signal into a modulated baseband signal, and then performs signal recovery processing on the modulated baseband signal to obtain a baseband feedback signal. This effectively eliminates linear impairments such as frequency offset and phase offset introduced by the transmission channel, thereby ensuring that the baseband feedback signal can accurately reflect the gain and phase imbalance of the analog components (such as mixers, amplifiers, and local oscillators) of the IQ modulation channel. Based on this, predistortion parameters are calculated based on the reference baseband signal and the baseband feedback signal, which can comprehensively and accurately extract the gain and phase imbalance characteristics of the channel itself. Finally, the reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal, which is then used to calibrate the IQ modulation channel. This achieves deep compensation for IQ imbalance, significantly improves the image rejection ratio and error vector amplitude performance of the output signal, and thus improves calibration accuracy.

[0067] Fourthly, this application provides a zero-IF transceiver, comprising: a signal transmitter as described in the third aspect.

[0068] Since the zero-IF transceiver includes a signal transmitter, the unbalanced calibration of the IQ modulation channel in the signal transmitter effectively compensates for gain and phase mismatch, reduces signal image interference and modulation error, and improves the quality of the RF signal output by the signal transmitter.

[0069] Fifthly, this application provides an electronic device, including: a zero-IF transceiver as described in the fourth aspect.

[0070] Since the electronic equipment includes a zero-IF transceiver, which includes a signal transmitter, the unbalanced calibration of the IQ modulation channel in the signal transmitter effectively compensates for gain and phase mismatch, reduces signal image interference and modulation error, improves signal purity and demodulation accuracy, enables the electronic equipment to maintain stable performance under complex working conditions, enhances adaptability to hardware differences and environmental changes, and improves communication quality. Attached Figure Description

[0071] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0072] Figure 1 This is an architectural diagram of an electronic device in the related technology;

[0073] Figure 2 This is a diagram of the architecture of the IQ modulation channel;

[0074] Figure 3 An architectural diagram of an electronic device provided in an embodiment of this application;

[0075] Figure 4 This is an architectural diagram of the signal reconstruction unit provided in an embodiment of this application;

[0076] Figure 5 This is an architectural diagram of the predistortion processing unit provided in the embodiments of this application;

[0077] Figure 6 An architectural diagram of another electronic device provided in an embodiment of this application;

[0078] Figure 7 A schematic flowchart of the unbalanced calibration method for the IQ modulation channel provided in the embodiments of this application;

[0079] Figure 8 A schematic diagram of the multi-cycle reference baseband signal and baseband feedback signal provided in the embodiments of this application;

[0080] Figure 9 The radio frequency signal spectrum before calibration provided in the embodiments of this application;

[0081] Figure 10 The calibrated radio frequency signal spectrum provided for the embodiments of this application.

[0082] Explanation of reference numerals in the attached figures:

[0083] 100 - Electronic devices;

[0084] 1-Zero IF transceiver;

[0085] 11-Signal transmitter; 12-Signal receiver; 13-Computing device;

[0086] 111-Baseband signal generator; 112-IQ modulation channel; 113-Signal acquisition unit; 114-Processor;

[0087] 1121 - First mixer; 1122 - Second mixer; 1123 - First adder;

[0088] 1141 - Signal acquisition unit; 1142 - Judgment unit; 1143 - Signal reconstruction unit; 1144 - Correction coefficient estimation unit; 1145 - Predistortion processing unit;

[0089] 11431 - Frequency offset compensation subunit; 11432 - Symbol synchronization subunit; 11433 - Joint phase error compensation subunit; 11434 - Signal-to-noise ratio optimization subunit; 11435 - Gain calibration subunit;

[0090] 11451 - First complex filter; 11452 - Second complex filter; 11453 - Local oscillator leakage suppressor; 11454 - Second adder. Detailed Implementation

[0091] The IQ imbalance calibration method provided in this application can be applied to electronic devices such as signal generators, analyzers, spectrum analyzers, oscilloscopes, and radio frequency signal generation and reception devices.

[0092] like Figure 1 This is an architectural diagram of an electronic device in the related technology, such as... Figure 1 As shown, the electronic device 100 includes a zero-IF transceiver 1, which can directly modulate a reference baseband signal into a radio frequency (RF) signal and directly down-convert the RF signal back to the reference baseband signal without any intermediate frequency (IF) processing. The reference baseband signal is the original signal generated in the digital domain for modulating the carrier wave. It consists of two orthogonal components: an in-phase component (I) and a quadrature component (Q). The I and Q signals contain all the information to be transmitted, and their frequencies are concentrated near the zero frequency (baseband), without frequency shifting. The RF signal is the signal after IQ modulation, up-conversion, and power amplifier amplification; its frequency has been shifted to the specified frequency band.

[0093] The zero intermediate frequency transceiver 1 includes a signal transmitter 11 and a signal receiver 12. The output terminal of the signal transmitter 11 is connected to the input terminal of the signal receiver 12. The signal transmitter 11 is used to modulate the reference baseband signal into a radio frequency signal and input the radio frequency signal to the signal receiver 12. The signal receiver 12 is used to restore the radio frequency signal to the original reference baseband signal.

[0094] Specifically, the signal transmitter 11 includes a baseband signal generator 111 and an IQ modulation channel 112. The output of the baseband signal generator 111 is connected to the input of the IQ modulation channel 112. The baseband signal generator 111 generates a reference baseband signal and inputs it to the IQ modulation channel 112. The IQ modulation channel 112 modulates the reference baseband signal to generate a radio frequency (RF) signal and outputs the RF signal to the signal receiver 12. The baseband signal generator 111 has been described in detail in related technologies and will not be repeated here.

[0095] Figure 2 The architecture diagram of the IQ modulation channel is as follows: Figure 2As shown, the IQ modulation channel 112 includes a first mixer 1121, a second mixer 1122, and a first adder 1123. The input terminal of the first mixer 1121 is connected to the first output terminal of the baseband signal generator 111, the input terminal of the second mixer 1122 is connected to the second output terminal of the baseband signal generator 111, the output terminal of the first mixer 1121 is connected to the first input terminal of the first adder 1123, and the output terminal of the second mixer 1122 is connected to the second input terminal of the first adder 1123. The first mixer 1121 is located in the I-channel, and the second mixer 1122 is located in the Q-channel. A common local oscillator signal provides a carrier for both the I-channel and Q-channel signals. The local oscillator signal is 90 degrees out of phase with the second mixer 1122. The working principle of the IQ modulation channel 112 is to multiply and mix two orthogonal carriers (sine and cosine with a 90-degree phase difference) with the in-phase component (I-channel) and quadrature component (Q-channel) of the reference baseband signal, respectively. Then, the product results of the two channels are added together using the first adder 1123, thereby modulating the spectrum of the reference baseband signal into a radio frequency signal.

[0096] However, in the existing technology, due to various factors affecting the amplitude and phase of the I and Q paths, the amplitude and phase of the I and Q paths are mismatched, which cannot accurately cancel the image signals on both sides of the local oscillator. Unwanted images will fall directly into the frequency spectrum of the radio frequency signal, resulting in an increase in the bit error rate and a deterioration of the error vector amplitude.

[0097] To address the problems of increased bit error rate and deteriorated error vector amplitude caused by image interference due to amplitude and phase mismatch between the I and Q paths, this application provides a signal transmitter. The following detailed description is provided with reference to specific embodiments.

[0098] Figure 3 An architecture diagram of an electronic device provided in this application embodiment, such as... Figure 3 As shown, in some embodiments, the electronic device 100 includes a zero-IF transceiver 1, which includes a signal transmitter 11 and a signal receiver 12. The signal transmitter 11 includes a baseband signal generator 111, an IQ modulation channel 112, a signal acquisition unit 113, and a processor 114. The first output terminal of the baseband signal generator 111 is connected to the first input terminal of the IQ modulation channel 112 and the first input terminal of the processor 114. The input terminal of the signal acquisition unit 113 is connected to the output terminal of the IQ modulation channel 112, and the output terminal of the signal acquisition unit 113 is connected to the second input terminal of the processor 114. The output terminal of the processor 114 is connected to the input terminal of the IQ modulation channel 112.

[0099] In some embodiments, the signal acquisition unit 113 is used to acquire the radio frequency signal output from the IQ modulation channel 112, down-convert the radio frequency signal and convert it from analog to digital to a modulated baseband signal, and send the modulated baseband signal to the processor 114. The signal acquisition unit 113 can be one of a vector signal analyzer, an envelope detector, or a down-conversion acquisition board. If the signal acquisition unit 113 is a vector signal analyzer, the vector signal analyzer is located outside the signal transmitter 11. The vector signal analyzer is used to process the input radio frequency IQ modulation signal (such as attenuation / amplification), and then convert it into an analog baseband I / Q signal through quadrature down-conversion (or superheterodyne down-conversion plus quadrature demodulation). After analog-to-digital conversion and subsequent digital signal processing (such as filtering and correction), the modulated baseband signal (I+jQ data stream) is finally obtained. If the signal acquisition unit 113 is an envelope detector, located inside the signal transmitter 11, the envelope detector is used to rectify and low-pass filter the RF output signal at the transmitter end, extract the envelope of its signal amplitude, and output a low-frequency analog voltage signal proportional to the instantaneous RF power. This signal only contains amplitude information and completely loses the phase and quadrature modulation information of the signal. Therefore, it is usually only used for transmitter output power monitoring, overload protection, or simple presence detection, and cannot be used for modulation quality analysis or digital demodulation. Therefore, a narrowband analog-to-digital converter needs to be set after the envelope detector to convert the analog signal into a modulated baseband signal. If the signal acquisition unit 113 is a down-conversion acquisition board, located inside the signal transmitter 11, the down-conversion acquisition board is used to couple out the RF output signal, down-convert it through the internally integrated local oscillator and mixer to convert it into a baseband signal, and then sample it through the onboard wideband analog-to-digital converter to finally output the modulated baseband signal.

[0100] Specifically, the processor 114 includes a signal acquisition unit 1141, a judgment unit 1142, a signal reconstruction unit 1143, a correction coefficient estimation unit 1144, and a predistortion processing unit 1145. The input terminal of the signal acquisition unit 1141 is connected to the output terminal of the signal acquisition unit 113; the input terminal of the judgment unit 1142 is connected to the output terminal of the signal acquisition unit 1141; the input terminal of the signal reconstruction unit 1143 is connected to the output terminal of the judgment unit 1142; the first input terminal of the correction coefficient estimation unit 1144 is connected to the output terminal of the signal reconstruction unit 1143; the second input terminal of the correction coefficient estimation unit 1144 is connected to the second output terminal of the baseband signal generator 111; the first input terminal of the predistortion processing unit 1145 is connected to the output terminal of the correction coefficient estimation unit 1144; the second input terminal of the predistortion processing unit 1145 is connected to the third output terminal of the baseband signal generator 111; and the output terminal of the predistortion processing unit 1145 is connected to the second input terminal of the IQ modulation channel 112. The signal acquisition unit 1141 is used to acquire the modulated baseband signal from the signal acquisition unit 113. The judgment unit 1142 is used to determine whether the correction coefficient of the predistortion processing unit 1145 needs to be updated based on the modulated baseband signal. If the judgment unit 1142 determines that the correction coefficient needs to be updated, the current modulated baseband signal is input to the signal reconstruction unit 1143. The signal reconstruction unit 1143 is used to reconstruct the radio frequency signal and generate a baseband feedback signal to characterize the distortion of the IQ modulation channel 112. The correction coefficient estimation unit 1144 is used to calculate the predistortion parameter to compensate for the IQ imbalance based on the difference between the reference baseband signal and the baseband feedback signal. The predistortion processing unit 1145 is used to perform predistortion processing on the reference baseband signal using the predistortion parameter to generate a baseband predistortion signal and input the baseband predistortion signal to the IQ modulation channel 112 to perform imbalance calibration on the IQ modulation channel 112.

[0101] In this embodiment, by setting a signal acquisition unit 113 and a processor 114 in the signal transmitter 11, the signal acquisition unit 113 acquires the radio frequency signal output from the IQ modulation channel 112 and down-converts the radio frequency signal and converts it from analog to digital to a modulation baseband signal. The processor 114 acquires the reference baseband signal and the modulation baseband signal, and reconstructs (restores) the modulation baseband signal to obtain the baseband feedback signal. This effectively eliminates the linearity impairments such as frequency offset and phase offset introduced by the IQ modulation channel 112, thereby ensuring that the baseband feedback signal can accurately feed back the IQ signal. The amplitude and phase imbalance of the analog components (such as mixers) in the modulation channel 112 are determined. Based on this, predistortion parameters are calculated using the reference baseband signal and the baseband feedback signal, which can comprehensively and accurately extract the gain and phase imbalance characteristics of the IQ modulation channel 112 itself. Finally, the reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal and calibrate the IQ modulation channel 112, thereby compensating for the imbalance of the IQ modulation channel 112 and improving the image rejection ratio and error vector amplitude performance of the output signal.

[0102] In some embodiments, the signal reconstruction unit 1143 preprocesses the modulated baseband signal to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel 112. The preprocessing includes at least one of frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio (SNR) optimization, and complex gain calibration. Frequency offset compensation refers to compensating for the frequency offset generated during signal transmission in the channel. Signal synchronization involves identifying the starting position of each complete cycle signal in the signal to perform period truncation and subsequent averaging and noise reduction. Joint phase error compensation involves jointly compensating for residual carrier frequency offset, sampling rate clock frequency deviation, and timing deviation in the signal. SNR optimization utilizes the periodic structure of the signal to perform time-domain averaging of multiple complete cycles after signal synchronization to suppress random noise and improve the SNR, thereby providing a high-quality signal for accurate estimation of subsequent IQ imbalance parameters. Complex gain calibration aligns the complex gain of the SNR-optimized signal to ensure its amplitude and phase are consistent with the signal on a global scale, thereby eliminating the common gain offset introduced by the loopback channel and ensuring the accuracy of subsequent IQ imbalance parameter estimation. In this embodiment, since the signal reconstruction unit 1143 performs frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration on the modulated baseband signal, it can eliminate linear distortions such as frequency offset and phase offset introduced by the IQ modulation channel 112, and convert the noisy and interference-laden modulated baseband signal into a fidelity baseband feedback signal. This improves the accuracy, robustness, and refinement of the distortion characterization of the IQ modulation channel 112, providing data support for subsequent improvements to the performance of the communication system.

[0103] Specifically, Figure 4The architecture diagram of the signal reconstruction unit is as follows: Figure 4 As shown, the signal reconstruction unit 1143 includes a frequency offset compensation subunit 11431, a symbol synchronization subunit 11432, a joint phase error compensation subunit 11433, a signal-to-noise ratio optimization subunit 11434, and a gain calibration subunit 11435. The input terminal of the frequency offset compensation subunit 11431 is connected to the output terminal of the judgment unit 1142. The input terminal of the symbol synchronization subunit 11432 is connected to the output terminal of the frequency offset compensation subunit 11431. The input terminal of the joint phase error compensation subunit 11433 is connected to the output terminal of the symbol synchronization subunit 11432. The input terminal of the signal-to-noise ratio optimization subunit 11434 is connected to the output terminal of the joint phase error compensation subunit 11433. The input terminal of the gain calibration subunit 11435 is connected to the output terminal of the signal-to-noise ratio optimization subunit 11434. The output terminal of the gain calibration subunit 11435 is connected to the first input terminal of the correction coefficient estimation unit 1144. Frequency offset compensation subunit 11431 is used to compensate for the frequency offset of the modulated baseband signal to obtain a first processed signal; symbol synchronization subunit 11432 is used to perform symbol synchronization and period truncation on the first processed signal to obtain a second processed signal; joint phase error compensation subunit 11433 is used to perform joint phase error compensation for time delay, frequency offset, and sampling rate error on the second processed signal to obtain a third processed signal; signal-to-noise ratio optimization subunit 11434 is used to optimize the signal-to-noise ratio of the third processed signal to obtain a denoised third processed signal; gain calibration subunit 11435 is used to perform complex gain calibration on the denoised third processed signal to generate a baseband feedback signal. In the above preprocessing process, since the linear differences such as frequency offset and phase offset caused by the separate transmission of in-phase and quadrature components are eliminated, the difference between the baseband feedback signal and the reference baseband signal can be used subsequently to determine the gain and phase imbalance of the in-phase and quadrature components caused by the analog components (such as mixers, amplifiers, and local oscillators) of the IQ modulation channel 112.

[0104] Specifically, Figure 5 This is an architecture diagram of the predistortion processing unit. (See diagram below.) Figure 4As shown, the predistortion processing unit 1145 may include a first complex filter 11451, a second complex filter 11452, a local oscillator leakage suppressor 11453, and a second adder 11454. The input terminals of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453 are connected to the output terminal of the baseband signal generator 111. The output terminals of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453 are all connected to the input terminal of the second adder 11454. The output terminal of the second adder 11454 is connected to the second input terminal of the IQ modulation channel 112. The first complex filter 11451 is used to compensate for the linear distortion (including amplitude non-flatness and phase non-linearity) generated when the reference baseband signal is transmitted on the main channel. Specifically, the first complex filter 11451 is used to adjust the in-phase and quadrature components of the reference baseband signal, thereby performing pre-equalization within the main signal frequency band to ensure that the reference baseband signal maintains ideal characteristics after passing through the IQ modulation channel 112. Simultaneously, due to the imperfections of the IQ modulation channel 112 (such as amplitude imbalance between the I / Q channels and non-quadrature phase), a carrier-centered image frequency band signal will be generated. The second complex filter is used to predict the image interference signal based on the reference baseband signal and generate its inverse signal. When this inverse image signal is superimposed on the actual image interference generated by the subsequent IQ modulation channel 112, the two can cancel each other out. Furthermore, due to the DC bias mismatch of the IQ modulator, the local oscillator signal leaks directly to the RF output, manifesting as an undesirable spike at the carrier frequency in the spectrum. Therefore, in the predistortion processing unit 1145, the local oscillator leakage suppressor 11453 is used to suppress local oscillator leakage. It mainly generates a DC correction by multiplying its internal suppression coefficient by the absolute mean of the input signal. After the DC correction is applied to the reference baseband signal, it can first cancel the DC bias of the two paths in the IQ modulation channel 112, thereby suppressing the generation of local oscillator leakage. The second adder 11454 is used to vector-add the output signals of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453 to synthesize the baseband predistortion signal. Thus, the baseband signal generator 111 can input the reference baseband signal to the first complex filter 11451, input the conjugate signal of the reference baseband signal to the second complex filter 11452, and input the mean of the absolute values ​​of the reference baseband signal over the entire cycle to the local oscillator leakage suppressor 11453. The local oscillator leakage suppressor 11453 stores the local oscillator leakage suppression coefficient, which is used to multiply it by the mean of the absolute values ​​of the reference baseband signal over the entire cycle to suppress local oscillator leakage. The three input terminals of the second adder 11454 are respectively connected to the outputs of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453.The second adder 11454 vector-adds the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453.

[0105] Figure 6 For the architecture diagram of another electronic device, such as Figure 6 As shown, in some other embodiments, the processor 114 in the signal transmitter 11 includes a predistortion processing unit 1145, which is used to predistort the reference baseband signal using predistortion parameters to generate a baseband predistortion signal. The external computing device 13 (such as a computer) also includes a signal acquisition unit 1141, a judgment unit 1142, a signal reconstruction unit 1143, and a correction coefficient estimation unit 1144. The input terminal of the signal acquisition unit 1141 is connected to the output terminal of the signal acquisition unit 113, the input terminal of the judgment unit 1142 is connected to the output terminal of the signal acquisition unit 1141, the input terminal of the signal reconstruction unit 1143 is connected to the output terminal of the judgment unit 1142, the first input terminal of the correction coefficient estimation unit 1144 is connected to the output terminal of the signal reconstruction unit 1143, and the second input terminal of the correction coefficient estimation unit 1144 is connected to the second output terminal of the baseband signal generator 111. The signal acquisition unit 1141 is used to acquire the modulated baseband signal from the signal acquisition unit 113. The judgment unit 1142 is used to determine whether the correction coefficient of the predistortion processing unit 1145 needs to be updated based on the modulated baseband signal. If the judgment unit 1142 determines that the correction coefficient needs to be updated, it inputs the current modulated baseband signal to the signal reconstruction unit 1143. The signal reconstruction unit 1143 is used to reconstruct the radio frequency signal and generate a baseband feedback signal to characterize the distortion of the IQ modulation channel 112. The correction coefficient estimation unit 1144 is used to calculate the predistortion parameter to compensate for the IQ imbalance based on the difference between the reference baseband signal and the baseband feedback signal. The predistortion processing unit 1145 is used to perform predistortion processing on the reference baseband signal using the predistortion parameter to generate a baseband predistortion signal and input the baseband predistortion signal to the IQ modulation channel 112 to perform imbalance calibration on the IQ modulation channel 112.

[0106] Please refer to Figure 7 , Figure 7 for Figure 3 or Figure 6 The diagram shows a flowchart of the unbalanced calibration method for the IQ modulation channel of the electronic device. Figure 7 As shown, the method includes the following steps:

[0107] S71: Acquires the reference baseband signal and the RF signal output by the IQ modulation channel based on the reference baseband signal, and converts the RF signal into a modulated baseband signal.

[0108] The reference baseband signal is the signal output by the baseband signal generator 111. The reference baseband signal can be a wideband signal, such as an OFDM (Orthogonal Frequency Division Multiplexing) signal with a sampling rate of 625MHz and a bandwidth of 500MHz.

[0109] The radio frequency signal is the signal generated after the reference baseband signal is input to the IQ modulation channel 112, specifically, as follows: Figure 2 As shown, the I-channel signal of the reference baseband signal is input to the first mixer 1121 and the Q-channel signal is input to the second mixer 1122. A common local oscillator signal provides a carrier for the I-channel and Q-channel signals. The local oscillator and the second mixer 1122 are 90 degrees out of phase. The two orthogonal carriers (sine and cosine with a 90-degree phase difference) are multiplied and mixed with the in-phase component (I-channel) and the quadrature component (Q-channel) of the reference baseband signal, respectively. Then, the two product results are added using the first adder 1123, thereby modulating the spectrum of the reference baseband signal into a radio frequency signal.

[0110] The modulated baseband signal is obtained by down-converting the radio frequency signal and performing analog-to-digital conversion using the signal acquisition unit 113. Its signal characteristics are similar to those of the reference baseband signal, and will not be described in detail here.

[0111] Specifically, since the input terminal of the signal acquisition unit 113 is connected to the output terminal of the IQ modulation channel 112, and the input terminal of the processor 114 is connected to the output terminal of the baseband signal generator 111, the signal acquisition unit 113 can be used to acquire the radio frequency signal output by the IQ modulation channel 112, and the processor 114 can be used to obtain the reference baseband signal. The signal acquisition unit 113 then down-converts the radio frequency signal and converts it from analog to digital to a modulated baseband signal, and sends the modulated baseband signal to the processor 114.

[0112] S72: Reconstruct the modulated baseband signal to generate a baseband feedback signal used to characterize the distortion of the IQ modulation channel;

[0113] Specifically, since the processor 114 includes a signal acquisition unit 1141, a judgment unit 1142, a signal reconstruction unit 1143, and a correction coefficient estimation unit 1144, before reconstructing the modulated baseband signal, the signal acquisition unit 1141 needs to acquire the modulated baseband signal. Simultaneously, the judgment unit 1142 determines whether pre-distortion processing of the baseband reference signal input to the IQ modulation channel 112 is required based on the stability or signal strength of the modulated baseband signal. When the judgment unit 1142 determines that pre-distortion processing of the baseband reference signal input to the IQ modulation channel 112 is required, the modulated baseband signal is input to the signal reconstruction unit 1143, which reconstructs the RF signal to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel 112.

[0114] The reconstruction here refers to a series of preprocessing steps performed on the acquired RF signal to eliminate linear distortions in the IQ modulation channel 112 (such as frequency offset, time delay, sampling rate error, etc.), thereby extracting the effective baseband feedback signal that only reflects the amplitude and phase mismatch of the I and Q channels. Therefore, the baseband feedback signal is the baseband digital signal obtained after compensating for distortions such as frequency offset, time delay, and sampling rate error in the RF signal. It only reflects the amplitude and phase mismatch of the I and Q channels, that is, only reflects the distortion of the IQ modulation channel 112, in order to compare it with the reference baseband signal.

[0115] In some embodiments, the modulated baseband signal is reconstructed to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel 112, including:

[0116] The modulated baseband signal is preprocessed to generate a baseband feedback signal for characterizing the distortion of the IQ modulation channel 112. The preprocessing may include at least one of frequency offset compensation, signal synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration.

[0117] Frequency offset compensation refers to the compensation for the frequency offset generated during the transmission of the modulated baseband signal in the channel, eliminating the systematic carrier frequency deviation introduced during signal generation, modulation, or transmission.

[0118] Signal synchronization refers to identifying the starting position of each full-cycle signal in the modulated baseband signal within the first processed signal (i.e., the acquired radio frequency signal) in order to perform period truncation and subsequent averaging noise reduction. Period truncation refers to cutting independent, equal-length symbol segments from the continuous signal stream after symbol synchronization is completed, based on the known symbol period length, to remove the cyclic prefix at the beginning of each symbol, retaining only the part carrying the valid information.

[0119] Joint phase error compensation refers to the joint compensation of residual carrier frequency offset, sampling rate clock frequency deviation, and timing deviation in the modulated baseband signal.

[0120] Signal-to-noise ratio (SNR) optimization refers to using the periodic structure of the modulated baseband signal to perform time-domain averaging of multiple complete periodic signals after signal synchronization, in order to suppress random noise and improve the SNR, thereby providing a high-quality baseband feedback signal for the accurate estimation of subsequent IQ imbalance parameters.

[0121] Complex gain calibration refers to performing complex gain alignment on the baseband feedback signal after signal-to-noise ratio optimization, so that its amplitude and phase are consistent with the modulated baseband signal on an overall scale, thereby eliminating the common gain offset introduced by the loopback channel and ensuring the accuracy of subsequent IQ imbalance parameter estimation.

[0122] The above preprocessing methods can be performed partially or fully. When performing full preprocessing on the modulated baseband signal, the steps can be as follows: frequency offset compensation -> signal synchronization -> joint phase error compensation -> signal-to-noise ratio optimization -> complex gain calibration. This can also eliminate linear distortion in the IQ modulation channel 112.

[0123] Through the above preprocessing, the noisy and interference-laden radio frequency signal is transformed into a high-fidelity and robust baseband feedback signal, thereby improving the accuracy, robustness, and refinement of the distortion characterization of the IQ modulation channel 112, and providing data support for subsequent improvement of the communication system performance.

[0124] In some cases, preprocessing of the modulated baseband signal includes:

[0125] Frequency offset compensation is performed on the modulated baseband signal to obtain the first processed signal;

[0126] The first processed signal is symbol-synchronized and periodically truncated to obtain the second processed signal;

[0127] The second processed signal is subjected to joint phase error compensation for time delay, frequency offset, and sampling rate error to obtain the third processed signal;

[0128] The signal-to-noise ratio is optimized and the complex gain is calibrated on the third processed signal to generate the baseband feedback signal.

[0129] Time delay refers to the overall time lag or lead of the modulated baseband signal relative to the reference baseband signal. Frequency offset refers to the residual frequency offset remaining after the first step of frequency offset compensation. Sampling rate error refers to the small deviation between the actual sampling frequency of the analog-to-digital converter at the receiver and the clock frequency of the digital-to-analog converter at the transmitter. Signal-to-noise ratio (SNR) optimization refers to suppressing out-of-band noise and narrowband interference through adaptive filtering; or using symbol decision feedback to clean the original signal with the estimated value after initial demodulation, iteratively improving the SNR. Complex gain calibration includes both amplitude scaling and phase rotation to eliminate the amplitude non-uniformity and fixed phase offset introduced by devices in IQ imbalance.

[0130] Since frequency offset compensation refers to compensating for the frequency of the modulated baseband signal to match the frequency of the reference baseband signal, the first processed signal is a baseband signal after preliminary coarse frequency correction. Because symbol synchronization and period truncation are performed on the first processed signal, the second processed signal consists of a series of aligned and truncated independent symbol units. Since joint phase error compensation involves simultaneously estimating the coupled error parameters of time delay, frequency offset, and sampling rate errors using an algorithm, and then performing unified correction, the third processed signal is a baseband signal synchronized in terms of time, frequency, and sampling clock. The compensation operation can include adjusting the sampling time through interpolation resampling to correct time delay and sampling rate errors, as well as performing subtle frequency offset rotation. Because signal-to-noise ratio optimization and complex gain calibration are performed on the third processed signal, the generated baseband feedback signal is comparable to the reference baseband signal.

[0131] Specifically, since the signal reconstruction unit 1143 includes a frequency offset compensation subunit 11431, a symbol synchronization subunit 11432, a joint phase error compensation subunit 11433, a signal-to-noise ratio optimization subunit 11434, and a gain calibration subunit 11435, when the judgment unit determines that the baseband reference signal input to the IQ modulation channel 112 needs to be pre-distorted, the frequency offset compensation subunit 11431 in the signal reconstruction unit 1143 can be used to compensate for the frequency offset of the modulated baseband signal to obtain a first processed signal. The symbol synchronization subunit 11432 can be used to synchronize the symbol and truncate the period of the first processed signal to obtain a second processed signal. The joint phase error compensation subunit 11433 can be used to compensate for the joint phase error of the second processed signal for time delay, frequency offset, and sampling rate error to obtain a third processed signal. The signal-to-noise ratio of the third processed signal can be optimized by the signal-to-noise ratio optimization subunit 11434, and the complex gain of the noise-reduced third processed signal can be calibrated by the gain calibration subunit 11435 to generate a baseband feedback signal.

[0132] By sequentially performing frequency offset compensation, symbol synchronization and period truncation, joint time-frequency domain phase error compensation, signal-to-noise ratio optimization, and complex gain calibration, various complex impairments introduced during signal transmission and acquisition are systematically eliminated. This allows for the gradual recovery of modulated baseband signals with carrier frequency offset, timing deviation, sampling rate error, and noise interference into high-quality baseband feedback signals that are strictly aligned with the reference baseband signal in time, frequency, amplitude, and phase. It overcomes the performance degradation of traditional calibration methods caused by the coupling of various impairments in broadband systems, laying a precise and reliable signal foundation for the subsequent high-precision and high-efficiency extraction of the intrinsic distortion parameters of the IQ modulation channel 112, thus ensuring the final performance and robustness of the entire unbalanced calibration system.

[0133] Specifically, the frequency offset compensation subunit 11431 in the signal reconstruction unit 1143 is used to perform frequency offset compensation on the modulated baseband signal to obtain the first processed signal, including: acquiring multiple cycles of modulated baseband signal, calculating the phase difference of the multiple cycles of modulated baseband signal under the same index symbol; correcting the signal frequency offset of the modulated baseband signal based on the phase difference to obtain the first processed signal.

[0134] The baseband signal generator 111 can output multiple periods of reference baseband signals, for example, transmitting these multiple periods of reference baseband signals to the IQ modulation channel 112 respectively. Each period of the reference baseband signal can include multiple reference signals. Multiple reference signals can correspond to multiple subcarriers. For example, one reference signal can correspond to one subcarrier. The reference baseband signal has a cyclic prefix, therefore the modulated baseband signal has a cyclic prefix, so two signal segments separated by the length of the symbol body within a symbol should theoretically be the same. The method for determining the same index symbol includes: the frequency offset compensation subunit 11431 performs autocorrelation on the signal to find its correlation peak with its own delay (these correlation peaks identify the boundaries of each symbol), then the symbols with the same relative position in different periods found by autocorrelation are the same index symbol, wherein multiple reference signals in each period can each correspond to one index symbol.

[0135] To illustrate, assume the reference baseband signal includes 500 reference signals. The index symbol of the first reference signal could be, for example, "001", the index symbol of the second reference signal could be, for example, "002", and so on. It can be understood that each index symbol can correspond to the same reference signal in different periods of the reference baseband signal, as well as the feedback signal of that reference signal in the baseband feedback signal. Assuming no distortion or other effects, and that the baseband feedback signal is also a periodic signal, the feedback signal with index symbol "001" in one period of the baseband feedback signal corresponds to the reference signal with index symbol "001" in the reference baseband signal.

[0136] Please refer to Figure 8 , Figure 8 This is a schematic diagram of a multi-cycle reference baseband signal and a baseband feedback signal. Figure 8The diagram illustrates a 100-cycle reference baseband signal, with each cycle comprising 500 reference signals: d0, d1, ..., d499. The index symbol for the first reference signal d0 in each cycle is "001", the index symbol for the second reference signal d1 is "002", ..., and the index symbol for the 500th reference signal d499 is "500". Correspondingly, for example, signal acquisition unit 113 can acquire 100 cycles of baseband feedback signals. Each cycle of baseband feedback signals may include s0, s1, ..., s499. The index symbol for the first signal in the baseband feedback signals is "001", the index symbol for the second signal is "002", and the index symbol for the 500th signal is "500". For example, the first reference signal corresponding to index symbol "001" corresponds to the first signal in the baseband feedback signals.

[0137] The frequency offset compensation subunit 11431 can determine the phase difference between the reference signal and the radio frequency signal corresponding to the same index symbol in different periods. Then, the frequency offset calibration parameters are determined based on these phase differences. For example, the frequency offset calibration parameters can be determined based on the average of the phase differences corresponding to multiple periods. The specific process is as follows: Assume the radio frequency signal Sx(n) satisfies the following formula:

[0138] (1);

[0139] Among them, phase difference for: The phase difference of a symbol in two cycles (e.g., the i-th cycle and the j-th cycle) can be determined by comparing the phases of the same symbol in those two cycles. .

[0140] Specifically, the processor 114 can use the above formula (1) to calculate the phase difference corresponding to the same index symbol in multiple cycles, and then calculate the average phase difference. It can be determined by the average phase difference. Calculate the frequency offset .

[0141] The aforementioned average phase difference can be either a statistical mean or an arithmetic mean.

[0142] The aforementioned frequency offset compensation subunit 11431, after obtaining the frequency offset Then, the estimated carrier frequency offset can be used. In the time domain, the received signal reference baseband signal is multiplied by a reverse-rotated compensation factor sequence (e.g., a complex exponential sequence) to generate a carrier frequency-aligned signal, which is the first processed signal.

[0143] In these implementations, frequency offset calibration parameters are determined by comprehensively considering the phase difference of multi-cycle reference baseband signals to perform frequency offset calibration, thereby obtaining a more accurate frequency offset calibration result.

[0144] In some cases, symbol synchronization of radio frequency signals includes: identifying the start signal of each full cycle signal in the first processed signal by utilizing the correlation between the reference baseband signal and the first processed signal; and periodically truncating the first processed signal based on the start signal.

[0145] An integer cycle signal refers to the smallest time unit in the reference baseband signal that has a complete repeating structure. This unit repeats periodically in the time domain and contains complete modulation information (such as a frame structure composed of one or more OFDM signals). The start signal is the first signal in the integer cycle signal. Periodic truncation refers to extracting an integer number of complete integer cycle signals from the first processed signal based on the identified start signal position, discarding incomplete preamble or tail portions.

[0146] Understandably, without transmit / receive mismatch, the first signal in the first processed signal should correspond to the first reference signal in the reference baseband signal. However, transmit / receive mismatch and IQ imbalance can cause signal asynchrony. The correlation between each reference signal and its corresponding signal in the RF signal is higher than the correlation between that reference signal and other signals in the RF signal. Therefore, the correlation between any reference signal in the reference baseband signal and its corresponding signal in the RF signal can be determined by the correlation between signals. For example, the discrete cross-correlation function can be used to calculate the correlation between signals.

[0147] In one example, the symbol synchronization subunit 11432 can use the entire cycle of the reference baseband signal as a reference template and perform a cross-correlation operation with the first processed signal; then, it determines the maximum value of the cross-correlation result, and the time delay position corresponding to this maximum value is the starting position of the entire cycle signal; the signal corresponding to the starting position is the starting signal. Starting from the starting signal, the symbol synchronization subunit 11432 extracts p consecutive complete cycles of signal from the first processed signal for subsequent processing, where p is an integer greater than 1.

[0148] The symbol synchronization subunit 11432 synchronizes signals by utilizing the correlation between the reference baseband signal and the first processed signal. This enables precise identification of the starting point of the periodic broadband signal and accurate period truncation accordingly. It effectively overcomes the performance degradation of traditional synchronization methods under low signal-to-noise ratio or complex channel impairments, providing a well-structured and clearly defined data foundation for subsequent time-frequency alignment, multi-cycle averaging, and high-precision parameter estimation. This ensures the reliability of the entire calibration process and the final compensation accuracy.

[0149] In some cases, joint phase error compensation is performed on the second processed signal for time delay, frequency offset, and sampling rate error, including:

[0150] The joint phase error compensation subunit 11433 constructs a time-frequency grid equation based on the reference baseband signal and the second processed signal, so as to perform joint phase error compensation through the time-frequency grid equation.

[0151] The time-frequency grid equation is a mathematical model used to jointly estimate and compensate for multiple linear impairments of a signal in the time and frequency domains. The time-frequency grid equation models the three mutually coupled impairments—time delay, residual frequency offset, and sampling rate error—as a linear relationship on a two-dimensional phase plane with respect to time and frequency, and corrects these impairments all at once by solving a system of equations. The processor 114 can construct the time-frequency grid equation as shown in formula (2) based on the second processed signal and the reference baseband signal.

[0152] In one example, the aforementioned joint phase error compensation subunit 11433 can use the second processed signal obtained after the radio frequency signal has undergone the aforementioned frequency offset processing and symbol synchronization processing, and the reference baseband signal to perform time-frequency grid equations to calibrate the residual frequency offset, symbol timing deviation and sampling rate error.

[0153] Specifically, the above time-frequency grid equation can be characterized, for example, by the following formula (2):

[0154] (2);

[0155] Where k is the subcarrier index, representing different frequency components in the reference baseband signal; k is an integer greater than or equal to 0 (where the index corresponding to the DC subcarrier can be 0). The index of the reference signal (the I-th OFDM number) is an integer greater than or equal to 1. Indicates the first transmitter The symbol corresponding to the kth subcarrier of the reference signal. For the receiver to receive and The corresponding signal. Indicates the residual carrier frequency offset (unit: Hz). Indicates the effective duration of each OFDM symbol (in seconds). , indicating up to the At the end of each symbol, due to frequency offset Total phase rotation resulting from accumulation (unit: cycles, multiplied by) (Then it becomes radians). Symbol timing deviation is the fractional part of the total time delay of a signal from the transmitter to the receiver along the transmission path, i.e., the delay (in seconds). It represents the frequency difference between two adjacent subcarriers, also known as the subcarrier spacing (unit: Hertz). Indicates for the first There are subcarriers with frequencies of . Due to time delay The resulting phase shift is precisely - ×(frequency)×(delay). The sampling clock frequency deviation (or sampling rate error) represents the relative error between the actual sampling rate and the nominal sampling rate of the analog-to-digital converter at the receiving end. It causes a slight difference in the scaling of the time base axis between the receiving and transmitting ends. ϵ Characterizes the phase rotation on the OFDM time-frequency two-dimensional grid that varies with time and frequency coupling caused by sampling clock frequency deviation.

[0156] The joint phase error compensation subunit 11433 can process the observed values ​​of the second processed signal. Substituting into formula (2) above, we can solve for the following: , and Then, based on the solution... , and Inverse compensation is performed. For example, the signal reconstruction unit 1143 can first compensate for the sampling clock error, because it simultaneously couples the distortion in the time and frequency domains; then compensate for the time delay; and finally compensate for the residual frequency offset. The entire compensation process can be integrated into one step.

[0157] In these implementations, the joint phase error compensation subunit 11433 performs joint compensation for residual carrier frequency offset, sampling rate clock frequency deviation, and timing deviation through time-frequency grid equations. By performing a single fitting, it can simultaneously (jointly) estimate the predistortion parameters and compensations corresponding to different frequencies, shortening calibration time and improving calibration efficiency. Furthermore, by utilizing a two-dimensional time-frequency signal structure, the coupling effects of various distortions are mathematically decoupled, avoiding error propagation and making parameter estimation more accurate. This lays a precise foundation for subsequent compensation in the IQ modulation channel 112.

[0158] In some cases, signal-to-noise ratio optimization and complex gain calibration are performed on the third-processed signal, including:

[0159] The third-processed signal is averaged and noise reduced in units of signal period.

[0160] Complex gain alignment is performed based on the denoised third-processed signal and the reference baseband signal.

[0161] After symbol synchronization and time-frequency grid equation processing, the reference baseband signal yields a third-processed signal. Time-domain averaging of multiple cycles of this third-processed signal produces a signal-to-noise ratio (SNR) optimized RF signal. For example, M one-dimensional reference baseband signals (L reference signals per cycle) are reorganized into a two-dimensional matrix of M rows × L columns according to cycle length L. M is an integer greater than or equal to 1, and L is an integer greater than or equal to 1. Each row represents a complete signal cycle, and each column represents a sampling point at the same relative time position across all cycles. An arithmetic mean is calculated for each column (L columns in total) of the matrix; for the i-th sampling position (i=0,1,…,L-1), its denoised value… It can be represented by the following formula (3):

[0162] (3);

[0163] in, This represents the complex value of the i-th sampling point in the m-th period.

[0164] The signal-to-noise ratio optimization subunit 11434 obtains a new sequence of length L through average noise reduction, which is the single-cycle signal after average noise reduction. This sequence is the direct input of the gain calibration subunit.

[0165] For the denoised third processed signal, the signal-to-noise ratio optimization subunit 11434 can calculate the complex gain between the reference baseband signal and the third processed signal based on the least squares method; then, the third processed signal is calibrated by amplitude scaling and phase rotation according to the complex gain to achieve complex gain alignment.

[0166] For example, the signal-to-noise ratio optimization subunit 11434 can use least-squares estimation to compare the relationship between the third processed signal and the reference baseband signal over multiple sampling points or a complete cycle to calculate a complex gain factor that includes information on the amplitude ratio and phase difference. Illustratively, if the overall amplitude of the third processed signal is A times the amplitude of the reference baseband signal, the overall phase leads... Then the complex gain G can be characterized by the following formula (4):

[0167] (4);

[0168] In one example, the signal-to-noise ratio optimization subunit 11434 can apply the inverse (or conjugate) of the complex gain to the third processed signal to perform amplitude scaling and phase rotation calibration on the third processed signal.

[0169] By averaging the aligned signal period and combining it with the reference baseband signal for complex gain calibration, on the one hand, period averaging effectively suppresses random noise, significantly improves the signal-to-noise ratio, and enhances the stability and anti-interference capability of subsequent parameter estimation; on the other hand, complex gain alignment eliminates the overall amplitude scaling and constant phase shift introduced during signal transmission, ensuring that the feedback signal and the reference signal are accurately comparable in the complex domain.

[0170] S73: Based on the reference baseband signal and the baseband feedback signal, the predistortion parameters used to compensate for IQ imbalance are obtained.

[0171] The signal reconstruction unit 1143 can transmit the baseband feedback signal obtained by reconstructing the reference baseband signal to the correction coefficient estimation unit 1144. The correction coefficient estimation unit 1144 can process the reference baseband signal and the baseband feedback signal to obtain the predistortion parameters.

[0172] Specifically, the correction coefficient estimation unit 1144 can calculate the reference baseband signal. The aforementioned predistortion parameters are determined based on the differences described above. These predistortion parameters may include, for example, filter parameters for filtering the reference baseband signal, filter coefficients for filtering the conjugate signal of the reference baseband signal, and correction coefficients for suppressing local oscillator leakage spurious signals.

[0173] The difference between the aforementioned calculated reference baseband signal and the baseband feedback signal refers to the system parameters characterizing the distortion of the IQ modulation channel 112 extracted from their relationship through mathematical modeling and parameter estimation algorithms.

[0174] Assumption Baseband feedback signal caused by distortion of IQ modulation channel 112 The model is represented by the following formula (5):

[0175] (5);

[0176] in, , Main roadbed gain, The main path phase (combined response of I and Q paths). , For mirror path gain, The phase of the mirror roadbed is denoted by d (caused by IQ amplitude / phase imbalance). d is the local oscillator leakage, which manifests as DC offset (complex constant). for The conjugate of . The correction coefficient estimation unit 1144 can be through and baseband feedback signal Using least squares fitting and d. The above d is the reference baseband signal. With baseband feedback signal The differences between them were quantified.

[0177] The pre-distortion objective is to design a pre-compensation model to perform digital domain inverse distortion on the reference baseband signal to offset the distortion in the analog domain. The pre-compensation model can be characterized by the following formula (6):

[0178] (6);

[0179] , These are the filter coefficients. This represents the local oscillator leakage suppression coefficient. The pre-distortion parameters of the above pre-compensation model include... , and .

[0180] The above pre-compensation model H(n) needs to satisfy the following formula (7) to achieve digital domain inverse distortion:

[0181] (7);

[0182] Ignoring higher-order terms, the approximate solution to the above formula (6) is:

[0183] ; ; ;

[0184] Specifically, the correction coefficient estimation unit 1144 can determine the parameters of the first complex filter 11451 and the second complex filter 11452, as well as the intrinsic leakage suppression coefficient of the local oscillator leakage suppressor 11453, based on the difference between the calculated reference baseband signal and the baseband feedback signal.

[0185] The predistortion parameters refer to a set of filter coefficients and intrinsic leakage suppression coefficients calculated by the processor 114 through an algorithm, which are used to perform inverse distortion processing on the transmitted signal in the digital baseband domain.

[0186] By using a reference baseband signal and a baseband feedback signal, predistortion parameters are calculated. These predistortion parameters may include the parameters of the first complex filter 11451 corresponding to the reference baseband signal, the parameters of the second complex filter 11452 corresponding to the conjugate signal of the reference baseband signal, and the intrinsic leakage suppression coefficient of the local oscillator leakage suppressor 11453. This allows for high-precision and high-efficiency compensation of frequency-selective IQ imbalance in broadband systems, improves the image rejection ratio, and enhances signal quality.

[0187] In some embodiments, step S73 includes: the predistortion processing unit 1145 calculating predistortion parameters based on the reference baseband signal and the baseband feedback signal using a nonlinear optimization algorithm. For example, suppose the reference baseband signal can be expressed as: = +j ); where, the I-channel signal is The Q-channel signal is The calculation methods corresponding to the nonlinear optimization algorithm can be shown by the following equations (8), (9), and (10):

[0188] (8);

[0189] (9);

[0190] (10);

[0191] in, For the number of iterations, This is the transpose of the predistortion parameters obtained from the k-th iteration. Table transpose, The channel response of the total transmission channel in the process from the transmitter in a broadband zero-IF architecture to obtaining the modulated baseband signal; This is the transpose of the predistortion parameters obtained from the calibration information obtained in the k-th iteration. This is the preprocessing procedure for the modulated baseband signal. To compensate for errors, Autocorrelation array for reference signal The inverse matrix, through multiple iterations, makes... Convergence is achieved, completing the precise compensation for IQ imbalance. Through multiple iterations, this process is completed. Convergence completes the precise compensation for broadband IQ imbalance.

[0192] In one example, the iteration termination condition includes at least one of the following: the output signal quality of the IQ modulation channel 112 meets a preset threshold, the number of iterations reaches a preset maximum number, and the change in the predistortion parameter calculated in two adjacent iterations is less than a preset tolerance.

[0193] The iteration process can be terminated when the above iteration termination condition is met. The coefficients of the first complex filter 11451 and the coefficients of the second complex filter 11452 obtained at the end of the iteration are used as their respective predistortion parameters.

[0194] S74: Use predistortion parameters to predistort the reference baseband signal to generate a baseband predistortion signal.

[0195] Predistortion processing refers to an active, inverse process of superimposing nonlinear and linear distortions on the original baseband signal to be transmitted in the digital baseband domain, so that the output signal after passing through the actual IQ modulation channel 112 is as close as possible to the ideal distortion-free signal. The baseband predistorted signal can be obtained by digitally distorting a reference baseband signal using predistortion parameters, with the aim of canceling the distortion introduced by subsequent analog channels.

[0196] In step S74 above, the predistortion processing unit 1145 uses predistortion parameters to filter the reference baseband signal and its conjugate signal to generate a baseband predistortion signal. Specifically, since IQ channel imbalance mathematically causes its output signal to include not only the original signal but also its conjugate (mirror) component, to cancel it out, the predistortion processing unit 1145 needs to include a first complex filter 11451 corresponding to the reference baseband signal, a second complex filter 11452 corresponding to the conjugate signal of the reference baseband signal, a local oscillator leakage suppressor 11453, and a second adder 11454. The filter coefficients can include the parameters of the first complex filter 11451 and the parameters of the second complex filter 11452. Specifically, the first complex filter 11451 is used to adjust the in-phase and quadrature components of the reference baseband signal, thereby performing pre-equalization within the main signal frequency band to ensure that the reference baseband signal can still maintain ideal characteristics after passing through the IQ modulation channel 112. Meanwhile, due to the imperfections of the IQ modulation channel 112 (such as amplitude imbalance and phase non-orthogonality between the I / Q channels), a carrier-centered image band signal is generated. The second complex filter 11452 is used to predict the image interference signal based on the reference baseband signal and generate its inverse signal. When this inverse image signal is superimposed with the image interference actually generated by the subsequent IQ modulation channel 112, the two can cancel each other out. In addition, due to the DC bias mismatch of the IQ modulator, the local oscillator signal leaks directly to the RF output, which manifests as an undesirable spike at the carrier frequency in the spectrum. Therefore, in the predistortion processing unit 1145, the local oscillator leakage suppressor 11453 is used to suppress local oscillator leakage. It mainly generates a DC correction by multiplying the suppression coefficient within it by the absolute mean of the input signal. After the DC correction is applied to the reference baseband signal, it can first cancel the DC bias of the two channels in the IQ modulation channel 112, thereby suppressing the generation of local oscillator leakage. The second adder 11454 is used to vector-add the output signals of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453 to synthesize a baseband predistortion signal. Thus, the baseband signal generator 111 can input a reference baseband signal to the first complex filter 11451, input the conjugate signal of the reference baseband signal to the second complex filter 11452, and input the mean of the absolute values ​​of the full-cycle reference baseband signal to the local oscillator leakage suppressor 11453. The local oscillator leakage suppressor 11453 stores a local oscillator leakage suppression coefficient, which is used to multiply by the mean of the absolute values ​​of the full-cycle reference baseband signal to suppress local oscillator leakage. The three input terminals of the second adder 11454 are respectively connected to the outputs of the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453.The second adder 11454 vector-adds the first complex filter 11451, the second complex filter 11452, and the local oscillator leakage suppressor 11453.

[0197] The predistortion parameters (filter coefficients and intrinsic leakage suppression coefficients) can be input to the predistortion processing unit 1145 to update the filter coefficients and the intrinsic leakage suppression coefficients of the local oscillator leakage suppression unit in the predistortion processing unit 1145. In one example, the correction coefficient estimation unit 1144 can input the predistortion parameters obtained in step S73 to the predistortion processing unit 1145 to update the coefficients of the predistortion processing unit 1145. For example, this updates the coefficients of the first complex filter 11451, the coefficients of the second complex filter 11452, and the intrinsic leakage compensation coefficients of the intrinsic leakage compensation module.

[0198] The predistortion processing unit 1145 performs conjugate processing and absolute value and average processing on the reference baseband signal input therein, and inputs the predistortion parameters into the reference baseband signal, the conjugate reference baseband signal and the reference baseband signal after absolute value and average processing, respectively, and finally obtains the baseband predistortion signal.

[0199] In some implementations, a baseband predistortion signal is generated by filtering the reference baseband signal and its conjugate signal using predistortion parameters, including:

[0200] Using predistortion parameters, the reference baseband signal and its conjugate signal are filtered in parallel. The superimposed signal of the two filtering results is then combined with the local oscillator leakage suppression coefficient to generate the baseband predistortion signal.

[0201] The analog path of the IQ modulator has a DC offset, which is treated as a fixed baseband signal by the modulation channel and up-converted to RF, generating a carrier leakage signal. To suppress the carrier leakage signal, the baseband signal generator 111 can pre-apply a local oscillator leakage suppression coefficient to the reference baseband signal before inputting it into the IQ modulation channel 112, thereby correcting for the local oscillator leakage spurious signal.

[0202] The baseband signal generator 111 can simultaneously input the reference baseband signal to the first complex filter 11451 and input the conjugate signal of the reference baseband signal to the second complex filter 11452, and perform parallel filtering processing on the reference baseband signal and its conjugate signal.

[0203] The mean of the absolute value of the reference baseband signal is input to the local leakage compensation module. In the local leakage compensation module, the mean of the absolute value of the reference baseband signal is multiplied by the intrinsic leakage compensation coefficient to obtain a first product. In the first complex filter 11451, the reference baseband signal is convolved with the first complex filter 11451 to obtain a first result. In the second complex filter 11452, the conjugate signal of the reference baseband signal is convolved with the second complex filter 11452 to obtain a second result. The first result, the second result, and the first product are added together and used as the output of the predistortion processing unit 1145. The above process can be characterized by the following formula (11):

[0204] (11);

[0205] in, This is the baseband predistortion signal. For reference baseband signal, The conjugate signal of the reference baseband signal, The mean of the absolute values ​​of the reference baseband signal, For the first complex filter 11451, For the second complex filter 11452, " is the symbol for convolution calculation. This is the local oscillator leakage suppression coefficient.

[0206] S75: Performs unbalanced calibration on the IQ modulation channel based on the baseband predistortion signal.

[0207] The predistortion processing unit 1145 can input the baseband predistortion signal into the IQ modulation channel 112, replacing the reference baseband signal, to achieve unbalanced calibration of the IQ modulation channel 112. Specifically, since the baseband predistortion signal is a complex signal, the real part of the baseband predistortion signal is input into the I path of the IQ modulation channel 112, and the imaginary part is input into the Q path of the IQ modulation channel 112. Because the predistortion signal compensates for both the amplitude and phase generated after the signal is output to the IQ modulation channel 112, when the baseband predistortion signal is input into the IQ modulation channel 112, no amplitude or phase imbalance will occur in the I and Q paths of the IQ modulation channel 112, thus achieving effective calibration of the IQ modulation channel.

[0208] After the predistortion parameters of the predistortion processing unit 1145 are no longer updated, all subsequent signals are predistorted by the predistortion processing unit 1145. In one example, the method further includes:

[0209] The predistortion processing unit 1145 inputs the baseband predistortion signal to the IQ modulation channel 112, and after performing unbalanced calibration on the IQ modulation channel 112, includes the following iterative steps:

[0210] The signal reconstruction unit 1143 performs signal recovery processing on the output new modulated baseband signal to obtain the updated baseband feedback signal;

[0211] The correction coefficient estimation unit 1144 recalculates the predistortion parameters based on the reference baseband signal and the updated baseband feedback signal. The predistortion processing unit 1145 uses the recalculated predistortion parameters to generate a new round of baseband predistortion signal. The iterative optimization steps are repeated until the preset iteration termination condition is met. Through multiple iterations, the determined predistortion parameters can be more accurately used for pre-compensation of the reference baseband signal, thereby further improving the accuracy of the unbalanced calibration of the IQ modulation channel 112.

[0212] In this embodiment, by acquiring the reference baseband signal and the modulated baseband signal output by the IQ modulation channel 112, and performing signal recovery processing on the modulated baseband signal to obtain the baseband feedback signal, the linear impairments such as frequency offset and phase offset introduced by the transmission channel are effectively eliminated. This ensures that the baseband feedback signal can accurately reflect the gain and phase imbalance of the analog components (such as mixers, amplifiers, and local oscillators) of the IQ modulation channel 112. Based on this, predistortion parameters are calculated based on the reference baseband signal and the baseband feedback signal, which can comprehensively and accurately extract the gain and phase imbalance characteristics of the channel itself. Finally, the reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal and calibrate the IQ modulation channel 112. This achieves deep compensation for IQ imbalance, significantly improves the image rejection ratio and error vector amplitude performance of the output signal, and thus improves the calibration accuracy. Furthermore, using a reference baseband signal can simultaneously compensate for distortions generated by signals of different frequencies in the transmission channel, which improves the calibration efficiency of the IQ modulation channel 112 compared to calibrating multiple frequencies separately.

[0213] The calculation process of the predistortion parameter is illustrated below with an example. The RF output of the signal transmitter 11 is connected to the signal acquisition unit 113 (vector signal analyzer). Reference baseband signal T... x (n) is an OFDM signal with a sampling rate of 625MHz and a bandwidth of 500MHz. After digital-to-analog conversion and low-pass filtering, it is modulated to a frequency of 6GHz. The IQ radio frequency signal is acquired by a vector signal analyzer and down-converted and analog-to-digital converted to obtain the radio frequency signal t.

[0214] Step 1: Compensate for the frequency shift generated during signal transmission by calculating the phase difference of the same index symbol in different periods of signal t, denoted as signal t1.

[0215] Step 2: Use cross-correlation to synchronize t1 with the reference baseband signal. Extract N (N≥3) positive period OFDM signals from the acquired symbols and denote them as signal t2.

[0216] Step 3: Based on signal t2 and reference baseband signal T x (n), construct a time-frequency grid equation system, calculate the phase error, and further compensate for the signal's time delay, frequency offset, and sampling rate error, denoted as signal t3.

[0217] Step 4: Perform average noise reduction on t3 in periodic units, and base the noise-reduced signal on the reference baseband signal T. x (n) Perform complex gain alignment to obtain the baseband reference signal R. x (n).

[0218] Step 5: Employ the Newton-least-squares iterative algorithm based on R. x (n) and reference baseband signal R x (n) Calculate the IQ unbalanced filter coefficients based on R... x (n) and reference baseband signal T x The mean value of (n) is used to calculate the local oscillator leakage stray emission suppression correction coefficient.

[0219] The reference signal is pre-distorted using the aforementioned coefficients to correct the IQ modulation distortion generated during the modulation process. To clearly demonstrate the calibration effect, a single-sided bandwidth OFDM signal is transmitted for verification. Figure 9 The broadband waveform spectrum before calibration. Figure 10 This is the calibrated broadband waveform spectrum. For example... Figure 9 As shown, before calibration, there was a strong mirror signal on the left (the power of the mirror signal was greater than 10 dBm), such as... Figure 10 As shown, after calibration, the mirror signal on the left (the power of the mirror signal is less than 0dBm) is reduced. Therefore, by using the above method to perform unbalanced calibration on the IQ modulation channel 112, the mirror signal can be effectively suppressed, thereby improving the calibration accuracy.

[0220] The signal transmitter 11 of this application embodiment includes, in addition to the processor 114 of the above embodiment, a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor 114, it implements the steps of the unbalanced calibration method of the IQ modulation channel as described in the above embodiment.

[0221] Each of the above modules or units can be implemented through software, hardware, or a combination of both. In this application, "implemented through software" means that the processor reads and executes program instructions stored in memory to implement the functions corresponding to the above modules or units. Here, the processor refers to a processing circuit capable of executing program instructions, including but not limited to at least one of the following: a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a microcontroller unit (MCU), or an artificial intelligence processor, or other processing circuits capable of running program instructions. In some embodiments, the processor may also include circuits with other processing functions (such as hardware circuits for hardware acceleration, bus and interface circuits, etc.). Processors can be presented as integrated chips, for example, as integrated chips whose processing functions only include executing software instructions, or they can be presented as SoCs (system on a chip), that is, on a single chip, in addition to the processing circuits (usually called "cores") that can run program instructions, there are also other hardware circuits for implementing specific functions (of course, these hardware circuits can also be implemented separately based on ASICs or FPGAs). Correspondingly, in addition to executing software instructions, the processing functions can also include various hardware acceleration functions (such as AI calculations, encoding and decoding, compression and decompression, etc.).

[0222] In this application, "implemented in hardware" means that the functions of the above-mentioned modules or units are implemented through hardware processing circuits that do not have program instruction processing capabilities. These hardware processing circuits can be composed of discrete hardware components or integrated circuits. To reduce power consumption and size, integrated circuits are typically used. Hardware processing circuits can include ASICs (application-specific integrated circuits) or PLDs (programmable logic devices); PLDs can include FPGAs (field-programmable gate arrays), CPLDs (complex programmable logic devices), and so on. These hardware processing circuits can be a single packaged semiconductor chip (e.g., packaged as an ASIC); or they can be integrated with other circuits (e.g., CPUs, DSPs) and packaged into a single semiconductor chip. For example, multiple hardware circuits and a CPU can be formed on a silicon substrate and packaged into a single chip; this type of chip is also called a SoC. Alternatively, circuits for implementing FPGA functions and a CPU can be formed on a silicon substrate and encapsulated into a single chip; this type of chip is also called a SoPC (system on a programmable chip).

[0223] It should be noted that when this application is implemented through software, hardware, or a combination of both, different software or hardware can be used, and it is not limited to using only one type of software or hardware. For example, one module or unit can be implemented using a CPU, while another module or unit can be implemented using a DSP. Similarly, when implemented using hardware, one module or unit can be implemented using an ASIC, while another module or unit can be implemented using an FPGA. Of course, it is not limited to using the same software (e.g., all through a CPU) or the same hardware (e.g., all through an ASIC) to implement some or all modules or units. Furthermore, those skilled in the art will understand that software is generally more flexible but less performant than hardware, while hardware is the opposite. Therefore, those skilled in the art can choose software, hardware, or a combination of both based on actual needs.

[0224] The above-described preferred embodiments have further illustrated the purpose, technical solutions, and advantages of the present invention. It should be understood that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of calibration of IQ imbalance, characterized by, include: Acquire a reference baseband signal and an RF signal output by the IQ modulation channel based on the reference baseband signal, and convert the RF signal into a modulated baseband signal; The modulated baseband signal is preprocessed to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel; wherein, the preprocessing includes frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration; Based on the reference baseband signal and the baseband feedback signal, predistortion parameters for compensating IQ imbalance are obtained; The reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal; Based on the baseband predistortion signal, the IQ modulation channel is subjected to unbalanced calibration.

2. The imbalance calibration method according to claim 1, characterized in that, The step of converting the radio frequency signal into a modulated baseband signal includes: The radio frequency signal is down-converted and then converted from analog to digital to a modulated baseband signal.

3. The method according to claim 1, characterized in that, The preprocessing of the modulated baseband signal includes: The modulation baseband signal is frequency offset compensated to obtain the first processed signal; The first processed signal is symbol-synchronized and periodically truncated to obtain the second processed signal; The second processed signal is subjected to joint phase error compensation for time delay, frequency offset, and sampling rate error to obtain the third processed signal; The signal-to-noise ratio is optimized and the complex gain is calibrated on the third processed signal to generate the baseband feedback signal.

4. The method according to claim 3, characterized in that, The step of performing frequency offset compensation on the modulated baseband signal to obtain a first processed signal includes: Acquire the modulated baseband signal for multiple cycles, and calculate the phase difference of the modulated baseband signal for multiple cycles under the same index symbol; The signal frequency offset of the modulated baseband signal is corrected based on the phase difference to obtain the first processed signal.

5. The imbalance calibration method according to claim 3, characterized in that, The symbol synchronization and periodic truncation of the first processed signal includes: By utilizing the correlation between the reference baseband signal and the first processed signal, the starting signal of each integer cycle signal in the first processed signal is identified; Based on the starting signal, the first processed signal is periodically truncated.

6. The imbalance calibration method according to claim 3, characterized in that, The joint phase error compensation for time delay, frequency offset, and sampling rate error of the second processed signal includes: Based on the reference baseband signal and the second processed signal, a time-frequency grid equation is constructed to perform the joint phase error compensation.

7. The method according to claim 3, characterized in that, The signal-to-noise ratio optimization and complex gain calibration of the third processed signal include: The third processed signal is averaged and noise reduced in units of signal period; Complex gain alignment is performed based on the denoised third processed signal and the reference baseband signal.

8. The method according to any one of claims 1 to 7, characterized in that, The process of obtaining predistortion parameters for compensating IQ imbalance based on the reference baseband signal and the baseband feedback signal includes: Based on the reference baseband signal and the baseband feedback signal, calculate the predistortion parameters; The step of predistorting the reference baseband signal using the predistortion parameters to generate a baseband predistorted signal includes: Using the predistortion parameters, the reference baseband signal and its conjugate signal are filtered to generate a baseband predistortion signal.

9. The method according to claim 8, characterized in that, The step of using the predistortion parameters to filter the reference baseband signal and its conjugate signal to generate a baseband predistortion signal includes: Using the predistortion parameters, the reference baseband signal and its conjugate signal are subjected to parallel filtering, and the superimposed signal of the two filtering results is combined with the local oscillator leakage suppression coefficient to generate the baseband predistortion signal.

10. The method according to claim 8, characterized in that, The reference baseband signal and the baseband feedback signal are used as the basis. Calculate the predistortion parameters, including: Based on the reference baseband signal and the baseband feedback signal, the predistortion parameters are calculated using a nonlinear optimization algorithm.

11. An imbalance calibration device, characterized in that, include: The signal acquisition unit is used to acquire the modulated baseband signal from the signal acquisition unit; The modulation baseband signal is obtained by converting the radio frequency signal output by the IQ modulation channel based on the reference baseband signal by the signal acquisition unit. The reference baseband signal is generated by the signal generator. A signal reconstruction unit is used to preprocess the modulated baseband signal to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel; wherein, the preprocessing includes frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration; a correction coefficient estimation unit is used to obtain predistortion parameters for compensating for IQ imbalance based on the reference baseband signal and the baseband feedback signal. A predistortion processing unit is used to perform predistortion processing on the reference baseband signal using the predistortion parameters to generate a baseband predistortion signal; The calibration execution unit is used to perform unbalanced calibration on the IQ modulation channel based on the baseband predistortion signal.

12. A signal transmitter, characterized in that, include: Signal generator, IQ modulation channel, signal acquisition unit, and processor; The signal generator is connected to the IQ modulation channel and the processor respectively, and is used to generate and send reference baseband signals to the IQ modulation channel and the processor respectively; The signal acquisition unit is connected to the processor and is used to acquire the radio frequency signal output by the IQ modulation channel based on the reference baseband signal, and convert the radio frequency signal into a modulation baseband signal and send it to the processor. The processor is used for: Receive the modulated baseband signal from the signal acquisition unit; The modulated baseband signal is preprocessed to generate a baseband feedback signal characterizing the distortion of the IQ modulation channel; wherein, the preprocessing includes frequency offset compensation, symbol synchronization, joint phase error compensation, signal-to-noise ratio optimization, and complex gain calibration; based on the reference baseband signal and the baseband feedback signal, predistortion parameters for compensating for IQ imbalance are calculated; The reference baseband signal is predistorted using the predistortion parameters to generate a baseband predistortion signal; The predistortion processing unit in the processor controls the baseband predistortion signal to be input to the IQ modulation channel to perform unbalance calibration on the IQ modulation channel.

13. A zero-IF transceiver, characterized in that, include: The signal transmitter as described in claim 12.

14. An electronic device, characterized in that, include: The zero-IF transceiver as described in claim 13.

Images