Baseband radio frequency signal data feedback processing method based on vector modulation

Abstract

The invention provides a baseband radio frequency signal data feedback processing method based on vector modulation, and the method comprises the steps: firstly receiving a baseband signal data set transmitted by a target radio frequency link in a preset frequency band, and the baseband signal data set comprises a plurality of signal frame original modulation parameters and corresponding radio frequency feedback signal waveforms; calculating a feedback error index composed of a phase offset, an amplitude distortion factor and a spectrum leakage factor according to a radio frequency feedback signal waveform, quantifying a modulation error, carrying out time-frequency domain conjoint analysis on the feedback signal waveform, extracting an error feature vector, inputting the error feature vector into a pre-training model for cross-frame correlation learning, and carrying out the correlation learning; according to the method, a dynamic adjustment parameter sequence containing a time sequence dependency relationship is generated, and finally a closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and original modulation parameters, so that precise processing and optimal control of baseband radio frequency signals are realized, and the signal transmission performance is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a baseband radio frequency signal data feedback processing method based on vector modulation. Background Art

[0002] In the field of communications technology, baseband RF signal processing is crucial for achieving efficient and accurate information transmission. As communication systems continue to demand higher signal quality and transmission efficiency, effectively processing baseband RF signals to reduce transmission errors and improve signal performance has become a hot research topic.

[0003] For example, when analyzing signals, most existing technologies process signals based on a single frame, failing to fully recognize the temporal dependencies and correlations between signal frames. Because signals in actual communications are continuous and dynamically changing, this approach, which ignores inter-frame relationships, makes signal processing unable to adapt to the dynamic nature of signals and respond effectively to signal changes in a timely manner, reducing the real-time and effectiveness of signal processing.

[0004] Furthermore, existing signal control methods are mostly open-loop or employ fixed-parameter adjustment mechanisms. Open-loop control methods cannot make real-time adjustments based on the actual signal transmission conditions, while fixed-parameter adjustment mechanisms struggle to adapt to varying communication environments and signal characteristics. They lack the flexibility to adapt to complex and ever-changing application scenarios, resulting in unstable signal transmission quality and failing to meet the high-precision and high-reliability signal processing requirements of modern communication systems. Summary of the Invention

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a baseband radio frequency signal data feedback processing method based on vector modulation, the method comprising: receiving a baseband signal dataset transmitted by a target radio frequency link within a preset frequency band, wherein the baseband signal dataset includes original modulation parameters of a plurality of signal frames arranged in a time series, and a radio frequency feedback signal waveform corresponding to each signal frame; Calculating a corresponding feedback error index based on the RF feedback signal waveform, wherein the feedback error index is composed of a phase offset, an amplitude distortion, and a spectrum leakage factor, and is used to quantify a modulation error of a signal frame during transmission; Based on the feedback error indicator, a time-frequency domain joint analysis is performed on the RF feedback signal waveform of each signal frame to extract the error feature vector of each signal frame, and the error feature vector is input into a pre-trained error prediction model for cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing a timing dependency relationship; A closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and the original modulation parameters of the multiple signal frames.

[0006] On the other hand, an embodiment of the present invention also provides a communication service system, including a processor and a machine-readable storage medium, wherein the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0007] Based on the above aspects, the embodiment of the present application receives a baseband signal dataset containing original modulation parameters and RF feedback signal waveforms, and then defines feedback error indicators consisting of phase offset, amplitude distortion, and spectral leakage factor. This can highly accurately quantify the modulation error of the signal frame during transmission. Next, a joint time-frequency domain analysis combined with a pre-trained error prediction model is used for cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing temporal dependencies, fully exploiting the characteristics of the signal in different domains. The error prediction model is then used for cross-frame learning to effectively capture the dynamic changes between frames, thereby generating a dynamic adjustment parameter sequence that can reflect the overall change trend of the signal. Finally, a closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and the original modulation parameters, realizing intelligent and adaptive control of baseband RF signal transmission. This closed-loop feedback mechanism can automatically adjust relevant parameters in real time according to the transmission error and dynamic changes of the signal, effectively reducing the modulation error during signal transmission, improving the accuracy and stability of signal transmission, and significantly improving the overall performance and reliability of the baseband RF signal processing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 The present invention provides a schematic diagram of the execution flow of a baseband RF signal data feedback processing method based on vector modulation according to an embodiment of the present invention.

[0009] Figure 2 FIG. 4 is a schematic diagram of exemplary hardware and software components of a communication service system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 The present invention provides a flow chart of a method for processing baseband RF signal data feedback based on vector modulation according to an embodiment of the present invention. The method for processing baseband RF signal data feedback based on vector modulation is introduced in detail below.

[0011] Step S110 , receiving a baseband signal dataset transmitted by a target radio frequency link within a preset frequency band, wherein the baseband signal dataset includes original modulation parameters of a plurality of signal frames arranged in a time series, and a radio frequency feedback signal waveform corresponding to each signal frame.

[0012] The target RF link refers to the RF link responsible for transmitting baseband signals within a specific frequency band in a wireless communication system. For example, the RF link in a wireless communication base station is responsible for converting baseband signals into RF signals and transmitting them. The preset frequency band refers to the specific frequency range in which the target RF link operates. For example, the 2.4 GHz frequency band is one of the commonly used frequency bands in wireless communications. The signal frame is the basic unit in the baseband signal data set. Each signal frame contains data of a certain length, which can be transmitted with specific modulation parameters. The original modulation parameters are used to define the parameters for converting digital information into a signal form suitable for RF transmission. For example, the initial setting values ​​of phase and amplitude, etc. The RF feedback signal waveform is the RF signal waveform obtained at a specific feedback point of the RF link after the RF feedback signal is transmitted through the RF link. It carries various information about the RF signal during the actual transmission process, such as interference, nonlinear effects of devices, etc.

[0013] Specifically, consider a scenario where a wireless communication base station transmits data to multiple mobile terminals. The target radio frequency link in the wireless communication base station operates in a specific preset frequency band, such as the 2.4 GHz band. Within this 2.4 GHz band, the base station continuously transmits a baseband signal dataset containing various information. This baseband signal dataset includes the original modulation parameters of multiple signal frames arranged in a time series. These original modulation parameters define how to convert digital information into a signal form suitable for radio frequency transmission. For example, the original modulation parameters may include information such as initial settings for phase and amplitude. The baseband signal dataset also includes the radio frequency feedback signal waveform corresponding to each signal frame. This radio frequency feedback signal waveform is acquired from a specific feedback point in the radio frequency link after the signal is transmitted. For example, a feedback collection point is set after the power amplifier at the base station transmitter. When the radio frequency signal passes through this feedback collection point, its waveform is collected as the radio frequency feedback signal waveform. This radio frequency feedback signal waveform carries various information about the actual transmission process of the radio frequency feedback signal, such as interference encountered and nonlinear effects of components.

[0014] Step S120 , calculating a corresponding feedback error index according to the RF feedback signal waveform, wherein the feedback error index is composed of a phase offset, an amplitude distortion, and a spectrum leakage factor, and is used to quantify a modulation error of a signal frame during transmission.

[0015] The phase offset is the difference between the actual transmission signal phase and the ideal transmission signal phase, reflecting the change in signal phase during transmission. The amplitude distortion is the deviation between the actual transmission signal amplitude and the ideal transmission signal amplitude, reflecting the change in signal amplitude during transmission. The spectrum leakage factor is the degree to which the energy distribution of the signal on the spectrum does not conform to the ideal condition, such as spectrum leakage caused by an imperfect filter or imperfections in the modulation process.

[0016] For example, in the aforementioned wireless communication base station scenario, the corresponding feedback error indicator needs to be calculated for each received RF feedback signal waveform. Taking the signal frame transmitted by the base station as an example, the RF feedback signal waveform is first subjected to orthogonal demodulation. The orthogonal demodulation process can be understood as decomposing the RF signal into a baseband in-phase component sequence and a baseband quadrature component sequence. Then, based on the obtained baseband in-phase component sequence and baseband quadrature component sequence, the instantaneous phase sequence and instantaneous amplitude sequence of each signal frame are calculated. For example, when calculating the instantaneous phase sequence, the relationship between the in-phase component and the quadrature component is used to derive the phase value of each sampling point.

[0017] Next, the instantaneous phase sequence is subjected to point-by-point difference calculations against a preset reference phase sequence to generate a phase difference sequence. This reference phase sequence is pre-determined based on an ideal signal transmission model and represents the phase change of the signal in the absence of any errors. This phase difference sequence is then averaged using a sliding window to obtain the phase offset for each signal frame. For example, if within a signal frame, the actual phase deviates from the reference phase due to instability in certain components in the RF link (such as an oscillator), this deviation is quantified as a phase offset.

[0018] To calculate amplitude distortion, the instantaneous amplitude sequence is normalized against a preset reference amplitude sequence to generate an amplitude difference sequence. This reference amplitude sequence is also based on ideal transmission. In actual transmission, if the nonlinear characteristics of the power amplifier cause changes in signal amplitude, these changes will be reflected in the amplitude difference sequence. Statistical variance analysis is then performed on the amplitude difference sequence to determine the amplitude distortion for each signal frame.

[0019] Finally, a windowed Fourier transform is performed on the RF feedback signal waveform to extract the ratio of mainlobe energy to sidelobe energy. Based on this ratio, the spectrum leakage factor for each signal frame is calculated. In wireless communications, if the signal's energy distribution on the spectrum is not ideal, for example, spectrum leakage may be caused by an imperfect filter or modulation process, the spectrum leakage factor can quantify this phenomenon. By combining the phase offset, amplitude distortion, and spectrum leakage factor, a feedback error metric is generated that comprehensively reflects the modulation error during the signal frame's transmission.

[0020] In step S130, based on the feedback error index, a time-frequency domain joint analysis is performed on the RF feedback signal waveform of each signal frame, an error feature vector of each signal frame is extracted, and the error feature vector is input into a pre-trained error prediction model for cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing a timing dependency.

[0021] Joint time-frequency analysis is a method that analyzes signals from both the time and frequency domains. Time-domain analysis focuses on how the signal changes over time, while frequency-domain analysis focuses on the energy distribution of the signal in the frequency domain. The error feature vector, extracted through joint time-frequency analysis, describes the error characteristics of the signal frame and includes error information from different perspectives (such as phase, amplitude, and spectrum).

[0022] The pre-trained error prediction model is a model that has been trained using a large amount of data and is capable of predicting signal frame errors. For example, a long short-term memory (LSTM) network can be used. Cross-frame correlation learning leverages the temporal correlation between signal frames, analyzing the error characteristics of historical signal frames to predict future signal frame errors. The dynamically adjusted parameter sequence is generated based on the feedback error indicator and the results of cross-frame correlation learning. It contains a parameter sequence with temporal dependencies and is used to dynamically adjust the transmission parameters of the RF link to optimize signal transmission quality.

[0023] For example, based on the previously calculated feedback error metrics, a joint time-frequency domain analysis can be performed on the RF feedback signal waveform for each signal frame. First, a phase error curve is generated based on the difference between the phase offset and a preset reference phase trajectory. This reference phase trajectory represents the path of the signal phase over time under ideal transmission conditions. Simultaneously, a sliding window normalization process is performed on the amplitude distortion to produce an amplitude error distribution histogram. This amplitude error distribution histogram intuitively displays the distribution of amplitude error across different parts of the signal frame.

[0024] The phase error variation curve, amplitude error distribution histogram, and spectrum leakage factor are concatenated using multi-channel features to generate an initial error feature vector. This initial error feature vector contains information describing the signal error from different perspectives. A convolutional neural network is then used to enhance the local features of the initial error feature vector. For example, in a convolutional neural network, multi-channel convolution is performed on the initial error feature vector. Convolution kernels of different scales are used to perform parallel convolution operations on the phase error variation curve, amplitude error distribution histogram, and spectrum leakage factor in the initial error feature vector, generating a first feature map containing different frequency response characteristics. Convolution kernels of different scales can capture the error characteristics of different frequency components.

[0025] Next, spatial pyramid pooling is performed on the first feature map to extract local maxima at different spatial levels, generating a second feature map with multi-resolution features. This step is similar to observing features in an image at different resolutions and can obtain more comprehensive error feature information. Next, the second feature map is subjected to frequency band separation of high-frequency and low-frequency error components based on a frequency domain decomposition filter bank, generating high-frequency and low-frequency feature sub-maps. This is like separating a sound signal into high-pitched and low-pitched components for more detailed analysis.

[0026] The high-frequency and low-frequency feature submaps are fed into a cross-fusion layer. A coupled feature map of the high-frequency and low-frequency error components is generated through element-by-element multiplication and channel-wise attention weighting. This approach highlights the interrelationships between error components at different frequencies. A depthwise separable convolution is performed on the coupled feature map, sliding the convolution kernel along both the time and frequency axes to integrate multi-scale contextual dependencies and generate a multi-scale fused feature map. Finally, a residual connection is performed between the multi-scale fused feature map and the initial error feature vector. After normalization using an activation function, the output is an error feature vector containing multi-scale error features.

[0027] Finally, the error feature vectors are fed into a pre-trained error prediction model for cross-frame correlation learning. Taking the long short-term memory network as an example, the error feature vectors are fed into the network in the order of the signal frames. The phase offset and amplitude distortion trends of the next signal frame are predicted to generate preliminary adjustment parameters. Simultaneously, a set of error feature vectors of the target RF link within the historical transmission cycle is obtained, and the association weights between the current error feature vector and the historical error feature vectors are calculated using an attention mechanism. For example, if the error characteristics of the current signal frame are similar to those at a specific moment in the past, the error feature vector at that historical moment is assigned a higher association weight. Based on the association weights, the historical adjustment parameters are weighted and fused to generate historically dependent compensation corrections. The preliminary adjustment parameters are then superimposed with the compensation corrections to generate a dynamic adjustment parameter sequence containing temporal dependencies.

[0028] Step S140: constructing a closed-loop feedback control link based on the dynamic adjustment parameter sequence and the original modulation parameters of the plurality of signal frames.

[0029] Among them, the closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and the original modulation parameters. It is a control link that can provide real-time feedback and adjust the signal transmission quality, thereby realizing intelligent and adaptive control of baseband RF signal transmission.

[0030] Based on the above steps, the embodiment of the present application receives a baseband signal dataset containing original modulation parameters and RF feedback signal waveforms, then defines feedback error indicators consisting of phase offset, amplitude distortion, and spectral leakage factors, which can highly accurately quantify the modulation error of the signal frame during transmission. Next, a joint time-frequency domain analysis combined with a pre-trained error prediction model is used to perform cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing temporal dependencies, fully exploiting the characteristics of the signal in different domains. The error prediction model is then used for cross-frame learning to effectively capture the dynamic changes between frames, thereby generating a dynamic adjustment parameter sequence that can reflect the overall change trend of the signal. Finally, a closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and the original modulation parameters, realizing intelligent and adaptive control of baseband RF signal transmission. This closed-loop feedback mechanism can automatically adjust relevant parameters in real time according to the transmission error and dynamic changes of the signal, effectively reducing the modulation error during signal transmission, improving the accuracy and stability of signal transmission, and significantly improving the overall performance and reliability of the baseband RF signal processing system.

[0031] In a possible implementation, step S120 includes: Step S121 : performing quadrature demodulation on the RF feedback signal waveform to generate a baseband in-phase component sequence and a baseband quadrature component sequence.

[0032] In this embodiment, in the radio frequency link of the base station, the radio frequency feedback signal contains complex information. Orthogonal demodulation is a process based on mathematical principles and algorithms. For example, through a specific circuit structure or digital signal processing algorithm, the radio frequency feedback signal can be decomposed into a baseband in-phase component sequence and a baseband orthogonal component sequence. Taking a specific signal frame as an example, assuming that the radio frequency feedback signal is a complex waveform with various changes after transmission, the baseband in-phase component sequence obtained after orthogonal demodulation may be represented as a series of discrete numerical values, which reflect the component characteristics of the radio frequency feedback signal in the in-phase direction; similarly, the baseband orthogonal component sequence also exists in a similar discrete numerical form, which corresponds to the in-phase component sequence in time, and together describes the characteristics of the radio frequency feedback signal at the baseband.

[0033] Step S122 : Calculate the instantaneous phase sequence and instantaneous amplitude sequence of each signal frame according to the baseband in-phase component sequence and the baseband quadrature component sequence.

[0034] When calculating the instantaneous phase sequence, the relationship between the baseband in-phase component and the baseband quadrature component can be exploited. For example, for each sampling point, the instantaneous phase value at that sampling point is obtained by calculating a mathematical operation such as the inverse tangent function. These instantaneous phase values, arranged in chronological order, form the instantaneous phase sequence. For calculating the instantaneous amplitude sequence, the instantaneous amplitude value at each sampling point can be obtained by performing a mathematical operation such as taking the square root of the sum of the squares of the in-phase and quadrature components. These instantaneous amplitude values, arranged in chronological order, form the instantaneous amplitude sequence.

[0035] Step S123 , performing point-by-point difference calculation on the instantaneous phase sequence and a preset reference phase sequence to generate a phase difference sequence, and performing sliding window averaging processing on the phase difference sequence to obtain a phase offset of each signal frame.

[0036] In the ideal transmission model of the base station, the preset reference phase sequence is pre-set and represents the phase change pattern of the signal in the absence of any interference and error. When the instantaneous phase sequence in actual transmission is calculated point by point with the reference phase sequence, a phase difference sequence can be obtained. For example, if at a certain moment, the value of the reference phase sequence is a certain angle, and the value of the actual instantaneous phase sequence at that moment deviates from it, the deviation is recorded in the phase difference sequence. In order to obtain a more stable and representative phase offset, the phase difference sequence is averaged using a sliding window. Assuming that the size of the sliding window is a certain number of sampling points, the phase difference is averaged within the sliding window, and the result is the phase offset of the signal frame. The phase offset reflects the average phase deviation of the signal during the entire signal frame transmission process, which may be caused by unstable oscillator frequency in the RF link or other factors.

[0037] Step S124 , performing normalized ratio calculation on the instantaneous amplitude sequence and a preset reference amplitude sequence to generate an amplitude difference sequence, and performing statistical variance analysis on the amplitude difference sequence to obtain the amplitude distortion of each signal frame.

[0038] In this embodiment, the preset reference amplitude sequence is also based on the ideal transmission setting. In actual transmission, devices such as power amplifiers may have nonlinear characteristics, resulting in changes in the amplitude of the signal. The amplitude difference sequence is obtained by calculating the normalized ratio of the instantaneous amplitude sequence to the reference amplitude sequence. For example, if the instantaneous amplitude of a certain sampling point is 1.2 times or 0.8 times the reference amplitude, the ratio is recorded in the amplitude difference sequence. The amplitude difference sequence is subjected to statistical variance analysis, and the variance can reflect the degree of dispersion of the amplitude difference within the entire signal frame. If the variance is large, it means that the amplitude fluctuates greatly within the signal frame, that is, the amplitude distortion is high; otherwise, the amplitude distortion is low.

[0039] Step S125 , performing a windowed Fourier transform on the RF feedback signal waveform, extracting the ratio of main lobe energy to side lobe energy, and calculating the spectrum leakage factor of each signal frame based on the ratio.

[0040] In this embodiment, the windowed Fourier transform is a mathematical tool that can analyze the characteristics of a signal in the frequency domain. By performing a windowed Fourier transform on the RF feedback signal waveform, the energy distribution of the signal in the frequency domain can be obtained, including the main lobe energy and the side lobe energy. The main lobe energy represents the frequency band where the main energy of the signal is concentrated, and the side lobe energy is the energy distribution outside the main lobe. After calculating the ratio of the main lobe energy to the side lobe energy, the spectrum leakage factor is calculated based on the ratio. For example, if the main lobe energy is very strong and the side lobe energy is relatively weak, then the spectrum leakage factor may be small, indicating that the energy distribution of the signal on the spectrum is more concentrated in the main lobe and the spectrum leakage is small; conversely, if the side lobe energy is not negligible relative to the main lobe energy, the spectrum leakage factor is large, which may be due to an imperfect filter or an imperfection in the modulation process.

[0041] Step S126 , fusing the phase offset, amplitude distortion, and spectrum leakage factor to generate the feedback error indicator.

[0042] For example, the fusion process can be based on a weighted algorithm or mathematical formula. For example, different weights may be assigned to each factor based on its importance in the overall error. The phase offset, amplitude distortion, and spectrum leakage factor are then weighted and summed or otherwise calculated based on the weights to produce a feedback error indicator that comprehensively reflects the modulation error during signal frame transmission. This feedback error indicator serves as an important basis for subsequent analysis and processing.

[0043] In a possible implementation, step S130 includes: Step S131 : generating a phase error variation curve according to the difference between the phase offset and a preset reference phase trajectory, and performing sliding window normalization processing on the amplitude distortion to obtain an amplitude error distribution histogram.

[0044] In this embodiment, the preset reference phase trajectory is the complete change path of the signal phase over time under ideal transmission conditions. Therefore, by calculating the difference between the phase offset and the reference phase trajectory, the change of the phase error over time within the entire signal frame can be obtained. These differences are plotted in chronological order to generate a phase error change curve. For the amplitude distortion, sliding window normalization processing is performed. For example, the amplitude distortion is normalized within a sliding window of a certain length so that the amplitude distortion within different windows is comparable. The frequency of occurrence of each normalized amplitude distortion value is then counted, and an amplitude error distribution histogram is plotted with these frequencies as the vertical coordinates and the amplitude distortion value range as the horizontal coordinate. This amplitude error distribution histogram can intuitively display the distribution of the amplitude error within different amplitude value ranges within the signal frame.

[0045] Step S132 , performing multi-channel feature splicing on the phase error variation curve, the amplitude error distribution histogram, and the spectrum leakage factor to generate an initial error feature vector.

[0046] During this process, these three pieces of information describing the error from different perspectives can be combined in a specific order and format. For example, the values ​​in the phase error variation curve, the frequency values ​​in the amplitude error distribution histogram, and the spectral leakage factor values ​​can be arranged in a certain order into a vector. This vector becomes the initial error feature vector, which contains a preliminary comprehensive description of the signal frame error, but further processing is required to extract more comprehensive error features.

[0047] Step S133 , performing local feature enhancement on the initial error feature vector through a convolutional neural network, extracting the coupling relationship between the high-frequency error component and the low-frequency error component, and generating the error feature vector containing multi-scale error features.

[0048] In a possible implementation, step S133 includes: Step S1331, performing multi-channel convolution processing on the initial error feature vector, using convolution kernels of different scales to perform parallel convolution operations on the phase error change curve, amplitude error distribution histogram and spectrum leakage factor in the initial error feature vector to generate a first feature map containing different frequency response characteristics.

[0049] In this embodiment, convolution kernels of different scales can be understood as filters of different sizes, which can capture the error characteristics of different frequency components. For example, a smaller-scale convolution kernel may pay more attention to the error characteristics of the high-frequency part, and be more sensitive to the rapidly changing part in the phase error change curve, the local fluctuations in the amplitude error distribution histogram, and the high-frequency changes in the spectrum leakage factor; while a larger-scale convolution kernel focuses more on the error characteristics of the low-frequency part, and can capture the changing trend of the error in the signal frame on a longer time scale. Through parallel convolution operations, the first feature map obtained contains a wealth of different frequency response features, which describe the error situation from different frequency perspectives.

[0050] Step S1332: Perform spatial pyramid pooling on the first feature map to extract local area maxima at different spatial levels to generate a second feature map with multi-resolution features.

[0051] In this process, local maxima in different regions are extracted from the first feature map. For example, at a finer spatial level, local extreme values ​​of the error feature within a small range may be captured. These extreme values ​​may correspond to significant errors at certain specific moments or frequency points within the signal frame; while at a coarser spatial level, the overall trend of the error feature within a larger range can be obtained. The local maxima of these different spatial levels are combined to form a second feature map with multi-resolution features. This second feature map can more comprehensively reflect the distribution of error features at different spatial scales.

[0052] Step S1333: performing frequency band separation of high-frequency error components and low-frequency error components on the second feature graph based on a frequency domain decomposition filter bank to generate a high-frequency feature subgraph and a low-frequency feature subgraph.

[0053] The frequency-domain decomposition filter bank separates the high-frequency and low-frequency components of the second feature map based on pre-set frequency limits. The high-frequency feature submap primarily contains high-frequency components of the error signature, such as those that may reflect rapidly changing phase errors within a signal frame, short-term amplitude fluctuations, and high-frequency components of spectral leakage. The low-frequency feature submap primarily contains low-frequency components, such as slowly changing phase errors within a signal frame and long-term amplitude fluctuations.

[0054] In step S1334, the high-frequency feature sub-graph and the low-frequency feature sub-graph are input into a cross-fusion layer, and a coupling feature graph of the high-frequency error component and the low-frequency error component is generated by element-by-element multiplication and channel attention weighting.

[0055] Element-wise multiplication emphasizes the relationship between high-frequency and low-frequency error components, while channel attention weighting adjusts this relationship based on the importance of different channels (i.e., high-frequency and low-frequency channels). For example, if a specific correlation pattern exists between high-frequency and low-frequency error components in a particular signal frame, this method can highlight this correlation pattern, and the generated coupling feature map can more accurately describe this correlation relationship.

[0056] Step S1335 , performing depth-wise separable convolution processing on the coupled feature map, sliding the convolution kernel along the time axis and the frequency axis respectively, integrating multi-scale context dependencies, and generating a multi-scale fusion feature map.

[0057] In this process, depthwise separable convolution can more effectively process data in coupled feature maps. Sliding the convolution kernel along the time and frequency axes can be understood as scanning error features in different time and frequency directions. In this way, contextual dependencies at different scales can be integrated. For example, the relationship between high-frequency and low-frequency error components at different times and at different frequencies can be comprehensively considered. The resulting multi-scale fused feature map can more comprehensively reflect the comprehensive characteristics of the error at multiple scales.

[0058] Step S1336: Perform a residual connection on the multi-scale fusion feature map and the initial error feature vector, and output the error feature vector containing multi-scale error features after normalization through an activation function.

[0059] In this embodiment, residual connections preserve some of the original information in the initial error feature vector while integrating new features from the multi-scale fusion feature map. Activation function normalization adjusts the resulting error feature vector to make it numerically more consistent with subsequent processing requirements. This final error feature vector, which incorporates error features extracted from multiple scales and angles, provides more comprehensive and accurate input information for subsequent operations such as cross-frame correlation learning.

[0060] In a possible implementation, step S130 further includes: Step S135 , inputting the error feature vector into the long short-term memory network in the order of the signal frames, predicting the phase offset trend and amplitude distortion trend of the next signal frame, and generating preliminary adjustment parameters.

[0061] Specifically, each signal frame has a corresponding error feature vector, which contains error information about that signal frame derived from multi-scale and multi-faceted analysis. A long short-term memory (LSTM) network is a neural network architecture specifically designed for processing sequential data. Error feature vectors, arranged sequentially by signal frame, are input into the LSTM, which utilizes its internal memory units and gating mechanisms to process the sequential data. For example, to predict phase offset trends, the LSTM analyzes phase-related information from the error feature vectors of previous signal frames, such as the characteristics of the phase error variation curve. By learning the patterns of this information over the time series, it predicts the likely phase offset trend for the next signal frame. Similarly, to predict amplitude distortion trends, the LSTM integrates amplitude-related information, such as the amplitude error distribution histogram, from the error feature vectors of previous signal frames to derive a prediction for the amplitude distortion trend for the next signal frame. Based on these predictions, preliminary adjustment parameters are generated. These preliminary adjustment parameters are preliminary estimates of the phase and amplitude adjustments for the next signal frame. For example, they may include values ​​such as the direction and approximate amplitude of the phase adjustment, and the amplitude adjustment ratio. These values ​​serve as the basis for subsequent adjustments.

[0062] Step S136: Obtain a set of error feature vectors of the target radio frequency link in a historical transmission period, and calculate the association weight between the current error feature vector and the historical error feature vector through an attention mechanism.

[0063] In this embodiment, a large number of error feature vectors from historical transmission cycles are accumulated during the base station's long-term transmission process. These historical error feature vectors reflect the error conditions of the target RF link under different transmission conditions. The attention mechanism is a mechanism that dynamically assigns weights based on current input. In this scenario, the attention mechanism considers multiple factors when calculating the association weight between the current error feature vector and historical error feature vectors. For example, if certain features in the current error feature vector are very similar to features at a specific historical moment in the historical error feature vector, such as the shape of the phase error variation curve or the pattern of the amplitude error distribution histogram, the error feature vector at that historical moment will be assigned a higher association weight. Specifically, the association weight may be determined by calculating a distance metric (such as Euclidean distance) or a mathematical method such as a correlation coefficient between the two. If the distance is close or the correlation is high, the association weight is large, indicating a strong correlation between the current error feature vector and the historical error feature vector. Conversely, if the distance is far or the correlation is low, the association weight is small.

[0064] Step S137 , weighted fusion is performed on the historical adjustment parameters according to the association weights to generate history-dependent compensation corrections, and the preliminary adjustment parameters are superimposed on the compensation corrections to generate the dynamic adjustment parameter sequence.

[0065] Historical adjustment parameters correspond to historical error eigenvectors and are adjustment parameters for different signal frames during previous transmissions. These historical adjustment parameters are weighted and fused using associated weights. For example, if the historical error eigenvector corresponding to a certain historical adjustment parameter has a higher association weight with the current error eigenvector, then that historical adjustment parameter will be given a higher weight in the weighted fusion process. Assuming that the historical adjustment parameters include different adjustment values ​​for phase and amplitude, the resulting historically dependent compensation correction after weighted fusion can reflect the impact of historical transmission experience on the current signal frame adjustment. The preliminary adjustment parameters are superimposed on the compensation correction amount. The preliminary adjustment parameters are obtained by LSTM prediction based on the error feature vector before the current signal frame, and the compensation correction amount is a weighted fusion result based on historical transmission experience. After the two are superimposed, a dynamic adjustment parameter sequence is generated. The dynamic adjustment parameter sequence combines the prediction results of the current signal frame and historical transmission experience, and contains complete parameter information for phase and amplitude adjustment. These parameters have a time-dependent relationship and can provide an accurate adjustment basis for the subsequent construction of a closed-loop feedback control link, thereby optimizing the base station's RF signal transmission, reducing problems such as phase offset and amplitude distortion, improving the quality and stability of signal transmission, and ensuring that the mobile terminal can accurately receive data sent by the base station.

[0066] In a possible implementation, step S140 includes: Step S141, based on the phase compensation weight and amplitude compensation ratio in the dynamically adjusted parameter sequence, multi-level vector modulation compensation is performed on the phase component and amplitude component in the original modulation parameters to generate a target modulation parameter set including frame-by-frame correction parameters, wherein each correction parameter includes a phase compensation curve and an amplitude compensation coefficient matrix dynamically adjusted according to the error characteristics of adjacent signal frames.

[0067] In detail, consider a signal frame. Within this signal frame, due to certain device characteristics in the RF link or slight external interference, a small phase offset may occur. This offset may be approximately linear, for example, the phase changes at a relatively stable rate over time. At this time, the linear compensation coefficient of the phase offset within the signal frame comes into play. The amount of compensation required can be calculated based on the trend of this offset, thereby making a preliminary correction to the phase component in the original modulation parameters. At the same time, within this signal frame, the amplitude may be distorted due to reasons such as the nonlinearity of the power amplifier. The mean correction amount of the amplitude distortion can adjust the amplitude component in the original modulation parameters based on the average of the amplitude distortion. For example, if the average value of the amplitude within the signal frame deviates from the ideal value, the mean correction amount can bring the average value back to a level close to the ideal value.

[0068] In step S142, the correction parameters in the target modulation parameter set are jointly encoded with the corresponding RF feedback signal waveform to generate a parameter optimization sample set, and the parameter optimization sample set is iteratively optimized for multiple rounds through a feedback control network to output real-time modulation parameters that meet preset error convergence conditions. The feedback control network uses a gradient descent-based weight update mechanism to perform nonlinear correction on the phase compensation curve and the amplitude compensation coefficient matrix.

[0069] Step S143: Load the real-time modulation parameters into the parameter configuration interface of the vector modulator, drive the vector modulator to perform multi-carrier orthogonal modulation and waveform reconstruction on the baseband signal according to the real-time modulation parameters, generate an RF output signal with suppressed phase noise and amplitude distortion, and synchronously feed back the waveform characteristics of the RF output signal to the error compensation module of the target RF link to form a closed-loop feedback control link.

[0070] In a possible implementation, step S141 includes: Step S1411, decomposing the phase compensation weights and amplitude compensation ratios in the dynamic adjustment parameter sequence into primary compensation parameters, intermediate compensation parameters and advanced compensation parameters according to preset compensation levels, wherein the primary compensation parameters include the linear compensation coefficient of the phase offset within the signal frame and the mean correction amount of the amplitude distortion, the intermediate compensation parameters include the gradient suppression coefficient of the phase jump between adjacent signal frames and the boundary constraint value of the amplitude fluctuation range, and the advanced compensation parameters include the phase equalization factor based on the time-frequency distribution matrix and the amplitude distortion compensation weight.

[0071] Step S1412, performing preliminary linear compensation on the phase component and amplitude component in the original modulation parameters of the current signal frame based on the primary compensation parameters to obtain primary correction parameters, wherein the primary correction parameters include the average offset correction value of the phase component and the normalized scaling coefficient of the amplitude component.

[0072] In the base station's signal processing flow, for the phase component, the average offset correction value is calculated based on the linear compensation coefficient of the phase offset within the signal frame. For example, if the linear compensation coefficient indicates that the phase needs to be adjusted by a certain angle, the adjustment value is the average offset correction value, which directly corrects the phase component in the original modulation parameters. For the amplitude component, the normalization scaling factor is obtained based on the mean correction of the amplitude distortion. Assuming that the original amplitude deviates from the normal proportional relationship due to distortion, the normalization scaling factor can be used to adjust the amplitude to a relatively reasonable range, so that the proportional relationship between the amplitude component and other related parameters is more in line with the ideal signal transmission requirements.

[0073] Step S1413: Input the primary correction parameters into the nonlinear correction module corresponding to the intermediate compensation parameters, perform smoothing filtering on the instantaneous jump of the phase component according to the gradient suppression coefficient, and perform truncation correction on the extreme value of the amplitude component in combination with the boundary constraint value to generate the intermediate correction parameters.

[0074] When a signal switches from one frame to the next, a transient phase jump may occur. This jump may be caused by a change in signal type or a sudden change in link status. During signal processing at the base station, the gradient suppression coefficient in the intermediate compensation parameters smooths out these transient phase jumps. For example, if a sudden, large phase jump occurs, the gradient suppression coefficient uses a specific filtering algorithm based on the amplitude and direction of the jump to smooth the change, making the phase change more continuous and gradual, thereby preventing the sudden jump from adversely affecting subsequent signal processing and transmission. Furthermore, various factors can cause the amplitude component to fluctuate significantly, even reaching extreme values ​​outside the normal range. The amplitude fluctuation range boundary constraint in the intermediate compensation parameters comes into play, truncating extreme amplitude values. For example, if the maximum amplitude exceeds the set boundary constraint, it is corrected to the boundary constraint value to prevent excessive amplitude fluctuations from seriously impacting signal quality. This process generates the intermediate correction parameters.

[0075] In step S1414, the intermediate correction parameters and the advanced compensation parameters are input into a time-frequency domain joint optimization module, the frequency domain distribution of the phase component is energy-balanced according to the phase equalization factor, and the time domain fluctuation of the amplitude component is weightedly suppressed by the amplitude distortion compensation weight to generate advanced correction parameters.

[0076] During base station RF signal transmission, the signal phase may have uneven energy distribution in the frequency domain. The phase equalization factor within the advanced compensation parameters addresses this issue within the time-frequency domain joint optimization module. For example, the phase energy at certain frequencies may be too high or too low. The phase equalization factor adjusts the frequency-domain energy distribution of the phase component based on the phase energy at different locations in the frequency domain, making the energy distribution more uniform and improving the signal's frequency-domain characteristics. Furthermore, the amplitude component may fluctuate in the time domain, potentially affecting signal stability. The amplitude distortion compensation weight suppresses these time-domain fluctuations. For example, if the amplitude fluctuates significantly within certain time periods, the amplitude distortion compensation weight uses a specific weighting algorithm based on the degree and characteristics of the fluctuation to suppress the fluctuation, making the amplitude more stable in the time domain. This processing results in the generation of the advanced correction parameters.

[0077] Step S1415, superimpose and integrate the primary correction parameters, intermediate correction parameters and advanced correction parameters in the order of signal frames to generate a target modulation parameter set containing frame-by-frame correction parameters, wherein the correction parameters of each signal frame include the three-level compensation superposition result of the phase component and the multi-level constrained fusion value of the amplitude component.

[0078] During the base station's signal modulation process, for each signal frame, the three-level compensation superposition result of the phase component is obtained by sequentially superimposing the phase average offset correction value in the primary correction parameter, the phase value after smoothing filtering of the intermediate correction parameter, and the phase value after frequency domain energy equalization adjustment of the advanced correction parameter. This superposition result integrates multiple adjustments, including preliminary linear compensation within the signal frame, smoothing between frames, and frequency domain optimization, to more comprehensively correct the phase component. For the amplitude component, the multi-level constraint fusion value is obtained by fusing the normalized scaling coefficient in the primary correction parameter, the amplitude value after extreme value truncation correction of the intermediate correction parameter, and the amplitude value after time domain fluctuation weighted suppression of the advanced correction parameter. This fusion value takes into account the adjustment of the amplitude at different levels, thereby more effectively optimizing the amplitude component. These frame-by-frame correction parameters are combined to form the target modulation parameter set, providing more accurate modulation parameters for subsequent signal processing.

[0079] In a possible implementation, step S142 includes: Step S1421: Divide the parameter optimization sample set into a training set and a validation set, perform feature mapping on the samples in the training set through the fully connected layer of the feedback control network, and output intermediate optimization parameters.

[0080] For example, each sample in the training set contains various previously processed information, such as correction parameters and RF feedback signal waveforms. The fully connected layer performs complex mathematical operations on this information based on pre-set connection weights and neuron structures, mapping the input sample information into intermediate optimization parameters. These parameters transform and refine the original sample information, containing feature information relevant to error optimization.

[0081] Step S1422 , calculating the mean square error between the intermediate optimization parameter and the corresponding RF feedback signal waveform in the validation set, and updating the weight parameters of the feedback control network using a back propagation algorithm based on the mean square error.

[0082] During base station optimization, mean squared error (MSE) is an important metric for measuring the difference between intermediate optimization parameters and the actual RF feedback signal waveform. For example, for each RF feedback signal waveform in the validation set, the mean squared error (MSE) between it and the corresponding intermediate optimization parameter is calculated. This calculation involves mathematical operations such as summing the squared differences at each sampling point. Based on this MSE, the weight parameters of the feedback control network are updated using the backpropagation algorithm. The backpropagation algorithm is an optimization algorithm based on the principle of gradient descent. It adjusts the weight parameters according to a specific step size based on the gradient of the MSE relative to the weight parameters, updating the weight parameters in a direction that reduces the MSE. For example, if a weight parameter has a significant impact on the MSE and the current direction is increasing the MSE, the backpropagation algorithm will decrease the value of that weight parameter, and vice versa. This continuously optimizes the weight parameters of the feedback control network, enabling it to better process samples and reduce error.

[0083] Step S1423: When the decreasing rate of the mean square error in the consecutive preset number of iterations is less than a preset threshold, the iteration is terminated and the real-time modulation parameters are output; otherwise, the number of hidden layer nodes is increased and parameter mapping is performed again.

[0084] During the iterative optimization process of a base station, the preset number of rounds and the preset threshold are important pre-set parameters. If the rate of decrease of the mean squared error is less than the preset threshold during consecutive iterations of the preset number of rounds, this means that the error reduction has been very slow and the optimization process is close to convergence. For example, assuming the preset number of rounds is 10 and the preset threshold is 0.01, if the rate of decrease of the mean squared error is less than 0.01 during 10 consecutive iterations, it is considered that a relatively optimal state has been reached. At this time, the iteration is terminated and the real-time modulation parameters are output. However, if the rate of decrease of the mean squared error has not reached the preset threshold, it means that the feedback control network may not have fully learned the information in the sample and the network complexity needs to be increased. Increasing the number of hidden layer nodes can increase the network's expressive power. Parameter mapping can then be re-performed and iterative optimization repeated until the preset error convergence conditions are met.

[0085] In a possible implementation, step S143 includes: Step S1431: Parse the real-time modulation parameters into multiple subcarrier modulation parameter groups according to the frequency band division rule of orthogonal subcarriers, and perform adaptive mapping of bandwidth and center frequency for each subcarrier modulation parameter group according to a preset subcarrier frequency band allocation strategy.

[0086] For example, in a multi-carrier communication system, different subcarriers have different frequency bands and communication requirements. The real-time modulation parameters contain the modulation information for the entire signal and need to be decomposed into each subcarrier. According to the frequency band division rules of orthogonal subcarriers, the relevant information in the real-time modulation parameters is parsed into multiple subcarrier modulation parameter groups. Each subcarrier modulation parameter group contains specific modulation information for that subcarrier. Then, according to the preset subcarrier frequency band allocation strategy, such as some frequency bands are suitable for high-speed data transmission and some frequency bands are suitable for low-power transmission, the bandwidth and center frequency of each subcarrier modulation parameter group are adaptively mapped. This means that according to the function and requirements of the subcarrier, the appropriate bandwidth and center frequency are set for each subcarrier to ensure that the subcarrier can transmit signals within the optimal frequency band range.

[0087] In step S1432, the orthogonal modulation unit of the vector modulator generates corresponding baseband IQ modulation signals based on the phase compensation curve and amplitude compensation coefficient matrix in each subcarrier modulation parameter group, and orthogonally superimposes the baseband IQ modulation signals of all subcarriers to generate a multi-carrier baseband orthogonal modulation signal.

[0088] In a base station's vector modulator, the quadrature modulation unit is a key component for signal modulation. For each subcarrier modulation parameter set, the phase compensation curve and amplitude compensation coefficient matrix provide key information for adjusting the phase and amplitude of that subcarrier. Based on the phase compensation curve, the vector modulator's quadrature modulation unit adjusts the phase of the baseband signal to generate the I-channel modulated signal; based on the amplitude compensation coefficient matrix, it adjusts the amplitude of the baseband signal to generate the Q-channel modulated signal. For example, if the phase compensation curve for a specific subcarrier indicates a gradual increase in phase over a certain period of time, the phase will be adjusted according to this phase compensation curve when generating the I-channel modulated signal. Similarly, if the amplitude compensation coefficient matrix specifies an amplitude scaling ratio, the amplitude will be adjusted accordingly when generating the Q-channel modulated signal. Then, the baseband IQ modulation signals of all subcarriers are orthogonally superimposed, that is, the I and Q modulation signals of each subcarrier are superimposed in an orthogonal manner to generate a multi-carrier baseband orthogonal modulation signal. The multi-carrier baseband orthogonal modulation signal contains the information of multiple subcarriers, and the validity and anti-interference performance of the signal are guaranteed by orthogonal superposition.

[0089] Step S1433: Perform digital-to-analog conversion on the multi-carrier baseband quadrature modulated signal to generate an analog baseband waveform, and up-convert the analog baseband waveform to a preset RF carrier frequency through a mixer to generate an initial RF signal waveform.

[0090] During the base station's signal conversion process, the multi-carrier baseband quadrature modulated signal is in digital form and needs to be converted to an analog signal for RF transmission. Digital-to-analog conversion converts the digital multi-carrier baseband quadrature modulated signal into an analog baseband waveform. This process involves algorithms and circuits that perform digital-to-analog conversion on the digital signal samples. The resulting analog baseband waveform is a baseband analog signal with a relatively low frequency range. To transmit the signal into the RF band, a mixer up-converts the analog baseband waveform to a preset RF carrier frequency. For example, if the preset RF carrier frequency is 2.4 GHz, the mixer mixes the analog baseband waveform with a 2.4 GHz carrier signal, raising the frequency of the analog baseband waveform to 2.4 GHz and generating the initial RF signal waveform. This initial RF signal waveform has frequency characteristics suitable for RF transmission.

[0091] Step S1434: Perform predistortion correction processing on the initial RF signal waveform according to the nonlinear predistortion compensation coefficient in the real-time modulation parameters, suppress the out-of-band radiation component of the initial RF signal waveform, and generate a RF output signal after predistortion correction.

[0092] During base station RF signal processing, the initial RF signal waveform may generate out-of-band radiation components due to the nonlinear characteristics of components such as power amplifiers. These out-of-band radiation components can interfere with signals in other frequency bands, impacting the performance of the entire communication system. The nonlinear predistortion compensation coefficient within the real-time modulation parameters is designed to compensate for this nonlinearity. Using a specific predistortion correction algorithm and circuit structure, the initial RF signal waveform is predistorted according to the nonlinear predistortion compensation coefficient. For example, if the nonlinear predistortion compensation coefficient indicates that a certain frequency band of the signal needs to be attenuated to suppress out-of-band radiation, the initial RF signal waveform can be processed accordingly according to the nonlinear predistortion compensation coefficient in that frequency band. This suppresses the out-of-band radiation components of the initial RF signal waveform, generating a predistorted RF output signal with improved spectral characteristics and reduced interference with signals in other frequency bands.

[0093] And, in step S1435, the time domain waveform characteristics and frequency domain energy distribution characteristics of the RF output signal are synchronously fed back to the error compensation module of the target RF link, and error correlation analysis is performed with the RF feedback signal waveform of the baseband signal data set of the next transmission cycle to update the real-time modulation parameters.

[0094] In the base station's closed-loop feedback control mechanism, the time-domain waveform characteristics and frequency-domain energy distribution characteristics of the RF output signal are synchronously fed back to the error compensation module of the target RF link. The module then performs error correlation analysis with the RF feedback signal waveform of the baseband signal dataset for the next transmission cycle. For example, information such as the temporal variation of the signal amplitude and phase fluctuations in the time-domain waveform characteristics, as well as the energy levels in different frequency bands in the frequency-domain energy distribution characteristics, are compared and analyzed with the corresponding information in the RF feedback signal waveform for the next transmission cycle. This error correlation analysis reveals signal variation trends and error sources during transmission. Based on these analysis results, the real-time modulation parameters are updated to enable more accurate signal modulation and optimization in the next transmission cycle, thereby continuously improving signal transmission quality and reducing issues such as phase noise and amplitude distortion.

[0095] In one possible implementation, the implementation process of the closed-loop feedback control link includes: Step S210 , collecting waveform sampling data of the RF output signal in real time, and calculating the instantaneous error between the waveform sampling data and the ideal reference waveform.

[0096] The RF output signal carries the information that the base station wants to transmit to the mobile terminal. Its waveform sampling data includes key characteristics such as the signal's amplitude and phase at different moments. To measure the accuracy of signal transmission, it is necessary to calculate the instantaneous error between these waveform sampling data and the ideal reference waveform. The ideal reference waveform is based on theoretically perfect signal transmission and represents the waveform that the RF output signal should have in the absence of any interference or distortion. When calculating the instantaneous error, for example, at each sampling moment, a specific mathematical formula is used to calculate the amplitude and phase difference between the waveform sampling data and the ideal reference waveform at that moment. These differences are combined to form the instantaneous error. This instantaneous error reflects the degree to which the RF output signal deviates from its ideal state at each moment.

[0097] Step S220: When the instantaneous error exceeds the dynamic threshold, a fast compensation mechanism is triggered to extract the correction parameters of the latest N frames from the target modulation parameter set, perform weighted average calculation, and generate emergency compensation parameters.

[0098] The dynamic threshold is a limit value dynamically set based on system performance requirements and actual operating conditions. In a base station's signal transmission environment, instantaneous errors may fluctuate within a certain range due to various factors, such as environmental interference and equipment aging. When the instantaneous error exceeds the dynamic threshold, the signal deviation has reached a level requiring immediate adjustment. At this point, the rapid compensation mechanism is triggered, extracting correction parameters from the target modulation parameter set for the most recent N frames and performing a weighted average calculation to generate emergency compensation parameters. The target modulation parameter set contains frame-by-frame correction parameters obtained after multi-level vector modulation compensation for each signal frame. These correction parameters are derived based on previous signal analysis and adjustments. The correction parameters for the most recent N frames are extracted. For example, if N is 5, the correction parameters for the five most recent frames are selected. The weighted average calculation of these correction parameters may assign different weights to each frame based on its importance or relevance to the current situation. For example, correction parameters for frames closer to the current time may be assigned higher weights because they better reflect current signal trends. The emergency compensation parameters derived from this weighted average calculation can quickly compensate for current signal deviations.

[0099] Step S230: insert the emergency compensation parameter into the head of the queue of the real-time modulation parameter, and preferentially load it into the vector modulator for waveform reconstruction.

[0100] In the base station's signal processing flow, real-time modulation parameters are a parameter queue used to control the vector modulator's modulation of the baseband signal. Inserting the emergency compensation parameters at the head of the queue ensures that the vector modulator receives them first and uses them first for waveform reconstruction. After receiving the emergency compensation parameters, the vector modulator quickly adjusts the baseband signal to be transmitted based on the phase and amplitude adjustment information in the emergency compensation parameters. For example, if the emergency compensation parameters include a phase correction value and an amplitude adjustment ratio, the vector modulator will immediately make corresponding changes to the phase and amplitude of the baseband signal, thereby reconstructing an RF output signal waveform that is closer to the ideal state, thereby reducing the impact of instantaneous errors on signal transmission.

[0101] In one possible implementation, the method further includes: Step S310: Spectrum monitoring is performed on the radio frequency output signal to extract energy values ​​and harmonic distribution characteristics of out-of-band interference frequencies.

[0102] When a base station transmits data to a mobile terminal, spectrum monitoring of the RF output signal is performed to ensure signal quality in the frequency domain. This spectrum monitoring process extracts the energy and harmonic distribution characteristics of out-of-band interference frequencies in the RF output signal. The spectrum of the RF output signal contains the energy distribution of the signal at different frequencies. Out-of-band interference frequencies are frequencies outside the normal signal frequency band but exhibit abnormal energy due to various reasons (such as harmonics generated by nonlinear devices or external interference). Spectrum monitoring equipment or algorithms can accurately detect the energy levels of these out-of-band interference frequencies. For example, if the energy level detected at a specific frequency is higher than the normal background noise level, the frequency is identified as an out-of-band interference frequency and its energy value is recorded. Harmonics are also extracted. Harmonics are frequency components that are integer multiples of the original signal frequency, generated by the signal passing through nonlinear devices (such as power amplifiers). By analyzing the distribution of these harmonics, including the energy level and phase relationship of different harmonics, a comprehensive understanding of the frequency domain characteristics of the RF output signal can be obtained.

[0103] Step S320: If it is detected that the energy value of any frequency point exceeds a threshold value, a parameter configuration instruction of a notch filter is generated according to the harmonic distribution characteristics.

[0104] The threshold is a pre-set energy limit used to determine whether out-of-band interference is severe enough to require action. When the energy value of an out-of-band interference frequency exceeds the threshold, it indicates that the interference at that frequency may have a significant impact on signal transmission. At this point, the parameter configuration instructions for the notch filter are generated based on the harmonic distribution characteristics. For example, if the harmonic distribution characteristics show that the harmonic energy at and near a certain interference frequency is strong and the phase relationship is complex, these factors need to be considered when generating the parameter configuration instructions for the notch filter. The center frequency and frequency bandwidth of the interference frequency whose energy value exceeds the threshold in the harmonic distribution characteristics are determined and used as the initial stopband parameters of the notch filter. The center frequency is the core frequency of the interference frequency, and the frequency bandwidth represents the affected frequency range of the interference frequency and its surroundings.

[0105] In a possible implementation, step S320 includes: Step S321 , determining the center frequency and frequency bandwidth of the interference frequency point whose energy value in the harmonic distribution feature exceeds the threshold value, and using the center frequency and frequency bandwidth as initial stopband parameters of the notch filter.

[0106] Spectrum monitoring equipment or algorithms analyze the spectrum of the RF output signal in detail. Each frequency point has a corresponding energy measurement result. When the energy value of a frequency point exceeds a preset threshold, it is identified as an interference frequency. For example, suppose communications are occurring near the 2.4 GHz band, and the energy value at 2.45 GHz is detected to be abnormally high, exceeding the threshold. In this case, 2.45 GHz is the center frequency of the interference frequency point. Furthermore, the frequency bandwidth of the interference frequency point must be determined. The frequency bandwidth reflects the frequency range affected by the interference frequency point and its surrounding areas. This can be determined by analyzing the energy distribution of the interference frequency point along the frequency axis. Assuming that significant interference occurs in the range from 2.43 GHz to 2.47 GHz, the frequency bandwidth of the interference frequency point is 0.04 GHz. Using a center frequency of 2.45 GHz and a frequency bandwidth of 0.04 GHz as the initial stopband parameters for the notch filter means that the notch filter is initially configured to suppress interference within this frequency range.

[0107] Step S322 , calculating the stopband attenuation slope and phase compensation factor of the notch filter according to the amplitude ratio and phase difference between the main harmonic and the subharmonic in the harmonic distribution characteristics, and generating a stopband suppression enhancement parameter.

[0108] In the harmonic distribution of an RF output signal, main harmonics and subharmonics are of great significance. For example, assume the main harmonic has a frequency of f1 and an amplitude of A1, while the subharmonic has a frequency of f2 and an amplitude of A2. There is a phase difference θ between the main and subharmonics. Using a specific mathematical algorithm, the notch filter's stopband attenuation slope and phase compensation factor are calculated based on the amplitude ratio A2 / A1 of the main and subharmonics and the phase difference θ. If the value of A2 / A1 is large, it indicates that the subharmonic energy is stronger than the main harmonic, and a steeper stopband attenuation slope may be required to suppress the subharmonics. The specific calculation may involve complex formulas based on RF signal processing theory and filter design principles. Furthermore, the phase difference θ also affects the calculation of the phase compensation factor. During the filtering process, different phase relationships may cause signal distortion or changes, requiring adjustment using the phase compensation factor. The stopband attenuation slope and phase compensation factor obtained through this calculation constitute the stopband suppression enhancement parameter, which can further optimize the notch filter's interference suppression effectiveness.

[0109] Step S323 : Based on the initial stopband parameters and the stopband suppression enhancement parameters, dynamically offset compensation is performed on the cutoff frequency boundary of the notch filter to generate a cutoff frequency range after the transition band is optimized.

[0110] In notch filter design, the initial stopband parameters (center frequency and bandwidth) and the stopband rejection enhancement parameters (stopband attenuation slope and phase compensation factor) jointly influence the determination of the cutoff frequency boundary. For example, based on the initial center frequency of 2.45 GHz and bandwidth of 0.04 GHz, as well as the calculated stopband attenuation slope and phase compensation factor, the notch filter's cutoff frequency boundary is adjusted. If the stopband attenuation slope is large, the cutoff frequency boundary may need to be offset toward the interference frequency to ensure more effective interference suppression within the stopband. Furthermore, to optimize the characteristics of the transition band (the frequency band from passband to stopband), this offset is dynamic and takes multiple factors into consideration. For example, to avoid excessive signal fluctuations or distortion within the transition band, the cutoff frequency boundary offset is fine-tuned based on factors such as the phase compensation factor in the stopband rejection enhancement parameters. This dynamic offset compensation ultimately generates an optimized cutoff frequency range within the transition band, which better meets the notch filter's requirements for interference suppression and signal quality.

[0111] Step S324 : determining the passband ripple coefficient and filter order of the notch filter according to the cutoff frequency range and the preset maximum group delay constraint.

[0112] In base station signal transmission, after the cutoff frequency range is determined, the passband ripple coefficient and filter order must be determined based on the preset maximum group delay constraint. The maximum group delay constraint ensures that the signal does not incur excessive delay when passing through the notch filter, as excessive delay may affect signal synchronization and transmission quality. For example, assume the cutoff frequency range is 2.43 GHz - 2.47 GHz, and the preset maximum group delay is T. The passband ripple coefficient reflects the fluctuation of the signal amplitude within the notch filter's passband (i.e., the frequency band for normal signal transmission). Based on the cutoff frequency range and the maximum group delay constraint, the passband ripple coefficient is determined using a specific filter design algorithm. If the cutoff frequency range is wide, the passband ripple coefficient may need to be appropriately reduced to minimize the impact on signals within the passband. The filter order is also related to the filter complexity and filtering performance. A higher filter order generally provides better filtering results, but also increases computational complexity and potential signal delay. While satisfying the maximum group delay constraint, the appropriate filter order is determined based on factors such as the cutoff frequency range and the passband ripple coefficient. For example, if a more precise filtering effect is required and the cutoff frequency range allows, a higher filter order might be chosen, but ensuring that this does not cause the group delay to exceed a preset value.

[0113] Step S325 , iteratively adjusting the combination of the passband ripple coefficient and the filter order so that the out-of-band suppression depth of the notch filter reaches the energy suppression requirement of the subharmonics in the harmonic distribution characteristics.

[0114] When determining the parameters of a notch filter, it may not be possible to determine the passband ripple coefficient and filter order that meet the subharmonic energy suppression requirements all at once. For example, an initial passband ripple coefficient and filter order are set, and then the out-of-band suppression depth of the notch filter is calculated. Assuming the initial passband ripple coefficient is Rp1 and the filter order is N1, the calculated out-of-band suppression depth is D1. If D1 is less than the subharmonic energy suppression requirement specified in the harmonic distribution characteristics, the passband ripple coefficient and filter order need to be adjusted based on the calculated results. For example, increase the filter order to N2, adjust the passband ripple coefficient to Rp2, and recalculate the out-of-band suppression depth to obtain D2. If D2 still does not meet the requirements, continue adjusting and iterating until the out-of-band suppression depth meets the subharmonic energy suppression requirement specified in the harmonic distribution characteristics. This process requires precise calculations and a deep understanding of filter performance to ensure that the notch filter can effectively suppress out-of-band interference.

[0115] Step S326 , encapsulating the finally determined cutoff frequency range, stopband attenuation slope, phase compensation factor, passband ripple coefficient, and filter order into the parameter configuration instruction.

[0116] Specifically, the parameter configuration instruction contains all the parameter information required for effective filtering by the notch filter. In the base station's signal processing system, these parameters are packaged into a single instruction for accurate transmission to the vector modulator's post-filter unit. For example, the parameter configuration instruction may be packaged in a specific data format, with each parameter encoded in a specified order and format, ensuring that the post-filter unit can correctly parse and configure the notch filter based on these parameters.

[0117] Step S330: Send the parameter configuration instruction to the post-filter unit of the vector modulator, and superimpose a reverse compensation component on the modulation parameters of the next signal frame to offset out-of-band interference.

[0118] In this embodiment, after receiving the parameter configuration instruction, the post-filter unit will configure the notch filter according to these parameters, so that it can effectively suppress out-of-band interference. At the same time, the reverse compensation component is superimposed on the modulation parameters of the next signal frame to offset the out-of-band interference. During the signal modulation process of the base station, the modulation parameters of the next signal frame will be adjusted according to the filtering effect of the notch filter. The purpose of superimposing the reverse compensation component is to further compensate for the impact that may be caused to the signal due to out-of-band interference. For example, if out-of-band interference causes the amplitude or phase of the signal at a certain frequency to change, by superimposing the reverse compensation component, the amplitude and phase of the signal at this frequency can be adjusted back to a state close to normal, thereby improving the quality of signal transmission and reducing the impact of out-of-band interference on signal transmission.

[0119] In one possible implementation, the method further includes: Step S410: Calculate the channel quality index of the target radio frequency link according to the error statistics accumulated in the closed-loop feedback control link.

[0120] During operation, the closed-loop feedback control link continuously collects and accumulates various error information, such as statistics of error indicators such as phase offset, amplitude distortion, and spectral leakage. These error statistics reflect the accuracy and stability of signal transmission within the RF link. These error statistics are combined using a specific algorithm to calculate the channel quality index (CQI). For example, different weights may be assigned based on the importance of different error indicators to signal quality. The CQI is then calculated by weighted summation or other mathematical operations based on the weighted sums or other mathematical operations to comprehensively reflect the communication quality of the target RF link.

[0121] Step S420: When the channel quality index is lower than a first threshold, the modulation order is reduced and the error correction coding redundancy is increased based on a first optimization strategy; and when the channel quality index is higher than a second threshold, the modulation order is increased based on a second optimization strategy and compressed sensing technology is used to optimize spectrum efficiency.

[0122] When a base station transmits data to a mobile terminal, the first threshold is a pre-set limit used to determine whether channel quality is poor. When the channel quality index falls below the first threshold, the channel quality has deteriorated to the point where measures are needed to improve signal transmission reliability. In this case, the modulation order is reduced based on the first optimization strategy. The modulation order is related to the signal transmission rate and interference immunity. A higher modulation order achieves a higher data transmission rate, but is more susceptible to interference when channel quality is poor. Reducing the modulation order means sacrificing a certain data transmission rate in exchange for stronger interference immunity. For example, if the original modulation scheme used was 64QAM (64-bit Quadrature Amplitude Modulation), the modulation order might be reduced to 16QAM. At the same time, the error correction code redundancy is increased. Error correction code is coded information added to detect and correct errors during signal transmission. Increasing redundancy means adding more error correction code information to the signal. This allows the mobile terminal to use this error correction code information to correct errors in the received signal even in the case of poor channel quality. For example, increasing the original error correction coding rate from 1 / 2 to 3 / 4 increases the redundant information in the signal.

[0123] The second threshold is also a pre-set limit used to determine whether the channel quality is good. When the channel quality index is higher than the second threshold, it indicates that the channel quality is good, and this advantage can be used to improve the data transmission rate and spectrum efficiency. Based on the second optimization strategy, the modulation order is increased, for example, from 16QAM to 64QAM or even higher modulation orders, thereby improving the data transmission rate. At the same time, compressed sensing technology is used to optimize spectrum efficiency. In the radio frequency transmission of the base station, spectrum resources are limited. Compressed sensing technology can improve spectrum efficiency without increasing bandwidth by utilizing the sparsity of the signal. For example, the baseband signal is processed through a compressed sensing algorithm, making the signal distribution on the spectrum more compact, thereby enabling more data to be transmitted within the same spectrum range.

[0124] Step S430: Encapsulate the adjusted modulation strategy parameters into control instructions, and update the configuration parameters of the vector modulator in real time through the closed-loop feedback control link.

[0125] During the base station's signal processing, adjusted modulation strategy parameters (such as the modulation order and error correction code redundancy) must be accurately transmitted to the vector modulator (VMM) to update its configuration. These parameters are encapsulated into control instructions using a specific data format and protocol. These instructions contain all the information necessary for the VMM to correctly configure itself. Through a closed-loop feedback control link, these control instructions are sent in real time to the VMM's configuration interface. Upon receiving these control instructions, the VMM updates its configuration based on the parameters contained in the instructions. For example, if the control instruction includes information about increasing the modulation order from 16QAM to 64QAM, the VMM adjusts its internal modulation module accordingly to modulate the baseband signal according to the new modulation order. Similarly, if the instruction includes information about adjusting the error correction code redundancy, the error correction code module is also updated accordingly. By updating the VMM's configuration parameters in real time, the signal transmission strategy can be adjusted in response to changes in channel quality, improving the performance and efficiency of the entire wireless communication system.

[0126] Figure 2 The diagram shows exemplary hardware and software components of a communication service system 100 that can implement the concepts of the present invention according to some embodiments of the present invention. For example, the processor 120 can be used in the communication service system 100 to perform the functions of the present invention.

[0127] The communication service system 100 can be a general-purpose server or a special-purpose server, both of which can be used to implement the baseband RF signal data feedback processing method based on vector modulation of the present invention. Although only one server is shown in the present invention, for convenience, the functions described in the present invention can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0128] For example, the communication service system 100 may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. For example, the communication service system 100 may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present invention may be implemented based on these program instructions. The communication service system 100 also includes an input / output (I / O) interface 150 between the computer and other input / output devices.

[0129] For ease of explanation, only one processor is described in the communication service system 100. However, it should be noted that the communication service system 100 in the present invention may also include multiple processors, so the steps performed by one processor described in the present invention may also be performed jointly or individually by multiple processors. For example, if the processor of the communication service system 100 performs step A and step B, it should be understood that step A and step B may also be performed jointly by two different processors or individually in one processor. For example, the first processor performs step A and the second processor performs step B, or the first processor and the second processor perform steps A and B together.

[0130] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer-executable instructions are preset. When the processor executes the computer-executable instructions, the baseband RF signal data feedback processing method based on vector modulation as described above is implemented.

[0131] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, multiple features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. A baseband radio frequency signal data feedback processing method based on vector modulation, characterized in that: The method comprises: receiving a baseband signal dataset transmitted by a target radio frequency link within a preset frequency band, wherein the baseband signal dataset includes original modulation parameters of a plurality of signal frames arranged in a time series, and a radio frequency feedback signal waveform corresponding to each signal frame; Calculating a corresponding feedback error index based on the RF feedback signal waveform, wherein the feedback error index is composed of a phase offset, an amplitude distortion, and a spectrum leakage factor, and is used to quantify the modulation error of the signal frame during transmission; Based on the feedback error indicator, a time-frequency domain joint analysis is performed on the RF feedback signal waveform of each signal frame to extract the error feature vector of each signal frame, and the error feature vector is input into a pre-trained error prediction model for cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing a timing dependency relationship; A closed-loop feedback control link is constructed based on the dynamic adjustment parameter sequence and the original modulation parameters of the multiple signal frames.

2. The baseband RF signal data feedback processing method based on vector modulation according to claim 1, characterized in that: The calculating a corresponding feedback error index according to the RF feedback signal waveform includes: Performing quadrature demodulation on the RF feedback signal waveform to generate a baseband in-phase component sequence and a baseband quadrature component sequence; Calculating the instantaneous phase sequence and instantaneous amplitude sequence of each signal frame according to the baseband in-phase component sequence and the baseband quadrature component sequence; Performing point-by-point difference calculation on the instantaneous phase sequence and a preset reference phase sequence to generate a phase difference sequence, and performing sliding window averaging processing on the phase difference sequence to obtain a phase offset for each signal frame; Performing normalized ratio calculation on the instantaneous amplitude sequence and a preset reference amplitude sequence to generate an amplitude difference sequence, and performing statistical variance analysis on the amplitude difference sequence to obtain an amplitude distortion of each signal frame; Performing a windowed Fourier transform on the RF feedback signal waveform, extracting a ratio of mainlobe energy to sidelobe energy, and calculating a spectrum leakage factor of each signal frame based on the ratio; The phase offset, amplitude distortion and spectrum leakage factor are combined to generate the feedback error index; The step of performing a time-frequency domain joint analysis on the RF feedback signal waveform of each signal frame according to the feedback error indicator to extract an error feature vector of each signal frame includes: Generating a phase error variation curve according to the difference between the phase offset and a preset reference phase trajectory, and performing sliding window normalization processing on the amplitude distortion to obtain an amplitude error distribution histogram; Performing multi-channel feature splicing on the phase error variation curve, the amplitude error distribution histogram, and the spectrum leakage factor to generate an initial error feature vector; Performing local feature enhancement on the initial error feature vector through a convolutional neural network, extracting the coupling relationship between the high-frequency error component and the low-frequency error component, and generating the error feature vector containing multi-scale error features; The step of performing local feature enhancement on the initial error feature vector by using a convolutional neural network, extracting the coupling relationship between the high-frequency error component and the low-frequency error component, and generating the error feature vector containing multi-scale error features includes: Performing multi-channel convolution processing on the initial error feature vector, using convolution kernels of different scales to perform parallel convolution operations on the phase error change curve, the amplitude error distribution histogram, and the spectrum leakage factor in the initial error feature vector to generate a first feature map containing different frequency response characteristics; Performing spatial pyramid pooling on the first feature map to extract local area maxima at different spatial levels to generate a second feature map with multi-resolution features; Performing frequency band separation of high-frequency error components and low-frequency error components on the second feature graph based on a frequency domain decomposition filter bank to generate a high-frequency feature subgraph and a low-frequency feature subgraph; Input the high-frequency feature subgraph and the low-frequency feature subgraph into the cross fusion layer, and generate a coupled feature graph of the high-frequency error component and the low-frequency error component by element-by-element multiplication and channel attention weighting; Performing depth-separable convolution processing on the coupled feature map, sliding the convolution kernel along the time axis and the frequency axis respectively, integrating multi-scale context dependencies, and generating a multi-scale fusion feature map; The multi-scale fusion feature map is residually connected with the initial error feature vector, and the error feature vector containing multi-scale error features is output after normalization through an activation function.

3. The baseband RF signal data feedback processing method based on vector modulation according to claim 1, characterized in that: The step of inputting the error feature vector into a pre-trained error prediction model for cross-frame correlation learning to generate a dynamic adjustment parameter sequence containing a temporal dependency relationship includes: Inputting the error feature vector into the long short-term memory network in the order of signal frames, predicting the phase offset trend and amplitude distortion trend of the next signal frame, and generating preliminary adjustment parameters; Obtaining a set of error feature vectors of the target radio frequency link in a historical transmission period, and calculating an association weight between a current error feature vector and a historical error feature vector through an attention mechanism; The historical adjustment parameters are weightedly integrated according to the associated weights to generate a history-dependent compensation correction amount, and the preliminary adjustment parameters are superimposed on the compensation correction amount to generate the dynamic adjustment parameter sequence.

4. The baseband RF signal data feedback processing method based on vector modulation according to claim 1, characterized in that: The step of constructing a closed-loop feedback control link based on the dynamic adjustment parameter sequence and the original modulation parameters of the plurality of signal frames comprises: Based on the phase compensation weights and amplitude compensation ratios in the dynamically adjusted parameter sequence, multi-level vector modulation compensation is performed on the phase components and amplitude components in the original modulation parameters to generate a target modulation parameter set including frame-by-frame correction parameters, wherein each correction parameter includes a phase compensation curve and an amplitude compensation coefficient matrix dynamically adjusted according to error characteristics of adjacent signal frames; The correction parameters in the target modulation parameter set are jointly encoded with the corresponding RF feedback signal waveform to generate a parameter optimization sample set, and the parameter optimization sample set is iteratively optimized through a feedback control network for multiple rounds to output real-time modulation parameters that meet preset error convergence conditions, wherein the feedback control network uses a gradient descent-based weight update mechanism to perform nonlinear correction on the phase compensation curve and the amplitude compensation coefficient matrix; The real-time modulation parameters are loaded into the parameter configuration interface of the vector modulator, and the vector modulator is driven to perform multi-carrier orthogonal modulation and waveform reconstruction on the baseband signal according to the real-time modulation parameters to generate an RF output signal with suppressed phase noise and amplitude distortion, and the waveform characteristics of the RF output signal are synchronously fed back to the error compensation module of the target RF link to form a closed-loop feedback control link.

5. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 4, characterized in that: The method of performing multi-level vector modulation compensation on the phase component and the amplitude component in the original modulation parameters based on the phase compensation weight and the amplitude compensation ratio in the dynamic adjustment parameter sequence to generate a target modulation parameter set including frame-by-frame correction parameters includes: Decomposing the phase compensation weights and amplitude compensation ratios in the dynamically adjusted parameter sequence into primary compensation parameters, intermediate compensation parameters, and advanced compensation parameters according to preset compensation levels, wherein the primary compensation parameters include a linear compensation coefficient for the phase offset within a signal frame and a mean correction for the amplitude distortion; the intermediate compensation parameters include a gradient suppression coefficient for the phase jump between adjacent signal frames and a boundary constraint value for the amplitude fluctuation range; and the advanced compensation parameters include a phase equalization factor based on a time-frequency distribution matrix and an amplitude distortion compensation weight; Performing preliminary linear compensation on the phase component and the amplitude component of the original modulation parameters of the current signal frame based on the primary compensation parameters to obtain primary correction parameters, wherein the primary correction parameters include an average offset correction value of the phase component and a normalized scaling coefficient of the amplitude component; Inputting the primary correction parameter into the nonlinear correction module corresponding to the intermediate compensation parameter, performing smoothing filtering on the instantaneous jump of the phase component according to the gradient suppression coefficient, and performing truncation correction on the extreme value of the amplitude component in combination with the boundary constraint value to generate the intermediate correction parameter; Inputting the intermediate correction parameters and the advanced compensation parameters into a time-frequency domain joint optimization module, performing energy balance adjustment on the frequency domain distribution of the phase component according to the phase balance factor, and weightedly suppressing the time domain fluctuation of the amplitude component using the amplitude distortion compensation weight, to generate advanced correction parameters; The primary correction parameters, intermediate correction parameters and advanced correction parameters are superimposed and integrated in the order of signal frames to generate a target modulation parameter set containing frame-by-frame correction parameters, wherein the correction parameters of each signal frame include the three-level compensation superposition result of the phase component and the multi-level constrained fusion value of the amplitude component.

6. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 4, characterized in that: The method of performing multiple rounds of iterative optimization on the parameter optimization sample set through the feedback control network to output real-time modulation parameters that meet preset error convergence conditions includes: Dividing the parameter optimization sample set into a training set and a validation set, performing feature mapping on the samples in the training set through the fully connected layer of the feedback control network, and outputting intermediate optimization parameters; Calculating the mean square error between the intermediate optimization parameter and the corresponding RF feedback signal waveform in the validation set, and updating the weight parameters of the feedback control network using a back propagation algorithm based on the mean square error; When the decreasing rate of the mean square error in the iteration of the preset number of consecutive rounds is less than the preset threshold, the iteration is terminated and the real-time modulation parameter is output; otherwise, the number of hidden layer nodes is increased and the parameter mapping is re-performed.

7. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 4, characterized in that: The method of loading the real-time modulation parameters into a parameter configuration interface of a vector modulator, driving the vector modulator to perform multi-carrier orthogonal modulation and waveform reconstruction on a baseband signal according to the real-time modulation parameters, and generating a radio frequency output signal with suppressed phase noise and amplitude distortion, includes: Parsing the real-time modulation parameters into multiple subcarrier modulation parameter groups according to the frequency band division rule of orthogonal subcarriers, and performing adaptive mapping of bandwidth and center frequency for each subcarrier modulation parameter group according to a preset subcarrier frequency band allocation strategy; The orthogonal modulation unit of the vector modulator generates corresponding baseband IQ modulation signals based on the phase compensation curve and amplitude compensation coefficient matrix in each subcarrier modulation parameter group, and orthogonally superimposes the baseband IQ modulation signals of all subcarriers to generate a multi-carrier baseband orthogonal modulation signal; Performing digital-to-analog conversion on the multi-carrier baseband quadrature modulated signal to generate an analog baseband waveform, and up-converting the analog baseband waveform to a preset radio frequency carrier frequency through a mixer to generate an initial radio frequency signal waveform; performing predistortion correction processing on the initial RF signal waveform according to the nonlinear predistortion compensation coefficient in the real-time modulation parameter, suppressing the out-of-band radiation component of the initial RF signal waveform, and generating a RF output signal after predistortion correction; In addition, the time domain waveform characteristics and frequency domain energy distribution characteristics of the RF output signal are synchronously fed back to the error compensation module of the target RF link, and error correlation analysis is performed with the RF feedback signal waveform of the baseband signal data set of the next transmission cycle to update the real-time modulation parameters.

8. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 4, characterized in that: The implementation process of the closed-loop feedback control link includes: Collecting waveform sampling data of the radio frequency output signal in real time, and calculating the instantaneous error between the waveform sampling data and the ideal reference waveform; When the instantaneous error exceeds a dynamic threshold, a fast compensation mechanism is triggered to extract the correction parameters of the latest N frames from the target modulation parameter set, perform weighted average calculation, and generate emergency compensation parameters; The emergency compensation parameters are inserted into the head of the queue of the real-time modulation parameters and are preferentially loaded into the vector modulator for waveform reconstruction.

9. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 4, characterized in that: The method further comprises: Performing spectrum monitoring on the radio frequency output signal to extract the energy value and harmonic distribution characteristics of the out-of-band interference frequency point; If it is detected that the energy value of any frequency point exceeds a threshold value, a parameter configuration instruction of a notch filter is generated according to the harmonic distribution characteristics; Sending the parameter configuration instruction to the post-filter unit of the vector modulator, and superimposing a reverse compensation component in the modulation parameters of the next signal frame to offset out-of-band interference; The parameter configuration instruction for generating a notch filter according to the harmonic distribution characteristics includes: Determine the center frequency and frequency bandwidth of the interference frequency point whose energy value exceeds the threshold value in the harmonic distribution characteristics, and use the center frequency and frequency bandwidth as initial stopband parameters of the notch filter; Calculating the stopband attenuation slope and phase compensation factor of the notch filter based on the amplitude ratio and phase difference between the main harmonic and the subharmonic in the harmonic distribution characteristics, and generating a stopband suppression enhancement parameter; Based on the initial stopband parameters and the stopband suppression enhancement parameters, dynamically offset compensation is performed on the cutoff frequency boundary of the notch filter to generate a cutoff frequency range after the transition band is optimized; Determining a passband ripple coefficient and a filter order of the notch filter according to the cutoff frequency range and a preset maximum group delay constraint; By iteratively adjusting the combination of the passband ripple coefficient and the filter order, the out-of-band suppression depth of the notch filter reaches the energy suppression requirement of the subharmonics in the harmonic distribution characteristics; The finally determined cutoff frequency range, stopband attenuation slope, phase compensation factor, passband ripple coefficient and filter order are packaged into the parameter configuration instruction.

10. The baseband radio frequency signal data feedback processing method based on vector modulation according to claim 1, characterized in that: The method further comprises: Calculating a channel quality index of the target radio frequency link based on error statistics accumulated in the closed-loop feedback control link; When the channel quality index is lower than a first threshold, reducing the modulation order and increasing error correction coding redundancy based on a first optimization strategy; and when the channel quality index is higher than a second threshold, increasing the modulation order and optimizing spectrum efficiency using compressed sensing technology based on a second optimization strategy; The adjusted modulation strategy parameters are encapsulated as control instructions, and the configuration parameters of the vector modulator are updated in real time through the closed-loop feedback control link.

Images