Multi-band processing system for electronic communication

Abstract

The invention relates to the technical field of information transmission, in particular to a multi-band processing system for electronic communication, which comprises a signal separation module, a frequency deviation compensation module, an amplitude correction module, a phase correction module and a dynamic feedback adjustment module. According to the invention, a plurality of frequency bands are extracted to execute band-pass filtering to separate signal components, short-time power spectral density is calculated to describe energy distribution, normalization coefficient operation is carried out to enhance frequency domain balance, high-energy component screening is carried out to reduce noise interference, signal effectiveness is improved, and frequency offset compensation calculates a change rate based on adjacent phase data and eliminates abnormal fluctuation. A time domain coefficient is adjusted to correct a frequency offset error, amplitude correction depends on short-time Fourier transform to screen the maximum amplitude to optimize signal consistency, phase adjustment fits a group delay error curve to optimize compensation parameters, dynamic feedback optimizes a strategy according to a signal-to-noise ratio to enhance compensation precision, and signal transmission quality and stability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information transmission, and in particular to a multi-band processing system for electronic communications. Background Art

[0002] The field of information transmission technology includes the design and implementation of wireless or wired transmission methods and related equipment for data, voice, image and other information. The core content of this technical field includes signal modulation and demodulation, multiplexing, channel coding, waveform design and transmission medium adaptation to ensure that data can be efficiently and stably transmitted from the sender to the receiver. Information transmission technology covers traditional analog communication, digital communication, optical communication and wireless communication, and is widely used in radio and television, satellite communication, mobile communication, the Internet and other fields. In the development of modern communication technology, information transmission has gradually evolved towards high bandwidth, low latency and multi-band compatibility to meet the data transmission needs in different application scenarios.

[0003] Among them, the multi-band processing system of electronic communications refers to a system that can receive, convert and process signals of different frequency bands. Aiming at the multi-band compatibility problem in the wireless communication environment, the system uses the RF front-end circuit to filter and amplify the signals of different frequency bands, generates the required local oscillator signal through the frequency synthesizer to complete the mixing conversion, and uses the baseband processing unit to perform signal demodulation, error correction and multi-channel data synthesis. The system can also combine spectrum sensing technology to realize dynamic monitoring of signals and frequency band allocation to meet the needs of multi-band parallel transmission in complex wireless communication environments.

[0004] The existing technology lacks a dynamic balancing strategy for signal frequency band processing, and the short-term power spectrum analysis is insufficient, which can easily cause an imbalance in energy distribution. Signals in some frequency bands may be overloaded or attenuated, affecting transmission quality. Frequency offset correction relies on fixed compensation and fails to dynamically adjust based on adjacent frame data, resulting in the accumulation of frequency offset errors and signal distortion. The amplitude control method does not fully consider the characteristics of the short-term power spectrum, and the amplification process is prone to introduce noise. Information loss is serious during attenuation, affecting signal integrity. Phase adjustment uses static compensation and lacks adaptive adjustment for different signal characteristics. Error accumulation exacerbates phase distortion. The dynamic adjustment strategy is fixed and does not combine the real-time signal status to optimize parameters. Adjustment lag affects system stability, which may lead to insufficient compensation or excessive correction, reducing transmission efficiency. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a multi-band processing system for electronic communications.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: A multi-band processing system for electronic communication comprises: The signal separation module obtains the input signal, extracts multiple frequency bands, performs bandpass filtering, separates the components of each frequency band, calculates the short-time power spectrum density, finds the normalization coefficient based on the spectrum energy, obtains the instantaneous power value, determines the frequency domain distribution, calculates the normalization ratio, verifies the signal balance state, filters the high-energy components, and generates a multi-band normalized signal; The frequency offset compensation module collects adjacent phases, calculates the phase change rate, calculates the mean of continuous frames, removes abnormal fluctuations, compares the frequency offset between frames, obtains the compensation factor, adjusts the time domain coefficient, and generates a time domain offset compensation signal based on the multi-band normalized signal; The amplitude correction module collects short-time Fourier transform data based on the time domain offset compensation signal, extracts power spectrum density, calculates attenuation factor, analyzes amplitude range, selects maximum amplitude, normalizes amplitude value, and generates normalized amplitude signal; Based on the normalized amplitude signal, the phase correction module collects group delay data, extracts phase trend, fits the error curve, determines the second-order change rate, selects compensation parameters, adjusts the phase, and generates a phase compensation signal.

[0007] As a further solution of the present invention, the multi-band normalized signal includes a low-frequency normalized output, a medium-frequency normalized output, and a high-frequency normalized output; the time domain offset compensation signal includes a phase smoothness value, a time domain correction amount, and an inter-frame equalization index; the normalized amplitude signal includes an amplitude standard amount, an amplitude equalization index, and an amplitude reference value; the phase compensation signal includes a phase correction value, a trend balance amount, and an error correction index.

[0008] As a further solution of the present invention, the signal separation module includes: The signal bandpass filtering submodule obtains the input signal, extracts multiple preset frequency bands, calls the filter to perform bandpass filtering, adjusts the filtering parameters according to the frequency range, adjusts the filter order, calculates the amplitude spectrum of the filtered signal, and compares the signal spectrum before and after filtering, extracts the energy distribution of the multi-band signal, and obtains the multi-band signal components; The short-time power spectrum calculation submodule calculates the short-time power spectrum density based on the multi-band signal components, obtains the signal power value in each time window, and determines the spectrum energy distribution using the formula: ; Obtain instantaneous power value through calculation, determine the energy proportion of differentiated frequency bands, and obtain spectrum energy distribution; in, Represents the instantaneous power value, Representative The signal components of the frequency band, Representative The center frequency of the band, Represents the number of frequency bands, represents the sum of the center frequencies used for normalization calculation, a represents the sub-band index, and b represents the sub-band index used for normalization calculation; The multi-band normalization submodule calculates the normalization ratio based on the spectrum energy distribution, adjusts the multi-band signal power, and calculates the normalized signal amplitude distribution. At the same time, it compares the signal power difference before and after the adjustment, screens the frequency bands with high energy proportion, synthesizes the screened signals, and obtains the multi-band normalized signals.

[0009] As a further solution of the present invention, the frequency offset compensation module includes: The phase change rate calculation submodule collects adjacent phases based on the multi-band normalized signal, calculates the phase change rate, extracts the phase change rate data of continuous frames, and calculates the mean to obtain the phase change rate mean; The abnormal fluctuation elimination submodule calls the phase change rate mean, counts the change range of continuous frames, eliminates abnormal fluctuation data, calculates the inter-frame frequency deviation, and obtains the frequency deviation data after eliminating the abnormal fluctuation; The time domain compensation calculation submodule calls the frequency deviation data after removing abnormal fluctuations, compares the frequency deviation between frames, obtains the compensation factor, and adjusts the time domain coefficient using the formula: ; Obtaining a time domain offset compensation factor by operation, adjusting the time domain signal, and generating a time domain offset compensation signal; in, represents the time domain offset compensation factor, Representative The frequency deviation value of the frame, Representative The weight parameters of the frame, Representative The phase value of the frame, Representative The signal amplitude of the frame, Representative The time coefficient of the frame, Represents the total number of frames. represents the frame number of phase change calculation, Represents the frame number of signal amplitude and time coefficient statistics, and c, d, and e represent frame indexes.

[0010] As a further solution of the present invention, the amplitude correction module includes: The short-time Fourier transform submodule calculates the frequency component of the signal in the short-time window based on the time domain offset compensation signal, intercepts the signal sequence using a sliding window, performs Fourier transform on the signal in each window, calculates the complex spectrum value after the transform, obtains the spectrum information on each time segment, and calls the modulus value of the transformed spectrum at the same time to generate short-time Fourier transform data; The power spectrum density extraction submodule calls the short-time Fourier transform data, calculates the power values ​​corresponding to the multi-frequency components, normalizes the power spectra of all time segments, calculates the power density curve by integration, obtains the energy distribution, and screens the key power distribution within the characteristic frequency band to generate power spectrum density data; The amplitude normalization submodule calls the power spectrum density data to calculate the attenuation factor of the signal within the characteristic frequency band using the formula: ; Calculate the weighted amplitude of each frequency component, extract the amplitude range, filter the maximum amplitude value, and perform normalization to obtain a normalized amplitude signal; in, Represents the amplitude value of the current frequency component, represents the attenuation factor, Represents frequency component The corresponding attenuation factor is, Represents the frequency components within the characteristic frequency band, represents the phase angle, represents the bias term associated with the frequency component, represents the total number of frequency components, represents the cosine value of the phase angle, and f represents the frequency component index.

[0011] As a further solution of the present invention, the phase correction module includes: The group delay acquisition submodule collects the group delay data of the signal at the differentiated frequency based on the normalized amplitude signal, calculates the group delay values ​​corresponding to multiple frequency points, and sorts the data to filter out the noise interference data to obtain the effective group delay data; The phase trend analysis submodule calls the effective group delay data, extracts the phase change trend based on the group delay data of multiple frequency points, calculates the second-order change rate, analyzes the error curve, and uses the formula: ; Calculate the phase adjustment amount and obtain the phase adjustment parameter; in, represents the phase adjustment amount, Representative The phase value at the frequency point, Representative The group delay value at the frequency point, Representative The adjustment coefficient of the frequency point, represents the total number of frequency points involved in the calculation, and g represents the frequency point index; The phase correction calculation submodule calls the phase adjustment parameter, selects a compensation strategy, performs phase adjustment on the normalized amplitude signal, and calculates and obtains a phase compensation signal.

[0012] As a further solution of the present invention, the system further includes: The dynamic feedback adjustment module collects the offset compensation value and the amplitude coefficient based on the phase compensation signal, extracts the phase coefficient, calculates the signal-to-noise ratio trend, compares the error gradient in the past, determines the update step size, filters out-of-range parameters, and generates a dynamic optimization compensation signal; The dynamic optimization compensation signal includes feedback correction parameters, error optimization amount, and signal-to-noise balance index.

[0013] As a further solution of the present invention, the dynamic feedback adjustment module includes: The offset compensation calculation submodule obtains the phase compensation signal, collects the offset compensation value and the amplitude coefficient, calculates the normalized ratio between the offset compensation value and the amplitude coefficient, compares the past data to obtain the offset change rate, and simultaneously screens the offset change rate exceeding the limit parameter, removes the exceeding limit data, and calculates the offset compensation equilibrium value; The signal-to-noise ratio trend analysis submodule extracts the phase coefficient based on the offset compensation equalization value, calculates the current signal-to-noise ratio, calls the signal-to-noise ratio data, calculates the current signal-to-noise ratio change rate, obtains the signal-to-noise ratio gradient trend based on the change rate, calculates the error amplitude according to the gradient trend, and generates the signal-to-noise ratio change trend; The dynamic optimization compensation generation submodule calls the signal-to-noise ratio change trend, compares the error gradient, calculates the optimization update step, removes out-of-range parameters, and uses the formula: ; Calculating a dynamic optimization coefficient, adjusting the phase compensation signal, and generating a dynamic optimization compensation signal; in, Represents the calculation result of the dynamic optimization compensation signal, represents the adjustment factor, represents the derivative of the rate of change of the signal-to-noise ratio with respect to time, represents the out-of-range error parameter, Represents the compensation constant.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, during the signal separation process, multiple frequency bands are extracted and bandpass filtering is performed to clearly separate the signal components. Short-time power spectral density calculation accurately depicts the energy distribution, and normalized coefficient calculation enhances frequency domain balance. High-energy component screening reduces noise interference and improves signal effectiveness. Frequency offset compensation calculates the rate of change through adjacent phase data, removes abnormal fluctuations in combination with the continuous frame mean, and adjusts the time domain coefficient to correct the frequency offset error. Amplitude correction relies on short-time Fourier transform to analyze the power spectrum, screens the maximum amplitude to optimize the signal amplitude consistency. Phase adjustment fits the error curve through group delay data, and adjusts the compensation parameters according to the rate of change to reduce the cumulative error. Dynamic feedback adjustment combines the signal-to-noise ratio trend to optimize the compensation strategy, screens out-of-range parameters according to the error gradient, and enhances compensation accuracy and system stability. In the process of multi-band signal processing, the synergy of energy balance, frequency offset correction, amplitude adjustment, phase optimization and dynamic compensation improves the signal transmission quality and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a system flow chart of the present invention; Figure 2 This is a flow chart of the signal separation module of the present invention; Figure 3 This is a flow chart of the frequency offset compensation module of the present invention; Figure 4 This is a flow chart of the amplitude correction module of the present invention; Figure 5 This is a flow chart of the phase correction module of the present invention; Figure 6 This is a flow chart of the dynamic feedback adjustment module of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0018] Embodiment 1 See also Figure 1 , a multi-band processing system for electronic communications comprises: The signal separation module obtains the input signal, extracts multiple frequency bands, performs bandpass filtering, separates the components of each frequency band, calculates the short-time power spectrum density, finds the normalization coefficient based on the spectrum energy, obtains the instantaneous power value, determines the frequency domain distribution, calculates the normalization ratio, verifies the signal balance state, filters the high-energy components, and generates a multi-band normalized signal; The frequency offset compensation module is based on multi-band normalized signals, collects adjacent phases, calculates the phase change rate, counts the continuous frame average, removes abnormal fluctuations, compares the frequency offset between frames, obtains the compensation factor, adjusts the time domain coefficient, and generates a time domain offset compensation signal; The amplitude correction module collects short-time Fourier transform data based on the time domain offset compensation signal, extracts the power spectrum density, calculates the attenuation factor, analyzes the amplitude range, selects the maximum amplitude, normalizes the amplitude value, and generates a normalized amplitude signal; The phase correction module collects group delay data based on the normalized amplitude signal, extracts the phase trend, fits the error curve, determines the second-order change rate, selects compensation parameters, adjusts the phase, and generates a phase compensation signal; The dynamic feedback adjustment module collects the offset compensation value and amplitude coefficient based on the phase compensation signal, extracts the phase coefficient, calculates the signal-to-noise ratio trend, compares the previous error gradient, determines the update step size, filters out-of-range parameters, and generates a dynamic optimization compensation signal.

[0019] The multi-band normalized signal includes a low-frequency normalized output, a medium-frequency normalized output, and a high-frequency normalized output. The time domain offset compensation signal includes a phase stationary value, a time domain correction amount, and an inter-frame equalization index. The normalized amplitude signal includes an amplitude standard amount, an amplitude equalization index, and an amplitude reference value. The phase compensation signal includes a phase correction value, a trend balance amount, and an error correction index. The dynamic optimization compensation signal includes feedback correction parameters, an error optimization amount, and a signal-to-noise balance index.

[0020] See also Figure 2 , the signal separation module includes: The signal bandpass filtering submodule obtains the input signal, extracts multiple preset frequency bands, calls the filter to perform bandpass filtering, adjusts the filtering parameters according to the frequency range, adjusts the filter order, calculates the amplitude spectrum of the filtered signal, and compares the signal spectrum before and after filtering, extracts the energy distribution of the multi-band signal, and obtains the multi-band signal components; First, perform frequency domain analysis on the input signal to determine its main energy distribution area, and configure the filter parameters according to different frequency ranges, and set the bandwidth of each frequency band. For example, for the processing of speech signals, the frequency band can be set to 300Hz-3.4kHz, and for biomedical signals (such as EEG signals), different frequency bands in the range of 0.5Hz-100Hz can be selected. Then, call the filter for bandpass filtering. Taking FIR filter as an example, the filter order needs to be set during design. The determination of the order is based on the filter accuracy requirements and computational complexity of the target frequency band. For example, if a steeper passband response is required, it can be used. A higher order (such as 64th order) can be used, otherwise a lower order (such as 16th order) can be used to reduce the computational burden. The filtered signal is Fourier transformed to obtain the amplitude spectrum and compared with the spectrum of the signal before filtering. By comparing the energy distribution of signals in different frequency bands, it is confirmed whether the signal components meet the set filtering characteristics. If it is found that the signal energy in a certain frequency band accounts for a small proportion, the filtering parameters can be appropriately adjusted to optimize the frequency band division. For example, in EEG signal processing, if the energy of the low frequency band (0.5Hz-4Hz) is too low, the passband range of the filter can be appropriately adjusted to improve the low frequency band energy capture capability, and finally extract multi-band signal components.

[0021] The short-time power spectrum calculation submodule calculates the short-time power spectrum density based on the multi-band signal components, obtains the signal power value in each time window, and determines the spectrum energy distribution using the formula: ; Obtain instantaneous power value through calculation, determine the energy proportion of differentiated frequency bands, and obtain spectrum energy distribution; in, Represents the instantaneous power value, Representative The signal components of the frequency band, Representative The center frequency of the band, Represents the number of frequency bands, represents the sum of the center frequencies used for normalization calculation, a represents the sub-band index, and b represents the sub-band index used for normalization calculation; First, set the short-time window length. For example, a 25ms window is often used in speech signal processing, and a 500ms window can be set in EEG signal analysis. Then, perform power calculation on the signal in each window, obtain the signal energy of each time window by calculating the square sum, and calculate the instantaneous power based on the spectrum energy distribution, using the formula: ; in, For the The signal components of the frequency band, represents the center frequency of the frequency band, is the total number of frequency bands. When calculating, the squares of the signals in each frequency band are summed and weighted and normalized. To illustrate with an actual example, suppose a signal is divided into three frequency bands, corresponding to center frequencies of 100 Hz, 500 Hz, and 1000 Hz, and the signal component energies are 0.2, 0.8, and 1.5, respectively. The calculation is as follows: ; The energy of each frequency band is calculated as follows: ; ; ; Get instantaneous power If this value is within a certain set threshold (for example, the preset threshold 1.5), the signal power in the window is determined to be high, otherwise the signal is considered to be weak. Through this calculation, the energy proportion of different frequency bands is determined for subsequent signal screening.

[0022] The multi-band normalization submodule calculates the normalization ratio based on the spectrum energy distribution, adjusts the multi-band signal power, and calculates the normalized signal amplitude distribution. At the same time, it compares the signal power difference before and after the adjustment, screens the frequency bands with high energy proportion, synthesizes the screened signals, and obtains the multi-band normalized signals.

[0023] First, set the normalization target value, such as setting the maximum energy component not to exceed 1.0 to avoid over-amplification of certain frequency bands and affecting the balance of the overall signal. When calculating the normalization ratio, take the current maximum energy value. As a baseline, calculate the normalization coefficient for each frequency band: ; Assuming the maximum component value of a signal energy is 2.0, the normalized ratio is: ; Then, all frequency band signal components are multiplied by the normalization ratio to ensure that the adjusted energy meets the preset range. After calculation, the signal power difference before and after adjustment is compared to select the frequency band with higher energy proportion and set the screening threshold. If it is set to 0.4, the low-frequency signal has low energy and is discarded, and only the mid- and high-frequency signals are retained. Finally, the filtered signals are synthesized to obtain multi-band normalized signals.

[0024] Table 1. Multi-band signal energy normalization results As shown in Table 1, the energy of different frequency bands is adjusted through normalization to maintain the overall balance of the signal. On this basis, the frequency bands with high energy proportion are screened, and finally the normalized signal is synthesized to meet the subsequent signal processing requirements.

[0025] See also Figure 3 , the frequency offset compensation module includes: The phase change rate calculation submodule collects adjacent phases based on the multi-band normalized signal, calculates the phase change rate, extracts the phase change rate data of consecutive frames, and calculates the mean to obtain the phase change rate mean; First, the phase values ​​of adjacent frames are obtained. In the specific execution process, multiple frames of data are extracted and the phase difference is calculated frame by frame. The phase value of the frame is , then the phase change rate can be expressed as ,in, It is obtained through frequency domain analysis, and normalization processing is performed on signals of different frequency bands so that their phase values ​​can be calculated under the same benchmark. By calculating the phase difference, the distribution of the phase change rate can be obtained. In the calculation process, the window sliding method is adopted, and each group of windows contains 10 frames of signal data (which can be adjusted according to the specific application scenario). The phase change rate in the window is averaged and defined as ,in Represents the number of window frames. In the process of calculating the mean, in order to reduce noise interference, abnormal data can be filtered out before the mean calculation, and the change rate threshold can be set. If the phase change rate of a frame exceeds the set threshold (for example, more than 3 times the standard deviation of the mean), the frame data is excluded, and finally the corrected phase change rate mean is obtained. This result shows that by calculating the mean of the phase change rate, the average phase change trend between frames can be obtained, which provides basic data for subsequent frequency deviation calculation, so that subsequent calculations are based on stable data rather than instantaneous phase changes interfered by noise.

[0026] The abnormal fluctuation elimination submodule calls the phase change rate mean, counts the change range of continuous frames, eliminates abnormal fluctuation data, calculates the inter-frame frequency deviation, and obtains the frequency deviation data after eliminating abnormal fluctuations; First, the phase change rate of all frames is analyzed and the range of change is calculated. If the range exceeds the set upper and lower limits, the abnormal fluctuation data is further screened, and the median de-extreme method is used to remove the extreme value frames that exceed the set upper and lower limits. The removal method uses dynamic threshold setting, such as setting the abnormal threshold ,in and They represent the mean and standard deviation of the rate of change within the normal range respectively, and the inter-frame frequency deviation is finally calculated. The formula is expressed as ,in is the frame period. During the calculation process, 1000 frames of data are used for statistical analysis. The example is as follows: Table 2 Frequency deviation calculation data table As shown in Table 2, by calculating the inter-frame frequency offset and eliminating abnormal fluctuations, stable frequency offset data is obtained to avoid the influence of extreme values ​​on the calculation accuracy. The result shows that the data after eliminating abnormal fluctuations can effectively reduce the interference of extreme values ​​on the frequency offset calculation, making the final obtained frequency offset data more stable, ensuring the accuracy of frequency offset compensation, and improving the reliability of subsequent time domain compensation calculation.

[0027] The time domain compensation calculation submodule calls the frequency deviation data after removing abnormal fluctuations, compares the frequency deviation between frames, obtains the compensation factor, and adjusts the time domain coefficient using the formula: ; Obtaining a time domain offset compensation factor by operation, adjusting the time domain signal, and generating a time domain offset compensation signal; in, represents the time domain offset compensation factor, Representative The frequency deviation value of the frame, Representative The weight parameters of the frame, Representative The phase value of the frame, Representative The signal amplitude of the frame, Representative The time coefficient of the frame, Represents the total number of frames. represents the frame number of phase change calculation, Represents the frame number of signal amplitude and time coefficient statistics, and c, d, and e represent frame indexes.

[0028] First, the weighted frequency deviation is calculated using a weighted average method with the weight parameter Determined according to the frame energy distribution, the calculation formula is as follows: ; Among them, the weight parameter It can be set according to the frame amplitude, such as setting , then the frame with larger amplitude has higher weight. In addition, in order to further improve the compensation accuracy, the inter-frame phase accumulation is introduced: ; Compute the normalized sum of squared magnitudes of a signal: ; Finally, the time domain offset compensation factor is calculated as: ; Among them, the time coefficient Based on the inter-frame stability calculation, such as ,in Represents the time normalization coefficient. Assuming the sampling frequency of a frame is 1000 Hz, the calculation is as follows: ; The calculation results are: ; ; Finally, the time domain compensation factor is obtained , adjust the time domain signal and generate a time domain offset compensation signal. The result shows that the calculated time domain compensation factor indicates the comprehensive influence of the inter-frame frequency offset. The larger the value, the greater the change of the inter-frame frequency offset, and the stronger the time domain compensation is needed. If it is in a smaller range (such as close to 1), it indicates that the frequency deviation is small and the system runs stably. Through the calculation of the compensation factor, it can be directly used to adjust the time domain characteristics of the signal, thereby compensating for the frequency offset error and ensuring the accuracy of signal synchronization.

[0029] See also Figure 4 , the amplitude correction module includes: The short-time Fourier transform submodule calculates the frequency components of the signal in the short-time window based on the time domain offset compensation signal, uses a sliding window to intercept the signal sequence, performs Fourier transform on the signal in each window, calculates the complex spectrum value after the transform, obtains the spectrum information on each time segment, and calls the modulus value of the transformed spectrum at the same time to generate short-time Fourier transform data; First, the input signal is sampled and the time domain signal is divided into fixed time windows. The length of each window is set to , each window interval is set to The window uses a Hanning window to reduce spectrum leakage. After intercepting the signal, a discrete Fourier transform (DFT) is performed on the signal in each window. sampling points, the signal is expressed as , and its transformation formula is: ; in, is the spectrum value after transformation, The value is 1024 to ensure sufficient frequency resolution. The DFT operation of each window generates a set of complex spectrum values, each of which contains amplitude and phase information. For example, for the signal After the transformation, the spectrum amplitudes at 50 Hz and 120 Hz are approximately 2 and 1.5 respectively. By repeatedly calculating the sliding window, the spectrum changes of the signal on the entire time axis are obtained, and the complete short-time Fourier transform data is obtained, and the spectrum information of each time segment is recorded, as shown in Table 3.

[0030] Table 3 Short-time Fourier transform data As shown in Table 3, the spectrum amplitudes in different time windows vary slightly, indicating that the instantaneous frequency components of the signal can be reflected by the short-time Fourier transform data, which can be used for subsequent power spectrum density calculations. This result shows that the short-time Fourier transform can accurately obtain the frequency components of the signal on the time axis and provide a data basis for power spectrum calculations. These changes in spectrum amplitudes can be used to analyze the energy distribution characteristics of the signal for subsequent calculation of the power spectrum density and normalization processing.

[0031] The power spectrum density extraction submodule calls short-time Fourier transform data, calculates the power values ​​corresponding to multiple frequency components, normalizes the power spectra of all time segments, calculates the power density curve using integration, obtains the energy distribution, and screens the key power distribution within the characteristic frequency band to generate power spectrum density data; Call the short-time Fourier transform data to calculate the power values ​​corresponding to the multiple frequency components. The power calculation formula for each frequency component is: ; in, For frequency The power value, is the amplitude of the frequency component, is the window length, let , for the 50Hz frequency point, the power is calculated as follows: ; All frequency points are calculated and normalized. The normalization calculation method is as follows: ; in, is the total power of all frequency components, is the total number of frequency components to ensure that the power is distributed in For the 50Hz frequency point, the normalized power is calculated as follows: ; After calculating the power spectrum of all time segments, the power density curve is calculated by integration. The calculation method is: ; in, is the target characteristic frequency band range, and the characteristic frequency band is , calculate the integral value to obtain the energy distribution in the characteristic frequency band, and select the frequency range with the highest energy density to obtain the power spectrum density data. This result shows that the power spectrum density calculation can effectively extract the energy changes in the characteristic frequency band and provide power data for amplitude normalization processing. The normalized power distribution data determines the main energy contribution of the signal characteristic frequency band and can be used for further processing of the signal amplitude.

[0032] The amplitude normalization submodule calls the power spectrum density data to calculate the attenuation factor of the signal within the characteristic frequency band using the formula: ; Calculate the weighted amplitude of each frequency component, extract the amplitude range, filter the maximum amplitude value, and perform normalization to obtain a normalized amplitude signal; in, Represents the amplitude value of the current frequency component, represents the attenuation factor, Represents frequency component The corresponding attenuation factor is, Represents the frequency components within the characteristic frequency band, represents the phase angle, represents the bias term associated with the frequency component, represents the total number of frequency components, represents the cosine value of the phase angle, and f represents the frequency component index.

[0033] The specific calculation method is as follows: ; in, represents the attenuation coefficient of the frequency component, let , represents the amplitude of each frequency component, is the corresponding frequency component, is the phase angle, is the bias term, and the frequency components and amplitudes in a certain time window are as shown in Table 4: Table 4 Frequency components and amplitudes Calculate the amplitude value after attenuation: ; Calculate the final normalized magnitude: ; The maximum amplitude value in the characteristic frequency band is screened and normalized to obtain the normalized amplitude signal. The result shows that the normalized amplitude signal has removed the power attenuation effect in the characteristic frequency band, making the contribution of the signal on different frequency components more balanced. At the same time, the signal adjusted by amplitude normalization can be used for subsequent signal analysis and classification processing to improve the analysis accuracy.

[0034] See also Figure 5 , the phase correction module includes: The group delay acquisition submodule collects the group delay data of the signal at the differentiated frequency based on the normalized amplitude signal, calculates the group delay values ​​corresponding to multiple frequency points, and sorts the data to filter out the noise interference data to obtain the effective group delay data; First, for the input normalized amplitude signal, the group delay data is collected at different frequency points. The specific method is to collect the group delay data of the set frequency point set. Measure each frequency point one by one The corresponding group delay It can be calculated by phase response, and the specific calculation method is: ; in, is the phase value at this frequency point, The system excites multiple frequency points with signals, measures the corresponding phase response by scanning, and calculates the group delay value. In practical applications, a linear frequency modulation pulse signal can be used to excite the system, and the phase changes at different frequency points can be recorded to obtain the corresponding group delay data. Then, for the collected group delay data, a data preprocessing method is used to eliminate abnormal data. First, the mean value of the group delay data is calculated. and standard deviation , then filter the data and set the threshold range , remove the data beyond this range to reduce the impact of noise, and finally obtain the effective group delay data set , as shown in Table 5.

[0035] Table 5 Group delay data collection and screening results The results show that after data screening, the abnormal group delay data with large deviations (such as 7.8ns at the 2.0GHz frequency point) are removed, and the effective group delay data can more accurately reflect the delay characteristics of the system at different frequency points, thereby providing more reliable input data for subsequent phase trend analysis.

[0036] The phase trend analysis submodule calls the effective group delay data, extracts the phase change trend based on the group delay data of multiple frequency points, calculates the second-order change rate, and analyzes the error curve using the formula: ; Calculate the phase adjustment amount and obtain the phase adjustment parameter; in, represents the phase adjustment amount, Representative The phase value at the frequency point, Representative The group delay value at the frequency point, Representative The adjustment coefficient of the frequency point, represents the total number of frequency points involved in the calculation, and g represents the frequency point index; First, the phase change trend is extracted based on the group delay data of multiple frequency points. The corresponding method is to calculate the phase increment of adjacent frequency points. , and further calculate the second-order rate of change , calculated as follows: ; in, For the The phase value at each frequency point, is the group delay value at this frequency point, is the adjustment coefficient, which is used to measure the impact of this frequency point on the overall trend. Or adaptively adjust based on signal quality. If the signal-to-noise ratio of a certain frequency point is low, reduce its weight, such as setting To reduce its impact, set Perform calculations: Frequency points , , , , corresponding to the group delay ns, ns, ns, ns Substituting into the formula: ; ; This result shows that the calculated phase adjustment It reflects the overall deviation of the phase change trend. The larger the value, the more uneven the phase change is. This value will be used as a key adjustment parameter in the subsequent phase compensation process to correct the phase deviation and improve the consistency of the signal.

[0037] The phase correction calculation submodule calls the phase adjustment parameters, selects the compensation strategy, performs phase adjustment on the normalized amplitude signal, and calculates and obtains the phase compensation signal.

[0038] The phase of the normalized amplitude signal is adjusted. The specific operation is to compensate the phase of the input signal: ; in is the compensation factor, which is set according to system performance, such as for high-precision applications , and for general scene settings ,use To perform compensation calculations:- - - - ; The adjusted phase compensation signal is ,Right now .

[0039] The results show that after the phase compensation calculation, the original phase data has been corrected, making the phase change smoother and reducing the nonlinear phase error caused by group delay changes. The adjusted phase value will provide a more stable phase response during signal transmission and processing, ensuring stable system performance.

[0040] See also Figure 6 , the dynamic feedback adjustment module includes: The offset compensation calculation submodule obtains the phase compensation signal, collects the offset compensation value and the amplitude coefficient, calculates the normalized ratio between the offset compensation value and the amplitude coefficient, compares the previous data to obtain the offset change rate, and simultaneously screens the offset change rate exceeding the limit parameter. After eliminating the exceeding limit data, the offset compensation equilibrium value is calculated; First, the phase compensation signal is received to obtain the offset compensation value and amplitude coefficient measured in the current system. The offset compensation value can be obtained by collecting the signal amplitude in the time domain through a high-precision sampler, and the amplitude coefficient is determined by calculating the ratio of the maximum value to the minimum value of the signal. Then, the offset compensation value and the amplitude coefficient are normalized. The calculation method of the normalized ratio is to divide the offset compensation value by the amplitude coefficient, so as to ensure that the signal offsets under different amplitudes can be compared under the same scale to improve the stability of the calculation. For example, if the current offset compensation value is 0.03V and the amplitude coefficient is 1.2V, the normalized ratio is calculated as follows: ; Then, the system extracts the past normalized ratio sequence from the historical data and calculates its rate of change. The rate of change is calculated by dividing the difference between the current normalized ratio and the normalized ratio at the previous moment by the normalized ratio at the previous moment. For example, if the normalized ratio at the previous moment is 0.028, the rate of change is calculated as follows: ; After obtaining the rate of change, the system further screens out the over-limit parameters, that is, determines whether the rate of change exceeds the set threshold. Assuming the threshold range is set to ±0.1, the current rate of change -0.107 exceeds the allowable range, so the data will be eliminated. After eliminating the over-limit data, the system calculates the average value of the remaining data to obtain the offset compensation balance value. Assuming that the remaining normalized ratio after eliminating the abnormal value is , the equilibrium value is calculated as follows: ; The results show that the offset compensation value of the current system remains relatively stable within the range of historical data, and after the normalized ratio calculation, its equilibrium value remains at 0.025. This means that in this step, after screening historical data and calculating the normalized ratio, a relatively reliable offset compensation benchmark value is obtained, which provides a basis for subsequent signal-to-noise ratio trend analysis.

[0041] The signal-to-noise ratio trend analysis submodule extracts the phase coefficient based on the offset compensation equalization value, calculates the current signal-to-noise ratio, calls the signal-to-noise ratio data, calculates the current signal-to-noise ratio change rate, and obtains the signal-to-noise ratio gradient trend based on the change rate. The error amplitude is calculated based on the gradient trend to generate the signal-to-noise ratio change trend; The specific method is to perform Fourier transform on the signal, obtain its phase information, and calculate the offset angle relative to the reference signal. For example, if the calculated phase offset angle is 5 degrees, the phase coefficient can be expressed as and ,Right now: ; ; Then, the current signal-to-noise ratio is calculated based on the phase coefficient. The signal-to-noise ratio calculation method is the ratio of the effective power of the signal to the noise power, that is: ; Assuming the signal power is 2.5mW and the noise power is 0.1mW, the signal-to-noise ratio is calculated as follows: ; Then, the system calls the historical signal-to-noise ratio data to calculate the current signal-to-noise ratio change rate. The change rate is calculated by dividing the difference between the current signal-to-noise ratio and the signal-to-noise ratio at the previous moment by the signal-to-noise ratio at the previous moment. Assuming that the signal-to-noise ratio at the previous moment is 13.8dB, the change rate is calculated as follows: ; Finally, the signal-to-noise ratio gradient trend is calculated based on the rate of change. The calculation method of the gradient trend is the derivative of the rate of change with respect to time, that is: ; Assumptions ,but ; The result shows that the current signal-to-noise ratio has a low rate of change per second, which means that the current signal quality of the system changes relatively smoothly. In this step, the calculated signal-to-noise ratio gradient trend provides input parameters for subsequent dynamic optimization compensation, ensuring that the compensation adjustment can accurately track the changing trend of signal quality.

[0042] The dynamic optimization compensation generation submodule calls the signal-to-noise ratio change trend, compares the error gradient, calculates the optimization update step size, and removes out-of-range parameters using the formula: ; Calculate the dynamic optimization coefficient, adjust the phase compensation signal, and generate a dynamic optimization compensation signal; in, represents the calculation result of the dynamic optimization compensation signal, represents the adjustment factor, represents the derivative of the rate of change of the signal-to-noise ratio with respect to time, represents the out-of-range error parameter, Represents the compensation constant.

[0043] Based on the trend of signal-to-noise ratio changes, the optimization update step size is calculated and out-of-range parameters are eliminated. The calculation formula for the optimization update step size is: ; Among them, the adjustment coefficient , maximum signal-to-noise ratio change rate , out-of-range error parameter , compensation constant , substitute into the calculation and get: ; The result shows that the update step size of the current optimized compensation signal is 3.0, which means that in this step, the system can dynamically adjust the phase compensation signal by calculating the update step size to match the current trend of the signal-to-noise ratio change and ensure the stability of the signal quality. The calculated update step size will be used to generate the final dynamic optimization compensation signal and continuously adjust the compensation amount to keep the signal-to-noise ratio stable within the expected range.

[0044] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A multi-band processing system for electronic communications, characterized in that: The system comprises: The signal separation module obtains the input signal, extracts multiple frequency bands, performs bandpass filtering, separates the components of each frequency band, calculates the short-time power spectrum density, finds the normalization coefficient based on the spectrum energy, obtains the instantaneous power value, determines the frequency domain distribution, calculates the normalization ratio, verifies the signal balance state, filters the high-energy components, and generates a multi-band normalized signal; The frequency offset compensation module collects adjacent phases, calculates the phase change rate, calculates the mean of continuous frames, removes abnormal fluctuations, compares the frequency offset between frames, obtains the compensation factor, adjusts the time domain coefficient, and generates a time domain offset compensation signal based on the multi-band normalized signal; The amplitude correction module collects short-time Fourier transform data based on the time domain offset compensation signal, extracts power spectrum density, calculates attenuation factor, analyzes amplitude range, selects maximum amplitude, normalizes amplitude value, and generates normalized amplitude signal; Based on the normalized amplitude signal, the phase correction module collects group delay data, extracts phase trend, fits the error curve, determines the second-order change rate, selects compensation parameters, adjusts the phase, and generates a phase compensation signal.

2. The multi-band processing system for electronic communication according to claim 1, characterized in that: The multi-band normalized signal includes a low-frequency normalized output, a medium-frequency normalized output, and a high-frequency normalized output; the time domain offset compensation signal includes a phase smoothness value, a time domain correction amount, and an inter-frame equalization index; the normalized amplitude signal includes an amplitude standard amount, an amplitude equalization index, and an amplitude reference value; the phase compensation signal includes a phase correction value, a trend balance amount, and an error correction index.

3. The multi-band processing system for electronic communication according to claim 2, characterized in that: The signal separation module comprises: The signal bandpass filtering submodule obtains the input signal, extracts multiple preset frequency bands, calls the filter to perform bandpass filtering, adjusts the filtering parameters according to the frequency range, adjusts the filter order, calculates the amplitude spectrum of the filtered signal, and compares the signal spectrum before and after filtering, extracts the energy distribution of the multi-band signal, and obtains the multi-band signal components; The short-time power spectrum calculation submodule calculates the short-time power spectrum density based on the multi-band signal components, obtains the signal power value in each time window, and determines the spectrum energy distribution using the formula: ； Obtain instantaneous power value through calculation, determine the energy proportion of differentiated frequency bands, and obtain spectrum energy distribution; in, Represents the instantaneous power value, Representative The signal components of the frequency band, Representative The center frequency of the band, Represents the number of frequency bands, represents the sum of the center frequencies used for normalization calculation, a represents the sub-band index, and b represents the sub-band index used for normalization calculation; The multi-band normalization submodule calculates the normalization ratio based on the spectrum energy distribution, adjusts the multi-band signal power, and calculates the normalized signal amplitude distribution. At the same time, it compares the signal power difference before and after the adjustment, screens the frequency bands with high energy proportion, synthesizes the screened signals, and obtains the multi-band normalized signals.

4. The multi-band processing system for electronic communication according to claim 3, characterized in that: The frequency offset compensation module comprises: The phase change rate calculation submodule collects adjacent phases based on the multi-band normalized signal, calculates the phase change rate, extracts the phase change rate data of continuous frames, and calculates the mean to obtain the phase change rate mean; The abnormal fluctuation elimination submodule calls the phase change rate mean, counts the change range of continuous frames, eliminates abnormal fluctuation data, calculates the inter-frame frequency deviation, and obtains the frequency deviation data after eliminating the abnormal fluctuation; The time domain compensation calculation submodule calls the frequency deviation data after removing abnormal fluctuations, compares the frequency deviation between frames, obtains the compensation factor, and adjusts the time domain coefficient using the formula: ； Obtaining a time domain offset compensation factor by operation, adjusting the time domain signal, and generating a time domain offset compensation signal; in, represents the time domain offset compensation factor, Representative The frequency deviation value of the frame, Representative The weight parameters of the frame, Representative The phase value of the frame, Representative The signal amplitude of the frame, Representative The time coefficient of the frame, Represents the total number of frames. represents the frame number of phase change calculation, Represents the frame number of signal amplitude and time coefficient statistics, and c, d, and e represent frame indexes.

5. The multi-band processing system for electronic communication according to claim 4, characterized in that: The amplitude correction module comprises: The short-time Fourier transform submodule calculates the frequency component of the signal in the short-time window based on the time domain offset compensation signal, intercepts the signal sequence using a sliding window, performs Fourier transform on the signal in each window, calculates the complex spectrum value after the transform, obtains the spectrum information on each time segment, and calls the modulus value of the transformed spectrum at the same time to generate short-time Fourier transform data; The power spectrum density extraction submodule calls the short-time Fourier transform data, calculates the power values ​​corresponding to the multi-frequency components, normalizes the power spectra of all time segments, calculates the power density curve by integration, obtains the energy distribution, and screens the key power distribution within the characteristic frequency band to generate power spectrum density data; The amplitude normalization submodule calls the power spectrum density data to calculate the attenuation factor of the signal within the characteristic frequency band using the formula: ； Calculate the weighted amplitude of each frequency component, extract the amplitude range, filter the maximum amplitude value, and perform normalization to obtain a normalized amplitude signal; in, Represents the amplitude value of the current frequency component, represents the attenuation factor, , represents the phase angle, represents the bias term associated with the frequency component, represents the total number of frequency components, represents the cosine value of the phase angle, and f represents the frequency component index.

6. The multi-band processing system for electronic communication according to claim 5, characterized in that: The phase correction module comprises: The group delay acquisition submodule collects the group delay data of the signal at the differentiated frequency based on the normalized amplitude signal, calculates the group delay values ​​corresponding to multiple frequency points, and sorts the data to filter out the noise interference data to obtain the effective group delay data; The phase trend analysis submodule calls the effective group delay data, extracts the phase change trend based on the group delay data of multiple frequency points, calculates the second-order change rate, analyzes the error curve, and uses the formula: ； Calculate the phase adjustment amount and obtain the phase adjustment parameter; in, represents the phase adjustment amount, Representative The phase value at the frequency point, Representative The group delay value at the frequency point, Representative The adjustment coefficient of the frequency point, represents the total number of frequency points involved in the calculation, and g represents the frequency point index; The phase correction calculation submodule calls the phase adjustment parameter, selects a compensation strategy, performs phase adjustment on the normalized amplitude signal, and calculates and obtains a phase compensation signal.

7. The multi-band processing system for electronic communication according to claim 6, characterized in that: The system further comprises: The dynamic feedback adjustment module collects the offset compensation value and the amplitude coefficient based on the phase compensation signal, extracts the phase coefficient, calculates the signal-to-noise ratio trend, compares the error gradient in the past, determines the update step size, filters out-of-range parameters, and generates a dynamic optimization compensation signal; The dynamic optimization compensation signal includes feedback correction parameters, error optimization amount, and signal-to-noise balance index.

8. The multi-band processing system for electronic communication according to claim 7, characterized in that: The dynamic feedback adjustment module comprises: The offset compensation calculation submodule obtains the phase compensation signal, collects the offset compensation value and the amplitude coefficient, calculates the normalized ratio between the offset compensation value and the amplitude coefficient, compares the past data to obtain the offset change rate, and simultaneously screens the offset change rate exceeding the limit parameter, removes the exceeding limit data, and calculates the offset compensation equilibrium value; The signal-to-noise ratio trend analysis submodule extracts the phase coefficient based on the offset compensation equalization value, calculates the current signal-to-noise ratio, calls the signal-to-noise ratio data, calculates the current signal-to-noise ratio change rate, obtains the signal-to-noise ratio gradient trend based on the change rate, calculates the error amplitude according to the gradient trend, and generates the signal-to-noise ratio change trend; The dynamic optimization compensation generation submodule calls the signal-to-noise ratio change trend, compares the error gradient, calculates the optimization update step, removes out-of-range parameters, and uses the formula: ； Calculating a dynamic optimization coefficient, adjusting the phase compensation signal, and generating a dynamic optimization compensation signal; in, Represents the calculation result of the dynamic optimization compensation signal, represents the adjustment factor, represents the derivative of the rate of change of the signal-to-noise ratio with respect to time, represents the out-of-range error parameter, Represents the compensation constant.

Images