Digital communication signal enhancement processing method

Abstract

The invention provides a digital communication signal enhancement processing method, which relates to the technical field of signal processing, and comprises the following specific steps: constructing a digital communication signal data set; constructing an adaptive basis function by combining the local instantaneous phase and the energy center-of-gravity offset; the key sampling points are dynamically focused through Gaussian weighting, digital communication signals are coded to an adaptive basis function, and a coding matrix is constructed; performing noise suppression on the coding matrix by using a dynamic threshold and nonlinear compression; fusing multi-scale back projection and global energy balance, and performing final enhancement processing on the digital communication signal; meanwhile, a digital communication signal enhancement processing model is provided, and digital communication signal enhancement processing is efficiently completed through the model.

Description

Technical Field

[0001] The present invention belongs to the technical field of signal processing, and in particular relates to a digital communication signal enhancement processing method. Background Art

[0002] With the continuous development of information technology, digital communication systems have been widely used in all walks of life; however, with the complexity of the communication environment and the increase of signal noise, the existing digital communication signal processing methods face many challenges; traditional digital communication signal enhancement technology often relies on global processing methods, ignoring the local characteristics of signals in different frequency bands and instantaneous changes, resulting in poor recovery of digital communication signals, and the clarity and reliability of digital communication signals are difficult to be effectively guaranteed, affecting the overall performance of the communication system.

[0003] In addition, existing noise suppression technologies usually use static threshold settings and cannot be dynamically adjusted according to the characteristics and real-time changes of digital communication signals, thereby reducing the accuracy and adaptability of noise suppression; traditional digital communication signal enhancement methods also find it difficult to balance high-precision signal recovery and anti-interference capabilities, which often leads to distortion and information loss in the digital communication signal reconstruction process; therefore, how to effectively improve the quality of digital communication signals, reduce noise interference and ensure the stable transmission of digital communication signals in complex communication environments has become a major problem in current digital communication technology. Summary of the invention

[0004] The present invention provides a digital communication signal enhancement processing method, aiming to propose a digital communication signal enhancement processing model, wherein the digital communication signal is subjected to window processing, and an adaptive basis function is constructed based on the instantaneous phase and the energy center of gravity offset, so as to accurately capture the local characteristics of the digital communication signal; the key sampling points in each window are focused by using the Gaussian weighting technology, so as to enhance the detail performance of the digital communication signal and construct a coding matrix; the dynamic threshold and nonlinear compression technology driven by the signal-to-noise ratio are used to further suppress the noise, retain the core information of the digital communication signal, and ensure the clarity of the digital communication signal; the digital communication signal is reconstructed and enhanced by using the multi-scale back-projection and global energy balancing technology, so as to improve the robustness and stability of the digital communication signal; and the digital communication signal enhancement processing model is used to complete the enhancement processing of the digital communication signal.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a method for enhancing digital communication signals, the specific steps of which are as follows: S1, collect and construct digital communication signal data set; S2, performing window processing on the digital communication signal, and constructing an adaptive basis function based on the local instantaneous phase and energy center of gravity offset; S3, dynamically focusing on key sampling points by Gaussian weighting, encoding the digital communication signal into an adaptive basis function, and constructing a coding matrix; S4, performing noise suppression on the coding matrix using a dynamic threshold and nonlinear compression driven by a signal-to-noise ratio; S5, fusing multi-scale back-projection and global energy balance to reconstruct the signal and perform final enhancement processing on the digital communication signal; S6. Build a digital communication signal enhancement processing model, input the digital communication signal data set into the model, execute S2 to S5 in sequence and perform model training to complete the digital communication signal enhancement processing.

[0006] Preferably, in step S1, for the digital communication signal data set, a high-precision digital communication receiving device is used to collect digital communication signals in real time, including cellular networks, satellite communications and shortwave communication signals, covering typical noise scenarios of Gaussian white noise, multipath fading and pulse interference, and stored in complex form as a binary file; the sampling rate and the total number of sampling points are set, the training set and the test set are divided proportionally, each sample is labeled with the true signal-to-noise ratio, modulation mode and noise type, and a digital communication signal data set is constructed.

[0007] Preferably, in step S2, the specific steps of constructing the adaptive basis function are: the input is the original digital communication signal , the sampling point index is , the total number of signal sampling points is P, the input is divided into windows, the window length is W, , the window slides point by point, the step size is 1, and the total number of windows is , by calculating the instantaneous phase of each window and the energy center of gravity offset to generate an adaptive basis function; Adaptive basis function The mathematical model is: ; Where w is the window index, , is the total number of windows, b is the adaptive basis function index, , is the number of adaptive basis functions, is the instantaneous phase, is the energy center of gravity offset, is the zero-proof constant, m is the index; Instantaneous phase The mathematical model is: ; In the formula, is the average frequency of the signal in the window; Energy center of gravity offset The mathematical model is: .

[0008] Preferably, in step S2, the design of the adaptive basis function aims to effectively cope with the variability of instantaneous characteristics and noise interference in digital communication signals, and to improve the recovery effect of digital communication signals in complex environments; first, by performing window processing on the digital communication signal and calculating the instantaneous phase and energy center of gravity offset of each window, the fluctuation characteristics of the digital communication signal in the local range can be captured, providing accurate local signal information for constructing the adaptive basis function; then, based on the local characteristics, the generated adaptive basis function can dynamically adjust the signal processing strategy to avoid the distortion caused by the global processing method and ensure the enhanced accuracy; the overall design can flexibly adjust the enhancement strategy according to the characteristics of the digital communication signal, improve the recovery effect in a high noise environment, reduce noise interference, and thus ensure the stability and reliability of the communication quality.

[0009] Preferably, in step S3, the specific steps of constructing the coding matrix are: projecting the input original digital communication signal to the adaptive basis function, calculating the projection coefficient for each window and the adaptive basis function, weighting the signal points in the window by the Gaussian kernel function, focusing on the key sampling points of the energy center of gravity, and using the calculated projection coefficients as matrix elements to construct the coding matrix ; The projection coefficient of the wth window on the bth adaptive basis function The mathematical model is: ; In the formula, is the Gaussian kernel function, is the signal fluctuation intensity of the wth window; The signal fluctuation intensity of the wth window The mathematical model is: ; In the formula, is the mean of the w-th window.

[0010] Preferably, in step S3, the construction of the coding matrix aims to further improve the quality and enhancement effect of the digital communication signal through refined signal processing; first, by projecting the original digital communication signal into the adaptive basis function space and calculating the projection coefficient of each window, the local characteristics of the digital communication signal are accurately captured; then, the signal points in each window are weighted by combining the Gaussian kernel function, focusing on the key sampling points of the energy center of gravity, effectively enhancing the core characteristics of the digital communication signal, and avoiding information loss caused by the global processing process; the overall design constructs a coding matrix to ensure the integrity of the digital communication signal while effectively improving the clarity and accuracy of the signal, reducing noise interference, and improving the transmission effect of the digital communication signal, providing a solid foundation for subsequent noise suppression and signal reconstruction.

[0011] Preferably, in step S4, the specific steps of performing noise suppression on the coding matrix are: the input is the coding matrix, combined with the projection coefficient, and the dynamic threshold is calculated for each adaptive basis function; nonlinear compression is performed on each projection coefficient, and all projection coefficients after nonlinear compression processing constitute the enhanced coding matrix; the dynamic threshold of the bth adaptive basis function The mathematical model is: ; In the formula, is the signal-to-noise ratio of the original digital communication signal, which is defined as the ratio of the original digital communication signal power to the noise power. is the original digital communication signal power, is the noise power, is a noise-free reference signal; Enhanced projection coefficient The mathematical model is: ; In the formula, is the steepness coefficient of the bth adaptive basis function, .

[0012] Preferably, in step S4, the specific step of noise suppression is aimed at effectively reducing the noise interference in the coding matrix and further improving the quality of the digital communication signal; first, the dynamic threshold of each adaptive basis function is calculated by combining the projection coefficient, and the noise suppression strength is dynamically adjusted according to the signal-to-noise ratio of the digital communication signal, and the noise suppression effect is adaptively adjusted according to different signal characteristics to ensure that important information in the digital communication signal is retained; then, the projection coefficient is processed by nonlinear compression, which can effectively reduce the impact of noise on the signal, while retaining the key features of the digital communication signal and improving the clarity of the digital communication signal; the overall design dynamically adjusts the noise suppression parameters to achieve adaptive noise suppression under different signal-to-noise ratio conditions, optimizes the recovery effect of the digital communication signal, enhances the system's anti-interference ability in the communication environment, and provides a reliable foundation for subsequent signal enhancement and reconstruction.

[0013] Preferably, in step S5, the specific steps of performing final enhancement processing on the digital communication signal are: inputting the enhanced coding matrix For each window, a time domain digital communication signal is generated through multi-scale back projection; the window reconstruction results are spliced ​​through impulse functions to obtain a reconstructed digital communication signal; energy balancing is performed on the global signal, and the final enhanced digital communication signal is obtained by combining the original digital communication signal; the time domain digital communication signal The mathematical model is: ; In the formula, To prevent zero constant; Reconstructing digital communication signals The mathematical model is: ; In the formula, is the impulse function, which is 1 when p=w, otherwise it is 0; Final enhanced digital communication signal The mathematical model is: ; In the formula, is the energy balance smoothing coefficient, is the original digital communication signal, To prevent zero constant.

[0014] Preferably, in step S5, the final enhancement processing of the digital communication signal aims to effectively improve the quality and stability of the digital communication signal through multi-scale back projection and global energy balance; first, a time-domain digital communication signal is generated through multi-scale back projection, and the signal information at different frequencies and scales is fused to restore the detail information in the digital communication signal, enhance the detail performance of the signal, and reduce the distortion occurring in the signal reconstruction process; then, the reconstructed digital communication signal is spliced ​​through an impulse function to ensure the continuity and integrity of the time domain signal; finally, the overall strength of the digital communication signal is optimized in combination with global energy balance, the energy distribution of the signal is adjusted, the phenomenon of local over-strength or over-weakness is avoided, and the stability of the digital communication signal is ensured; the overall design effectively improves the robustness and stability of the digital communication signal, enhances the fidelity of the digital communication signal, and ensures the efficient transmission and accurate recovery of the digital communication signal through the combination of multi-scale processing and global energy balance.

[0015] Preferably, in step S6, for the digital communication signal enhancement processing model, firstly, a digital communication signal data set is input, and each processing step is executed in sequence, and the specific steps are: in the digital communication signal windowing stage, the input digital communication signal is windowed according to the characteristics of the signal, and the instantaneous phase and energy center of gravity offset of each window are calculated to construct an adaptive basis function; then, the signal of each window is projected into the adaptive basis function space by using Gaussian weighted dynamic focusing key sampling points, and the projection coefficients are calculated to generate a coding matrix; in the noise suppression stage, the coding matrix is ​​processed based on the dynamic threshold and nonlinear compression technology driven by the signal-to-noise ratio, so as to effectively reduce noise interference and retain the key information in the digital communication signal; finally, the digital communication signal is reconstructed and enhanced by using multi-scale back projection and global energy balancing technology to ensure the clarity and stability of the digital communication signal; through the training and parameter optimization of the model, the model can automatically adapt to the enhancement requirements under different signal conditions, effectively improve the quality of the digital communication signal and ensure stable transmission, and finally output the enhanced digital communication signal processing result.

[0016] Compared with the prior art, the present invention has the following technical effects: the technical solution provided by the present invention proposes a digital communication signal enhancement processing model, wherein the digital communication signal is subjected to window processing, and an adaptive basis function is constructed based on the instantaneous phase and the energy center of gravity offset to accurately capture the local characteristics of the digital communication signal; the Gaussian weighting technology is used to focus on the key sampling points in each window, the detail performance of the digital communication signal is enhanced, and a coding matrix is ​​constructed; based on the dynamic threshold and nonlinear compression technology driven by the signal-to-noise ratio, the noise is further suppressed, the core information of the digital communication signal is retained, and the clarity of the digital communication signal is ensured; the digital communication signal is reconstructed and enhanced through multi-scale back projection and global energy balancing technology to improve the robustness and stability of the digital communication signal; the digital communication signal enhancement processing model is used to complete the enhancement processing of the digital communication signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention provides a flowchart of a digital communication signal enhancement processing method.

[0018] Figure 2 It is a flow chart of constructing an adaptive basis function provided by the present invention.

[0019] Figure 3 It is a flow chart of constructing a coding matrix provided by the present invention.

[0020] Figure 4 It is a flow chart of performing noise suppression on a coding matrix provided by the present invention.

[0021] Figure 5It is a flow chart of the final enhancement processing of the digital communication signal provided by the present invention.

[0022] Figure 6 It is the digital communication signal provided by the present invention that has not been processed by the digital communication signal enhancement processing model.

[0023] Figure 7 It is a digital communication signal processed by the digital communication signal enhancement processing model provided by the present invention. DETAILED DESCRIPTION

[0024] The present invention aims to propose a digital communication signal enhancement processing method and a digital communication signal enhancement processing model, wherein the digital communication signal is subjected to window processing and an adaptive basis function is constructed based on the instantaneous phase and the energy center of gravity offset to accurately capture the local characteristics of the digital communication signal; the key sampling points in each window are focused by using Gaussian weighting technology to enhance the detail performance of the digital communication signal and to construct a coding matrix; the dynamic threshold and nonlinear compression technology driven by the signal-to-noise ratio are used to further suppress noise, retain the core information of the digital communication signal, and ensure the clarity of the digital communication signal; the digital communication signal is reconstructed and enhanced by multi-scale back projection and global energy balancing technology to improve the robustness and stability of the digital communication signal; and the digital communication signal enhancement processing model is used to complete the enhancement processing of the digital communication signal.

[0025] See also Figure 1 As shown, a digital communication signal enhancement processing method in an embodiment of the present application, the specific steps are as follows.

[0026] S1. Collect and construct digital communication signal dataset.

[0027] Furthermore, in step S1, for the digital communication signal dataset, a software-defined radio module is used to collect digital communication signals in an actual communication environment, including cellular networks, satellite communications, and shortwave communication signals, covering typical noise scenarios of Gaussian white noise, multipath fading, and pulse interference, with a signal-to-noise ratio ranging from -10 dB to 20 dB, and digital communication signals are stored in complex form as binary files; the sampling rate Set to 10MHz, quantization bit is 16 bits, each signal segment duration is 1 millisecond, corresponding to the total number of sampling points P=10 4 , the total number of samples in the digital communication signal dataset is 10 5 , the training set and the test set are divided into 8:2, and each sample is labeled with the true signal-to-noise ratio, modulation method and noise type to construct a digital communication signal dataset.

[0028] S2. Perform window processing on the digital communication signal and construct an adaptive basis function based on the local instantaneous phase and energy center of gravity offset.

[0029] Furthermore, in step S2, an adaptive basis function is constructed, and the process is as follows: Figure 2 As shown, the specific steps of constructing the adaptive basis function are:

[0030] Input is the original digital communication signal , the sampling point index is , the total number of signal sampling points is P, P=10 4 , divide the input into windows with a window length of W, =1000, the window slides point by point, the step size is 1, and the total number of windows is , =P-W+1=9001, the adaptive basis function is generated by calculating the instantaneous phase and energy center of gravity offset of each window; adaptive basis function The mathematical model is: ; Where w is the window index, , is the total number of windows, b is the adaptive basis function index, , is the number of adaptive basis functions, is the instantaneous phase, is the energy center of gravity offset, m is the index, To prevent zero constant, =9, ; Instantaneous phase The mathematical model is: ; In the formula, is the average frequency of the signal in the window; Energy center of gravity offset The mathematical model is: .

[0031] S3. By dynamically focusing on key sampling points with Gaussian weighting, the digital communication signal is encoded into an adaptive basis function to construct a coding matrix.

[0032] Furthermore, in step S3, the process for constructing the encoding matrix is ​​as follows: Figure 3 As shown, the specific steps of constructing the coding matrix are:

[0033] The input original digital communication signal is projected onto the adaptive basis function, and the projection coefficient is calculated for each window and adaptive basis function. The signal points in the window are weighted by the Gaussian kernel function to focus on the key sampling points of the energy center of gravity. The calculated projection coefficient is used as the matrix element to construct the coding matrix ; The projection coefficient of the wth window on the bth adaptive basis function The mathematical model is: ; In the formula, is the Gaussian kernel function, is the signal fluctuation intensity of the wth window; The signal fluctuation intensity of the wth window The mathematical model is: ; In the formula, is the mean of the w-th window.

[0034] S4. Perform noise suppression on the coding matrix using a dynamic threshold driven by the signal-to-noise ratio and nonlinear compression.

[0035] Further, in step S4, noise suppression is performed on the coding matrix, and the process is as follows: Figure 4 As shown, the specific steps of performing noise suppression on the coding matrix are:

[0036] The input is the encoding matrix, combined with the projection coefficients, and the dynamic threshold is calculated for each adaptive basis function; nonlinear compression is performed on each projection coefficient, and all projection coefficients after nonlinear compression constitute the enhanced encoding matrix; the dynamic threshold of the bth adaptive basis function The mathematical model is: ; In the formula, is the signal-to-noise ratio of the original digital communication signal, which is defined as the ratio of the original digital communication signal power to the noise power. is the original digital communication signal power, is the noise power, is a noise-free reference signal, and a noise-free reference signal is generated by simulation; Noise-free reference signal The mathematical model is: ; In the formula, is the symbol amplitude of the quadrature phase shift keying (QPSK) modulation, which is taken as the normalized level. , is the carrier frequency, =1MHz, is the phase state of QPSK modulation, , is the sampling interval, the sampling rate =10MHz, , K is the total number of modulation symbols, K=1000; Enhanced projection coefficient The mathematical model is: ; In the formula, is the steepness coefficient of the bth adaptive basis function, .

[0037] S5. Fusion of multi-scale back-projection and global energy balance reconstructs the signal and performs final enhancement processing on the digital communication signal.

[0038] Further, in step S5, the digital communication signal is subjected to final enhancement processing, and the process is as follows: Figure 5 As shown, the specific steps of performing final enhancement processing on the digital communication signal are as follows.

[0039] The input is the enhanced encoding matrix For each window, a time domain digital communication signal is generated through multi-scale back projection; the window reconstruction results are spliced ​​through impulse functions to obtain a reconstructed digital communication signal; energy balancing is performed on the global signal, and the final enhanced digital communication signal is obtained by combining the original digital communication signal; the time domain digital communication signal The mathematical model is: ; In the formula, To prevent zero constant, ; Reconstructing digital communication signals The mathematical model is: ; In the formula, is the impulse function, which is 1 when p=w, otherwise it is 0; Final enhanced digital communication signal The mathematical model is: ; In the formula, is the energy balance smoothing coefficient, is the original digital communication signal, To prevent zero constant, , .

[0040] S6. Build a digital communication signal enhancement processing model, input the digital communication signal data set into the model, execute S2 to S5 in sequence and perform model training to complete the digital communication signal enhancement processing.

[0041] Furthermore, in step S6, for the digital communication signal enhancement processing model, firstly, a digital communication signal data set is input, and each processing step is executed in sequence. The specific steps are: in the digital communication signal windowing stage, the input digital communication signal is windowed according to the characteristics of the signal, and the instantaneous phase and energy center of gravity offset of each window are calculated to construct an adaptive basis function; then, the signal of each window is projected into the adaptive basis function space by using Gaussian weighted dynamic focusing key sampling points, and the projection coefficients are calculated to generate a coding matrix; in the noise suppression stage, the coding matrix is ​​processed based on the dynamic threshold and nonlinear compression technology driven by the signal-to-noise ratio, which effectively reduces noise interference and retains the key information in the digital communication signal; finally, the digital communication signal is reconstructed and enhanced by using multi-scale back projection and global energy balancing technology to ensure the clarity and stability of the digital communication signal; through model training and parameter optimization, the model can automatically adapt to the enhancement requirements under different signal conditions, effectively improve the quality of digital communication signals and ensure stable transmission, and finally output the enhanced digital communication signal processing results.

[0042] Furthermore, in step S6, for the digital communication signal enhancement processing model, it is developed based on the Python programming language, using the PyTorch framework, the input is the digital communication signal dataset, the batch size is set to 64, the Adam optimizer is used, the learning rate is set to 0.001, the number of training times is 100, and after the iterative training is completed, the digital communication signal enhancement processing model outputs the digital communication signal enhancement processing result.

[0043] Furthermore, in step S6, the digital communication signal is input into the digital communication signal enhancement processing model for processing, and the effects before and after the processing are as follows: Figure 6 and Figure 7 As shown in the figure, the horizontal axis represents time in milliseconds, and the vertical axis represents the amplitude of the digital communication signal in volts; before the model processing, the digital communication signal fluctuates greatly in the time domain, and the noise interference is obvious, resulting in unstable signal amplitude, affecting the clarity and availability of the digital communication signal; after the model enhancement processing, the noise is effectively suppressed, the amplitude fluctuation of the digital communication signal is significantly reduced, and the overall digital communication signal becomes smoother and clearer; the processing results shown in the figure show that the digital communication signal enhancement processing model can significantly improve the quality and stability of the digital communication signal, thereby realizing more reliable digital communication signal transmission, verifying the effectiveness of the model proposed in this paper.

[0044] The above are only preferred embodiments of the present invention. It should be pointed out that a person skilled in the art can make several modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A digital communication signal enhancement processing method, characterized in that: The specific steps include: S1, collect and construct digital communication signal data set; S2, performing window processing on the digital communication signal, and constructing an adaptive basis function based on the local instantaneous phase and energy center of gravity offset; S3, dynamically focusing on key sampling points by Gaussian weighting, encoding the digital communication signal into an adaptive basis function, and constructing a coding matrix; S4, performing noise suppression on the coding matrix using a dynamic threshold and nonlinear compression driven by a signal-to-noise ratio; S5, fusing multi-scale back-projection and global energy balance to reconstruct the signal and perform final enhancement processing on the digital communication signal; S6. Build a digital communication signal enhancement processing model, input the digital communication signal data set into the model, execute S2 to S5 in sequence and perform model training to complete the digital communication signal enhancement processing.

2. A digital communication signal enhancement processing method according to claim 1, characterized in that: In the step S2, the specific steps of constructing the adaptive basis function are: the input is the original digital communication signal , the sampling point index is , the total number of signal sampling points is P, and the input is divided into The window length is W, and the window slides point by point. The adaptive basis function is generated by calculating the instantaneous phase and energy center of gravity offset of each window; the adaptive basis function The mathematical model is: ； Where w is the window index, , is the total number of windows, b is the adaptive basis function index, , is the number of adaptive basis functions, is the instantaneous phase, is the energy center of gravity offset, is the anti-zero constant and m is the index.

3. A digital communication signal enhancement processing method according to claim 2, characterized in that: In the step S3, the specific steps of constructing the coding matrix are: projecting the input original digital communication signal to the adaptive basis function, calculating the projection coefficient for each window and the adaptive basis function, weighting the signal points in the window by the Gaussian kernel function, focusing on the key sampling points of the energy center of gravity, and using the calculated projection coefficients as matrix elements to construct the coding matrix ; The projection coefficient of the wth window on the bth adaptive basis function The mathematical model is: ； In the formula, is the signal fluctuation intensity of the wth window.

4. A digital communication signal enhancement processing method according to claim 3, characterized in that: In the step S4, the specific steps of performing noise suppression on the coding matrix are: the input is the coding matrix, combined with the projection coefficient, and the dynamic threshold is calculated for each adaptive basis function; nonlinear compression is performed on each projection coefficient, and all projection coefficients after nonlinear compression constitute the enhanced coding matrix; the dynamic threshold of the bth adaptive basis function The mathematical model is: ； Where SNR is the signal-to-noise ratio of the original digital communication signal; Enhanced projection coefficient The mathematical model is: ； In the formula, is the steepness coefficient of the bth adaptive basis function.

5. A digital communication signal enhancement processing method according to claim 4, characterized in that: In the step S5, the specific steps of performing the final enhancement processing on the digital communication signal are as follows: the input is the enhanced coding matrix For each window, a time domain digital communication signal is generated through multi-scale back projection; the window reconstruction results are spliced ​​through impulse functions to obtain a reconstructed digital communication signal; energy balancing is performed on the global signal, and the final enhanced digital communication signal is obtained by combining the original digital communication signal; the time domain digital communication signal The mathematical model is: ； In the formula, To prevent zero constant; Reconstructing digital communication signals The mathematical model is: ； In the formula, is the impulse function; Final enhanced digital communication signal The mathematical model is: ； In the formula, is the energy balance smoothing coefficient, is the original digital communication signal, To prevent zero constant.

Images