Modulation format identification method and device of modulation signal

Abstract

The invention discloses a modulation format identification method and device of a modulation signal. The method comprises the following steps: acquiring a modulation signal transmitted by a satellite-ground link; mapping the modulation signal into a binary gray scale constellation diagram; a plurality of signal features in the binarized gray scale constellation diagram are extracted, feature fusion is carried out on the plurality of signal features, fusion features of the modulation signals are obtained, and the signal features comprise at least one of fine-grained difference box dimension features, information dimension features and gray scale co-occurrence matrix features; and analyzing the fusion feature of the modulation signal by using the classification model to obtain a modulation format of the modulation signal. According to the method and the device, the technical problem that the accuracy of the identification result is relatively poor due to the fact that the modulation format of the signal is obtained by analyzing the single-dimensional signal features in the related technology is solved.

Description

Technical Field

[0001] This application relates to the field of satellite communication technology, and more specifically, to a method and apparatus for identifying the modulation format of a modulation signal. Background Technology

[0002] With the rapid development of 6G, cloud computing, and the Internet of Things, society's demand for bandwidth and spectrum utilization is constantly increasing. Satellite communication systems are expanding their channel capacity at an unprecedented rate to cope with the explosive growth in data traffic. In the future, satellite communication systems will evolve into a flexible and versatile platform capable of carrying diverse signals and equipped with various modulation formats, aiming to serve a wide range of users and diverse application scenarios. Therefore, digital receivers play a crucial role. They not only need to receive signals, but more importantly, they must be able to accurately determine the modulation format of these signals. This is because different modulation formats encode information in different ways, directly determining the strategy and efficiency of subsequent demodulation processing.

[0003] Currently, modulation format identification methods mainly rely on feature extraction techniques such as High-Order Cumulants (HOC), Histogram of Oriented Gradients (HOG), and Box-Counting Fractal Dimension (BC-FD) to extract single-dimensional features. These features are then analyzed using classification models to determine the signal's modulation format. However, in complex channel environments, especially with low signal-to-noise ratios (e.g., SNR ≤ 5dB) or channel attenuation (e.g., severe weather conditions, satellite-to-ground link losses), relying on single-dimensional features for modulation format identification fails to comprehensively characterize the signal. Furthermore, the features extracted by these techniques cannot effectively and stably distinguish high-order modulation signals, leading to frequent identification errors and limiting the accuracy of demodulation.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides a method and apparatus for identifying the modulation format of a modulated signal, which at least solves the technical problem that related technologies obtain the modulation format of a signal by analyzing the signal features of a single dimension, resulting in poor accuracy of the identification results.

[0006] According to one aspect of the embodiments of this application, a method for identifying the modulation format of a modulation signal is provided, comprising: acquiring a modulation signal transmitted via a satellite-to-ground link; mapping the modulation signal into a binary grayscale constellation diagram; extracting multiple signal features from the binary grayscale constellation diagram and fusing the multiple signal features to obtain fused features of the modulation signal, wherein the signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and gray-level co-occurrence matrix features; and analyzing the fused features of the modulation signal using a preset classification model to obtain the modulation format of the modulation signal.

[0007] Optionally, mapping the modulated signal to a binary grayscale constellation diagram includes: performing a first preprocessing operation on the modulated signal, wherein the first preprocessing operation includes at least one of the following: filtering processing and signal synchronization processing; performing down-conversion processing on the preprocessed modulated signal to obtain the real and imaginary parts of the corresponding baseband signal; sampling the real and imaginary parts of the baseband signal according to a preset sampling frequency to obtain the real and imaginary parts of multiple sampling points respectively; mapping the real and imaginary parts of the multiple sampling points respectively onto the horizontal and vertical axes of the complex plane to obtain a constellation diagram; and performing a second preprocessing operation on the constellation diagram to obtain a binary grayscale constellation diagram, wherein the second preprocessing operation includes at least: grayscale processing and binarization processing.

[0008] Optionally, multiple signal features are extracted from the binarized grayscale constellation image, including: dividing the binarized grayscale constellation image into multiple sub-images; for each sub-image, determining whether the box containing the maximum grayscale value and the box containing the minimum grayscale value are the same when the sub-image is covered by boxes of different sizes; if the box containing the maximum grayscale value and the box containing the minimum grayscale value are the same when the sub-image is covered by boxes of the target size, determining that the number of boxes when the sub-image is covered by boxes of the target size is 1; if the box containing the maximum grayscale value and the box containing the minimum grayscale value are different when the sub-image is covered by boxes of the target size, determining the difference between the index of the box containing the maximum grayscale value and the index of the box containing the minimum grayscale value, and determining the target size and... The quotient of half the difference between the maximum and minimum gray values ​​within a sub-image is used as the nearest integer not less than the ratio of the difference to the quotient. This is then used as the number of boxes when the sub-image is covered by boxes of the target size. Based on the number of boxes when each sub-image in the binarized gray-scale constellation image is covered by boxes of different sizes, the total number of boxes when the binarized gray-scale constellation image is covered by boxes of different sizes is determined. Based on the logarithm of the total number of boxes when multiple sub-images in the binarized gray-scale constellation image are covered by boxes of different sizes, and the logarithm of the inverse of each size, a first linear regression model is constructed. The least squares method is used to solve the first linear regression model to obtain the first slope of the first best-fit line. This first slope is then used as the fine-grained difference box dimension feature of the binarized gray-scale constellation image.

[0009] Optionally, extracting multiple signal features from the binarized grayscale constellation image further includes: determining the frequency of occurrence of the ratio of the number of pixels in each box to the total number of pixels in the binarized grayscale constellation image when the binarized grayscale constellation image is covered by boxes of different sizes, and determining the information entropy of the corresponding size based on the frequency of occurrence of each ratio; constructing a second linear regression model based on the information entropy of different sizes and the logarithmic value of the reciprocal of different sizes, and solving the second linear regression model using the least squares method to obtain the second slope of the second best-fit line, and using the second slope as the information dimension feature of the binarized grayscale constellation image.

[0010] Optionally, extracting multiple signal features from the binarized grayscale constellation image further includes: determining the grayscale co-occurrence matrix corresponding to the binarized grayscale constellation image; analyzing the grayscale co-occurrence matrix to obtain multiple texture features of the grayscale co-occurrence matrix, wherein the texture features include at least: second-order angular moment, contrast, correlation, and entropy; calculating the mean and variance of the multiple texture features, and using the mean and variance as grayscale co-occurrence matrix features of the binarized grayscale constellation image.

[0011] Optionally, determining the gray-level co-occurrence matrix corresponding to the binary gray-level constellation image includes: traversing each pixel in the binary gray-level constellation image, determining multiple neighboring pixels of the current pixel at a preset distance and a preset direction, and forming multiple gray-level value pairs by the current pixel and each neighboring pixel; constructing a gray-level co-occurrence matrix with each gray-level value as the row and column, and the occurrence frequency of each gray-level value pair as the matrix element.

[0012] Optionally, the gray-level co-occurrence matrix is ​​analyzed to obtain multiple texture features of the gray-level co-occurrence matrix, including: calculating the second angular moments based on the gray-level co-occurrence matrix according to the following formulas: Based on the gray-level co-occurrence matrix, the contrast ratio is calculated according to the following formula: Based on the gray-level co-occurrence matrix, the correlation is calculated according to the following formula: Based on the gray-level co-occurrence matrix, the entropy is calculated according to the following formula: In the above formula, i and j represent the row and column numbers of the gray-level co-occurrence matrix, respectively, and , This represents the matrix element in the i-th row and j-th column of the gray-level co-occurrence matrix. , , , , where log represents the logarithmic function.

[0013] According to another aspect of the embodiments of this application, a modulation format identification device for a modulation signal is also provided, comprising: an acquisition module for acquiring a modulation signal transmitted via a satellite-to-ground link; an image processing module for mapping the modulation signal into a binary grayscale constellation diagram; a feature processing module for extracting multiple signal features from the binary grayscale constellation diagram and performing feature fusion on the multiple signal features to obtain fused features of the modulation signal, wherein the signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and gray-level co-occurrence matrix features; and an identification model for analyzing the fused features of the modulation signal using a preset classification model to obtain the modulation format of the modulation signal.

[0014] According to another aspect of the embodiments of this application, a computer program product is also provided, the computer program product comprising: a computer program, wherein when the computer program is executed by a processor, it implements the above-described modulation format recognition method for modulation signals.

[0015] According to another aspect of the embodiments of this application, an electronic device is also provided, the electronic device including: a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the modulation format identification method of the modulation signal described above through the computer program.

[0016] In this embodiment, by mapping the modulated signal to a binary grayscale constellation diagram, multiple signal features such as fine-grained differential box dimension features, information dimension features, and gray-level co-occurrence matrix features are extracted and fused to obtain fused features. Finally, a classification model is used to analyze the fused features to determine the modulation format of the modulated signal. This achieves the technical effect of high-precision identification of the modulation format of satellite single-modulation signals under low signal-to-noise ratio and channel attenuation conditions, thereby improving the demodulation performance and robustness of satellite communication systems. It effectively solves the technical problem that related technologies obtain the modulation format of signals by analyzing single-dimensional signal features, resulting in poor accuracy of the identification results. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 This is a flowchart illustrating an optional modulation format identification method for a modulation signal according to an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of an optional constellation diagram according to an embodiment of this application;

[0020] Figure 3This is a schematic diagram of a constellation of modulated signals in 8PSK format at different optional signal-to-noise ratios according to an embodiment of this application;

[0021] Figure 4 This is a schematic diagram of a constellation diagram of an 8PSK format modulated signal under different optional loss parameters and weather conditions according to an embodiment of this application;

[0022] Figure 5 This is a schematic diagram of the structure of an optional modulation format identification device for a modulation signal according to an embodiment of this application;

[0023] Figure 6 This is a hardware structure block diagram of an electronic device for implementing a modulation format identification method for a modulated signal, according to an embodiment of this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0026] To better understand the embodiments of this application, the following is a translation and explanation of some nouns or terms that appear in the description of the embodiments of this application:

[0027] Constellation diagrams are visual representations of signals and their relationships in digital communication. They are implemented by representing digital signals on a complex plane. Based on the complex form of digital signals, they are decomposed into in-phase (I) and quadrature (Q) components through an equivalent low-pass signal, corresponding to the amplitude modulation of cosine and sine carriers, respectively. In digital modulation, the constellation diagram maps the data bitstream to constellation points on the complex plane using the IQ coordinate system. In PSK (Phase Shift Keying) format, constellation points are distributed on the unit circle, while QAM (Quadrature Amplitude Modulation) format includes amplitude and phase variations to achieve higher spectral efficiency. Gray code encoding ensures that adjacent constellation points differ by only 1 bit, reducing the bit error rate.

[0028] Fractal dimension (FD) is a key feature in fractal geometry used to estimate the surface roughness of an entire image. It is the fractal dimension of a bounded set B that exhibits self-similarity in geometric space. It can be represented as: In the formula, r represents the length of one side of the box. Let r represent the number of boxes that cover the bounded set B. Therefore, the fractal dimension mentioned above means determining the box dimension of set B by covering the bounded set B with boxes of side length r.

[0029] Ordinary Least Squares (ORS): also known as the least squares method, is a mathematical optimization technique that finds the best function match for data by minimizing the sum of squared errors. It is used for parameter estimation and curve fitting.

[0030] Example 1

[0031] According to an embodiment of this application, a method for identifying the modulation format of a modulation signal is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0032] Figure 1 This is a flowchart illustrating a modulation format identification method for a modulated signal according to an embodiment of this application, as shown below. Figure 1 As shown, the method includes the following steps:

[0033] Step S102: Obtain the modulation signal transmitted via the satellite-to-ground link.

[0034] Step S104: Map the modulated signal into a binary grayscale constellation diagram.

[0035] Step S106: Extract multiple signal features from the binarized grayscale constellation image and perform feature fusion on the multiple signal features to obtain the fused features of the modulated signal. The signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features.

[0036] Step S108: Analyze the fusion characteristics of the modulated signal using a preset classification model to obtain the modulation format of the modulated signal.

[0037] Based on the scheme defined in steps S102 to S108 above, it can be understood that in this embodiment of the application, by mapping the modulated signal to a binary grayscale constellation diagram, and then extracting multiple signal features such as fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features, and performing feature fusion to obtain fused features, and finally using a preset classification model to analyze the fused features to determine the modulation format of the modulated signal, the technical effect of high-precision identification of the modulation format of satellite single modulation signals under low signal-to-noise ratio and channel attenuation conditions is achieved, thereby improving the demodulation performance and robustness of the satellite communication system.

[0038] The following describes each step of the modulation format identification method for modulated signals in conjunction with a specific implementation process.

[0039] In the technical solution provided in step S102 above, the digital receiver (also known as the "signal receiver") can receive radio frequency signals transmitted from the source ground station and relayed via satellite, including but not limited to various satellite signals received through the downlink of the satellite communication system. This signal may experience different channel conditions during transmission, such as Gaussian white noise, satellite-to-ground link loss, and attenuation caused by weather factors, leading to changes in signal quality.

[0040] In the technical solution provided in step S104 above, the digital receiver can map the modulated signal into a binary grayscale constellation diagram to transform the signal processing problem into an image analysis problem, which is convenient for subsequent processing and analysis.

[0041] As an optional implementation, in the technical solution provided in step S104 above, the method may include:

[0042] Step S1041: Perform the first preprocessing operation on the modulated signal.

[0043] Specifically, the first preprocessing operation described above aims to improve signal quality and ensure the accuracy of subsequent processing. Therefore, the first preprocessing operation includes, but is not limited to, filtering and signal synchronization. Specifically, filtering is used to remove noise and interference from the modulated signal, such as using a low-pass filter to remove high-frequency noise, ensuring signal purity; signal synchronization is used to synchronize the carrier and symbols of the modulated signal. Carrier synchronization involves extracting the carrier phase information from the received modulated signal to align the local oscillator of the digital receiver with the carrier phase of the modulated signal, while symbol synchronization determines the start and end times of symbols to correctly extract symbol information.

[0044] Step S1042: Perform down-conversion processing on the preprocessed modulation signal to obtain the real and imaginary parts of the corresponding baseband signal.

[0045] Specifically, downconversion is the process in a communication system that converts a modulated signal into a baseband signal. This process involves multiplying the received modulated signal with a signal generated by the local oscillator of the digital receiver to reduce the frequency of the modulated signal from radio frequency to baseband.

[0046] The following are the detailed steps and principles of downconversion processing:

[0047] If the expression of the preprocessed modulated signal is written as: In the formula, Indicates the amplitude of the modulated signal. This represents the carrier angular frequency of the modulated signal, and , Indicates the carrier frequency of the modulated signal. This indicates the phase of the modulated signal.

[0048] The signal generated by the local oscillator of a digital receiver is typically a sine or cosine wave with the same carrier frequency as the modulated signal, and its expression can be written as: or, .

[0049] Therefore, the modulated signal Signal generated by the local oscillator Multiplying them together yields the down-converted signal: ,in:

[0050] (1) If a cosine wave is used as the signal generated by the local oscillator, then the down-converted signal can be written as follows: Furthermore, it can be written as the following formula: .

[0051] (2) If a sine wave is used as the signal generated by the local oscillator, then the down-converted signal can be written as follows: Furthermore, it can be written as the following formula: .

[0052] Therefore, the down-converted signal includes two frequency components: a high-frequency component. or Its frequency is Low-frequency components or Its frequency is 0 (i.e., baseband signal).

[0053] To extract the baseband signal, a low-pass filter can be used to remove high-frequency components. Therefore, the cutoff frequency of the low-pass filter should be lower than [the specified frequency]. Then, the baseband signal obtained after passing through the low-pass filter is: ,or .

[0054] The baseband signal described above can be represented in complex form as follows: In the formula, This represents the amplitude (i.e., the real part) of the baseband signal, which corresponds to the in-phase component (I) of the modulation signal. It represents the phase (i.e., imaginary part) of the baseband signal, which corresponds to the quadrature component (O) of the modulation signal.

[0055] Step S1043: Sample the real and imaginary parts of the baseband signal according to the preset sampling frequency to obtain the real and imaginary parts of multiple sampling points.

[0056] Specifically, before sampling the baseband signal, it is necessary to determine the symbol rate (i.e., the number of symbols transmitted per second) and the synchronization time, where:

[0057] Regarding the symbol rate, it can be calculated by taking the autocorrelation function of the baseband signal and observing the distance between the main peaks of the autocorrelation function, and using that distance as the symbol rate; or, by analyzing the spectrum of the baseband signal, taking the position of the first zero in the spectrum as the symbol rate of the baseband signal.

[0058] Regarding synchronization time, the synchronization time of baseband signals can be analyzed using methods such as early-late gating algorithms and phase-locked loops.

[0059] Therefore, the reciprocal of the symbol rate is used as the sampling period, and uniform sampling is performed at the center point of each sampling period to obtain the real and imaginary parts of multiple sampling points.

[0060] In step S1044, the real and imaginary parts of each of the multiple sampling points are mapped onto the horizontal axis (i.e., the X-axis) and vertical axis (i.e., the Y-axis) of the complex plane respectively to obtain the constellation diagram.

[0061] Specifically, the constellation diagrams of modulated signals with different modulation formats are different, and the number of clustered regions of vector points in the constellation diagrams varies with the modulation order. For example, Figure 2 This is a schematic diagram of an optional constellation diagram according to an embodiment of this application, such as... Figure 2 As shown, Figure 2 The first image (a) in the first row from left to right is in BPSK (Binary Phase Shift Keying) format, whose constellation diagram consists of two points with opposite phases located on the I-axis of the complex plane, involving only two phase states; the second image (b) is in QPSK (Quadrature Phase Shift Keying) format, whose constellation diagram consists of four points evenly distributed across the four quadrants of the unit circle, forming a square; the third image (c) is in 8PSK (8-Phase Shift Keying) format, whose constellation diagram consists of eight points evenly distributed across the circumference of the unit circle, forming an octagonal pattern; the first image (d) and the second image (e) in the second row from left to right are both in QAM (Quadrature Amplitude) format. The first figure (d) is a 16QAM modulation, and its constellation diagram consists of 16 evenly distributed points forming a 4×4 rectangular grid. The second figure (e) is a 64QAM modulation, and its constellation diagram consists of 64 points forming an 8×8 rectangular grid. Therefore, the distribution of points in the constellation diagram of the QAM format depends on the modulation order. The higher the order, the more points there are on the constellation diagram, and the denser the distribution.

[0062] When the modulated signal is transmitted through the satellite-to-ground link, its vector point distribution in the constellation diagram changes with the variation in signal loss. For example, Figure 3 This is a schematic diagram of a constellation of modulated signals in 8PSK format at different optional signal-to-noise ratios according to an embodiment of this application, such as... Figure 3 As shown, Figure 3 The three figures from left to right are constellation diagrams of the 8PSK modulated signal at 0dB, 10dB, and 20dB. It's easy to see that as the signal-to-noise ratio (SNR) gradually decreases, the vector points of the 8PSK modulated signal gradually disperse; when the SNR is 0, the vector points all cluster in the same area. Furthermore, Figure 4 This is a schematic diagram of a constellation diagram of an 8PSK format modulated signal under optional loss parameters and weather conditions according to an embodiment of this application, such as... Figure 4 As shown, with changes in weather (such as dark clouds and thunderclouds) and loss parameters (such as m=5 and m=2) in the satellite-to-ground link, the vector points in the constellation diagram will shift, causing vector points in different regions to intersect.

[0063] Step S1045: Perform a second preprocessing operation on the constellation image to obtain a binary grayscale constellation image.

[0064] The second preprocessing operation includes at least two steps: grayscale processing and binarization. Grayscale processing converts the constellation diagram into a grayscale image, facilitating subsequent image feature extraction algorithms. Binarization further simplifies the image by converting it into a binary image with only black and white grayscale levels, which highlights the distribution characteristics of signal points while reducing computational complexity.

[0065] In the above embodiments, by processing the modulated signal transmitted via the satellite-to-ground link to convert it into a processable binary grayscale constellation diagram, it not only helps to extract signal features through image processing algorithms, but also improves the robustness and efficiency of the recognition algorithm under various conditions, while simplifying the requirements for computing resources.

[0066] In the technical solution provided in step S106 above, the digital receiver can extract multiple signal features from the binarized grayscale constellation diagram. These signal features include, but are not limited to: fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features.

[0067] As an optional implementation, in the technical solution provided in step S106 above, the fine-grained difference box dimension features in the binarized grayscale constellation image can be extracted according to the following method:

[0068] Step 1: Divide the binarized grayscale constellation image into multiple sub-images. The division criterion is that any two adjacent sub-images overlap only at the same pixel on their common boundary. For example, if the resolution is M... The binary grayscale constellation image of M is divided into segments of size m. If there are multiple subgraphs of m, then the number of overlapping pixels between any two adjacent subgraphs is m.

[0069] Step 2: For each subgraph, repeat the following steps in a loop:

[0070] First, determine whether the box containing the maximum gray value (whose box number can be denoted as i) and the box containing the minimum gray value (whose box number can be denoted as j) are the same when covering the subimage with boxes of different sizes.

[0071] If, when a subimage is covered by boxes of the target size, the boxes containing the maximum and minimum grayscale values ​​within the subimage are the same, then the number of boxes covered by the subimage according to the target size is determined to be 1.

[0072] If, when the subimage is covered by boxes of the target size, the boxes containing the maximum grayscale value and the boxes containing the minimum grayscale value are different, then the difference between the index of the box containing the maximum grayscale value and the index of the box containing the minimum grayscale value (i.e., ij) is determined. Then, the quotient of the target size and half the difference between the maximum and minimum grayscale values ​​in the subimage (i.e., m / G, where G represents half the difference between the maximum and minimum grayscale values ​​in the subimage) is determined. Finally, the nearest integer not less than the ratio of the difference to the quotient is used as the number of boxes when the subimage is covered by boxes of the target size.

[0073] Step 3: Based on the number of boxes when each sub-image in the binarized grayscale constellation image is covered by boxes of different sizes, determine the total number of boxes when the binarized grayscale constellation image is covered by boxes of different sizes.

[0074] Step 4: Based on the logarithm of the total number of boxes when multiple sub-images within the binary grayscale constellation image are covered by boxes of different sizes, and the logarithm of the inverse of different sizes, construct the first linear regression model, and solve the first linear regression model using the least squares method to obtain the first slope of the first best-fit line (that is, the rate at which the number of boxes increases as the size of the boxes decreases), and use the first slope as the fine-grained difference box dimension feature of the binary grayscale constellation image.

[0075] The first linear regression model mentioned above can be expressed as: In the formula, r represents the size of the box. The log represents the total number of boxes when the binary grayscale constellation image is covered by boxes of size r, and the log is the common logarithm with base 10. Therefore, the first slope of the first best-fit line of the first linear regression model is mathematically related to the fractal dimension of the image, which quantifies the characteristic of image complexity changing with scale.

[0076] In the above embodiments, based on the concept of box dimension in fractal geometry, the box-covering method is used to evaluate the complexity and self-similarity of the binarized grayscale constellation image. Under low signal-to-noise ratio or channel attenuation conditions, the point set distribution of the constellation image may become non-uniform and complex, while fine-grained differential box-dimensional features can accurately quantify this complexity through local analysis. The choice of box size and the covering strategy directly affect the feature extraction accuracy. The box-covering method can capture the non-uniform details inside the constellation image, thereby providing richer information for subsequent modulation format recognition.

[0077] As an optional implementation, in the technical solution provided in step S106 above, the information dimension features of the binarized grayscale constellation image can be extracted according to the following method:

[0078] Step 1: Determine the frequency of occurrence of the ratio of the number of pixels in each box to the total number of pixels in the binary grayscale constellation image when the binary grayscale constellation image is covered by boxes of different sizes, and determine the information entropy of the corresponding size based on the frequency of occurrence of each ratio.

[0079] The second step is to construct a second linear regression model based on the information entropy of different sizes and the logarithmic value of the reciprocal of different sizes, and solve the second linear regression model using the least squares method to obtain the second slope of the second best-fit line (i.e., the change of image complexity as the box size decreases, which quantifies the sensitivity of image complexity to scale), and use the second slope as the information dimension feature of the binary grayscale constellation image.

[0080] The second linear regression model mentioned above can be expressed as: In the formula, r represents the size of the box, and i represents the frequency of the i-th ratio. Let ln represent the frequency of occurrence of the i-th ratio of the number of pixels in the box to the total number of pixels in the binary grayscale constellation image when the box of size r is used for coverage. ln represents the natural logarithm, and its base is the irrational number e, with a value of approximately 2.71828.

[0081] In the above embodiments, considering that the details and local changes of the constellation diagram may be difficult to identify directly under low signal-to-noise ratio or channel attenuation conditions, the distribution complexity of signal points in the constellation diagram is quantified by information entropy based on information theory and fractal geometry.

[0082] As an optional implementation, in the technical solution provided in step S106 above, the gray-level co-occurrence matrix features in the binary gray-level constellation image can be extracted according to the following method:

[0083] Step 1: Determine the gray-level co-occurrence matrix corresponding to the binarized gray-level constellation image.

[0084] Optionally, in the technical solution provided in the first step above, the method for constructing the gray-level co-occurrence matrix may include: traversing each pixel in the binary gray-level constellation map, determining multiple neighboring pixels of the current pixel at a preset distance and a preset direction, and forming multiple gray-level value pairs by the current pixel and each neighboring pixel; constructing the gray-level co-occurrence matrix by using each gray-level value as the row and column, and using the occurrence frequency of each gray-level value pair as the matrix element.

[0085] Therefore, the expression for the matrix elements of the gray-level co-occurrence matrix can be written as: In the formula, Represents the coordinates of pixels within a binary grayscale constellation image; Represents grayscale value, and The possible values ​​include the following combinations: (0,0), (0,1), (1,0), and (1,1); These represent the position offset; d represents the preset generation step size. This indicates the preset generation direction, which can be 0°, 45°, 90°, or 135°. Additionally, to facilitate subsequent processing and calculations, the elements within the gray-level co-occurrence matrix can be normalized to convert the element values ​​into probabilistic forms, i.e., the probability of gray-level value pairs appearing.

[0086] Step 2: Analyze the gray-level co-occurrence matrix to obtain multiple texture features of the gray-level co-occurrence matrix, including:

[0087] (1) Angular Second Moment: also known as energy, it is a measure of the uniformity of gray-level distribution and texture coarseness in an image. Therefore, the larger the value of the angular second moment, the more uniform the distribution of gray-level values ​​in the image.

[0088] (2) Contrast: It is used to measure the difference in gray values ​​between adjacent pixels in an image. Therefore, the larger the contrast value, the greater the gray value difference between adjacent pixels in the image, and the higher the contrast of the image.

[0089] (3) Correlation: used to reflect the linear dependence of pixel gray values ​​in an image.

[0090] (4) Entropy: An important indicator used to measure the randomness or uncertainty of the distribution of gray values ​​in an image. It reflects the complexity of the image. Therefore, the larger the entropy value, the more random the distribution of gray values ​​in the image and the higher the uncertainty.

[0091] Optionally, in the technical solution provided in the second step above, the method may include:

[0092] Based on the gray-level co-occurrence matrix, calculate the second angular moments using the following formulas: ;

[0093] Based on the gray-level co-occurrence matrix, the contrast ratio is calculated using the following formula: ;

[0094] Based on the gray-level co-occurrence matrix, the correlation is calculated using the following formula: ;

[0095] Based on the gray-level co-occurrence matrix, the entropy is calculated according to the following formula: .

[0096] In the above formula, i and j represent the row and column numbers of the gray-level co-occurrence matrix, respectively, and , This represents the matrix element in the i-th row and j-th column of the gray-level co-occurrence matrix. , , , , where log represents the logarithmic function.

[0097] In addition, the texture features mentioned above may also include: the inverse differential moment (IDM) matrix, which reflects the clarity and regularity of the image texture. Generally, the clearer, more regular, and easier to describe the image texture, the larger the value of the corresponding inverse differential moment matrix.

[0098] Step 3: Calculate the mean and variance of multiple texture features, and use the mean and variance as the gray-level co-occurrence matrix features of the binary gray-level constellation map.

[0099] Specifically, since the texture features of second moment, contrast, correlation, and entropy focus on different aspects of an image—namely, the uniformity of pixel values, the intensity of grayscale value changes, the linear relationship between grayscale values, and the randomness of grayscale value distribution—this embodiment of the application can comprehensively reflect the average condition and trend of an image across these different dimensions by calculating the mean and variance of these texture features, providing a richer and more comprehensive image description.

[0100] Furthermore, in the technical solution provided in step S106 above, the digital receiver can fuse multiple extracted signal features to obtain the fused features of the modulated signal.

[0101] Specifically, the feature fusion process may include the following steps:

[0102] Step 1: Standardize the extracted signal features to ensure all features are on the same scale, preventing any one feature from dominating the fusion result due to its large numerical range. Standardization methods include, but are not limited to, min-max scaling and Z-score standardization.

[0103] The second step is to stitch together the standardized signal features in a certain order to form a longer fused feature. For example, if the feature vector length of the fine-grained difference box dimension feature is 64, the feature vector length of the information dimension feature is also 64, and the feature vector length of the gray-level co-occurrence matrix feature is 80, then the length of the fused feature vector will be 208.

[0104] It should be noted that during the fusion process, different types of signal features can also be weighted to reflect their relative importance in the signal recognition task. This application does not impose specific restrictions on this.

[0105] In the technical solution provided in step S108 above, the digital receiver can use a preset classification model to analyze the fusion characteristics of the modulated signal to obtain the modulation format of the modulated signal. The modulation format includes, but is not limited to, BPSK, QPSK, 8PSK, QAM and other formats.

[0106] Specifically, the aforementioned classification model can be obtained by iteratively training a pre-defined classifier using multiple sets of pre-defined training sample data. Each set of training sample data can include: fusion features of the demodulated modulated signal as training samples and modulation format of the demodulated modulated signal as sample labels; and the architecture of the classifier can be a support vector machine.

[0107] Therefore, during the training of the classifier using multiple sets of training sample data, the classifier continuously attempts to find the optimal hyperplane that maximizes the boundary between data points of different modulation formats. This means that the classifier learns how to distinguish different modulation formats based on the fused features of the input. Furthermore, cross-validation is used during training to evaluate and adjust the model's performance. This typically involves splitting the multiple sets of training sample data multiple times, using a portion as a temporary validation set and the remainder as the training set. This helps prevent overfitting and ensures the classifier's performance on unknown data.

[0108] Furthermore, during the training process of the aforementioned classification model, the classifier can be loaded into memory. For example, the raw data of the classifier can be loaded from non-volatile memory into volatile memory so that the processor can run the classifier. The raw data of the classifier refers to unprocessed data, which typically includes the classifier's parameters and structural data. The structural data can be the computational relationships based on the parameters, such as the forward propagation computational relationships between intermediate layers and between neurons. Specifically, the structural data can include the structure-related code of the classifier, such as the code used to perform related calculations between intermediate layers and between neurons.

[0109] In one implementation, a region can be partitioned in memory for loading the classifier, which may include a structure data storage area and a parameter storage area. The structure data storage area stores structure-related code, and the parameters referenced by it can be pointed to by pointers to the addresses of specific parameters in the parameter storage area. During the training of the classifier, it may be necessary to frequently update the parameters, which can be done by updating the parameter values ​​in the parameter storage area.

[0110] In the modulation format identification method described above, by extracting multi-dimensional features such as fine-grained differential box dimension features, information dimension features, and gray-level co-occurrence matrix features from the binarized grayscale constellation diagram, a more comprehensive and refined signal description can be provided. Furthermore, through multi-feature fusion and classifier training, not only can common modulation formats such as BPSK, QPSK, 8PSK, 16QAM, and 64QAM be accurately identified, but a foundation is also laid for identifying a wider range of signal modulation formats. This better adapts to new modulation technologies that may emerge in future satellite communication systems, maintaining the technology's forward-looking nature.

[0111] Example 2

[0112] According to an embodiment of this application, a modulation format identification device for a modulation signal is also provided for implementing the modulation format identification method for the modulation signal in Embodiment 1, such as... Figure 5 As shown, the modulation format recognition device for the modulation signal includes at least: an acquisition module 52, an image processing module 54, a feature processing module 56, and a recognition model 58, wherein:

[0113] Acquisition module 52 is used to acquire the modulation signal transmitted via the satellite-to-ground link.

[0114] Image processing module 54 is used to map the modulated signal into a binary grayscale constellation diagram.

[0115] The feature processing module 56 is used to extract multiple signal features from the binary grayscale constellation map and perform feature fusion on the multiple signal features to obtain the fused features of the modulated signal. The signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features.

[0116] Identification model 58 is used to analyze the fusion characteristics of the modulated signal using a preset classification model to obtain the modulation format of the modulated signal.

[0117] The following section describes the functions of each module of the modulation format recognition device for modulation signals, based on a specific implementation process.

[0118] Optionally, the image processing module is further configured to map the modulated signal into a binary grayscale constellation diagram according to the following method: performing a first preprocessing operation on the modulated signal, wherein the first preprocessing operation includes at least one of the following: filtering processing, signal synchronization processing; performing down-conversion processing on the preprocessed modulated signal to obtain the real and imaginary parts of the corresponding baseband signal; sampling the real and imaginary parts of the baseband signal according to a preset sampling frequency to obtain the real and imaginary parts of multiple sampling points respectively; mapping the real and imaginary parts of the multiple sampling points respectively onto the horizontal and vertical axes of the complex plane to obtain a constellation diagram; performing a second preprocessing operation on the constellation diagram to obtain a binary grayscale constellation diagram, wherein the second preprocessing operation includes at least: grayscale processing, binarization processing.

[0119] Optionally, the feature processing module is further configured to extract fine-grained difference box-dimensional features from the binarized grayscale constellation image as follows: The binarized grayscale constellation image is divided into multiple sub-images; for each sub-image, it is determined whether the box containing the maximum grayscale value and the box containing the minimum grayscale value are the same when the sub-image is covered with boxes of different sizes; if the boxes containing the maximum grayscale value and the minimum grayscale value are the same when the sub-image is covered with boxes of the target size, the number of boxes covered by the target size is determined to be 1; if the boxes containing the maximum grayscale value and the minimum grayscale value are different when the sub-image is covered by boxes of the target size, the difference between the index of the box containing the maximum grayscale value and the index of the box containing the minimum grayscale value is determined. The quotient of the target size and half the difference between the maximum and minimum grayscale values ​​within the sub-image is determined. The nearest integer not less than the ratio of the difference to the quotient is then used as the number of boxes when the sub-image is covered by boxes of the target size. Based on the number of boxes when each sub-image in the binarized grayscale constellation image is covered by boxes of different sizes, the total number of boxes when the binarized grayscale constellation image is covered by boxes of different sizes is determined. Based on the logarithm of the total number of boxes when multiple sub-images in the binarized grayscale constellation image are covered by boxes of different sizes, and the logarithm of the reciprocal of different sizes, a first linear regression model is constructed. The least squares method is used to solve the first linear regression model to obtain the first slope of the first best-fit line, and the first slope is used as the fine-grained difference box dimension feature of the binarized grayscale constellation image.

[0120] Optionally, the feature processing module is also used to extract the information dimension features of the binary grayscale constellation image as follows: determine the frequency of occurrence of the ratio of the number of pixels in each box to the total number of pixels in the binary grayscale constellation image when the binary grayscale constellation image is covered by boxes of different sizes, and determine the information entropy of the corresponding size based on the frequency of occurrence of each ratio; construct a second linear regression model based on the information entropy of different sizes and the logarithmic value of the reciprocal of different sizes, and solve the second linear regression model using the least squares method to obtain the second slope of the second best-fit line, and use the second slope as the information dimension feature of the binary grayscale constellation image.

[0121] Optionally, the feature processing module is also used to extract gray-level co-occurrence matrix features from the binarized gray-level constellation image as follows: determine the gray-level co-occurrence matrix corresponding to the binarized gray-level constellation image; analyze the gray-level co-occurrence matrix to obtain multiple texture features of the gray-level co-occurrence matrix, wherein the texture features include at least: second moment of angle, contrast, correlation and entropy; calculate the mean and variance of the multiple texture features, and use the mean and variance as gray-level co-occurrence matrix features of the binarized gray-level constellation image.

[0122] Optionally, the feature processing module is also used to determine the gray-level co-occurrence matrix corresponding to the binary gray-level constellation map according to the following method: traversing each pixel in the binary gray-level constellation map, determining multiple neighboring pixels of the current pixel at a preset distance and a preset direction, and forming multiple gray-level value pairs by the current pixel and each neighboring pixel; constructing the gray-level co-occurrence matrix with each gray-level value as the row and column, and the occurrence frequency of each gray-level value pair as the matrix element.

[0123] Optionally, the feature processing module can determine multiple texture features of the gray-level co-occurrence matrix as follows:

[0124] Based on the gray-level co-occurrence matrix, calculate the second angular moments using the following formulas: ;

[0125] Based on the gray-level co-occurrence matrix, the contrast ratio is calculated using the following formula: ;

[0126] Based on the gray-level co-occurrence matrix, the correlation is calculated using the following formula: ;

[0127] Based on the gray-level co-occurrence matrix, the entropy is calculated according to the following formula: .

[0128] In the above formula, i and j represent the row and column numbers of the gray-level co-occurrence matrix, respectively, and , This represents the matrix element in the i-th row and j-th column of the gray-level co-occurrence matrix. , , , , where log represents the logarithmic function.

[0129] It should be noted that each module in the modulation format identification device of the modulation signal in this embodiment corresponds one-to-one with each implementation step of the modulation format identification method of the modulation signal in Embodiment 1. Since Embodiment 1 has been described in detail, some details not shown in this embodiment can be referred to Embodiment 1, and will not be elaborated further here.

[0130] Example 3

[0131] According to an embodiment of this application, a computer program product is also provided, which includes a computer program, wherein when the computer program is executed by a processor, it implements the modulation format recognition method for the modulation signal in Embodiment 1.

[0132] According to an embodiment of this application, a non-volatile storage medium is also provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the modulation format identification method of the modulation signal in Embodiment 1 by running the computer program.

[0133] According to an embodiment of this application, a processor is also provided for running a computer program, wherein the computer program executes the modulation format recognition method for the modulation signal in Embodiment 1.

[0134] According to an embodiment of this application, an electronic device is also provided, comprising: a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the modulation format identification method for the modulation signal in Embodiment 1 through the computer program.

[0135] Specifically, the computer program executes the following steps during runtime: acquiring the modulated signal transmitted via the satellite-to-ground link; mapping the modulated signal into a binary grayscale constellation diagram; extracting multiple signal features from the binary grayscale constellation diagram and fusing these features to obtain the fused features of the modulated signal, wherein the signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features; and analyzing the fused features of the modulated signal using a pre-defined classification model to obtain the modulation format of the modulated signal.

[0136] As an alternative implementation, the above-mentioned electronic device may exist in the form of a mobile terminal, a computer terminal, or a similar computing device. Figure 6 A hardware block diagram of an electronic device for implementing a modulation format identification method for modulated signals is shown. Figure 6As shown, the electronic device 60 may include one or more processors 602 (shown as 602a, 602b, ..., 602n in the figure) 602 (processor 602 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 604 for storing data, and a transmission device 606 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 6 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, electronic device 60 may also include components that are more... Figure 6 The more or fewer components shown, or having the same Figure 6 The different configurations shown.

[0137] It should be noted that the aforementioned one or more processors 602 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element of the electronic device 60. As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0138] The memory 604 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the modulation format identification method of the modulation signal in this embodiment. The processor 602 executes various functional applications and data processing by running the software programs and modules stored in the memory 604, thereby implementing the above-mentioned vulnerability detection method for the application. The memory 604 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 604 may further include memory remotely located relative to the processor 602, and these remote memories can be connected to the electronic device 60 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0139] The transmission device 606 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the electronic device 60. In one example, the transmission device 606 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 606 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0140] The display may be, for example, a touchscreen liquid crystal display (LCD), which allows the user to interact with the user interface of the electronic device 60.

[0141] The sequence numbers of the above embodiments are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0142] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0143] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0144] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0145] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0146] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0147] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for identifying the modulation format of a modulated signal, characterized in that, include: Acquire the modulated signal transmitted via the satellite-to-ground link; The modulated signal is mapped into a binary grayscale constellation diagram; Multiple signal features are extracted from the binarized grayscale constellation image, and feature fusion is performed on the multiple signal features to obtain the fused features of the modulated signal, wherein the signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and gray-level co-occurrence matrix features; The modulation format of the modulation signal is obtained by analyzing the fusion characteristics of the modulation signal using a preset classification model.

2. The method according to claim 1, characterized in that, Mapping the modulated signal into a binary grayscale constellation diagram includes: The modulated signal is subjected to a first preprocessing operation, wherein the first preprocessing operation includes at least one of the following: filtering processing and signal synchronization processing; The preprocessed modulation signal is down-converted to obtain the real and imaginary parts of the corresponding baseband signal; The real and imaginary parts of the baseband signal are sampled according to a preset sampling frequency to obtain the real and imaginary parts of multiple sampling points. The real and imaginary parts of each of the multiple sampling points are mapped onto the horizontal and vertical axes of the complex plane, respectively, to obtain the constellation diagram; The constellation diagram is subjected to a second preprocessing operation to obtain the binarized grayscale constellation diagram, wherein the second preprocessing operation includes at least: grayscale processing and binarization processing.

3. The method according to claim 1, characterized in that, Extracting multiple signal features from the binarized grayscale constellation image, including: The binarized grayscale constellation image is divided into multiple sub-images; For each sub-image, determine whether the box containing the maximum grayscale value and the box containing the minimum grayscale value are the same when the sub-image is covered by boxes of different sizes; if the box containing the maximum grayscale value and the box containing the minimum grayscale value are the same when the sub-image is covered by boxes of the target size, determine that the number of boxes when the sub-image is covered by boxes of the target size is 1; if the box containing the maximum grayscale value and the box containing the minimum grayscale value are different when the sub-image is covered by boxes of the target size, determine the difference between the index of the box containing the maximum grayscale value and the index of the box containing the minimum grayscale value, and determine the quotient of the target size and half of the difference between the maximum and minimum grayscale values ​​in the sub-image, and then take the nearest integer not less than the ratio of the difference to the quotient as the number of boxes when the sub-image is covered by boxes of the target size. Based on the number of boxes when each sub-image in the binarized grayscale constellation image is covered by boxes of different sizes, determine the total number of boxes when the binarized grayscale constellation image is covered by boxes of different sizes. A first linear regression model is constructed based on the logarithm of the total number of boxes when multiple sub-images within the binary grayscale constellation image are covered by boxes of different sizes, and the logarithm of the inverse of different sizes. The first linear regression model is solved using the least squares method to obtain the first slope of the first best-fit line, and the first slope is used as the fine-grained difference box dimension feature of the binary grayscale constellation image.

4. The method according to claim 1, characterized in that, Extracting multiple signal features from the binarized grayscale constellation image also includes: When the binarized grayscale constellation image is covered by boxes of different sizes, determine the frequency of occurrence of the ratio of the number of pixels in each box to the total number of pixels in the binarized grayscale constellation image, and determine the information entropy of the corresponding size based on the frequency of occurrence of each ratio. A second linear regression model is constructed based on the information entropy of different sizes and the logarithmic value of the reciprocal of different sizes. The second linear regression model is solved by the least squares method to obtain the second slope of the second best-fit line. The second slope is used as the information dimension feature of the binary grayscale constellation image.

5. The method according to claim 1, characterized in that, Extracting multiple signal features from the binarized grayscale constellation image also includes: Determine the gray-level co-occurrence matrix corresponding to the binarized gray-level constellation image; The gray-level co-occurrence matrix is ​​analyzed to obtain multiple texture features of the gray-level co-occurrence matrix, wherein the texture features include at least: second angular moment, contrast, correlation and entropy; Calculate the mean and variance of multiple texture features, and use the mean and variance as the gray-level co-occurrence matrix features of the binarized gray-level constellation map.

6. The method according to claim 5, characterized in that, Determining the gray-level co-occurrence matrix corresponding to the binarized gray-level constellation image includes: Traverse each pixel in the binary grayscale constellation image, determine multiple neighboring pixels of the current pixel at a preset distance and a preset direction, and form multiple grayscale value pairs with the current pixel and each of the neighboring pixels; The gray-level co-occurrence matrix is ​​constructed by using each gray-level value as a row and column, and using the frequency of occurrence of each gray-level value pair as a matrix element.

7. The method according to claim 5, characterized in that, Analyzing the gray-level co-occurrence matrix yields multiple texture features, including: Based on the gray-level co-occurrence matrix, the second moment of the angle is calculated according to the following formulas: ; The contrast ratio is calculated based on the gray-level co-occurrence matrix and according to the following formula: ; Based on the gray-level co-occurrence matrix, the correlation is calculated according to the following formula: ; Based on the gray-level co-occurrence matrix, the entropy is calculated according to the following formula: ; In the above formula, i and j represent the row number and column number of the gray-level co-occurrence matrix, respectively, and , This represents the matrix element in the i-th row and j-th column of the gray-level co-occurrence matrix. , , , , where log represents the logarithmic function.

8. A modulation format identification device for a modulation signal, characterized in that, include: The acquisition module is used to acquire the modulated signal transmitted via the satellite-to-ground link; The image processing module is used to map the modulated signal into a binary grayscale constellation diagram; The feature processing module is used to extract multiple signal features from the binarized grayscale constellation diagram and perform feature fusion on the multiple signal features to obtain the fused features of the modulated signal, wherein the signal features include at least one of the following: fine-grained differential box dimension features, information dimension features, and grayscale co-occurrence matrix features; The identification model is used to analyze the fusion features of the modulated signal using a preset classification model to obtain the modulation format of the modulated signal.

9. A computer program product, characterized in that, include: A computer program, wherein when executed by a processor, the computer program implements the modulation format identification method for the modulation signal according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: A memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute, via the computer program, the modulation format identification method for the modulation signal according to any one of claims 1 to 7.

Images