Model training and signal detection method and device, medium and electronic equipment

Abstract

The embodiment of the invention discloses a model training and signal detection method and device, a medium and electronic equipment, and relates to the technical field of signal processing, a signal detection model based on deep learning is trained to realize signal detection, detection is carried out through image recognition to avoid the influence of an electromagnetic environment and signal noise, and the detection accuracy is improved. By collecting data describing signal I / Q, that is, in-phase component data and orthogonal component data describing signal phase and amplitude, and performing multi-frame segmentation on the data, each framed signal is approximate to a stable signal, so that spectral analysis is facilitated, and the accuracy of signal detection and identification is improved. The method comprises the following steps: firstly, carrying out linear mapping on a frequency domain, then converting the frequency domain data into frequency domain data by utilizing fast Fourier transform, sequentially storing the frequency domain data into a buffer area, obtaining a data matrix, generating a spectrogram by utilizing the data matrix, training the spectrogram obtained under linear mapping, and marking signal information on the spectrogram. The signal detection model obtained through training can extract image features and realize rapid and accurate recognition of signals based on images in an application stage, and after the features of the signals are recognized on the images, the features are mapped into parameters of the signals, so that detection and recognition of the signals can be realized.

Description

Technical Field

[0001] The present application relates to the field of signal processing technology, and in particular to a model training and signal detection method, device, medium and electronic equipment. Background Art

[0002] Radar, short for "Radio Detection and Ranging", is an electronic device that uses electromagnetic waves to detect targets. It emits electromagnetic waves to illuminate the target and receives its echo, thereby obtaining information such as the distance from the target to the electromagnetic wave emission point, the rate of change of distance (radial velocity), azimuth, and altitude. The typical frequency band occupied by radar signals is from 500 MHz to 18 GHz. It is a technical signal that uses electromagnetic waves to detect and track targets. Traditional radar signal analysis is usually based on intra-pulse modulation recognition algorithms in the time domain and frequency domain to identify radar signals and estimate parameters. It can solve most intra-pulse modulation recognition problems, but the accuracy of detection still needs to be improved. For example, in complex electromagnetic environments, electromagnetic interference will affect the accuracy of recognition. When the signal-to-noise ratio is low, the accuracy of parameter demodulation will also be reduced. Summary of the invention

[0003] The main purpose of this application is to provide a model training and signal detection method, device, medium and electronic equipment, aiming to solve the problem of reduced accuracy of detection and identification of radar signals in the prior art.

[0004] To achieve the above objectives, the technical solutions adopted in the embodiments of the present application are as follows:

[0005] In a first aspect, an embodiment of the present application provides a model training method, comprising the following steps:

[0006] Divide the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data;

[0007] Performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data;

[0008] The frequency domain data are sequentially stored into the data buffer to obtain a data matrix;

[0009] Linearly map the points of the data matrix into color information to obtain a spectrum;

[0010] Mark the signal information on the spectrum graph to obtain a training sample image;

[0011] Based on the training sample images, a signal detection model based on deep learning is trained.

[0012] In a possible implementation manner of the first aspect, dividing the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data includes:

[0013] The sample in-phase component data and the sample orthogonal component data are divided into multiple frames and overlapped to obtain first sample in-phase component data and first sample orthogonal component data.

[0014] In a possible implementation manner of the first aspect, after the sample in-phase component data and the sample orthogonal component data are segmented into multiple frames to obtain the first sample in-phase component data and the first sample orthogonal component data, the model training method further includes:

[0015] Performing windowing processing on the first sample in-phase component data and the first sample orthogonal component data to obtain second sample in-phase component data and second sample orthogonal component data;

[0016] Performing a fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data, including:

[0017] Perform fast Fourier transform on the second sample in-phase component data and the second sample orthogonal component data to obtain frequency domain data.

[0018] In a possible implementation of the first aspect, before linearly mapping the points of the data matrix to color information and obtaining the spectrogram, the model training method further includes:

[0019] Determine the relationship between the height of the data matrix and the height threshold, and obtain a determination result;

[0020] Linearly map the points of the data matrix to color information to obtain a spectrum, including:

[0021] The points of the data matrix are linearly mapped to color information, and a spectrum is obtained based on the judgment results.

[0022] In a second aspect, an embodiment of the present application provides a signal detection method, comprising the following steps:

[0023] Segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data;

[0024] Performing fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data;

[0025] The target frequency domain data are sequentially stored into the data buffer to obtain a target data matrix;

[0026] Linearly map the points of the target data matrix into color information to obtain the target spectrum;

[0027] The target spectrum graph is input into a signal detection model to obtain an output result to calculate the signal parameters of the target signal; wherein the signal detection model is obtained according to the model training method provided in any one of the first aspects above.

[0028] In a possible implementation of the second aspect, inputting the target spectrogram into a signal detection model and obtaining an output result to calculate a signal parameter of the target signal includes:

[0029] Input the target spectrum into the signal detection model to obtain the output result;

[0030] The output result is mapped into time information and frequency information, and the signal parameters of the target signal are calculated according to the time information and the frequency information.

[0031] In a third aspect, an embodiment of the present application provides a model training device, comprising:

[0032] A first segmentation module, the first segmentation module is used to segment the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data;

[0033] A first transformation module, the first transformation module is used to perform a fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data;

[0034] A first matrix module, the first matrix module is used to store the frequency domain data into the data buffer in sequence to obtain a data matrix;

[0035] A first mapping module, the first mapping module is used to linearly map the points of the data matrix into color information to obtain a spectrum diagram;

[0036] The annotation module is used to annotate signal information on the spectrum graph to obtain a training sample image;

[0037] The training module is used to train a signal detection model based on deep learning based on training sample images.

[0038] In a fourth aspect, an embodiment of the present application provides a signal detection device, including:

[0039] A second segmentation module, the second segmentation module is used to segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data;

[0040] A second transformation module, the second transformation module is used to perform a fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data;

[0041] A second matrix module, the second matrix module is used to store the target frequency domain data into the data buffer in sequence to obtain a target data matrix;

[0042] A second mapping module, the second mapping module is used to linearly map the points of the target data matrix into color information to obtain a target spectrum diagram;

[0043] A detection module, the detection module is used to input the target spectrum into a signal detection model, obtain an output result to calculate the signal parameters of the target signal; wherein the signal detection model is obtained according to the model training method provided in any one of the first aspects above.

[0044] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium storing a computer program. When the computer program is loaded and executed by a processor, it implements a model training method as provided in any one of the first aspects above or a signal detection method as provided in any one of the second aspects above.

[0045] In a sixth aspect, an embodiment of the present application provides an electronic device, including a processor and a memory, wherein:

[0046] The memory is used to store computer programs;

[0047] The processor is used to load and execute a computer program so that the electronic device performs a model training method as provided in any one of the first aspects above or a signal detection method as provided in any one of the second aspects above.

[0048] Compared with the prior art, the beneficial effects of this application are:

[0049] The embodiments of the present application propose a model training and signal detection method, device, medium and electronic device, and the training method includes: dividing the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain the first sample in-phase component data and the first sample orthogonal component data; performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data; storing the frequency domain data into a data buffer in sequence to obtain a data matrix; linearly mapping the points of the data matrix into color information to obtain a spectrogram; marking the signal information on the spectrogram to obtain a training sample image; based on the training sample image, training to obtain a signal detection model based on deep learning. The present application trains a signal detection model based on deep learning to realize signal detection, and performs detection through image recognition to avoid the influence of electromagnetic environment and signal-noise, so as to improve the accuracy of signal detection and recognition. By collecting data describing the signal I / Q, that is, the in-phase component data and orthogonal component data describing the signal phase and amplitude, these data are divided into multiple frames so that each frame signal is approximately a stationary signal, which is convenient for spectrum analysis. Then, the data is converted into frequency domain data using fast Fourier transform and stored in a buffer in sequence to obtain a data matrix and use it to generate a spectrogram. The spectrogram obtained under linear mapping is trained. Since the signal information is marked on the spectrogram, the trained signal detection model can extract image features and realize fast and accurate recognition of image-based signals in the application stage. After the signal features are recognized on the image, they are mapped into signal parameters to realize signal detection and recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic diagram of the structure of an electronic device of a hardware operating environment involved in an embodiment of the present application;

[0051] Figure 2 A flowchart of a model training method provided in an embodiment of the present application;

[0052] Figure 3 A schematic diagram of multi-frame segmentation of data in the model training method provided in an embodiment of the present application;

[0053] Figure 4 A schematic diagram of mapping through a data matrix in the model training method provided in an embodiment of the present application;

[0054] Figure 5 A schematic diagram of a flow chart of a model training method provided in an embodiment of the present application in one implementation manner;

[0055] Figure 6 A schematic diagram of a flow chart of a signal detection method provided in an embodiment of the present application;

[0056] Figure 7 A schematic diagram of mapping parameter information obtained in the signal detection method provided in an embodiment of the present application;

[0057] Figure 8 A schematic diagram of a flow chart of a signal detection method provided in an embodiment of the present application in one implementation manner;

[0058] Fig. 9 A schematic diagram of a module of a model training device provided in an embodiment of the present application;

[0059] Fig.10 A schematic diagram of a module of a signal detection device provided in an embodiment of the present application;

[0060] Markings in the figure: 101 - processor, 102 - communication bus, 103 - network interface, 104 - user interface, 105 - memory. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0062] See attached Figure 1 , attached Figure 1 This is a schematic diagram of the structure of an electronic device of the hardware operating environment involved in the embodiment of the present application. The electronic device may include: a processor 101, such as a central processing unit (CPU), a communication bus 102, a user interface 104, a network interface 103, and a memory 105. Among them, the communication bus 102 is used to realize the connection and communication between these components. The user interface 104 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 104 may also include a standard wired interface and a wireless interface. The network interface 103 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface). The memory 105 may optionally be a storage device independent of the aforementioned processor 101. The memory 105 may be a high-speed random access memory (RAM) memory, or it may be a stable non-volatile memory (NVM), such as at least one disk storage; the processor 101 may be a general-purpose processor, including a central processing unit, a network processor, etc., or it may be a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component.

[0063] Those skilled in the art will appreciate that Figure 1 The structure shown in the figure does not constitute a limitation on the electronic device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0064] As attached Figure 1 As shown, the memory 105 as a storage medium may include an operating system, a network communication module, a user interface module, and a model training device or a signal detection device.

[0065] In the attached Figure 1 In the electronic device shown, the network interface 103 is mainly used for data communication with a network server; the user interface 104 is mainly used for data interaction with a user; the processor 101 and the memory 105 in the present application can be set in the electronic device, and the electronic device calls the model training device or signal detection device stored in the memory 105 through the processor 101, and executes the model training method or signal detection method provided in the embodiment of the present application.

[0066] Refer to the attached Figure 2 Based on the hardware device of the foregoing embodiment, the embodiment of the present application provides a model training method, comprising the following steps:

[0067] S10: Divide the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data.

[0068] In the specific implementation process, for non-stationary signals, the signal spectrum may be unstable for a long time. After the sample data is equally divided into multiple frames, the signal in each frame is approximately a stationary signal, which is convenient for spectrum analysis. The sample data is the I and Q data of the sample baseband, which can be regarded as the original digital signal data, where I represents the in-phase component and Q represents the orthogonal component. These are used as sample in-phase component data and sample orthogonal component data and divided into multiple frames of signals. The divided data are recorded as the first sample in-phase component data and the first sample orthogonal component data. Specifically: the sample in-phase component data and the sample orthogonal component data are divided into multiple frames to obtain the first sample in-phase component data and the first sample orthogonal component data, including:

[0069] The sample in-phase component data and the sample orthogonal component data are divided into multiple frames and overlapped to obtain first sample in-phase component data and first sample orthogonal component data.

[0070] In the specific implementation process, the data is divided into the same f frame data. In order to improve the frequency resolution and reduce the spectrum leakage, the data is overlapped during the data segmentation. Figure 3The method shown is a 50% overlapping processing method, that is, the starting frame of the segmented data is the middle frame of the previous segmented data, and the ending frame is the middle frame of the next segmented data. In other embodiments, the proportion of the overlapping part can be adjusted according to actual needs; the data after the overlapping processing is recorded as the first sample in-phase component data and the first sample orthogonal component data, respectively, and the signal is divided into multiple overlapping small segments. Each small segment signal in the time domain can be independently transformed by FFT to obtain the frequency domain information corresponding to the segment, which can capture the changing characteristics of the signal in time and frequency to a certain extent, and better observe the dynamic evolution of the signal frequency component over time.

[0071] S20: Performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data.

[0072] In the specific implementation process, the fast Fourier transform (FFT) is an efficient algorithm of the discrete Fourier transform (DFT) and is one of the most basic methods in the time domain to frequency domain transformation analysis. The fast Fourier transform converts the signal from the time domain to the frequency domain and decomposes it into its component frequencies, so as to better analyze, synthesize and process the signal.

[0073] In one embodiment, after the sample in-phase component data and the sample orthogonal component data are segmented into multiple frames to obtain the first sample in-phase component data and the first sample orthogonal component data, the model training method further includes:

[0074] The first sample in-phase component data and the first sample orthogonal component data are subjected to windowing processing to obtain second sample in-phase component data and second sample orthogonal component data.

[0075] In the specific implementation process, after the data is divided into multiple segments, if FFT transformation is performed directly, leakage may occur in the spectrum. Windowing can improve the truncation effect and suppress spectrum leakage. Windowing is a technical means in signal processing. By selecting appropriate window functions and parameters, the characteristic information in the signal can be effectively extracted to improve the accuracy and effect of signal processing. In practical applications, it is necessary to flexibly select and use the windowing method according to the specific signal characteristics and processing requirements. Its main function is to alleviate the Gibbs effect, reduce leakage and improve the accuracy of the processing results. The data after windowing is recorded as the second sample in-phase component data and the second sample orthogonal component data.

[0076] Based on the above steps, performing a fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data includes:

[0077] Perform fast Fourier transform on the second sample in-phase component data and the second sample orthogonal component data to obtain frequency domain data.

[0078] S30: storing the frequency domain data into the data buffer in sequence to obtain a data matrix.

[0079] In the specific implementation process, the m-point fast Fourier transform is performed in the above steps, that is, the m data points are subjected to high-frequency spectrum analysis to be converted into frequency domain data, which are then stored in the data buffer in sequence. At this time, the buffer is a data matrix of size f*m. The data matrix of the buffer is used to generate a waterfall diagram, also called a spectrogram, which is an intuitive way to show the change of signal frequency over time. The waterfall diagram can help quickly identify frequency drift, modulation mode or other frequency characteristics, and is used to extract meaningful frequency information from time domain signals to support signal analysis and detection. At the same time, it can provide visualization tools to help understand signal characteristics.

[0080] S40: Linearly map the points of the data matrix into color information to obtain a spectrum.

[0081] In the specific implementation process, each point of the data matrix is ​​linearly mapped to color information to obtain one or more spectrograms, that is, time-frequency graphs. The number of spectrograms obtained by mapping can be determined according to the height of the matrix and the preset height threshold, that is, before linearly mapping the points of the data matrix to color information and obtaining the spectrogram, the model training method also includes:

[0082] The relationship between the height of the data matrix and the height threshold is determined to obtain a determination result.

[0083] Linearly map the points of the data matrix to color information to obtain a spectrum, including:

[0084] The points of the data matrix are linearly mapped to color information, and a spectrum is obtained based on the judgment results.

[0085] In the specific implementation process, the height threshold is set to 992 in this embodiment. If the height of the data matrix is ​​not greater than 992, the data matrix is ​​directly mapped to a spectrum graph. If it is greater than 992, the data matrix is ​​split into multiple spectrum graphs. The Y-axis direction of two adjacent spectrum graphs has a certain degree of overlap, as shown in the attached figure. Figure 4 The figure shows a method of mapping to obtain a spectrum graph, specifically, when the height of the data matrix is ​​greater than 992, the overlap is set to 30% to map to obtain multiple spectrum graphs.

[0086] S50: Marking signal information on the spectrum graph to obtain a training sample image.

[0087] In the specific implementation process, the spectrogram is saved as a bitmap as a training sample. For example, in the implementation method of the aforementioned steps, the spectrogram is saved as a bitmap of 992*992 size, that is, an image that can be recognized by deep learning. N groups of sample data are collected and saved as pictures as needed, where the N groups of sample data contain radar signals of different modulation types. The signal information is marked on the spectrogram, that is, the spectrogram is marked after being saved as a bitmap. The marked information includes the signal type, center coordinates, and width and height of the signal. After the marking is completed, the training sample image can be divided into a training set and a test set in proportion, such as a 9:1 ratio between the training set and the test set.

[0088] S60: Based on the training sample images, a signal detection model based on deep learning is obtained through training.

[0089] In the specific implementation process, deep learning is used for training, and existing model architectures can be used, such as the YOLOv8s model. This model is a typical target detection model, and the overall architecture consists of a basic network, a feature extraction layer, a prediction layer, and an output layer. The basic network is mainly responsible for extracting the underlying features of the image, the feature extraction layer is used to further extract higher-level semantic features, the prediction layer is responsible for predicting the location and category of the target, and the output layer is used to output the final detection results. According to the training results, the training cycle, pre-training weights and other parameters are adjusted until the model converges on the training samples to obtain the training model yoloModel, which is also the signal detection model.

[0090] In this embodiment, a signal detection model based on deep learning is trained to realize signal detection, and detection is performed through image recognition to avoid the influence of electromagnetic environment and signal-noise, so as to improve the accuracy of signal detection and recognition. Data describing the signal I / Q, that is, the in-phase component data and the orthogonal component data describing the phase and amplitude of the signal are collected, and these data are divided into multiple frames so that each frame signal is approximately a stationary signal, which is convenient for spectrum analysis. Then, the data is converted into frequency domain data by fast Fourier transform and stored in a buffer in sequence to obtain a data matrix and use it to generate a spectrogram. The spectrogram obtained under linear mapping is trained. Since the signal information is marked on the spectrogram, the trained signal detection model can extract image features and realize fast and accurate recognition of image-based signals in the application stage. After the signal features are recognized on the image, they are mapped into signal parameters to realize detection and recognition of the signal.

[0091] Refer to the attached Figure 5 , in the attached Figure 5 The model training method provided in this application is further described below:

[0092] First, the sample I and Q data are obtained, and then they are divided into f frames and converted into frequency domain data through FFT. They are stored in the data buffer in sequence to obtain a data matrix. Then, the points of the data matrix are linearly mapped to color information to obtain a spectrogram. The spectrogram is saved as a bitmap and then data is annotated and divided to form a training sample image. Finally, it is input into YOLOv8s for iterative training. After the training is completed, the training model is exported to obtain a signal detection model that can be used to implement signal detection.

[0093] See attached Figure 6 Based on the same inventive concept as in the above-mentioned embodiment, the embodiment of the present application further provides a signal detection method, comprising the following steps:

[0094] S100: Segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data;

[0095] S200: Performing fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data;

[0096] S300: storing the target frequency domain data into the data buffer in sequence to obtain a target data matrix;

[0097] S400: Linearly map the points of the target data matrix into color information to obtain a target spectrum graph;

[0098] S500: Input the target spectrum graph into a signal detection model to obtain an output result to calculate the signal parameters of the target signal; wherein the signal detection model is obtained according to the model training method provided in an embodiment of the present application.

[0099] In the specific implementation process, the beneficial effects of the embodiments of the present application can refer to the aforementioned embodiments. In the application stage of the model, in the detection process of the signal, the processing process before the target spectrogram is input into the signal detection model is similar to the process of model training. The target signal is the signal to be detected, and the IQ signal to be detected with a collection time length of Ts, that is, the in-phase component data and the orthogonal component data, is divided into f frames of data according to the model training process, and then fast Fourier transform is performed, stored in the data buffer, and mapped into a spectrogram, that is, the target spectrogram. During this period, the data can also be overlapped and windowed according to the implementation method of model training. At the same time, it is necessary to record the starting time tf of each frame of the segmented f frames of data, and the calculation method is tf=i*Ts / fi=0,1,2...f.

[0100] The signal detection model obtained by the training method provided in the embodiment of the present application is called for reasoning to obtain a set of prediction results. It should be noted that the target spectrum here also needs to be saved as a bitmap recognizable by the model for input reasoning. The result set contains the category of each detected signal, the width and height of the signal box, the pixel offset of the x and y axes of the center of the signal box, and the prediction confidence. The pixel information of each prediction result in the result set is mapped to the parameter information of the signal. The mapping method is as shown in the attached figure. Figure 7 As shown. The parameter information includes center frequency, signal bandwidth, signal pulse and pulse interval. It is known that during the measurement process, the center frequency centFreq of the collected IQ data, the analysis bandwidth span, and the acquisition time Tm are used to calculate the center frequency SigcentFreq of the signal using the following formula:

[0101]

[0102] Among them, x is the Figure 7 The x-axis coordinate of the center point of the middle signal frame 1 (the entire image can be regarded as a coordinate axis area, the upper left corner of the image represents the origin of the coordinate axis, the coordinates of the lower right corner of the image are (img.Width, img.Height), and the unit of the coordinate axis is pixel). img.Width represents the image width of the target spectrum graph in pixels.

[0103] Use the following formula to get the signal start time SigStartTime:

[0104]

[0105] Among them, y is the Figure 7 The y-axis coordinate of the center point of signal frame 1, result.Height represents the height of signal frame 1, that is, SigHeight, and img.Height represents the image height of the target spectrum graph, in pixels.

[0106] Similarly, the frequency and time information corresponding to the four corners of the signal frame can be obtained, and then the parameters of the signal can be calculated, that is, the target spectrum is input into the signal detection model, and the output result is obtained to calculate the signal parameters of the target signal, including:

[0107] Input the target spectrum into the signal detection model to obtain the output result;

[0108] The output result is mapped into time information and frequency information, and the signal parameters of the target signal are calculated according to the time information and the frequency information.

[0109] As attached Figure 7As shown in the figure, the signal pulse width is the frequency difference corresponding to the width of the signal frame, the signal duration is the time difference corresponding to the height of the signal frame, the signal interval is the time difference corresponding to the edges of the two adjacent signal frames, (x, y) is the coordinate of the center point of signal frame 1, this coordinate point can be output by the detection model, SigWidth and WigHeight respectively represent the width and height of signal frame 1. Width and Height respectively represent the width and height of the target spectrum, which are img.Width and img.Height in the above formula.

[0110] Refer to the attached Figure 8 , in the attached Figure 8 The signal detection method provided by the present application is further described below in the following embodiments:

[0111] First, measure the I and Q data. According to a similar processing method during model training, divide it into f frames, convert it into frequency domain data through FFT, and store it in the data buffer in sequence to obtain a data matrix. At the same time, record the start time of each frame of data. Then, the matrix height, that is, the number of matrix rows, can be determined. Set the threshold to 992 to determine whether the number of matrix rows is greater than 992. If not, directly map the data matrix to a single spectrogram. If so, split the matrix and map it to multiple spectrograms. Then use the YOLO model for inference and prediction, that is, the signal detection model obtained by the training method provided in the embodiment of the present application, to obtain the output result and map it to time and frequency, calculate the signal parameters to complete the recognition, and finally use the waterfall chart to refresh and display the recognition result in real time.

[0112] Refer to the attached Fig. 9 Based on the same inventive concept as in the above-mentioned embodiment, the embodiment of the present application further provides a model training device, comprising:

[0113] A first segmentation module, the first segmentation module is used to segment the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data;

[0114] A first transformation module, the first transformation module is used to perform a fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data;

[0115] A first matrix module, the first matrix module is used to store the frequency domain data into the data buffer in sequence to obtain a data matrix;

[0116] A first mapping module, the first mapping module is used to linearly map the points of the data matrix into color information to obtain a spectrum diagram;

[0117] The annotation module is used to annotate signal information on the spectrum graph to obtain a training sample image;

[0118] The training module is used to train a signal detection model based on deep learning based on training sample images.

[0119] See attached Fig.10 Based on the same inventive concept as in the above-mentioned embodiment, the embodiment of the present application further provides a signal detection device, comprising:

[0120] A second segmentation module, the second segmentation module is used to segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data;

[0121] A second transformation module, the second transformation module is used to perform a fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data;

[0122] A second matrix module, the second matrix module is used to store the target frequency domain data into the data buffer in sequence to obtain a target data matrix;

[0123] A second mapping module, the second mapping module is used to linearly map the points of the target data matrix into color information to obtain a target spectrum diagram;

[0124] The detection module is used to input the target spectrum into the signal detection model to obtain the output result to calculate the signal parameters of the target signal; wherein, the signal detection model is obtained according to the model training method provided in the embodiment of the present application.

[0125] Those skilled in the art should understand that the division of the various modules in the embodiment is merely a division of logical functions, and in actual application, all or part of them can be integrated into one or more actual carriers, and these modules can be implemented entirely in the form of software called by a processing unit, or entirely in the form of hardware, or in the form of a combination of software and hardware. It should be noted that the modules in the model training device and the signal detection device in this embodiment correspond one-to-one to the steps in the model training method and the signal detection method in the aforementioned embodiment. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned embodiment, and will not be repeated here.

[0126] Based on the same inventive concept as in the aforementioned embodiments, an embodiment of the present application further provides a computer-readable storage medium storing a computer program. When the computer program is loaded and executed by a processor, a model training method or a signal detection method as provided in the embodiments of the present application is implemented.

[0127] Based on the same inventive concept as in the above-mentioned embodiment, an embodiment of the present application further provides an electronic device, including a processor and a memory, wherein:

[0128] The memory is used to store computer programs;

[0129] The processor is used to load and execute a computer program so that the electronic device executes a model training method or a signal detection method as provided in the embodiments of the present application.

[0130] In some embodiments, the computer readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or various devices including one or any combination of the above memories. The computer may be various computing devices including intelligent terminals and servers.

[0131] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0132] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0133] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices located at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0134] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0135] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0136] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a multimedia terminal device (which can be a mobile phone, a computer, a television receiver, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0137] In summary, the present application provides a model training and signal detection method, device, medium and electronic device, and the training method includes: dividing the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data; performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data; storing the frequency domain data into a data buffer in sequence to obtain a data matrix; linearly mapping the points of the data matrix into color information to obtain a spectrogram; marking the signal information on the spectrogram to obtain a training sample image; based on the training sample image, training to obtain a signal detection model based on deep learning. The present application trains a signal detection model based on deep learning to realize signal detection, and performs detection through image recognition to avoid the influence of electromagnetic environment and signal-noise, so as to improve the accuracy of signal detection and recognition. By collecting data describing the signal I / Q, that is, the in-phase component data and orthogonal component data describing the signal phase and amplitude, these data are divided into multiple frames so that each frame signal is approximately a stationary signal, which is convenient for spectrum analysis. Then, the data is converted into frequency domain data using fast Fourier transform and stored in a buffer in sequence to obtain a data matrix and use it to generate a spectrogram. The spectrogram obtained under linear mapping is trained. Since the signal information is marked on the spectrogram, the trained signal detection model can extract image features and realize fast and accurate recognition of image-based signals in the application stage. After the signal features are recognized on the image, they are mapped into signal parameters to realize signal detection and recognition.

[0138] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A model training method, characterized in that: The following steps are involved: Divide the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data; Performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data; Storing the frequency domain data in a data buffer in sequence to obtain a data matrix; Linearly mapping the points of the data matrix into color information to obtain a spectrum; Marking signal information on the spectrogram to obtain a training sample image; Based on the training sample images, a signal detection model based on deep learning is obtained through training.

2. The model training method according to claim 1, characterized in that: The step of dividing the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data includes: The sample in-phase component data and the sample orthogonal component data are divided into multiple frames and overlapped to obtain first sample in-phase component data and first sample orthogonal component data.

3. The model training method according to claim 1, characterized in that: After the sample in-phase component data and the sample orthogonal component data are segmented into multiple frames to obtain the first sample in-phase component data and the first sample orthogonal component data, the model training method further comprises: Performing windowing processing on the first sample in-phase component data and the first sample orthogonal component data to obtain second sample in-phase component data and second sample orthogonal component data; The performing fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data includes: Perform fast Fourier transform on the second sample in-phase component data and the second sample orthogonal component data to obtain frequency domain data.

4. The model training method according to claim 1, characterized in that: Before linearly mapping the points of the data matrix into color information to obtain a spectrum, the model training method further includes: Determine the size relationship between the height of the data matrix and the height threshold, and obtain a determination result; The step of linearly mapping the points of the data matrix into color information to obtain a spectrum diagram comprises: The points of the data matrix are linearly mapped into color information, and a spectrum diagram is obtained according to the judgment result.

5. A signal detection method, characterized in that: The following steps are involved: Segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data; Performing fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data; Storing the target frequency domain data in a data buffer in sequence to obtain a target data matrix; Linearly mapping the points of the target data matrix into color information to obtain a target spectrum graph; The target spectrogram is input into a signal detection model to obtain an output result to calculate the signal parameters of the target signal; wherein the signal detection model is obtained according to the model training method described in any one of claims 1-4.

6. The signal detection method according to claim 5, characterized in that: The step of inputting the target spectrogram into a signal detection model and obtaining an output result to calculate a signal parameter of the target signal comprises: Inputting the target spectrogram into a signal detection model to obtain an output result; The output result is mapped into time information and frequency information, and a signal parameter of the target signal is calculated according to the time information and the frequency information.

7. A model training device, characterized in that: include: A first segmentation module, the first segmentation module is used to segment the sample in-phase component data and the sample orthogonal component data into multiple frames to obtain first sample in-phase component data and first sample orthogonal component data; A first transformation module, the first transformation module is used to perform a fast Fourier transform on the first sample in-phase component data and the first sample orthogonal component data to obtain frequency domain data; A first matrix module, the first matrix module is used to store the frequency domain data into a data buffer in sequence to obtain a data matrix; A first mapping module, the first mapping module is used to linearly map the points of the data matrix into color information to obtain a spectrum diagram; A labeling module, the labeling module is used to label signal information on the spectrum graph to obtain a training sample image; A training module is used to train a signal detection model based on deep learning based on the training sample images.

8. A signal detection device, characterized in that: include: A second segmentation module, the second segmentation module is used to segment the in-phase component data and the orthogonal component data of the target signal into multiple frames to obtain first in-phase component data and first orthogonal component data; A second transformation module, the second transformation module is used to perform a fast Fourier transform on the first in-phase component data and the first orthogonal component data to obtain target frequency domain data; A second matrix module, the second matrix module is used to store the target frequency domain data into a data buffer in sequence to obtain a target data matrix; A second mapping module, the second mapping module is used to linearly map the points of the target data matrix into color information to obtain a target spectrum diagram; A detection module, wherein the detection module is used to input the target spectrum into a signal detection model to obtain an output result to calculate the signal parameters of the target signal; wherein the signal detection model is obtained according to the model training method described in any one of claims 1-4.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is loaded and executed by the processor, the model training method as described in any one of claims 1 to 4 or the signal detection method as described in any one of claims 5 to 6 is implemented.

10. An electronic device, characterized in that: comprising a processor and a memory, wherein: The memory is used to store computer programs; The processor is used to load and execute the computer program so that the electronic device performs the model training method as described in any one of claims 1-4 or the signal detection method as described in any one of claims 5-6.

Images