A method for processing real short wave protocol sample data and application

Abstract

The application discloses a kind of real shortwave protocol sample data processing methods, comprising: obtaining real shortwave protocol sample data set;By anti-fading energy detection method, the burst signal in real shortwave protocol sample data set is sequentially intercepted, and burst signal set is formed;If it is judged that the burst signal is real sample, the center frequency of the real sample is obtained, and the real sample is converted into I / Q burst signal by frequency conversion according to the center frequency;The sampling rate of I / Q burst signal is obtained, and the corresponding low-pass filter is selected according to the sampling rate, and the I / Q burst signal is filtered out by low-pass filter to remove the out-of-band redundant information;The signal amplitude and signal phase between I and Q of I / Q burst signal are sequentially calculated, and the training data is formed according to the arrangement mode of amplitude and phase interleaving.It can solve the problem that the data specification of real shortwave communication protocol in the prior art is not standard, the data amount is not balanced, and the training difficulty is great.

Description

A method for processing sample data of shortwave protocol and its application Technical Field

[0001] This application relates to the fields of communication and deep learning technology, and more specifically, to a method for processing sample data of shortwave protocol and a training device for a shortwave protocol intelligent recognition model. Background Technology

[0002] In the training and implementation of intelligent shortwave protocol recognition models, the standardization of actual protocol samples and the extraction of burst signals have always been challenging aspects of model training. Specifically, this manifests in several ways: due to differences in source and purpose, the signal start positions, center frequencies, and data specifications in the sample library formed by actual protocol samples are not fixed; the varying communication frequencies of different shortwave protocols lead to an imbalance in the amount of data among different types of actual protocol samples; and multipath fading caused by multipath effects severely impacts the signal energy of actual protocol samples.

[0003] Currently, common methods for extracting and detecting burst signals from real-world protocol samples include time-consuming and labor-intensive manual screening and labeling, and synchronous demodulation methods which have a certain false detection rate and cannot extract protocol signals with unknown synchronization information. Given these limitations, real-world protocol samples are difficult to use as training data for recognition models in actual training processes. Therefore, proposing a standardized method for processing real-world protocol sample data and establishing a suitable model training system based on this standardized data is an urgent problem to be solved. Summary of the Invention

[0004] To address at least one deficiency or improvement need in the prior art, this invention provides a method for processing sample data of shortwave protocol and a training device for a shortwave protocol intelligent recognition model, aiming to solve the problems of non-standard specifications, unbalanced data volume, and high training difficulty of shortwave communication protocol data in the prior art.

[0005] To achieve the above objectives, according to a first aspect of the present invention, a method for processing sample data of a shortwave protocol obtained from actual sampling is provided, comprising: acquiring a dataset of sample data of a shortwave protocol obtained from actual sampling; sequentially extracting burst signals from the dataset of sample data of the shortwave protocol obtained from actual sampling using an anti-fading energy detection method to form a set of burst signals; if it is determined that the burst signals in the set of burst signals are real samples, then acquiring the center frequency of the real samples, and converting the real samples into I / Q burst signals by down-conversion according to the center frequency; acquiring the sampling rate of the I / Q burst signals, selecting a corresponding low-pass filter according to the sampling rate, and filtering out out-of-band redundant information from the I / Q burst signals by passing the low-pass filter; sequentially calculating the signal amplitude and signal phase between the I and Q paths of the I / Q burst signals, and forming training data according to an alternating arrangement of amplitude and phase.

[0006] In one embodiment of the present invention, the acquisition of the actual shortwave protocol sample dataset includes: a series of .wav format data obtained by broadband reception and narrowband sampling through a shortwave receiver; the sample sampling rate adopts one of the standard levels of 9.6KHz, 24KHz, and 48KHz; the sample data indicates the signal bandwidth, and if the saved format is a real number signal, the signal center frequency is also indicated.

[0007] In one embodiment of the present invention, the anti-fading energy detection method includes the following steps: reading the shortwave protocol sample data to be detected and windowing it; sequentially calculating the signal energy under each window and selecting a continuous relatively high-energy window sequence; for the relatively low-energy window sequence, loading a neural network trained for anti-fading and performing energy compensation on it; re-selecting a continuous relatively high-energy window sequence to obtain the burst signal detected by energy.

[0008] In one embodiment of the present invention, the down-conversion formula is: ;in, For real number signals, This is the signal after down-conversion. The center frequency of the signal. The sampling rate.

[0009] In one embodiment of the present invention, the relationship between the cutoff frequency of the low-pass filter and the sample sampling rate is as follows: a sampling rate of 9.6KHz corresponds to a cutoff frequency of 3KHz; or a sampling rate of 24KHz corresponds to a cutoff frequency of 12KHz; or a sampling rate of 48KHz corresponds to a cutoff frequency of 24KHz.

[0010] In one embodiment of the present invention, the signal amplitudes of the I-path and Q-path of the I / Q burst signal are expressed as follows: The signal phase is represented as: ;in, For signal amplitude, For signal phase, For I-channel signal, This is the Q-channel signal.

[0011] In one embodiment of the present invention, the signal amplitude and signal phase are arranged as follows: Where n is the number of shortwave protocol samples actually collected. For signal amplitude, This refers to the signal phase.

[0012] According to a second aspect of the present invention, a shortwave protocol intelligent recognition model training device is also provided, comprising: a data loading module for batch reading actual shortwave protocol sample data from a specified directory and performing data augmentation on protocol categories with fewer than a preset number of sample data; a training parameter configuration module for adjusting the training hyperparameters and training modes required for the training process; a training module for initializing and establishing or reading a network recognition model and adjusting the recognition model parameters through iterative training; and a result output module for generating a recognition accuracy change curve for monitoring the training process and outputting the recognition result model.

[0013] In one embodiment of the present invention, the data enhancement includes: frequency domain enhancement: adding a random frequency offset within the range of 200Hz to -200Hz to the sample data; time domain enhancement: adding blank data of random length within the range of 0 to 75ms before and after the sample data; and noise enhancement: superimposing random Gaussian noise with an intensity within the range of 0 to 2dB onto the sample data.

[0014] In one embodiment of the present invention, the configuration of the training hyperparameters and training mode includes: training rounds, used to adjust the number of model parameter iterations; training batches, used to configure the number of samples processed simultaneously in each training iteration; a base learning rate, used to configure the degree of influence of each training iteration on parameter changes; a pre-trained model loading mode, used to configure whether to load a pre-trained model; and a result saving mode, used to configure the result model to be saved as either the best model or all models.

[0015] In another aspect, embodiments of the present invention also provide an electronic device, comprising: a memory and one or more processors connected to the memory, the memory storing a computer program, and the processors being used to execute the computer program to implement the real-time shortwave protocol sample data processing method as described in any of the above embodiments.

[0016] In another aspect, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions for performing the shortwave protocol sample data processing method as described in any of the above embodiments.

[0017] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0018] The method and model training device for processing shortwave protocol sample data provided in this invention improve shortwave protocol data by detecting the location of burst signals through energy detection and enhancing a small amount of protocol data. Specific training is then performed using a parameter-configurable training device, thereby improving the usability of the sampled data for shortwave protocol recognition model training tasks. This solves the problems of non-standard specifications, unbalanced data volume, and high training difficulty in current shortwave communication protocol data. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 is a flowchart of the method for processing sample shortwave protocol data provided in an embodiment of this application;

[0021] Figure 2 is a structural diagram of the model training device provided in an embodiment of this application;

[0022] Figure 3 is a diagram of the parameter configuration interface of the model training device provided in the embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0024] The terms "first," "second," "third," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. 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 includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0025] As shown in Figure 1, the first embodiment of the present invention proposes a method for processing sample data of a shortwave protocol, including: step S1, acquiring a sample dataset of shortwave protocol data; step S2, sequentially extracting burst signals from the sample dataset of shortwave protocol data using an anti-fading energy detection method to form a burst signal set; step S3, if the burst signal in the burst signal set is determined to be a real sample, then acquiring the center frequency of the real sample, and converting the real sample into an I / Q burst signal by down-conversion according to the center frequency; step S4, acquiring the sampling rate of the I / Q burst signal, selecting a corresponding low-pass filter according to the sampling rate, and filtering out out-of-band redundant information from the I / Q burst signal through the low-pass filter; step S5, sequentially calculating the signal amplitude and signal phase between the I and Q paths of the I / Q burst signal, and forming training data according to the alternating arrangement of amplitude and phase.

[0026] In step S1, for example, a shortwave protocol sample dataset is acquired using a shortwave receiver in a designated controlled frequency band. Specifically, this includes a series of .wav format data obtained by wideband reception and narrowband decimation sampling using a shortwave receiver; the sample sampling rate is one of the standard levels of 9.6 kHz, 24 kHz, and 48 kHz; the sample data indicates the signal bandwidth, and if the saved format is a real number signal, the signal center frequency is also indicated.

[0027] In step S2, the acquired shortwave protocol sample data is used to detect burst signals in the sample data using an anti-fading energy detection method. The anti-fading energy detection method includes: reading the shortwave protocol sample data to be detected and windowing it; calculating the signal energy under each window sequentially and selecting a continuous relatively high-energy window sequence; for the relatively low-energy window sequence, loading a neural network trained on anti-fading technology to perform energy compensation; and re-selecting a continuous relatively high-energy window sequence to obtain the burst signal detected by energy.

[0028] In step S3, if the actual burst signal is a real number signal, then the center frequency of the signal needs to be read and down-converted to a baseband I / Q signal, as shown in the following formula:

[0029] ;

[0030] The down-conversion formula is as follows: ;

[0031] in, For real number signals, This is the signal after down-conversion. The center frequency of the signal. The sampling rate.

[0032] In step S4, the sampling rate of the I / Q burst signal is obtained, and a corresponding low-pass filter is selected based on the sampling rate. The I / Q burst signal is then passed through the low-pass filter to remove out-of-band redundant information. The cutoff frequency and coefficients of the filter are determined by the sample sampling rate.

[0033] Optionally, the relationship between the cutoff frequency of the low-pass filter and the sample sampling rate is as follows: a sampling rate of 9.6 kHz corresponds to a cutoff frequency of 3 kHz; or a sampling rate of 24 kHz corresponds to a cutoff frequency of 12 kHz; or a sampling rate of 48 kHz corresponds to a cutoff frequency of 24 kHz.

[0034] In step S5, the signal amplitude and phase between the I and Q channels of the I / Q burst signal are calculated sequentially, and the training data is formed according to the arrangement of [amplitude-phase-amplitude-phase…]. The signal amplitudes of the I and Q channels of the I / Q burst signal are expressed as: The signal phase is represented as: ;in, For signal amplitude, For signal phase, For I-channel signal, This is the Q-channel signal.

[0035] The arrangement of the signal amplitude and signal phase is as follows:

[0036] ;

[0037] Where n is the number of shortwave protocol samples actually used. For signal amplitude, This refers to the signal phase.

[0038] In this way, the shortwave protocol sample data processing method provided by the present invention improves the shortwave protocol data by detecting the location of burst signals through energy detection and enhancing the small amount of protocol data. Through specific training by a parameter-configurable training device, the availability of the sampled data in the shortwave protocol recognition model training task is improved, thereby solving the problems of non-standard specifications, unbalanced data volume, and high training difficulty of the current shortwave communication protocol data.

[0039] Furthermore, as shown in Figure 2, the second embodiment of the present invention proposes a shortwave protocol intelligent recognition model training device, which includes, for example, a data loading module, a training parameter configuration module, a training module, and a result output module. The data loading module is used to batch read real-world shortwave protocol sample data from a specified directory and perform data augmentation on protocol categories with fewer than a preset number of sample data. The training parameter configuration module is used to adjust the training hyperparameters and training modes required for the training process. The training module is used to initialize and establish or read the network recognition model and adjust the recognition model parameters through iterative training. The result output module generates a recognition accuracy change curve for monitoring the training process and outputs the recognition result model.

[0040] Specifically, the model training device is deployed on a high-performance server equipped with GPUs, and its operation process includes the following steps:

[0041] Step 1: When the device is running, the user first needs to configure the training parameters and select the training mode on the user interface of the training device. The user's configuration information will be transmitted to the training module through the training parameter configuration module.

[0042] Step 2: The device will then read the training data in batches from the specified configuration "training data directory" and process the training data through the data loading and processing module to form the training data.

[0043] If the sample data is small (less than 1000 samples for a certain category), the data augmentation module will be automatically activated. The data augmentation module includes frequency domain augmentation, time domain augmentation, and noise augmentation.

[0044] Frequency domain enhancement: Adds a random frequency offset within the range of 200Hz to -200Hz to the sample data;

[0045] The specific method is as follows:

[0046]

[0047] in For the sample data to be augmented, For the augmented sample data, The center frequency of the signal. To increase the random frequency offset within the range of 200Hz to -200Hz, The sampling rate.

[0048] Temporal augmentation: Add blank data of random length ranging from 0 to 75 ms before and after the sample data;

[0049] Noise enhancement involves superimposing random Gaussian noise with an intensity ranging from 0 to 2 dB onto the sample data.

[0050] Step 3: After processing the training data, the device will automatically load the built-in model and initialize the model's internal network parameters. Then, it will automatically begin training and iterating the model according to the configured training parameters until the maximum number of training epochs is reached. The training progress will be displayed in real-time on the interactive interface.

[0051] Step 4: The trained model will be saved to the specified output directory via the results saving module. The resulting model includes a training checkpoint model of type ".chkpt" and an application model of type ".onnx". The training variation curve and the recognition accuracy confusion matrix generated during training will also be saved to the output directory via the results saving module.

[0052] Figure 3 shows the parameter configuration interface of the model training device provided by the present invention. The model training device can interact with the user through the parameter configuration interface to control the training mode. The configurable parameters are as follows:

[0053] Training data directory: Configures the storage location for the live shortwave protocol dataset. The directory location can be changed manually in the window or by clicking the "Select Directory" button.

[0054] Results Save Directory: Configure the location where training results will be stored. The directory location can be changed manually in the window or by clicking the "Select Directory" button.

[0055] Training rounds: This is a configurable number of training rounds, which adjusts the number of iterations for updating model parameters. A larger value means more training iterations.

[0056] Training batch size: This can be configured to process the number of samples simultaneously in each training iteration. The larger the value, the higher the training efficiency and the better the convergence stability, but the larger the GPU memory usage.

[0057] Base learning rate: This configurable value shows the degree to which each training iteration affects parameter changes. A larger value results in faster training but makes convergence more difficult; a smaller value makes convergence easier but results in slower training and a greater likelihood of getting stuck in local optima.

[0058] Pre-trained model loading mode: This can be configured to load a pre-trained model. When it is "yes", it means that a pre-trained model is loaded for this training, which is generally used when resuming training after a breakpoint is required; when it is "no", it means that a new model is trained.

[0059] Results saving mode: The result model can be saved in either the best model or all models. When set to "best model", only the training result model with the highest protocol classification accuracy is saved in each training round; when set to "all models", all models are saved in each training round.

[0060] The shortwave protocol intelligent identification model training device disclosed in the second embodiment of the present invention is applicable to implementing the shortwave protocol sample data processing method described in the first embodiment above. The specific method can be found in the content of the first embodiment, and therefore will not be described in detail here. Optionally, each module and the other operations or functions described above are for implementing the method described in the first embodiment, and the beneficial effects of this embodiment are the same as those of the first embodiment above. For the sake of brevity, they will not be repeated here.

[0061] The third embodiment of the present invention also proposes an electronic device, for example including: at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the computer program is executed by the processing unit, the processing unit performs the method described in the first embodiment, and the beneficial effects of the electronic device provided in this embodiment are the same as the beneficial effects of the real-time shortwave protocol sample data processing method provided in the first embodiment.

[0062] The fourth embodiment of the present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method. The beneficial effects of the computer-readable storage medium provided in this embodiment are the same as those of the real-time shortwave protocol sample data processing method provided in the first embodiment.

[0063] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0064] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0065] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only 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 coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0066] 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 network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0067] 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.

[0068] 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 device (CMD). 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 memory 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 described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0069] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0070] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of embodiments of this disclosure upon considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

[0071] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for processing sample data of a shortwave protocol, characterized in that, Includes the following steps: Acquire a real-world shortwave protocol sample dataset; sequentially extract burst signals from the real-world shortwave protocol sample dataset using an anti-fading energy detection method to form a burst signal set; if the burst signal in the burst signal set is determined to be a real sample, obtain the center frequency of the real sample, and convert the real sample into an I / Q burst signal using down-conversion based on the center frequency; obtain the sampling rate of the I / Q burst signal, select the corresponding low-pass filter based on the sampling rate, and filter out out-of-band redundant information from the I / Q burst signal using the low-pass filter; sequentially calculate the signal amplitude and signal phase between the I and Q paths of the I / Q burst signal, and form the training data according to the alternating amplitude and phase arrangement; wherein, the down-conversion formula is: ;in, For real number signals, This is the signal after down-conversion. The center frequency of the signal. The sampling rate is denoted as ; whereby the signal amplitudes of the I-channel and Q-channel of the I / Q burst signal are expressed as: The signal phase is represented as: ;in, For signal amplitude, For signal phase, For I-channel signal, This is a Q-channel signal; wherein the signal amplitude and signal phase are arranged as follows: Where n is the number of shortwave protocol samples actually collected. For signal amplitude, This refers to the signal phase.

2. The method for processing shortwave protocol sample data as described in claim 1, characterized in that, The acquisition of the shortwave protocol sample dataset includes: a series of .wav format data obtained by broadband reception and narrowband sampling through a shortwave receiver; the sample sampling rate adopts one of the standard levels of 9.6KHz, 24KHz, and 48KHz; the sample data indicates the signal bandwidth, and if the format is a real number signal, the signal center frequency is also indicated.

3. The method for processing shortwave protocol sample data as described in claim 1, characterized in that, The anti-fading energy detection method includes the following steps: reading the actual shortwave protocol sample data to be detected and applying a window; The signal energy under each window is calculated sequentially, and a continuous sequence of relatively high-energy windows is selected. For the relatively low-energy window sequence, a neural network trained to resist fading is loaded to compensate for its energy. A new sequence of continuous relatively high-energy windows is selected to obtain the burst signal detected by energy.

4. The method for processing shortwave protocol sample data as described in claim 1, characterized in that, The relationship between the cutoff frequency of the low-pass filter and the sample sampling rate is as follows: a sampling rate of 9.6 kHz corresponds to a cutoff frequency of 3 kHz; or a sampling rate of 24 kHz corresponds to a cutoff frequency of 12 kHz; or a sampling rate of 48 kHz corresponds to a cutoff frequency of 24 kHz.

5. A shortwave protocol intelligent identification model training device, applicable to the shortwave protocol sample data processing method described in any one of claims 1-4, characterized in that, include: The data loading module is used to read shortwave protocol sample data from a specified directory in batches and to perform data augmentation on protocol categories with fewer than a preset number of sample data. The training parameter configuration module is used to adjust the training hyperparameters and training modes required during the training process; The training module is used to initialize and build or read the network recognition model, and to adjust the recognition model parameters through iterative training. The results output module generates a graph showing the change in recognition accuracy to monitor the training process and outputs the recognition result model.

6. The shortwave protocol intelligent recognition model training device as described in claim 5, characterized in that, The data enhancement includes: frequency domain enhancement: adding random frequency offsets within the range of 200Hz to -200Hz to the sample data; time domain enhancement: adding blank data of random length within the range of 0 to 75ms before and after the sample data; noise enhancement: superimposing random Gaussian noise with an intensity of 0 to 2dB on the sample data.

7. The shortwave protocol intelligent recognition model training device as described in claim 5, characterized in that, The configuration of the training hyperparameters and training modes includes: number of training rounds, used to adjust the number of model parameter iterations; number of training batches, used to configure the number of samples processed simultaneously in each training iteration; base learning rate, used to configure the degree of influence of each training iteration on parameter changes; pre-trained model loading mode, used to configure whether to load the pre-trained model; and result saving mode, used to configure the method of saving the result model as either saving the best model or saving all models.

Images