CN119199792A - Radar signal recognition method, device and equipment based on software and deep learning - Google Patents
Radar signal recognition method, device and equipment based on software and deep learning Download PDFInfo
- Publication number
- CN119199792A CN119199792A CN202411708788.1A CN202411708788A CN119199792A CN 119199792 A CN119199792 A CN 119199792A CN 202411708788 A CN202411708788 A CN 202411708788A CN 119199792 A CN119199792 A CN 119199792A
- Authority
- CN
- China
- Prior art keywords
- radar signal
- radar
- neural network
- convolutional neural
- network model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/417—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section involving the use of neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
本发明涉及基于软件与深度学习的雷达信号识别方法、装置和设备，通过首先构建了基于深度学习的卷积神经网络模型，再结合软件无线电的方便灵活性给卷积神经网络模型喂入大量的不同类别的雷达信号数据进行训练，让卷积神经网络模型自动进行信号的特征提取，从而极大的提高雷达信号的识别准确率。
The present invention relates to a radar signal recognition method, device and equipment based on software and deep learning. A convolutional neural network model based on deep learning is first constructed, and then a large amount of radar signal data of different categories is fed into the convolutional neural network model for training in combination with the convenience and flexibility of software radio, so that the convolutional neural network model automatically extracts signal features, thereby greatly improving the recognition accuracy of radar signals.
Description
Technical Field
The invention belongs to the technical field of radar signal processing, and relates to a radar signal identification method, device and equipment based on software and deep learning.
Background
The identification of radar signals plays an important role in electronic countermeasure, and can effectively acquire value information transmitted by the opposite radar, so that the opposite radar can be favorably deployed by the opposite party, wherein the main task is to identify the type of the opposite radar signals so as to determine the parameter information of the radar signals. However, in reality, radar signals are various in types and systems, pulse signals and continuous wave signals are separated from the duration of transmission, and frequency modulation (such as chirp, nonlinear chirp, step frequency and frequency agility) and phase modulation (bi-phase modulation and multi-phase modulation) are generally used in the modulation mode, the nonlinear chirp is further divided into cosine modulation, tangential modulation, arctan modulation, hyperbolic modulation and parabolic modulation, the two-phase code sequence is further divided into barker code, M sequence and L sequence, and the multi-phase code sequence includes frank code and P-phase code (including P1, P2, P3 and P4-phase code). The phase is divided into coherent and non-coherent signals from the aspect of phase, the pulse repetition frequency variation mode is divided into repetition frequency jitter, repetition frequency sliding, repetition frequency dispersion and the like, the signals of the types can generate various radar signals through free combination, and the accuracy of identification is still difficult to improve by using a traditional radar signal identification method in the face of such complex and various radar signals.
Disclosure of Invention
Aiming at the problems in the traditional radar signal identification, the invention provides a radar signal identification method based on software and deep learning, a radar signal identification device based on software and deep learning and a computer device, which can greatly improve the accuracy of radar signal identification.
In order to achieve the above object, the embodiment of the present invention adopts the following technical scheme:
In one aspect, a method for identifying radar signals based on software and deep learning is provided, including the steps of:
acquiring a radar signal to be identified through software radio frequency equipment;
Inputting the radar signal to be identified into a trained radar signal identification model, outputting the signal type of the radar signal to be identified and determining the signal parameter of the radar signal to be identified, wherein the training and acquiring process of the radar signal identification model comprises the following steps:
acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
The method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels as a training set into a convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from a radar signal database, and inputting the radar signals without labels as a test set into the convolutional neural network model for model test;
and if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as a radar signal recognition model.
On the other hand, still provide a radar signal recognition device based on software and deep learning, include:
the signal acquisition module is used for acquiring radar signals to be identified through the software radio frequency equipment;
The radar signal recognition module is used for inputting the radar signal to be recognized into the trained radar signal recognition model, outputting the signal type of the radar signal to be recognized and determining the signal parameter of the radar signal to be recognized, wherein the training acquisition process of the radar signal recognition model comprises the following steps:
acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
The method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels as a training set into a convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from a radar signal database, and inputting the radar signals without labels as a test set into the convolutional neural network model for model test;
and if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as a radar signal recognition model.
In yet another aspect, a computer device is provided, including a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the software and deep learning based radar signal identification method described above when the processor executes the computer program.
One of the above technical solutions has the following advantages and beneficial effects:
According to the radar signal identification method, device and equipment based on software and deep learning, the convolutional neural network model based on deep learning is firstly constructed, and then a large amount of radar signal data of different categories are fed into the convolutional neural network model for training by combining the convenience and flexibility of software radio, so that the convolutional neural network model automatically performs characteristic extraction of signals, and the identification accuracy of radar signals is greatly improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments or the conventional techniques of the present invention, the drawings required for the descriptions of the embodiments or the conventional techniques will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the drawings without inventive effort for those skilled in the art.
FIG. 1 is a flow chart of a software and deep learning based radar signal identification method in one embodiment;
FIG. 2 is a flow diagram of a training process of a radar signal identification model in one embodiment;
FIG. 3 is a schematic diagram of a training system of a radar signal recognition method based on software and deep learning in one embodiment;
fig. 4 is a schematic block diagram of a radar signal identifying apparatus based on software and deep learning in an embodiment.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
It is noted that reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those skilled in the art will appreciate that the embodiments described herein may be combined with other embodiments. The term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
Embodiments of the present invention will be described in detail below with reference to the attached drawings in the drawings of the embodiments of the present invention.
In one embodiment, as shown in fig. 1, a method for identifying radar signals based on software and deep learning is provided, which may include the following processing steps S12 and S14:
s12, acquiring a radar signal to be identified through software radio frequency equipment;
S14, inputting the radar signal to be identified into the trained radar signal identification model, outputting the signal type of the radar signal to be identified and determining the signal parameters of the radar signal to be identified. As shown in fig. 2, the training acquisition process of the radar signal identification model includes steps S101 to S109:
S101, acquiring radar signals received by an antenna through a software radio frequency device, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types, wherein the radar signals can be acquired through the receiving antenna through down-conversion and then acquiring baseband signals, the signal types of the radar signals are priori knowledge, and after the signal types are determined by a transmitting end, different parameters of the signals need to be adjusted as much as possible so as to acquire different radar signals. Labeling the acquired signals by labeling the labels, namely labeling the acquired signals by corresponding signal categories. The radar signal database may include the absolute paths of the collected data of the radar signal and the corresponding signal types, as well as the duration of the collection (e.g., a fixed duration set according to the need of the collection) and the number of corresponding IQ samples (which refers to the result of down-converting a Radio Frequency (RF) signal to a lower Intermediate Frequency (IF) or baseband and then sampling it when the system receives the signal).
And S103, constructing a one-dimensional convolutional neural network model according to the set radar signal identification index, wherein the convolutional neural network model specifically comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer in the direction from data input to output, and when the existing network layers in the field are used, specific network parameters of the existing network layers can be set according to the actual application requirements, so long as the existing network layers can be used for executing required identification tasks.
S105, taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels into a convolutional neural network model as a training set to perform model training.
And S107, after the convolutional neural network model is trained to be converged to a set degree, taking out the radar signals without labels from the radar signal database as a test set to input the radar signals into the convolutional neural network model for model test, wherein the set degree of model convergence can be selected according to actual training requirements, for example, factors such as model training rounds, output effects, training time consumption and the like are comprehensively considered, and the corresponding convergence degree when the factors can be well balanced to achieve the optimal training effect is selected.
S109, if the recognition accuracy of the convolutional neural network model reaches a set threshold, finishing model training and taking the convolutional neural network model as a radar signal recognition model;
That is, if the accuracy of the convolutional neural network model output at its Softmax (multi-classification) layer reaches a set threshold (e.g., 99.5%), model training is ended and the convolutional neural network model is taken as a radar signal recognition model.
In some embodiments, if the recognition accuracy of the convolutional neural network model does not reach the set threshold, the method returns to step S105 to perform iterative training until the recognition accuracy of the convolutional neural network model reaches the set threshold, and then a radar signal recognition model is obtained.
It can be understood that by adopting a software-defined radio technology, various radar signal waveforms are generated by using software programming on an upper computer, waveform data are loaded into a unified hardware platform (existing software radio frequency equipment in the field) for real-time transmission, the power amplifier and the horn antenna are connected and then radiated on an air interface, and the other equipment can collect the received radar signals and tag the radar signals according to the signal types so as to train an identification model of the radar signals in a supervised learning mode.
Specifically, in a first step, various radar signals are transmitted based on software radio and various acquired signal files are stored according to signal categories.
And secondly, constructing a one-dimensional convolutional neural network model, wherein the network structure can be correspondingly innovated according to specific application requirements, such as, but not limited to, adding a migration learning layer, a attention mechanism layer, other functions and the like, and mainly adjusting parameters of the network model in the follow-up process to adapt to the current application scene. By adopting the convolutional neural network model structure of the embodiment, the problem of network degradation is alleviated by introducing the residual layer, and the problems of gradient elimination and explosion are avoided by introducing the batch normalization layer, so that the model training is faster, the feature extraction is deeper, and the classification effect is better.
Feeding the signal file marked in the first step to the model for large-scale training, testing the trained model, calculating the accuracy of identification, evaluating the generalization capability of the model by splitting the training set and the test set data in the existing cross folding mode, and returning to the second step if the generalization capability does not meet the given design requirement.
According to the radar signal identification method based on software and deep learning, the convolutional neural network model based on deep learning is firstly constructed, and then a large amount of radar signal data of different categories are fed into the convolutional neural network model for training by combining the convenience and flexibility of software radio, so that the convolutional neural network model automatically performs characteristic extraction of signals, and the identification accuracy of radar signals is greatly improved. Compared with the traditional technology, the scheme fully utilizes the flexibility of the software radio, is convenient to use various radar signal types, has rich signal parameters, such as the bandwidth, time width, signal to noise ratio and the like of the signal, can be adjusted at any time, is most important for a deep-learning convolutional neural network model, and directly influences the effect of final signal identification, so that the flexibility of the software radio can be combined to well meet the training requirements.
The convolutional neural network model has achieved a great deal of success in the field of image recognition, so that the advantages of the convolutional neural network model are used for radar signal recognition, various convolutional neural networks such as a residual neural network, hole convolution (dilation convolution), attention mechanisms and the like can be conveniently utilized, and the characteristics of radar signals required to be recognized in an actual application scene can be combined for improvement. The radar signal data used for training the model are real signal data which are directly acquired from an air interface and are not simulated, so that the method has good practical reference value for real radar signal identification.
In one embodiment, as shown in fig. 3, a radar signal received by a (directional receiving) antenna is transmitted by a target device through an air interface, and the target device comprises a host computer, a software radio frequency device, a power amplifier (i.e. a power amplifier) and a directional transmitting antenna which are sequentially connected, wherein a chirp signal generator, a non-chirp signal generator, a multi-phase coded signal generator, a step-frequency signal generator and a frequency agile signal generator are arranged in the host computer.
It can be understood that the data transmission can be performed between the upper computer and the software radio frequency device through interfaces such as USB3.0, gigabit network port or gigabit optical fiber, and the like, and similarly, the data transmission can also be performed between the signal acquisition device and the software radio frequency device through interfaces such as USB3.0, gigabit network port or gigabit optical fiber, and the like, at the receiving and identifying side of the radar signal, wherein the signal acquisition device, the CNN network model (namely, the convolutional neural network model) and the model evaluation module can be located on the same computer/server so as to be respectively used for bearing the functions of acquiring the radar signal in the radar signal identification method based on software and deep learning (the specific implementation of the acquisition function can be understood by referring to the radar signal acquisition process existing in the art), the training and running of the convolutional neural network model, the accuracy evaluation of the model, and the like.
The radar signal data for training is directly sent from the air interface through the target equipment, rather than simulated signal data, various radar signals corresponding to actual real radar signal identification scenes can be generated through combination of various types of signal generators deployed in an upper computer in a software programming mode, the signal types are flexibly configured and various, various real radar signal data can be provided without depending on complex various hardware circuit implementations, and therefore the obtained radar signal identification model can be used for model training based on real radar signals, and is more suitable for identification processing of real radar signals in the real application scenes, and finally accuracy of radar signal identification is further improved.
In one embodiment, the convolutional neural network model also introduces a layer of migration learning and attention mechanisms.
It can be appreciated that in this embodiment, a migration learning and attention mechanism layer is also introduced for the convolutional neural network model to achieve a stronger migration application capability for different signal recognition, and to enable the model to process input data more flexibly, focusing on specific information or features. Specifically, the model may be trimmed on the new recognition task data set by replacing the top layer of the pre-trained convolutional neural network model with a structure suitable for the new recognition task, such as a classification layer, to accommodate the requirements of the new recognition task, and may then use the new recognition task data set to train the model, adjusting parameters to optimize its performance. Attention modules can also be inserted between specific layer or layers in the backbone network of the convolutional neural network model, or attention mechanisms can be integrated into the whole structure of the model, so that the model can process input data more flexibly, and specific information or characteristics are focused. A
It should be understood that, although the steps in the flowcharts of fig. 1 and 2 are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 1 and 2 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor does the order in which the sub-steps or stages are performed necessarily occur in sequence, but may be performed alternately or alternately with at least a portion of the other steps or sub-steps of other steps.
In one embodiment, as shown in fig. 4, a radar signal identifying device based on software and deep learning is further provided, and the radar signal identifying device comprises a signal acquiring module 11 and an identifying output module 13. The signal acquisition module 11 is used for acquiring a radar signal to be identified through a software radio frequency device. The recognition output module 13 is used for inputting the radar signal to be recognized into the trained radar signal recognition model, outputting the signal type of the radar signal to be recognized and determining the signal parameters of the radar signal to be recognized.
The training acquisition process of the radar signal identification model comprises the following steps:
acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
The method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels as a training set into a convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from a radar signal database, and inputting the radar signals without labels as a test set into the convolutional neural network model for model test;
and if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as a radar signal recognition model.
According to the radar signal identification device based on software and deep learning, the convolutional neural network model based on deep learning is firstly constructed, and then a large amount of radar signal data of different categories are fed into the convolutional neural network model for training by combining the convenience and flexibility of software radio, so that the convolutional neural network model automatically performs characteristic extraction of signals, and the identification accuracy of radar signals is greatly improved.
In one embodiment, the training acquisition process of the radar signal identification model further comprises:
If the recognition accuracy of the convolutional neural network model does not reach the set threshold, returning to the step of taking out the radar signal with the tag from the radar signal database as a training set to input the radar signal with the tag into the convolutional neural network model for model training for iterative training until the recognition accuracy of the convolutional neural network model reaches the set threshold, and obtaining the radar signal recognition model.
In one embodiment, the convolutional neural network model also introduces a layer of migration learning and attention mechanisms.
In one embodiment, the radar signal received by the antenna is transmitted by the target device through the air interface, and the target device comprises a host computer, a software radio frequency device, a power amplifier and a directional transmitting antenna which are sequentially connected, wherein a signal generator for linear frequency modulation, a signal generator for non-linear frequency modulation, a signal generator for multi-phase coding, a signal generator for step frequency and a signal generator for frequency agility are arranged in the host computer.
For specific limitations of the software and deep learning based radar signal recognition device, reference may be made to the above limitation of the software and deep learning based radar signal recognition method, and the description thereof will not be repeated here. The above modules may be embedded in hardware or may be independent of a device with radar signal processing function, or may be stored in software in a memory of the device, so that the processor may call and execute operations corresponding to the above modules, where the device may be, but is not limited to, various radar signal processing devices existing in the art.
In one embodiment, a computer device is provided, comprising a memory storing a computer program and a processor that when executing the computer program performs the processing steps of obtaining radar signals to be identified by a software radio frequency device, inputting the radar signals to be identified into a trained radar signal identification model, outputting the signal type of the radar signals to be identified and determining the signal parameters of the radar signals to be identified.
The training acquisition process of the radar signal identification model comprises the following steps:
acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
The method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels as a training set into a convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from a radar signal database, and inputting the radar signals without labels as a test set into the convolutional neural network model for model test;
and if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as a radar signal recognition model.
In one embodiment, the processor may further implement the steps or sub-steps added to the embodiments of the software and deep learning based radar signal identification method described above when executing the computer program.
In one embodiment, a computer readable storage medium is provided having a computer program stored thereon, which when executed by a processor performs the processing steps of obtaining radar signals to be identified by a software radio frequency device, inputting the radar signals to be identified into a trained radar signal identification model, outputting a signal type of the radar signals to be identified and determining signal parameters of the radar signals to be identified.
The training acquisition process of the radar signal identification model comprises the following steps:
acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
The method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels as a training set into a convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from a radar signal database, and inputting the radar signals without labels as a test set into the convolutional neural network model for model test;
and if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as a radar signal recognition model.
In one embodiment, the computer program may be executed by the processor to implement the steps or sub-steps added in the embodiments of the software and deep learning based radar signal identification method.
Those skilled in the art will appreciate that implementing all or part of the above-described methods may be accomplished by way of a computer program, which may be stored on a non-transitory computer readable storage medium and which, when executed, may comprise the steps of the above-described embodiments of the methods. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (SYNCHLINK) DRAM (SLDRAM), memory bus dynamic random access memory (Rambus DRAM, RDRAM for short), and interface dynamic random access memory (DRDRAM), among others.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it is possible for those skilled in the art to make several variations and modifications without departing from the spirit of the present invention, which fall within the protection scope of the present invention. The scope of the invention should therefore be pointed out in the appended claims.
Claims (8)
1. A radar signal identification method based on software and deep learning is characterized by comprising the following steps:
acquiring a radar signal to be identified through software radio frequency equipment;
Inputting the radar signal to be identified into a trained radar signal identification model, outputting the signal type of the radar signal to be identified and determining the signal parameters of the radar signal to be identified, wherein the training and acquiring process of the radar signal identification model comprises the following steps:
Acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
the method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
Taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels serving as a training set into the convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from the radar signal database, and inputting the radar signals without labels into the convolutional neural network model as a test set for model test;
And if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as the radar signal recognition model.
2. The software and deep learning based radar signal identification method of claim 1, wherein the training acquisition process of the radar signal identification model further comprises:
if the recognition accuracy of the convolutional neural network model does not reach the set threshold, returning to the step of taking out the radar signal with the tag from the radar signal database as a training set and inputting the radar signal with the tag into the convolutional neural network model for model training for iterative training until the recognition accuracy of the convolutional neural network model reaches the set threshold, and obtaining the radar signal recognition model.
3. The software and deep learning based radar signal identification method of claim 1 or 2, wherein the convolutional neural network model further introduces a layer of migration learning and attention mechanisms.
4. A radar signal identification device based on software and deep learning, comprising:
the signal acquisition module is used for acquiring radar signals to be identified through the software radio frequency equipment;
The identification output module is used for inputting the radar signal to be identified into a trained radar signal identification model, outputting the signal type of the radar signal to be identified and determining the signal parameter of the radar signal to be identified, wherein the training acquisition process of the radar signal identification model comprises the following steps:
Acquiring radar signals received by an antenna through software radio frequency equipment, acquiring the radar signals for a fixed duration, labeling tags, and storing the radar signals in a radar signal database according to signal types;
the method comprises the steps of constructing a one-dimensional convolutional neural network model according to a set radar signal identification index, wherein the convolutional neural network model sequentially comprises a one-dimensional convolutional layer, a pooling layer, a batch normalization layer, a plurality of groups of residual error layers, an excitation layer, an average pooling layer and a full connection layer from data input to data output;
Taking out the radar signals with the labels from the radar signal database, and inputting the radar signals with the labels serving as a training set into the convolutional neural network model for model training;
After the convolutional neural network model is trained to be converged to a set degree, taking out radar signals without labels from the radar signal database, and inputting the radar signals without labels into the convolutional neural network model as a test set for model test;
And if the recognition accuracy of the convolutional neural network model reaches a set threshold, ending model training and taking the convolutional neural network model as the radar signal recognition model.
5. The software and deep learning based radar signal recognition device of claim 4, wherein the training acquisition process of the radar signal recognition model further comprises:
if the recognition accuracy of the convolutional neural network model does not reach the set threshold, returning to the step of taking out the radar signal with the tag from the radar signal database as a training set and inputting the radar signal with the tag into the convolutional neural network model for model training for iterative training until the recognition accuracy of the convolutional neural network model reaches the set threshold, and obtaining the radar signal recognition model.
6. The software and deep learning based radar signal recognition apparatus of claim 4 or 5, wherein the convolutional neural network model further incorporates a layer of migration learning and attention mechanisms.
7. The software and deep learning based radar signal identification device of claim 6, wherein the radar signal received by the antenna is transmitted by a target device through an air interface, the target device comprises a host computer, a software radio frequency device, a power amplifier and a directional transmitting antenna which are sequentially connected, and a chirp signal generator, a non-chirp signal generator, a multi-phase coded signal generator, a step frequency signal generator and a frequency agile signal generator are arranged in the host computer.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the software and deep learning based radar signal identification method of any one of claims 1 to 3 when the computer program is executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411708788.1A CN119199792A (en) | 2024-11-27 | 2024-11-27 | Radar signal recognition method, device and equipment based on software and deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411708788.1A CN119199792A (en) | 2024-11-27 | 2024-11-27 | Radar signal recognition method, device and equipment based on software and deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119199792A true CN119199792A (en) | 2024-12-27 |
Family
ID=94056973
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411708788.1A Pending CN119199792A (en) | 2024-11-27 | 2024-11-27 | Radar signal recognition method, device and equipment based on software and deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119199792A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112098957A (en) * | 2020-09-15 | 2020-12-18 | 西安电子科技大学 | Complex radar radiation source identification method based on one-dimensional self-walking convolution neural network |
| CN112232120A (en) * | 2020-09-10 | 2021-01-15 | 中国人民解放军海军工程大学 | A software radio-based radar radiation source signal classification system and method |
| US20210199797A1 (en) * | 2019-12-26 | 2021-07-01 | Samsung Electronics Co., Ltd. | Method and device to process radar signal |
| CN113406588A (en) * | 2021-05-14 | 2021-09-17 | 北京理工大学 | Joint modulation type identification and parameter estimation method for cognitive radar signals |
| CN113869121A (en) * | 2021-08-26 | 2021-12-31 | 中国人民解放军海军工程大学 | Radar waveform classification method and system based on effective region identification |
| CN115390037A (en) * | 2022-09-06 | 2022-11-25 | 中国人民解放军海军工程大学 | Multi-category, unknown radar emitter pulse signal sorting system |
-
2024
- 2024-11-27 CN CN202411708788.1A patent/CN119199792A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210199797A1 (en) * | 2019-12-26 | 2021-07-01 | Samsung Electronics Co., Ltd. | Method and device to process radar signal |
| CN112232120A (en) * | 2020-09-10 | 2021-01-15 | 中国人民解放军海军工程大学 | A software radio-based radar radiation source signal classification system and method |
| CN112098957A (en) * | 2020-09-15 | 2020-12-18 | 西安电子科技大学 | Complex radar radiation source identification method based on one-dimensional self-walking convolution neural network |
| CN113406588A (en) * | 2021-05-14 | 2021-09-17 | 北京理工大学 | Joint modulation type identification and parameter estimation method for cognitive radar signals |
| CN113869121A (en) * | 2021-08-26 | 2021-12-31 | 中国人民解放军海军工程大学 | Radar waveform classification method and system based on effective region identification |
| CN115390037A (en) * | 2022-09-06 | 2022-11-25 | 中国人民解放军海军工程大学 | Multi-category, unknown radar emitter pulse signal sorting system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wei et al. | PRI modulation recognition based on squeeze-and-excitation networks | |
| US11321966B2 (en) | Method and apparatus for human behavior recognition, and storage medium | |
| CN113065525B (en) | Age identification model training method, face age identification method and related device | |
| CN109408555B (en) | Data type identification method and device and data storage method and device | |
| CN111428448B (en) | Text generation method, device, computer equipment and readable storage medium | |
| CN112818891A (en) | Intelligent identification method for communication interference signal type | |
| CN110890088A (en) | Voice information feedback method and device, computer equipment and storage medium | |
| CN114529765B (en) | Data processing method, device and computer readable storage medium | |
| CN112395941B (en) | Radio signal radiation source individual identification method, computer device and storage medium | |
| CN115098556A (en) | User demand matching method and device, electronic equipment and storage medium | |
| CN116306574A (en) | Big data mining method and server applied to intelligent wind control task analysis | |
| CN116150651A (en) | AI-based depth synthesis detection method and system | |
| CN119199792A (en) | Radar signal recognition method, device and equipment based on software and deep learning | |
| CN118471268B (en) | Method, system and storage medium for intelligently identifying environmental noise category | |
| CN111931932A (en) | Method and device for generating countermeasure sample, electronic equipment and readable storage medium | |
| CN116739795A (en) | Knowledge graph-based insurance risk assessment method and device and electronic equipment | |
| CN113535815B (en) | Business operation behavior big data mining method and system suitable for electronic commerce | |
| CN117131329A (en) | Radio frequency fingerprint identification method based on homodromous orthogonal signals and contrast learning | |
| CN116306676A (en) | Information processing method for online business application and artificial intelligent service system | |
| CN112989603A (en) | Workflow adjusting method and device | |
| CN113707184B (en) | Method and device for determining emotion characteristics, electronic equipment and storage medium | |
| CN118035638B (en) | A drone detection method, device and equipment based on data-enhanced machine learning | |
| CN118709792B (en) | Multi-source telemetry data fusion method, device, electronic equipment and storage medium | |
| CN118094469B (en) | Fusion method, device, equipment and storage medium of multi-source heterogeneous data | |
| CN114202673B (en) | Training method of certificate classification model, certificate classification method, device and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |