CN111638488B - LSTM network-based radar interference signal identification method - Google Patents
LSTM network-based radar interference signal identification method Download PDFInfo
- Publication number
- CN111638488B CN111638488B CN202010281219.9A CN202010281219A CN111638488B CN 111638488 B CN111638488 B CN 111638488B CN 202010281219 A CN202010281219 A CN 202010281219A CN 111638488 B CN111638488 B CN 111638488B
- Authority
- CN
- China
- Prior art keywords
- interference signal
- lstm network
- radar interference
- value
- gate
- 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.)
- Active
Links
Images
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
- G01S7/021—Auxiliary means for detecting or identifying radar signals or the like, e.g. radar jamming signals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar Systems Or Details Thereof (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
Abstract
The invention relates to a radar interference signal identification method based on an LSTM network, which comprises the following steps: s1: and simulating to generate a time-frequency domain sampling sequence of a plurality of different radar interference signals as a data set, dividing the data set into a training sample, a test sample and a verification sample, wherein each sample is provided with a tag value. S2: constructing an LSTM network and initializing the LSTM network; s3: training the LSTM according to the training sample to obtain a radar interference signal identification model; s4: and respectively inputting the test sample and the verification sample into the radar interference signal identification model to test the performance of the radar interference signal identification model. S5: and inputting the time-frequency domain sequence of the radar interference signal to be identified into a radar interference signal identification model, so as to realize the identification and judgment of the radar interference signal. The LSTM network-based radar interference signal identification method does not need complex signal preprocessing work such as filtering, GWT conversion, carrier frequency estimation and the like on an interference signal.
Description
Technical Field
The invention belongs to the technical field of radar signal identification, and particularly relates to a radar interference signal identification method based on an LSTM network.
Background
With the development of the electronic information field, electronic countermeasure plays an important role in electronic information reconnaissance, electronic support and threat alert systems, and in the increasingly serious background of modern electronic war, accurate judgment of the type of an interference signal has important significance for implementing purposeful anti-interference measures on electronic systems such as radars. Along with the implementation of interference and the development of anti-interference technology, a radar system is required to be capable of rapidly and accurately identifying interference signals. The existing interference identification method based on the neural network mainly comprises a BP network, a decision tree and a deep learning method for introducing the convolutional neural network.
For the traditional identification method, the characteristic value is usually manually extracted by experience for identification, so that a large amount of workload for signal preprocessing and characteristic extraction is required, and the identification accuracy is greatly affected by noise. The interference signal identification method based on the convolutional neural network needs to perform various transformations on signals to generate image data as samples, and has huge data set and easy overfitting phenomenon. When the deep learning method is applied to interference signal recognition, the accuracy of the recognition algorithm is greatly improved compared with that of the prior neural network algorithm, but the signal characteristic value still needs to be extracted, and the time-frequency domain image of the signal needs to be obtained through conversion as input, so that a great deal of manpower and time cost are required to be input in the early stage, and the requirement of combat instantaneity cannot be met.
Therefore, the radar interference signal identification method with accurate and efficient identification and simple input training sample and without complex signal preprocessing in the early stage is a problem to be solved urgently.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a radar interference signal identification method based on an LSTM network. The technical problems to be solved by the invention are realized by the following technical scheme:
the invention provides a radar interference signal identification method based on an LSTM network, which comprises the following steps:
s1: and simulating to generate a time-frequency domain sampling sequence of a plurality of different radar interference signals as a data set, dividing the data set into a training sample, a test sample and a verification sample, wherein each sample is provided with a tag value.
S2: constructing an LSTM network and initializing the LSTM network;
s3: training the LSTM according to the training sample to obtain a radar interference signal identification model;
s4: and respectively inputting the test sample and the verification sample into the radar interference signal identification model to verify the performance of the radar interference signal identification model.
S5: and inputting the time-frequency domain sequence of the radar interference signal to be identified into the radar interference signal identification model to realize the identification and judgment of the radar interference signal.
In one embodiment of the invention, the number of different radar interference signals includes: broadband blocking interference signals, narrowband aiming interference signals, sweep frequency interference signals, dressing spectrum interference signals, on-off modulation interference signals and smart noise interference signals; the dividing ratio of the training sample, the test sample and the verification sample is 8:1:1.
In one embodiment of the invention, the LSTM network includes an input gate, a forget gate, and an output gate connected in sequence;
initializing the LSTM network includes: setting the initial learning rate of the LSTM network to be 0.001, setting the number of hidden layer units of the LSTM network to be 50, setting the maximum iteration number of the LSTM network to be 40, and adopting an Adam algorithm to be a gradient descent algorithm to control the learning rate.
In one embodiment of the present invention, the S3 includes:
s31: performing an iteration according to the training samples, inputting a group of training samples into an input gate of the LSTM network, and calculating an activation value i of the input gate t State candidate value of memory cell at time t
The activation function and the state candidate function of the input gate are as follows:
i t ＝σ(w i x t +u i h t-1 +b i )，
wherein x is t Indicating the input of the memory unit at the time t, h t-1 Representing the input value of the input gate at time t-1, sigma representing the sigmoid activation function, tanh representing the hyperbolic tangent activation function, w i ,u i Respectively represent the weights of the input gate activation functions, b i Representing the bias of the input gate activation function, w c ,u c Weights, b, respectively representing input gate state candidate functions c Representing the bias of the input gate state candidate function;
s32: calculating the information to be discarded or stored by the memory unit, and forgetting the gate according to the activation value i of the input gate t State candidate value of memory cell at time t
Calculating to obtain a new state value C t The calculation formula is as follows:
f t ＝σ(w f x t +u f h t-1 +b f )，
wherein C is t Representing a new state value, f t Indicating the activation value of the forgetting gate, w f ,u f Weights respectively representing the activation functions of the forgetting gate, b f A bias representing a forgetting gate activation function;
s33: according to the new state value C t Calculating the activation value o of the output gate at the time t t Output value h t The calculation formula is as follows:
o t ＝σ(w o x t +u o h t-1 +b o )，
h t ＝o t *tanh(C t )，
wherein w is o ,u o Respectively representing the control weights of the output gate activation functions, b o Representing the bias of the output gate activation function;
s34: according to the output value h t Calculating a loss function of the LSTM network with the label value of the training sample, if the iteration number is smaller than the maximum iteration number, updating the weight and the bias of the LSTM network, adding 1 to the iteration number, and repeating the steps S31-S34 until the recognition accuracy of the LSTM network model to the training sample is greater than 99% or is overlappedAnd stopping iteration if the generation times are greater than or equal to the maximum iteration times, and obtaining the radar interference signal identification model.
In one embodiment of the present invention, the S5 includes:
s51: sampling the signals received by the radar system to obtain time domain sequence data, and calculating the time domain sequence data through Fourier transformation;
s52: and inputting the time domain sequence data and the frequency domain sequence data into the radar interference signal identification model to obtain the label of the output interference signal.
Compared with the prior art, the invention has the beneficial effects that:
1. compared with the traditional interference signal identification method based on the neural network, the radar interference signal identification method based on the LSTM does not need complex signal preprocessing work such as filtering, GWT conversion, carrier frequency estimation and the like on the interference signal.
2. According to the LSTM network-based radar interference signal identification method, for LSTM network training, network training can be completed by only utilizing the time domain and the frequency domain sampling sequence of signals, and the sample is simple and the data complexity is low.
The foregoing description is only an overview of the present invention, and is intended to be implemented in accordance with the teachings of the present invention, as well as the preferred embodiments thereof, together with the following detailed description of the invention, given by way of illustration only, together with the accompanying drawings.
Drawings
Fig. 1 is a schematic flow chart of a radar interference signal identification method based on an LSTM network according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of LSTM network training provided by an embodiment of the invention;
FIG. 3 is a graph showing the variation of a loss function with the number of iterations provided by an embodiment of the present invention;
FIG. 4 is a graph showing the variation of recognition accuracy with iteration number according to the embodiment of the present invention;
fig. 5 is a graph showing the change of the recognition probability of the radar interference signal recognition model provided by the embodiment of the invention to different types of interference signals along with the dry-noise ratio.
Detailed Description
In order to further explain the technical means and effects adopted by the invention to achieve the preset aim, the following describes a radar interference signal identification method based on an LSTM network according to the invention in detail with reference to the attached drawings and the detailed description.
The foregoing and other features, aspects, and advantages of the present invention will become more apparent from the following detailed description of the preferred embodiments when taken in conjunction with the accompanying drawings. The technical means and effects adopted by the present invention to achieve the intended purpose can be more deeply and specifically understood through the description of the specific embodiments, however, the attached drawings are provided for reference and description only, and are not intended to limit the technical scheme of the present invention.
Example 1
Referring to fig. 1 and fig. 2 in combination, fig. 1 is a schematic flow chart of a radar interference signal identifying method based on an LSTM network according to an embodiment of the present invention, and fig. 2 is a schematic flow chart of LSTM network training according to an embodiment of the present invention. As shown in the figure, the radar interference signal identification method based on the LSTM network in the embodiment of the invention includes:
s1: and simulating to generate a time-frequency domain sampling sequence of a plurality of different radar interference signals as a data set, dividing the data set into a training sample, a test sample and a verification sample, wherein each sample is provided with a tag value.
Specifically, in this embodiment, the several different radar interference signals include: broadband blocking interference signals, narrowband aiming interference signals, sweep interference signals, vanity spectrum interference signals, on-off modulation interference signals, and smart noise interference signals, i.e., class 6 radar interference signals. 200 x 900 signal sample data are generated by each type of interference, wherein 200 represents that 200 interference signals are repeatedly generated by each type of interference as training, 900 represents sampling values, and then the data set of each type of radar interference signals is finally obtained as a 3-dimensional matrix, the first dimension is signal category, the second dimension is time domain real number sequence data of signals, and the third dimension is signal frequency domain real number sequence data.
Further, the dividing ratio of the training sample, the test sample and the verification sample is 8:1:1. In order to accelerate the training speed of the LSTM network and avoid the situation of being in a local optimal solution, in the training process, the training samples are divided into a plurality of subsets, the size of each subset is 40, and data of one subset is selected for training in each iteration, so that the training speed can be accelerated.
S2: constructing an LSTM network and initializing the LSTM network;
specifically, the LSTM network includes an input gate, a forget gate, and an output gate connected in sequence. The gate is used for allowing the memory unit of the LSTM network to store the interference signal sequence information for a long time, so that the problem of gradient disappearance can be effectively reduced. Wherein the input gate control determines information that is allowed to be updated; the forget gate controls the information to be saved or abandoned; the output gate then decides that information can be output.
In this implementation, initializing the LSTM network includes: setting the initial learning rate of the LSTM network to be 0.001, setting the number of hidden layer units of the LSTM network to be 50, setting the maximum iteration number of the LSTM network to be 40, and adopting an Adam algorithm to be a gradient descent algorithm to control the learning rate. Typically, the LSTM network has an input gate size equal to the input sample data dimension and an output gate equal to the number of categories to be identified.
S3: training the LSTM according to the training sample to obtain a radar interference signal identification model;
preferably, the data is randomly shuffled to reduce the overfitting phenomenon as each subset is trained. Then inputting the LSTM network, and outputting a category with the highest probability by using a softmax regression mode when the LSTM network is trained each time. Specifically, the LSTM network calculates the probability that the interference signal belongs to a certain class at the output layer, selects one with the largest probability as the output class, compares with the label of the actual signal and adjusts the LSTM network to retrain. And stopping training when the accuracy of the LSTM network on the training sample meets the requirement (more than 99 percent) or the maximum iteration number is reached, and storing the trained LSTM network. It should be noted that when the LSTM network reaches the maximum number of iterations but the accuracy is not sufficient, it is generally necessary to modify the number of hidden layer units to readjust the training network.
Specifically, the S3 includes:
s31: performing an iteration based on the training samples, inputting a set (a subset) of the training samples into the input gates of the LSTM network to calculate the activation value i of the input gates t State candidate value of memory cell at time t
The activation function and the state candidate function of the input gate are as follows:
i t ＝σ(w i x t +u i h t-1 +b i ) (1)，
wherein x is t Indicating the input of the memory unit at the time t, h t-1 Representing the input value of the input gate at time t-1, sigma representing the sigmoid activation function, tanh representing the hyperbolic tangent activation function, w i ,u i Respectively represent the weights of the input gate activation functions, b i Representing the bias of the input gate activation function, w c ,u c Weights, b, respectively representing input gate state candidate functions c Representing the bias of the input gate state candidate function;
s32: calculating the information to be discarded or stored by the memory unit, and forgetting the gate according to the activation value i of the input gate t State candidate value of memory cell at time t
Calculating to obtain a new state value C t The calculation formula is as follows:
f t ＝σ(w f x t +u f h t-1 +b f ) (4)，
wherein C is t Representing a new state value, f t Indicating the activation value of the forgetting gate, w f ,u f Weights respectively representing the activation functions of the forgetting gate, b f A bias representing a forgetting gate activation function;
in this embodiment, the forgetting gate needs to read the input x of the memory cell at time t t And an input value h at time t-1 t-1 And outputting a value between 0 and 1 as the state of the memory cell at the previous time, and then forgetting the gate and passing through the activation value i of the input gate obtained in the step S31 t State candidate value of memory cell at time t
And calculating to obtain a new state value.
S33: according to the new state value C t Calculating the activation value o of the output gate at the time t t Output value h t The calculation formula is as follows:
o t ＝σ(w o x t +u o h t-1 +b o ) (5)，
h t ＝o t *tanh(C t ) (6)，
wherein w is o ,u o Respectively representing the control weights of the output gate activation functions, b o Representing the bias of the output gate activation function;
in an embodiment, the activation value o of the gate is output t The latest state value C at the moment of t t Multiplied by the tanh function (hyperbolic tangent activation function) to finally obtain the output value h of the output gate t 。
S34: according to the output value h t Calculating a loss function of the LSTM network with the label value of the training sample if iteratingAnd when the number of times is smaller than the maximum iteration number, updating the weight and the bias of the LSTM network, adding 1 to the iteration number, and repeating the steps S31-S34 until the recognition accuracy of the LSTM network model to the training sample is greater than 99% or the iteration number is greater than or equal to the maximum iteration number, stopping iteration, and obtaining the radar interference signal recognition model.
Referring to fig. 3 and fig. 4 in combination, fig. 3 is a plot of a loss function versus iteration number according to an embodiment of the present invention; fig. 4 is a graph showing a change of recognition accuracy with iteration number according to an embodiment of the present invention. As shown in the figure, in the actual training process, the LSTM network of this embodiment divides the whole training sample data set into several subsets to train in batches, so that many "burrs" will occur on the loss function curve, but the overall variation trend of the loss function curve with the increase of the iteration number is gradually reduced and finally reaches to about 0, as shown in fig. 3. The overall trend of the corresponding recognition accuracy curve increases gradually with the increase of the iteration number, as shown in fig. 4.
It should be noted that, when the network parameters of the LSTM network are not suitable for a specific problem, the recognition accuracy may decrease with the increase of the number of iterations, and at this time, it is necessary to choose whether to terminate training in advance by observing the loss function curve and the change condition of the recognition accuracy curve.
S4: respectively inputting the test sample and the verification sample into the radar interference signal identification model to verify the performance of the radar interference signal identification model;
in this embodiment, the test sample is used to test various hyper-parameter values in the learning process, and the verification sample is mainly used to verify the generalization ability of the model.
S5: and inputting the time-frequency domain sequence of the radar interference signal to be identified into the radar interference signal identification model to realize the identification and judgment of the radar interference signal.
Specifically, the method comprises the following steps:
s51: sampling the signals received by the radar system to obtain time domain sequence data, and calculating the time domain sequence data through Fourier transformation;
in this embodiment, the simulation software is used to generate an interference plus noise signal without a tag, which is used to simulate the signal received by the radar system.
S52: and inputting the time domain sequence data and the frequency domain sequence data into the radar interference signal identification model to obtain the label of the output interference signal.
Referring to fig. 5, fig. 5 is a plot of the recognition probability of different types of interference signals according to the interference-to-noise ratio by the radar interference signal recognition model according to the embodiment of the present invention. In practical application, the noise signal has the greatest influence on the identification of the interference signal, so the validity of the radar interference signal identification model is checked through the change of the dry-to-noise ratio, and the radar interference signal identification model of the embodiment has better identification accuracy on different types of interference signals.
Compared with the traditional interference signal identification method based on the neural network, the radar interference signal identification method based on the LSTM does not need complex signal preprocessing work such as filtering, GWT conversion, carrier frequency estimation and the like on the interference signal. And when the LSTM network is trained, the network training can be completed by only using the time domain and the frequency domain sampling sequences of the signals, and the sample is simple and the data complexity is low.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (4)
1. The radar interference signal identification method based on the LSTM network is characterized by comprising the following steps of:
s1: simulating to generate a time-frequency domain sampling sequence of a plurality of different radar interference signals as a data set, dividing the data set into a training sample, a test sample and a verification sample, wherein each sample is provided with a tag value; wherein the plurality of different radar interference signals includes: broadband blocking interference signals, narrowband aiming interference signals, sweep frequency interference signals, dressing spectrum interference signals, on-off modulation interference signals and smart noise interference signals;
s2: constructing an LSTM network and initializing the LSTM network;
s3: training the LSTM according to the training sample to obtain a radar interference signal identification model; the step S3 comprises the following steps:
s31: performing an iteration according to the training samples, inputting a group of training samples into an input gate of the LSTM network, and calculating an activation value i of the input gate t State candidate value of memory cell at time t
The activation function and the state candidate function of the input gate are as follows:
i t ＝σ(w i x t +u i h t-1 +b i )，
wherein x is t Indicating the input of the memory unit at the time t, h t-1 Representing the input value of the input gate at time t-1, sigma representing the sigmoid activation function, tanh representing the hyperbolic tangent activation function, w i ,u i Respectively represent the weights of the input gate activation functions, b i Representing the bias of the input gate activation function, w c ,u c Weights, b, respectively representing input gate state candidate functions c Representing the bias of the input gate state candidate function;
s32: calculating the information to be discarded or stored by the memory unit, and forgetting the gate according to the activation value i of the input gate t State candidate value of memory cell at time t
Calculating to obtain a new state value C t The calculation formula is as follows:
f t ＝σ(w f x t +u f h t-1 +b f )，
wherein C is t Representing a new state value, f t Indicating the activation value of the forgetting gate, w f ,u f Weights respectively representing the activation functions of the forgetting gate, b f A bias representing a forgetting gate activation function;
s33: according to the new state value C t Calculating the activation value o of the output gate at the time t t Output value h t The calculation formula is as follows:
o t ＝σ(w o x t +u o h t-1 +b o )，
h t ＝o t *tanh(C t )，
wherein w is o ,u o Respectively representing the control weights of the output gate activation functions, b o Representing the bias of the output gate activation function;
s34: according to the output value h t Calculating a loss function of the LSTM network with the label value of the training sample, updating the weight and the bias of the LSTM network if the iteration number is smaller than the maximum iteration number, adding 1 to the iteration number, and repeating the steps S31-S34 until the recognition accuracy of the LSTM network model to the training sample is greater than 99% or the iteration number is greater than or equal to the maximum iteration number, and stopping iteration to obtain the radar interference signal recognition model;
s4: respectively inputting the test sample and the verification sample into the radar interference signal identification model to verify the performance of the radar interference signal identification model;
s5: and inputting the time-frequency domain sequence of the radar interference signal to be identified into the radar interference signal identification model to realize the identification and judgment of the radar interference signal.
2. The LSTM network-based radar interference signal identification method according to claim 1, wherein a division ratio of the training sample, the test sample, and the verification sample is 8:1:1.
3. The method for identifying radar interference signals based on LSTM network as claimed in claim 1,
the LSTM network comprises an input door, a forget door and an output door which are connected in sequence;
initializing the LSTM network includes: setting the initial learning rate of the LSTM network to be 0.001, setting the number of hidden layer units of the LSTM network to be 50, setting the maximum iteration number of the LSTM network to be 40, and adopting an Adam algorithm to be a gradient descent algorithm to control the learning rate.
4. The LSTM network-based radar interference signal identification method according to claim 1, wherein the S5 includes:
s51: sampling the signals received by the radar system to obtain time domain sequence data, and calculating the time domain sequence data through Fourier transformation;
s52: and inputting the time domain sequence data and the frequency domain sequence data into the radar interference signal identification model to obtain the label of the output interference signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010281219.9A CN111638488B (en) | 2020-04-10 | 2020-04-10 | LSTM network-based radar interference signal identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010281219.9A CN111638488B (en) | 2020-04-10 | 2020-04-10 | LSTM network-based radar interference signal identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111638488A CN111638488A (en) | 2020-09-08 |
| CN111638488B true CN111638488B (en) | 2023-05-30 |
Family
ID=72330583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010281219.9A Active CN111638488B (en) | 2020-04-10 | 2020-04-10 | LSTM network-based radar interference signal identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111638488B (en) |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112202527B (en) * | 2020-10-01 | 2022-09-13 | 西北工业大学 | Intelligent electromagnetic signal identification system interference method based on momentum gradient disturbance |
| CN112415510B (en) * | 2020-11-05 | 2023-08-04 | 深圳大学 | Dual-station radar gesture recognition method, device, system and storage medium |
| CN112491442B (en) * | 2020-11-17 | 2021-12-28 | 中山大学 | Self-interference elimination method and device |
| CN112986950B (en) * | 2020-12-25 | 2024-09-06 | 南京理工大学 | Single-pulse laser radar echo feature extraction method based on deep learning |
| CN113067805B (en) * | 2021-03-15 | 2022-05-24 | 衢州学院 | Internet of things weak electromagnetic interference attack detection method and system based on edge calculation |
| CN113030902B (en) * | 2021-05-08 | 2022-05-17 | 电子科技大学 | Twin complex network-based few-sample radar vehicle target identification method |
| CN113645700A (en) * | 2021-07-15 | 2021-11-12 | 中国移动通信集团广东有限公司广州分公司 | Deep learning-based resource allocation method and device for improving SCMA system performance |
| CN113419225B (en) * | 2021-08-19 | 2023-02-28 | 北京九天利建信息技术股份有限公司 | Radar interference effect online evaluation method |
| CN113887506B (en) * | 2021-10-22 | 2024-04-30 | 吉林大学 | MDS-CNN-based interference signal classification and identification method and system |
| CN114492583B (en) * | 2021-12-28 | 2022-12-20 | 中国电子科技集团公司第二十九研究所 | LSTM-based radar search mode identification method |
| CN114509730B (en) * | 2022-01-14 | 2024-07-16 | 西安电子科技大学 | ELM-based interference frequency domain state online prediction method |
| CN114938232B (en) * | 2022-06-15 | 2023-04-07 | 北京邮电大学 | LSTM-based simultaneous co-frequency full-duplex digital domain self-interference suppression method |
| CN115840875A (en) * | 2022-11-10 | 2023-03-24 | 北京擎天信安科技有限公司 | Millimeter wave radar abnormal signal detection method and system based on analog transducer |
| CN115865129B (en) * | 2022-12-01 | 2024-03-29 | 电子科技大学 | Narrowband interference intelligent elimination method based on denoising self-encoder |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826630A (en) * | 2019-11-08 | 2020-02-21 | 哈尔滨工业大学 | Radar interference signal feature level fusion identification method based on deep convolutional neural network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10691962B2 (en) * | 2017-09-22 | 2020-06-23 | Toyota Motor Engineering & Manufacturing North America, Inc. | Systems and methods for rear signal identification using machine learning |
-
2020
- 2020-04-10 CN CN202010281219.9A patent/CN111638488B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826630A (en) * | 2019-11-08 | 2020-02-21 | 哈尔滨工业大学 | Radar interference signal feature level fusion identification method based on deep convolutional neural network |
Non-Patent Citations (1)
| Title |
|---|
| 基于LSTM的雷达辐射源识别技术;刘括然;《舰船电子工程》(第12期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111638488A (en) | 2020-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111638488B (en) | LSTM network-based radar interference signal identification method | |
| Zhang et al. | A fault diagnosis method for wind turbines gearbox based on adaptive loss weighted meta-ResNet under noisy labels | |
| CN109597043B (en) | Radar signal identification method based on quantum particle swarm convolutional neural network | |
| CN110133599B (en) | Intelligent radar radiation source signal classification method based on long-time and short-time memory model | |
| CN110222748B (en) | OFDM radar signal identification method based on 1D-CNN multi-domain feature fusion | |
| CN111126134B (en) | Radar radiation source deep learning identification method based on non-fingerprint signal eliminator | |
| CN110334741A (en) | Radar range profile's recognition methods based on Recognition with Recurrent Neural Network | |
| CN109766791B (en) | Communication signal modulation identification method based on self-encoder | |
| CN108052968B (en) | QSFLA-SVM perception intrusion detection method | |
| CN113761805B (en) | Controllable source electromagnetic data denoising method, system, terminal and readable storage medium based on time domain convolution network | |
| CN112949383B (en) | Waveform agility radar radiation source identification method based on HyDeep-Att network | |
| CN113435247A (en) | Intelligent identification method, system and terminal for communication interference | |
| CN110082738B (en) | Radar target identification method based on Gaussian mixture and tensor recurrent neural network | |
| CN108957418A (en) | A kind of radar target identification method based on Recognition with Recurrent Neural Network model | |
| CN114527441A (en) | Radar signal identification method of LSTM network based on multi-head attention mechanism | |
| CN110889207B (en) | Deep learning-based intelligent assessment method for credibility of system combination model | |
| Zhang et al. | A deep learning approach for modulation recognition | |
| CN113378910B (en) | Poisoning attack method for identifying electromagnetic signal modulation type based on pure label | |
| CN113406574B (en) | Online clustering method for multifunctional radar working mode sequence | |
| CN113048012B (en) | Wind turbine generator yaw angle identification method and device based on Gaussian mixture model | |
| CN117592370A (en) | PINN-based two-dimensional forming field solving method, system and storage medium | |
| CN116340846A (en) | Aliasing modulation signal identification method for multi-example multi-label learning under weak supervision | |
| CN114463576B (en) | Network training method based on re-weighting strategy | |
| CN114372495A (en) | Electric energy quality disturbance classification method and system based on deep space residual error learning | |
| CN110457863B (en) | Parameter estimation method for wind driven generator pitch subsystem based on ellipsoid contraction filtering |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |