CN113723358A - Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment - Google Patents
Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment Download PDFInfo
- Publication number
- CN113723358A CN113723358A CN202111080309.2A CN202111080309A CN113723358A CN 113723358 A CN113723358 A CN 113723358A CN 202111080309 A CN202111080309 A CN 202111080309A CN 113723358 A CN113723358 A CN 113723358A
- Authority
- CN
- China
- Prior art keywords
- signal
- sample
- countermeasure
- classification result
- reconstruction
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 99
- 238000012549 training Methods 0.000 claims abstract description 203
- 238000000034 method Methods 0.000 claims abstract description 50
- 238000013145 classification model Methods 0.000 claims abstract description 23
- 238000012360 testing method Methods 0.000 claims description 72
- 230000003042 antagnostic effect Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 14
- 230000008054 signal transmission Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 19
- 238000003860 storage Methods 0.000 description 15
- 238000004590 computer program Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000003062 neural network model Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000013136 deep learning model Methods 0.000 description 4
- 238000009826 distribution Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000282412 Homo Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000009982 effect on human Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application discloses a method and a device for detecting a countermeasure signal based on a generation countermeasure network and electronic equipment. The method comprises the following steps: acquiring a signal training sample; generating a first reconstruction sample corresponding to the signal training sample by using a generation countermeasure network based on the signal training sample; classifying the signal training sample and the first reconstruction sample by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample; determining a confrontation signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample; the countermeasure signal is detected based on a countermeasure signal detection threshold. The countermeasure signal detection method based on the generation countermeasure network can accurately detect the countermeasure signal, reduces misleading and loss caused by the existence of the countermeasure signal in the signal demodulation process, effectively reduces risks brought by the countermeasure signal, and strengthens safety and secrecy of the signal transmission process.
Description
Technical Field
The present application relates to the field of signal detection technologies, and in particular, to a method and an apparatus for detecting a countermeasure signal based on a generation countermeasure network, and an electronic device.
Background
With the development of scientific progress and hardware technology, the big data era has come along, and topics such as artificial intelligence and machine learning have become hot spots for the next discussion. Especially, in recent years, deep learning models have achieved great success in the aspect of feature extraction of data.
Relevant studies have shown that it is possible to design slightly perturbed, antagonistic samples for known deep learning models, such that the models are identified and classified incorrectly, and although these forged samples have no significant effect on human judgment, they are fatal misleading for deep learning models, often causing deep models to produce human unexpected results. For example, in the field of signal modulation type classification, the classification model judges the label of the QAM16 modulated signal as the label of the QAM64 modulation type. Recently, a series of antagonistic attacks successfully implemented in the real world have demonstrated that this problem is a safety hazard for all deep learning based systems. Therefore, researches on the sample detection technology attract more and more attention of researchers in the field of machine learning and safety, and particularly, researches and exploration on deep learning have good guiding effects on future application and practice.
Although Deep Neural Networks (DNNs) perform well on some complex problems such as speech recognition, signal modulation type classification, etc., they are susceptible to well-designed perturbations. Typically, these perturbations are imperceptible to humans, but they can make the model misjudge with a higher confidence. For example, in practical applications, a signal transmitting base station transmits radio signals to a target base station, and the signals have great application value, and if the signals are intercepted maliciously in the middle and are judged by a machine learning means to be transmitted again after being adjusted elaborately (the signals are called countermeasure signals), the signals will bring great potential threats to receivers of the signals, and therefore, it is important to detect which signals are countermeasure signals.
Disclosure of Invention
In view of the above, a main object of the present application is to provide a method, an apparatus and an electronic device for detecting a countermeasure signal based on a generated countermeasure network, so as to solve the technical problem in the prior art that detection of the countermeasure signal is not accurate enough.
According to a first aspect of the present application, there is provided a method for detecting a countermeasure signal based on generation of a countermeasure network, comprising:
acquiring a signal training sample;
generating a first reconstruction sample corresponding to the signal training sample by using the generation countermeasure network based on the signal training sample;
classifying the signal training sample and the first reconstruction sample by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
determining a confrontation signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample;
detecting a counter signal according to the counter signal detection threshold.
Optionally, the generating of the countermeasure network is trained by:
acquiring the signal training sample, wherein the signal training sample comprises a training data label;
generating a second reconstruction sample with the generator generating a countermeasure network based on the training data labels and random noise data;
judging the signal training sample and the second reconstruction sample by using a discriminator in the generated countermeasure network to obtain a judgment result;
and updating the parameters of the generated countermeasure network according to the judgment result to obtain the trained generated countermeasure network.
Optionally, the training data labels comprise training data labels of a plurality of signal training samples, and the generating, with the generator for generating a countermeasure network, a second reconstruction sample based on the training data labels and random noise data comprises:
initializing a plurality of random noise values corresponding to a target signal training sample for a training data label of the target signal training sample, wherein the target signal training sample is any one of a plurality of signal training samples;
generating a plurality of second reconstruction samples corresponding to the target signal training samples by using the generator for generating the countermeasure network based on a plurality of random noise values corresponding to the target signal training samples;
and determining a final second reconstruction sample corresponding to the target signal training sample according to the similarity between the target signal training sample and each second reconstruction sample.
Optionally, the classification result of the signal training sample includes predicted values of a plurality of signal training samples, the classification result of the first reconstructed sample includes predicted values of a plurality of first reconstructed samples, and the determining the counter signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstructed sample includes:
respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values;
an average of the plurality of error values is determined as the countermeasure signal detection threshold.
Optionally, the detecting a counter signal according to the counter signal detection threshold comprises:
acquiring a signal test sample;
generating a third reconstructed sample by using the generator for generating the countermeasure network according to the signal test sample;
classifying the signal test sample and the third reconstruction sample by using the preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample;
and detecting the countermeasure signal in the signal test sample by using the countermeasure signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
Optionally, the classification result of the signal test sample and the classification result of the third reconstructed sample both include predicted values, and the detecting, according to the classification result of the signal test sample and the classification result of the third reconstructed sample, the countermeasure signal in the signal test sample using the countermeasure signal detection threshold includes:
determining an error value of the predicted value of the signal test sample and the predicted value of the third reconstructed sample;
determining the signal test sample as a challenge signal if the error value is greater than the challenge signal detection threshold;
determining that the signal test sample is a non-countermeasure signal if the error value is not greater than the countermeasure signal detection threshold.
According to a second aspect of the present application, there is provided a countermeasure signal detection apparatus based on generation of a countermeasure network, comprising:
the acquisition unit is used for acquiring a signal training sample;
a generating unit, configured to generate, based on the signal training sample, a first reconstruction sample corresponding to the signal training sample by using the generative countermeasure network;
the classification unit is used for classifying the signal training sample and the first reconstruction sample respectively by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
a determining unit, configured to determine a countermeasure signal detection threshold according to a classification result of the signal training sample and a classification result of the first reconstruction sample;
a detection unit for detecting a countermeasure signal according to the countermeasure signal detection threshold.
Optionally, the classification result of the signal training samples includes predicted values of a plurality of signal training samples, and the determining unit is specifically configured to:
respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values;
an average of the plurality of error values is determined as the countermeasure signal detection threshold.
Optionally, the detection unit is specifically configured to:
acquiring a signal test sample;
generating a third reconstructed sample by using the generator for generating the countermeasure network according to the signal test sample;
classifying the signal test sample and the third reconstruction sample by using the preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample;
and detecting the countermeasure signal in the signal test sample by using the countermeasure signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
In accordance with a third aspect of the present application, there is provided an electronic device comprising: a processor, a memory storing computer-executable instructions,
the executable instructions, when executed by the processor, implement the aforementioned countermeasure signal detection method based on generation of a countermeasure network.
According to a fourth aspect of the present application, there is provided a computer-readable storage medium storing one or more programs which, when executed by a processor, implement the foregoing countermeasure signal detection method based on generation of a countermeasure network.
The beneficial effect of this application is: according to the countermeasure signal detection method based on the generated countermeasure network, a signal training sample is obtained firstly; then based on the signal training sample, generating a first reconstruction sample corresponding to the signal training sample by using the generation countermeasure network; classifying the signal training sample and the first reconstruction sample by using a preset classification model respectively to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample; then determining a counter signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample; finally, a counter signal is detected according to the counter signal detection threshold. The countermeasure signal detection method based on the generation countermeasure network can accurately detect the countermeasure signal, reduces misdirection and loss caused by the existence of the countermeasure signal in the signal demodulation process, effectively reduces risks brought by the countermeasure signal, and strengthens safety and secrecy of the signal transmission process.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
fig. 1 is a schematic flow chart of a countermeasure signal detection method based on generation of a countermeasure network according to an embodiment of the present application;
FIG. 2 is a schematic diagram of an infrastructure of a deep neural network model according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a training process for generating a countermeasure network according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a countermeasure signal detection flow in an embodiment of the present application;
fig. 5 is a block diagram of a countermeasure signal detection apparatus based on a countermeasure network generation in an embodiment of the present application;
fig. 6 is a schematic structural diagram of an electronic device in an embodiment of the present application.
Detailed Description
Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. These embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein.
Fig. 1 is a schematic flowchart illustrating a method for detecting a countermeasure signal based on a generated countermeasure network according to an embodiment of the present application, and referring to fig. 1, the method for detecting a countermeasure signal based on a generated countermeasure network according to an embodiment of the present application includes the following steps S110 to S150:
step S110, a signal training sample is obtained.
In the embodiment of the application, when the countermeasure signal is detected, a signal training sample is required to be obtained first, where the signal training sample refers to signal data which is labeled in advance and used for training to generate the countermeasure network, and the signal may be an electromagnetic signal.
And step S120, generating a first reconstruction sample corresponding to the signal training sample by using the generation countermeasure network based on the signal training sample.
Generation of a countermeasure network (GAN) is a deep learning model, and is one of the most promising methods for unsupervised learning in complex distribution in recent years. The model passes through at least two modules in the framework: mutual game learning of a generator (genetic Model) and a discriminator (Discriminative Model) can generate quite good output, and therefore the game learning method is widely applied to the aspects of enriching sample data quantity and the like.
The method and the device for generating the confrontation network aim at minimizing the difference between the distribution of the reconstructed sample and the training sample by generating the confrontation network to reconstruct the sample data, and meanwhile, the discriminator can accurately discriminate the reconstructed sample from the training sample. Therefore, in the embodiment of the application, the obtained signal training sample is input into the generation countermeasure network which is trained in advance, so that the first reconstruction sample which has a small difference with the signal training sample can be generated.
Step S130, classifying the signal training sample and the first reconstruction sample respectively by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample.
After the first reconstructed sample corresponding to the signal training sample is obtained, classification prediction needs to be performed on the signal training sample and the first reconstructed sample respectively by using a classification model trained in advance, so that a classification result of the signal training sample and a classification result of the first reconstructed sample are obtained. The classification model can be trained as follows:
training samples X ═ X using the signal1,x2,…,xnClass label Z ═ Z for X1,z2,…,znTraining a deep neural network model, and defining a deep neural network model function and a loss function as follows:
zi＝argmax(F(xi)),i＝1,2,…,n (1)
Lm＝-(zilog(max(F(xi)))+(1-zi)log(1-max(F(xi)))),i＝1,2,…,n (2)
wherein F (·) represents a deep neural network model, F (x)i) Representing model pairs xiOutput vector of xiDenotes the ith sample, ziIndicates the label corresponding to the ith sample, LmAnd representing the loss value between the predicted value and the real value of the sample by the deep neural network model.
The classification model can be obtained by training an infrastructure of the deep neural network model, and as shown in fig. 2, a schematic diagram of an infrastructure of the deep neural network model in the embodiment of the present application is provided, where the schematic diagram includes an input layer, a hidden layer, an output layer, and the like.
Of course, which deep neural network model is specifically adopted can be flexibly selected by those skilled in the art according to actual requirements, and is not specifically limited herein.
Step S140, determining a countermeasure signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample.
In step S150, a countermeasure signal is detected based on the countermeasure signal detection threshold.
After obtaining the classification structure of the signal training sample and the classification result of the first reconstructed sample, it is necessary to compare the deviation between the signal training sample and the classification result of the corresponding first reconstructed sample, where for a normal signal, i.e. a non-countermeasure signal, the deviation between the signal training sample and the classification result of the corresponding first reconstructed sample should be very small, and for a countermeasure signal, the deviation between the signal training sample and the classification result of the corresponding first reconstructed sample should be relatively large.
Thus, here, a counter signal detection threshold may be determined based on a comparison of the classification results of the signal training samples and the first reconstruction samples, which may serve as a basis for a subsequent detection of the counter signal.
The countermeasure signal detection method based on the generation countermeasure network can accurately detect the countermeasure signal, reduces misdirection and loss caused by the existence of the countermeasure signal in the signal demodulation process, effectively reduces risks brought by the countermeasure signal, and strengthens safety and secrecy of the signal transmission process.
In one embodiment of the present application, the generation of the countermeasure network is trained by: acquiring a signal training sample, wherein the signal training sample comprises a training data label; generating a second reconstruction sample with a generator that generates a countermeasure network based on the training data labels and the random noise data; judging the signal training sample and the second reconstruction sample by using a discriminator in the generated countermeasure network to obtain a judgment result; and updating parameters of the generated countermeasure network according to the judgment result to obtain the trained generated countermeasure network.
As shown in fig. 3, a schematic diagram of a training process for generating a countermeasure network in the embodiment of the present application is provided. When the countermeasure network is generated in the training process, the embodiment of the present application may first obtain a signal training sample, where the signal training sample is the same as the signal training sample of the foregoing embodiment and includes a training sample and a corresponding training data label.
Initializing a plurality of random noise data for each class of training samples, then generating a second reconstruction sample corresponding to each training sample by a generator G for generating a countermeasure network, then respectively distinguishing each training sample and the second reconstruction sample corresponding to each training sample by a discriminator D for generating the countermeasure network to obtain a distinguishing result, and determining the generation Loss G-Loss of the generator G and the distinguishing Loss D-Loss of the discriminator D according to the distinguishing result.
Based on the generation Loss G-Loss of the generator G, the generator G is updated by a gradient descent method, and the formula can be expressed as follows:
wherein G (-) is a generator, θgModel parameters representing the generator, ri (j)Is the jth random noise value of the ith sample.
Based on the discrimination Loss D-Loss of the discriminator D, the discriminator D is updated by gradient ascent, and the formula can be expressed as:
wherein D (-) is a discriminator, θdModel parameters, x, representing discriminatorsiDenotes the ith sample, ri (j)Is the jth random noise value of the ith sample.
Through the iterative training of the process, the generation countermeasure network with better output effect is finally obtained.
In one embodiment of the application, the training data labels comprise training data labels of a plurality of signal training samples, and generating the second reconstruction samples with the generator for generating the countermeasure network based on the training data labels and the random noise data comprises: initializing a plurality of random noise values corresponding to a target signal training sample for a training data label of the target signal training sample, wherein the target signal training sample is any one of a plurality of signal training samples; generating a plurality of second reconstruction samples corresponding to the target signal training samples by using a generator for generating a countermeasure network based on a plurality of random noise values corresponding to the target signal training samples; and determining a final second reconstruction sample corresponding to the target signal training sample according to the similarity between the target signal training sample and each second reconstruction sample.
In actual training, the signal training samples comprise a plurality of signal training samples and corresponding training data labels, and when a generator for generating a countermeasure network generates a second reconstruction sample based on the training data labels and random noise data, m random values r can be initialized randomly for each target signal training samplei (1),…,ri (m)Then, a plurality of second reconstruction samples corresponding to the target signal training sample are generated by a generator for generating a countermeasure network, and the original signal training sample x is selected from the second reconstruction samplesiAnd taking one reconstructed sample with the closest distribution as a second reconstructed sample of the training sample of the target signal. In particular, it is here possible to pass the distance d between the samplesiTo measure the similarity between samples, the similarity calculation formula is as follows:
wherein x isiRepresents the ith training sample, ri (j)Is denoted by xiJ sample of label generation, diIndicating the similarity between the two samples.
In one embodiment of the present application, the classification result of the signal training sample includes a prediction value of a plurality of signal training samples, the classification result of the first reconstructed sample includes a prediction value of a plurality of first reconstructed samples, and determining the counter signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstructed sample includes: respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values; an average of the plurality of error values is determined as a challenge signal detection threshold.
The classification result of the signal training samples obtained in the embodiment of the present application includes the predicted values of a plurality of signal training samples, and the classification result of the first reconstruction sample includes the predicted values of a plurality of first reconstruction samples, and when determining the countermeasure signal detection threshold, the classification result of the signal training samples can be obtained in the following manner:
wherein T represents a counter signal detection threshold, F (r)i (j)) Representing a preset classification model on a first reconstructed sample ri (j)F (x) is the predicted value ofi) Representing model to signal training sample xiThe predicted value of (2).
According to the embodiment of the application, the absolute error values between the predicted values of the plurality of signal training samples and the predicted values of the corresponding first reconstruction samples are averaged to be used as the standard for detecting the countermeasure signal subsequently, so that the accuracy of the detection of the countermeasure signal can be further improved.
In one embodiment of the present application, detecting the counter signal according to the counter signal detection threshold includes: acquiring a signal test sample; generating a third reconstructed sample with a generator for generating a countermeasure network from the signal test sample; classifying the signal test sample and the third reconstruction sample by using a preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample; and detecting the counter signal in the signal test sample by using the counter signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
As shown in fig. 4, a schematic diagram of a countermeasure signal detection flow in the embodiment of the present application is provided. In the embodiment of the application, when the countermeasure signal is detected by using the countermeasure signal detection threshold, the signal test sample X can be obtained firstt＝{xt1,xt2,…,xtnThen, the generator which is trained by the foregoing embodiment and generates the countermeasure network generates the third reconstruction sample R corresponding to the signal test samplet＝{rt1,rt2,…,rtnAnd then, respectively carrying out classification detection on the signal test sample and the third reconstructed sample by using a preset classification model to obtainTo the classification result of the signal test sample and the classification result of the third reconstructed sample.
Whether the signal test sample is the countermeasure signal is detected using the countermeasure signal detection threshold determined in the foregoing embodiment by comparing the error between the classification result of the signal test sample and the classification result of the third reconstructed sample.
In an embodiment of the present application, the classification result of the signal test sample and the classification result of the third reconstructed sample both include predicted values, and the detecting the counter signal in the signal test sample by using the counter signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstructed sample includes: determining an error value of the predicted value of the signal test sample and the predicted value of the third reconstructed sample; determining the signal test sample as a countermeasure signal if the error value is greater than the countermeasure signal detection threshold; in the event that the error value is not greater than the challenge signal detection threshold, the signal test sample is determined to be a non-challenge signal.
The classification result of the signal test sample and the classification result of the third reconstruction sample in the embodiment of the present application both include a predicted value, and when the countermeasure signal is detected, the absolute error value E between the predicted value of the signal test sample and the predicted value of the third reconstruction sample may be calculated in the following manner:
E＝|max(F(xti))-max(F(rti))|,i＝1,2,…,n (7)
wherein E represents an absolute error, F (r)ti) Representing the preset classification model pair generated data rtiF (x) is the predicted value ofti) Representing model pairs xtiThe predicted value of (2).
After calculating the absolute error value E between the predicted value of the signal test sample and the predicted value of the third reconstructed sample, the absolute error value E may be compared with the countermeasure signal detection threshold T, if the absolute error value E is greater than the countermeasure signal detection threshold T, it is indicated that the category deviation between the signal test sample and its reconstructed sample is large, and therefore the signal test sample may be determined as a countermeasure signal, and if the absolute error value E is not greater than the countermeasure signal detection threshold T, it is indicated that the category deviation between the signal test sample and its reconstructed sample is very small, and therefore the signal test sample may be determined as a non-countermeasure signal, i.e., a normal signal.
To further illustrate the training method and training effect of the present application for generating an anti-confrontation network, here, for example, the signal training samples of the present application may specifically include 12 small categories of phase shift keying modulation, frequency shift keying modulation, quadrature amplitude modulation, and pulse amplitude modulation: BPSK, QPSK, 8PSK, OQPSK, 2FSK, 4FSK, 8FSK, 16QAM, 32QAM, 64QAM, 4PAM and 8 PAM. The original signal data is randomly generated to ensure equal probability of transmitting bits. The pulse shaping filter adopts a raised cosine filter and a roll coefficient, and a random value is extracted within the range of [0.2 and 0.7 ]. The phase deviation is randomly selected within the range of [ -pi, pi ], and the normalized carrier frequency offset is randomly selected within the range of [ -0.1,0.1 ]. The signal-to-noise ratio for each modulation class is evenly distributed from-20 dB to 30 dB. Each data sample is an IQ signal, comprising 64 symbols, with 8 sample points per symbol, and thus 512 sample points per sample. The training set and the test set are 312,000 and 156,000 in size, respectively, and the amount of samples of each type of modulation signal is the same.
Obtaining a confrontation sample based on a gradient attack method, obtaining a final detection result of the confrontation sample by using the method, and finally using an ACC (Accuracy) index as a basis for judging the training effect.
Table 1 shows ACC indicators of the Method for detecting a challenge signal based on a generated challenge network in the embodiment of the present application, which use an Alexnet network architecture and an FGSM attack algorithm (Fast Gradient Sign Method). As can be seen from table 1, the accuracy of the countermeasure signal detection method based on generation of the countermeasure network according to the embodiment of the present application can reach 88.68%, and therefore, the countermeasure sample can be effectively detected.
TABLE 1
| Model | Attack | ACC |
| Alexnet | FGSM | 88.68％ |
Table 2 shows a countermeasure signal detection method (Detect-GAN) based on a generated countermeasure network according to an embodiment of the present application, and ACC index comparisons based on a Bayesian Uncertainty detection method (BUE for short), a Kernel Density detection method (KDE for short), and a Local Intrinsic dimension method (LID for short).
TABLE 2
| Method | Detect-GAN | LID | KDE | BUE |
| ACC | 88.68％ | 87.20％ | 61.31％ | 54.08％ |
As can be seen from tables 1 and 2, the method for detecting the countermeasure signal based on the generation countermeasure network according to the present application has a better detection effect than the existing detection method.
The method belongs to the technical concept similar to the method for detecting the countermeasure signal based on the generated countermeasure network, and the embodiment of the application also provides a device for detecting the countermeasure signal based on the generated countermeasure network. Fig. 5 shows a block diagram of a countermeasure signal detection apparatus based on a countermeasure network generation in an embodiment of the present application, and referring to fig. 5, a countermeasure signal detection apparatus 500 based on a countermeasure network generation includes: an acquisition unit 510, a generation unit 520, a classification unit 530, a determination unit 540, and a detection unit 550. Wherein the content of the first and second substances,
an obtaining unit 510, configured to obtain a signal training sample;
a generating unit 520, configured to generate, based on the signal training samples, first reconstructed samples corresponding to the signal training samples by using the generation countermeasure network;
the classification unit 530 is configured to classify the signal training sample and the first reconstructed sample respectively by using a preset classification model, so as to obtain a classification result of the signal training sample and a classification result of the first reconstructed sample;
a determining unit 540, configured to determine a counter signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample;
a detection unit 550 for detecting the countermeasure signal according to the countermeasure signal detection threshold.
In one embodiment of the present application, the generation of the countermeasure network is trained by: acquiring a signal training sample, wherein the signal training sample comprises a training data label; generating a second reconstruction sample with a generator that generates a countermeasure network based on the training data labels and the random noise data; judging the signal training sample and the second reconstruction sample by using a discriminator in the generated countermeasure network to obtain a judgment result; and updating parameters of the generated countermeasure network according to the judgment result to obtain the trained generated countermeasure network.
In one embodiment of the present application, the training data labels include training data labels of a plurality of signal training samples, and the generation countermeasure network is trained by: initializing a plurality of random noise values corresponding to a target signal training sample for a training data label of the target signal training sample, wherein the target signal training sample is any one of a plurality of signal training samples; generating a plurality of second reconstruction samples corresponding to the target signal training samples by using a generator for generating a countermeasure network based on a plurality of random noise values corresponding to the target signal training samples; and determining a final second reconstruction sample corresponding to the target signal training sample according to the similarity between the target signal training sample and each second reconstruction sample.
In an embodiment of the present application, the classification result of the signal training samples includes predicted values of a plurality of signal training samples, and the determining unit 540 is specifically configured to: respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values; an average of the plurality of error values is determined as a challenge signal detection threshold.
In an embodiment of the present application, the detecting unit 550 is specifically configured to: acquiring a signal test sample; generating a third reconstructed sample with a generator for generating a countermeasure network from the signal test sample; classifying the signal test sample and the third reconstruction sample by using a preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample; and detecting the counter signal in the signal test sample by using the counter signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
In an embodiment of the present application, the classification result of the signal test sample and the classification result of the third reconstructed sample both include predicted values, and the detection unit 550 is specifically configured to: determining an error value of the predicted value of the signal test sample and the predicted value of the third reconstructed sample; determining the signal test sample as a countermeasure signal if the error value is greater than the countermeasure signal detection threshold; in the event that the error value is not greater than the challenge signal detection threshold, the signal test sample is determined to be a non-challenge signal.
It should be noted that:
fig. 6 illustrates a schematic structural diagram of an electronic device. Referring to fig. 6, at a hardware level, the electronic device includes a memory and a processor, and optionally further includes an interface module, a communication module, and the like. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may also include a non-volatile Memory, such as at least one disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, the interface module, the communication module, and the memory may be connected to each other via an internal bus, which may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 6, but that does not indicate only one bus or one type of bus.
A memory for storing computer executable instructions. The memory provides computer executable instructions to the processor through the internal bus.
A processor executing computer executable instructions stored in the memory and specifically configured to perform the following operations:
acquiring a signal training sample;
generating a first reconstruction sample corresponding to the signal training sample by using a generation countermeasure network based on the signal training sample;
classifying the signal training sample and the first reconstruction sample by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
determining a confrontation signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample;
the countermeasure signal is detected based on a countermeasure signal detection threshold.
The functions performed by the countermeasure signal detection apparatus based on generation of the countermeasure network as disclosed in the embodiment of fig. 5 of the present application may be applied to or implemented by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The electronic device may further perform the steps performed by the countermeasure signal detection method based on the generated countermeasure network in fig. 1, and implement the functions of the countermeasure signal detection method based on the generated countermeasure network in the embodiment shown in fig. 1, which are not described herein again.
An embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs that, when executed by a processor, implement the foregoing countermeasure signal detection method based on a generation countermeasure network, and are specifically configured to perform:
acquiring a signal training sample;
generating a first reconstruction sample corresponding to the signal training sample by using a generation countermeasure network based on the signal training sample;
classifying the signal training sample and the first reconstruction sample by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
determining a confrontation signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample;
the countermeasure signal is detected based on a countermeasure signal detection threshold.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
The present application is described in terms of flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) characterized by computer-usable program code.
The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (10)
1. A countermeasure signal detection method based on a generation countermeasure network is characterized by comprising the following steps:
acquiring a signal training sample;
generating a first reconstruction sample corresponding to the signal training sample by using the generation countermeasure network based on the signal training sample;
classifying the signal training sample and the first reconstruction sample by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
determining a confrontation signal detection threshold according to the classification result of the signal training sample and the classification result of the first reconstruction sample;
detecting a counter signal according to the counter signal detection threshold.
2. The method of claim 1, wherein the generating the antagonistic network is trained by:
acquiring the signal training sample, wherein the signal training sample comprises a training data label;
generating a second reconstruction sample with the generator generating a countermeasure network based on the training data labels and random noise data;
judging the signal training sample and the second reconstruction sample by using a discriminator in the generated countermeasure network to obtain a judgment result;
and updating the parameters of the generated countermeasure network according to the judgment result to obtain the trained generated countermeasure network.
3. The method of claim 2, wherein the training data labels comprise training data labels of a plurality of signal training samples, and wherein generating second reconstruction samples with the generator that generates the countermeasure network based on the training data labels and random noise data comprises:
initializing a plurality of random noise values corresponding to a target signal training sample for a training data label of the target signal training sample, wherein the target signal training sample is any one of a plurality of signal training samples;
generating a plurality of second reconstruction samples corresponding to the target signal training samples by using the generator for generating the countermeasure network based on a plurality of random noise values corresponding to the target signal training samples;
and determining a final second reconstruction sample corresponding to the target signal training sample according to the similarity between the target signal training sample and each second reconstruction sample.
4. The method of claim 1, wherein the classification result of the signal training samples comprises predicted values of a plurality of signal training samples, wherein the classification result of the first reconstructed sample comprises predicted values of a plurality of first reconstructed samples, and wherein determining the counter signal detection threshold according to the classification result of the signal training samples and the classification result of the first reconstructed sample comprises:
respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values;
an average of the plurality of error values is determined as the countermeasure signal detection threshold.
5. The method of claim 1, wherein the detecting a counter signal according to the counter signal detection threshold comprises:
acquiring a signal test sample;
generating a third reconstructed sample by using the generator for generating the countermeasure network according to the signal test sample;
classifying the signal test sample and the third reconstruction sample by using the preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample;
and detecting the countermeasure signal in the signal test sample by using the countermeasure signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
6. The method of claim 5, wherein the classification result of the signal test sample and the classification result of the third reconstructed sample each comprise a predicted value, and wherein detecting the counter signal in the signal test sample using the counter signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstructed sample comprises:
determining an error value of the predicted value of the signal test sample and the predicted value of the third reconstructed sample;
determining the signal test sample as a challenge signal if the error value is greater than the challenge signal detection threshold;
determining that the signal test sample is a non-countermeasure signal if the error value is not greater than the countermeasure signal detection threshold.
7. A countermeasure signal detection apparatus based on generation of a countermeasure network, comprising:
the acquisition unit is used for acquiring a signal training sample;
a generating unit, configured to generate, based on the signal training sample, a first reconstruction sample corresponding to the signal training sample by using the generative countermeasure network;
the classification unit is used for classifying the signal training sample and the first reconstruction sample respectively by using a preset classification model to obtain a classification result of the signal training sample and a classification result of the first reconstruction sample;
a determining unit, configured to determine a countermeasure signal detection threshold according to a classification result of the signal training sample and a classification result of the first reconstruction sample;
a detection unit for detecting a countermeasure signal according to the countermeasure signal detection threshold.
8. The apparatus according to claim 7, wherein the classification result of the signal training samples comprises predicted values of a plurality of signal training samples, and the determining unit is specifically configured to:
respectively determining error values of the predicted value of each signal training sample and the predicted value of the corresponding first reconstruction sample to obtain a plurality of error values;
an average of the plurality of error values is determined as the countermeasure signal detection threshold.
9. The apparatus according to claim 7, wherein the detection unit is specifically configured to:
acquiring a signal test sample;
generating a third reconstructed sample by using the generator for generating the countermeasure network according to the signal test sample;
classifying the signal test sample and the third reconstruction sample by using the preset classification model to obtain a classification result of the signal test sample and a classification result of the third reconstruction sample;
and detecting the countermeasure signal in the signal test sample by using the countermeasure signal detection threshold according to the classification result of the signal test sample and the classification result of the third reconstruction sample.
10. An electronic device, comprising: a processor, a memory storing computer-executable instructions,
the executable instructions, when executed by the processor, implement the countermeasure signal detection method based on generation of a countermeasure network of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111080309.2A CN113723358A (en) | 2021-09-15 | 2021-09-15 | Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111080309.2A CN113723358A (en) | 2021-09-15 | 2021-09-15 | Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113723358A true CN113723358A (en) | 2021-11-30 |
Family
ID=78683906
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111080309.2A Pending CN113723358A (en) | 2021-09-15 | 2021-09-15 | Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113723358A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114760132A (en) * | 2022-04-14 | 2022-07-15 | 中国电信股份有限公司 | Signal sender identity authentication method and device, storage medium and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110969242A (en) * | 2019-11-27 | 2020-04-07 | 浙江工业大学 | Defense method for generating general inverse disturbance based on generative confrontation |
| US20200143079A1 (en) * | 2018-11-07 | 2020-05-07 | Nec Laboratories America, Inc. | Privacy-preserving visual recognition via adversarial learning |
| CN112347844A (en) * | 2020-09-21 | 2021-02-09 | 浙江工业大学 | Signal countermeasure sample detector design method based on LID |
| US20210192357A1 (en) * | 2018-05-17 | 2021-06-24 | Magic Leap, Inc. | Gradient adversarial training of neural networks |
| CN113378643A (en) * | 2021-05-14 | 2021-09-10 | 浙江工业大学 | Signal countermeasure sample detection method based on random transformation and wavelet reconstruction |
-
2021
- 2021-09-15 CN CN202111080309.2A patent/CN113723358A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210192357A1 (en) * | 2018-05-17 | 2021-06-24 | Magic Leap, Inc. | Gradient adversarial training of neural networks |
| US20200143079A1 (en) * | 2018-11-07 | 2020-05-07 | Nec Laboratories America, Inc. | Privacy-preserving visual recognition via adversarial learning |
| CN110969242A (en) * | 2019-11-27 | 2020-04-07 | 浙江工业大学 | Defense method for generating general inverse disturbance based on generative confrontation |
| CN112347844A (en) * | 2020-09-21 | 2021-02-09 | 浙江工业大学 | Signal countermeasure sample detector design method based on LID |
| CN113378643A (en) * | 2021-05-14 | 2021-09-10 | 浙江工业大学 | Signal countermeasure sample detection method based on random transformation and wavelet reconstruction |
Non-Patent Citations (1)
| Title |
|---|
| 付强;王华伟;熊明兰;: "基于GANBP的航空发动机性能退化预测模型", 人类工效学, no. 01 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114760132A (en) * | 2022-04-14 | 2022-07-15 | 中国电信股份有限公司 | Signal sender identity authentication method and device, storage medium and electronic equipment |
| CN114760132B (en) * | 2022-04-14 | 2023-10-31 | 中国电信股份有限公司 | Signal sender identity authentication method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11909563B2 (en) | Method and apparatus for modulation recognition of signals based on cyclic residual network | |
| CN108600135B (en) | Method for identifying signal modulation mode | |
| Wu et al. | Robust automatic modulation classification under varying noise conditions | |
| US9923912B2 (en) | Learning detector of malicious network traffic from weak labels | |
| US20140270489A1 (en) | Learned mid-level representation for contour and object detection | |
| CN109344257B (en) | Text emotion recognition method and device, electronic equipment and storage medium | |
| CN111461243B (en) | Classification method, classification device, electronic equipment and computer-readable storage medium | |
| CN103632159B (en) | Method and system for training classifier and detecting text area in image | |
| Krawczyk et al. | Dynamic classifier selection for one-class classification | |
| CN101893704A (en) | Rough set-based radar radiation source signal identification method | |
| CN111510408B (en) | Signal modulation mode identification method and device, electronic equipment and storage medium | |
| CN112347844B (en) | LID-based signal countermeasure sample detector design method | |
| CN104038792B (en) | For the video content analysis method and apparatus of IPTV supervision | |
| CN112364851B (en) | Automatic modulation recognition method and device, electronic equipment and storage medium | |
| CN112395382A (en) | Ship abnormal track data detection method and device based on variational self-encoder | |
| Zhang et al. | Open set recognition of communication signal modulation based on deep learning | |
| CN113723358A (en) | Countermeasure signal detection method and device based on generation of countermeasure network and electronic equipment | |
| CN115294615A (en) | Radio frequency fingerprint identification method based on machine learning | |
| CN113791389B (en) | Method and system for identifying unknown radar radiation source individuals | |
| CN101996312A (en) | Method and device for tracking targets | |
| CN111177403A (en) | Sample data processing method and device | |
| CN116109907A (en) | Target detection method, target detection device, electronic equipment and storage medium | |
| CN113378643B (en) | Signal countermeasure sample detection method based on random transformation and wavelet reconstruction | |
| CN117523218A (en) | Label generation, training of image classification model and image classification method and device | |
| CN115577239A (en) | Electromagnetic signal open set identification method and device, electronic equipment and storage 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 |