CN115378777A - Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment - Google Patents

Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment Download PDF

Info

Publication number
CN115378777A
CN115378777A CN202211022794.2A CN202211022794A CN115378777A CN 115378777 A CN115378777 A CN 115378777A CN 202211022794 A CN202211022794 A CN 202211022794A CN 115378777 A CN115378777 A CN 115378777A
Authority
CN
China
Prior art keywords
entropy
communication signal
characteristic value
value
signal
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
Application number
CN202211022794.2A
Other languages
Chinese (zh)
Inventor
王茂法
巩超
李文欣
钱高峰
薛静泽
杨武
朱振经
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN202211022794.2A priority Critical patent/CN115378777A/en
Publication of CN115378777A publication Critical patent/CN115378777A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/0012—Modulated-carrier systems arrangements for identifying the type of modulation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B13/00—Transmission systems characterised by the medium used for transmission, not provided for in groups H04B3/00 - H04B11/00
    • H04B13/02—Transmission systems in which the medium consists of the earth or a large mass of water thereon, e.g. earth telegraphy
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Noise Elimination (AREA)

Abstract

The invention discloses a method for identifying a modulation mode of an underwater communication signal in an alpha stable distribution noise environment, which comprises the following steps: s2, preprocessing the communication signal S (n) collected in the step S1 to obtain a preprocessed communication signal S o (n); s3, extracting the communication signal S in the step S2 o (n) the signal feature vector comprises a characteristic value E of the Shannon entropy of the singular spectrum Ssse The characteristic value of the singular spectral index entropy is E Ssee The characteristic value of the wavelet energy spectrum entropy is E Wese The characteristic value of the Shannon entropy of the sum power spectrum is E Seps (ii) a S4, defining four-dimensional input parameters X = { E = Ssse ,E Ssee ,E Wese ,E Seps }；S5, inputting the four-dimensional input parameter X defined in the step S4 into a GWO-SVM classification model; by adopting the technical scheme, as the influence of sharp pulses can be better inhibited by preprocessing, a plurality of signal characteristic vectors are used as the input of the classifier, and the high identification success rate is achieved; the GWO-SVM classifier is adopted to avoid the influence of punishment parameters of the SVM and width parameter values of the RBF basis functions on SVM classification results, and the functions adopt the optimal accuracy of five-fold cross validation.

Description

Method for identifying modulation mode of underwater communication signal in alpha stable distribution noise environment
Technical Field
The invention relates to the technical field of communication signal modulation mode identification, in particular to an underwater communication signal modulation mode identification method in an alpha stable distributed noise environment.
Background
The modulation and identification of the underwater acoustic communication signals are important links for recovering the information of the underwater acoustic communication signals under the non-cooperative receiving condition, have important significance for the development and utilization of ocean resources, the improvement of underwater reconnaissance early warning capability and the like, and can realize the further demodulation and interpretation of the received signals only by correctly identifying the modulation mode of the received underwater acoustic communication signals.
Modulation identification research dates back to 1969 as early as a academic report made by c.s.waver et al at the university of stanford in the united states, which first proposed a modulation identification technique, and proposed that differences in signal spectrum be used to identify different modulation type signals in the report; since then, more and more scholars are engaged in the research on modulation recognition, and the modulation recognition technology has been developed over decades to generate various recognition algorithms, but whatever the algorithm, it can be classified into two types of methods: a likelihood function-based decision theory method and a feature extraction-based statistical pattern recognition method; the statistical pattern recognition method based on feature extraction has the advantages that the features representing the signal characteristics are extracted, so that the robustness is higher, the calculation complexity is lower, in addition, the requirement on prior knowledge is lower, and the statistical pattern recognition method based on feature extraction has a higher application value in engineering application.
The statistical pattern recognition method based on feature extraction mainly comprises two processes: firstly, various signal analysis methods are adopted to extract the characteristics of the modulation signals, and then a classifier is adopted to classify the modulation signals.
However, in actual underwater acoustic communication, due to the influences of marine biological activity, ocean current, tide, sea surface wave, seismic activity, traffic and shipping and the like, noise of the underwater acoustic communication often contains significant spike pulses, statistical characteristics often deviate from Gaussian distribution under the influence of the spike pulses, and impulse noise does not have higher-order cumulant, second-order cyclic spectrum and other second-order or higher-order statistics, so that a modulation identification algorithm based on the statistics is inevitably invalid and cannot meet the requirements of modulation mode identification of water communication signals, therefore, the existing algorithm is expanded and improved, and an important significance is achieved in researching a proper method to inhibit alpha distribution noise to adapt to modulation identification of the underwater acoustic communication; secondly, the quality of the classifier is directly related to the recognition accuracy, the problem of establishing the current LSSVM estimation model, namely the selection problem of the hyper-parameters, is inappropriately selected, and the reliability of the prediction result is reduced.
Disclosure of Invention
The invention aims to provide an underwater communication signal modulation mode identification method in an alpha stable distribution noise environment, which is suitable for the alpha stable distribution environment and can improve the identification accuracy.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a method for identifying a modulation mode of an underwater communication signal in an alpha stable distribution noise environment comprises the following steps:
s1, acquiring a communication signal S (n) in an alpha stable distributed noise environment;
s2, the amplitude value of the communication signal S (n) collected in the step S1 is larger than the threshold tau r The part of the signal S is subjected to nonlinear suppression preprocessing to obtain a preprocessed communication signal S o (n)；
S3, extracting the communication signal S in the step S2 o The signal characteristic vector of (n) comprises a characteristic value of time domain information entropy and a characteristic value of frequency domain information entropy, and the characteristic value of singular spectrum Shannon entropy in the time domain characteristic is E Ssse Singular spectral exponential entropyHas a characteristic value of E Ssee And the characteristic value of the entropy of the wavelet energy spectrum is E Wese The characteristic value of the Shannon entropy of the power spectrum in the frequency domain characteristic is E Seps ；
S4, constructing a signal feature vector;
the signal feature vector extracted according to the step S3 comprises a feature value of singular spectrum Shannon entropy, a feature value of singular spectrum index entropy, a feature value of wavelet energy spectrum entropy and a feature value of power spectrum Shannon entropy, and is defined as a four-dimensional input parameter X = { E = { E = Ssse ,E Ssee ,E Wese ,E Seps }；
S5, training a GWO-SVM classifier;
and (4) inputting the four-dimensional input parameter X defined in the step S4 into a GWO-SVM classification model for training and optimization, and identifying a modulation mode.
Preferably, the signal is S o (n) pairs of signals can be expressed as:
Figure BDA0003814699900000031
the above-mentioned tau r Can be represented by the formula tau r ＝(1+2τ o )τ Q Obtaining;
the above tau o Is a constant coefficient, is set to 1.5, tau Q Is the second quarter-bit value of the modulus | S (n) | of the signal S (n), S o And (n) is the communication signal after nonlinear suppression preprocessing.
Preferably, the signal feature vector comprises the feature value of the time domain information entropy and the feature value of the frequency domain information entropy, and the extracting steps of the feature value of the time domain information entropy and the feature value of the frequency domain information entropy are as follows:
(1) the communication signal S o (n) is obtained in step S2 by performing nonlinear suppression preprocessing on the communication signal S (n) acquired in step S1, where n =1,2, \ 8230;, k; k represents a signal sample data length;
(2) setting window length M, and transmitting the communication signal S in the step (1) o (n) partitioning to generate matrix a:
Figure BDA0003814699900000032
(3) performing Singular Value Decomposition (SVD) on the matrix A in the step (2) to obtain a non-zero singular value sigma 1 ≥σ 2 ≥…≥σ M ；
(4) Characteristic value E of Shannon entropy of singular spectrum Ssse The calculation formula of (c) is as follows:
Figure BDA0003814699900000033
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000034
(5) characteristic value E of singular spectral index entropy Ssee The calculation formula of (c) is as follows:
Figure BDA0003814699900000035
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000041
(6) characteristic value E of the wavelet energy spectrum entropy Wese The steps of (1) are as follows;
for the time signal S after nonlinear suppression pretreatment o Performing wavelet transformation to obtain wavelet coefficients on n scales, wherein the calculation formula is as follows:
Figure BDA0003814699900000042
when the scale is i, the energy value is m i ,i＝1,2,…,n；
(7) Characteristic value E of Shannon entropy of wavelet energy spectrum Wese The calculation formula of (c) is as follows:
Figure BDA0003814699900000043
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000044
(8) characteristic value E of the power spectrum entropy Seps ；
Communication signal time sequence S after nonlinear suppression preprocessing o ＝{S o (i) Is defined as S 'and its power spectrum is' o (ω):
Figure BDA0003814699900000045
In the formula S o (ω) is the time series S o Fourier transform of (1);
(9) characteristic value E of Shannon entropy of power spectrum entropy value Seps The calculation formula of (c):
Figure BDA0003814699900000046
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000047
preferably, the nonlinear suppression preprocessing comprises truncation and zero setting processing in a threshold suppression method.
By adopting the technical scheme, the influence of sharp pulses can be better inhibited through the cut-off and zero-setting neutralization in the threshold inhibition method, the signal characteristic vector comprises four characteristics of a characteristic value of time domain information entropy and a characteristic value of frequency domain information entropy as the input of the classifier, and the classifier has higher identification success rate compared with single characteristic input; the GWO-SVM classifier avoids the problem that punishment parameters of an SVM and width parameter values of RBF basis functions influence SVM classification results, global optimization of the parameters is carried out by using a GWO algorithm, the target function adopts the optimal accuracy of five-fold cross validation, and a classification model with higher accuracy is obtained.
Drawings
Fig. 1 is a flow chart of a communication signal modulation scheme identification method according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made with reference to the accompanying drawings. It should be noted that the description of the embodiments is provided to help understanding of the present invention, and is not intended to limit the present invention. In addition, the technical features involved in the respective embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
As shown in fig. 1, an identification method of an underwater communication signal modulation mode in an α stable distributed noise environment of the invention includes the following steps:
s1, acquiring a communication signal S (n) in an alpha stable distributed noise environment;
s2, the amplitude value of the communication signal S (n) collected in the step S1 is larger than the threshold tau r The part of the signal S is subjected to nonlinear suppression preprocessing to obtain a preprocessed communication signal S o (n) suppressing the influence of sharp pulses on feature extraction;
s3, extracting the communication signal S in the step S2 o The signal characteristic vector of (n) comprises a characteristic value of time domain information entropy and a characteristic value of frequency domain information entropy, and the characteristic value of singular spectrum Shannon entropy in the time domain characteristic is E Ssse The characteristic value of the singular spectral index entropy is E Ssee And the characteristic value of the entropy of the wavelet energy spectrum is E Wese The characteristic value of the Shannon entropy of the power spectrum in the frequency domain characteristic is E Seps And the complementarity and the relevance among the four characteristics improve the classification performance of the classifier and have higher identification success rate.
S4, constructing a signal feature vector;
the signal feature vector extracted according to the step S3 comprises a feature value of singular spectrum Shannon entropy, a feature value of singular spectrum index entropy, a feature value of wavelet energy spectrum entropy and a feature value of power spectrum Shannon entropy, and is defined as a four-dimensional input parameter X = { E = { E = Ssse ,E Ssee ,E Wese ,E Seps The signal feature vector has a plurality of features, and complementarity and correlation exist among the featuresThe classification performance of the classifier is improved, and the success rate of identification is improved;
s5, training a GWOO-SVM classifier;
and (5) inputting the four-dimensional input parameter X defined in the step (S4) into a GWOO-SVM classification model for training optimization, identifying a modulation mode, avoiding the problem that the punishment parameter of the SVM and the width parameter value of the RBF basis function influence the SVM classification result, improving the influence of the punishment parameter and the width parameter value of the RBF basis function on the SVM classification result by the GWOO optimization algorithm, and improving the accuracy of the classification model.
Otherwise, the signal is S o (n) pairs of signals can be expressed as:
Figure BDA0003814699900000061
τ r can be represented by the formula tau r ＝(1+2τ o )τ Q Obtaining;
τ o is a constant coefficient and is set to 1.5, tau Q Is the second quarter-bit value of the modulus | S (n) | of the signal S (n), S o And (n) is the communication signal after nonlinear suppression pretreatment.
Further, the signal feature vector comprises the feature value of the time domain information entropy and the feature value of the frequency domain information entropy, and the extraction steps are as follows:
(1) communication signal S o (n) is obtained by performing nonlinear suppression preprocessing on the communication signal S (n) acquired in the step S1 in the step S2, wherein n =1,2, \ 8230;, k; k represents a signal sample data length;
(2) setting window length M, and transmitting the communication signal S in the step (1) o (n) partitioning to generate matrix a:
Figure BDA0003814699900000062
(3) performing Singular Value Decomposition (SVD) on the matrix A in the step (2) to obtain a non-zero singular value sigma 1 ≥σ 2 ≥…≥σ M ；
(4) Characteristic value E of Shannon entropy of singular spectrum Ssse The calculation formula of (a) is as follows:
Figure BDA0003814699900000071
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000072
(5) characteristic value E of singular spectral exponential entropy Ssee The calculation formula of (a) is as follows:
Figure BDA0003814699900000073
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000074
(6) characteristic value E of wavelet energy spectrum entropy Wese The steps of (1) are as follows;
for the time signal S after the nonlinear suppression pretreatment o Performing wavelet transformation to obtain wavelet coefficients on n scales, wherein the calculation formula is as follows:
Figure BDA0003814699900000075
when the scale is i, the energy value is m i ,i＝1,2,…,n；
(7) Characteristic value E of wavelet energy spectrum Shannon entropy Wese The calculation formula of (a) is as follows:
Figure BDA0003814699900000076
in the formula (I), the compound is shown in the specification,
Figure BDA0003814699900000077
(8) characteristic value E of power spectrum entropy Seps ；
Communication signal time sequence S after nonlinear suppression preprocessing o ＝{S o (i) Is defined as S 'and its power spectrum is' o (ω):
Figure BDA0003814699900000078
In the formula S o (ω) is the time series S o Fourier transform of (1);
(9) characteristic value E of Shannon entropy of power spectrum entropy value Seps The calculation formula of (2):
Figure BDA0003814699900000079
in the formula (I), the compound is shown in the specification,
Figure BDA00038146999000000710
in addition, the nonlinear inhibition pretreatment comprises truncation and zero setting treatment in a threshold inhibition method, so that a better sharp pulse inhibition effect is achieved, the influence of alpha stable distribution noise on signals is smaller, and the accuracy is improved.
The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It will be apparent to those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, and the scope of protection is still within the scope of the invention.

Claims (4)

1. A method for identifying a modulation mode of an underwater communication signal in an alpha stable distribution noise environment is characterized by comprising the following steps:
s1, acquiring a communication signal S (n) in an alpha stable distribution noise environment;
s2, the amplitude value of the communication signal S (n) collected in the step S1 is larger than the threshold tau r Performing nonlinear suppression preprocessing on the part to obtain a preprocessed communication signal S o (n)；
S3, extracting the communication signal S in the step S2 o The signal characteristic vector of (n) comprises a characteristic value of time domain information entropy and a characteristic value of frequency domain information entropy, and the characteristic value of singular spectrum Shannon entropy in the time domain characteristic is E Ssse The characteristic value of the singular spectral index entropy is E Ssee And the characteristic value of the entropy of the wavelet energy spectrum is E Wese The characteristic value of the Shannon entropy of the power spectrum in the frequency domain characteristic is E Seps ；
S4, constructing a signal feature vector;
the signal feature vector extracted according to the step S3 includes a feature value of singular spectrum shannon entropy, a feature value of singular spectrum index entropy, a feature value of wavelet energy spectrum entropy and a feature value of power spectrum shannon entropy, and is defined as a four-dimensional input parameter X = { E = { (E) } Ssse ,E Ssee ,E Wese ,E Seps }；
S5, training a GWOO-SVM classifier;
and (4) inputting the four-dimensional input parameter X defined in the step S4 into a GWO-SVM classification model for training and optimization, and identifying a modulation mode.
2. The method for identifying the modulation mode of the underwater communication signal in the environment of the alpha stable distributed noise according to claim 1, wherein: the signal is S o (n) the pair of signals can be represented as:
Figure FDA0003814699890000011
the above-mentioned tau r Can be represented by the formula tau r ＝(1+2τ o )τ Q Obtaining;
the above-mentioned tau o Is a constant coefficient, is set to 1.5, tau Q Is the second quarter-bit value of the modulus | S (n) | of the signal S (n), S o And (n) is the communication signal after nonlinear suppression pretreatment.
3. The method for identifying the modulation mode of the underwater communication signal in the environment of the alpha stable distributed noise according to claim 1, wherein:
the signal characteristic vector comprises the characteristic value of time domain information entropy and the characteristic value of frequency domain information entropy, and the extraction steps are as follows:
(1) the communication signal S o (n) is obtained by performing nonlinear suppression preprocessing on the communication signal S (n) acquired in the step S1 in the step S2, wherein n =1,2, \ 8230;, k; k represents a signal sample data length;
(2) setting window length M, and transmitting the communication signal S in the step (1) o (n) partitioning to generate a matrix A:
Figure FDA0003814699890000021
(3) performing Singular Value Decomposition (SVD) on the matrix A in the step (2) to obtain a non-zero singular value sigma 1 ≥σ 2 ≥…≥σ M ；
(4) Characteristic value E of Shannon entropy of singular spectrum Ssse The calculation formula of (c) is as follows:
Figure FDA0003814699890000022
in the formula (I), the compound is shown in the specification,
Figure FDA0003814699890000023
k＝1,2,…,M。
(5) characteristic value E of singular spectrum exponential entropy Ssee The calculation formula of (a) is as follows:
Figure FDA0003814699890000024
in the formula (I), the compound is shown in the specification,
Figure FDA0003814699890000025
k＝1,2,…,M。
(6) characteristic value E of the entropy of the wavelet energy spectrum Wese The steps of (1) are as follows;
for the time signal S after the nonlinear suppression pretreatment o Performing wavelet transformation to obtain wavelet coefficients on n scales, wherein the calculation formula is as follows:
Figure FDA0003814699890000031
when the scale is i, the energy value is m i ,i＝1,2,…,n；
(7) Characteristic value E of Shannon entropy of wavelet energy spectrum Wese The calculation formula of (a) is as follows:
Figure FDA0003814699890000032
in the formula (I), the compound is shown in the specification,
Figure FDA0003814699890000033
k＝1,2,…,n。
(8) characteristic value E of the power spectrum entropy Seps ；
Communication signal time sequence S after nonlinear suppression pretreatment o ＝{S o (i) Is defined as S 'and its power spectrum is' o (ω):
Figure FDA0003814699890000034
In the formula S o (ω) is the time series S o Fourier transform of (3);
(9) characteristic value E of Shannon entropy of power spectrum entropy value Seps The calculation formula of (c):
Figure FDA0003814699890000035
in the formula (I), the compound is shown in the specification,
Figure FDA0003814699890000036
k＝1,2,…,n。
4. the method for identifying the modulation mode of the underwater communication signal in the environment of the alpha stable distributed noise according to any one of claims 1 to 3, wherein: the nonlinear suppression preprocessing comprises truncation and zero setting processing in a threshold suppression method.
CN202211022794.2A 2022-08-25 2022-08-25 Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment Pending CN115378777A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211022794.2A CN115378777A (en) 2022-08-25 2022-08-25 Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211022794.2A CN115378777A (en) 2022-08-25 2022-08-25 Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment

Publications (1)

Publication Number Publication Date
CN115378777A true CN115378777A (en) 2022-11-22

Family

ID=84066886

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211022794.2A Pending CN115378777A (en) 2022-08-25 2022-08-25 Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment

Country Status (1)

Country Link
CN (1) CN115378777A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116582195A (en) * 2023-06-12 2023-08-11 浙江瑞通电子科技有限公司 Unmanned aerial vehicle signal spectrum recognition algorithm based on artificial intelligence

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116582195A (en) * 2023-06-12 2023-08-11 浙江瑞通电子科技有限公司 Unmanned aerial vehicle signal spectrum recognition algorithm based on artificial intelligence
CN116582195B (en) * 2023-06-12 2023-12-26 浙江瑞通电子科技有限公司 Unmanned aerial vehicle signal spectrum identification method based on artificial intelligence

Similar Documents

Publication Publication Date Title
CN112087774B (en) Communication radiation source individual identification method based on residual error neural network
CN111723701B (en) Underwater target identification method
CN106529428A (en) Underwater target recognition method based on deep learning
CN113014524B (en) Digital signal modulation identification method based on deep learning
CN111680737B (en) Radar radiation source individual identification method under differential signal-to-noise ratio condition
CN111010356A (en) Underwater acoustic communication signal modulation mode identification method based on support vector machine
CN107180259B (en) STAP training sample selection method based on system identification
CN104751183B (en) Classification of Polarimetric SAR Image method based on tensor MPCA
CN115378777A (en) Method for identifying underwater communication signal modulation mode in alpha stable distribution noise environment
CN116047427B (en) Small sample radar active interference identification method
CN112115821A (en) Multi-signal intelligent modulation mode identification method based on wavelet approximate coefficient entropy
CN112014801A (en) Composite interference identification method based on SPWVD and improved AlexNet
Petrov et al. Development of nonlinear filtering algorithms of digital half-tone images
CN114024808A (en) Modulation signal identification method and system based on deep learning
Dong et al. Radar signal modulation recognition based on bispectrum features and deep learning
Socheleau et al. Detection of mysticete calls: a sparse representation-based approach
CN111446998A (en) Direction-of-arrival estimation method based on deep learning
Bagwe et al. Automated Radar Signal Analysis Based on Deep Learning
CN115267713A (en) Semantic segmentation based intermittent sampling interference identification and suppression method
Gao et al. Supervised Contrastive Learning-Based Modulation Classification of Underwater Acoustic Communication
CN115037437A (en) Side channel attack method and system based on deep learning by using SpecAugment technology
CN115270891A (en) Method, device, equipment and storage medium for generating signal countermeasure sample
CN115310083A (en) Malicious software robust identification method based on contrast learning
CN114584441A (en) Digital signal modulation identification method based on deep learning
CN112434716B (en) Underwater target data amplification method and system based on condition countermeasure neural network

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