CN108268837A - Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic - Google Patents
Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic Download PDFInfo
- Publication number
- CN108268837A CN108268837A CN201711494014.3A CN201711494014A CN108268837A CN 108268837 A CN108268837 A CN 108268837A CN 201711494014 A CN201711494014 A CN 201711494014A CN 108268837 A CN108268837 A CN 108268837A
- Authority
- CN
- China
- Prior art keywords
- emitter
- analysis
- characteristic
- entropy
- feature
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/148—Wavelet transforms
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Complex Calculations (AREA)
Abstract
Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic, is related to specific emitter identification field.Include the following steps：1) Wavelet Entropy feature extraction is carried out to emitter Signals, obtains characteristic parameter；2) Chaos Feature Analysis is carried out to emitter Signals, extracts Chaos characteristic parameter, small echo entropy feature vector and chaos characteristic vector are combined, obtain assemblage characteristic vector, input feature vector grader realizes the identification to radiation source individual.Overcome the defects of currently used Time-Frequency Analysis Method does not account for emitter Signals non-linear nature, both the strong time-frequency for having played wavelet package transforms differentiates force characteristic, extract the multiple dimensioned local feature of signal, more pass through this nonlinear analysis method of chaos analysis, consider the nonlinear situation of signal entirety, so as to more accurately reflect the feature of emitter Signals, make the characteristic parameter extracted that there is stronger distinction.
Description
Technical field
The present invention relates to specific emitter identification field, more particularly, to a kind of radiation based on Wavelet Entropy and chaotic characteristic
Source method for extracting fingerprint feature.
Background technology
" fingerprint " of so-called signal of communication just refers to table of the fine feature of communication radiation source individual using signal as carrier
It is existing, it is commonly referred to as due to the hardware device difference between radiation source individual, and attached on the signal launched can be with
It is different from the personal feature of other radiation sources.In sea warfare is modernized, Underwater Targets Recognition is that first enemy has found and effectively to enemy
Hydroacoustic electronic warfare is carried out, first enemy is using weapon attacking, the premise gained mastery over the enemy.If it can effectively be extracted in complicated underwater environment
Go out the fingerprint characteristic of radiation source, different radiation sources can just be equipped and distinguished, it, can be further real by the analysis and identification to individual
Existing communication structure, the judgement of strategic plan provide crucial foundation for the corresponding plan of action deployment of my army, have important reality
Meaning.
Currently used Emitter Fingerprint feature extracting method mainly analyzes the time-frequency domain of emitter Signals, lacks
It is trapped in and handles the approximately linear signal of emitter Signals in this kind of method, do not account for the non-linear nature of signal, therefore
It can not reflect its nonlinear characteristic well.Chaos is a kind of important motion state of nonlinear system, can be characterized well
In Nonlinear Random Process particular law, chaos analysis can preferably embody as a kind of nonlinear analysis method
The substantive characteristics of nonlinear properties.
Wavelet transformation is a kind of effective Time-Frequency Analysis Method, can subtly extract letter of the signal under different scale
Breath, and wavelet package transforms have stronger time-frequency resolving power compared to wavelet transformation.Comentropy is metric signal uncertainty
Important parameter combines wavelet package transforms and the Analysis of Entropy, can reflect the multiple dimensioned local characteristics of signal and embody letter
The complexity of number entirety.
Invention content
It is an object of the invention to the Emitter Fingerprint feature extracting method based on time frequency analysis is overcome not reflect radiation
The defects of source signal non-linear nature, provides a kind of Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic.
The present invention includes the following steps：
1) Wavelet Entropy feature extraction is carried out to emitter Signals, obtains characteristic parameter；
2) Chaos Feature Analysis is carried out to emitter Signals, extracts Chaos characteristic parameter, by small echo entropy feature vector and mixed
Ignorant feature vector combines, and obtains assemblage characteristic vector, and input feature vector grader realizes the identification to radiation source individual.
It is described that Wavelet Entropy feature extraction is carried out to emitter Signals in step 1), obtain the specific method of characteristic parameter
Can be：N-layer wavelet packet decomposition is carried out to emitter Signals first, obtains 2nA sub-band calculates the comentropy of each sub-band, obtains
To 2nDimensional feature vector.
It is described that four parts are included to emitter Signals progress Chaos Feature Analysis in step 2)：Correlation dimension analysis,
The analysis of Kolmogorov entropys, Lyapunov index analysis and Hurst index analysis form 4 dimension chaos characteristic vectors；It docks first
The one-dimensional emitter Signals time series received carries out phase space reconfiguration, and correlation dimension is calculated on the basis of to reconstruct phase space analysis
Number, Kolmogorov entropys and Lyapunov indexes, Hurst indexes to emitter Signals time series then by using rescaled range
Method acquires.
Not the defects of emitter Signals non-linear nature not being accounted for the present invention overcomes currently used Time-Frequency Analysis Method,
Both the strong time-frequency for having played wavelet package transforms differentiates force characteristic, extracts the multiple dimensioned local feature of signal, more pass through chaos analysis
This nonlinear analysis method considers the nonlinear situation of signal entirety, so as to more accurately reflect the feature of emitter Signals,
Make the characteristic parameter extracted that there is stronger distinction.
Description of the drawings
Fig. 1 is the Emitter Fingerprint feature extracting method block diagram provided by the invention based on Wavelet Entropy and chaotic characteristic.
Fig. 2 is Wavelet Entropy feature extraction operation flow chart provided by the invention.
Fig. 3 is characteristic curve diagram of the emitter Signals of the embodiment of the present invention after correlation dimension is analyzed.
Fig. 4 is characteristic curve diagram of the emitter Signals of the embodiment of the present invention after the analysis of Kolmogorov entropys.
Fig. 5 is characteristic curve diagram of the emitter Signals of the embodiment of the present invention after Lyapunov index analysis.
Fig. 6 is characteristic curve diagram of the emitter Signals of the embodiment of the present invention after Hurst index analysis.
Specific embodiment
For the purpose of the present invention, scheme and advantage is more clearly understood, following embodiment will be with reference to attached drawing to the present invention
It is further described.
Fig. 1 is the Emitter Fingerprint feature extraction side based on Wavelet Entropy and chaotic characteristic according to invention embodiment
Method realizes block diagram.Emitter Signals are extracted with Wavelet Entropy feature and Chaos characteristic parameter respectively.The wherein extraction of Wavelet Entropy feature
Step is：N-layer wavelet packet decomposition is carried out to emitter Signals, obtains 2nA sub-band calculates the comentropy of each sub-band, obtains
2nDimensional feature vector.Chaos Feature Analysis include correlation dimension analysis, Lyapunov index analysis, Kolmogorov entropys analyze and
Hurst index analysis extracts 4 dimension chaos characteristics vector.Finally by Wavelet Entropy characteristic parameter and Chaos characteristic parameter joint group
Into 2n+ 4 dimensional feature vectors input grader carries out individual identification.
Fig. 2 is Wavelet Entropy feature extraction operation flow chart.First to emitter Signals sequence X (k) (k=1,2 ..., N)
N-layer wavelet packet decomposition is carried out, obtains 2nA sub-band reconstruction signal xi(t) (i=0,1,2 ..., 2n- 1, t=1,2 ..., m),
Calculate each discrete point ENERGY E of reconstruction signali(t)=| xi(t)|2, and then calculate the gross energy of each sub-band reconstruction signalEach sub-band information entropy is calculated by equation below：
pi(t) ratio of gross energy, S are accounted for for each discrete point energy of sub-band signaliI-th as required of sub-band
Comentropy.The comentropy of each sub-band is acquired by the above method, that is, forms 2nTie up small echo entropy feature vector.
Before carrying out Chaos Feature Analysis to emitter Signals, phase space reconfiguration, the higher-dimension after reconstruct are carried out to signal first
Spacing wave contains abundant nonlinear transformations.By choosing appropriate delay time T and Embedded dimensions m, by primary radiation
Source signal sequence X (k) (k=1,2 ..., N) it is reconstructed into higher dimensional space sequence Xi=[x (i), x (i+1) ..., x (i+ (m-1)
× τ)], i=1,2 ..., N- (m-1) × τ.
Emitter Signals are associated with dimensional analysis and the analysis of Kolmogorov entropys, by calculating correlation integral, is drawn
The method of correlation integral curve realizes that step is as follows：
(1) after m dimension phase space reconfigurations are carried out to signal, critical distance r is set, calculates any two points (X in phase spacei,Xj)
The distance between | | Xi,Xj| |, if less than r, retain the phase point pair.The step is repeated, phase point pair of the statistical distance less than r
Number calculates it with phase point to the ratio between sum, and then obtains correlation integral function：
Wherein m is Embedded dimensions；M=N- (m-1) τ represents phase point sum；θ is Heaviside functions, and expression formula is
(2) correlation dimension function D (m) is obtained by the following formula：
Different Embedded dimensions m is taken to carry out phase space reconfiguration to signal, under each Embedded dimensions, is taken different critical
Distance r calculates corresponding correlation integral C (r, m), draws ln [C (r, m)]-lnr curves.As shown in figure 3, curve is from upper in figure
Represent that Embedded dimensions m increases to 10 correlation integral curve from 1 successively under, with the increase of m, the slope of curve gradually tends to be steady
Fixed, corresponding slope is equal to correlation dimension D at this time.
(3) Kolmogorov entropys are calculated using the following formula, is denoted as K：
Embedded dimensions m=1, setting initial criticality distance r are set firstij, correlation integral C (r are calculated according to step (1)ij,
M), constantly reduce critical distance rijValue, until C (rij, m) value not with rijWhen reducing and changing, by C at this time
(rij, m) and value is denoted as C (r, m).It is incremented by the value of m, repeats the above steps, C (r, m+1) is calculated, and calculate K's according to formula (5)
Value, when the value of K changes no longer as m is incremented by, K values at this time are the value of Kolmogorov entropys, as shown in Figure 4.
The calculating of maximum Lyapunov exponent realizes that step is as follows using small data sets arithmetic：
(1) to one-dimensional emitter Signals sequence X (k) (k=1,2 ... N) do FFT transform, calculate P average period；
(2) to original signal sequence by each point X in the phase space that m ties up after reconstructingi, search its nearest neighbor point
The distance of nearest neighbor point pairIt is denoted as di(0), meet following condition：
di(0)=min (| | Xi-Xv| |), (i, v=1,2 ..., M；| i-v | ＞ P) (6)
(3) for each nearest neighbor point to XiWithCalculate its distance after j discrete time walks：
(4) j is walked for each discrete time, calculates the j step pitches of all nearest neighbor points pair from di(j) it is flat after taking the logarithm
Mean value y (j)：
Wherein q is non-zero di(j) number, Δ t are the sampling interval of primary radiation source signal sequence.
(5) choose different Embedded dimensions m, repeat step (2)~(4), draw y (j) in the case of different Embedded dimensions-
J curves.Cluster y (j)-j curves when being illustrated in figure 5 Embedded dimensions from 1 to 5, each curve have one section of approximately parallel portion
Point, regression straight line is made with least square method to this partial trace, the slope of the straight line is exactly maximum Lyapunov exponent, is denoted as
λ1。
The calculating of Hurst indexes uses rescaled range method, and step is as follows：
(1) one-dimensional emitter Signals sequence X (k) (k=1,2 ..., N) is averagely divided into the adjacent son that s length is l
Section yi(i=1,2 ..., s), N=ls；U of i-th of subinterval are denoted as yi,u(u=1,2 ..., l).
(2) each subinterval y is calculatediMean value Ei, accumulated deviation Zi, very poor RiWith variance Si：
Ri=max { Zi}-min{Zi} (11)
(3) R in all subintervals is calculatedi/SiMean value, be denoted as (R/S)l, have：
(4) different subinterval length l is taken, obtains different (R/S)l, draw lnl-ln (R/S)lCurve, with minimum two
Multiplication is fitted the slope of curve to get to the value H of Hurst indexes, as shown in Figure 6.
The above content is specific embodiment further descriptions made for the present invention, it is impossible to assert the present invention's
Specific implementation value is confined to these explanations.
Claims (8)
1. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic, it is characterised in that include the following steps：
1) Wavelet Entropy feature extraction is carried out to emitter Signals, obtains characteristic parameter；
2) Chaos Feature Analysis is carried out to emitter Signals, extracts Chaos characteristic parameter, small echo entropy feature vector and chaos is special
Sign vector combines, and obtains assemblage characteristic vector, and input feature vector grader realizes the identification to radiation source individual.
2. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
Described to carry out Wavelet Entropy feature extraction to emitter Signals in step 1), the specific method for obtaining characteristic parameter is：It is right first
Emitter Signals carry out n-layer wavelet packet decomposition, obtain 2nA sub-band calculates the comentropy of each sub-band, obtains 2nDimensional feature to
Amount.
3. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
It is described that four parts are included to emitter Signals progress Chaos Feature Analysis in step 2)：Correlation dimension analysis,
The analysis of Kolmogorov entropys, Lyapunov index analysis and Hurst index analysis form 4 dimension chaos characteristic vectors；It docks first
The one-dimensional emitter Signals time series received carries out phase space reconfiguration, and correlation dimension is calculated on the basis of to reconstruct phase space analysis
Number, Kolmogorov entropys and Lyapunov indexes, Hurst indexes to emitter Signals time series then by using rescaled range
Method acquires.
4. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
The Wavelet Entropy feature extraction operation flow is as follows：N-layer is carried out to emitter Signals sequence X (k) (k=1,2 ..., N) first
WAVELET PACKET DECOMPOSITION obtains 2nA sub-band reconstruction signal xi(t) (i=0,1,2 ..., 2n- 1, t=1,2 ..., m) calculate weight
Each discrete point ENERGY E of structure signali(t)=| xi(t)|2, and then calculate the gross energy of each sub-band reconstruction signal
Each sub-band information entropy is calculated by equation below：
pi(t) ratio of gross energy, S are accounted for for each discrete point energy of sub-band signaliThe information of i-th as required of sub-band
Entropy；The comentropy of each sub-band is acquired by the above method, that is, forms 2nTie up small echo entropy feature vector.
5. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
Chaos Feature Analysis is carried out to emitter Signals, phase space reconfiguration, the higher dimensional space signal packet after reconstruct are carried out to signal first
Containing abundant nonlinear transformations；By choosing appropriate delay time T and Embedded dimensions m, by primary radiation source signal sequence X
(k) (k=1,2 ..., N) it is reconstructed into higher dimensional space sequence：
Xi=[x (i), x (i+1) ..., x (i+ (m-1) × τ)], i=1,2 ..., N- (m-1) × τ.
6. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
Emitter Signals are associated with dimensional analysis and the analysis of Kolmogorov entropys, by calculating correlation integral, draws correlation integral
The method of curve, step are as follows：
(1) after m dimension phase space reconfigurations are carried out to signal, critical distance r is set, calculates any two points (X in phase spacei,Xj) between
Distance | | Xi,Xj| |, if less than r, retaining the phase point pair, repeat the step, statistical distance is less than the number of the phase point pair of r,
It is calculated with phase point to the ratio between sum, and then obtains correlation integral function：
Wherein m is Embedded dimensions；M=N- (m-1) τ represents phase point sum；θ is Heaviside functions, and expression formula is
(2) correlation dimension function D (m) is obtained by the following formula：
Different Embedded dimensions m is taken to carry out phase space reconfiguration to signal, under each Embedded dimensions, takes different critical distance
R calculates corresponding correlation integral C (r, m), draws ln [C (r, m)]-lnr curves, and curve represents embedded dimension successively from top to bottom
Number m increases to 10 correlation integral curve from 1, and with the increase of m, the slope of curve gradually tends towards stability, at this time corresponding slope
It is equal to correlation dimension D；
(3) Kolmogorov entropys are calculated using the following formula, is denoted as K：
Embedded dimensions m=1, setting initial criticality distance r are set firstij, correlation integral C (r are calculated according to step (1)ij, m), no
It is disconnected to reduce critical distance rijValue, until C (rij, m) value not with rijWhen reducing and changing, by C (r at this timeij, m) and value
It is denoted as C (r, m)；Be incremented by m value, repeat the above steps, C (r, m+1) be calculated, according to calculate K value, when K value no longer
Change as m is incremented by, K values at this time are the value of Kolmogorov entropys.
7. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
The calculating of maximum Lyapunov exponent uses small data sets arithmetic, and step is as follows：
(1) to one-dimensional emitter Signals sequence X (k) (k=1,2 ... N) do FFT transform, calculate P average period；
(2) to original signal sequence by each point X in the phase space that m ties up after reconstructingi, search its nearest neighbor pointArest neighbors
The distance of point pairIt is denoted as di(0), meet following condition：
di(0)=min (| | Xi-Xv| |), (i, v=1,2 ..., M；| i-v | ＞ P)
(3) for each nearest neighbor point to XiWithCalculate its distance after j discrete time walks：
(4) j is walked for each discrete time, calculates the j step pitches of all nearest neighbor points pair from di(j) the average value y after taking the logarithm
(j)：
Wherein q is non-zero di(j) number, Δ t are the sampling interval of primary radiation source signal sequence；
(5) different Embedded dimensions m is chosen, repeats step (2)~(4), draws y (j)-j songs in the case of different Embedded dimensions
Line, cluster y (j)-j curves when Embedded dimensions are from 1 to 5, each curve has one section of approximately parallel part, to this partial trace
Regression straight line is made with least square method, the slope of the straight line is exactly maximum Lyapunov exponent, is denoted as λ1。
8. the Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic as described in claim 1, it is characterised in that
The calculating of Hurst indexes uses rescaled range method, and step is as follows：
(1) one-dimensional emitter Signals sequence X (k) (k=1,2 ..., N) is averagely divided between the adjacent subarea that s length is l
yi(i=1,2 ..., s), N=ls；U of i-th of subinterval are denoted as yi,u(u=1,2 ..., l)；
(2) each subinterval y is calculatediMean value Ei, accumulated deviation Zi, very poor RiWith variance Si：
Ri=max { Zi}-min{Zi}
(3) R in all subintervals is calculatedi/SiMean value, be denoted as (R/S)l, have：
(4) different subinterval length l is taken, obtains different (R/S)l, draw lnl-ln (R/S)lCurve uses least square method
The slope of curve is fitted to get to the value H of Hurst indexes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711494014.3A CN108268837B (en) | 2017-12-31 | 2017-12-31 | Radiation source fingerprint feature extraction method based on wavelet entropy and chaotic characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711494014.3A CN108268837B (en) | 2017-12-31 | 2017-12-31 | Radiation source fingerprint feature extraction method based on wavelet entropy and chaotic characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108268837A true CN108268837A (en) | 2018-07-10 |
| CN108268837B CN108268837B (en) | 2020-07-03 |
Family
ID=62772932
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711494014.3A Active CN108268837B (en) | 2017-12-31 | 2017-12-31 | Radiation source fingerprint feature extraction method based on wavelet entropy and chaotic characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108268837B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948731A (en) * | 2019-03-29 | 2019-06-28 | 成都大学 | A kind of communication station individual discrimination method, system, storage medium and terminal |
| CN111695444A (en) * | 2020-05-21 | 2020-09-22 | 哈尔滨工业大学 | Radiation source individual feature extraction method based on wave atomic transformation |
| CN112711026A (en) * | 2020-12-19 | 2021-04-27 | 西北工业大学 | Target detection and signal frequency estimation method based on variable differential step length relative change entropy |
| CN113255545A (en) * | 2021-06-03 | 2021-08-13 | 中国人民解放军国防科技大学 | Communication radiation source individual identification method combining artificial features and depth features |
| CN116257780A (en) * | 2023-02-22 | 2023-06-13 | 中国人民解放军国防科技大学 | Unsupervised feature extraction and self-adaptive DBSCAN clustering method based on blind signal separation |
| CN117851808A (en) * | 2023-12-29 | 2024-04-09 | 中国人民解放军国防科技大学 | Robust nonlinear fingerprint feature extraction method based on differential phase space |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101650429A (en) * | 2009-09-04 | 2010-02-17 | 西北工业大学 | Method for extracting chaotic characteristic parameter from low-signal-to-noise ratio underwater target signal |
| CN101732110A (en) * | 2009-11-19 | 2010-06-16 | 杭州电子科技大学 | Method for fusing brain electricity and muscle electricity signal chaos characteristics for hand motion identification |
| US20120143568A1 (en) * | 2009-05-20 | 2012-06-07 | The Trustees Of The University Of Pennsylvania | Self-adaptive bio-signal and modulation device |
| CN103927556A (en) * | 2014-05-07 | 2014-07-16 | 天津工业大学 | ECG signal classifying method based on wavelet packet and approximate entropy |
| CN106529478A (en) * | 2016-11-08 | 2017-03-22 | 中国人民解放军空军工程大学 | Radar radiation source signal identification method according to three-dimensional entropy characteristic |
| CN106991393A (en) * | 2017-03-30 | 2017-07-28 | 上海电机学院 | Emitter Fingerprint feature extracting method based on Multifractal Dimension |
-
2017
- 2017-12-31 CN CN201711494014.3A patent/CN108268837B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120143568A1 (en) * | 2009-05-20 | 2012-06-07 | The Trustees Of The University Of Pennsylvania | Self-adaptive bio-signal and modulation device |
| CN101650429A (en) * | 2009-09-04 | 2010-02-17 | 西北工业大学 | Method for extracting chaotic characteristic parameter from low-signal-to-noise ratio underwater target signal |
| CN101732110A (en) * | 2009-11-19 | 2010-06-16 | 杭州电子科技大学 | Method for fusing brain electricity and muscle electricity signal chaos characteristics for hand motion identification |
| CN103927556A (en) * | 2014-05-07 | 2014-07-16 | 天津工业大学 | ECG signal classifying method based on wavelet packet and approximate entropy |
| CN106529478A (en) * | 2016-11-08 | 2017-03-22 | 中国人民解放军空军工程大学 | Radar radiation source signal identification method according to three-dimensional entropy characteristic |
| CN106991393A (en) * | 2017-03-30 | 2017-07-28 | 上海电机学院 | Emitter Fingerprint feature extracting method based on Multifractal Dimension |
Non-Patent Citations (3)
| Title |
|---|
| 周宏: "基于左右手运动想象的脑电信号的分类研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 李冬梅: "基于非线性动力学特征指标的癫痫脑电信号分类、定位与预测方法研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
| 潘伟: "硫化矿石碓自热过程的非线性及数值仿真研究", 《中国博士学位论文全文数据库 工程科技I辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948731A (en) * | 2019-03-29 | 2019-06-28 | 成都大学 | A kind of communication station individual discrimination method, system, storage medium and terminal |
| CN111695444A (en) * | 2020-05-21 | 2020-09-22 | 哈尔滨工业大学 | Radiation source individual feature extraction method based on wave atomic transformation |
| CN111695444B (en) * | 2020-05-21 | 2023-06-23 | 哈尔滨工业大学 | Wave atom transformation-based radiation source individual feature extraction method |
| CN112711026A (en) * | 2020-12-19 | 2021-04-27 | 西北工业大学 | Target detection and signal frequency estimation method based on variable differential step length relative change entropy |
| CN113255545A (en) * | 2021-06-03 | 2021-08-13 | 中国人民解放军国防科技大学 | Communication radiation source individual identification method combining artificial features and depth features |
| CN113255545B (en) * | 2021-06-03 | 2021-09-21 | 中国人民解放军国防科技大学 | Communication radiation source individual identification method combining artificial features and depth features |
| CN116257780A (en) * | 2023-02-22 | 2023-06-13 | 中国人民解放军国防科技大学 | Unsupervised feature extraction and self-adaptive DBSCAN clustering method based on blind signal separation |
| CN117851808A (en) * | 2023-12-29 | 2024-04-09 | 中国人民解放军国防科技大学 | Robust nonlinear fingerprint feature extraction method based on differential phase space |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108268837B (en) | 2020-07-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108268837A (en) | Emitter Fingerprint feature extracting method based on Wavelet Entropy and chaotic characteristic | |
| CN104113789B (en) | On-line video abstraction generation method based on depth learning | |
| CN109633588A (en) | Recognition Method of Radar Emitters based on depth convolutional neural networks | |
| CN102148987B (en) | Compressed sensing image reconstructing method based on prior model and 10 norms | |
| CN107132576B (en) | The seismic data Time-Frequency Analysis Method of wavelet transformation is squeezed based on second order sync | |
| CN102279390A (en) | Intra-pulse modulation and recognition method of low signal-to-noise radar radiation source signal | |
| CN109117747A (en) | Radar signal classification method based on cyclo-stationary Yu depth convolutional neural networks | |
| CN109471074A (en) | Recognition Method of Radar Emitters based on singular value decomposition Yu one-dimensional CNN network | |
| CN108052863A (en) | Electrical energy power quality disturbance recognition methods based on the maximum variance method of development | |
| CN104887263A (en) | Identity recognition algorithm based on heart sound multi-dimension feature extraction and system thereof | |
| CN106570183A (en) | Color picture retrieval and classification method | |
| CN102799892A (en) | Mel frequency cepstrum coefficient (MFCC) underwater target feature extraction and recognition method | |
| CN108042132A (en) | Brain electrical feature extracting method based on DWT and EMD fusions CSP | |
| CN111680737B (en) | Radar radiation source individual identification method under differential signal-to-noise ratio condition | |
| CN105991492A (en) | Frequency hopping (FH) signal identification method | |
| CN111010356A (en) | Underwater acoustic communication signal modulation mode identification method based on support vector machine | |
| CN109918417A (en) | Time series data adaptive segmentation, dimensionality reduction and characterizing method and application based on wavelet transformation | |
| CN102184451A (en) | Multi-wavelet fusion feature based neuron action feature extraction method | |
| CN114757224A (en) | Specific radiation source identification method based on continuous learning and combined feature extraction | |
| Li et al. | Research on feature importance evaluation of wireless signal recognition based on decision tree algorithm in cognitive computing | |
| CN104751470A (en) | Image quick-matching method | |
| CN104361339A (en) | Palm image extracting and identification method | |
| CN103778439A (en) | Body contour reconstruction method based on dynamic time-space information digging | |
| CN104680176B (en) | A kind of brain wave (EEG) Modulation recognition method based on the selection of non-gaussian neutrality vector characteristics | |
| CN110472540A (en) | A kind of Φ-OTDR vibration signal classification algorithm based on LMD-ICA-PNN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |