CN113657138B - Radiation source individual identification method based on equipotential star map - Google Patents
Radiation source individual identification method based on equipotential star map Download PDFInfo
- Publication number
- CN113657138B CN113657138B CN202010401692.6A CN202010401692A CN113657138B CN 113657138 B CN113657138 B CN 113657138B CN 202010401692 A CN202010401692 A CN 202010401692A CN 113657138 B CN113657138 B CN 113657138B
- Authority
- CN
- China
- Prior art keywords
- signal
- radiation source
- signal data
- equipotential
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000005855 radiation Effects 0.000 title claims abstract description 47
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000010586 diagram Methods 0.000 claims abstract description 34
- 238000004040 coloring Methods 0.000 claims abstract description 14
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 14
- 238000004364 calculation method Methods 0.000 claims abstract description 12
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000000605 extraction Methods 0.000 claims abstract description 7
- 238000005070 sampling Methods 0.000 claims description 23
- 239000003086 colorant Substances 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000001228 spectrum Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R21/00—Arrangements for measuring electric power or power factor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R29/00—Arrangements for measuring or indicating electric quantities not covered by groups G01R19/00 - G01R27/00
- G01R29/08—Measuring electromagnetic field characteristics
-
- 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
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- 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
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Measurement Of Radiation (AREA)
Abstract
A radiation source individual identification method based on an equipotential star map comprises the following steps: s1: collecting signal data of a plurality of radiation source individuals with the same model; s2: performing data preprocessing on the signal data to obtain a signal constellation diagram; s3: calculating the point density of the data points in the signal constellation diagram, and coloring the signal constellation diagram according to the point density calculation result to obtain an equipotential star map; s4: performing feature extraction and classification on the equipotential star map by using a convolutional neural network; s5: and outputting the identification result. According to the invention, IQ two paths of signal data are converted from a signal domain to a graph domain based on the equipotential star map, deep features of a radiation source individual carried by signals are automatically and effectively extracted through the convolutional neural network, each equipotential star map corresponds to the signal data, and the signal data corresponds to the radiation source individual, so that the radiation source individual is more effectively identified and classified.
Description
Technical Field
The invention relates to a radiation source individual identification method based on an equipotential star map, and belongs to the technical field of radiation source signal identification.
Background
Specific radiation source Identification (SEI) refers to a technique of performing characteristic measurement on a received electromagnetic signal, and determining an individual radiation source generating the signal according to existing prior information. Meanwhile, with the advent of the 5G and Internet of things times, the security of wireless networks has also met a great challenge. The radiation source individual identification technology can be applied to the projects such as wireless access authentication, electromagnetic environment supervision and the like, and the safety of a wireless network can be effectively improved. In the civil field, the radiation source individual identification technology is one of key technologies in the fields of cognitive radio, radio positioning, communication equipment fault detection and identification and the like, and plays a very important role.
With the development of science and technology, the electromagnetic signal environment where the signal is located is increasingly complex, the density of the radiation source is increased by times, and the electromagnetic signal patterns are complex and changeable, which all provide the radiation source with unprecedented challenges for identification. The common radiation source identification method is difficult to realize accurate, stable and reliable radiation source identification in a complex electromagnetic environment. Therefore, research and development of new radiation source identification methods and identification devices are urgently needed.
Disclosure of Invention
The technical problem to be solved by the invention is to provide the radiation source individual identification method based on the equipotential star map, which is characterized in that IQ two paths of signal data are converted from a signal domain to a map domain based on the equipotential star map, deep features of the radiation source individual carried by the signals are automatically and effectively extracted through a convolutional neural network, each equipotential star map corresponds to the signal data, and the signal data corresponds to the radiation source individual, so that the radiation source individual identification and classification can be more effectively realized.
The technical problems to be solved by the invention are realized by the following technical scheme:
the invention provides a radiation source individual identification method based on an equipotential star map, which comprises the following steps:
s1: collecting signal data of a plurality of radiation source individuals with the same model;
s2: performing data preprocessing on the signal data to obtain a signal constellation diagram;
S3: calculating the point density of the data points in the signal constellation diagram, and coloring the signal constellation diagram according to the point density calculation result to obtain an equipotential star map;
s4: performing feature extraction and classification on the equipotential star map by using a convolutional neural network;
S5: and outputting the identification result.
Preferably, the signal data includes I-way signal data and Q-way signal data.
Preferably, the data preprocessing is performed on the signal data to obtain a signal constellation specifically:
And taking the I-path signal data in the signal data of each sampling point of the plurality of radiation source individuals as the abscissa of a signal constellation diagram, and taking the Q-path signal data in the signal data of each sampling point as the ordinate of the signal constellation diagram, so that the signal data of the plurality of radiation source individuals are mapped into the signal constellation diagram.
Preferably, the point density calculation is performed using a window function.
Preferably, the window function is a density window function, and the normalized point density ρ at the sampling point is:
Wherein ρ (i) represents the normalized dot density of the ith sampling point, H (i) represents the abscissa of the ith sampling point on the signal constellation, V (i) represents the ordinate of the ith sampling point on the signal constellation, r represents half of the side length of the square area of the calculated normalized dot density, N represents the total number of sampling points in the signal constellation, and f (x) satisfies the following expression:
Preferably, the r=sqrt ((range (x)/30)/(2+ (range (y)/30)),
Where x represents the I-path signal data, y represents the Q-path signal data, range function represents the difference between the maximum and minimum values in the solution vector, and sqrt represents the square root function.
Preferably, the coloring processing of the signal constellation according to the point density calculation result specifically includes:
and selecting different colors, and coloring the data points with different normalized point density values or different normalized point density ranges in the signal constellation diagram.
Preferably, the colors in the gradient color bars using yellow to green to blue represent the normalized dot density from high to low in sequence.
Preferably, the convolutional neural network is AlexNet.
In summary, the IQ two-way signal data are converted from the signal domain to the graph domain based on the equipotential star map, deep features of the radiation source individuals carried by the signals are automatically and effectively extracted through the convolutional neural network, each equipotential star map corresponds to the signal data, and the signal data correspond to the radiation source individuals, so that the radiation source individuals are more effectively identified and classified.
The technical scheme of the invention is described in detail below with reference to the accompanying drawings and specific embodiments.
Drawings
FIG. 1 is a block diagram of an equipotential star map-based radiation source individual identification method of the present invention;
fig. 2 is a diagram showing a comparison of an equipotential star map-based radiation source individual identification method and a conventional feature extraction method according to the present invention.
Detailed Description
FIG. 1 is a block diagram of an equipotential star map-based radiation source individual identification method of the present invention. As shown in fig. 1, the invention provides a radiation source individual identification method based on an equipotential star map, which comprises the following steps:
s1: collecting signal data of a plurality of radiation source individuals with the same model;
s2: performing data preprocessing on the signal data to obtain a signal constellation diagram;
S3: calculating the point density of the data points in the signal constellation diagram, and coloring the signal constellation diagram according to the point density calculation result to obtain an equipotential star map;
s4: performing feature extraction and classification on the equipotential star map by using a convolutional neural network;
S5: and outputting the identification result.
In the S1, the signal data includes I-way signal data and Q-way signal data.
In the step S2, the data preprocessing is performed on the signal data to obtain a signal constellation diagram, specifically, an I-path signal data in the signal data of each sampling point of the plurality of radiation source individuals is taken as an abscissa of the signal constellation diagram, and a Q-path signal data in the signal data of each sampling point is taken as an ordinate of the signal constellation diagram, so that the signal data of the plurality of radiation source individuals is mapped into the signal constellation diagram. It should be added that the above examples are only illustrative, and the present invention is not limited to a specific manner of obtaining a signal constellation through data preprocessing, and those skilled in the art may select a suitable existing method to process signal data into a signal constellation according to practical situations.
In the signal constellation, different regions have different sampling point densities. In the present invention, a window function is illustratively employed for the point density calculation described in S3.
Specifically, in the present invention, the window function is preferably a density window function, and when the density window function slides on the signal constellation, the density window function counts the number of points in different area windows, and divides the number of points by the number of sampling points of the whole signal constellation to obtain a normalized point density ρ at the sampling points, namely
Where ρ (i) represents the normalized dot density of the ith sampling point, H (i) represents the abscissa of the ith sampling point on the signal constellation, V (i) represents the ordinate of the ith sampling point on the signal constellation, r represents half of the side length of the square area of the calculated normalized dot density, N represents the total number of sampling points in the signal constellation, and f (x) satisfies the following expression:
Preferably, r=sqrt ((range (x)/30)/(range (y)/30)/(2) where x represents I-path signal data, y represents Q-path signal data, range function represents the difference between the maximum value and the minimum value in the solution vector, and sqrt represents a square root function).
By the above method, the point density at each data point in the signal constellation can be obtained, and of course, the invention is not limited to the calculation mode of the point density, and those skilled in the art can select other well-known point density calculation modes.
In the step S3, the coloring processing of the signal constellation according to the point density calculation result specifically includes: and selecting different colors, and coloring the data points with different normalized point density values or different normalized point density ranges in the signal constellation diagram. The present invention is not limited to specific rules of coloring, for example, coloring the signal constellation by a preset color bar containing a corresponding relationship between a dot density and a color. Preferably, the colors in the gradient color bars using yellow to green to blue represent the normalized dot density from high to low in sequence.
By the coloring process in S3, an image having RGB information, that is, an equipotential star map can be obtained can be formed. The traditional signal constellation diagram has defects when being used for extracting signal characteristics, for example, in the actual data acquisition process, the internal noise of a device can seriously pollute radio frequency signals, so that the acquired constellation diagrams of different radio frequency signals have the same graph, compared with the traditional signal constellation diagram, the equipotential star diagram provided by the invention has the coloring step, the characteristic of a deeper level of an individual can be displayed, which is equivalent to the addition of a dimension and the more display of the difference among the individuals, and therefore, the invention can be more accurate when the individual identification is carried out.
In S4, the present invention does not limit the type of the convolutional neural network as long as it can achieve picture recognition classification. Preferably, convolutional neural network AlexNet is used in the present invention to automatically extract features and classify the equivalent polar star map. Since AlexNet is a public network, the specific process in S4 is not described here again.
Fig. 2 is a diagram showing a comparison of an equipotential star map-based radiation source individual identification method and a conventional feature extraction method according to the present invention. In fig. 2, the method of combining convolutional neural network AlexNet with an equipotential star map is compared with the conventional Welch, yule-Walker, and Burg methods to extract power spectrum features from signal data. The data set used in this comparative experiment is presented below: eight power amplifier output data of the same model and the same batch are collected by using the baseband signal receiver as a research object, the working center frequency of the amplifier module is 433MHz, each individual uses 100 samples for research, and each sample has 20000 sample points. When the power spectrum feature extraction is carried out, 2048 is used for FFT point number, PCA is used for dimension reduction after the power spectrum is obtained, and finally a K neighbor classifier is used for classification and identification.
As can be seen from fig. 2, the identification accuracy of the radiation source individual identification method using the convolutional neural network and the equipotential star map of the present invention is superior to that of the conventional method.
In summary, the IQ two-way signal data are converted from the signal domain to the graph domain based on the equipotential star map, deep features of the radiation source individuals carried by the signals are automatically and effectively extracted through the convolutional neural network, each equipotential star map corresponds to the signal data, and the signal data correspond to the radiation source individuals, so that the radiation source individuals are more effectively identified and classified.
Claims (4)
1. The method for identifying the radiation source individuals based on the equipotential star map is characterized by comprising the following steps of:
s1: collecting signal data of a plurality of radiation source individuals with the same model;
s2: performing data preprocessing on the signal data to obtain a signal constellation diagram;
S3: calculating the point density of the data points in the signal constellation diagram, and coloring the signal constellation diagram according to the point density calculation result to obtain an equipotential star map;
s4: performing feature extraction and classification on the equipotential star map by using a convolutional neural network;
S5: outputting a recognition result;
the signal data comprise I-path signal data and Q-path signal data;
the data preprocessing is performed on the signal data, and the signal constellation diagram is specifically:
Taking the I-path signal data in the signal data of each sampling point of a plurality of radiation source individuals as the abscissa of a signal constellation diagram, and taking the Q-path signal data in the signal data of each sampling point as the ordinate of the signal constellation diagram, so that the signal data of the plurality of radiation source individuals are mapped into the signal constellation diagram;
Performing the point density calculation using a window function;
the window function is a density window function, and the normalized point density ρ at the sampling point is:
Wherein ρ (i) represents the normalized dot density of the ith sampling point, H (i) represents the abscissa of the ith sampling point on the signal constellation, V (i) represents the ordinate of the ith sampling point on the signal constellation, r represents half of the side length of the square area of the calculated normalized dot density, N represents the total number of sampling points in the signal constellation, and f (x) satisfies the following expression:
the r=sqrt ((range (x)/30)/(2+ (range (y)/30)),
Where x represents the I-path signal data, y represents the Q-path signal data, range function represents the difference between the maximum and minimum values in the solution vector, sqrt represents the open root number.
2. The method for identifying an individual radiation source based on an equipotential star map according to claim 1, wherein the coloring process of the signal constellation according to the point density calculation result specifically comprises:
and selecting different colors, and coloring the data points with different normalized point density values or different normalized point density ranges in the signal constellation diagram.
3. The method of claim 2, wherein the normalized dot density from high to low is represented sequentially using colors in a yellow to green to blue gradient color bar.
4. The method for identifying individuals as sources of radiation based on an equipotential star map of claim 1 wherein said convolutional neural network is AlexNet.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010401692.6A CN113657138B (en) | 2020-05-12 | 2020-05-12 | Radiation source individual identification method based on equipotential star map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010401692.6A CN113657138B (en) | 2020-05-12 | 2020-05-12 | Radiation source individual identification method based on equipotential star map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113657138A CN113657138A (en) | 2021-11-16 |
| CN113657138B true CN113657138B (en) | 2024-05-21 |
Family
ID=78476753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010401692.6A Active CN113657138B (en) | 2020-05-12 | 2020-05-12 | Radiation source individual identification method based on equipotential star map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113657138B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116628472B (en) * | 2023-04-28 | 2024-08-30 | 哈尔滨工程大学 | Characteristic correlation-based radiation source individual signal diagram structure mapping method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4516263A (en) * | 1982-04-30 | 1985-05-07 | The Charles Stark Draper Laboratory, Inc. | Spatially integral, video signal processor |
| JP2006010418A (en) * | 2004-06-24 | 2006-01-12 | Nec Engineering Ltd | Fixed star sensor |
| JP2008040683A (en) * | 2006-08-03 | 2008-02-21 | Matsushita Electric Works Ltd | Signal identification method and signal identification device |
| CN101278316A (en) * | 2005-08-02 | 2008-10-01 | 美国西门子医疗解决公司 | System and method for automatic segmentation of vessels in breast MR sequences |
| CN102231720A (en) * | 2011-07-25 | 2011-11-02 | 南京信息工程大学 | Wavelet blind equalization method for fusing spline function Renyi entropy and time diversity |
| JP2017150924A (en) * | 2016-02-24 | 2017-08-31 | 三菱電機株式会社 | Angle measurement device |
| EP3232371A1 (en) * | 2016-04-15 | 2017-10-18 | Ricoh Company, Ltd. | Object recognition method, object recognition device, and classifier training method |
| CN108427987A (en) * | 2018-03-08 | 2018-08-21 | 四川大学 | A kind of Modulation Mode Recognition method based on convolutional neural networks |
| CN110427893A (en) * | 2019-08-06 | 2019-11-08 | 西安电子科技大学 | A kind of specific emitter identification method, apparatus and computer storage medium based on convolutional neural networks |
-
2020
- 2020-05-12 CN CN202010401692.6A patent/CN113657138B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4516263A (en) * | 1982-04-30 | 1985-05-07 | The Charles Stark Draper Laboratory, Inc. | Spatially integral, video signal processor |
| JP2006010418A (en) * | 2004-06-24 | 2006-01-12 | Nec Engineering Ltd | Fixed star sensor |
| CN101278316A (en) * | 2005-08-02 | 2008-10-01 | 美国西门子医疗解决公司 | System and method for automatic segmentation of vessels in breast MR sequences |
| JP2008040683A (en) * | 2006-08-03 | 2008-02-21 | Matsushita Electric Works Ltd | Signal identification method and signal identification device |
| CN102231720A (en) * | 2011-07-25 | 2011-11-02 | 南京信息工程大学 | Wavelet blind equalization method for fusing spline function Renyi entropy and time diversity |
| JP2017150924A (en) * | 2016-02-24 | 2017-08-31 | 三菱電機株式会社 | Angle measurement device |
| EP3232371A1 (en) * | 2016-04-15 | 2017-10-18 | Ricoh Company, Ltd. | Object recognition method, object recognition device, and classifier training method |
| CN108427987A (en) * | 2018-03-08 | 2018-08-21 | 四川大学 | A kind of Modulation Mode Recognition method based on convolutional neural networks |
| CN110427893A (en) * | 2019-08-06 | 2019-11-08 | 西安电子科技大学 | A kind of specific emitter identification method, apparatus and computer storage medium based on convolutional neural networks |
Non-Patent Citations (3)
| Title |
|---|
| Wireless Physical-Layer Identification Assisted 5G Network Security;J. Chang, Y. Xiao and Z. Zhang;《IEEE INFOCOM 2019》;第1-5页 * |
| 基于星座图的混合MPSK信号盲识别算法;赵春晖;杜宇;;黑龙江大学工程学报;20121130(04);全文 * |
| 基于星座图的通信辐射源个体识别方法;王威;李诗娴;王新;;湖南城市学院学报(自然科学版)(05);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113657138A (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Reus-Muns et al. | Trust in 5G open RANs through machine learning: RF fingerprinting on the POWDER PAWR platform | |
| CN111832417B (en) | Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning | |
| WO2018176889A1 (en) | Method for automatically identifying modulation mode for digital communication signal | |
| CN106130942A (en) | A kind of wireless communication signals Modulation Identification based on Cyclic Spectrum and method for parameter estimation | |
| CN104966085A (en) | Remote sensing image region-of-interest detection method based on multi-significant-feature fusion | |
| CN112749633B (en) | Separate and reconstructed individual radiation source identification method | |
| CN113657138B (en) | Radiation source individual identification method based on equipotential star map | |
| CN114422311A (en) | Signal modulation identification method and system combining deep neural network and expert prior characteristics | |
| CN111310719B (en) | Unknown radiation source individual identification and detection method | |
| CN111260645A (en) | Method and system for detecting tampered image based on block classification deep learning | |
| CN115294615A (en) | Radio frequency fingerprint identification method based on machine learning | |
| Wang et al. | A cooperative spectrum sensing method based on signal decomposition and K-medoids algorithm | |
| CN115186714A (en) | Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition | |
| Wang et al. | Specific emitter identification based on deep adversarial domain adaptation | |
| Ying et al. | Differential complex-valued convolutional neural network-based individual recognition of communication radiation sources | |
| Zhang et al. | Data augmentation aided few-shot learning for specific emitter identification | |
| CN118260571A (en) | Satellite signal individual identification method based on multi-mode feature fusion and deep learning | |
| CN111310680B (en) | Radiation source individual identification method based on deep learning | |
| Hu et al. | Feature fusion convolution-aided transformer for automatic modulation recognition | |
| CN115809426A (en) | Radiation source individual identification method and system | |
| CN115409056A (en) | Automatic modulation identification method for large dynamic signal-to-noise ratio | |
| CN114727293A (en) | Identity counterfeit attack identification method based on radio frequency fingerprint and unsupervised learning | |
| Zhou et al. | Application of Deep Learning in Signal Modulation Recognition System for Wireless Communication | |
| Luo et al. | Extraction and recognition of fingerprint characteristic of mobile terminal’s transient signal | |
| CN118568476B (en) | Radiation source individual identification method based on fusion of domain knowledge and depth 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |