CN115186714A - Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition - Google Patents
Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition Download PDFInfo
- Publication number
- CN115186714A CN115186714A CN202210854425.3A CN202210854425A CN115186714A CN 115186714 A CN115186714 A CN 115186714A CN 202210854425 A CN202210854425 A CN 202210854425A CN 115186714 A CN115186714 A CN 115186714A
- Authority
- CN
- China
- Prior art keywords
- feature
- spectral
- frequency spectrum
- radiation source
- characteristic
- 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
- 238000001228 spectrum Methods 0.000 title claims abstract description 85
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 37
- 238000003199 nucleic acid amplification method Methods 0.000 title claims abstract description 37
- 230000003321 amplification Effects 0.000 title claims abstract description 36
- 230000005855 radiation Effects 0.000 claims abstract description 114
- 238000004364 calculation method Methods 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 claims abstract description 4
- 230000003595 spectral effect Effects 0.000 claims description 72
- 238000012545 processing Methods 0.000 claims description 15
- 238000004458 analytical method Methods 0.000 claims description 14
- 230000003044 adaptive effect Effects 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 12
- 230000009286 beneficial effect Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 238000002156 mixing Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 230000009467 reduction Effects 0.000 claims description 2
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 claims 1
- 229910052739 hydrogen Inorganic materials 0.000 claims 1
- 239000001257 hydrogen Substances 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 10
- 238000012935 Averaging Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 239000003086 colorant Substances 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000037237 body shape Effects 0.000 description 2
- 238000012512 characterization method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 230000016507 interphase Effects 0.000 description 1
- 230000001050 lubricating effect Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000007784 solid electrolyte Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- 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
- Image Analysis (AREA)
Abstract
The invention belongs to the technical field of radiation source fingerprint identification, and particularly relates to a network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition, which specifically refers to the steps of carrying out high-precision frequency estimation on a plurality of network card signals received in a wireless network signal detection process, extracting frequency spectrum fingerprint features, solving the problem of small individual feature difference of radiation sources among network card devices, and realizing detail amplification. The method comprises the steps of extracting a local frequency spectrum with richer fingerprint information as fingerprint features, obtaining a public part of the frequency spectrum features of the radiation sources by carrying out weighting and averaging on the frequency spectrum features of the multiple radiation sources, obtaining main body components and stray components of the frequency spectrum features based on self-adaptive decomposition, realizing detail amplification in the frequency spectrum features based on relevant operation in a calculation process, eliminating intentional modulation amount, obtaining amplified final frequency spectrum features, and laying a foundation for subsequently improving the identification success rate of the network card equipment.
Description
Technical Field
The invention belongs to the technical field of radiation source fingerprint identification, and particularly relates to a network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition.
Background
Radiation source fingerprinting, also called Specific Emitter Identification (SEI), refers to a technique for extracting features in an electromagnetic signal that can represent differences in the hardware of a radiation source transmitter, thereby identifying a Specific radiation source device. The hardware difference characteristic is irrelevant to the transmitted signal pattern parameter, is independent of the transmission information, cannot be forged and cannot be avoided. This radiation source hardware specific information is carried on the signal in an unintentionally modulated manner.
For radiation source fingerprinting, a core problem is the accurate characterization of the radiation source hardware differences. The key information for identification is in fact the unintentional modulation about the device that is embedded in the signal. Compared to intentional modulation, unintentional modulation is not a major component and is subtle and imperceptible. Moreover, the calculated characteristics of signals having the same intended modulation are highly similar, and the accompanying fingerprint information, which is capable of representing the unintended modulation of their transmission sources, is very weak in practice.
There is thus an important problem in radiation source fingerprinting: hardware differences of different radiation sources are small in practice, most of extracted feature information represents commonalities of equipment, only slight individual differences exist, and great trouble is brought to subsequent classification and identification.
In order to solve this problem, some studies have been directed to using a classifier with higher accuracy, for example, a neural network for identifying subtle differences (CHEN P B, GUO Y L, LI g. Differential adaptive networks for specific evaluation, electronics Letters,2020,56 (1)); some methods estimate and eliminate the intentional modulation of the Signal (MERCHENT K, REVAY S, STANTCHEV G, NOUSAIN B, deep Learning for RF Device Fingerprinting in Cognitive Communication Networks, IEEE Journal of Selected Topics in Signal Processing,2018.12 (1): 160-167.doi: 10.1109/jstsp.2018.2796446.), however, the estimation accuracy affects the unintentional modulation information and introduces new errors; some methods of suppressing the main Signal components by Wavelet transformation (WU L W, NIU J P, WANG Z, et al. Primary Signal compression Based on synchronized queuezed Wavelet Transform, journal of Electronics & Information Technology,2019,42 (8): 2045-2052.) do, however, directly process the Signal waveform in the time domain.
Compared with other domain characteristics, the fingerprint characteristics of the frequency domain radiation source are relatively stable and are less influenced by noise environment. But the difference between the signal spectra is small, especially the signal spectra of the same modulation pattern, modulation parameter are highly similar. The fingerprint features of the radiation source extracted on this basis are very subtle and require further processing.
Although the same modulated signal from different devices is spectrally similar in general, subtle differences always exist, the latter being of interest for radiation source fingerprinting. Through specific processing and signal self-adaptive decomposition means, the main body component and the difference component which represents the information of the radiation source equipment can be decomposed. Variational Modal Decomposition (VMD) is an adaptive, completely non-recursive method of modal Variational and signal processing. The method has the advantages that the modal decomposition number can be determined, the self-adaptability of the method is shown in that the modal decomposition number of a given sequence is determined according to the actual situation, the optimal center frequency and the limited bandwidth of each Mode can be matched in a self-adaptive Mode in the subsequent searching and solving processes, the effective separation of inherent modal components (IMF) and the frequency domain division of signals can be realized, the effective decomposition components of the given signals are obtained, and the optimal solution of the variation problem is finally obtained.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: in the problem of individual identification of the radiation source, it is most critical to extract the unintentional modulation information contained in the signal. Unintentional modulation information tends to be lower in energy than intentional modulation, "swamped" among intentional modulation information. If the modulation modes of signals emitted by different radiation source devices are the same, the modulation parameters are the same, and even the information is the same, the intentional modulation is completely the same under the condition, the extracted features are highly similar due to the influence of the intentional modulation, the difference between individuals is reflected only in a slight place, and the difficulty of fingerprint identification is greatly improved. In order to better exert the main function of unintentional modulation in SEI and fully utilize the advantages of the frequency spectrum characteristic in fingerprint characteristic characterization, the invention provides a network card frequency spectrum fingerprint characteristic amplification method based on characteristic correlation and adaptive decomposition.
The existing radiation source fingerprint identification method is mainly divided into two steps, wherein the first step is to extract features based on a large amount of data of known labels; and the second step is to extract and analyze the characteristics of the received signals without the labels based on the existing characteristics as a database, and compare the extracted characteristics with the existing characteristic data in the database to complete the identification of the target of the unknown label.
Compared with the processing flow of the existing method, the method needs to additionally add a step to analyze the characteristic components and eliminate the parts which are not beneficial to the radiation source fingerprint identification, so the method is mainly divided into three steps, which are respectively as follows:
(1) Data processing and spectrum feature extraction: the method is consistent with the first step in the prior method, and has the difference that the preprocessing precision and efficiency are improved, and a new spectrum characteristic is extracted.
(2) Extracting frequency spectrum reference features and analyzing feature components: this is a newly added step of the present invention. To achieve feature amplification, fingerprint features need to be analyzed based on data of known tags. Spectral reference characteristics and an intentional modulation amount unrelated to the SEI are obtained in this step. This step need only be performed once.
(3) Extracting the amplification characteristics of the label-free data: and extracting features of the newly received data without the label and realizing target identification, which corresponds to the second step of the existing method, namely finishing the identification of the radiation source aiming at the newly received data.
The invention adopts the technical scheme that a network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition comprises the following steps:
s1: data processing and spectrum feature extraction;
this step is mainly intended to process and extract the spectral features to be amplified from the signal received at the receiver. The final output of this step is the spectral feature to be amplified, denoted by the symbol f (k). The method comprises the following specific steps:
s1.1, data reception and preprocessing
After receiving the signal, the receiver performs preprocessing such as detection, filtering, noise reduction and the like on the original signal to obtain a signal sample x to be processed 0 (t)；
According to the IEEE802.11 series standards, a wireless network signal has a specific frame header structure (Preamble), and the method mainly extracts the fingerprint characteristics of the radiation source based on the signal frame header part.
Signal sample x to be processed 0 (t) analyzing and detecting to obtain frame header part x of the signal sample to be processed 1 (t) (pretreatment method of Signal frame header is described in L.Sun, X.Wang, A.Yang and Z.Huang, "Radio Frequency knowledge Extraction Based on Multi-Dimension addition Engine," in IEEE Signal Processing Letters, vol.27, pp.471-475, 2020.).
S1.2, estimating and eliminating signal frequency offset
The positions of frequency bands where actually transmitted signals are located are different, actual reception also causes a certain frequency offset, and the existence of the frequency offset value causes a certain deviation to exist in the range of subsequently extracted spectral features, so that a frame header part x of a signal sample to be processed needs to be accurately estimated 1 Frequency deviation f of (t) 0 . Due to frames of signal samples to be processedHead part x 1 (t) is the signal sample x to be processed 0 (t) a part which is truncated, thus x 0 (t) and x 1 (t) the frequency offset is the same, based on x being longer in length in the frequency offset estimation 0 (t) the estimation precision can be improved by carrying out frequency offset estimation; the method comprises the following specific steps:
s1.2.1 Fourier interpolation algorithm-based pair x 0 (t) frequency offset value f 0 Estimating (specific estimation method is shown in Absutanois E, mulgrew B. Iterative frequency estimation by interpolation on Fourier coefficients. IEEE Transactions on signal processing 2005; 1237-42.);
s1.2.2 according to f 0 X is to be 1 (t) moving to zero frequency to obtain a baseband signal and eliminating the influence of frequency offset;
s1.2.3, performing mean value removal on baseband signals, and performing energy normalization to obtain a signal x to be subjected to feature extraction 2 (t)；
S1.3, spectral feature calculation
Fingerprint information is not uniformly distributed on spectral features, and distribution differences are more obvious near the center Frequency (lubricating Sun, xiaong Wang, zhitao Huang, and Baoguo Li, radio Frequency Fingerprint Extraction based on Feature induction. IEEE Internet of Things Journal,2022. DOI.
Therefore, the invention only reserves the frequency spectrum with a certain width near the main lobe by extracting the local frequency spectrum with most abundant fingerprint information in the frequency spectrum as the fingerprint characteristic of the radiation source, rather than relying on the whole frequency spectrum information. The calculation method is simple and convenient, and compared with the method of utilizing all frequency spectrum information, the signal characteristic dimension is greatly reduced.
The method comprises the following specific steps:
s1.3.1 calculating the Signal x to be feature extracted at 0 frequency 2 Frequency spectrum f of (t) 0 (k)：
Wherein,
representing a fourier transform; k represents a frequency index value of Fourier transform and is also a dimension index value of the spectrum characteristic; n is a radical of FFT Points representing a Fourier transform; i- | denotes calculating the amplitude value.
S1.3.2 Signal x to be subjected to feature extraction 2 Frequency spectrum f of (t) 0 (k) Has a main lobe of 3dB width W B The bandwidth of the left and right of the peak value is taken as lambda 1 W B As preliminary spectral feature value f 1 (k),k＝1,...,K 2 Wherein λ is 1 Weighting parameter, K, representing the spectral range 2 Indicating the length of the spectrum after truncation, i.e. preliminary spectral characteristic f 1 (k) Of the length of (c). Experiments show that 1 =2 can satisfy the identification requirement.
S1.3.3 pairs of f 1 (k) Calculating the second derivative
Select to satisfy
Minimum value k of min Wherein λ is 2 Denotes the weight coefficient, λ 2 ＞0。
Selection of f 1 (k) Upper [ k ] min ,K 2 ]The segment serves as the spectral feature f (k) to be amplified.
S2: extracting frequency spectrum reference features and analyzing feature components;
the step is mainly to comprehensively utilize the signal characteristics of all known tags of all radiation sources extracted in the step S1, calculate common components of the signals as frequency spectrum reference characteristics, amplify details based on correlation operation, and finally complete component analysis based on adaptive decomposition to obtain an intentional modulation amount which is not beneficial to radiation source fingerprint identification.
The output of this step is the spectral reference feature
And an intentional modulation amount h (j). In practical applications, the main purpose of this stage is to obtain the spectral basisQuasi-feature
And an intentional modulation h (j) which is obtained based on the entire training data of all the radiation sources and which only needs to be performed once. The method comprises the following specific steps:
s2.1, calculating frequency spectrum reference characteristics
S1 provides a method for calculating the spectral feature f (k) of a signal sample to be amplified. Multiple signal samples of multiple radiation sources are to be processed in the SEI problem. The reference feature of the present invention is a common feature of all training samples of multiple radiation sources, and thus requires processing of multiple signal samples of multiple radiation sources.
Assuming that M wireless network cards are needed to carry out individual identification analysis of radiation sources, the mth radiation source has N m The signal sample, the spectral feature to be amplified of the ith sample can be represented as f i m (k),i＝1,...,N m ,m＝1，…,M。
Here, f i m (k) I.e. the output f (k) of S1, which is information representing the radiation source and the signal sample, the upper corner mark m represents the label of the corresponding radiation source, and the lower corner mark i represents the ith signal sample of the current radiation source.
S2.1.1 according to the same method as S1, calculating to obtain spectral characteristics to be amplified of all samples of all radiation sources and forming a set { f } i m (k)},i＝1,...,N m ,m＝1，…,M。
S2.1.2 calculating the characteristic mean value of each radiation source
S2.1.3, obtaining a frequency spectrum reference characteristic by calculating a weighted average value
Wherein alpha is m Is the characteristic weight of the mth radiation source, satisfies
And the radiation source characteristic weight is determined according to the data scale and the characteristic distribution condition of each radiation source. Under normal conditions
Spectral reference feature
The common part obtained by calculating all training samples of all radiation sources is considered as the common part of the radiation sources, i.e. the intentional modulation which is not beneficial for individual identification of the radiation sources, and is suppressed below.
S2.2, spectral reference feature detail amplification
Spectral reference signature obtained for S2.1
Performing autocorrelation operation to realize detail amplification to obtain standard frequency spectrum characteristics
Wherein,
representing an autocorrelation operation. The characteristic dimension here varies, J =2K-1.
In this step
Is to a weighted average value-frequency spectrum reference characteristic based on multiple radiation source signals
The obtained radiation source after the autocorrelation operation is bilaterally symmetrical in distribution, and can reflect the commonality of the radiation source. To match the spectral reference characteristics in S2.1
The distinguishing is carried out by the following steps,
referred to as standard spectral signature.
S2.3, adaptive decomposition of standard spectral features
Standard spectrum characteristics obtained from S2.2
VMD decomposition is performed resulting in different components, which are called IMF. The VMD Decomposition specifically implements the process reference (DRAGOMIRETSKIY K, ZOSSO D. Variational Mode Decomposion, IEEE trans. On Signal Processing,2014,62 (3): 531-544. DOI.
Assuming that the number of decomposition layers is L, the obtained decomposition result can be expressed as:
wherein, V, V, omega are respectively the time domain combination, the frequency domain combination and the center frequency of IMF.
The time domain combination V of IMFs is expressed as:
V＝[v 1 (j),v 2 (j),…,v L (j)],j＝1,2,...,J
wherein v is l (j) L =1, 2.. And L is the single IMF component
S2.4, spectral feature analysis
Performing spectral component analysis on the L IMF components obtained by VMD decomposition in S2.3, specifically as follows:
s2.4.1 calculating different IMF components v l (j) Energy value of (d):
s2.4.2 according to { E l Sorting the IMF, wherein the IMF component with the highest energy is the principal component
Low energy as stray component
S2.4.3 selected principal component
And stray component
Is represented as
Obtaining an intentional modulation feature h (j):
s2, calculating the obtained frequency spectrum reference characteristics
Standard spectral features
Are cumulatively weighted based on a plurality of samples from a plurality of radiation sources and are therefore considered to be a common component common to all radiation sources.
The intentional modulation feature h (j) is calculated on the basis of these common components, is considered to be common to the signals of different radiation sources, is an irrelevant component which is not of concern to the SEI, and needs to be removed. Obtaining h (j) means that the analysis of the spectral feature is completed.
S3: extracting the amplification characteristics of the label-free data;
the true purpose of the SEI is to enable the identification of the radiation source for newly received non-tagged signals. Therefore, the newly received signal is mainly analyzed and processed in the stage, and the amplified final spectrum fingerprint features are extracted, so that the individual identification performance of the radiation source is improved.
Note that each newly received signal needs to first repeat the data processing and spectral feature extraction of S1, which is the basis for feature amplification.
Then, the spectral feature of the new signal is compared with the spectral reference feature
And (4) carrying out correlation operation, and finally, eliminating the intentional modulation amount h (j) to obtain the amplified final spectrum characteristic.
The output of this stage is the final spectral feature after amplification
Since here the radiation source label to which the signal corresponds is unknown, denoted by the symbol n. The method comprises the following specific steps:
s3.1, correlation transformation of samples to be identified
S3.1.1 calculating the spectral feature f of the newly received signal according to S1 i n (k) In that respect In the identification phase, the radiation source tag is denoted by n, since the corresponding network card tag is unknown. For matching the spectral features f to be amplified with unambiguous label information calculated in S2 i m (k) To distinguish, the spectral characteristic to be identified for the ith sample of the nth radiation source is here denoted by f i n (k) And (4) showing. f (k), f i m (k)、f i n (k) All are calculated based on the same method, and the subscript represents the radiation source label corresponding to the signal and the sample number information.
S3.1.2 mixing of f i n (k) And spectral reference characteristics
Performing cross-correlation operation to obtain new characteristics
And (3) realizing difference amplification:
it is noted that the features herein
Despite the formally similar standard spectral characteristics obtained in S2
Similar, but standard spectral characteristics
The spectrum characteristic analysis method is a common component of all known samples, does not refer to the characteristics of a certain radiation source, only needs to calculate once, only needs to be carried out once in spectrum characteristic component analysis, and therefore does not have subscripts. And the new features obtained here
Is the current signal sample f i n (k) Thus following the feature f i n (k) N denotes the label of the radiation source and i denotes the sample number.
S3.2, calculating the final spectrum characteristics after amplification
New characteristics in S3.1
Subtracting the intentional modulation characteristic quantity h (j) obtained by the step S2.4.3 to obtain an amplified final spectrum characteristic:
The invention relates to an amplification technology of a wireless network card frequency spectrum fingerprint characteristic. The method specifically comprises the steps of carrying out high-precision frequency estimation on a plurality of network card signals received in the wireless network signal detection process, extracting frequency spectrum fingerprint characteristics, solving the problem that the individual characteristic difference of radiation sources among network card devices is small, and realizing detail amplification. The method specifically comprises the steps of extracting a local frequency spectrum with richer fingerprint information as fingerprint features, obtaining a public part of the frequency spectrum features of the radiation sources by carrying out weighting and averaging on the frequency spectrum features of the multiple radiation sources, obtaining main components and stray components of the frequency spectrum features based on self-adaptive decomposition, realizing detail amplification in the frequency spectrum features based on relevant operation in a calculation process, eliminating intentional modulation amount, obtaining amplified final frequency spectrum features, and laying a foundation for subsequently improving the identification success rate of the network card equipment.
The invention has the following technical effects: the invention realizes the amplification of the subtle difference characteristics of the network card signals on the frequency spectrum, which is mainly embodied as follows:
(1) Extracting a local frequency spectrum with most abundant fingerprint information in a signal frequency spectrum as a radiation source fingerprint characteristic instead of relying on all information, particularly removing a main lobe and only reserving a frequency spectrum with a certain width near the main lobe; the calculation method is simple and convenient, and compared with the full length, the characteristic dimension is greatly reduced.
(2) The method comprises the steps of obtaining common components of fingerprint features of different radiation sources based on weighted average values of a large number of samples of a plurality of radiation sources, defining spectral reference features and standard spectral features, and providing a corresponding calculation method.
(3) And decomposing the newly extracted fingerprint characteristics of the radiation source by using a self-adaptive decomposition VMD algorithm, separating components which are not beneficial to SEI (solid electrolyte interphase), including main components and spurious components which are intentionally modulated, and eliminating the components.
(4) And amplifying the slight difference in the fingerprint characteristics of the radiation source by utilizing correlation operation.
(5) The calculation complexity is low, and the real-time performance is high.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention;
FIG. 2 is a signal spectral feature image;
FIG. 3 is a raw spectral signature;
FIG. 4 is a spectral reference feature autocorrelation;
FIG. 5 is a standard spectral feature decomposition result;
FIG. 6 is a cross-correlation result of spectral features of different radiation sources with reference features;
FIG. 7 is a final feature diagram of the radiation source after differential amplification of fingerprint features;
fig. 8 is a characteristic partial diagram after the fingerprint difference of the radiation source is enlarged.
Detailed Description
The following further describes embodiments of the present invention with reference to the accompanying drawings.
Fig. 1 is a flow chart of the implementation of the present invention, and the present invention provides a network card spectrum fingerprint feature amplification method based on feature correlation and adaptive decomposition, which is divided into the following three major steps:
s1: data processing and spectrum feature extraction;
s2: extracting frequency spectrum reference features and analyzing feature components;
s3: extracting the amplification characteristics of the label-free data;
the method comprises the following specific steps:
firstly, processing received data by S1.1-S1.3 to obtain a spectrum characteristic to be amplified;
then, whether the frequency spectrum reference feature calculation is needed or not is judged, if so, S2.1-S2.4 are carried out to obtain the reference feature
And an intentional modulation amount h (j). Since the calculations of S2.1-S2.4 are based on a plurality of sample cumulative weighting data for a plurality of radiation sources, allCommon components common to radiation sources.
Finally, for the new signal received subsequently, steps S1.1-S1.3 need to be repeated, and steps S3.1 and S3.2 are performed to obtain the final spectrum feature after amplification.
Therefore, in the practical application process, S2.1-S2.4 only need to be carried out once to obtain corresponding reference characteristics
And the steps S2.1-S2.4 do not need to be repeated after the intentional modulation amount h (j), thereby improving the calculation efficiency.
FIG. 2 shows the calculated (S1.1-S1.3) signal spectral reference characteristics
The horizontal axis represents a feature dimension, and the vertical axis represents a corresponding feature value. The features here are from the right region of the main peak of the spectrum, with dimensions 4000. In the feature calculation process, λ in S1.3.2 1 W B λ of 1 λ in =2,S1.3.3 2 =10, the following chart calculates the parameters as consistent. The result is a common part calculated based on 300 signal samples of 3 wireless network cards, where each radiation source contains 100 signal samples, i.e. M =3,n m ＝100,m＝1,2,3。
Fig. 3 shows the original spectrum characteristics obtained by calculating the signals of multiple radiation sources, i.e. f (k) obtained through steps S1.1-S1.3, different colors represent different network card individuals, and the emission signals of different individuals are the same. The left hand panel shows the original spectral characteristics of the two radiation sources, the darker set of curves representing the characteristics of the radiation source 1 and the lighter representing the characteristics of the radiation source 2; the right-hand graph shows the addition of the features of radiation source 3 to the original radiation sources 1 and 2, which are shown as the second, outermost set of dark curves in the figure.
It can be seen that the signals of different radiation sources are different in the original features extracted by the present invention, and can be used for device identification. The main body shape of these characteristic curves is consistent, however, differing only slightly. These subtle differences are of real concern for radiation source fingerprinting. The subtlety of the difference causes higher similarity of characteristic curves and serious overlapping of characteristic curves along with the increase of the number of the individual radiation sources, and the identification difficulty is greatly increased.
FIG. 4 is a reference feature
Standard frequency spectrum characteristics obtained by amplifying detail information after autocorrelation
Graph (S2.2). The horizontal axis represents the feature dimension and the vertical axis represents the feature value, as in fig. 2 and 3. The calculation is based on the spectral reference features in FIG. 2
Characteristic of this time
The two are symmetrical about 4000, and the characteristic dimension is 7999.
Is the common part where individual differences of the radiation sources are eliminated, and its body shape can be used to represent the intentional modulation component, which is the input to the component analysis of the VMD decomposition in S2.3.
FIG. 5 is a standard spectral signature
Graph of IMF in time domain after VMD decomposition of L =3 (S2.3). As can be seen, after decomposition, the different components IMF 1-IMF 3 are decomposed, wherein IMF1 has the strongest energy, and is
And IMF3 has the lowest energy and the highest frequency, is the spurious component.
FIG. 6 is a graph of the feature f calculated by the steps S1.1-S1.3 for a newly received sample i n (k) And reference features
Results of performing cross-correlation
The two groups of curves with different colors in the left small graph in fig. 6 represent the characteristics of the radiation source 1 (dark color) and the radiation source 2 (light color), respectively, and have the same meaning as the curves in fig. 3; the newly added set of dark curves in the small figure on the right side represents the radiation source 3 compared to the left side figure.
From the left and right side symmetry of the waveform, it can be seen that the frequency spectrum is compared with the standard frequency spectrum characteristics
(figure 4) of the drawing,
and is not left-right symmetric. It can be clearly seen that the difference is amplified compared to the original spectral signature in fig. 3. And after the target number is increased, the characteristic difference among different network cards is still obvious.
FIG. 7 is the resulting spectral signature resulting from the fingerprint differential amplification after the method of the present invention
Graph (S3.2). The characteristic obtained here is that the main body component and the stray component h (j) which are obtained after the VMD decomposition and are not beneficial to the individual identification of the radiation source are eliminated, and only the unintentional modulation information which can represent the difference between individuals is reserved, so that the fingerprint information is amplified.
The meanings of the curves in the left and right small pictures in fig. 7 are the same as those in fig. 3 and 6. The curve sub-tables of different colors represent the amplified final spectra of different radiation sources
The two sets of curves in the left diagram represent the radiation source 1 (dark color) and the radiation source 2 (light color), respectively, and the right diagram represents the features of the radiation source 3 (top dark color) additionally plotted on the left basisCurve) is shown.
Compared with the original spectrum characteristics in fig. 3, the amplification effect is very obvious.
FIG. 8 is a feature after differential magnification of fingerprints
A partially enlarged view of (a). The meaning of the curves in the figure is consistent with that of figure 7. The value ranges of the characteristic curves of different targets are different, the fluctuation rules are different, and the distinctiveness is greatly enhanced.
The invention has the effect of carrying out verification analysis based on the really acquired signals transmitted by the three wireless network cards with the same type and the IEEE802.11b standard. The working frequency point of the signal is 2.4GHz, and the sampling rate is 10GHz. The header part of the obtained signal frame is detected, and the length of the data analyzed by the header part is 2500 points. Fig. 2 to 8 are all actual effect diagrams. The human eyes can clearly see that the original characteristic curves distributed together show obvious difference after being processed by the method, and the original slight difference is obviously amplified.
Claims (5)
1. A network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition is characterized by comprising the following steps:
s1: data processing and spectral feature extraction
The method comprises the following specific steps:
s1.1, data reception and preprocessing
After receiving the signal, the receiver performs detection, filtering and noise reduction pretreatment on the original signal to obtain a signal sample x to be processed 0 (t)；
Signal sample x to be processed 0 (t) analyzing and detecting to obtain a frame header part x of the signal sample to be processed 1 (t)；
S1.2, estimating and eliminating signal frequency offset
The method comprises the following specific steps:
s1.2.1 Fourier-based interpolation algorithm for x 0 (t) frequency offset value f 0 Carrying out estimation;
s1.2.2 according to f 0 X is to be 1 (t) shifting to zero frequency to obtainBaseband signal, eliminate the frequency offset influence;
s1.2.3, performing mean value removal on baseband signals, and performing energy normalization to obtain a signal x to be subjected to feature extraction 2 (t)；
S1.3, spectral feature calculation
The method comprises the following specific steps:
s1.3.1 calculating the Signal x to be feature extracted at 0 frequency 2 Frequency spectrum f of (t) 0 (k)：
Wherein,
representing a fourier transform; k represents a frequency index value of Fourier transform and is also a dimension index value of the spectrum characteristic; n is a radical of hydrogen FFT Points representing a Fourier transform; | represents the calculated amplitude value;
s1.3.2 Signal x to be characteristic extracted 2 Frequency spectrum f of (t) 0 (k) Has a main lobe of 3dB width W B The bandwidth of the left and right of the extracted peak is lambda 1 W B As preliminary spectral feature value f 1 (k),k＝1,...,K 2 Wherein λ is 1 Weighting parameter, K, representing the spectral range 2 Indicating the length of the spectrum after truncation, i.e. preliminary spectral characteristic f 1 (k) Length of (d);
s1.3.3 pairs of f 1 (k) Calculating the second derivative
Select to satisfy
Minimum value k of min Wherein λ is 2 Denotes a weight coefficient, λ 2 ＞0；
Selection f 1 (k) Upper [ k ] min ,K 2 ]The segment is used as a spectral feature f (k) to be amplified;
s2: spectral reference feature extraction and feature component analysis
The method comprises the following specific steps:
s2.1, calculating frequency spectrum reference characteristics
Assuming that M wireless network cards are needed to carry out individual identification analysis of radiation sources, the mth radiation source has N m The signal sample, the spectral feature to be amplified of the ith sample can be represented as f i m (k),i＝1,...,N m ,m＝1，...,M；
Here, f i m (k) That is, the output f (k) of S1 is information representing a radiation source and a signal sample, an upper corner mark m represents a label of the corresponding radiation source, and a lower corner mark i represents an ith signal sample of the current radiation source;
s2.1.1 according to the same method as S1, calculating and obtaining spectral characteristics to be amplified of all samples of all radiation sources and forming a set { f } i m (k)},i＝1,...,N m ,m＝1，...,M；
S2.1.2 calculating the characteristic mean value of each radiation source
S2.1.3, obtaining a frequency spectrum reference characteristic by calculating a weighted average value
Wherein alpha is m Is the characteristic weight of the mth radiation source, satisfies
Spectral reference feature
Is obtained byCalculating a common part obtained by all training samples of all radiation sources, wherein the common part is considered as a common part of the radiation sources, namely intentional modulation which is not beneficial to individual identification of the radiation sources is carried out, and the common part is suppressed;
s2.2, spectral reference feature detail amplification
Spectral reference signature obtained for S2.1
Performing autocorrelation operation to realize detail amplification to obtain standard frequency spectrum characteristics
Wherein,
representing an autocorrelation operation; the characteristic dimension here varies, J =2K-1;
in this step
Is a weighted average value based on multiple radiation source signals-frequency spectrum reference characteristic
The obtained radiation source is subjected to autocorrelation operation, is bilaterally symmetrical in distribution, and can reflect the commonality of the radiation source; to match the spectral reference characteristics in S2.1
The distinguishing is carried out by the following steps,
referred to as standard spectral features;
s2.3, adaptive decomposition of Standard spectral features
Standard spectrum characteristics obtained from S2.2
VMD decomposition is carried out to obtain different components which are called IMF;
assuming that the number of decomposition layers is L, the obtained decomposition result can be expressed as:
wherein, V, V, omega are the time domain combination, frequency domain combination and IMF center frequency of the IMF decomposed respectively;
the time domain combination V of IMFs is represented as:
wherein v is l (j) L =1, 2.., L is a single IMF component;
s2.4, spectral feature analysis
Performing spectral component analysis on the L IMF components obtained by VMD decomposition in S2.3, specifically as follows:
s2.4.1 calculating different IMF components v l (j) Energy value of (c):
s2.4.2 according to { E l Sorting the IMF, wherein the IMF component with the highest energy is the principal component
Low energy as stray component
S2.4.3 selecting the principal component components
And stray component
Is represented as
Obtaining an intentional modulation feature h (j):
h(j)＝∑ vl∈l v l (j)
s2, calculating the obtained frequency spectrum reference characteristics
Standard spectral signature
Is cumulatively weighted based on a plurality of samples from a plurality of radiation sources and is therefore considered to be a common component common to all radiation sources;
the intentional modulation characteristic quantity h (j) is calculated on the basis of the common components, is considered to be common to signals of different radiation sources, is an irrelevant component which is not concerned by SEI, and needs to be removed; obtaining h (j) means that the analysis of the spectral characteristic components is completed;
s3: label-free data amplification feature extraction
The method comprises the following specific steps:
s3.1, correlation transformation of samples to be identified
S3.1.1 calculating the spectral feature f of the New received Signal according to S1 i n (k) (ii) a In the identification phase, the radiation source label is represented by n because the corresponding network card label is unknown; for comparison with the spectral feature f to be amplified with unambiguous label information calculated in S2 i m (k) To distinguish, here the spectral characteristics to be identified of the ith sample of the nth radiation sourceBy f i n (k) Represents; f (k), f i m (k)、f i n (k) The data acquisition method is calculated based on the same method, and the subscripts represent the radiation source labels corresponding to the signals and the sample number information;
s3.1.2 mixing of f i n (k) And spectral reference characteristics
Performing cross-correlation operation to obtain new characteristics
And (3) realizing difference amplification:
s3.2, calculating the final spectrum characteristics after amplification
New characteristics in S3.1
Subtracting the intentional modulation characteristic quantity h (j) obtained by the step S2.4.3 to obtain an amplified final spectrum characteristic:
2. The network card spectrum fingerprint feature amplification method based on feature correlation and adaptive decomposition according to claim 1, characterized in that: s1.2 in frequency offset estimation, based on x with longer length 0 (t) performing frequency offset estimation may improve estimation accuracy.
3. The network card spectrum fingerprint feature amplification method based on feature correlation and adaptive decomposition according to claim 1, characterized in that: s1.3.2, weighting parameter λ of spectral range 1 ＝2。
4. The network card spectrum fingerprint feature amplification method based on feature correlation and adaptive decomposition according to claim 1, characterized in that: s2.1.3, characteristic weight alpha of mth radiation source m According to the data scale and the characteristic distribution of each radiation source.
5. The network card spectrum fingerprint feature amplification method based on feature correlation and adaptive decomposition according to claim 4, characterized in that: characteristic weight of mth radiation source
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210854425.3A CN115186714B (en) | 2022-07-15 | 2022-07-15 | Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210854425.3A CN115186714B (en) | 2022-07-15 | 2022-07-15 | Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115186714A true CN115186714A (en) | 2022-10-14 |
| CN115186714B CN115186714B (en) | 2023-07-21 |
Family
ID=83520227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210854425.3A Active CN115186714B (en) | 2022-07-15 | 2022-07-15 | Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115186714B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115510924A (en) * | 2022-11-17 | 2022-12-23 | 中铁第一勘察设计院集团有限公司 | Radio frequency fingerprint identification method based on improved variational modal decomposition |
| CN117765575A (en) * | 2024-01-12 | 2024-03-26 | 中国人民解放军国防科技大学 | Multi-receiver cooperation-based cross-receiving system radiation source fingerprint identification method |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8160163B1 (en) * | 2007-08-06 | 2012-04-17 | University Of South Florida | Method for OFDM signal identification and parameter estimation |
| CN106792706A (en) * | 2016-12-09 | 2017-05-31 | 中国人民解放军国防科学技术大学 | A kind of illegal invasion detection method based on wireless network signal spectrum non-stationary property |
| CN109150340A (en) * | 2018-07-28 | 2019-01-04 | 中国人民解放军战略支援部队信息工程大学 | Method for extracting fingerprint feature and identity detection method and Radar recognition bearing calibration |
| CN109307862A (en) * | 2018-07-05 | 2019-02-05 | 西安电子科技大学 | A kind of target radiation source individual discrimination method |
| CN110991378A (en) * | 2019-12-11 | 2020-04-10 | 中国人民解放军空军工程大学 | Power amplifier individual identification method and device |
| CN111814703A (en) * | 2020-07-13 | 2020-10-23 | 四川大学 | HB-based signal joint feature extraction method under non-reconstruction condition |
| CN113219415A (en) * | 2021-04-30 | 2021-08-06 | 北京理工大学 | Interference source individual identification method based on envelope fingerprint characteristics |
| CN113219431A (en) * | 2021-04-30 | 2021-08-06 | 北京理工大学 | Interference source individual identification method based on waveform fingerprints |
| CN114255487A (en) * | 2021-12-22 | 2022-03-29 | 安徽农业大学 | Internet of things equipment identity authentication method for semi-supervised radio frequency fingerprint extraction |
-
2022
- 2022-07-15 CN CN202210854425.3A patent/CN115186714B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8160163B1 (en) * | 2007-08-06 | 2012-04-17 | University Of South Florida | Method for OFDM signal identification and parameter estimation |
| CN106792706A (en) * | 2016-12-09 | 2017-05-31 | 中国人民解放军国防科学技术大学 | A kind of illegal invasion detection method based on wireless network signal spectrum non-stationary property |
| CN109307862A (en) * | 2018-07-05 | 2019-02-05 | 西安电子科技大学 | A kind of target radiation source individual discrimination method |
| CN109150340A (en) * | 2018-07-28 | 2019-01-04 | 中国人民解放军战略支援部队信息工程大学 | Method for extracting fingerprint feature and identity detection method and Radar recognition bearing calibration |
| CN110991378A (en) * | 2019-12-11 | 2020-04-10 | 中国人民解放军空军工程大学 | Power amplifier individual identification method and device |
| CN111814703A (en) * | 2020-07-13 | 2020-10-23 | 四川大学 | HB-based signal joint feature extraction method under non-reconstruction condition |
| CN113219415A (en) * | 2021-04-30 | 2021-08-06 | 北京理工大学 | Interference source individual identification method based on envelope fingerprint characteristics |
| CN113219431A (en) * | 2021-04-30 | 2021-08-06 | 北京理工大学 | Interference source individual identification method based on waveform fingerprints |
| CN114255487A (en) * | 2021-12-22 | 2022-03-29 | 安徽农业大学 | Internet of things equipment identity authentication method for semi-supervised radio frequency fingerprint extraction |
Non-Patent Citations (1)
| Title |
|---|
| 王翔: "Unintentional modulation microstructure enlargement", 《JOURNAL OF SYSTEMS ENGINEERING AND ELECTRONICS》, pages 522 - 533 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115510924A (en) * | 2022-11-17 | 2022-12-23 | 中铁第一勘察设计院集团有限公司 | Radio frequency fingerprint identification method based on improved variational modal decomposition |
| CN117765575A (en) * | 2024-01-12 | 2024-03-26 | 中国人民解放军国防科技大学 | Multi-receiver cooperation-based cross-receiving system radiation source fingerprint identification method |
| CN117765575B (en) * | 2024-01-12 | 2024-06-21 | 中国人民解放军国防科技大学 | Multi-receiver cooperation-based cross-receiving system radiation source fingerprint identification method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115186714B (en) | 2023-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tang et al. | FingerNet: An unified deep network for fingerprint minutiae extraction | |
| CN115186714B (en) | Network card frequency spectrum fingerprint feature amplification method based on feature correlation and self-adaptive decomposition | |
| CN114564982B (en) | Automatic identification method for radar signal modulation type | |
| CN111050315B (en) | Wireless transmitter identification method based on multi-core two-way network | |
| CN114118131A (en) | Attention mechanism-based multi-feature fusion wireless equipment radio frequency fingerprint extraction method | |
| CN112749633B (en) | Separate and reconstructed individual radiation source identification method | |
| CN114218984A (en) | Radio frequency fingerprint identification method based on sample multi-view learning | |
| CN116166936A (en) | Radar signal modulation type identification method based on wavelet transformation phase mutation detection | |
| Su et al. | Specific Emitter Identification Based on CNN via Variational Mode Decomposition and Bimodal Feature Fusion | |
| CN112883895B (en) | Illegal electromagnetic signal detection method based on self-adaptive weighted PCA and realization system thereof | |
| Liu et al. | An effective radio frequency signal classification method based on multi-task learning mechanism | |
| CN114580476B (en) | Unmanned aerial vehicle signal identification model construction method and corresponding identification method and system | |
| KR20190054744A (en) | Gas detection method using SVM classifier | |
| CN102880861A (en) | High-spectrum image classification method based on linear prediction cepstrum coefficient | |
| CN115169406B (en) | Instantaneous phase fingerprint feature enhancement method based on empirical mode decomposition | |
| Yan et al. | Non-intrusive fingerprints extraction from hyperspectral imagery | |
| CN109271902B (en) | Infrared weak and small target detection method based on time domain empirical mode decomposition under complex background | |
| CN115085831B (en) | Unmanned aerial vehicle remote control signal recognition system and method based on mixed time-frequency analysis | |
| CN113408420B (en) | Spread spectrum signal identification method based on multi-dimensional parameter extraction and support vector machine | |
| CN114580470B (en) | OFDM and UFDM multi-carrier signal identification method based on non-uniform quantization map features | |
| CN115809426A (en) | Radiation source individual identification method and system | |
| CN111404852B (en) | Modulation mode identification method based on amplitude and spectral amplitude characteristics | |
| CN113343802B (en) | Multi-wavelet-based radio frequency fingerprint image domain identification method | |
| CN116863327A (en) | Cross-domain small sample classification method based on cooperative antagonism of double-domain classifier | |
| 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 |