CN111601307A - Transient-steady state based partial superposition radio frequency fingerprint method - Google Patents
Transient-steady state based partial superposition radio frequency fingerprint method Download PDFInfo
- Publication number
- CN111601307A CN111601307A CN202010417579.7A CN202010417579A CN111601307A CN 111601307 A CN111601307 A CN 111601307A CN 202010417579 A CN202010417579 A CN 202010417579A CN 111601307 A CN111601307 A CN 111601307A
- Authority
- CN
- China
- Prior art keywords
- signal
- steady
- transient
- radio frequency
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/60—Context-dependent security
- H04W12/69—Identity-dependent
- H04W12/79—Radio fingerprint
Landscapes
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a transient-stable state-based partial superposition radio frequency fingerprint method, which is used for carrying out energy normalization, frequency offset and phase offset estimation and compensation pretreatment on an acquired signal; then dividing the preprocessed signal into a transient signal part and a steady-state signal part, and performing same symbol superposition on the steady-state signal to improve the signal-to-noise ratio; then, the signal is spliced with the transient signal, and the spliced signal is subjected to radio frequency fingerprint extraction and identification, so that the identification performance of the equipment in a low signal-to-noise ratio scene is improved. The invention effectively combines the identification advantages of the transient signal and the steady-state signal radio frequency fingerprint under different signal-to-noise ratios, thereby greatly improving the identification performance of the equipment under the condition of low signal-to-noise ratio and slightly improving the performance under the condition of high signal-to-noise ratio.
Description
Technical Field
The invention relates to the field of information security, in particular to a transient-steady state-based partial superposition radio frequency fingerprint method.
Background
With the rise of intelligent hardware technology, the development of the internet of things (IoT) presents an exponential growth situation, and the internet of everything interconnection has become a necessary trend for technology development and industrial application. Meanwhile, with the formal commercial use of the fifth generation mobile communication technology (5G), ultrahigh-speed data transmission, ultra-wide wireless coverage and ultra-low data delay promote the further popularization and expansion of the Internet of things in the fields of smart cities, smart homes, intelligent traffic, industrial internet, smart power grids, smart medical treatment and the like, and the large-scale deployment and application of the cellular Internet of things terminal are comprehensively promoted. Along with the rapid increase of the number of connections and the rapid increase of the access rate, massive sensitive confidential data can be transmitted on a wireless channel in the future. Once the information is leaked due to the malicious intrusion of an illegal attacker, direct property loss or personal injury can be brought to the user. Therefore, future wireless networks should have strong security and privacy.
In recent years, radio frequency fingerprint-based device identification and authentication techniques have gained widespread attention. The radio frequency fingerprint is an essential characteristic of the wireless equipment caused by hardware tolerance and process technology in the production and manufacturing process, has the characteristics of uniqueness and difficult cloning, can effectively resist attacks such as equipment counterfeiting, tampering and the like, and can be used for identity identification and authentication of equipment of the Internet of things.
The existing radio frequency fingerprint technology can be generally divided into a transient radio frequency fingerprint technology and a steady radio frequency fingerprint technology according to a target signal interval of extracted features of the existing radio frequency fingerprint technology. However, the existing methods have poor performance under a low signal-to-noise ratio (less than 10dB), and the document "On Radio Frequency Identification for DSSS Systems in LowSNR scenes" proposes that the signal-to-noise ratio can be improved by superimposing the same training sequence in a multi-frame scene, but how to improve the recognition rate in a single-frame scene is still to be researched.
Disclosure of Invention
Aiming at the problems in the prior art, the invention aims to provide a transient-steady state-based partial superposition radio frequency fingerprint method, which realizes the identification and authentication of wireless equipment in practical application. The method comprises the steps of firstly preprocessing acquired signals such as energy normalization, frequency deviation, phase deviation estimation and compensation, dividing the preprocessed signals into a transient signal part and a steady signal part, superposing the steady signals by the same symbol to improve the signal-to-noise ratio, splicing the steady signals with the transient signals, extracting and identifying the radio frequency fingerprint of the spliced signals, and improving the identification performance of the equipment in a low signal-to-noise ratio scene.
The purpose of the invention is realized by the following technical scheme:
a partial superposition radio frequency fingerprint method based on transient state-steady state is characterized in that: carrying out energy normalization, frequency offset, phase offset estimation and compensation pretreatment on the acquired signals; then dividing the preprocessed signal into a transient signal part and a steady-state signal part, and performing same symbol superposition on the steady-state signal to improve the signal-to-noise ratio; then, the signal is spliced with the transient signal, and the spliced signal is subjected to radio frequency fingerprint extraction and identification, so that the identification performance of the equipment in a low signal-to-noise ratio scene is improved; the method comprises the following steps:
(1) acquiring an output signal of wireless equipment in a high signal-to-noise ratio experimental environment;
(2) adding different Gaussian white noises into the acquired signals for simulation, acquiring wireless signals under different signal-to-noise ratios, and preprocessing the wireless signals;
(3) dividing the preprocessed signal into a transient signal section and a steady-state signal section, superposing the same symbols in the steady-state signal section, and splicing the superposed signals with the transient signal;
(4) randomly dividing signals spliced under all signal-to-noise ratios into a training set, a verification set and a test set according to a certain proportion;
(5) training a classifier according to the extracted radio frequency fingerprint features on a training set and a verification set;
(6) and (3) performing preprocessing in the step (2) and partial overlapping and splicing in the step (3) aiming at the actually received output signals of the wireless equipment, then extracting the radio frequency fingerprint characteristics, and performing classification and identification on the wireless equipment by using a trained classifier.
Further, the output signal of the low power consumption wireless device in the step (1) is acquired by connecting a direct coaxial line and an attenuator, or the signal is wirelessly received with high signal to noise ratio acquired in a short-distance and line-of-sight environment.
Further, the wireless signals collected in the step (1) are directly down-converted to baseband for subsequent processing.
Further, the range of the simulated signal-to-noise ratio in the step (2) covers the range of the actual signal-to-noise ratio of the identification scene, and the noise superposition is performed before the preprocessing.
Further, the pretreatment in the step (2) specifically comprises: signal detection and interception, energy normalization, and estimation and compensation of signal frequency offset and phase offset.
Further, the step (3) specifically comprises:
(3-1) dividing the preprocessed signal into transient signal segments StransientAnd a steady-state signal segment SsteadyThe partitioning of the transient and steady state signals includes, but is not limited to, empirical methods, differential threshold detection methods, and differential bayesian detection methods. And during division, if the starting point of the steady-state signal is not the starting sampling point of one symbol, the steady-state signal is pushed backwards to the starting point of the next symbol.
(3-2) performing same symbol superposition on the steady-state signal to improve the signal-to-noise ratio, namely starting from the single symbol, the steady-state signal can be expressed as Ssteady＝[s1,s2,…,sN]Wherein, N is the number of symbols contained in the steady-state signal, and the signal corresponding to each symbol is represented as sk＝RFFk(xk),xk∈ {1,2, …, K }, where xkIs the data value of the kth symbol of the steady-state signal, K represents the number of values of all symbols, RFFkAnd (-) represents the radio frequency fingerprint corresponding to the kth symbol. The signals corresponding to the same symbol in the steady-state signal may be superimposed, and the signal corresponding to the symbol i after the superimposition is ss (i) ═ mean (RFF)k(xkI)), i ∈ {1,2, …, K }, where mean () denotes the superposition averaging operation.
(3-3) splicing the superposed steady-state signal and transient-state signal to obtain a partial superposed signal Sps＝[Stransient,ss(1),ss(2),…,ss(K)]。
Further, the signal set S in step (4)Including all the pre-processed and partially superposed baseband signals under signal-to-noise ratio, and dividing the sample set S into a training set S according to a certain proportiontrainVerification set SvalidAnd test set StestMethods of partitioning include, but are not limited to, cross-validation, set-out, and self-service.
Further, the radio frequency fingerprint feature extraction and identification method in the step (5) includes but is not limited to one or more of manually selected features and corresponding machine learning classification algorithms, automatic feature extraction and classification methods based on deep learning, and the like.
Further, after preprocessing the actually received wireless device output signal in step (6) and performing the preprocessing required in step (2), performing the partial overlapping and splicing processing required in step (3), and then performing the classification and identification of the wireless device by using the radio frequency fingerprint extraction and identification method trained in step (5).
The invention has the beneficial effects that: compared with the prior art, the invention has the following remarkable advantages: by the method, the radio frequency fingerprint information contained in the transient signal and the steady-state signal can be well balanced, and the stability of the radio frequency fingerprint information of the steady-state signal is effectively improved by superposition. Through simulation and experiments, the method can greatly improve the identification performance of the wireless equipment in a low signal-to-noise ratio scene compared with the traditional transient method and steady-state method.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a diagram showing the result of improving the equipment identification rate at a low noise ratio by using the method of the present invention.
Detailed Description
The embodiment provides a transient-steady state based partial superposition radio frequency fingerprint method, as shown in fig. 1, including:
(1) the output signal of the wireless device under the high signal-to-noise ratio experimental environment is collected.
In this embodiment, 27 ZigBee wireless transmitting modules with power amplifiers are selected as the devices to be identified, and each device is numbered1-27. The wireless signal transmitted by the ZigBee wireless transmitting module is received in the wireless environment with short distance, visible distance and high received signal-to-noise ratio, and the wireless signal can also be collected through the connection of a direct coaxial line and an attenuator. The collected wireless signals are directly converted into baseband through down conversion and then are subjected to subsequent processing. The symbol rate of the ZigBee equipment that this embodiment adopted is 1Mbps, and the direct down-conversion of the wireless signal of gathering obtains the baseband signal:
where x (t) represents the complex baseband representation of the transmitted signal, af is the frequency offset remaining after down-conversion,
representing phase deviation. The baseband signal is oversampled and stored 10 times with a 10Msps sampling rate.
(2) Adding different white Gaussian noises into the collected signals for simulation, acquiring wireless signals under different signal-to-noise ratios, and preprocessing the wireless signals, wherein the preprocessing specifically comprises the following steps:
(2-1) in the embodiment, aiming at baseband signals y (t) of all ZigBee wireless modules collected in a short distance, Gaussian white noises with different powers are added to simulate different signal-to-noise ratio levels. The specific signal-to-noise ratio of the Gaussian white noise simulation is from-10 dB to 30dB, a group is formed every 5dB, and all collected signals are simulated once under each signal-to-noise ratio.
(2-2) after the noise superposition, the following preprocessing is performed: firstly, the starting and the ending of each frame signal are roughly judged by utilizing the power change point detection principle, then the energy normalization is carried out on the intercepted frame signals,
where A is the root mean square of the added noise y (t). The energy normalization can avoid the influence of signal transmitting power and distance on identification, and the method does not consider the identification of the distance of the target equipment through the received signal power as fingerprint characteristics.
(2-3) the signal y' (t) after energy normalization still has residual frequency offset and phase offset, and because the frequency offset and the phase offset are two parameters which are easy to counterfeit, and subsequent superposition processing cannot be performed to improve the signal-to-noise ratio under the condition of the existence of the frequency offset and the phase offset. In order to extract the fingerprint characteristics of the device independent of frequency offset and phase offset, precise frequency offset and phase offset estimation and compensation are firstly carried out before fingerprint extraction (refer to patent 201510797097.8 for a specific method).
(3) Dividing the preprocessed signal into a transient signal section and a steady-state signal section, superposing the same symbols in the steady-state signal section, and splicing the superposed signals with the transient signal, wherein the method specifically comprises the following steps:
(3-1) in this embodiment, the preprocessed signal is divided into transient signal segments StransientAnd a steady-state signal segment SsteadyThe partitioning of the transient and steady state signals includes, but is not limited to, empirical methods, differential threshold detection methods, and differential bayesian detection methods. And during division, if the starting point of the steady-state signal is not the starting sampling point of one symbol, the steady-state signal is pushed backwards to the starting point of the next symbol.
(3-2) in this embodiment, only the steady-state portion in the preamble signal is considered in the steady-state signal segment of the preprocessed ZigBee signal, and a difference threshold detection method is used to detect that the steady-state signal starting points of 27 devices are substantially in the second symbol of the preamble, so that the 3 rd to 8 th preamble symbols are divided into steady-state signals, and the 1 st to 2 nd symbols and 20 sampling points before the preamble are used as transient signals.
(3-3) in this embodiment, steady-state signals of 6 symbols are superposed by the same symbol, so as to improve the signal-to-noise ratio. Since the leading symbol data of the ZigBee signal are all 0x0, the steady state signal can be expressed as Ssteady＝[s3,s4,…,s8]The signal corresponding to each symbol is denoted as sk＝RFFk(x0) Wherein x is0Representing the same data symbol 0x0, RFF in the preamblekAnd (-) represents the radio frequency fingerprint corresponding to the kth symbol. The radio frequency fingerprints corresponding to the same symbol in the steady-state signal are the same, so that superposition can be carried out, and the signal obtained after superposition is
RFFs(-) represents the preamble steady-state signal single symbol radio frequency fingerprint.
(3-4) splicing the superposed steady-state signal and transient-state signal to obtain a partial superposed signal Sps＝[Stransient,ss]。
(4) Randomly dividing a signal set S spliced under all signal-to-noise ratios into S according to a certain proportiontrainVerification set SvalidAnd test set Stest。
In this embodiment, the signal set S includes signals spliced under all signal-to-noise ratios, and the sample set S is divided into the training set S according to a ratio of 3:1:1trainVerification set SvalidAnd test set StestMethods of partitioning include, but are not limited to, cross-validation issue, set-out method, and self-service method.
(5) And training a classifier according to the extracted radio frequency fingerprint features on the training set and the verification set.
In this embodiment, a Convolutional Neural Network (CNN) including two convolutional layers, two pooling layers, and two fully-connected layers is selected as the radio frequency fingerprint extraction and classifier, where the last fully-connected layer is a Softmax layer as the classifier, and the training is performed on all the snr signals in the training set during training, and the training is stopped when the recognition performance in the verification set does not increase any more. In this embodiment, the number of layers of convolution and pooling, and the convolution kernel parameter and pooling parameter are determined according to actual needs, but are not limited thereto. In addition, the radio frequency fingerprint feature extraction and identification method includes, but is not limited to, one or more of manually selected features and corresponding machine learning classification algorithms, automatic feature extraction and classification methods based on deep learning, and the like.
(6) And (3) aiming at the actually received wireless equipment output signal, preprocessing in the step (2) and partial overlapping and splicing in the step (3) are carried out, and then the trained convolutional neural network is used for carrying out radio frequency fingerprint extraction and classification identification on the wireless equipment.
By the method, the characteristics of large difference of transient signal radio frequency fingerprints and superposition of steady state signals to improve the signal to noise ratio can be utilized, and the identification advantages of the transient signal radio frequency fingerprints and the steady state signal radio frequency fingerprints under different signal to noise ratios are effectively combined, so that the equipment identification performance is greatly improved at low signal to noise ratio, and the performance can be slightly improved at high signal to noise ratio.
FIG. 2 shows a test set S using transient signals, steady-state signals, and non-superimposed whole-segment signals as input pairs of convolutional neural networks, with and without the present methodtestAnd (5) carrying out classified test results. It can be seen that when the method is used, under the condition of extremely low signal-to-noise ratio (-10 to 0dB), the identification accuracy is improved by about 9% compared with the transient signal scheme with the best performance in other three schemes, and is improved by about 5% when 5dB and 10dB are adopted. Under the high signal-to-noise ratio environment of 15dB or above, compared with the transient + steady-state signal method with the best performance in other three schemes, the performance can be improved by less than 1%.
The invention provides a method for extracting radio frequency fingerprints after steady state signal same symbol superposition and transient signal splicing based on the difference between transient and steady state signals, and greatly improves the radio frequency fingerprint identification rate under low signal-to-noise ratio.
The above description is only one preferred embodiment of the present invention, and should not be taken as limiting the scope of the present invention, so that the present invention is defined by the appended claims.
Claims (9)
1. A partial superposition radio frequency fingerprint method based on transient state-steady state is characterized in that: carrying out energy normalization, frequency offset, phase offset estimation and compensation pretreatment on the acquired signals; then dividing the preprocessed signal into a transient signal part and a steady-state signal part, and performing same symbol superposition on the steady-state signal to improve the signal-to-noise ratio; then, the signal is spliced with the transient signal, and the spliced signal is subjected to radio frequency fingerprint extraction and identification, so that the identification performance of the equipment in a low signal-to-noise ratio scene is improved; the method comprises the following steps:
(1) acquiring an output signal of wireless equipment in a high signal-to-noise ratio experimental environment;
(2) adding different Gaussian white noises into the acquired signals for simulation, acquiring wireless signals under different signal-to-noise ratios, and preprocessing the wireless signals;
(3) dividing the preprocessed signal into a transient signal section and a steady-state signal section, superposing the same symbols in the steady-state signal section, and splicing the superposed signals with the transient signal;
(4) randomly dividing signals spliced under all signal-to-noise ratios into a training set, a verification set and a test set according to a certain proportion;
(5) training a classifier according to the extracted radio frequency fingerprint features on a training set and a verification set;
(6) and (3) performing preprocessing in the step (2) and partial overlapping and splicing in the step (3) aiming at the actually received output signals of the wireless equipment, then extracting the radio frequency fingerprint characteristics, and performing classification and identification on the wireless equipment by using a trained classifier.
2. The transient-steady state based partially-superimposed radio frequency fingerprint method according to claim 1, wherein the output signal of the low power consumption wireless device in step (1) is obtained through a direct coaxial line plus attenuator connection, or is a high signal-to-noise ratio wireless received signal collected in a short-distance line-of-sight environment.
3. The transient-stationary based partially-superimposed radio frequency fingerprint method according to claim 1, wherein the wireless signals collected in step (1) are directly down-converted to baseband for subsequent processing.
4. The transient-stationary based partial overlap radio frequency fingerprinting method of claim 1 characterized in that the simulated signal-to-noise ratio range in step (2) covers the actual signal-to-noise ratio range of the identified scene and the noise overlap is done before the preprocessing.
5. The transient-stationary based partially-superimposed radio frequency fingerprint method according to claim 1, wherein the preprocessing in step (2) specifically comprises: the method comprises the following steps of signal detection and interception, energy normalization, and estimation and compensation of frequency offset and phase offset, and specifically comprises the following steps:
(2-1) aiming at baseband signals y (t) of all ZigBee wireless modules collected in a short distance, Gaussian white noises with different powers are added to simulate different signal-to-noise ratio levels;
(2-2) after the noise superposition, the following preprocessing is performed: firstly, the starting and the ending of each frame signal are roughly judged by utilizing the power change point detection principle, then the energy normalization is carried out on the intercepted frame signals,
wherein A is the root mean square of the noise y (t); the energy normalization can avoid the influence of signal transmission power and distance on identification;
(2-3) the signal y (t) after energy normalization also has residual frequency offset and phase offset, and because the frequency offset and the phase offset are two parameters which are easy to counterfeit, and subsequent superposition processing cannot be performed to improve the signal-to-noise ratio under the condition of the frequency offset and the phase offset; in order to extract the fingerprint characteristics irrelevant to the frequency offset and the phase offset of the equipment, accurate frequency offset and phase offset estimation and compensation are firstly carried out before fingerprint extraction.
6. The transient-steady state based partially-superimposed radio frequency fingerprint method according to claim 1, wherein the step (3) specifically comprises:
(3-1) dividing the preprocessed signal into transient signal segments StransientAnd a steady-state signal segment SsteadyThe transient signal and the steady-state signal are divided by an empirical method, a differential threshold detection method and a differential Bayesian detection method; during dividing, if the starting point of the steady-state signal is not the starting sampling point of one symbol, the steady-state signal is pushed backwards to the starting point position of the next symbol;
(3-2) performing same symbol superposition on the steady-state signal to improve the signal-to-noise ratio, namely starting from the single symbol, the steady-state signal can be expressed as Ssteady＝[s1,s2,…,sN]Wherein, N is the number of symbols contained in the steady-state signal, and the signal corresponding to each symbol is represented as sk＝RFFk(xk),xk∈ {1,2, …, K }, where xkIs the data value of the kth symbol of the steady-state signal, K represents the number of values of all symbols, RFFk() represents the radio frequency fingerprint corresponding to the kth symbol; in the steady-state signal, the signals corresponding to the same symbol are superposed, and the signal corresponding to the symbol i after superposition is ss (i) ═ mean (RFF)k(xkI)), i ∈ {1,2, …, K }, where mean () denotes a superposition averaging operation;
(3-3) splicing the superposed steady-state signal and transient-state signal to obtain a partial superposed signal Sps＝[Stransient,ss(1),ss(2),…,ss(K)]。
7. The transient-stationary state-based partial overlap radio frequency fingerprint method according to claim 1, wherein the signal set S in step (4) comprises baseband signals after preprocessing and partial overlap processing under all signal-to-noise ratios, and the sample set S is divided into a training set S according to a certain proportiontrainVerification set SvalidAnd test set StestThe dividing method comprises a cross-validation method, a leave method and a self-service method.
8. The transient-stationary based partial overlap radio frequency fingerprint method according to claim 1, wherein the radio frequency fingerprint feature extraction and identification method in step (5) is one or more of manually selected features and corresponding machine learning classification algorithm, and automatic feature extraction and classification method based on deep learning.
9. The transient-steady state based partial superposition radio frequency fingerprint method according to claim 1, wherein in step (6), after preprocessing the actually received wireless device output signal, partial superposition and splicing are performed, and then classification and identification of the wireless device are performed by using a trained radio frequency fingerprint extraction and identification method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010417579.7A CN111601307A (en) | 2020-05-18 | 2020-05-18 | Transient-steady state based partial superposition radio frequency fingerprint method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010417579.7A CN111601307A (en) | 2020-05-18 | 2020-05-18 | Transient-steady state based partial superposition radio frequency fingerprint method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111601307A true CN111601307A (en) | 2020-08-28 |
Family
ID=72191336
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010417579.7A Pending CN111601307A (en) | 2020-05-18 | 2020-05-18 | Transient-steady state based partial superposition radio frequency fingerprint method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111601307A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112312457A (en) * | 2020-10-29 | 2021-02-02 | 上海电机学院 | Communication radiation source individual identification method based on complex depth residual error network |
| CN113949986A (en) * | 2021-08-11 | 2022-01-18 | 厦门大学 | Positioning method, system and monitoring method based on WR and radio frequency fingerprint technology |
| CN114783006A (en) * | 2022-04-28 | 2022-07-22 | 东南大学 | Wireless device transient and steady state device fingerprint extraction and identification method based on signal difference |
| CN115017935A (en) * | 2022-04-22 | 2022-09-06 | 大连海事大学 | Radio frequency signal emission time detection method based on sliding window variance track |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108809355A (en) * | 2018-07-04 | 2018-11-13 | 南京东科优信网络安全技术研究院有限公司 | A method of in low signal-to-noise ratio extraction equipment physical fingerprint feature |
| CN109921886A (en) * | 2019-01-28 | 2019-06-21 | 东南大学 | A kind of low power consuming devices radio-frequency fingerprint recognition methods of robust |
-
2020
- 2020-05-18 CN CN202010417579.7A patent/CN111601307A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108809355A (en) * | 2018-07-04 | 2018-11-13 | 南京东科优信网络安全技术研究院有限公司 | A method of in low signal-to-noise ratio extraction equipment physical fingerprint feature |
| CN109921886A (en) * | 2019-01-28 | 2019-06-21 | 东南大学 | A kind of low power consuming devices radio-frequency fingerprint recognition methods of robust |
Non-Patent Citations (1)
| Title |
|---|
| JIABAO YU 等: "Radio Frequency Fingerprint Identification Based on Denoising Autoencoders", 《2019 INTERNATIONAL CONFERENCE ON WIRELESS AND MOBILE COMPUTING,NETWORKING AND COMMUNICATIONS》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112312457A (en) * | 2020-10-29 | 2021-02-02 | 上海电机学院 | Communication radiation source individual identification method based on complex depth residual error network |
| CN112312457B (en) * | 2020-10-29 | 2023-07-21 | 上海电机学院 | Communication radiation source individual identification method based on complex depth residual error network |
| CN113949986A (en) * | 2021-08-11 | 2022-01-18 | 厦门大学 | Positioning method, system and monitoring method based on WR and radio frequency fingerprint technology |
| CN113949986B (en) * | 2021-08-11 | 2024-03-29 | 厦门大学 | Positioning method, positioning system and monitoring method based on WR (Wireless Fidelity) and RF (radio frequency) fingerprint technology |
| CN115017935A (en) * | 2022-04-22 | 2022-09-06 | 大连海事大学 | Radio frequency signal emission time detection method based on sliding window variance track |
| CN115017935B (en) * | 2022-04-22 | 2024-04-26 | 大连海事大学 | Method for detecting radio frequency signal transmitting moment based on sliding window variance track |
| CN114783006A (en) * | 2022-04-28 | 2022-07-22 | 东南大学 | Wireless device transient and steady state device fingerprint extraction and identification method based on signal difference |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111601307A (en) | Transient-steady state based partial superposition radio frequency fingerprint method | |
| CN109921886B (en) | Robust low-power-consumption equipment radio frequency fingerprint identification method | |
| CN107947830B (en) | A kind of radio-frequency fingerprint recognition methods for resisting multi-path jamming | |
| Sa et al. | Specific emitter identification techniques for the internet of things | |
| CN106161297A (en) | In ofdm system, anti-pilot tone spoofing attack channel based on independent component analysis is estimated and recognition methods | |
| CN111027614B (en) | Noise enhanced radio frequency fingerprint identification method and device | |
| Chen et al. | LoRa radio frequency fingerprint identification based on frequency offset characteristics and optimized LoRaWAN access technology | |
| CN114417914A (en) | Radio frequency fingerprint extraction and identification method based on multi-channel convolutional neural network | |
| CN111565383B (en) | Method for eliminating channel characteristics and extracting radio frequency fingerprint of ZigBee device | |
| CN113343874A (en) | Large-scale radio signal identification method based on deep convolutional neural network | |
| CN113343868A (en) | Radiation source individual identification method and device, terminal and storage medium | |
| Yuan et al. | Wireless device identification based on improved convolutional neural network model | |
| Wang et al. | A radio frequency fingerprinting identification method based on energy entropy and color moments of the bispectrum | |
| CN113905383B (en) | IFF signal identification method, device and medium based on radio frequency fingerprint | |
| CN112566129B (en) | Radio frequency fingerprint extraction and identification method capable of resisting multipath interference | |
| CN112437030B (en) | LoRa data packet starting point alignment method and device | |
| Wang et al. | Method for identifying pseudo GPS signal based on radio frequency fingerprint | |
| Li et al. | The identification of power IoT devices on differential constellation trajectory map | |
| CN112996001A (en) | Physical layer secure communication method based on radio frequency fingerprint image scrambling | |
| Cheng et al. | Radio frequency emitter identification based on ensemble learning | |
| CN113286306A (en) | Wireless channel fingerprint filtering method and system based on channel equalization | |
| Wen et al. | RF transmitter identification and classification based on deep residual shrinkage network | |
| CN117956475A (en) | LoRa network security identity authentication method based on radio frequency fingerprint technology and CNN-BILSTM | |
| Guo et al. | Deep Learning Based RF Fingerprint Identification using Cyclic Shift Characteristic | |
| Wang et al. | Radiofrequency fingerprint feature extraction and recognition using a coordinate attention-guided deep residual shrinkage 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200828 |