CN106911603A - A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method - Google Patents
A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method Download PDFInfo
- Publication number
- CN106911603A CN106911603A CN201710131554.9A CN201710131554A CN106911603A CN 106911603 A CN106911603 A CN 106911603A CN 201710131554 A CN201710131554 A CN 201710131554A CN 106911603 A CN106911603 A CN 106911603A
- Authority
- CN
- China
- Prior art keywords
- signal
- data
- frequency
- spectrum
- real
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/0012—Modulated-carrier systems arrangements for identifying the type of modulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The present invention discloses a kind of broadband monitoring pattern Imitating modulated signal real-time identification method, comprises the following steps：Receive and radiofrequency signal and change output intermediate-freuqncy signal in the range of wide-band；A/D samplings are carried out to intermediate-freuqncy signal, Digital Down Convert DDC treatment exports I/Q data；I/Q data to being input into carries out Fast Fourier Transform (FFT) and obtains real time spectrum, detects and extract the signal center frequency in frequency range automatically, and extracts the frequency spectrum data and I/Q data of each signal correspondence length according to spectrum signature and calculate character pair parameter；Classification is carried out to time domain and frequency domain character parameter using machine learning algorithm to judge to recognize the modulation system of analog signal.Using technical scheme, all analog signal modulation types that can be in the range of Real time identification wide-band, with robustness and scalability.
Description
Technical field
The present invention relates to radio monitoring and SIGNT analysis field, and in particular to a kind of broadband monitoring pattern Imitating
Signal modulation style real-time identification method.
Background technology
Signal modulation pattern-recognition is the important topic of radio art, is that future radios monitor important development side
To often using pattern-recongnition method at present.Mode identification method is a kind of method based on signal data feature extraction, meter
Signal is calculated in time domain, frequency domain character parameter, signal prior information is independent of substantially, realize the blind recognition of signal modulation mode.Closely
Nian Lai, theoretical research personnel propose various effective characteristic parameters, also have a small amount of application in radio monitoring field.But it is existing
Some recognizers, are directed to arrowband individual signals and are identified, and are still deposited for wide-band multi signal and the Real time identification deposited
In limitation, especially face and be subject to after the influence such as multipath fading, loss what is be an actually-received in transmission signal communication process
There is larger distortion in signal data, and signal data it is on the low side when, Characteristic parameter identification degree declines, and signal modulate occurs larger
Deviation, without robustness.
The content of the invention
The object of the invention is directed to during radio monitoring, due to signal transmission during real-time monitoring wide-band range signal
The interference that process is present causes signal characteristic abstraction distortion, it is impossible to the problem for effectively recognizing, there is provided a kind of modulated-analog signal is wide
Frequency range real-time identification method.The real-time broadband radiofrequency signal data that methods described will be received, obtain by data prediction
Time domain I/Q data and frequency domain power modal data, the corresponding time domain data of each signal of counting statistics and frequency spectrum data feature are joined respectively
Number, is realized to the Real time identification of signal using based on random forest machine learning algorithm.
To achieve the above object, the present invention is adopted the following technical scheme that：
A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method comprises the following steps：
Step a：Obtain broadband real-time radio frequency signal and be converted into intermediate-freuqncy signal；
Step b：A/D samplings are carried out to intermediate-freuqncy signal, Digital Down Convert (DDC) treatment exports I/Q data；
Step c：The complete frequency spectrum data that Fast Fourier Transform (FFT) (FFT) is obtained in frequency range are carried out to I/Q data；
Step d：Frequency domain noise threshold value is calculated automatically using wavelet data processing module, and judges thresholding as signal
threshold；
Step e：The signal number n and respective centre frequency Fi more than thresholding are detected using wavelet data processing module；
Step f：Calculate the frequency interval Fd of signal and both sides adjacent signals, and will wherein smaller value as M signal number
According to extraction length Lo；
Step g：The frequency spectrum data of each signal is extracted according to data length Lo and signal frequency Fi, and calculates signal spectrum
Kurtosis K；
Step h：The size of comparison signal kurtosis K and thresholding K_threshold, adjustment signal data extracts length Lo_i；
Step i：The corresponding frequency spectrum data of each signal and IQ numbers are extracted according to data length Lo_i and signal frequency again
According to, and normalized is done to frequency spectrum data；
Step j：According to the I/Q data after stage extraction, signal time domain charactreristic parameter is calculated：Instantaneous amplitude irrelevance R_d,
Using wavelet transformation data processing module to I/Q data noise reduction process, instantaneous phase irrelevance P_d is calculated；
Step k：According to the signal spectrum data after stage extraction, signal frequency domain characteristic parameter is calculated：Frequency spectrum kurtosis K, root
Spectrum peak points Peak_num is calculated according to the frequency spectrum data after normalization；
Step l：Classification judgement is carried out to the signal time domain and frequency domain parameter that are input into using random forests algorithm module；
Step m：Classification is completed, output result.
Preferably, deterministic process is as follows in step l；
AM：P_d<0.001、R_d>0.2、K>10、Peak_num<5；
CW：P_d<0.001、R_d<0.001、K>20th, Peak_num=1
FM：P_d>0.2、R_d<0.001、K<1.5、Peak_num>5.
Beneficial effects of the present invention are as follows：
The invention has the advantages that realizing the real-time grading identification to all modulated-analog signals in broad frequency range.Wherein
Using the automatic of wavelet transformation data processing modules implement pectrum noise THRESHOLD ESTIMATION and signal frequency and fragmented spectrum data
Extract, it is by calculating the time domain and frequency domain character parameter of signal and special to multiple signals using random forests algorithm processing module
Levy parameter and do classification judgement, realize the broadband real-time grading identification of online simulation modulated signal.This technology invention uses machine
Random forests algorithm in study automatically creates decision-making woodlot and realizes classification forecast function, has more than general single decision tree
Good robustness.Increase for data signal calculation of characteristic parameters to realize the Classification and Identification to data signal simultaneously, with good
Good autgmentability and versatility.
Brief description of the drawings
Fig. 1 is of the invention to realize schematic flow sheet
Specific embodiment
The present invention is described in further detail with reference to embodiment
As shown in figure 1, the present invention provides a kind of modulated-analog signal wide-band real-time identification method, step is implemented such as
Under：
Step a：Obtain broadband real-time radio frequency signal and be converted into intermediate-freuqncy signal
Step b：A/D samplings are carried out to intermediate-freuqncy signal, Digital Down Convert (DDC) treatment exports I/Q data
Step c：The complete frequency spectrum data that Fast Fourier Transform (FFT) (FFT) is obtained in frequency range are carried out to I/Q data
Step d：Frequency domain noise threshold value threshold is calculated automatically using wavelet data processing module, and is sentenced as signal
Disconnected thresholding
Step e：The signal number n and centre frequency Fi more than thresholding are detected using wavelet data processing module
Step f：The frequency interval Fd between adjacent signals is calculated, and wherein will extract length as signal data by smaller value
Lo
Step g：The frequency spectrum data of each signal is extracted according to data length Lo and signal frequency Fi, and calculates signal spectrum
Kurtosis K
Step h：The size of comparison signal kurtosis K and thresholding K_threshold, adjustment signal data extracts length Lo_i
Step i：The corresponding frequency spectrum data of each signal and IQ numbers are extracted according to data length Lo_i and signal frequency again
According to, and normalized is done to frequency spectrum data
Step j：According to the I/Q data after stage extraction, signal time domain charactreristic parameter is calculated：Instantaneous amplitude irrelevance R_d,
Using wavelet transformation data processing module to I/Q data noise reduction process, instantaneous phase irrelevance P_d is calculated.
Step k：According to the signal spectrum data after stage extraction, signal frequency domain characteristic parameter is calculated：Frequency spectrum kurtosis K, root
Spectrum peak points Peak_num is calculated according to the frequency spectrum data after normalization.
Step l：Decision-making woodlot is automatically created using random forests algorithm grader, signal time domain and frequency domain parameter are carried out
Classification judgement, general judging characteristic
-AM：P_d<0.001、R_d>0.2、K>10、Peak_num<5；
-CW：P_d<0.001、R_d<0.001、K>20th, Peak_num=1
-FM：P_d>0.2、R_d<0.001、K<1.5、Peak_num>5
Step m：Classification terminates, output result.
The present invention be introduced primarily into wavelet transformation data processing module realize signal bottom make an uproar thresholding estimation and more than door
The signal center frequency of limit is extracted, and counts the corresponding data length that the next dynamic of frequency spectrum data kurtosis adjusts each signal.Root
According to the parameter for having identification higher during modulated-analog signal feature selecting actual propagation.Make after the completion of calculation of characteristic parameters
The classification for realizing analog signal modulation type with random forests algorithm grader judges.
Broadband monitoring pattern Imitating modulated signal real-time identification method of the invention, comprises the following steps：Receive wideband
Radiofrequency signal and output intermediate-freuqncy signal is changed in segment limit；A/D samplings are carried out to intermediate-freuqncy signal, Digital Down Convert DDC treatment is defeated
Go out I/Q data；I/Q data to being input into carries out Fast Fourier Transform (FFT) and obtains real time spectrum, detects automatically and extracts in frequency range
Signal frequency, and the frequency spectrum data and I/Q data of each signal correspondence length are extracted according to spectrum signature and character pair ginseng is calculated
Number, the modulation system of analog signal is judged using machine learning algorithm according to time domain and frequency domain character parameter to classify.Using this
The technical scheme of invention, all analog signal modulation types that can be in the range of Real time identification wide-band, with robustness and can
Autgmentability.
Claims (2)
1. a kind of broadband monitoring pattern Imitating signal modulation style real-time identification method, it is characterised in that comprise the following steps：
Step a：Obtain broadband real-time radio frequency signal and be converted into intermediate-freuqncy signal；
Step b：A/D samplings are carried out to intermediate-freuqncy signal, Digital Down Convert (DDC) treatment exports I/Q data；
Step c：The complete frequency spectrum data that Fast Fourier Transform (FFT) (FFT) is obtained in frequency range are carried out to I/Q data；
Step d：Frequency domain noise threshold value is calculated automatically using wavelet data processing module, and judges thresholding as signal
threshold；
Step e：The signal number n and respective centre frequency Fi more than thresholding are detected using wavelet data processing module；
Step f：The frequency interval Fd of signal and both sides adjacent signals is calculated, and wherein smaller value will be carried as M signal data
Take length Lo；
Step g：The frequency spectrum data of each signal is extracted according to data length Lo and signal frequency Fi, and calculates signal spectrum kurtosis
K；
Step h：The size of comparison signal kurtosis K and thresholding K_threshold, adjustment signal data extracts length Lo_i；
Step i：The corresponding frequency spectrum data of each signal and I/Q data are extracted according to data length Lo_i and signal frequency again, and
Normalized is done to frequency spectrum data；
Step j：According to the I/Q data after stage extraction, signal time domain charactreristic parameter is calculated：Instantaneous amplitude irrelevance R_d, uses
Wavelet transformation data processing module calculates instantaneous phase irrelevance P_d to I/Q data noise reduction process；
Step k：According to the signal spectrum data after stage extraction, signal frequency domain characteristic parameter is calculated：Frequency spectrum kurtosis K, according to returning
Frequency spectrum data after one change calculates spectrum peak points Peak_num；
Step l：Classification judgement is carried out to the signal time domain and frequency domain parameter that are input into using random forests algorithm module；
Step m：Classification is completed, output result.
2. broadband monitoring pattern Imitating signal modulation style real-time identification method as claimed in claim 1, it is characterised in that
Deterministic process is as follows in step l；
AM：P_d<0.001、R_d>0.2、K>10、Peak_num<5；
CW：P_d<0.001、R_d<0.001、K>20th, Peak_num=1；
FM：P_d>0.2、R_d<0.001、K<1.5、Peak_num>5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710131554.9A CN106911603A (en) | 2017-03-07 | 2017-03-07 | A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710131554.9A CN106911603A (en) | 2017-03-07 | 2017-03-07 | A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106911603A true CN106911603A (en) | 2017-06-30 |
Family
ID=59186655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710131554.9A Pending CN106911603A (en) | 2017-03-07 | 2017-03-07 | A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106911603A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154164A (en) * | 2017-11-15 | 2018-06-12 | 上海微波技术研究所（中国电子科技集团公司第五十研究所） | Signal of communication modulation classification system and method based on deep learning |
| CN111585671A (en) * | 2020-04-15 | 2020-08-25 | 国网河南省电力公司郑州供电公司 | Electric power LTE wireless private network electromagnetic interference monitoring and identifying method |
| CN112422465A (en) * | 2019-10-09 | 2021-02-26 | 上海矢元电子有限公司 | Signal modulation identification equipment |
| CN113872710A (en) * | 2021-10-27 | 2021-12-31 | 中北大学 | Programmable radio signal real-time monitoring alarm system and method |
| CN114666236A (en) * | 2022-03-29 | 2022-06-24 | 北京扬铭科技发展有限责任公司 | Full-automatic signal detection, identification and alarm method |
| CN116582195A (en) * | 2023-06-12 | 2023-08-11 | 浙江瑞通电子科技有限公司 | Unmanned aerial vehicle signal spectrum recognition algorithm based on artificial intelligence |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102629879A (en) * | 2012-03-21 | 2012-08-08 | 华南理工大学 | Underwater acoustic communication method based on mode frequency modulation |
| CN103472141A (en) * | 2013-09-09 | 2013-12-25 | 北京工业大学 | Signal demodulation method for recognizing vibrant noise modulation mechanism |
| CN104796366A (en) * | 2015-04-10 | 2015-07-22 | 长春理工大学 | Communication signal system identification system and method |
| US20160294504A1 (en) * | 2015-03-31 | 2016-10-06 | Allen-Vanguard Corporation | System and Method for Classifying Signal Modulations |
| CN106330805A (en) * | 2016-08-29 | 2017-01-11 | 重庆会凌电子新技术有限公司 | Automatic signal modulation mode identification method and system |
-
2017
- 2017-03-07 CN CN201710131554.9A patent/CN106911603A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102629879A (en) * | 2012-03-21 | 2012-08-08 | 华南理工大学 | Underwater acoustic communication method based on mode frequency modulation |
| CN103472141A (en) * | 2013-09-09 | 2013-12-25 | 北京工业大学 | Signal demodulation method for recognizing vibrant noise modulation mechanism |
| US20160294504A1 (en) * | 2015-03-31 | 2016-10-06 | Allen-Vanguard Corporation | System and Method for Classifying Signal Modulations |
| CN104796366A (en) * | 2015-04-10 | 2015-07-22 | 长春理工大学 | Communication signal system identification system and method |
| CN106330805A (en) * | 2016-08-29 | 2017-01-11 | 重庆会凌电子新技术有限公司 | Automatic signal modulation mode identification method and system |
Non-Patent Citations (3)
| Title |
|---|
| K.HASSAN等: "Automatic Modulation Recognition Using Wavelet Transform and Neural Networks in Wireless Systems", 《NEUTRAL INFORMATION PROCESSING》 * |
| KENLAU等: "Modulation recognition in the 868 MHz band using classification trees and random forests", 《INTERNATIONAL JOURNAL OF ELECTRONICS AND COMMUNICATIONS》 * |
| 王鑫等: "基于随机森林的认知网络主用户信号调制类型识别算法", 《东北大学学报》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154164A (en) * | 2017-11-15 | 2018-06-12 | 上海微波技术研究所（中国电子科技集团公司第五十研究所） | Signal of communication modulation classification system and method based on deep learning |
| CN112422465A (en) * | 2019-10-09 | 2021-02-26 | 上海矢元电子有限公司 | Signal modulation identification equipment |
| CN112422465B (en) * | 2019-10-09 | 2021-10-22 | 上海矢元电子股份有限公司 | Signal modulation identification equipment |
| CN111585671A (en) * | 2020-04-15 | 2020-08-25 | 国网河南省电力公司郑州供电公司 | Electric power LTE wireless private network electromagnetic interference monitoring and identifying method |
| CN111585671B (en) * | 2020-04-15 | 2022-06-10 | 国网河南省电力公司郑州供电公司 | Electric power LTE wireless private network electromagnetic interference monitoring and identifying method |
| CN113872710A (en) * | 2021-10-27 | 2021-12-31 | 中北大学 | Programmable radio signal real-time monitoring alarm system and method |
| CN114666236A (en) * | 2022-03-29 | 2022-06-24 | 北京扬铭科技发展有限责任公司 | Full-automatic signal detection, identification and alarm method |
| CN116582195A (en) * | 2023-06-12 | 2023-08-11 | 浙江瑞通电子科技有限公司 | Unmanned aerial vehicle signal spectrum recognition algorithm based on artificial intelligence |
| CN116582195B (en) * | 2023-06-12 | 2023-12-26 | 浙江瑞通电子科技有限公司 | Unmanned aerial vehicle signal spectrum identification method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106911603A (en) | A kind of broadband monitoring pattern Imitating signal modulation style real-time identification method | |
| Pan et al. | Specific emitter identification based on deep residual networks | |
| CN107124381B (en) | Automatic identification method for digital communication signal modulation mode | |
| CN110855591B (en) | QAM and PSK signal intra-class modulation classification method based on convolutional neural network structure | |
| Güner et al. | Automatic digital modulation classification using extreme learning machine with local binary pattern histogram features | |
| CN107038421A (en) | Modulation Types recognition methods based on sparse storehouse own coding | |
| CN112347871B (en) | Interference signal modulation identification method for communication carrier monitoring system | |
| CN103067325A (en) | Cooperative modulation identification method based on multi-class characteristic parameters and evidence theory | |
| CN102571230A (en) | Distributed collaborative signal identification method based on blind estimation of higher order statistics and signal to noise ratio | |
| CN108052956A (en) | Wireless light communication subcarrier modulation constellation recognition methods under a kind of atmospheric turbulance | |
| WO2021103206A1 (en) | Machine learning algorithm-based wireless radio frequency device identity recognition method and system | |
| CN114358064B (en) | Interference detection device and method based on deep support vector data description | |
| CN111027614B (en) | Noise enhanced radio frequency fingerprint identification method and device | |
| CN111510255A (en) | Frequency hopping signal blind detection and parameter estimation method based on broadband frequency spectrum data | |
| CN114422311B (en) | Signal modulation recognition method and system combining deep neural network and expert priori features | |
| CN110826425B (en) | VHF/UHF frequency band radio signal modulation mode identification method based on deep neural network | |
| CN104408474B (en) | Emitter Recognition and device | |
| CN112014801B (en) | SPWVD and improved AlexNet based composite interference identification method | |
| Wang et al. | Specific emitter identification based on deep adversarial domain adaptation | |
| CN104883328A (en) | Self-adaptive shortwave frequency hopping system channel estimation method | |
| Hazza et al. | Robustness of digitally modulated signal features against variation in HF noise model | |
| CN102082619B (en) | Transmission adaptive method based on double credible evaluations | |
| CN113242201B (en) | Wireless signal enhanced demodulation method and system based on generation classification network | |
| CN115964631A (en) | Signal modulation type identification method based on machine learning | |
| CN104392252A (en) | Method and device of identifying radiation sources |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170630 |
|
| WD01 | Invention patent application deemed withdrawn after publication |