CN108427987A - A kind of Modulation Mode Recognition method based on convolutional neural networks - Google Patents
A kind of Modulation Mode Recognition method based on convolutional neural networks Download PDFInfo
- Publication number
- CN108427987A CN108427987A CN201810190856.8A CN201810190856A CN108427987A CN 108427987 A CN108427987 A CN 108427987A CN 201810190856 A CN201810190856 A CN 201810190856A CN 108427987 A CN108427987 A CN 108427987A
- Authority
- CN
- China
- Prior art keywords
- signal
- modulation
- neural networks
- convolutional neural
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Abstract
The present invention proposes a kind of Modulation Mode Recognition method based on convolutional neural networks, belongs to field of image recognition.Its feature is to include the following steps：1）The generation or acquisition of radiofrequency signal data；2）Radiofrequency signal data are classified and are arranged according to modulation system；3）By the I/Q two-way radiofrequency signal data of acquisition, with the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）For the longitudinal axis, corresponding signal constellation (in digital modulation) figure is generated；4）A small amount of electromagnetic signal planisphere picture is input to convolutional neural networks and carries out model construction, what category of model exported is the modulation system of signal；5）A small amount of radiofrequency signal planisphere verification picture is input in training pattern, the classification accuracy of model is verified；6）Finally, test pictures are input to convolutional neural networks and carry out Classification and Identification.The present invention is high to the classification accurate rate of rf-signal modulation mode, and model construction does not need mass data, and important in inhibiting is identified to radiofrequency signal.
Description
Technical field
The Modulation Mode Recognition method based on convolutional neural networks that the present invention relates to a kind of, belongs to field of image recognition, can
It is identified for rf-signal modulation mode.
Background technology
Deep learning is a kind of based on the method for carrying out representative learning to data in machine learning.The concept of deep learning is
It was proposed in 2006 by Hinton et al..The concept of deep learning is derived from deep neural network, contains the more of multiple hidden layers
Layer perceptron is exactly a kind of deep learning structure.Deep learning is higher by feature by what combination low-level feature formation was more abstracted,
To find that the distributed nature of data indicates.Convolutional neural networks have been developed in recent years efficient identification method.20th century
The sixties, Hubel and Wiese have found its uniqueness when being used for the neuron of local sensitivity and set direction in studying cat cortex
Network structure can be effectively reduced the complexity of Feedback Neural Network, then propose convolutional neural networks.Convolutional Neural
Network includes two layers, and one is characterized extract layer, secondly being characterized mapping layer.Realize that the feature of input picture carries by first layer
It takes, structural nonlinear is made by the second layer.
Traditional rf-signal modulation mode automatic identifying method mainly has statistical pattern recognition method and decision theory to know
Other method.Recognition methods is mainly made of Signal Pretreatment, feature extraction and type identification three parts, traditional automatic identifying method
Accuracy rate is relatively low, operand is big, identification is difficult.Deep learning convolutional neural networks have feature extraction functions.Naturally by convolution
Neural network and signal modulation mode identification connect.In digital communicating field, often by digital signal table on a complex plane
Show, intuitively to indicate the relationship between signal and signal.The distribution map of signal phasor endpoint is called planisphere.Planisphere
For judging that modulation system has very intuitive effectiveness.The data of radiofrequency signal storage are I/Q two-way Time Domain Amplitude data, by I/Q
Two paths of data is with the roads I（Same phase）Data are horizontal axis, the roads Q（It is orthogonal）Data are the longitudinal axis, generate corresponding signal constellation (in digital modulation) figure.By radio frequency
Signal constellation (in digital modulation) figure picture is sent into convolutional neural networks, and the modulation system of radiofrequency signal can be obtained, and is modulated to radiofrequency signal recognition
Mode important in inhibiting.
Invention content
The present invention proposes a kind of Modulation Mode Recognition method based on convolutional neural networks, and this method is intended to using a small amount of
Radiofrequency signal planisphere picture constructs convolutional neural networks model, with realizing the high-accuracy of rf-signal modulation mode knowledge of classifying
Not.
The technical solution adopted in the present invention is as follows：
Modulation Mode Recognition method based on convolutional neural networks, is as follows：
Step 1：The generation or acquisition of radiofrequency signal data；
Step 2：Radiofrequency signal data are classified and are arranged according to modulation system；
Step 3：By the I/Q two-way radiofrequency signal data of acquisition, with the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）For the longitudinal axis, generate
Corresponding signal constellation (in digital modulation) figure；
Step 4：A small amount of radiofrequency signal planisphere picture is input to convolutional neural networks and carries out model construction, category of model output
Be signal modulation system；
Step 5：A small amount of radiofrequency signal planisphere verification picture is input in training pattern, the classification accuracy of model is verified；
Step 6：Test pictures are input to convolutional neural networks and carry out Classification and Identification；
Advantageous effect：The present invention carries out feature extraction and classification using convolutional neural networks to radiofrequency signal planisphere picture, real
The now image classification of quick pinpoint accuracy.The present invention has the characteristics that structure model data amount is few and accuracy is high, is suitble to use
It is identified in accurately rf-signal modulation mode, to electronic countermeasure important in inhibiting.
Description of the drawings
Fig. 1 is present system schematic diagram
Fig. 2 is radiofrequency signal data generation module figure of the present invention
Fig. 3 is the roads IQ of the present invention signal and planisphere relationship definition graph
Fig. 4 is radiofrequency signal planisphere of the present invention
Fig. 5 is convolutional neural networks schematic diagram of the present invention
Fig. 6 is training performance figure of the present invention to radiofrequency signal planisphere
Fig. 7 is Modulation Mode Recognition design sketch of the present invention.
Specific implementation mode
The present invention is described in further detail With reference to embodiment, it should be understood that described below
Preferred embodiment is only for the purpose of illustrating and explaining the present invention and is not intended to limit the present invention：
1. the generation of radiofrequency signal data
Using various modulated signals in software radio software GNU Radio emulation actual life, as AM-DSB, CPFSK,
QAM64 etc..Analogue system includes raw data module, transmitting modulation module, channel simulator module, data acquisition memory module.
Raw data module selects simulation signal generator or derived digital signal, analog signal choosing respectively for analog-modulated and digital modulation
The audio file of .mp3 formats is selected, digital signal selects text file as signal source.Transmitting modulation module shares 11 kinds of differences
Modulation system, analog-modulated has three ways, such as WBFM, AM-DSB, AM-SSB, digital modulation have BPSK, QPSK, 8PSK,
8 kinds of modulation systems of PAM4, QAM16, QAM64, GFSK, CPFSK.Data acquire memory module using in GNU radio
Dynamic Channel Model modules, by the way that sampling frequency deviation, carrier deviation, selection degenerated mode, additivity is arranged
The parameters such as white Gaussian noise, the actual channel transmission during simulation is real.Data acquire memory module to output stream time-domain signal into
128 sequential sampling of row I/Q two-way.Each sampled data is stored in the form of 2 × 128.The corresponding moment sequential sampling signal of storage
Signal-to-noise ratio（SNR）And modulation system（MOD）.Final all data are packaged to be stored with .dat formatted files.
2. the classification and arrangement of radiofrequency signal data
Test a total of 11 kinds of modulation systems of data generated（3 kinds of analog-modulateds and 8 kinds of digital modulations）, signal-to-noise ratio coverage area
From -20dB ~+18dB.2 × 128 radiofrequency signal data of shared I/Q two-way 220000.Classify according to signal-to-noise ratio ,-
20, -18, -16 ...+16 ,+18 totally 20 classes separately include 11 kinds of modulation systems, under same signal-to-noise ratio, often together per class
Modulation system includes 1000 data.SNR and modulation system label are stamped to every data（The needs of Training below）.
3. generating signal constellation (in digital modulation) figure
Radiofrequency signal can resolve into one group of relatively independent component, i.e., same to phase（I）With it is orthogonal（Q）Component, the two components are
Orthogonal, and it is mutually incoherent.With the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）A rectangular coordinate system is established for the longitudinal axis, by I/Q
Two-way radiofrequency signal data are mapped in I, Q plane, and each data of acquisition become 128 points in I, Q plane, these points claim
Make constellation, such figure is referred to as planisphere.At each SNR, corresponding signal constellation (in digital modulation) figure is generated according to modulation system respectively.
4. model construction
Choose AM-DSB, AM-SSB, CPFSk, GFSK, PAM4, QAM64 totally 6 kinds of modulation system signals, -20dB, -18dB, -
16dB ...+16dB ,+18dB are in the case of totally 20 Signal to Noise Ratio (SNR).Under a certain Signal to Noise Ratio (SNR), 2700 training figures are shared
Piece and 2700 verification pictures, each modulation system respectively have 450 trained pictures, 450 verification pictures, are per pictures size
150 × 120 pixels.Using Lenet neural networks, model is by 3 convolutional layers, 3 pond layers, 1 flat layer and 1 output
Layer is constituted, and optimizer selects Adam, loss function to select categorical_crossentropy functions.Last output layer selection
Softmax classification activation primitives, other layer choosings select relu activation primitives.There are six nodes for last output layer, correspond to 6 kinds of tune respectively
Mode processed.
5. model is verified
Under a certain signal-to-noise ratio, with 600 kinds of radiofrequency signal planisphere pictures, each 100 of each modulation system, test model classification
Accuracy rate.Use signal constellation (in digital modulation) figure picture under the different signal-to-noise ratio in -20dB ~+18dB ranges, test model classification accurate again
Rate.Record test the time it takes.The feasibility of model is verified in comparative analysis.
6. category of model
The modulation system of signal in signal constellation (in digital modulation) figure input model to be identified, will be accurately identified, adjust affiliated per pictures is exported
Mode type processed,
The method of the present invention passes through under 20 SNR, per the lower 6 kinds of modulation systems of SNR, 600 constellation pictures are verified, in height totally
Under SNR, classification accuracy is high.Wherein, identify that the type of planisphere picture number and modulation system can be added arbitrarily.
Attached drawing is described in detail
Fig. 1 is systematic schematic diagram, it schematically illustrates entire workflow.Generation or acquisition, radio frequency letter including radiofrequency signal
The classification and arrangement of number generate corresponding planisphere, deep learning frame model training constellation according to radiofrequency signal data
Figure image data, according to modulation system Classification and Identification.The convolutional neural networks model that deep learning frame model oneself is built.
Fig. 2 is radiofrequency signal data generation module definition graph, it is imitated by raw data module, transmitting modulation module, channel
True module, 4 part of data acquisition collection module composition.Raw data module is by two kinds of simulation signal generator and derived digital signal；Hair
Penetrating modulation module has 3 kinds of analog-modulateds and 8 kinds of digital modulations；Channel simulator module is inclined comprising sampling frequency deviation, centre frequency
It moves, 4 selection degenerated mode, additive white Gaussian noise module compositions.
Fig. 3 is IQ two paths of signals and radio frequency signal amplitude phase relation definition graph.As can be seen from the figure：
。
Fig. 4 is under identical Signal to Noise Ratio (SNR), and the planisphere of different modulating mode radiofrequency signal is different.Four shown in figure
Subgraph is the radiofrequency signal planisphere of tetra- kinds of modulation systems of CPFSK, PAM4, QAM64, GFSK when SNR is equal to 2dB.It can from figure
To find out that the radiofrequency signal planisphere shape feature in a manner of different modulating is different.
Fig. 5 is the convolutional neural networks built（CNN）Structural model.The model include 3 convolutional layers, 3 pond layers,
Full articulamentum, output layer etc..Prototype network input layer is the planisphere picture of 150 × 120 sizes, and convolution kernel size is 3 × 3,
Maximum pond core size is 2 × 2.Finally export 6 nodes, corresponding 6 kinds of different modulating modes.
When Fig. 6 is that SNR is equal to 6dB, model training performance figure.Horizontal axis is training epoch periods in figure, and the longitudinal axis is loss
Value or mistake.One epoch period indicates that all trained pictures of training set complete primary training, and penalty values indicate true tag
The deviation of value and realistic model output.Training process is exactly by the process of continuous adjustment model parameter so that penalty values constantly subtract
It is small, while paying attention to preventing over-fitting from occurring.As can be seen from the figure with the increase of epoch, model training accuracy rate is tested
Card accuracy rate is gradually increased, and penalty values are gradually reduced.
Fig. 7 is Modulation Mode Recognition design sketch.Horizontal axis is Signal to Noise Ratio (SNR), and the longitudinal axis is Classification and Identification accuracy rate.SNR covers
Lid -20dB ~+18dB ranges, from the point of view of entire SNR coverage areas, with the raising of SNR, Modulation Mode Recognition accuracy rate by
Gradually improve.When Signal to Noise Ratio (SNR) is below -8dB, recognition accuracy is low, to no effect.In -6dB, recognition accuracy reaches SNR
70% or more.For SNR in 0dB, recognition accuracy reaches 90% or more.When SNR is 0dB or more, recognition accuracy is higher, works as SNR
For 14dB or more when, Classification and Identification accuracy rate levels off to 100%.
Claims (7)
1. a kind of Modulation Mode Recognition method based on convolutional neural networks, characterized in that radiofrequency signal recognition modulation system letter
Folk prescription is just and recognition accuracy is high, including following steps：
The generation or acquisition of step 1, radiofrequency signal data
Using various modulated signals in software radio software GNU Radio emulation actual life, as AM-DSB, CPFSK,
QAM64 etc., analogue system include transmitting modulation module, channel simulator module, data acquisition module, can also utilize wireless universal
Electric receiving device（USRP）The various radiofrequency signals in reality are received with GNU Radio softwares（Such as FM fm broadcast signals）；
A total of 11 kinds of modulation systems of data that step 2, radiofrequency signal data are classified according to modulation system and arrangement experiment generates
（3 kinds of analog-modulateds and 8 kinds of digital modulations）, signal-to-noise ratio coverage area from -20dB ~+18dB, penetrate by shared I/Q two-way 2 × 128
Frequency signal data 220000, classifies according to signal-to-noise ratio, -20, -18, -16 ...+16 ,+18 totally 20 classes, per class
11 kinds of modulation systems are separately included, under same signal-to-noise ratio, includes often 1000 data with modulation system, is stamped to every data
SNR and modulation system label；
Step 3, the I/Q two-way radiofrequency signal data by acquisition, with the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）For the longitudinal axis, generate
Corresponding signal constellation (in digital modulation) figure radiofrequency signal can resolve into one group of relatively independent component, i.e., same to phase（I）With it is orthogonal（Q）Component,
The two components are orthogonal, and mutually incoherent, with the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）A right angle is established for the longitudinal axis to sit
Mark system, I/Q two-way radiofrequency signal data is mapped in I, Q plane, each data of acquisition become 128 in I, Q plane
Point, these points are referred to as constellation, such figure is referred to as planisphere；
A small amount of radiofrequency signal planisphere picture is input to convolutional neural networks progress model construction, category of model output by step 4
Be signal modulation system choose 6 kinds of modulation system signals share 2700 training datas and 2700 under same signal-to-noise ratio
Verify data is opened, each modulation system respectively has 450 trained pictures, 450 verification pictures, and model is by 3 convolutional layers, 3 ponds
Change layer, 1 flat layer and 1 output layer to constitute, optimizer selects Adam, loss function to select categorical_
Crossentropy functions, last output layer selection softmax classification activation primitives, other layer choosings select relu activation primitives, most
There are six node, corresponding 6 kinds of modulation systems for output layer afterwards；
A small amount of radiofrequency signal planisphere verification picture is input in training pattern by step 5, verifies the classification accuracy of model
Under identical signal-to-noise ratio, with 600 kinds of radiofrequency signal planisphere pictures, each 100 of each modulation system, test model point
Class accuracy rate, record test the time it takes；
Test pictures are input to convolutional neural networks progress Classification and Identification by signal constellation (in digital modulation) figure input mould to be identified by step 6
In type, the modulation system of signal is accurately identified.
2. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
The generation or acquisition of radiofrequency signal data in rapid 1 carry out signal modulation, channel simulator, signal data using GNU Radio softwares
Processing and preservation, the RF electromagnetic signal that simulating realistic USRP is received.
3. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
The signal data of generation is carried out taxonomic revision, and stamps correspondence by the difference of the SNR of basis signal and modulation system in rapid 2
SNR, modulation system label.
4. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
By I/Q two-way radiofrequency signal data in rapid 3, with the roads I（Same phase）For horizontal axis, the roads Q（It is orthogonal）For the longitudinal axis, corresponding signal star is generated
Seat figure, constellation sizes are 150*120 pixels.
5. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
A small amount of planisphere training picture is input to convolutional neural networks by rapid 4 carries out model construction, and category of model output is signal
Modulation system, network structure use Lenet models, oneself successively add every layer, learning rate 0.5.
6. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
Rapid 5 are input to a small amount of verification picture in the model generated in step 4, by calculating planisphere picture recognition rf-signal modulation
The accuracy rate of mode verifies the feasibility of model.
7. a kind of Modulation Mode Recognition method based on convolutional neural networks according to claim 1, it is characterised in that：Step
Rapid 6 are input to picture to be tested in the model of structure, export the recognition result of planisphere, realize rf-signal modulation mode
Identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810190856.8A CN108427987A (en) | 2018-03-08 | 2018-03-08 | A kind of Modulation Mode Recognition method based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810190856.8A CN108427987A (en) | 2018-03-08 | 2018-03-08 | A kind of Modulation Mode Recognition method based on convolutional neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108427987A true CN108427987A (en) | 2018-08-21 |
Family
ID=63157629
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810190856.8A Pending CN108427987A (en) | 2018-03-08 | 2018-03-08 | A kind of Modulation Mode Recognition method based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108427987A (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109246048A (en) * | 2018-10-30 | 2019-01-18 | 广州海格通信集团股份有限公司 | A kind of safety of physical layer communication means and system based on deep learning |
| CN109787929A (en) * | 2019-02-20 | 2019-05-21 | 深圳市宝链人工智能科技有限公司 | Signal modulate method, electronic device and computer readable storage medium |
| CN109889212A (en) * | 2019-02-01 | 2019-06-14 | 华侨大学 | A kind of blind demodulation method based on deep learning and software radio |
| CN109886075A (en) * | 2018-12-27 | 2019-06-14 | 成都数之联科技有限公司 | A kind of signal modulation pattern recognition methods based on planisphere |
| CN110086737A (en) * | 2019-03-13 | 2019-08-02 | 西安电子科技大学 | A kind of recognition methods of the modulation mode of communication signal based on figure neural network |
| CN110191073A (en) * | 2019-06-28 | 2019-08-30 | 华侨大学 | A kind of Modulation Mode Recognition method based on deep learning suitable for scene change |
| CN110309854A (en) * | 2019-05-21 | 2019-10-08 | 北京邮电大学 | A kind of signal modulation mode recognition methods and device |
| CN110324080A (en) * | 2019-06-28 | 2019-10-11 | 北京邮电大学 | A kind of method, apparatus of optical information networks, electronic equipment and medium |
| CN110532671A (en) * | 2019-08-26 | 2019-12-03 | 北京航空航天大学 | A kind of sample data generation method for electromagnetic signal Classification and Identification |
| CN110601764A (en) * | 2019-09-16 | 2019-12-20 | 西南交通大学 | Radio frequency modulation format identification method based on optical assistance |
| CN111079347A (en) * | 2019-12-26 | 2020-04-28 | 华侨大学 | Signal-to-noise ratio estimation method based on deep learning by using constellation diagram |
| WO2020087293A1 (en) * | 2018-10-30 | 2020-05-07 | 华为技术有限公司 | Communication receiver and method for processing signal |
| CN111277526A (en) * | 2020-03-01 | 2020-06-12 | 西北工业大学 | Modulation identification method of constellation diagram identical signals based on compressed sensing |
| CN111585923A (en) * | 2020-03-23 | 2020-08-25 | 成都奥特为科技有限公司 | Modulation mode recognition device and system based on convolutional neural network |
| CN111585925A (en) * | 2020-04-18 | 2020-08-25 | 西北工业大学 | Robust real-time radio frequency signal modulation identification method based on deep learning |
| CN111884962A (en) * | 2020-06-01 | 2020-11-03 | 山东师范大学 | Signal modulation type classification method and system based on convolutional neural network |
| CN111935043A (en) * | 2020-08-05 | 2020-11-13 | 四川大学 | Phase modulation signal modulation mode identification method based on phase statistical chart |
| CN113361433A (en) * | 2021-06-16 | 2021-09-07 | 中国人民解放军国防科技大学 | Modulation signal identification method based on neural network and application thereof |
| CN113518050A (en) * | 2021-06-24 | 2021-10-19 | 华东交通大学 | Modulation identification method, system, readable storage medium and device |
| CN113657138A (en) * | 2020-05-12 | 2021-11-16 | 哈尔滨工程大学 | Radiation source individual identification method based on equipotential planet chart |
| CN113822162A (en) * | 2021-08-24 | 2021-12-21 | 北京邮电大学 | Convolutional neural network modulation identification method based on pseudo constellation diagram |
| US20210405109A1 (en) * | 2020-06-25 | 2021-12-30 | Rohde & Schwarz GbmH & Co. KG | Method and system for acquiring a measurement related dataset |
| CN114900407A (en) * | 2022-07-12 | 2022-08-12 | 南京科伊星信息科技有限公司 | Modulation mode automatic identification and countermeasure method based on data enhancement |
| CN114978827A (en) * | 2022-04-22 | 2022-08-30 | 深圳市人工智能与机器人研究院 | Modulation identification method for correcting frequency offset based on constellation diagram phase anomaly ratio |
| CN115150237A (en) * | 2022-06-09 | 2022-10-04 | 姚辰熙 | Radio signal identification technology based on deep learning algorithm |
| CN115173936A (en) * | 2022-06-30 | 2022-10-11 | 烽火通信科技股份有限公司 | Optical module identification marking method and device |
| CN115333902A (en) * | 2021-05-10 | 2022-11-11 | 陕西尚品信息科技有限公司 | Communication signal modulation identification method and device |
| CN115333905A (en) * | 2022-10-12 | 2022-11-11 | 南通中泓网络科技有限公司 | Signal modulation mode identification method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101764784A (en) * | 2009-12-11 | 2010-06-30 | 西安电子科技大学 | Quadrature amplitude modulation within-class identification method based on image processing under multipath channel |
| US20140269492A1 (en) * | 2013-03-12 | 2014-09-18 | Antonio Forenza | Systems and methods for exploiting inter-cell multiplexing gain in wireless cellular systems via distributed input distributed output technology |
| CN107220606A (en) * | 2017-05-22 | 2017-09-29 | 西安电子科技大学 | The recognition methods of radar emitter signal based on one-dimensional convolutional neural networks |
| CN107342962A (en) * | 2017-07-03 | 2017-11-10 | 北京邮电大学 | Deep learning intelligence Analysis On Constellation Map method based on convolutional neural networks |
-
2018
- 2018-03-08 CN CN201810190856.8A patent/CN108427987A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101764784A (en) * | 2009-12-11 | 2010-06-30 | 西安电子科技大学 | Quadrature amplitude modulation within-class identification method based on image processing under multipath channel |
| US20140269492A1 (en) * | 2013-03-12 | 2014-09-18 | Antonio Forenza | Systems and methods for exploiting inter-cell multiplexing gain in wireless cellular systems via distributed input distributed output technology |
| CN107220606A (en) * | 2017-05-22 | 2017-09-29 | 西安电子科技大学 | The recognition methods of radar emitter signal based on one-dimensional convolutional neural networks |
| CN107342962A (en) * | 2017-07-03 | 2017-11-10 | 北京邮电大学 | Deep learning intelligence Analysis On Constellation Map method based on convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| XIAOYU LIU ET AL: "Deep Neural Network Architectures for Modulation Classification", 《ARXIV:1712.00443V3》 * |
Cited By (38)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109246048B (en) * | 2018-10-30 | 2021-02-02 | 广州海格通信集团股份有限公司 | Physical layer secure communication method and system based on deep learning |
| CN109246048A (en) * | 2018-10-30 | 2019-01-18 | 广州海格通信集团股份有限公司 | A kind of safety of physical layer communication means and system based on deep learning |
| WO2020087293A1 (en) * | 2018-10-30 | 2020-05-07 | 华为技术有限公司 | Communication receiver and method for processing signal |
| CN109886075A (en) * | 2018-12-27 | 2019-06-14 | 成都数之联科技有限公司 | A kind of signal modulation pattern recognition methods based on planisphere |
| CN109889212A (en) * | 2019-02-01 | 2019-06-14 | 华侨大学 | A kind of blind demodulation method based on deep learning and software radio |
| CN109787929A (en) * | 2019-02-20 | 2019-05-21 | 深圳市宝链人工智能科技有限公司 | Signal modulate method, electronic device and computer readable storage medium |
| CN110086737A (en) * | 2019-03-13 | 2019-08-02 | 西安电子科技大学 | A kind of recognition methods of the modulation mode of communication signal based on figure neural network |
| CN110086737B (en) * | 2019-03-13 | 2021-07-02 | 西安电子科技大学 | Communication signal modulation mode identification method based on graph neural network |
| CN110309854A (en) * | 2019-05-21 | 2019-10-08 | 北京邮电大学 | A kind of signal modulation mode recognition methods and device |
| CN110191073B (en) * | 2019-06-28 | 2021-08-31 | 华侨大学 | Modulation mode identification method based on deep learning and suitable for changing scene |
| CN110191073A (en) * | 2019-06-28 | 2019-08-30 | 华侨大学 | A kind of Modulation Mode Recognition method based on deep learning suitable for scene change |
| CN110324080A (en) * | 2019-06-28 | 2019-10-11 | 北京邮电大学 | A kind of method, apparatus of optical information networks, electronic equipment and medium |
| CN110532671A (en) * | 2019-08-26 | 2019-12-03 | 北京航空航天大学 | A kind of sample data generation method for electromagnetic signal Classification and Identification |
| CN110601764A (en) * | 2019-09-16 | 2019-12-20 | 西南交通大学 | Radio frequency modulation format identification method based on optical assistance |
| CN111079347B (en) * | 2019-12-26 | 2023-08-01 | 华侨大学 | Signal-to-noise ratio estimation method based on deep learning by using constellation diagram |
| CN111079347A (en) * | 2019-12-26 | 2020-04-28 | 华侨大学 | Signal-to-noise ratio estimation method based on deep learning by using constellation diagram |
| CN111277526B (en) * | 2020-03-01 | 2021-06-11 | 西北工业大学 | Modulation identification method of constellation diagram identical signals based on compressed sensing |
| CN111277526A (en) * | 2020-03-01 | 2020-06-12 | 西北工业大学 | Modulation identification method of constellation diagram identical signals based on compressed sensing |
| CN111585923A (en) * | 2020-03-23 | 2020-08-25 | 成都奥特为科技有限公司 | Modulation mode recognition device and system based on convolutional neural network |
| CN111585925A (en) * | 2020-04-18 | 2020-08-25 | 西北工业大学 | Robust real-time radio frequency signal modulation identification method based on deep learning |
| CN113657138A (en) * | 2020-05-12 | 2021-11-16 | 哈尔滨工程大学 | Radiation source individual identification method based on equipotential planet chart |
| CN111884962A (en) * | 2020-06-01 | 2020-11-03 | 山东师范大学 | Signal modulation type classification method and system based on convolutional neural network |
| US11567117B2 (en) * | 2020-06-25 | 2023-01-31 | Rohde & Schwarz Gmbh & Co. Kg | Method and system for acquiring a measurement related dataset |
| US20210405109A1 (en) * | 2020-06-25 | 2021-12-30 | Rohde & Schwarz GbmH & Co. KG | Method and system for acquiring a measurement related dataset |
| CN111935043A (en) * | 2020-08-05 | 2020-11-13 | 四川大学 | Phase modulation signal modulation mode identification method based on phase statistical chart |
| CN115333902A (en) * | 2021-05-10 | 2022-11-11 | 陕西尚品信息科技有限公司 | Communication signal modulation identification method and device |
| CN113361433A (en) * | 2021-06-16 | 2021-09-07 | 中国人民解放军国防科技大学 | Modulation signal identification method based on neural network and application thereof |
| CN113518050A (en) * | 2021-06-24 | 2021-10-19 | 华东交通大学 | Modulation identification method, system, readable storage medium and device |
| CN113518050B (en) * | 2021-06-24 | 2021-12-24 | 华东交通大学 | Modulation identification method, system, readable storage medium and device |
| CN113822162A (en) * | 2021-08-24 | 2021-12-21 | 北京邮电大学 | Convolutional neural network modulation identification method based on pseudo constellation diagram |
| CN113822162B (en) * | 2021-08-24 | 2023-10-13 | 北京邮电大学 | Convolutional neural network modulation identification method based on pseudo constellation diagram |
| CN114978827A (en) * | 2022-04-22 | 2022-08-30 | 深圳市人工智能与机器人研究院 | Modulation identification method for correcting frequency offset based on constellation diagram phase anomaly ratio |
| CN115150237A (en) * | 2022-06-09 | 2022-10-04 | 姚辰熙 | Radio signal identification technology based on deep learning algorithm |
| CN115173936A (en) * | 2022-06-30 | 2022-10-11 | 烽火通信科技股份有限公司 | Optical module identification marking method and device |
| CN114900407B (en) * | 2022-07-12 | 2022-10-14 | 南京科伊星信息科技有限公司 | Modulation mode automatic identification and countermeasure method based on data enhancement |
| CN114900407A (en) * | 2022-07-12 | 2022-08-12 | 南京科伊星信息科技有限公司 | Modulation mode automatic identification and countermeasure method based on data enhancement |
| CN115333905A (en) * | 2022-10-12 | 2022-11-11 | 南通中泓网络科技有限公司 | Signal modulation mode identification method |
| CN115333905B (en) * | 2022-10-12 | 2023-01-03 | 南通中泓网络科技有限公司 | Signal modulation mode identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108427987A (en) | A kind of Modulation Mode Recognition method based on convolutional neural networks | |
| CN108234370B (en) | Communication signal modulation mode identification method based on convolutional neural network | |
| CN107979554B (en) | Radio signal Modulation Identification method based on multiple dimensioned convolutional neural networks | |
| CN110855591B (en) | QAM and PSK signal intra-class modulation classification method based on convolutional neural network structure | |
| CN110086737B (en) | Communication signal modulation mode identification method based on graph neural network | |
| Daldal et al. | Automatic determination of digital modulation types with different noises using convolutional neural network based on time–frequency information | |
| CN107547460A (en) | Radio communication Modulation Signals Recognition method based on deep learning | |
| CN107342962A (en) | Deep learning intelligence Analysis On Constellation Map method based on convolutional neural networks | |
| US11349743B2 (en) | Machine learning training system for identification or classification of wireless signals | |
| CN110598530A (en) | Small sample radio signal enhanced identification method based on ACGAN | |
| CN108540202A (en) | A kind of satellite communication signals Modulation Mode Recognition method, satellite communication system | |
| CN113378644B (en) | Method for defending signal modulation type recognition attack based on generation type countermeasure network | |
| CN107317778A (en) | BPSK modulating signal phase transition detection methods based on 1D CNN | |
| CN114943245A (en) | Automatic modulation recognition method and device based on data enhancement and feature embedding | |
| Liu et al. | Jamming recognition based on feature fusion and convolutional neural network | |
| Zhang et al. | A machine learning paradigm for Studying Pictorial realism: are constable's clouds more real than his contemporaries? | |
| CN116319210A (en) | Signal lightweight automatic modulation recognition method and system based on deep learning | |
| Fu et al. | A new method to solve the problem of facing less learning samples in signal modulation recognition | |
| CN114070439B (en) | Virtual-real combined channel mapping method and device and channel mapping system | |
| Cun et al. | Specific emitter identification based on eye diagram | |
| Pruengkarn et al. | An evaluation model for e-learning Websites in Thailand University | |
| US11507803B2 (en) | System for generating synthetic digital data for data multiplication | |
| CN114520758A (en) | Signal modulation identification method based on instantaneous characteristics | |
| CN114070688A (en) | Multi-standard underwater acoustic communication signal modulation identification method and system | |
| CN111935043A (en) | Phase modulation signal modulation mode identification method based on phase statistical chart |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180821 |