CN110728175A - Electromagnetic signal red and black recognition device and method for wireless network - Google Patents
Electromagnetic signal red and black recognition device and method for wireless network Download PDFInfo
- Publication number
- CN110728175A CN110728175A CN201910811601.3A CN201910811601A CN110728175A CN 110728175 A CN110728175 A CN 110728175A CN 201910811601 A CN201910811601 A CN 201910811601A CN 110728175 A CN110728175 A CN 110728175A
- Authority
- CN
- China
- Prior art keywords
- signal
- red
- black
- signals
- electromagnetic
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses an electromagnetic signal red and black recognition device and method for a wireless network, wherein the recognition device comprises an antenna, a receiver, a signal acquisition and processing unit, a calculation and analysis unit, a power module and a clock module; the receiver receives the wireless electromagnetic signal through the antenna, processes the signal and then sends the signal to the signal acquisition and processing unit; the signal acquisition and processing unit finishes the acquisition and processing work of the signal and outputs the signal to the calculation and analysis unit; and the calculation analysis unit identifies the wireless electromagnetic red and black signals by using the convolution neural network technology of artificial intelligence deep learning. The identification method is based on the identification device. The electromagnetic signal red and black recognition device and method for the wireless network solve the problem that the difference of red and black characteristics of electromagnetic signals is weak and extraction is difficult in actual environment, and break through the bottleneck that manual extraction of red and black characteristics is difficult.
Description
Technical Field
The invention relates to an electromagnetic signal red and black recognition device and method for a wireless network.
Background
With the development of information technology, wireless networks have covered various aspects of people's work and life, and network security issues are getting more and more concerned. Due to the open nature of the wireless communication network, radio waves are easily intercepted and analyzed or deciphered. An encryption algorithm is generally adopted to carry out encryption protection on information transmitted in a wireless network, and the encrypted signal is a black signal which is difficult to decipher after being intercepted by a person; the signal that is not encrypted is the red signal. If the information is transmitted without encryption (the information is clear), the electromagnetic transmission signal is intercepted or deciphered by people, and the red signal leakage can be caused. Therefore, the electromagnetic transmission signals of the wireless network need to be identified in red and black, so that the red and black states of the electromagnetic signals are monitored.
At present, under the condition of a known wireless communication protocol, a method for identifying red and black of an electromagnetic signal demodulates and decodes an encrypted signal, restores the demodulated and decoded encrypted signal into a 0-1 bit sequence, and then analyzes the 0-1 bit sequence and statistical characteristics (such as 0-1 equalization, run distribution characteristics, autocorrelation, information entropy and the like), thereby realizing the classification and identification of the red and black signals. The demodulation and decoding method of the known communication protocol only aims at single equipment and other network equipment which cannot be popularized, and the universality is poor; the wireless network terminals have various models, non-uniform communication systems and various modulation modes, and it is difficult to demodulate, decode and restore wireless electromagnetic signals into bit stream data under the condition of unknown wireless communication protocols.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides an electromagnetic signal red and black recognition device and an electromagnetic signal red and black recognition method for a wireless network, after a wireless electromagnetic signal is obtained, the wireless electromagnetic signal does not need to be demodulated, decoded and restored into bit stream data, the statistical characteristics of the wireless electromagnetic red and black signal do not need to be extracted manually, but a hierarchical multilayer structure of an artificial intelligent deep learning algorithm (convolutional neural network) is adopted, the characteristic extraction is completed step by step from a lower layer to a higher layer from an input signal, and finally the higher layer characteristics are mapped to a linearly separable characteristic space according to certain nonlinear combination to realize the signal red and black recognition. Meanwhile, the LeNet convolutional neural network (LeNet CNN) is applied to the wireless electromagnetic signal red and black recognition for the first time, and the LeNet CNN network is correspondingly improved in the application process, so that the accuracy of the convolutional neural network in the red and black signal classification recognition is improved.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an electromagnetic signal red and black recognition device for a wireless network comprises a receiver, a signal acquisition and processing unit, a calculation and analysis unit, a power module and a clock module;
the receiver receives the wireless electromagnetic signal, processes the signal and sends the signal to the signal acquisition and processing unit;
the signal acquisition and processing unit finishes the acquisition and processing work of the signal and outputs the signal to the calculation and analysis unit;
and the calculation analysis unit identifies the wireless electromagnetic red and black signals by using the convolution neural network technology of artificial intelligence deep learning.
An electromagnetic signal red and black identification method for a wireless network is based on the electromagnetic signal red and black identification device and comprises the following steps:
step one, a receiver receives a wireless electromagnetic signal;
step two, the receiver processes the received wireless electromagnetic signals, outputs broadband analog intermediate frequency signals and sends the signals to the signal acquisition and processing unit;
thirdly, the signal acquisition and processing unit finishes the acquisition and processing work of the broadband analog intermediate frequency signal, converts the intermediate frequency signal into a baseband signal and outputs the baseband signal to the calculation and analysis unit;
and step four, the calculation and analysis unit realizes the detection of the wireless electromagnetic red signal.
Compared with the prior art, the invention has the following positive effects:
aiming at the problem of safety detection of electromagnetic signals of a wireless network, the invention designs a novel method and a novel device for identifying and detecting red and black of electromagnetic signals, solves the problem that the difference of red and black characteristics of the electromagnetic signals is weak and difficult to extract in the actual environment, and breaks through the bottleneck that the characteristics of the red and black signals are difficult to extract manually. By adopting the technology, the machine can independently learn and extract the red and black signal characteristics by using the artificial intelligence deep learning, so that the method for identifying the wireless network electromagnetic red and black signals is realized, the accuracy of red and black signal identification and the adaptability to the parameter change of the red and black signals are improved, and the problem of real-time detection of red signal leakage in a wireless environment is solved.
Drawings
The invention will now be described, by way of example, with reference to the accompanying drawings, in which:
fig. 1 is a schematic diagram of an electromagnetic signal red and black identification device for a wireless network transmitting device according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of an improved LeNet convolutional neural network provided by an embodiment of the present invention.
Fig. 3 is a flowchart for training a convolutional neural network to recognize red and black signal features according to an embodiment of the present invention.
Fig. 4 is a flowchart of identifying red and black of electromagnetic signals of a wireless network according to an embodiment of the present invention.
Fig. 5 is a flowchart of an electromagnetic signal red-black identification method for a wireless network transmitting device according to an embodiment of the present invention.
Detailed Description
The invention provides an electromagnetic signal red and black recognition device for a wireless network transmitting device, which refers to fig. 1 and comprises an antenna, a receiver, a signal acquisition and processing unit, a calculation and analysis unit, a power supply module, a clock module, a display screen and a storage unit.
The receiver receives the electromagnetic signal of the wireless network transmitting device through the antenna, outputs a broadband analog intermediate frequency signal through processing of frequency conversion, filtering, amplification, gain control and the like, and sends the broadband analog intermediate frequency signal to the signal acquisition processing unit for processing. The signal acquisition and processing unit mainly finishes acquisition and processing of broadband analog intermediate frequency signals, and the processing unit mainly converts the analog intermediate frequency signals into time domain digital I/Q signals through ADC conversion and sends the time domain digital I/Q signals to the calculation and analysis unit. The calculation and analysis unit extracts the time domain and frequency domain basic features of the electromagnetic signals through preprocessing such as signal screening, filtering, extraction and FFT calculation, automatically extracts the signal features through feature learning and layering algorithm of a large sample by utilizing the convolutional neural network technology of artificial intelligent deep learning, classifies and identifies the red and black signals, displays the red and black identification result to a user interface on a display screen after detecting the red signals, and stores the red and black signal identification result in the storage unit.
The basic framework of the convolution neural network technology of artificial intelligence deep learning adopts a LeNet convolution neural network. Three improvements are made on the basis of a LeNet convolutional neural network:
1) a rectifying linear unit (ReLU) activation function is adopted to replace a Sigmoid activation function in a LeNet network; y ═ relu (x) ═ max (0, x)
The ReLU activation function has sparse activation characteristics, the calculation amount of the whole network can be reduced, and the problem of gradient disappearance of the Sigmoid function in back propagation is relieved.
2) Maximum pooling (Max-pooling) is adopted in the pooling layer to replace average pooling (Mean-pooling) in the LeNet network pooling layer, the maximum pooling can reduce estimated Mean shift caused by parameter errors of the convolutional layer, and more signal weak characteristic information is reserved.
3) During convolutional neural network training, a clipping technique (Dropout) is used for network nodes in convolutional layer C5 and fully-connected layer F6, i.e., the nodes in the network are discarded with a 50% probability, so that overfitting is prevented and training calculation amount is reduced.
The structural diagram of the improved LeNet convolutional neural network is shown in FIG. 2.
Specifically, the basic features of the red and black signals are selected as training set sample matrixes (6 thousands of the red and black signal sample matrixes 32x32 are selected, and 12 thousands of sample matrixes are calculated in total), and the convolutional neural network is trained in a supervised learning mode. Because the number of training set samples is large, the calculation amount for calculating the gradient descent for one time is large, therefore, the stochastic gradient descent algorithm is adopted to train the convolutional neural network: in each gradient descent calculation of the training algorithm, only one Batch (the number of sample matrixes is 50) is selected from the training set, and the red and black signal forward propagation and error backward propagation calculation processes are respectively completed. One iteration (Epoch) needs to traverse the entire training set, so 2400 gradient descent calculations and network parameter updates in total need to be completed.
(1) Forward propagation calculation process of red and black signals
Selecting 50 sample matrixes of red and black signals from a training set to form an input layer, alternately appearing twice through a convolutional layer and a pooling layer (namely comprising C1, S2, C3 and S4), then passing through a full-connection layer (comprising C5, F6 and an output layer), and finally calculating the error between the prediction result and the real result (sample label) of the current convolutional neural network model because the training data are marked by answers;
(2) error back propagation calculation process
Then, based on the error between the predicted value and the actual value obtained by forward propagation, the backward propagation algorithm calculates the gradient of each layer of network parameters in turn by using a random gradient descent algorithm according to the sequence of the output layer → F6 → C5 → S4 → C3 → S2 → C1 → the input layer, and updates each layer of network parameters by using the gradient, so as to enable the predicted result and the actual value of the neural network model on Batch to be closer. And continuously selecting a red and black signal sample matrix from the training set, and repeating the operation 2400 times until the whole training set is traversed, namely completing one iteration.
When the stochastic gradient descent algorithm is used, the whole training set needs to be traversed for each iteration, the cost function can be expected to be reduced along with the increase of the iteration times until the iteration times are converged to proper precision, the training process of the convolutional neural network is ended, and the parameter model of the convolutional neural network is stored. During training, the person skilled in the art can determine the required suitable convergence accuracy according to actual requirements.
A flow chart for training a convolutional neural network to recognize red and black signal features is shown in fig. 3.
The convolutional neural network is basically converged to an optimal network state through multiple iterations and optimization after being trained by red and black signals marked in a training set, and has the functions of completing extraction of red and black signal features layer by layer, mapping high-level features to a linearly separable feature space according to certain nonlinear combination, and finally accurately realizing classification and identification of the red and black signals.
The process of identifying the red and black of the electromagnetic signals of the wireless network is simpler than the training process, and the prediction result of the current convolutional neural network model is calculated only by carrying out the forward propagation calculation process on the samples of the red and black signals. Through tests, the red and black signal identification accuracy of wireless network electromagnetic signals received by the antenna under the condition of a certain signal-to-noise ratio (SNR is 10dB) is 96.8%.
The process of identifying red and black of the wireless network electromagnetic signal is shown in fig. 4.
Referring to fig. 5, the electromagnetic signal red and black identification method for a wireless network transmitting device according to the present invention, which uses the electromagnetic signal red and black identification device, includes the following steps:
step one, a receiver receives a wireless electromagnetic signal through an antenna;
step two, the receiver processes the received wireless electromagnetic signals, outputs broadband analog intermediate frequency signals and sends the signals to the signal acquisition and processing unit;
the receiver processes the received wireless electromagnetic signals by frequency conversion, filtering, amplification and gain control, wherein the frequency of the intermediate frequency signals is 140MHz, and the bandwidth of the intermediate frequency signals is not more than 40 MHz;
thirdly, the signal acquisition and processing unit finishes the acquisition and processing work of the broadband analog intermediate frequency signal, converts the intermediate frequency signal into a baseband signal and outputs the baseband signal to the calculation and analysis unit;
specifically, firstly, ADC conversion is carried out on an analog intermediate frequency signal, and the sampling clock of the ADC conversion is not higher than 250 MSPS; then the ADC data enters the FPGA for signal processing, after the processing of the DDC, filtering, FFT, multi-phase filtering channelization and the like, the intermediate frequency signal is converted into a baseband signal, and I, Q paths of data of the baseband are output to a calculation and analysis unit through a high-speed data interface;
step four, the calculation and analysis unit realizes the detection of the wireless electromagnetic red signal;
specifically, the above-mentioned artificial intelligence deep learning convolutional neural network technology is used to identify I, Q paths of data of the radio magnetic red and black signal, and detect whether the data is a red signal. And if the red signal is detected, displaying the red and black recognition result to a user interface on a display screen, and storing the red and black signal recognition result in a storage unit.
Claims (10)
1. An electromagnetic signal red and black recognition device for a wireless network is characterized by comprising a receiver, a signal acquisition and processing unit, a calculation and analysis unit, a power module and a clock module;
the receiver receives the wireless electromagnetic signal, processes the signal and sends the signal to the signal acquisition and processing unit;
the signal acquisition and processing unit finishes the acquisition and processing work of signals and outputs the signals to the calculation and analysis unit;
the calculation analysis unit identifies the wireless electromagnetic red and black signals by using a convolutional neural network technology of artificial intelligence deep learning.
2. The electromagnetic signal red and black identification device for the wireless network according to claim 1, further comprising an antenna, a display screen and a storage unit;
the receiver receives wireless electromagnetic signals through an antenna;
and after detecting the red signal, the calculation and analysis unit displays the red and black recognition result to a user interface on the display screen, and stores the red and black signal recognition result in the storage unit.
3. The electromagnetic signal red and black recognition device for the wireless network according to claim 1 or 2, wherein the basic framework of the artificial intelligence deep learning convolutional neural network technology adopts a LeNet convolutional neural network, and adopts a rectifying linear unit activation function to replace a Sigmoid activation function in the LeNet network, and the rectifying linear unit activation function is as follows:
y=ReLU(x)=max(0,x);
maximum pooling is adopted in the pooling layer to replace average pooling in the LeNet network pooling layer;
in the convolutional neural network training process, a cutting technique is adopted for network nodes in the convolutional layer C5 and the fully-connected layer F6.
4. The device for identifying the red and the black of the electromagnetic signals for the wireless network according to claim 3, wherein in the training process of the convolutional neural network, the device comprises the following steps:
s1, preparing a wireless electromagnetic red and black signal training set;
s2, selecting a Batch sample from the training set;
s3, calculating a network predicted value by signal forward propagation;
s4, calculating an error cost function according to the sample label;
s5, calculating the network parameter gradient through error back propagation;
s6, updating the network parameters according to the gradient;
s7, judging whether the whole training set is traversed or not, if so, carrying out the next step, and if not, repeating the steps S2 to S7;
and S8, judging whether the error cost function converges to proper precision, if so, saving the network model parameters, and if not, repeating the steps S2 to S8.
5. The apparatus for recognizing red and black signals of electromagnetic signals for wireless networks according to claim 4, wherein in step S5, the gradient of each layer of network parameters is calculated sequentially by using a random gradient descent algorithm.
6. An electromagnetic signal red and black identification method for a wireless network, which is based on the electromagnetic signal red and black identification device of any one of claims 1-5, and comprises the following steps:
step one, the receiver receives a wireless electromagnetic signal;
step two, the receiver processes the received wireless electromagnetic signals, outputs broadband analog intermediate frequency signals and sends the signals to the signal acquisition and processing unit;
step three, the signal acquisition and processing unit finishes the acquisition and processing work of broadband analog intermediate frequency signals, converts the intermediate frequency signals into baseband signals and outputs the baseband signals to the calculation and analysis unit;
and fourthly, the calculation and analysis unit realizes wireless electromagnetic red signal detection.
7. The method according to claim 6, wherein in the step two:
the receiver processing the received wireless electromagnetic signal comprises: frequency conversion, filtering, amplification and gain control.
8. The method according to claim 7, wherein in the second step, the intermediate frequency signal has a frequency of 140MHz, and the bandwidth of the intermediate frequency signal does not exceed 40 MHz.
9. The method according to claim 8, wherein in step three, ADC conversion is performed on the analog intermediate frequency signal, the sampling clock is not higher than 250MSPS, then the data of the ADC enters FPGA for signal processing, the intermediate frequency signal is converted into a baseband signal, and I, Q paths of data of the baseband are output to the computational analysis unit through a high-speed data interface;
the signal processing based on the FPGA comprises the steps of wideband DDC, filtering, FFT and polyphase filtering channelization.
10. The method according to claim 8, wherein in step four, the samples of the red and black signals are subjected to forward propagation calculation, and if the red signal is detected, the red and black recognition results are displayed on a display screen and stored.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910811601.3A CN110728175A (en) | 2019-08-30 | 2019-08-30 | Electromagnetic signal red and black recognition device and method for wireless network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910811601.3A CN110728175A (en) | 2019-08-30 | 2019-08-30 | Electromagnetic signal red and black recognition device and method for wireless network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110728175A true CN110728175A (en) | 2020-01-24 |
Family
ID=69218828
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910811601.3A Pending CN110728175A (en) | 2019-08-30 | 2019-08-30 | Electromagnetic signal red and black recognition device and method for wireless network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110728175A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117332410A (en) * | 2023-10-25 | 2024-01-02 | 北京航空航天大学 | Electromagnetic leakage red and black signal identification method based on information entropy characteristics |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107122738A (en) * | 2017-04-26 | 2017-09-01 | 成都蓝色起源科技有限公司 | Automatic Communication Signals Recognition based on deep learning model and its realize system |
CN108680796A (en) * | 2018-05-17 | 2018-10-19 | 集美大学 | Electromagnetic information leakage detecting system and method for computer display |
US20180314974A1 (en) * | 2017-04-27 | 2018-11-01 | Raytheon Company | Machine learning algorithm with binary pruning technique for automatic interpulse modulation recognition |
CN109145889A (en) * | 2018-11-02 | 2019-01-04 | 中国电子科技网络信息安全有限公司 | A kind of bright ciphertext Modulation recognition detection method carrying out blind estimate for wireless signal |
-
2019
- 2019-08-30 CN CN201910811601.3A patent/CN110728175A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107122738A (en) * | 2017-04-26 | 2017-09-01 | 成都蓝色起源科技有限公司 | Automatic Communication Signals Recognition based on deep learning model and its realize system |
US20180314974A1 (en) * | 2017-04-27 | 2018-11-01 | Raytheon Company | Machine learning algorithm with binary pruning technique for automatic interpulse modulation recognition |
CN108680796A (en) * | 2018-05-17 | 2018-10-19 | 集美大学 | Electromagnetic information leakage detecting system and method for computer display |
CN109145889A (en) * | 2018-11-02 | 2019-01-04 | 中国电子科技网络信息安全有限公司 | A kind of bright ciphertext Modulation recognition detection method carrying out blind estimate for wireless signal |
Non-Patent Citations (2)
Title |
---|
周长林 等: "基于PSO-SVM方法的电源线传导泄漏信号识别与还原", 《电子与信息学报》 * |
胡貌男 等: ""基于改进卷积神经网络的图像分类方法"", 《通信技术》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117332410A (en) * | 2023-10-25 | 2024-01-02 | 北京航空航天大学 | Electromagnetic leakage red and black signal identification method based on information entropy characteristics |
CN117332410B (en) * | 2023-10-25 | 2024-04-12 | 北京航空航天大学 | Electromagnetic leakage red and black signal identification method based on information entropy characteristics |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107124381B (en) | Automatic identification method for digital communication signal modulation mode | |
CN109657604B (en) | Satellite amplitude-phase signal identification and demodulation method and device based on cyclic neural network | |
CN110197127A (en) | Wireless signal detection and electromagnetic interference categorizing system and method based on deep learning | |
CN111832417B (en) | Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning | |
CN110099019B (en) | LoRa modulation signal detection method based on deep learning | |
CN106713190B (en) | MIMO transmitting antenna number blind estimation calculation method based on random matrix theory and characteristic threshold estimation | |
CN107819548B (en) | The symbol detection method of environment backscatter system based on Manchester's code | |
CN112949487A (en) | WiFi-based ship cab personnel number detection method and system | |
CN111817803A (en) | Frequency spectrum sensing method and system based on correlation coefficient and K-means clustering algorithm and computer readable storage medium | |
Smith et al. | A communication channel density estimating generative adversarial network | |
CN106230544A (en) | The monitoring identification of a kind of automobile remote-control interference signal and localization method | |
CN110728175A (en) | Electromagnetic signal red and black recognition device and method for wireless network | |
CN117713967A (en) | Intelligent frequency spectrum sensing anti-interference communication system for ad hoc network unmanned aerial vehicle | |
CN113259289B (en) | Single-channel aliasing signal modulation mode identification method based on residual error neural network | |
CN106357367A (en) | Non-coherent joint detecting and channel decoding method under massive MIMO (multiple input multiple output) system | |
CN104270210A (en) | Soft-decision spectrum sensing method based on compression non-reconstruction | |
Feng et al. | FCGCN: Feature Correlation Graph Convolution Network for Few-Shot Individual Identification | |
Wang et al. | A spatiotemporal multi-stream learning framework based on attention mechanism for automatic modulation recognition | |
Zhang et al. | Blind recognition of channel coding based on CNN-BLSTM | |
Chen et al. | Neuromorphic Split Computing with Wake-Up Radios: Architecture and Design via Digital Twinning | |
Wang et al. | Blind recognition method for non-cooperative communication signals based on aerial computing | |
Jiao et al. | Joint modulation and coding recognition using deep learning | |
Lijun et al. | A signal demodulation algorithm based on generative adversarial networks | |
CN111917674A (en) | Modulation identification method based on deep learning | |
Zheng et al. | Deep learning for cooperative radio signal classification |
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: 20200124 |