CN115296758A - Method, system, computer device and storage medium for identifying interference signal - Google Patents
Method, system, computer device and storage medium for identifying interference signal Download PDFInfo
- Publication number
- CN115296758A CN115296758A CN202210799581.4A CN202210799581A CN115296758A CN 115296758 A CN115296758 A CN 115296758A CN 202210799581 A CN202210799581 A CN 202210799581A CN 115296758 A CN115296758 A CN 115296758A
- Authority
- CN
- China
- Prior art keywords
- sequence data
- network
- data
- time
- frequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
- H04B17/345—Interference values
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Noise Elimination (AREA)
Abstract
The invention relates to the technical field of interference identification, and discloses an identification method, a system, computer equipment and a storage medium of an interference signal. The method comprises the following steps: IQ two-path sampling is carried out on the obtained baseband complex signals to obtain real part discrete sequence data and imaginary part discrete sequence data; performing spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain spectrum amplitude sequence data; performing time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data; and inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result. The invention not only exerts the advantage of processing corresponding data by a single network, but also fully utilizes the multi-dimensional information of the signal, and can improve the accuracy of identifying the interference signal.
Description
Technical Field
The present invention relates to the field of communication interference identification technologies, and in particular, to an interference signal identification method, system, computer device, and storage medium.
Background
With the development of communication technology, signal modulation modes are more and more, and the problems of self-interference and cross-interference are inevitably generated, so that when communication system equipment and network deployment are carried out, a certain communication anti-interference technology must be adopted to ensure the normal operation of wireless communication, and whether an interference signal can be accurately identified is the key for realizing the anti-interference technology.
At present, the main methods for identifying interference signals can be divided into a traditional method based on artificial design features and an autonomous feature extraction method based on network learning, however, the two methods have certain limitations, in the first method, because the features of the interference signals are often difficult to perceive or accurately express, and the feature expressions of different signals are different, so that the identification of the design features is difficult, corresponding feature parameters with definite physical meanings need to be designed aiming at specific signal types, the process has the problems of incomplete, non-independence and high complexity of features, and some important information can be ignored and lost, so that the final classification identification accuracy is not ideal; in the second method, the number of network layers needs to be increased to learn more details when the deep learning network faces more complex data, but this may cause the network to generate a situation of gradient explosion or gradient disappearance due to too deep layers during the backward propagation process of the network, which results in failure to update network parameters well, and thus the network may not learn normally.
Disclosure of Invention
In order to solve the above technical problems, an object of the present invention is to provide an interference signal identification method, system, computer device and storage medium, which performs comprehensive interference signal identification prediction by combining an LSTM network and a deep residual error network, processes data of different formats using different neural networks, fully utilizes network characteristics to extract feature information, and utilizes multi-dimensional information of signals to provide more feature details for the network, thereby improving accuracy of interference signal identification.
In a first aspect, the present invention provides a method for identifying an interference signal, where the method includes:
IQ two-path sampling is carried out on the obtained baseband complex signal to obtain real part discrete sequence data and imaginary part discrete sequence data;
performing spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain spectrum amplitude sequence data;
performing time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data;
and inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result, wherein the convolution cycle joint network comprises a long-short term memory network and a deep residual error network.
Further, the specific steps of inputting the real discrete sequence data, the imaginary discrete sequence data, the spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result include:
dividing the real part discrete sequence data, the imaginary part discrete sequence data and the frequency spectrum amplitude sequence data into three channels and inputting the three channels into a long-term and short-term memory network to extract data dynamic characteristics;
inputting the time-frequency amplitude image data into a deep residual error network, and extracting image static characteristics;
and integrating the data dynamic characteristics and the image static characteristics through a full connection layer to obtain characteristic data, and classifying and identifying the characteristic data to obtain the interference signal identification result.
Further, the spectral amplitude sequence data is calculated using the following formula:
in the formula, X (N) represents a baseband complex signal, N is a sequence length of the baseband complex signal, N is the number of samples of the baseband complex signal, and X (k) is converted spectrum amplitude sequence data.
Further, the time-frequency amplitude image data is calculated by adopting the following formula:
where m and k represent time and frequency variables, respectively, N represents the sequence length of each time segment, i.e., the length of the windowing, D represents the length of the window shift of the adjacent time segment, x (N) represents the baseband complex signal, W N A window function of length N points is represented.
Further, the network hidden layer of the long-short term memory network is a two-layer bidirectional long-short term memory, the deep layer residual error network includes 16 convolutional layers, a batch normalization layer and an active layer corresponding to each convolutional layer, and 6 residual error structures, two convolutional layers are spaced between the first convolutional layer and the first residual error structure, and two convolutional layers are sequentially spaced between each residual error structure.
In a second aspect, the present invention provides a system for identifying an interfering signal, the system comprising:
the two-path sampling module is used for carrying out IQ two-path sampling on the acquired baseband complex signal to obtain real part discrete sequence data and imaginary part discrete sequence data;
the frequency spectrum analysis module is used for carrying out frequency spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain frequency spectrum amplitude sequence data;
the time-frequency analysis module is used for carrying out time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data;
and the interference identification module is used for inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result, wherein the convolution cycle joint network comprises a long-short term memory network and a deep residual error network.
Further, the interference identification module further comprises:
the dynamic feature extraction module is used for dividing the real part discrete sequence data, the imaginary part discrete sequence data and the frequency spectrum amplitude sequence data into three channels and inputting the three channels into a long-term and short-term memory network to extract data dynamic features;
the static characteristic extraction module is used for inputting the time-frequency amplitude image data into a deep residual error network and extracting static characteristics of the image;
and the classification identification module is used for integrating the data dynamic characteristics and the image static characteristics through a full connection layer to obtain characteristic data, and performing classification identification on the characteristic data to obtain the interference signal identification result.
Further, the network hidden layer of the long-short term memory network is a two-layer bidirectional long-short term memory, the deep residual error network includes 16 convolutional layers, a batch normalization layer and an active layer corresponding to each convolutional layer, and 6 residual error structures, two convolutional layers are spaced between the first convolutional layer and the first residual error structure, and two convolutional layers are sequentially spaced between each residual error structure.
In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the foregoing method when executing the computer program.
In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the steps of the above method.
The invention provides an identification method, a system, computer equipment and a storage medium of interference signals. According to the method, the long-term and short-term memory network is used for processing the one-dimensional sequence data, the deep residual error network is used for processing the two-dimensional time-frequency image data, and the full-connection layer is used for predicting the interference signal after integrating the characteristics extracted by the single network.
Drawings
Fig. 1 is a schematic flowchart of an interference signal identification method according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of step S40 in FIG. 1;
FIG. 3 is a schematic diagram of a network architecture of a long term memory network according to an embodiment of the present invention;
FIG. 4 is a network architecture of a deep residual error network according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a network architecture of a convolution loop join network according to an embodiment of the present invention;
FIG. 6 is a table comparing accuracy data in an interference identification experiment of three neural networks according to an embodiment of the present invention;
FIG. 7 is a line comparison graph of accuracy versus drying ratio in an interference identification experiment for three neural networks in an embodiment of the present invention;
FIG. 8 is a diagram illustrating the generalization performance of the drying ratio of the deep residual network in the generalization performance experiment;
FIG. 9 is a graph showing the generalization performance of the drying ratio of the long-term and short-term memory network in a generalization performance experiment;
FIG. 10 is a graphical representation of the drying ratio generalization performance of the combined convolution and circulation network in a generalization performance experiment;
fig. 11 is a schematic structural diagram of an identification system of an interference signal according to an embodiment of the present invention;
fig. 12 is an internal structural view of a computer device in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a method for identifying interference information according to a first embodiment of the present invention includes steps S10 to S40:
and S10, IQ two-path sampling is carried out on the acquired baseband complex signal to obtain real discrete sequence data and imaginary discrete sequence data.
For the obtained communication signal, a series of signal processing is needed before the identification can be performed through a subsequent neural network, because the communication signal is a baseband complex signal and comprises a real part and an imaginary part, IQ two-path sampling is needed to obtain real part discrete sequence data and imaginary part discrete sequence data of the baseband complex signal, and if the length of the two-path sequence data is N, the two-path sequence data can jointly represent the baseband complex signal.
And S20, carrying out spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain spectrum amplitude sequence data.
And S30, carrying out time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data.
In order to facilitate the identification by using a neural network, besides performing two-way sampling, further processing needs to be performed on the baseband complex signal to extract frequency domain information and time-frequency diagram information of the signal, which specifically includes:
performing spectrum analysis on an input baseband complex signal by using discrete Fourier transform to obtain spectrum amplitude sequence data, wherein the transform formula is as follows:
in the formula, X (N) represents a baseband complex signal, N is a sequence length of the baseband complex signal, N is the number of samples of the baseband complex signal, and X (k) is converted spectrum amplitude sequence data.
Meanwhile, the baseband complex signal is subjected to time-frequency analysis by adopting short-time Fourier transform, so that two-dimensional time-frequency image data is obtained, wherein the adopted transform formula is as follows:
where m and k represent time and frequency variables, respectively, N represents the sequence length of each time segment, i.e. the length of the windowing, D represents the length of the window shift of the adjacent time segment, i.e. D = window length L-window overlap length, x (N) represents the baseband complex signal, W N A window function of length N points is represented.
In this embodiment, the number L of window length points added during the short-time fourier transform can be set to 128, the length D of window movement in an adjacent time period is set to 8, and a hamming window is selected for the window type, so that the finally obtained two-dimensional time-frequency amplitude image data has the specification length of (N-L)/D and the width of L, although the setting of the numerical value can be flexibly set according to the actual situation, and is not described in detail here.
And S40, inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result, wherein the convolution cycle joint network comprises a long-short term memory network and a deep residual error network.
Before the obtained data is processed in the next step, the neural network architecture adopted in the embodiment of the invention is explained, and because the existing common mode for identifying the interference signal is to directly extract and identify the characteristics of the one-dimensional sequence data or the two-dimensional time-frequency image data of the signal by using a single network, the prior art is not comprehensive in actually utilizing the sequence information of the signal, and the identification accuracy of the interference signal is influenced.
In order to solve the problem of low accuracy caused by the prior art, the method combines a long-short term memory network (LSTM) and a deep residual error network (ResNet), constructs a convolution cycle combination network, can simultaneously extract the characteristics of one-dimensional sequence data and two-dimensional image information of communication signals, and fully utilizes the multi-dimensional information of the signals to extract more detailed characteristics of the signals, so that the classification identification is more accurate, and the specific steps are as shown in figure 2:
step S401, dividing the real part discrete sequence data, the imaginary part discrete sequence data and the frequency spectrum amplitude sequence data into three channels and inputting the three channels into a long-term and short-term memory network, and extracting data dynamic characteristics.
Referring to fig. 3, the long-short term memory network LSTM used in this embodiment adopts two layers of bidirectional long-short term memory BILSTM as the hidden layer of the network, the BILSTM is formed by combining a front-to-back forward direction LSTM and a back-to-front reverse direction LSTM, and the network architecture in this embodiment can more fully mine context information of the sequence.
Based on the network architecture of the application, the acquired real part discrete sequence data, imaginary part discrete sequence data and spectral amplitude sequence data are divided into three channels to be input into the LSTM network, and through two-way sequence data extraction, time-related dynamic characteristics of the sequence data are automatically extracted by the LSTM network.
And S402, inputting the time-frequency amplitude image data into a deep residual error network, and extracting static features of the image.
For the obtained time-frequency amplitude image data, because the two-dimensional image data is obviously unsuitable for extracting feature information by using an LSTM network, in this embodiment, a deep residual error network ResNet is used to extract static features of an image, where a network architecture of ResNet is shown in fig. 4:
the main architecture of the deep residual error network in this embodiment includes 16 convolutional layers, each convolutional layer is sequentially connected to its corresponding batch normalization layer BN and active layer RELU, and from the first convolutional layer, one residual error structure is added every two convolutional layers, and 6 residual error structures are required to be added in total to construct an identity map.
Then, the obtained time-frequency amplitude image data is input into a built deep residual error network for static feature extraction, in the embodiment, the deep residual error network and the long-term and short-term memory network respectively comprise full connection layers with 50 nodes, that is, the ResNet network and the LSTM network respectively extract 50 static features and 50 dynamic features. It should be understood that the specific values related to the neural network architecture in the present embodiment are only preferred and not actually completely limiting.
And S403, integrating the data dynamic characteristics and the image static characteristics through a full connection layer to obtain characteristic data, and performing classification and identification on the characteristic data to obtain the interference signal identification result.
Referring to fig. 5, in the embodiment of the present invention, 50 static features and 50 dynamic features extracted by the ResNet network and the LSTM network respectively are integrated through a full connection layer, which is equivalent to 100 features being extracted altogether, and then final classification and identification are performed through a classifier.
In order to verify the effect of the convolution cycle joint network of this embodiment on interference signal identification, a simulation experiment is performed on an MATLAB platform, the convolution cycle joint network of this embodiment is compared with a deep residual error network and an LSTM network, a total of 36 signal types including six single interference types and aliasing combinations of the single types are set in the simulation experiment, and comparison data of identification accuracy obtained after identification is performed by three types of neural networks is shown in fig. 6 and 7.
As can be seen from the identification accuracy comparison results of fig. 6 and fig. 7, the interference identification result of the convolution cycle joint network adopted in the embodiment is significantly improved compared with that of a deep residual network, and compared with an LSTM network, although the identification rate is slightly reduced under the condition of a low drying ratio, since the influence of the time-frequency image in the residual network branch on the feature extraction is large when the noise is too large, the whole joint network is influenced, but the identification accuracy is improved by 1% to 2% when the drying ratio is greater than 5dB, which indicates that the convolution cycle joint network can integrate the advantages of a single network, fully utilize the information of signal dimensions, and improve the performance to a certain extent compared with the single network.
Meanwhile, the drying ratio of the convolution cycle joint network of the embodiment is obviously improved compared with the generalization performance of a single deep residual error network and an LSTM network, one definition of the generalization performance of the neural network means that the trained network has good generalization performance if the untrained data is not represented in the training of the network, and in order to better illustrate the comparison, three training data with different interference signal drying ratio distribution conditions are respectively set:
(1) the data range of the signal drying ratio is-10 dB to 20dB, the drying ratio interval of each group of data is 2dB, namely, 16 groups of training data with different drying ratios are total;
(2) the data range of the signal drying ratio is-10 dB to 20dB, the drying ratio interval of each group of data is 3dB, namely 11 groups of training data with different drying ratios are total;
(3) the data range of the signal drying ratio is-10 dB to 20dB, the drying ratio interval of each group of data is 5dB, namely, 7 groups of training data with different drying ratios are totally obtained.
And the distribution condition (1) is simultaneously used as test data to serve as a control group, three different neural networks are used for testing each group of data, and the obtained comparison results of the generalization performance of the drying ratio of the different networks are respectively shown in fig. 8, fig. 9 and fig. 10.
Through comparison of the results of fig. 8, fig. 9 and fig. 10, it can be clearly seen that the convolution and loop combination network has an excellent drying ratio generalization capability, and compared with a single deep residual error network and an LSTM network, the drying ratio generalization capability of the convolution and loop combination network has a very considerable improvement, so as to further prove that the features extracted by the convolution and loop combination network according to the embodiment of the present invention are more detailed, regular and representative than those extracted by the single network, and thus the accuracy of the interference signal identification can be further improved.
Referring to fig. 11, based on the same inventive concept, a system for identifying an interference signal according to a second embodiment of the present invention includes:
the two-path sampling module 10 is configured to perform IQ two-path sampling on the acquired baseband complex signal to obtain real discrete sequence data and imaginary discrete sequence data;
the spectrum analysis module 20 is configured to perform spectrum analysis on the baseband complex signal through discrete fourier transform to obtain spectrum amplitude sequence data;
the time-frequency analysis module 30 is configured to perform time-frequency analysis on the baseband complex signal through short-time fourier transform, so as to obtain time-frequency amplitude image data;
and an interference identification module 40, configured to input the real discrete sequence data, the imaginary discrete sequence data, the spectrum amplitude sequence data, and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification, so as to obtain an interference signal identification result, where the convolution cycle joint network includes a long-term and short-term memory network and a deep residual error network.
The technical features and technical effects of the system for identifying an interference signal according to the embodiment of the present invention are the same as those of the method according to the embodiment of the present invention, and are not described herein again. The modules in the above interference signal identification system may be implemented wholly or partially by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
Referring to fig. 12, in an embodiment, an internal structure of a computer device may specifically be a terminal or a server. The computer apparatus includes a processor, a memory, a network interface, a display, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program to run on the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of identifying an interfering signal. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on a shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those of ordinary skill in the art that the architecture shown in FIG. 12 is merely a block diagram of a portion of the architecture associated with the present application, and is not intended to limit the computing devices to which the present application may be applied, and that a particular computing device may include more or fewer components than shown in its own right, or may combine certain components, or have the same arrangement of components.
In addition, an embodiment of the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method when executing the computer program.
Furthermore, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the steps of the foregoing method.
In summary, in the method, the system, the computer device, and the storage medium for identifying an interference signal provided by this embodiment, IQ two-path sampling is performed on an acquired baseband complex signal to obtain real discrete sequence data and imaginary discrete sequence data; performing spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain spectrum amplitude sequence data; performing time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data; and inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude value sequence data and the time-frequency amplitude value image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result. Compared with the traditional interference signal identification method, the method only uses a single network to extract and identify the one-dimensional characteristics or the two-dimensional image characteristics of the sequence data, so that the problem of low accuracy of insufficient identification of information utilization is caused.
The embodiments in this specification are described in a progressive manner, and all the same or similar parts of the embodiments are directly referred to each other, and each embodiment is described with emphasis on differences from other embodiments. In particular, as for the system embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for relevant points. It should be noted that, the technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present description should be considered as being included in the present specification.
The above-mentioned embodiments only express several preferred embodiments of the present application, and the description thereof is specific and detailed, but not to be understood as limiting the scope of the invention. It should be noted that, for those skilled in the art, various modifications and substitutions can be made without departing from the technical principle of the present invention, and these should be construed as the protection scope of the present application. Therefore, the protection scope of the present patent application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for identifying an interfering signal, comprising:
IQ two-path sampling is carried out on the obtained baseband complex signal to obtain real part discrete sequence data and imaginary part discrete sequence data;
performing spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain spectrum amplitude sequence data;
performing time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data;
and inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result, wherein the convolution cycle joint network comprises a long-short term memory network and a deep residual error network.
2. The method according to claim 1, wherein the step of inputting the real discrete sequence data, the imaginary discrete sequence data, the spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain the interference signal identification result comprises:
dividing the real part discrete sequence data, the imaginary part discrete sequence data and the frequency spectrum amplitude sequence data into three channels and inputting the three channels into a long-term and short-term memory network to extract data dynamic characteristics;
inputting the time-frequency amplitude image data into a deep residual error network, and extracting image static characteristics;
and integrating the data dynamic characteristics and the image static characteristics through a full connection layer to obtain characteristic data, and classifying and identifying the characteristic data to obtain the interference signal identification result.
3. The method of identifying an interfering signal according to claim 1, wherein the spectral amplitude sequence data is calculated using the following formula:
in the formula, X (N) represents a baseband complex signal, N is a sequence length of the baseband complex signal, N is the number of samples of the baseband complex signal, and X (k) is converted spectrum amplitude sequence data.
4. The method according to claim 1, wherein the time-frequency amplitude image data is calculated by using the following formula:
where m and k represent time and frequency variables, respectively, N represents the sequence length of each time segment, i.e., the length of the windowing, D represents the length of the window shift of the adjacent time segment, x (N) represents the baseband complex signal, W N A window function of length N points is represented.
5. The method according to claim 1, wherein the network hidden layer of the long-short term memory network is a two-layer bi-directional long-short term memory, the deep residual network comprises 16 convolutional layers, a batch normalization layer and an active layer corresponding to each convolutional layer, and 6 residual structures, the first convolutional layer and the first residual structure are separated by two convolutional layers, and each residual structure is sequentially separated by two convolutional layers.
6. An identification system for interfering signals, comprising:
the two-path sampling module is used for carrying out IQ two-path sampling on the acquired baseband complex signal to obtain real part discrete sequence data and imaginary part discrete sequence data;
the frequency spectrum analysis module is used for carrying out frequency spectrum analysis on the baseband complex signal through discrete Fourier transform to obtain frequency spectrum amplitude sequence data;
the time-frequency analysis module is used for carrying out time-frequency analysis on the baseband complex signal through short-time Fourier transform to obtain time-frequency amplitude image data;
and the interference identification module is used for inputting the real part discrete sequence data, the imaginary part discrete sequence data, the frequency spectrum amplitude sequence data and the time-frequency amplitude image data into a trained convolution cycle joint network for signal identification to obtain an interference signal identification result, wherein the convolution cycle joint network comprises a long-short term memory network and a deep residual error network.
7. The system for identifying an interfering signal according to claim 6, wherein the interference identification module further comprises:
the dynamic feature extraction module is used for dividing the real part discrete sequence data, the imaginary part discrete sequence data and the frequency spectrum amplitude sequence data into three channels and inputting the three channels into a long-term and short-term memory network to extract data dynamic features;
the static characteristic extraction module is used for inputting the time-frequency amplitude image data into a deep residual error network and extracting static characteristics of the image;
and the classification identification module is used for integrating the data dynamic characteristics and the image static characteristics through a full connection layer to obtain characteristic data, and performing classification identification on the characteristic data to obtain the interference signal identification result.
8. The system of claim 6, wherein the network hidden layer of the long-short term memory network is a two-layer bidirectional long-short term memory, and the deep residual network comprises 16 convolutional layers, a batch normalization layer and an active layer corresponding to each convolutional layer, and 6 residual structures, wherein two convolutional layers are spaced between a first convolutional layer and a first residual structure, and two convolutional layers are sequentially spaced between each residual structure.
9. A computer arrangement comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 5 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 5.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210799581.4A CN115296758B (en) | 2022-07-07 | 2022-07-07 | Method, system, computer equipment and storage medium for identifying interference signals |
| ZA2023/05874A ZA202305874B (en) | 2022-07-07 | 2023-06-01 | Interference signal identification method and system, computer device, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210799581.4A CN115296758B (en) | 2022-07-07 | 2022-07-07 | Method, system, computer equipment and storage medium for identifying interference signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115296758A true CN115296758A (en) | 2022-11-04 |
| CN115296758B CN115296758B (en) | 2023-08-11 |
Family
ID=83823230
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210799581.4A Active CN115296758B (en) | 2022-07-07 | 2022-07-07 | Method, system, computer equipment and storage medium for identifying interference signals |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115296758B (en) |
| ZA (1) | ZA202305874B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108616470A (en) * | 2018-03-26 | 2018-10-02 | 天津大学 | Modulation Signals Recognition method based on convolutional neural networks |
| CN111695417A (en) * | 2020-04-30 | 2020-09-22 | 中国人民解放军空军工程大学 | Signal modulation pattern recognition method |
| CN112566174A (en) * | 2020-12-02 | 2021-03-26 | 中国电子科技集团公司第五十二研究所 | Abnormal I/Q signal identification method and system based on deep learning |
| US11222264B1 (en) * | 2020-08-28 | 2022-01-11 | Naval Aviation University of PLA | Method and device for recognizing space-frequency block code |
| CN114239657A (en) * | 2021-12-18 | 2022-03-25 | 中国科学技术大学 | Time sequence signal identification method based on complex value interference neural network |
-
2022
- 2022-07-07 CN CN202210799581.4A patent/CN115296758B/en active Active
-
2023
- 2023-06-01 ZA ZA2023/05874A patent/ZA202305874B/en unknown
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108616470A (en) * | 2018-03-26 | 2018-10-02 | 天津大学 | Modulation Signals Recognition method based on convolutional neural networks |
| CN111695417A (en) * | 2020-04-30 | 2020-09-22 | 中国人民解放军空军工程大学 | Signal modulation pattern recognition method |
| US11222264B1 (en) * | 2020-08-28 | 2022-01-11 | Naval Aviation University of PLA | Method and device for recognizing space-frequency block code |
| CN112566174A (en) * | 2020-12-02 | 2021-03-26 | 中国电子科技集团公司第五十二研究所 | Abnormal I/Q signal identification method and system based on deep learning |
| CN114239657A (en) * | 2021-12-18 | 2022-03-25 | 中国科学技术大学 | Time sequence signal identification method based on complex value interference neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| ZA202305874B (en) | 2023-12-20 |
| CN115296758B (en) | 2023-08-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Margoosian et al. | An accurate kernelized energy detection in Gaussian and non-Gaussian/impulsive noises | |
| CN109065027B (en) | Voice distinguishing model training method and device, computer equipment and storage medium | |
| CN110222748B (en) | OFDM radar signal identification method based on 1D-CNN multi-domain feature fusion | |
| CN108965001B (en) | Method and device for evaluating vehicle message data model | |
| CN109165510A (en) | Android malicious application detection method based on binary channels convolutional neural networks | |
| CN110702411A (en) | Residual error network rolling bearing fault diagnosis method based on time-frequency analysis | |
| US20210304889A1 (en) | A two-phase disease diagnosis system and method thereof | |
| CN112257263B (en) | Equipment residual life prediction system based on self-attention mechanism | |
| CN110138480A (en) | Method and system, frequency spectrum sensing method and the system of training frequency spectrum perception model | |
| CN112784519A (en) | PCB routing parameter setting device and method and storage medium | |
| CN114818828B (en) | Training method of radar interference perception model and radar interference signal identification method | |
| CN109376068A (en) | Software reliability prediction method, apparatus, computer equipment and storage medium | |
| CN112947080B (en) | Scene parameter transformation-based intelligent decision model performance evaluation system | |
| CN115296758A (en) | Method, system, computer device and storage medium for identifying interference signal | |
| CN117574202A (en) | Radar antenna scanning mode identification method based on multi-time sequence feature fusion | |
| CN111161789B (en) | Analysis method and device for key areas of model prediction | |
| CN112926054A (en) | Malicious file detection method, device, equipment and storage medium | |
| CN113343796B (en) | Knowledge distillation-based radar signal modulation mode identification method | |
| CN105117330A (en) | CNN (Convolutional Neural Network) code testing method and apparatus | |
| KR20220046467A (en) | Method for automatically pruning search space using machine learning in symbolic execution | |
| CN113269303A (en) | Data processing method and data processing device for deep learning inference framework | |
| CN114598403B (en) | Data link broadband noise electromagnetic signal interference prediction method and system | |
| CN113159100B (en) | Circuit fault diagnosis method, circuit fault diagnosis device, electronic equipment and storage medium | |
| Li et al. | Power Analysis Attack Based on Lightweight Convolutional Neural Network | |
| CN116680592A (en) | Training method and device for unmanned aerial vehicle individual recognition model and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |