Cross-receiver signal type identification method, device, equipment, medium and product
By matching and modifying the signals collected by different receivers, we ensure that they meet the independent and same distribution conditions. We also use deep learning technology to solve the problem of degradation of identification performance during cross-receiver signal type recognition, and achieve efficient and accurate signal type recognition.
Patent Information
- Application Number
- CN202510089740.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-21
- Publication Date
- 2025-05-13
AI Technical Summary
Due to the differences in equipment parameters between different receivers, the data characteristics of the collected signals are inconsistent, which in turn causes the recognition performance of the signal recognition model trained on a receiver to degrade when identifying signal types across receivers.
By preprocessing signals collected by at least two source receivers, a training data set is constructed, and a neural network model is trained based on the training data set, a signal recognition model is obtained. Then, the signal collected by the target receiver and the signal collected by the source receiver is matched to determine the feature difference information, and the characteristics of the signal collected by the target receiver are modified based on this information to ensure that it meets the independent and same distribution conditions as the training data as possible, and finally the modified signal is input to the signal type recognition model for identification.
By solving the problem of receiver parameter differences and combining time-frequency transformation and deep learning technology, the performance of signal type recognition across receivers is significantly improved, the recognition error is reduced, the limitations of traditional signal processing methods are reduced, and the robustness and accuracy of signal type recognition are enhanced.
Smart Images
Figure CN119989053A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of machine learning technology, and in particular to a method, apparatus, device, storage medium and computer product for cross-receiver signal type recognition. Background Art
[0002] With the rapid development of a new generation of artificial intelligence technology represented by deep learning, a large number of research institutions and engineering and technical personnel have introduced deep learning into the field of communications, especially signal recognition.
[0003] In the related art, when identifying the type of signal collected by a signal receiver, a signal recognition model trained by the characteristics of the signal collected by the signal receiver can be used for recognition.
[0004] However, deep learning can be effective only if the training data and test data are independent and identically distributed. Due to the differences in equipment parameters between different receivers, the data features of the signals collected by different receivers are different, which may not meet the independent and identically distributed conditions, resulting in a decrease in the recognition performance of the signal recognition model trained based on the data collected by one receiver when recognizing the data collected by other receivers (that is, when identifying the signal type across receivers). Summary of the invention
[0005] The main purpose of this application is to provide a method, device, equipment, storage medium and computer product for cross-receiver signal type identification, aiming to solve the technical problem of related technologies in which the identification performance is reduced due to inconsistent signal data characteristics collected by different receivers when identifying cross-receiver signal types.
[0006] To achieve the above object, the present application proposes a method for identifying the type of a signal across receivers, the method comprising:
[0007] The signal preprocessing stage performs signal preprocessing on the signals collected by at least two source receivers to construct a training data set for the signal recognition model;
[0008] In the model training phase, the neural network model is trained based on the training data set to obtain a signal recognition model;
[0009] The feature matching stage matches the signal features of the target receiver's acquisition signal with the source receiver's acquisition signal to determine the feature difference information between the target receiver's acquisition signal and the source receiver's acquisition signal;
[0010] Based on the characteristic difference information and the signal characteristics of the signal collected by the source receiver, the target receiver collection signal is modified to obtain a modified target receiver collection signal;
[0011] In the signal type identification stage, the modified target receiver acquisition signal is input into the signal type identification model to obtain the signal type identification result.
[0012] In one embodiment, the characteristic information of the collected signal includes at least the sampling frequency information of the target receiver and the sampling frequency information of the source receiver, and the characteristic matching link includes the steps of:
[0013] Determine, from the source receiver sampling frequency information set, source receiver sampling frequency information that is closest to the target receiver sampling frequency information;
[0014] Based on the sampling frequency information of the closest source receiver, the sampling frequency of the target receiver to acquire the signal is modified.
[0015] In one embodiment, the step of modifying the sampling frequency of the target receiver to acquire the signal based on the closest sampling frequency information of the source receiver includes:
[0016] If the sampling frequency information of the closest source receiver is greater than the sampling frequency of the target receiver acquisition signal, then the target receiver acquisition signal is upsampled;
[0017] If the sampling frequency information of the closest source receiver is less than the sampling frequency of the target receiver acquisition signal, the target receiver acquisition signal is downsampled.
[0018] In one embodiment, the signal preprocessing step includes the steps of:
[0019] Acquiring source receiver acquisition signals through source receivers with different sampling rate gears; wherein, for each sampling rate gear, acquiring at least two source receiver acquisition signals;
[0020] Perform time-frequency conversion processing on the collected signals of each source receiver to obtain a time-frequency conversion matrix corresponding to the number of sampling rate gears;
[0021] Based on the time-frequency transformation matrix, a training data set corresponding to the number of sampling rate gears is constructed.
[0022] In one embodiment, the model training process includes the following steps:
[0023] For each training data set, input the training data set into the neural network model to obtain the model weight file corresponding to the training data set;
[0024] Based on all model weight files and neural network models, a signal recognition model is built.
[0025] In one embodiment, the signal type identification step includes the steps of:
[0026] inputting the modified target receiver acquisition signal into a signal recognition model;
[0027] Signal type recognition processing is performed based on the model weight file corresponding to the sampling frequency of the modified target receiver to acquire the signal, so as to obtain a signal type recognition result.
[0028] In a second aspect, to achieve the above-mentioned purpose, the present application further provides a device for identifying a type of a signal across receivers, the device comprising:
[0029] A signal preprocessing module, used to perform signal preprocessing on signals collected by at least two source receivers to construct a training data set for a signal recognition model;
[0030] A model training module is used to train a neural network model based on a training data set to obtain a signal recognition model;
[0031] A feature matching module is used to match the signal characteristics of the target receiver acquisition signal with the source receiver acquisition signal to determine the feature difference information between the target receiver acquisition signal and the source receiver acquisition signal;
[0032] Based on the characteristic difference information and the signal characteristics of the signal collected by the source receiver, the target receiver collection signal is modified to obtain a modified target receiver collection signal;
[0033] The signal type recognition module is used to input the modified target receiver acquisition signal into the signal type recognition model to obtain the signal type recognition result.
[0034] On the third aspect, in order to achieve the above-mentioned purpose, the present application continues to provide a cross-receiver signal type identification device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the above-mentioned cross-receiver signal type identification method.
[0035] In a fourth aspect, in order to achieve the above-mentioned purpose, the present application continues to provide a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the above-mentioned cross-receiver signal type identification method are implemented.
[0036] In a fifth aspect, to achieve the above-mentioned purpose, the present application continues to provide a computer program product, the computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned cross-receiver signal type identification method are implemented.
[0037] One or more technical solutions proposed in this application have at least the following technical effects:
[0038] By solving the problem of parameter differences between different receivers and combining time-frequency transformation with deep learning technology, the performance of cross-receiver low signal-to-noise ratio signal type recognition is effectively improved. Specifically, this application first matches the difference in the acquired signals of the source receiver and the target receiver, and modifies the acquired signal characteristics of the target receiver according to the acquired signal characteristics of the source and target receivers, ensuring that the training data and the test data meet the basic conditions of independent distribution as much as possible. Then, by utilizing the powerful pattern recognition capabilities of deep learning, the signal type can be accurately classified even under low signal-to-noise ratio conditions. This method not only reduces the recognition error caused by differences in receiver hardware, but also reduces the limitations of traditional signal processing methods in low signal-to-noise ratio environments, significantly enhancing the robustness and accuracy of signal type recognition. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0041] Figure 1 Schematic diagram of the flow of a method for identifying a signal type across receivers in one embodiment of the present application.
[0042] Figure 2 It is a flowchart of a method for identifying a signal type across receivers in a specific implementation example of the present application.
[0043] Figure 3 This is a schematic diagram of the structure of the cross-receiver signal type identification device of the present application.
[0044] Figure 4 This is a schematic diagram of the structure of the cross-receiver signal type identification device of this application.
[0045] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0046] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.
[0047] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.
[0048] In the related art, when identifying the type of signal collected by a signal receiver, a signal recognition model trained with historical data collected by the signal receiver can be used for recognition.
[0049] However, deep learning can be effective only if the training data and test data are independent and identically distributed. Due to the differences in equipment parameters between different receivers, the data features of the signals collected by different receivers are different, which may not meet the independent and identically distributed conditions, resulting in a decrease in the recognition performance of the signal recognition model trained based on the data collected by one receiver when recognizing the data collected by other receivers (that is, when identifying the signal type across receivers).
[0050] In response to the problem of decreased recognition performance when identifying signal types across signal receivers in the related art, the present application provides a solution. By solving the problem of parameter differences between different receivers and combining time-frequency transformation with deep learning technology, the performance of low signal-to-noise ratio signal type recognition across receivers is effectively improved.
[0051] Based on this, the embodiment of the present application provides a method for identifying the type of a signal across receivers, referring to Figure 1 , Figure 1 This is a flowchart of a first embodiment of a method for identifying signal types across receivers of the present application.
[0052] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a cross-receiver signal type identification device, etc. The following takes the cross-receiver signal type identification device as an example to illustrate this embodiment and the following embodiments.
[0053] In this embodiment, the cross-receiver signal type identification method includes steps S10 to S50:
[0054] Step S10, signal preprocessing stage, performs signal preprocessing on signals collected by at least two source receivers to construct a training data set for a signal recognition model.
[0055] Step S20, model training phase, trains the neural network model based on the training data set to obtain a signal recognition model.
[0056] Step S30, feature matching step, performs signal feature matching on the target receiver acquisition signal and the source receiver acquisition signal to determine feature difference information between the target receiver acquisition signal and the source receiver acquisition signal.
[0057] Step S40, based on the characteristic difference information and the signal characteristics of the signal collected by the source receiver, modify the signal collected by the target receiver to obtain a modified signal collected by the target receiver.
[0058] Step S50, signal type identification step, inputs the modified target receiver acquisition signal into the signal type identification model to obtain a signal type identification result.
[0059] It should be noted that the source receiver can collect various types of signals. In order to ensure the recognition effect of the signal recognition model, it is necessary to construct a training data set based on the signals collected by various types of source receivers. The source receiver can be one or more.
[0060] It should also be noted that the signal recognition model is used to identify the signal collected by the target receiver. It can be understood that the signal type collected by the target receiver is generally included in the various signal types that can be collected by the source receiver. The signal characteristics of the signals collected by the source receiver and the target receiver can be sampling frequency, modulation mode, modulation rate, etc. The characteristic difference information is the difference in the collection method between the source receiver and the target receiver.
[0061] Specifically, in the signal preprocessing phase, different types of source receivers may collect signals with different transmission modes or frequency characteristics. The main purpose of signal preprocessing is to clean and prepare these signals for subsequent model training. In the model training phase, the preprocessed signals will be used to build a training data set, which is the basis of the signal recognition model. This stage uses the preprocessed training data set to train a neural network model. Each signal sample in the training data set contains its corresponding label (signal type). The neural network continuously adjusts parameters through the back-propagation algorithm to reduce the error between the predicted results and the true label, thereby obtaining accurate signal recognition capabilities. The model in the training process can be a convolutional neural network (CNN) or a recurrent neural network (RNN), depending on the characteristics of the signal. For example, a time series signal can use an RNN, while a spectrogram may be more suitable for a CNN. The ultimate goal is to build an accurate signal recognition model.
[0062] Due to the differences in sampling frequency, modulation mode or modulation rate between different receivers, the characteristics of the signals collected by the target receiver and the source receiver may also be different. In the feature matching link, the signal collected by the target receiver will be feature matched with the signal collected by the source receiver to identify the feature difference information between the signal collected by the target receiver and the signal collected by the source receiver. Through the feature difference information, the signal collected by the target receiver can be modified to meet the independent and identically distributed conditions, providing a basis for cross-receiver signal recognition.
[0063] The task of the signal type recognition link is to input the modified target receiver signal into the signal type recognition model to obtain the signal type recognition result.
[0064] It can be understood that by performing differential matching on the difference of the collected signals of the source receiver and the target receiver, and modifying the collected signal characteristics of the target receiver according to the collected signal characteristics of the source and target receivers, it is ensured that the training data and the test data meet the basic conditions of independent distribution as much as possible. Then, by using the powerful pattern recognition ability of deep learning, the signal type can be accurately classified even under low signal-to-noise ratio conditions. This embodiment not only reduces the recognition error caused by the difference in receiver hardware, but also reduces the limitations of traditional signal processing methods in low signal-to-noise ratio environments, and significantly enhances the robustness and accuracy of signal type recognition.
[0065] In a feasible implementation manner, the signal preprocessing step includes steps S11 to S13:
[0066] Step S11, obtaining source receiver acquisition signals through source receivers with different sampling rate gears; wherein, for each sampling rate gear, at least two source receiver acquisition signals are obtained.
[0067] Step S12, performing time-frequency conversion processing on the collected signals of each source receiver to obtain a time-frequency conversion matrix corresponding to the number of sampling rate gears.
[0068] Step S13: constructing a training data set corresponding to the number of sampling rate gears based on the time-frequency transformation matrix.
[0069] In this embodiment, the signal can be acquired through a source receiver including different sampling rate gears, and at least two signals are collected for each sampling rate gear. This method can provide basic signal data, covering a variety of sampling rates and signal characteristics, laying the foundation for subsequent training data construction to ensure that the signal recognition model has good recognition capabilities at different sampling rates and signal types.
[0070] Perform time-frequency transformation on each collected signal to generate a high-dimensional representation (time-frequency matrix) that describes the dynamic characteristics of the signal, providing structured features for subsequent model training. Specifically, the signal at each sampling rate level corresponds to a training data set, that is, each training data set covers multiple signal types at the same sampling rate, ensuring that the model has good generalization capabilities and is adaptable to different practical application scenarios. The model training process also includes steps S21 to S22:
[0071] For each training data set, input the training data set into the neural network model to obtain the model weight file corresponding to the training data set;
[0072] Based on all model weight files and neural network models, a signal recognition model is built.
[0073] The signal type identification process includes steps S51 to S52:
[0074] Step S51, inputting the modified target receiver acquisition signal to the signal recognition model.
[0075] Step S52: performing signal type recognition processing based on the model weight file corresponding to the modified sampling frequency of the target receiver to acquire the signal, and obtaining a signal type recognition result.
[0076] Specifically, each training data set is input into the neural network model in turn for training. In the neural network, the weights of the connection parameters of each layer of the neural network are adjusted according to the input training data set to minimize the gap between the predicted value and the true value, and the model weight file corresponding to each training data set is obtained. Each sampling rate gear corresponds to a training data set, and the model weight file corresponding to the sampling frequency of the target receiver data acquisition data can be selected to identify the signal type.
[0077] In this embodiment, the characteristic information of the collected signal includes at least the sampling frequency information of the target receiver and the sampling frequency information of the source receiver, and the characteristic matching process includes steps S31 to S41:
[0078] Step S31: determine the source receiver sampling frequency information closest to the target receiver sampling frequency information from the source receiver sampling frequency information set.
[0079] Step S41, based on the sampling frequency information of the closest source receiver, modify the sampling frequency of the target receiver to collect the signal.
[0080] Wherein, step S32 also includes steps A10 to A20:
[0081] Step A10: if the sampling frequency information of the closest source receiver is greater than the sampling frequency of the target receiver acquisition signal, then upsampling processing is performed on the target receiver acquisition signal.
[0082] In step A20, if the sampling frequency information of the closest source receiver is less than the sampling frequency of the target receiver acquisition signal, down-sampling processing is performed on the target receiver acquisition signal.
[0083] Specifically, in the feature matching link, the sampling frequency information set of the source receiver can be constructed based on the sampling frequency information corresponding to each training data set according to the training signal recognition model. When matching the sampling rate feature, first confirm the sampling frequency information of the target receiver acquisition signal to be identified, that is, the sampling rate gear corresponding to the target receiver when the signal is acquired, and determine the source receiver sampling frequency information closest to the target receiver sampling frequency information from the source receiver sampling frequency information set. Based on the closest source receiver sampling frequency information, modify the sampling frequency of the target receiver acquisition signal.
[0084] Specifically, if the sampling frequency information of the source receiver is consistent with the sampling frequency information of the target receiver, the target receiver acquisition signal can be directly used for signal type identification; if the source receiver sampling frequency is greater than the target receiver sampling frequency, the target receiver acquisition signal is upsampled; if the source receiver sampling frequency is less than the target receiver sampling frequency, the target receiver acquisition signal is downsampled, and the target receiver acquisition signal after upsampling or downsampling is used for signal type identification.
[0085] This implementation constructs a signal recognition model with high generalization ability and precise recognition performance through multi-sampling rate signal acquisition, time-frequency transformation processing and targeted neural network training. At the same time, the signal with different sampling rates is up-sampled / down-sampled through the feature matching link to ensure the consistency of data distribution, thereby achieving efficient and accurate cross-receiver signal type recognition, and significantly enhancing the adaptability and reliability of the signal recognition model in various application scenarios.
[0086] For example, to help understand the implementation process of the cross-receiver signal type identification method obtained in this embodiment, please refer to Figure 2 , Figure 2 A brief structural diagram of the cross-receiver signal type identification method in this implementation example is provided, specifically:
[0087] In this example, the various steps of cross-receiver signal type identification are handled as follows:
[0088] In the signal preprocessing stage, the in-phase and quadrature signal data (IQ data) are collected through the source receiver. The source receiver has i (i≤4) sampling rate gears, and the sampling rates are Fs s1 ,...,Fs si For each sampling rate gear, m (m≥2) types of signals are collected, and the signal-to-noise ratio SNR of the collected signals is in the range of 5dB≤SNR<8dB.
[0089] Perform time-frequency transformation on the IQ data of each sampling rate gear to obtain the rows H and columns W of the IQ data as the time-frequency matrix M.new =M[H*h1:H*h2,W*w1:W*w2] extracts the new time-frequency matrix M new , where h1(1≤h1≤h2) and h2(h1≤h2≤H) are row decimation factors, and w1(1≤w1≤w2) and w2(w1≤w2≤W) are column decimation factors. There are a total of i time-frequency matrices, that is, each sampling rate gear corresponds to a time-frequency matrix.
[0090] According to M norm =(M new -M newmin ) / (M newmax -M newmin ) to normalize each time-frequency matrix, where M newmin M new The minimum value, M newmax M new The maximum value of .
[0091] One training data set is constructed for each source receiver sampling rate gear. In each training data set, each type of signal corresponds to x (x ≥ 5) normalized time-frequency matrices in the low signal-to-noise ratio range. Each training data set has m*x time-frequency matrices, and there are i training data sets in total.
[0092] In the model training phase, a neural network model based on deep learning is used to train i training data sets respectively. After the training is completed, i weight files of the signal type recognition model are obtained. Each weight file corresponds to a sampling rate gear of a source signal receiver. The signal type recognition model is constructed based on all weight files.
[0093] Sampling rate matching stage: construct the source receiver sampling frequency information set Fs according to the sampling rate gear of the source receiver s1 ,...,Fs si , the signal-to-noise ratio obtained from the target receiver is 5dB≤SNR<8dB, and the sampling rate is Fs t If the source receiver sampling rate information set contains the same IQ data as the target receiver sampling rate Fs, t If the source receiver sampling rate information completely matches, the signal type identification link is entered. Otherwise, a source receiver sampling rate information that is closest to the source receiver sampling rate set is selected as the target receiver IQ data. The sampling rate that needs to be changed is recorded as Fs sk , and Fs sk ∈Fs s1 ,...,Fs si (i≤4).
[0094] Change the sampling rate of the target receiver IQ data. If Fs t >Fs skThen the target receiver IQ data is downsampled so that the sampling rate after downsampling is Fs t ′=Fs sk Otherwise, the target receiver IQ data is upsampled so that the upsampled sampling rate Fs t ″=Fs sk .
[0095] In the signal type identification phase, the IQ data collected by the target receiver is transformed into time-frequency data and the time-frequency matrix is normalized. The sampling rate information Fs of the source receiver is loaded from the signal type identification model. sk The corresponding weight file performs signal type recognition on the IQ data collected by the target receiver and outputs the recognition result of the signal type.
[0096] It can be understood that in this example, by solving the distribution of collected data caused by differences in different receivers, the need to retrain or fine-tune the signal type recognition model due to the replacement of different receivers is avoided. The time-frequency transformation is used to transform the collected signal from the time domain to the time-frequency domain. The deep learning algorithm can accurately identify the signal type when the signal-to-noise ratio of the collected signal is 5dB≤SNR<8dB, avoiding the problem of inaccurate signal parameter estimation under low signal-to-noise ratio conditions resulting in a decrease in signal type recognition performance. At the same time, after testing, although this example solves the problem of signal type recognition under low signal-to-noise ratios of 5dB≤SNR<8dB, it can still accurately identify the signal type under the condition of SNR≥8dB.
[0097] The present application also provides a device for identifying signal types across receivers. Figure 3 , the cross-receiver signal type identification device comprises:
[0098] The signal preprocessing module 10 is used to perform signal preprocessing on signals collected by at least two source receivers to construct a training data set for a signal recognition model.
[0099] The model training module 20 is used to train the neural network model based on the training data set to obtain a signal recognition model.
[0100] The feature matching module 30 is used to perform signal feature matching on the target receiver acquisition signal and the source receiver acquisition signal, and determine feature difference information between the target receiver acquisition signal and the source receiver acquisition signal.
[0101] Based on the characteristic difference information and the signal characteristics of the signal collected by the source receiver, the signal collected by the target receiver is modified to obtain a modified signal collected by the target receiver.
[0102] The signal type identification module 40 is used to input the modified target receiver acquisition signal into the signal type identification model to obtain a signal type identification result.
[0103] The cross-receiver signal type identification device provided by the present application adopts the cross-receiver signal type identification method in the above embodiment, which can solve the technical problem of reduced identification performance due to inconsistent signal data features collected by different receivers during cross-receiver signal type identification. Compared with the related art, the beneficial effects of the cross-receiver signal type identification device provided by the present application are the same as the beneficial effects of the cross-receiver signal type identification method provided by the above embodiment, and other technical features in the cross-receiver signal type identification device are the same as the features disclosed in the above embodiment method, which will not be repeated here.
[0104] The present application provides a cross-receiver signal type identification device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the cross-receiver signal type identification method in the above-mentioned embodiment.
[0105] Reference below Figure 4 , which shows a schematic diagram of the structure of a cross-receiver signal type identification device suitable for implementing the embodiment of the present application. The cross-receiver signal type identification device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The cross-receiver signal type identification device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0106] like Figure 4As shown, the cross-receiver signal type identification device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the cross-receiver signal type identification device are also stored in the RAM 1004. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 may allow the cross-receiver signal type recognition device to communicate wirelessly or wired with other devices to exchange data. Although the cross-receiver signal type recognition device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have instead.
[0107] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.
[0108] The cross-receiver signal type identification device provided by the present application adopts the cross-receiver signal type identification method in the above embodiment, which can solve the technical problem of reduced identification performance due to inconsistent signal data features collected by different receivers during cross-receiver signal type identification. Compared with the related art, the beneficial effects of the cross-receiver signal type identification device provided by the present application are the same as the beneficial effects of the cross-receiver signal type identification method provided by the above embodiment, and other technical features in the cross-receiver signal type identification device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.
[0109] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.
[0110] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0111] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the cross-receiver signal type identification method in the above-mentioned embodiment.
[0112] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.
[0113] The computer-readable storage medium may be included in the cross-receiver signal type identification device; or may exist independently without being assembled into the cross-receiver signal type identification device.
[0114] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the cross-receiver signal type identification device, the cross-receiver signal type identification device can implement the cross-receiver signal type identification method.
[0115] Computer program code for performing the operations of the present application may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0116] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0117] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.
[0118] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned cross-receiver signal type identification method, and can solve the technical problem of reduced recognition performance due to inconsistent signal data features collected by different receivers during cross-receiver signal type identification. Compared with the related art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the cross-receiver signal type identification method provided by the above-mentioned embodiment, and will not be repeated here.
[0119] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned cross-receiver signal type identification method when executed by a processor.
[0120] The computer program product provided by the present application can solve the technical problem of reduced recognition performance due to inconsistent signal data features collected by different receivers when identifying cross-receiver signal types. Compared with the related art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the cross-receiver signal type identification method provided by the above embodiment, and will not be described in detail here.
[0121] The above are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.
Claims
1. A method for identifying signal types across receivers, characterized in that: The method includes: The signal preprocessing stage performs signal preprocessing on the signals collected by at least two source receivers to construct a training data set for the signal recognition model; Model training phase, training the neural network model based on the training data set to obtain the signal recognition model; The feature matching step is to match the signal collected by the target receiver with the signal collected by the source receiver to determine the feature difference information between the signal collected by the target receiver and the signal collected by the source receiver; Based on the characteristic difference information and the signal characteristic of the source receiver acquisition signal, modify the target receiver acquisition signal to obtain a modified target receiver acquisition signal; In the signal type identification step, the modified target receiver acquisition signal is input into the signal type identification model to obtain a signal type identification result.
2. The method for identifying the type of a signal across receivers according to claim 1, wherein: The characteristic information of the collected signal at least includes the sampling frequency information of the target receiver and the sampling frequency information of the source receiver. The characteristic matching link includes the steps of: Determine, from the source receiver sampling frequency information set, source receiver sampling frequency information that is closest to the target receiver sampling frequency information; Based on the sampling frequency information of the closest source receiver, the sampling frequency of the target receiver for acquiring signals is modified.
3. The method for identifying the type of a signal across receivers as claimed in claim 2, wherein: The step of modifying the sampling frequency of the target receiver to acquire the signal based on the sampling frequency information of the closest source receiver includes: If the sampling frequency information of the closest source receiver is greater than the sampling frequency of the target receiver acquisition signal, then upsampling the target receiver acquisition signal; If the sampling frequency information of the closest source receiver is less than the sampling frequency of the target receiver acquisition signal, down-sampling processing is performed on the target receiver acquisition signal.
4. The method for identifying the type of a signal across receivers as claimed in claim 2, wherein: The signal preprocessing step includes the following steps: Acquire source receiver acquisition signals through source receivers with different sampling rate gears; wherein, for each sampling rate gear, acquire at least two source receiver acquisition signals; Performing time-frequency conversion processing on each source receiver acquisition signal to obtain a time-frequency conversion matrix corresponding to the number of sampling rate gears; Based on the time-frequency transformation matrix, a training data set corresponding to the number of sampling rate gears is constructed.
5. The method for identifying the type of a signal across receivers as claimed in claim 4, wherein: The model training process includes the following steps: For each of the training data sets, input the training data set into the neural network model to obtain a model weight file corresponding to the training data set; Based on all the model weight files and the neural network model, the signal recognition model is constructed.
6. The method for identifying the type of a signal across receivers as claimed in claim 5, characterized in that: The signal type identification process includes the following steps: Inputting the modified target receiver acquisition signal into the signal recognition model; Signal type recognition processing is performed based on the model weight file corresponding to the modified sampling frequency of the target receiver to acquire the signal, so as to obtain a signal type recognition result.
7. A device for identifying signal types across receivers, characterized in that: The device comprises: A signal preprocessing module, used to perform signal preprocessing on signals collected by at least two source receivers to construct a training data set for a signal recognition model; A model training module, used to train a neural network model based on the training data set to obtain the signal recognition model; A feature matching module, used to perform signal feature matching on the target receiver acquisition signal and the source receiver acquisition signal, and determine feature difference information between the target receiver acquisition signal and the source receiver acquisition signal; Based on the characteristic difference information and the signal characteristic of the source receiver acquisition signal, modify the target receiver acquisition signal to obtain a modified target receiver acquisition signal; The signal type identification module is used to input the modified target receiver acquisition signal into the signal type identification model to obtain a signal type identification result.
8. A device for identifying signal types across receivers, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the cross-receiver signal type identification method according to any one of claims 1 to 6.
9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the cross-receiver signal type identification method according to any one of claims 1 to 6 are implemented.
10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the cross-receiver signal type identification method according to any one of claims 1 to 6 are implemented.