Multi-frequency-hopping signal sorting method, system and device used in complex electromagnetic environment and storage medium

Abstract

The invention provides a multi-frequency-hopping signal sorting method, system and device used in a complex electromagnetic environment and a storage medium, and aims to realize multi-frequency-hopping signal sorting in the complex electromagnetic environment without any prior information. The method comprises the steps that multi-frequency-hopping signal data in a complex electromagnetic environment are acquired, and the multi-frequency-hopping signal data comprise effective signal data and noise signal data; performing time-frequency analysis processing on the multi-frequency-hopping signal data to obtain time-frequency data corresponding to the multi-frequency-hopping signal data; predicting an optimal radius parameter in a DBSCAN clustering algorithm by using a neural network learning method; based on the optimal radius parameter, performing DBSCAN clustering algorithm processing on the time-frequency data to obtain a cluster of effective signals; and extracting characteristic parameters of the effective signals based on the clusters of the effective signals.

Description

Technical Field

[0001] This application relates to the field of electromagnetic signal sorting, specifically to a method, system, device, and storage medium for sorting multi-frequency hopping signals in complex electromagnetic environments. Background Technology

[0002] Frequency-hopping communication technology, due to its good anti-interference performance, is widely used in shortwave and VHF radio communication fields. This also presents significant challenges to frequency-hopping signal reconnaissance and analysis, especially the reconnaissance and analysis of multi-frequency-hopping signals in complex electromagnetic environments. In the process of frequency-hopping signal reconnaissance and analysis, existing multi-frequency-hopping signal sorting methods mainly employ non-blind sorting methods based on partial prior information (number of signal sources or some characteristic parameters of the signal). This involves first extracting multi-frequency-hopping signals in complex electromagnetic environments using time-frequency analysis methods; then measuring and estimating the characteristic parameters of the frequency-hopping signals, including frequency parameters, period parameters, and transition times; and finally, sorting the multi-frequency-hopping signals based on the measured characteristic parameters and supported by prior information.

[0003] However, the prerequisite for achieving multi-frequency hopping signal sorting is to have certain prior information about multi-frequency hopping signals.

[0004] Therefore, how to achieve multi-frequency hopping signal sorting in complex electromagnetic environments without any prior information has become a research focus in the fields of reconnaissance and countermeasures. Summary of the Invention

[0005] This application provides a method for sorting multi-frequency hopping signals in complex electromagnetic environments, so as to adapt to the sorting of multi-frequency hopping signals in complex electromagnetic environments without any prior information.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] In a first aspect, this application provides a method for sorting multi-frequency hopping signals in a complex electromagnetic environment. The method includes: acquiring multi-frequency hopping signal data in a complex electromagnetic environment, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data.

[0008] Time-frequency analysis and processing are performed on the multi-frequency hopping signal data to obtain the corresponding time-frequency data.

[0009] Using neural network learning methods, the optimal radius parameter in the DBSCAN clustering algorithm is predicted;

[0010] Based on the optimal radius parameter, the time-frequency data is processed by the DBSCAN clustering algorithm to obtain clusters of effective signals;

[0011] Based on clusters of effective signals, feature parameters of the effective signals are extracted.

[0012] One possible design approach, the first aspect of the method, also includes acquiring multi-frequency hopping signal data through electromagnetic signal reconnaissance equipment and / or electromagnetic environment monitoring equipment.

[0013] One possible design scheme, the first aspect of the method, also includes a method for time-frequency analysis processing of multi-frequency hopping signal data, including at least one or more of the following: short-time Fourier transform, Wigner distribution, wavelet analysis.

[0014] One possible design approach, the first aspect of which also includes using neural network learning methods to predict the optimal radius parameter in the DBSCAN clustering algorithm, including:

[0015] Set multiple radius parameters and construct a scoring function;

[0016] Based on multiple radius parameters, the DBSCAN clustering algorithm is applied to the time-frequency data to obtain the clustering analysis results corresponding to the multiple radius parameters respectively;

[0017] Based on multiple radius parameters and the cluster analysis results corresponding to each radius parameter, the score function values ​​corresponding to each radius parameter are obtained.

[0018] A sample dataset is constructed based on multiple radius parameters and their corresponding score function values.

[0019] Based on the sample set, the parameters of the neural network are optimized so that the neural network can predict the clustering analysis results of the time-frequency data after performing the DBSCAN clustering algorithm according to the radius parameter, thereby enabling the neural network to predict the optimal radius parameter.

[0020] One possible design approach, the first aspect of which also includes optimizing the parameters of the neural network based on the sample set, includes:

[0021] A subset of samples in the sample set is input into the neural network model. The neural network obtains the output value corresponding to the radius parameter in the sample subset based on the radius parameter in the sample subset.

[0022] The loss function is calculated based on the output value corresponding to the radius parameter in the sample subset and the score function value in the sample subset;

[0023] The parameters of the neural network are optimized based on the loss function.

[0024] One possible design approach, the first aspect of which also includes optimizing the parameters of the neural network based on the loss function, includes:

[0025] Based on the loss function, the parameters of the neural network are optimized using mini-batch gradient descent and backpropagation algorithms.

[0026] One possible design approach, the first aspect of which also includes the characteristic parameters of the effective signal including center frequency, occupied bandwidth, maximum frequency hopping interval, minimum frequency hopping interval, and frequency hopping rate.

[0027] Secondly, this application provides a multi-frequency hopping signal sorting system for complex electromagnetic environments, the system comprising:

[0028] The acquisition module is used to acquire multi-frequency hopping signal data under complex electromagnetic environments, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data;

[0029] The processing module is used to perform time-frequency analysis on multi-frequency hopping signal data to obtain the corresponding time-frequency data. It is used to predict the optimal radius parameter in the DBSCAN clustering algorithm using a neural network learning method. It is also used to perform DBSCAN clustering on the time-frequency data based on the optimal radius parameter to obtain clusters of effective signals.

[0030] The extraction module is used to extract feature parameters of the effective signals based on clusters of effective signals.

[0031] Thirdly, a multi-frequency hopping signal sorting device for use in complex electromagnetic environments is provided, the multi-frequency hopping signal sorting device for use in complex electromagnetic environments includes a module for performing the method of the first aspect described above.

[0032] In one possible design, the multi-frequency hopping signal sorting device for complex electromagnetic environments in the third aspect may further include a transceiver. This transceiver can be a transceiver circuit or an interface circuit. The transceiver can be used by the multi-frequency hopping signal sorting device for complex electromagnetic environments in the third aspect to communicate with other devices.

[0033] In one possible design, the multi-frequency hopping signal sorting device for complex electromagnetic environments in the third aspect may further include a memory. This memory may be integrated with the processor or disposed separately. The memory may be used to store instructions relating to the method of the first aspect.

[0034] Fourthly, a multi-frequency hopping signal sorting device for use in complex electromagnetic environments is provided. The multi-frequency hopping signal sorting device for use in complex electromagnetic environments includes: a processor coupled to a memory, the processor executing instructions stored in the memory to cause the multi-frequency hopping signal sorting device for use in complex electromagnetic environments to perform the method of the first aspect.

[0035] In one possible design, the multi-frequency hopping signal sorting device for complex electromagnetic environments of the fourth aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used by the multi-frequency hopping signal sorting device for complex electromagnetic environments of the fourth aspect to communicate with other devices.

[0036] Fifthly, a multi-frequency hopping signal sorting device for complex electromagnetic environments is provided, comprising: a processor and a memory; the memory is used to store instructions, which, when executed by the processor, cause the multi-frequency hopping signal sorting device for complex electromagnetic environments to perform the method of the first aspect.

[0037] In one possible design, the multi-frequency hopping signal sorting device for complex electromagnetic environments of the fifth aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used by the multi-frequency hopping signal sorting device for complex electromagnetic environments of the fifth aspect to communicate with other devices.

[0038] In a sixth aspect, a computer-readable storage medium is provided, the computer-readable storage medium including storage of a computer program or instructions that, when executed, cause the multi-frequency hopping signal sorting method of the first aspect for complex electromagnetic environments to be performed.

[0039] In this embodiment, time-frequency analysis is first used to process multi-frequency hopping signals in complex electromagnetic environments. Then, a neural network learning method is used to predict the optimal radius parameter in the DBSCAN clustering algorithm. Based on the optimal radius parameter, the time-frequency data is processed by the DBSCAN clustering algorithm to obtain clusters of effective signals. Finally, based on the clusters of effective signals, the feature parameters of the effective signals are extracted. That is, multi-frequency hopping signals can be identified and sorted in complex electromagnetic environments without any prior information, ensuring accurate extraction of target signals while filtering out interference and noise, thus ensuring the quality and reliability of communication.

[0040] Other features and advantages of this application will be described in detail in the following detailed description section. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 A flowchart illustrating a multi-frequency hopping signal sorting method for complex electromagnetic environments provided in this application embodiment;

[0043] Figure 2 The results of cluster analysis on the radius parameters selected by human experience in the embodiments of this application;

[0044] Figure 3 A schematic diagram of the loss function provided in the embodiments of this application;

[0045] Figure 4 The neural network training convergence graph provided in the embodiments of this application;

[0046] Figure 5 The result of cluster analysis on the optimal radius parameters generated by the neural network provided in the embodiments of this application;

[0047] Figure 6 A schematic diagram of the structure of the multi-frequency hopping signal sorting device for complex electromagnetic environments provided in the embodiments of this application. Figure 1 ;

[0048] Figure 7 A schematic diagram of the structure of the multi-frequency hopping signal sorting device for complex electromagnetic environments provided in the embodiments of this application. Figure 2 . Detailed Implementation

[0049] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. At the same time, in the description of the embodiments of this application, the terms "first," "second," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0050] Figure 1 This is a flowchart illustrating a multi-frequency hopping signal sorting method for complex electromagnetic environments, provided in an embodiment of this application.

[0051] The process of this multi-frequency hopping signal sorting method for complex electromagnetic environments is as follows:

[0052] Step S101: Acquire multi-frequency hopping signal data under complex electromagnetic environment, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data.

[0053] Among them, multi-frequency hopping signal data is obtained through electromagnetic signal reconnaissance equipment and / or electromagnetic environment monitoring equipment. The above two methods are only examples. They can also be obtained through a variety of other technologies and equipment, such as software-defined radio (SDR), radar systems, communication monitoring systems, satellite signal receiving systems, etc. Each device and method has different characteristics and application scenarios, and is suitable for different electromagnetic environments and signal analysis needs. No restrictions are imposed here.

[0054] It should be noted that the aforementioned valid signal data typically contains important communication information (such as data packets, voice, and video), needs to be extracted from complex environments, and usually exhibits strong signal strength and specific frequency transition patterns. Noise signal data, on the other hand, typically does not contain valid communication information, and its spectrum is random or exhibits interference characteristics. The characteristics of noise signals are usually manifested in their wide spectral distribution and frequency irregularities.

[0055] Step S102: Perform time-frequency analysis processing on the multi-frequency hopping signal data to obtain the time-frequency data corresponding to the multi-frequency hopping signal data.

[0056] Methods for time-frequency analysis of multi-frequency hopping signal data include at least one or more of the following: Short-Time Fourier Transform (STFT), Wigner-Vile Distribution (WVD), and wavelet analysis.

[0057] Step S103: Use neural network learning methods to predict the optimal radius parameter in the DBSCAN clustering algorithm.

[0058] Before describing step S103 in detail, which uses neural network learning to predict the optimal radius parameter in the DBSCAN clustering algorithm, we will first explain how to process the time-frequency data in step S102 using the DBSCAN clustering algorithm to obtain clusters of effective signals when the optimal radius parameter is not obtained.

[0059] Cluster analysis primarily extracts valuable information from data by exploring its structural information and the similarities between data points. Following the principle of maximizing intra-cluster similarity and minimizing inter-cluster similarity, clustering algorithms group data points into clusters. Commonly used clustering algorithms include partition-based clustering, hierarchical clustering, density-based clustering, and spectral clustering. Due to its strong interpretability and ease of understanding, density-based clustering algorithms have received considerable attention. This embodiment uses the density-based spatial clustering of applications with noise (DBSCAN) algorithm to process time-frequency data of multi-frequency hopping signals, finding the final clustering result by identifying the largest set of density-connected data points.

[0060] A cluster mined by the DBSCAN algorithm consists of all data points located in a region separated by densely connected regions and sparsely connected regions. The density of a data point is defined as the number of data points within a spherical neighborhood centered at that data point and with radius r.

[0061] For any data point x1, its density ρ i For: ρ i =|N r (x i )|, where N r (x i )={x j ∈X|d ij ≤r},d ij It is x i and x j The Euclidean distance between them, where |N| represents the number of data points in set N.

[0062] The DBSCAN algorithm refers to data points located within clusters and those located at cluster edges as core points and boundary points, respectively. Based on the density of the data points, it sets a density threshold, Minpts, to identify core and boundary points in the dataset.

[0063] Definition 1: For any data point x in dataset X i If density ρ i If x is greater than or equal to the density threshold Minpts, then x i It is a core point; if density ρ i If x ≤ density threshold Minpts, then i It is a boundary point.

[0064] Definition 2, if x i It is the core point and x j ∈N r (x iIf x j From x i Direct density is achievable.

[0065] Definition 3, for any two data points x in dataset X i and x j If chain x exists a1 x a2 x aN , where x a1 =x i And x aN =x j , making There is x a(h+1) From x ah If the direct density is attainable, then x j From x i Density can be achieved.

[0066] The process of mining time-frequency data of multi-frequency hopping signals using the DBSCAN algorithm can be described as follows: First, arbitrarily select a time-frequency data point as the core point, and merge all density-reachable time-frequency data points from this core point to form the first cluster. Then, arbitrarily select another time-frequency data point from the remaining time-frequency data points as the core point, and merge all density-reachable time-frequency data points from this core point to form the second cluster, and so on, until no core point remains among the remaining time-frequency data points. Data points that are not density-reachable from any core point are identified as noise by the DBSCAN algorithm.

[0067] The specific steps for mining time-frequency data of multi-frequency hopping signals using the DBSCAN algorithm are as follows:

[0068] Input data: Multi-frequency hopping signal time-frequency dataset X = {x1, x2, ..., x n}, radius r, density threshold Minpts.

[0069] Output: Clusters C1, C2, ..., C c Noise C0, denoted by 'c', where 'c' represents the number of clusters in the dataset.

[0070] According to ρ i =|N r (x i )| Calculate the density of each data point in X; Identify the core points in X according to Definition 1, denoted as CP for the set of core points and R for the initial set; Set initial conditions: R←CP, c←0;

[0071] Perform the following iterative process until the condition is met. ( (representing the empty set)

[0072] (1) c←c+1;

[0073] (2)

[0074] (3)C c ←C c ∪{x j |x j From x i Density achievable;

[0075] (4) R←R\C c .

[0076] Finally, confirm the noise:

[0077] Furthermore, by processing the STFT time-frequency data of multi-frequency hopping signals using the DBSCAN algorithm, the following can be obtained: Figure 2 The results are shown. Furthermore, the DBSCAN algorithm indicates that the radius parameter in the input data... and density threshold An initial value needs to be set. (The above...) Figure 2 The radius parameter is a radius selected based on human experience. To set more suitable initial values ​​for the STFT time-frequency data of different multi-frequency hopping signals, this application utilizes a neural network learning method to predict the optimal radius parameter in the DBSCAN clustering algorithm. For details on how to use the neural network learning method to predict the optimal radius parameter in the DBSCAN clustering algorithm, please refer to the following steps S201-S205.

[0078] Step S201: Set multiple radius parameters and construct the scoring function.

[0079] Set multiple radius parameters r1, r2, ..., r K For k∈N, construct the scoring function:

[0080] in: v is the number of noise points in the cluster analysis result C0; α is the weight coefficient, α∈[0,1], used to determine the weight ratio of u and v in the scoring function; is the average distance between all points in the i-th cluster in the clustering analysis results under the current radius r and density threshold Minpts, and c is the number of clusters in the clustering analysis results.

[0081] When the density threshold Minpts is determined, the clustering analysis result is uniquely determined by the radius r. Simultaneously, u and v are also uniquely determined by the clustering analysis result; therefore, u and v are functions of r, and the scoring function g(u,v) is also a function of r. Since the analytical expressions for u and v with r cannot be accurately represented, the relationship between g(u,v) and r cannot be analytically expressed either. This embodiment introduces a neural network to describe the relationship between g(u,v) and r.

[0082] Since u decreases as v increases, g(u,v) must have an extremum. There is a functional relationship between g(u,v) and r, so there must exist an extremum point r. By solving for the extremum point of the neural network, an extremum point r can be determined. fit This makes g(u,v) achieve its maximum value g MAX The radius r that makes g(u,v) reach its maximum value fit That is the optimal radius value.

[0083] Step S202: Based on multiple radius parameters, perform DBSCAN clustering algorithm on the time-frequency data respectively to obtain the clustering analysis results corresponding to the multiple radius parameters.

[0084] Randomly select multiple radius parameters r1, r2, ..., r K For k∈N, the DBSCAN clustering algorithm is used on the time-frequency data to obtain clustering analysis results corresponding to multiple radius parameters.

[0085] Step S203: Based on the multiple radius parameters and the cluster analysis results corresponding to the multiple radius parameters, obtain the score function values ​​corresponding to the multiple radius parameters respectively.

[0086] That is, based on multiple radius parameters and the cluster analysis results corresponding to each radius parameter, g1, g2, ..., g are obtained. K ,k∈N.

[0087] Step S204: Construct a sample dataset based on multiple radius parameters and the score function values ​​corresponding to each radius parameter.

[0088] The above sample dataset can be represented as {(r1,g1),(r2,g2),…,(r K ,g K )}.

[0089] Step S205: Based on the sample set, optimize the parameters of the neural network so that the neural network can predict the clustering analysis results of the time-frequency data after performing the DBSCAN clustering algorithm according to the radius parameter, thereby enabling the neural network to predict the optimal radius parameter.

[0090] Specifically, in step S205 above, optimizing the parameters of the neural network based on the sample set includes:

[0091] Step S301: Input a subset of samples from the sample set into the neural network model. The neural network obtains the output value corresponding to the radius parameter in the sample subset based on the radius parameter in the sample subset.

[0092] The neural network is based on the radius parameters r1, r2, ..., r in the sample subset. K For k∈N, we obtain the output values ​​R(r1), R(r2), ..., R(r... k ), k∈N.

[0093] Step S302: Calculate the loss function based on the output value corresponding to the radius parameter in the sample subset and the score function value in the sample subset.

[0094] The above loss function is expressed as Loss k =[g k -R(r k )] 2 You can refer to this. Figure 3 Understood. The loss function is calculated according to this formula. k Loss characterizes the degree to which a neural network approximates actual mathematical relationships. k The smaller the value, the more realistic the neural network is.

[0095] Step S303: Optimize the parameters of the neural network based on the loss function.

[0096] Based on the value of the loss function, the parameters of the neural network are optimized using mini-batch gradient descent and backpropagation algorithms. Steps S301-S303 are repeated until the loss is reached. k convergence.

[0097] Through the above training process, the neural network model obtained in step S205 is an approximate function of g(u,v) and r. Given the radius r as input, the estimated value of the corresponding score can be obtained at the output. This enables the neural network to predict the clustering analysis result of the DBSCAN clustering algorithm on the time-frequency data based on the radius parameter, and thus enables the neural network to predict the optimal radius parameter r. fit This can be approximated as finding the extreme point that maximizes the output value R(r) of the neural network.

[0098] That is, r∈(0,1) is randomly selected as the initial value, and the optimal solution of the above optimization model is obtained on the trained neural network using the gradient ascent algorithm as the numerical solution method, and the maximum value R is obtained iteratively. MAX(r), the radius r corresponding to this maximum value is the optimal radius.

[0099] For example, this application embodiment uses a neural network architecture based on PyTorch-3.6 to build a scoring network, and trains the scoring network according to the above training process. Before training begins, 100 radii are randomly sampled, and the corresponding score values ​​are calculated using the clustering analysis results of the DBSCAN algorithm and the scoring function, constructing a sample dataset containing 100 samples. The scoring network is trained using this sample dataset, and the training results are as follows. Figure 4 Show.

[0100] Step S104: Based on the optimal radius parameter, the time-frequency data is processed by the DBSCAN clustering algorithm to obtain clusters of effective signals.

[0101] Optimal radius parameters calculated using neural networks Using these initial values, DBSCAN cluster analysis was performed on the multi-frequency hopping signal, and the clustering results are as follows: Figure 5 As shown. Furthermore, in comparison... Figure 5 and Figure 2 It can be seen that the clustering analysis results obtained by using the radius calculated by the neural network are basically similar to those obtained by human experience, indicating that the optimal radius calculated by the neural network is in line with human expectations.

[0102] Step S105: Extract the feature parameters of the effective signals based on the clusters of effective signals.

[0103] The characteristic parameters of an effective signal include center frequency, occupied bandwidth, maximum frequency hopping interval, minimum frequency hopping interval, and frequency hopping rate.

[0104] In summary, in this embodiment, firstly, time-frequency analysis is used to process multi-frequency hopping signals in complex electromagnetic environments; then, a neural network learning method is used to predict the optimal radius parameter in the DBSCAN clustering algorithm; next, based on the optimal radius parameter, the time-frequency data is processed by the DBSCAN clustering algorithm to obtain clusters of effective signals; finally, based on the clusters of effective signals, the feature parameters of the effective signals are extracted. That is, without any prior information, multi-frequency hopping signals can be identified and sorted in complex electromagnetic environments, ensuring accurate extraction of target signals while filtering out interference and noise, thus ensuring communication quality and reliability.

[0105] The above combination Figures 1-5 This application provides a detailed description of the multi-frequency hopping signal sorting method for complex electromagnetic environments, as provided in the embodiments of this application. The following details the implementation of the multi-frequency hopping signal sorting system for complex electromagnetic environments provided in the embodiments of this application.

[0106] The system specifically includes an acquisition module, a processing module, and an extraction module, as shown below.

[0107] The acquisition module is used to acquire multi-frequency hopping signal data under complex electromagnetic environments, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data;

[0108] The processing module is used to perform time-frequency analysis on multi-frequency hopping signal data to obtain the corresponding time-frequency data. It is used to predict the optimal radius parameter in the DBSCAN clustering algorithm using a neural network learning method. It is also used to perform DBSCAN clustering on the time-frequency data based on the optimal radius parameter to obtain clusters of effective signals.

[0109] The extraction module is used to extract feature parameters of the effective signals based on clusters of effective signals.

[0110] Furthermore, the specific implementation of the above system is basically similar to the method implementation, so the description is relatively simple. For relevant details, please refer to the description of the method implementation. Moreover, it should be noted that in the various modules of the system of this application, the components are logically divided according to the functions they are to perform. However, this application is not limited to this and can re-divide or combine the components as needed.

[0111] The above describes the multi-frequency hopping signal sorting method and system for complex electromagnetic environments provided by the embodiments of this application. The following, in conjunction with... Figures 6-7 This document provides a detailed description of the multi-frequency hopping signal sorting device for complex electromagnetic environments provided in the embodiments of this application.

[0112] Figure 6 This is a schematic diagram of the structure of the multi-frequency hopping signal sorting device for complex electromagnetic environments provided in the embodiments of this application. Figure 1 For example, such as Figure 6 As shown, the multi-frequency hopping signal sorting device 600 for complex electromagnetic environments includes a transceiver module 601 and a processing module 602. For ease of explanation, Figure 6 Only the main components of this multi-frequency hopping signal sorting device for complex electromagnetic environments are shown.

[0113] The transceiver module 601 is used to perform the transceiver function of the above-mentioned multi-frequency hopping signal sorting method for complex electromagnetic environments, and the processing module 602 is used to perform other functions of the above-mentioned multi-frequency hopping signal sorting method for complex electromagnetic environments besides the transceiver function.

[0114] Optionally, the transceiver module 601 may include a transmitting module ( Figure 6 (not shown in the image) and receiving module ( Figure 6(Not shown in the image). The transmitting module is used to implement the transmitting function of the multi-frequency hopping signal sorting device 600 for use in complex electromagnetic environments, and the receiving module is used to implement the receiving function of the multi-frequency hopping signal sorting device 600 for use in complex electromagnetic environments.

[0115] Optionally, the multi-frequency hopping signal sorting device 600 for use in complex electromagnetic environments may also include a storage module. Figure 6 (Not shown in the image), the storage module stores programs or instructions. When the processing module 602 executes the program or instructions, the multi-frequency hopping signal sorting device 600 for complex electromagnetic environments can perform the multi-frequency hopping signal sorting method for complex electromagnetic environments in the embodiments of this application.

[0116] The following is combined with Figure 7 Each component of the multi-frequency hopping signal sorting device 700 for use in complex electromagnetic environments will be described in detail:

[0117] The processor 701 is the control center of the multi-frequency hopping signal sorting device 700 used in complex electromagnetic environments. It can be a single processor or a collective term for multiple processing elements. For example, the processor 701 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement the embodiments of this application, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0118] Optionally, the processor 701 can execute various functions of the multi-frequency hopping signal sorting device 700 for complex electromagnetic environments by running or executing software programs stored in the memory 702 and calling data stored in the memory 702, such as executing the multi-frequency hopping signal sorting method for complex electromagnetic environments in the embodiments of this application.

[0119] In a specific implementation, as one example, the processor 701 may include one or more CPUs, for example... Figure 7 CPU0 and CPU1 are shown in the diagram.

[0120] In a specific implementation, as one example, the multi-frequency hopping signal sorting device 700 for complex electromagnetic environments may also include multiple processors, for example... Figure 7The processors 701 and 704 are shown in the diagram. Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, "processor" can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions). The memory 702 is used to store the software program executing the scheme of this application, and its execution is controlled by the processor 701. Specific implementation methods can be found in the above method embodiments, and will not be repeated here.

[0121] Optionally, the memory 702 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 702 may be integrated with the processor 701 or exist independently, and may be connected via an interface circuit of the multi-frequency hopping signal sorting device 700 for complex electromagnetic environments. Figure 7 (Not shown in the image) is coupled to the processor 701, but this application embodiment does not specifically limit this.

[0122] Transceiver 703 is used for communication with other communication devices. For example, in a multi-frequency hopping signal sorting device 700 for use in complex electromagnetic environments, which is the first device, transceiver 703 can be used to communicate with a second device or a third device.

[0123] Optionally, transceiver 703 may include a receiver and a transmitter. Figure 7 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0124] Optionally, the transceiver 703 can be integrated with the processor 701 or exist independently, and can be connected via the interface circuit of the multi-frequency hopping signal sorting device 700 for complex electromagnetic environments. Figure 7 (Not shown in the image) is coupled to the processor 701, but this application embodiment does not specifically limit this.

[0125] Understandable, Figure 7 The structure of the multi-frequency hopping signal sorting device 700 for complex electromagnetic environments shown in the figure does not constitute a limitation on the multi-frequency hopping signal sorting device for complex electromagnetic environments. Actual multi-frequency hopping signal sorting devices for complex electromagnetic environments may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0126] Furthermore, the technical effects of the multi-frequency hopping signal sorting device 700 used in complex electromagnetic environments can be referred to the technical effects of the method described in the above method embodiments, and will not be repeated here.

[0127] It should be understood that the processor in the embodiments of this application can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0128] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0129] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

Claims

1. A method for sorting multi-frequency hopping signals in complex electromagnetic environments, characterized in that, The method includes: Acquire multi-frequency hopping signal data under complex electromagnetic environments, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data; Time-frequency analysis processing is performed on the multi-frequency hopping signal data to obtain the time-frequency data corresponding to the multi-frequency hopping signal data; Using neural network learning methods, the optimal radius parameter in the DBSCAN clustering algorithm is predicted; Based on the optimal radius parameter, the time-frequency data is processed by the DBSCAN clustering algorithm to obtain the clusters of the effective signals; Based on the clusters of the effective signals, the feature parameters of the effective signals are extracted.

2. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 1, characterized in that, The multi-frequency hopping signal data is acquired through electromagnetic signal reconnaissance equipment and / or electromagnetic environment monitoring equipment.

3. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 1, characterized in that, The method for performing time-frequency analysis processing on the multi-frequency hopping signal data includes at least one or more of the following: short-time Fourier transform, Wigner distribution, and wavelet analysis.

4. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 1, characterized in that, The method of using neural network learning to predict the optimal radius parameter in the DBSCAN clustering algorithm includes: Set multiple radius parameters and construct a scoring function; Based on the multiple radius parameters, the DBSCAN clustering algorithm is applied to the time-frequency data to obtain the clustering analysis results corresponding to the multiple radius parameters respectively; Based on the multiple radius parameters and the cluster analysis results corresponding to the multiple radius parameters, the score function values ​​corresponding to the multiple radius parameters are obtained respectively; A sample dataset is constructed based on the plurality of radius parameters and the score function values ​​corresponding to the plurality of radius parameters; Based on the sample set, the parameters of the neural network are optimized so that the neural network can predict the clustering analysis results of the time-frequency data after performing the DBSCAN clustering algorithm according to the radius parameter, thereby enabling the neural network to predict the optimal radius parameter.

5. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 4, characterized in that, The optimization of the neural network parameters based on the sample set includes: A subset of samples in the sample set is input into the neural network model, and the neural network obtains the output value corresponding to the radius parameter in the sample subset based on the radius parameter in the sample subset. The loss function is calculated based on the output value corresponding to the radius parameter in the sample subset and the score function value in the sample subset; Based on the loss function, the parameters of the neural network are optimized.

6. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 1, characterized in that, The optimization of the parameters of the neural network based on the loss function includes: Based on the loss function, the parameters of the neural network are optimized using mini-batch gradient descent and backpropagation algorithms.

7. The multi-frequency hopping signal sorting method for complex electromagnetic environments according to claim 1, characterized in that, The characteristic parameters of the effective signal include center frequency, occupied bandwidth, maximum frequency hopping interval, minimum frequency hopping interval, and frequency hopping rate.

8. A multi-frequency hopping signal sorting system for complex electromagnetic environments, characterized in that, The system includes: The acquisition module is used to acquire multi-frequency hopping signal data under complex electromagnetic environments, wherein the multi-frequency hopping signal data includes effective signal data and noise signal data; The processing module is used to perform time-frequency analysis processing on the multi-frequency hopping signal data to obtain the time-frequency data corresponding to the multi-frequency hopping signal data; it is also used to predict the optimal radius parameter in the DBSCAN clustering algorithm using a neural network learning method; and it is further used to perform DBSCAN clustering algorithm processing on the time-frequency data based on the optimal radius parameter to obtain the clusters of the effective signals. An extraction module is used to extract feature parameters of the effective signals based on clusters of the effective signals.

9. A multi-frequency hopping signal sorting device for use in complex electromagnetic environments, characterized in that, The apparatus includes a module for performing the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a computer program or instructions that, when executed, cause the method as described in any one of claims 1-7 to be performed.

Images