Adaptive multi-density clustering interference signal sorting method capable of merging parameters
The adaptive multi-density clustering method is used to process interference signals, which solves the problem of interference signal sorting in complex electromagnetic environments, realizes efficient interference signal sorting, and improves sorting performance.
Patent Information
- Application Number
- CN202510385020.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-28
- Publication Date
- 2025-08-01
AI Technical Summary
The existing interference signal sorting methods are difficult to effectively process multi-density interference signals in complex electromagnetic environments, especially the interference signal patterns emitted by different interference sources are complex, which leads to difficulty in sorting and identification.
Adaptive multi-density clustering method for merging parameters is adopted, by receiving interference signals and extracting pulse description words, building pulse description words matrix, calculating K-dist curves and adaptive key parameters k, generating multi-density-related Eps sequences, using the iterated IterDBSCAN algorithm for clustering, merging or adding new clusters until the interfering signal sorting is completed.
It improves the interfering signal sorting effect in complex electromagnetic environments, can handle multi-density conditions, avoids performance losses caused by inaccurate density grading, and improves the sorting effect.
Smart Images
Figure CN120408049A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of interference perception, and particularly relates to a method for sorting interference signals with mergeable parameter adaptive multi-density clustering. Background Art
[0002] In interference perception applications, a monitoring receiver receives pulse signals generated by different interference radiation sources, and these signals are interleaved on the time scale, causing difficulties for subsequent interference signal recognition. The task of sorting interference pulse signals, as a prerequisite for sensing and identifying sensing signals, aims to accurately separate the pulse streams of different interferences. However, with the sharp increase in the density and complexity of the electromagnetic environment, the complexity and dexterity of intentional interference increase. Not only do the interference signals emitted by different interference sources vary, but the interference emitted by a single interference source also exhibits different patterns, making it increasingly difficult to sort and identify interference pulse signals.
[0003] Existing pulse signal sorting methods are divided into two categories, namely methods based on TOA conversion and methods based on multiple parameters. Methods based on TOA transformation have been widely applied to the deinterleaving of pulse signals with periodic PRIs, especially for pulse signals with constant PRIs. The basic idea of the TOA-based deinterleaving method is to first find the PRI period, and then search for the target radiation source pulses with determined values from the pulse stream. The cumulative difference histogram, sequence difference histogram, and difference of the TOA matrix can be used to determine the PRI period. However, these RSS methods based on TOA transformation are very sensitive to the distribution and measurement error of TOA. To overcome the dependence on TOA, the radiation source signal sorting method based on multiple parameters attempts to solve the RSS problem from the perspective of multiple PDW clustering, including K-means clustering, support vector clustering, fuzzy clustering, and density clustering. However, the K-means clustering method does not have the ability to handle non-spherical clusters, and the performance of support vector clustering and fuzzy clustering both severely depends on the previous initialization settings. Although density-based clustering methods, such as the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, can detect clusters of arbitrary shapes, it is difficult to find the optimal density and distance thresholds, namely the MinPts and Eps parameters. In the field of data mining, DBSCAN parameter adaptive selection methods have been studied, and the k-dist curve is mainly used. However, these methods are very sensitive to the selection of key parameters and the distribution of the dataset. Moreover, since the above methods often have only one set of parameters, it is difficult to show good performance for PDW datasets with multi-density in complex electromagnetic environments. Summary of the Invention
[0004] The technical objective of the present invention is to provide a method for sorting interference signals with mergeable parameter adaptive multi-density clustering to solve the problem of insufficient performance in sorting interference signals.
[0005] To solve the above problems, the technical solution of the present invention is as follows: A method for sorting interference signals of adaptive multi-density clustering with mergeable parameters, comprising the following steps: S1: Receive interference signals, and extract pulse description words from the interference signals. The pulse description words include information on carrier frequency, pulse width, pulse repetition interval, and direction of arrival; S2: During the interference perception period, construct a pulse description word matrix and perform normalization processing on the pulse description word matrix; S3: Model the problem of sorting interference signals based on the pulse description word matrix into a multi-density clustering problem; S4: Calculate the distance matrix of each pulse description word, calculate the K-dist curve, and determine the adaptive key parameter k; S5: Draw a histogram for the values of the k-column distance matrix, determine the number of density levels through the peaks of the histogram, and generate an Eps sequence related to multi-density; S6: Generate MinPts parameters based on the Eps sequence in ascending order, and use the iterative IterDBSCAN algorithm for clustering. During each iteration, according to the coincidence degree between the newly added clusters and the original clusters, new clusters are added or merged into the already clustered clusters, and the noise is used as the input for the next round of clustering. Repeat the above steps until all values in the Eps sequence are traversed, thereby completing the sorting of smart interference signals in a complex electromagnetic environment.
[0006] Among them, step S2 is specifically as follows: Within an interference perception period T, assuming that N pulses are collected, the corresponding pulse description word matrix is expressed as , where represents the th PWD vector, and normalization processing is performed on the pulse description word matrix according to the characteristics.
[0007] Among them, step 4 is specifically as follows: Obtain the corresponding Euclidean distance matrix based on the pulse description word matrix , and each element in the Euclidean distance matrix represents the Euclidean distance from to in the pulse description word matrix. The calculation formula is: Then, sort each row in the Euclidean distance matrix in ascending order to obtain the ascending distance matrix ; Represent the kth column of the ascending distance matrix as , then the candidate The calculation formula is: Based on each in value, select the average number of neighbors of all data points in the pulse descriptor matrix to calculate a series of values. The calculation formula is: where represents the number of neighbor samples when using parameters ; Input each pair of and parameters ( ) into the original DBSCAN algorithm to obtain the number of clusters ; If the number of clusters obtained remains unchanged three times, it is considered that this number of clusters is stable and the same as the true number of clusters in the dataset; In addition, the number of clusters decreases monotonically with , and the maximum that satisfies is the required adaptive key parameter k.
[0008] Among them, step S5 is specifically as follows: According to the adaptive key parameter k and the Euclidean distance matrix , obtain a series of values, where is the th element of the matrix ; According to the above values, obtain its frequency histogram at intervals of . The calculation formula is: where represents the count of the histogram interval ; To eliminate the influence of noise, filter the histogram to obtain the satisfied interval number : where represents the threshold of the filter; Obtain the number of density intervals by selecting the peaks of the filtered histogram; According to the number of values related to density, use the k-means method to cluster in to obtain the corresponding cluster center values; Arrange the cluster center values of clusters in ascending order: For the first For each cluster, use the maximum value of each cluster as the parameter related to multi-density parameter , , for the th cluster, use the average value of the values within the cluster to represent the parameter corresponding to the cluster; multi-density Eps sequence
[0009] Among them, step S6 is specifically as follows: Select the Eps related to multi-density in ascending order , and use the selected Eps value to calculate the average number of neighbors for each point to obtain the corresponding value. The calculation formula is: Use and to send the pulse descriptor matrix into the newly designed iterative IterDBSCAN algorithm for clustering. In each iteration of the IterDBSCAN algorithm, by judging whether the number of overlapping points between the new cluster and the existing clusters exceeds the threshold value. If it exceeds the threshold value, it is judged as an existing cluster and merged into the existing cluster; if it does not exceed the threshold value, it is judged as a new cluster and added to the cluster set , and finally obtain the clustering and noise. The calculation formula is: Then, update the clustering , The noise data is used as the input for the next iteration ; When all the Eps values in are traversed, the clustering process ends and returns the clustering results and noise .
[0010] An electronic device includes a processor and a memory. A computer instruction that can run on the processor is stored on the memory. The processor is used to call the computer instruction in the memory to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method as described above.
[0011] A storage medium is used to store computer instructions. When the computer instructions run, they control the device where the storage medium is located to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method as described above.
[0012] Due to the above technical solutions, the present invention has the following advantages and positive effects compared with the prior art: The present invention provides a method for sorting radar signals with parameter - adaptive multi - density clustering, which can process the sorting of different types of interference signals in a complex electromagnetic environment, handle the case where the PDW data set has multi - density levels between and within classes, avoid performance losses caused by inaccurate density grading, and improve the sorting effect of complex and dexterous interference pulse signals. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention.
[0014] Figure 1 FIG. 10 is a schematic flow chart of a method for sorting interference signals with mergeable parameter - adaptive multi - density clustering provided by the present invention; FIG. 2 (a - c) is a comparison chart of the sorting of the method for sorting interference signals with mergeable parameter - adaptive multi - density clustering of the present invention and other methods. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the specific embodiments of the present invention will be described below with reference to the accompanying drawings. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings and other embodiments can be obtained.
[0016] For the sake of simplicity of the drawings, only the parts related to the present invention are schematically shown in each drawing, and they do not represent the actual structure of the product. In addition, for the sake of simplicity and easy understanding of the drawings, in some drawings, parts with the same structure or function are only schematically shown one of them, or only one of them is marked. In this article, "one" not only means "only this one", but also can mean "more than one" situation.
[0017] The following further details a method for sorting interference signals with mergeable parameter - adaptive multi - density clustering proposed by the present invention with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will be clearer according to the following description and the claims.
[0018] Embodiment Referring to Figure 1 and FIG. 2, this embodiment provides a method for sorting interference signals with mergeable parameter - adaptive multi - density clustering, including the following steps: First, in step S1, a radar signal is received by an interference-aware receiver, and there is an interference signal in the radar signal. Pulse description words (PDWs) are extracted from the interference signal, including key information such as carrier frequency (RF), pulse width (RW), and pulse repetition interval (PRI).
[0019] Then, in step S2, within an interference-aware period T, the interference-aware period T can be set to 1 s. Assume that N pulses are collected at the receiving end within the interference-aware period T. Then, the corresponding PDW sequence, i.e., the PDW matrix, is expressed as , where represents the th PDW vector. Next, normalization processing is performed on the above PDW matrix according to features.
[0020] In step S3, the problem of sorting interference signals based on the PDW matrix is modeled as a multi-density clustering problem.
[0021] Furthermore, in step S4, by calculating the distance matrix of each PDW, the K-dist curve is then calculated, and the adaptive key parameter k is determined based on the K-dist curve.
[0022] To find the adaptive key parameter k, the corresponding Euclidean distance matrix is obtained based on the PDW sequence , and each element in the Euclidean distance matrix is , representing the Euclidean distance from to in the PDW sequence. The calculation formula is: Then, each row in the Euclidean distance matrix is sorted in ascending order to obtain the ascending distance matrix ; The kth column of the ascending distance matrix is denoted as , then the candidate is calculated as: Using in each value, a series of values are calculated by taking the average of the number of neighbors of all data points in the PDW matrix. The calculation formula is: where represents the number of neighbor samples of when using the parameter.
[0023] Then, each pair of and parameters ( ) Input into the original DBSCAN algorithm to obtain the number of clusters . If the number of clusters obtained remains unchanged three times, then this number of clusters is considered stable and is the same as the true number of clusters in the dataset. In addition, the number of clusters decreases monotonically with satisfying The maximum of is the required adaptive key parameter k.
[0024] Then, in step S5, draw a histogram for the values of the k-column distance matrix, determine the number of density levels through the peaks of the histogram, and perform k-means clustering on the values of the column distance matrix according to the number of density levels to generate a density-related parameter sequence. Specifically: According to the adaptive key parameter k and the Euclidean distance matrix , a series of values are obtained, where represents the th element of the matrix . According to the above values, obtain its frequency histogram at intervals of where represents the count of the histogram interval ; To eliminate the influence of noise, filter the histogram to obtain the satisfied interval sequence number : where represents the threshold of the filter. Then, obtain the number of density intervals by selecting the peaks of the filtered histogram. According to the number of values related to density, use the k-means method to cluster in to obtain the corresponding cluster center values. Arrange the cluster center values of the clusters in ascending order. For the first clusters, take the maximum value of each cluster as the multi-density related parameter , , for the th cluster, use the average value of the values within the cluster to represent the corresponding parameter; The multi-density Eps sequence
[0025] Finally, enter step S6. In ascending order of the Eps sequence, generate MinPts parameters using the Eps sequence in turn, and perform clustering using the iterative IterDBSCAN algorithm. Collect the generated clustering results, and use the noise as the input for the next round of clustering and repeat until all values in the Eps sequence are traversed, completing the sorting of radar signals in complex environments. Specifically: Select the related to multi-density in ascending order, and use the selected Eps value to calculate the average number of neighbors of each point to obtain the corresponding value. The calculation formula is: Next, use and to send the PDW data set into the newly designed iterative IterDBSCAN algorithm for clustering. In each round of iteration of the IterDBSCAN algorithm, by judging whether the number of overlapping points between the new cluster and the existing cluster exceeds the threshold value. If it exceeds the threshold value, it is judged as an existing cluster and merged into the existing cluster; if it does not exceed the threshold value, it is judged as a new cluster and added to the cluster set . Finally, obtain the clustering and noise. The calculation formula is: Then, update the clustering , and use the noise data as the input for the next iteration . When all the Eps values in are traversed, the clustering process ends, and the clustering results and noise are returned. Further, the clustering and merging process of this step S6 can be referred to Table 1 below: Table 1 As shown in Table 2 below, compare the method proposed in this embodiment with the improved AF-DBSCAN method and the multi-density AMD-DBSCAN method. The table consists of three items: the number of categories of clustering, the normalized mutual information index NMI, and the clustering accuracy rate. It can be seen from the table that the interference signal sorting method proposed in this embodiment has obtained the best sorting performance under three received PDW data sets.
[0026] Table 2 Among them, Figures 2(a-c) show the sorting results of this application and other methods under the conditions set in Experiment 4 in Table 2.
[0027] Preferably, this embodiment also provides an electronic device, which includes a processor and a memory. A computer instruction that can run on the processor is stored on the memory, and the processor is configured to call the computer instruction in the memory to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method as described above.
[0028] Further preferably, this embodiment also provides a storage medium for storing computer instructions. When the computer instructions run, they control the device where the storage medium is located to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method as described above.
[0029] The embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, provided that these changes fall within the scope of the claims of the present invention and their equivalent technologies, they still fall within the protection scope of the present invention.
Claims
1. A method for sorting interference signals of mergeable parameter adaptive multi-density clustering, characterized in that It includes the following steps: S1: Receive interference signals, and extract pulse description words from the interference signals. The pulse description words include information such as carrier frequency, pulse width, pulse repetition interval, and direction of arrival; S2: During the interference perception period, construct a pulse description word matrix and perform normalization processing on the pulse description word matrix; S3: Model the interference signal sorting problem based on the pulse description word matrix into a multi-density clustering problem; S4: Calculate the distance matrix of each pulse description word, calculate the K-dist curve, and determine the adaptive key parameter k; S5: Draw a histogram for the values of the k-column distance matrix, determine the number of density levels through the peaks of the histogram, and generate an Eps sequence related to multi-density; S6: Generate MinPts parameters based on the Eps sequence in ascending order, and use the iterative IterDBSCAN algorithm for clustering. In each round of the iterative process, according to the coincidence degree between the newly added cluster and the original cluster, a new cluster is added or merged into the existing clustered cluster, and the noise is used as the input for the next round of clustering. Repeat the above steps until all values in the Eps sequence are traversed, thereby completing the intelligent interference signal sorting in the complex electromagnetic environment.
2. The method for sorting interference signals of mergeable parameter adaptive multi-density clustering according to claim 1, wherein The specific content of step S2 is as follows: Within an interference perception period T, assuming that N pulses are collected, the corresponding pulse descriptor matrix is expressed as , where represents the th PWD vector. For the pulse descriptor matrix, normalization processing is performed according to the characteristics.
3. The method for sorting interference signals of adaptive multi-density clustering with mergeable parameters according to claim 2, characterized in that, The specific content of step 4 is as follows: Obtain the corresponding Euclidean distance matrix based on the pulse descriptor matrix , each element in the Euclidean distance matrix represents the Euclidean distance from to in the pulse descriptor matrix, and the calculation formula is: Then, sort each row in the Euclidean distance matrix in ascending order to obtain an ascending distance matrix ; Denote the k-th column of the ascending distance matrix as , then the candidate The calculation formula is: Based on each in value, select the average number of neighbors of all data points in the pulse descriptor matrix to calculate a series of values, and the calculation formula is: Among them, indicates the number of neighbor samples when using the parameter; Input each pair of and parameters ( ) into the original DBSCAN algorithm to obtain the number of clusters ; if the number of clusters obtained remains unchanged three times, then this number of clusters is considered stable and is the same as the true number of clusters in the dataset; in addition, the number of clusters decreases monotonically with , and the largest satisfying is the required adaptive key parameter k.
4. The method for sorting interference signals of mergeable parameter adaptive multi-density clustering according to claim 3, wherein The specific content of step S5 is as follows: According to the adaptive key parameter k and the Euclidean distance matrix , a series of values are obtained, where is the -th element of the matrix ; According to the above values, a frequency histogram with an interval of is obtained, and the calculation formula is: Among them, represents the count of the histogram interval ; To eliminate the influence of noise, filter the histogram to obtain the interval numbers that meet the requirements : Among them, represents the threshold of the filter; Obtain the number of density intervals by selecting the peaks of the filtered histogram ; According to the number of values related to density, using the k-means method to cluster in to obtain the corresponding cluster center values; for the cluster center values of the clusters are arranged in ascending order: For the first clusters, the maximum value of each cluster is used as the multi-density related parameter , , for the th cluster, the average value of the values within the cluster is used to represent the parameter corresponding to the cluster; the multi-density Eps sequence 。 5. The method for sorting interference signals of mergeable parameter adaptive multi-density clustering according to claim 4, wherein The specific content of step S6 is as follows: Select in ascending order those related to multi-density , and use the selected Eps value to average the number of neighbors for each point to obtain the corresponding value. The calculation formula is as follows: Utilize and Yes. Send the pulse description word matrix into the newly designed iterative IterDBSCAN algorithm for clustering. In each iteration of the IterDBSCAN algorithm, by judging whether the number of overlapping points between the new cluster and the existing clusters exceeds the threshold value. If it exceeds the threshold value, it is judged as an existing cluster and merged into the existing cluster; if it does not exceed the threshold value, it is judged as a new cluster and added to the cluster set , and finally the clustering and noise are obtained. The calculation formula is as follows: Then, update the clusters , and use the noise data as the input for the next iteration ; When all the Eps values in are traversed, the clustering process ends and the clustering results and noise are returned.
6. An electronic device, characterized in that, The electronic device includes a processor and a memory. A computer instruction that can run on the processor is stored on the memory. The processor is used to call the computer instruction in the memory to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method described in any one of claims 1-5.
7. A storage medium, characterized in that, The storage medium is used to store computer instructions, and when the computer instruction runs, it controls the device where the storage medium is located to execute the mergeable parameter adaptive multi-density clustering interference signal sorting method described in any one of claims 1-5.