Feature tree clustering based signal sorting method and device, storage medium and equipment

Abstract

The application relates to the field of radar signal sorting, and particularly discloses a signal sorting method and device based on feature tree clustering, a storage medium and equipment.The signal sorting method based on feature tree clustering comprises the following steps: after processing target feature parameters of each radar pulse signal into standard data, converting the standard data of each radar pulse signal into corresponding feature classes, wherein the feature classes are nodes of a feature tree; according to a merging condition of the feature classes, performing layer-by-layer clustering on the current feature classes until all the feature classes are merged into the same feature class, so that the feature tree is obtained, wherein the top layer of the feature tree is the same feature class; when the target feature class is determined on the feature tree based on a target feature class radius and / or a target sample threshold, determining a signal sorting result according to the target feature class.The application solves the technical problems that the traditional radar signal sorting method has poor accuracy and cannot cope with complex electromagnetic environments.

Description

Technical Field

[0001] This invention relates to the field of radar signal sorting, and more specifically, to a signal sorting method, apparatus, storage medium, and device based on feature tree clustering. Background Technology

[0002] In electromagnetic space, various radars and signals often intertwine. To obtain useful radar signals, it is essential to classify the diverse electromagnetic pulse signals entering the system according to their respective radiation sources before subsequent signal processing can proceed. With the widespread application and rapid development of electronic equipment, the space electromagnetic environment has become increasingly complex and variable. The dense array of various electromagnetic signals constitutes electronic interference signals for detection radars, affecting the efficiency of radiation source signal sorting. As operating frequency bands continue to widen and signal parameters overlap in multiple dimensions, various unconventional radar radiation source signals have adopted new technologies such as frequency agility and pulse compression, resulting in significant variations in characteristic parameters. This makes it difficult to guarantee the accuracy of traditional radiation source signal sorting methods.

[0003] Radar signal sorting is a technique that separates the pulse sequence of multiple radars under conditions of interleaved pulses, and estimates and identifies the parameters of each radar. Only through radar signal sorting can multiple real radar pulse sequences be obtained, allowing for further processing such as jamming, localization, or tracking. As the foundation of radar reconnaissance signal processing, the importance of radar signal sorting is self-evident. However, with the continuous development of radar technology, the rapidly increasing pulse current density, the ever-widening spectrum, the complex and varied modulation methods, and the continuously changing radar parameters have made the electromagnetic environment that radar signal sorting must cope with increasingly harsher, increasing the difficulty of sorting. Traditional radar signal sorting methods suffer from poor accuracy and are unable to handle complex electromagnetic environments.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a signal sorting method, apparatus, storage medium, and device based on feature tree clustering, to at least solve the technical problems of poor accuracy and inability to cope with complex electromagnetic environments in traditional radar signal sorting methods.

[0006] According to one aspect of the present invention, a signal sorting method based on feature tree clustering is provided, comprising: After processing the target feature parameters of each radar pulse signal into standard data, converting the standard data of each radar pulse signal into a corresponding feature class, wherein the feature class is a node of a feature tree; performing layer-by-layer clustering on the current feature class according to the merging conditions of the feature class, until all the feature classes are merged into the same feature class, obtaining the feature tree, wherein the top layer of the feature tree is the same feature class; when a target feature class is determined on the feature tree based on the target feature class radius and / or target sample threshold, determining the signal sorting result based on the target feature class.

[0007] Preferably, before processing the target feature parameters of each radar pulse signal into standard data, the method further includes: converting the received radar pulse signal into a pulse descriptor; using information entropy to determine the information content of each feature in the pulse descriptor; and determining the target feature parameters from the features based on the information content of each feature.

[0008] Preferably, the above-mentioned layer-by-layer clustering of the current feature class based on the above-mentioned feature class merging conditions includes: calculating the class distance between every two feature classes in all feature classes and determining the two feature classes with the closest distance; merging the two feature classes with the closest distance to obtain a two-level class; and using the above-mentioned two-level class to replace the above-mentioned two feature classes to continue to merge and cluster based on the class distance.

[0009] Preferably, after obtaining the feature tree, the method further includes: if the target feature class is not determined on the feature tree based on the target feature class radius and / or the target sample threshold, continuing to search for radar pulse signals.

[0010] Preferably, after continuing to search for radar pulse signals or determining the target feature class, the method further includes: updating the feature tree based on the received new radar pulse signals and the feature classes that exceed the time window, wherein the time window is a preset radar signal sorting duration.

[0011] Preferably, updating the feature tree based on the received new radar pulse signal includes: searching in the feature tree for the candidate feature class that is closest to the new feature class corresponding to the new radar pulse signal; replacing the candidate feature class with the new feature class when the distance between the candidate feature class and the new feature class is less than the radius of the target feature class, and merging the upper-level classes of the candidate feature class to update the feature tree.

[0012] Preferably, updating the feature tree based on feature classes that exceed the time window includes: searching for expired feature classes that exceed the time window in the feature tree based on the input time of the corresponding radar pulse signal; sequentially deleting the expired feature classes from the upper-level classes that include the feature classes at three or more levels; connecting the feature class closest to the expired feature class with the third-level class corresponding to the second-level class of the expired feature class, and deleting the expired feature class and the second-level class of the expired feature class to update the feature tree.

[0013] According to another aspect of the present invention, a signal sorting device based on feature tree clustering is also provided, comprising: a conversion unit, which, after processing the target feature parameters of each radar pulse signal into standard data, converts the standard data of each radar pulse signal into a corresponding feature class, wherein the feature class is a node of a feature tree; a clustering unit, which performs layer-by-layer clustering on the current feature class according to the merging conditions of the feature class, until all the feature classes are merged into the same feature class to obtain the feature tree, wherein the top layer of the feature tree is the same feature class; and a determination unit, which, when a target feature class is determined on the feature tree based on the target feature class radius and / or a target sample threshold, determines the signal sorting result based on the target feature class.

[0014] According to another aspect of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer program, wherein the computer program is configured to execute the above-described signal sorting method based on feature tree clustering at runtime.

[0015] According to another aspect of the present invention, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to execute the above-described signal sorting method based on feature tree clustering through the computer program.

[0016] In this embodiment of the invention, after processing the target feature parameters of each radar pulse signal into standard data, the standard data of each radar pulse signal is converted into a corresponding feature class. The feature class is a node in the feature tree. Based on the merging conditions of the feature classes, the current feature class is clustered layer by layer until all feature classes are merged into the same feature class, resulting in a feature tree. The top layer of the feature tree is the same feature class. When the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined based on the target feature class. By processing each radar pulse signal into a feature class and treating it as a node in the feature tree, all feature classes are merged into the same feature class at the top layer through layer-by-layer clustering to obtain the feature tree. The target class is determined from the feature tree based on the class radius or sample threshold as the signal sorting result. Thus, the feature tree is generated based on the radar pulse signal as the feature class, achieving the goal of accurately obtaining radar signal sorting results based on the feature tree. This improves the accuracy of radar signal sorting and enables the use of accurate radar signal sorting results to cope with complex electromagnetic environments, thereby solving the technical problems of poor accuracy and inability to cope with complex electromagnetic environments in traditional radar signal sorting methods. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0018] Figure 1 This is a flowchart illustrating an optional signal sorting method based on feature tree clustering according to an embodiment of the present invention;

[0019] Figure 2 This is a flowchart illustrating an optional signal sorting method based on feature tree clustering according to an embodiment of the present invention;

[0020] Figure 3 This is a schematic diagram of feature tree generation in an optional signal sorting method based on feature tree clustering according to an embodiment of the present invention;

[0021] Figure 4 This is a schematic diagram of an optional signal sorting device based on feature tree clustering according to an embodiment of the present invention;

[0022] Figure 5 This is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present invention. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] According to one aspect of the present invention, a signal sorting method based on feature tree clustering is provided, which is widely used.

[0026] As an optional implementation method, such as Figure 1 As shown, the signal sorting method based on feature tree clustering includes:

[0027] S102, after processing the target feature parameters of each radar pulse signal into standard data, the standard data of each radar pulse signal is converted into a corresponding feature class, where the feature class is a node of the feature tree;

[0028] S104. Based on the merging conditions of feature classes, perform layer-by-layer clustering on the current feature classes until all feature classes are merged into the same feature class to obtain a feature tree, wherein the top layer of the feature tree is the same feature class.

[0029] S106, when the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined according to the target feature class.

[0030] In this embodiment, after processing the target feature parameters of each radar pulse signal into standard data, the standard data of each radar pulse signal is converted into a corresponding feature class. The feature class is a node of the feature tree. According to the merging conditions of the feature classes, the current feature class is clustered layer by layer until all feature classes are merged into the same feature class to obtain the feature tree. The top layer of the feature tree is the same feature class. When the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined based on the target feature class. By processing each radar pulse signal into a feature class and treating it as a node of the feature tree, all feature classes are merged into the same feature class at the top layer through layer-by-layer clustering to obtain the feature tree. The target class is determined from the feature tree based on the class radius or sample threshold as the signal sorting result. Thus, the feature tree is generated based on the radar pulse signal as the feature class, achieving the purpose of accurately obtaining the radar signal sorting result based on the feature tree. This improves the accuracy of radar signal sorting and enables the use of accurate radar signal sorting results to cope with complex electromagnetic environments. This solves the technical problems of poor accuracy and inability to cope with complex electromagnetic environments in traditional radar signal sorting methods.

[0031] As an optional implementation, before processing the target characteristic parameters of each radar pulse signal into standard data, the following steps are also included:

[0032] S12, convert the received radar pulse signal into a pulse description word;

[0033] S14, use information entropy to determine the information content of each feature in the pulse description;

[0034] S16, determine the target feature parameters from each feature based on the information content of each feature.

[0035] As an optional implementation, based on the merging conditions of feature classes, the current feature classes are clustered layer by layer, including:

[0036] S104-2, calculate the class distance between any two feature classes in all feature classes, and determine the two feature classes with the closest distance;

[0037] S104-4, merge the two closest feature classes to obtain a two-level class;

[0038] S104-6, using a two-level class to replace the two feature classes and continue to merge and cluster based on class distance.

[0039] As an optional implementation, after obtaining the feature tree, the method further includes: if the target feature class is not determined on the feature tree based on the target feature class radius and / or the target sample threshold, continuing to search for radar pulse signals.

[0040] As an optional implementation, after continuing to search for radar pulse signals or determining the target feature class, the method further includes: updating the feature tree based on the received new radar pulse signals and the feature classes that exceed the time window, wherein the time window is a preset radar signal sorting duration.

[0041] As an optional implementation, updating the feature tree based on the received new radar pulse signal includes:

[0042] S22, Search the feature tree for the candidate feature class that is closest to the new feature class corresponding to the new radar pulse signal;

[0043] S24. If the distance between the candidate feature class and the new feature class is less than the radius of the target feature class, replace the candidate feature class with the new feature class and merge the upper-level classes of the candidate feature class to update the feature tree.

[0044] As an optional implementation, updating the feature tree based on feature classes that exceed the time window includes:

[0045] S32, based on the input time of the corresponding radar pulse signal, search for expired feature classes that are out of time in the feature tree;

[0046] S34, delete expired feature classes sequentially from the upper-level classes (including feature classes) at levels three or above;

[0047] S36, connect the third-level class corresponding to the second-level class of the expired feature class that is closest to the expired feature class, and delete the expired feature class and the second-level class of the expired feature class to update the feature tree.

[0048] In this embodiment, radar signal sorting based on feature tree hierarchical clustering solves the problems of poor accuracy and difficulty in dealing with complex environments in traditional radiation source signal sorting algorithms.

[0049] Specifically, the radar signal sorting process based on feature tree hierarchical clustering is not limited to two stages: a search stage and a tracking stage. In the search stage, feature determination and feature tree generation are performed. In the tracking stage, the feature tree is updated and the clustering results are calculated. If clustering fails, the process switches to the search state. If clustering is successful, the tracking state continues.

[0050] Radar signal sorting methods based on feature tree hierarchical clustering are not limited to, for example, Figure 2 As shown:

[0051] S1, Feature Determination. The first step is to process the received radar pulse signal into a Pulse Description Word (PDW). This is not limited to abstracting the radar pulse signal in the electromagnetic environment intercepted by the radar receiving system into a digital Pulse Description Word (PDW). The composition of the Pulse Description Word (PDW) is not limited to the following formula:

[0052]

[0053] In equation (1), DOA represents the angle of arrival of the signal, RF represents the carrier frequency of the signal, TOA represents the arrival time of the signal, PW represents the pulse width of the signal, PA represents the pulse amplitude of the signal, F represents the intra-pulse modulation type of the signal, and i represents the pulse number sorted according to the detection time.

[0054] The second step is to determine the information content of each feature in the pulse descriptor word (PDW) and to identify the target feature parameters. Specifically, this involves, but is not limited to, estimating the information content of each feature in the PDW using information entropy, defining features with low entropy values ​​as having higher importance. Features are then ranked according to their importance, and a threshold is set to determine the most important features for clustering. This is not limited to determining 2-3 feature parameters for clustering.

[0055] The third step is to standardize the target feature parameters. This is not limited to using the min-max normalization method to standardize the feature data. By ensuring that the data falls within a specific interval, the unit restrictions on the feature data are removed, and the data is transformed into dimensionless pure numerical values, which facilitates comparison and calculation between feature data of different units or magnitudes.

[0056] The definitions and operations related to feature classes are not limited to the following:

[0057] Suppose that the feature class contains column vectors of sample features in the form of: The composition of a feature class is not limited to the following expression:

[0058] C={Num, Centre, Sum, Sqsum, Ntoa} (2)

[0059] In equation (2), Num represents the number of samples in the feature class, Centre represents the centroid of the feature class, Sum represents the sum of the magnitudes of the feature vectors of each sample in the feature class, Sqsum represents the sum of the squares of the magnitudes of the feature vectors of each sample in the feature class, and Ntoa represents the sum of the squares of the magnitudes of the feature vectors of each sample in the feature class. TOA The maximum value.

[0060] The radius of a feature class is defined as the mean squared error of all samples, but is not limited to the following formula:

[0061]

[0062] Let G be the function that merges two feature classes A and B into C. Then the operation of C = G(A, B) is not limited to the following formula:

[0063] C Num =A Num +B Num

[0064]

[0065] C sum =A Sum +B Sum

[0066] C Sqsum =A Sqsum +B Sqsum

[0067] C Ntoa =max(A Ntoa B Ntoa (4)

[0068] Let S be the function that extracts feature class B from feature class A. Num >B Num Therefore, the operation of C = S(A, B) is not limited to the following formula:

[0069] C Num =A Num -B Num

[0070]

[0071] C Sum =A Sum -B Sum

[0072] C Sqsum =A Sqsum -B Sqsum

[0073]

[0074] The distance between two feature classes is defined as the Euclidean distance between the centroids of the two feature classes. Therefore, the distance between two feature classes A and B is not limited to the following formula:

[0075]

[0076] S2, Feature Tree Generation. During the search phase, a feature tree is constructed using a generation algorithm based on the pulse descriptors of radar pulse signals sampled over a period of time. The feature tree is not limited to a binary tree where all nodes belong to the feature class.

[0077] The algorithm for generating feature trees is not limited to the following:

[0078] 1) The sample set containing M samples obtained during the search phase Convert into M feature classes The transformed feature classes are then used as leaf nodes of the feature tree.

[0079] 2) Calculate the Euclidean distance between all feature classes, find the two feature classes with the closest distance, and denote them as C. p C q ;

[0080] 3) Place C p C q Merge into C k (Two-level class), and C k As C p and C q The parent node;

[0081] 4) Place C p C q Remove from the search scope of the feature class, and C k Add to the search scope;

[0082] 5) Repeat steps 2) to 4) above until all feature classes are merged into the same feature class at the top.

[0083] Taking 5 samples as an example, the feature tree generated by the above algorithm is not limited to the following: Figure 2 As shown, the five samples are x1, x2, x3, x4, and x5. The feature tree generation algorithm sequentially merges x2 and x3, and then merges x4 and x5 to form two parent nodes. Next, x1 is merged with x2 and x3 to get x1x2x3. Finally, x1x2x3 is merged with x4 and x5 to become the top node, thus generating the feature tree.

[0084] S3, Calculate the clustering results. Clustering results can be calculated after the feature tree is generated during both the search and tracking phases. Clustering is successful when a feature class that meets the target class radius and / or sample threshold is found based on the feature tree search. The target feature class obtained through clustering is the radar signal sorting result.

[0085] S4, Update Feature Tree. After the feature tree is generated, the search phase transitions to the tracking phase. During the tracking phase, the feature tree is updated with sample data corresponding to newly received radar pulse signals, while sample data that exceeds the time window is removed from the feature tree.

[0086] The algorithm for feature tree updating is not limited to the following:

[0087] 1) The received new pulse signal is processed into new sample data x i ;

[0088] 2) By calculating the search distance x in the feature tree i The most recent leaf feature class C k And record the Euclidean distance D between them;

[0089] 3) Compare D with the target class radius threshold R th If D>R th Then x i Discard, otherwise keep C k Replace the feature class contained within with x i At the same time, C k The ancestor nodes are merged sequentially.

[0090] The algorithms for feature tree removal are not limited to the following:

[0091] 1) Find the condition that satisfies the removal criteria: (T) now -N toa )>T win Feature class C k , among which, T now N represents the current time. toa T is the signal input time. win For statistical time windows;

[0092] 2) If a C that satisfies the removal condition is found k Then from C k Remove C from ancestor nodes k ;

[0093] 3) Place C k sibling nodes and C k The parent node is connected to its parent's child node;

[0094] 4) Delete C k and its parent node;

[0095] 5) Repeat steps 1)-4) above until the time for all feature classes no longer meets the removal criteria.

[0096] After updating the feature tree during the tracking phase, S3 is not limited to being executed again to calculate the clustering results. If the clustering is successful, the tracking state continues; if the clustering fails, the search state is entered.

[0097] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0098] According to another aspect of the present invention, a signal sorting apparatus based on feature tree clustering is also provided for implementing the above-described signal sorting method based on feature tree clustering. For example... Figure 4As shown, the device includes:

[0099] The conversion unit 402, after processing the target feature parameters of each radar pulse signal into standard data, converts the standard data of each radar pulse signal into a corresponding feature class, wherein the feature class is a node of the feature tree;

[0100] Clustering unit 404 is used to perform layer-by-layer clustering on the current feature class according to the merging conditions of the feature class, until all feature classes are merged into the same feature class to obtain a feature tree, wherein the top layer of the feature tree is the same feature class;

[0101] The determining unit 406 is used to determine the signal sorting result based on the target feature class when the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold.

[0102] Optionally, the signal sorting device based on feature tree clustering described above further includes a preprocessing unit, which converts the received radar pulse signal into a pulse descriptor before processing the target feature parameters of each radar pulse signal into standard data; determines the information content of each feature in the pulse descriptor using information entropy; and determines the target feature parameters from each feature based on the information content of each feature.

[0103] Optionally, the clustering unit 404 performs layer-by-layer clustering on the current feature class according to the merging conditions of the feature class, including: calculating the class distance between every two feature classes in all feature classes and determining the two feature classes with the closest distance; merging the two feature classes with the closest distance to obtain a two-level class; and using the two-level class to replace the two feature classes to continue merging and clustering according to the class distance.

[0104] Optionally, the signal sorting device based on feature tree clustering further includes a search unit, which, after obtaining the feature tree, continues to search for radar pulse signals if the target feature class is not determined on the feature tree based on the target feature class radius and / or the target sample threshold.

[0105] Optionally, the signal sorting device based on feature tree clustering further includes an update unit, which, after continuing to search for radar pulse signals or after determining the target feature class, further includes: updating the feature tree based on the received new radar pulse signals and the feature classes that exceed the time window, wherein the time window is a preset radar signal sorting duration.

[0106] Optionally, the above-mentioned updating unit updates the feature tree based on the received new radar pulse signal, including: searching in the feature tree for the candidate feature class that is closest to the new feature class corresponding to the new radar pulse signal; replacing the candidate feature class with the new feature class when the distance between the candidate feature class and the new feature class is less than the radius of the target feature class, and merging the various upper-level classes of the candidate feature class to update the feature tree.

[0107] Optionally, the above-mentioned update unit updates the feature tree based on feature classes that have exceeded the time window, including: searching for expired feature classes that have exceeded the time window in the feature tree based on the input time of the corresponding radar pulse signal; deleting expired feature classes sequentially from the upper-level classes that include feature classes at three or more levels; connecting the feature class closest to the expired feature class with the third-level class corresponding to the second-level class of the expired feature class, and deleting the expired feature class and the second-level class of the expired feature class to update the feature tree.

[0108] In this embodiment, after processing the target feature parameters of each radar pulse signal into standard data, the standard data of each radar pulse signal is converted into a corresponding feature class. The feature class is a node of the feature tree. According to the merging conditions of the feature classes, the current feature class is clustered layer by layer until all feature classes are merged into the same feature class to obtain the feature tree. The top layer of the feature tree is the same feature class. When the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined based on the target feature class. By processing each radar pulse signal into a feature class and treating it as a node of the feature tree, all feature classes are merged into the same feature class at the top layer through layer-by-layer clustering to obtain the feature tree. The target class is determined from the feature tree based on the class radius or sample threshold as the signal sorting result. Thus, the feature tree is generated based on the radar pulse signal as the feature class, achieving the purpose of accurately obtaining the radar signal sorting result based on the feature tree. This improves the accuracy of radar signal sorting and enables the use of accurate radar signal sorting results to cope with complex electromagnetic environments. This solves the technical problems of poor accuracy and inability to cope with complex electromagnetic environments in traditional radar signal sorting methods.

[0109] According to another aspect of the present invention, an electronic device for implementing the above-described signal sorting method based on feature tree clustering is also provided. This electronic device may be a terminal device or a server. Figure 5 As shown, the electronic device includes a memory 502 and a processor 504. The memory 502 stores a computer program, and the processor 504 is configured to execute the steps of any of the above method embodiments through the computer program.

[0110] Optionally, in this embodiment, the aforementioned electronic device may be located in at least one of a plurality of network devices in a computer network.

[0111] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0112] S1, after processing the target feature parameters of each radar pulse signal into standard data, convert the standard data of each radar pulse signal into a corresponding feature class, where the feature class is a node of the feature tree;

[0113] S2, based on the merging conditions of feature classes, perform layer-by-layer clustering on the current feature classes until all feature classes are merged into the same feature class, to obtain a feature tree, where the top layer of the feature tree is the same feature class;

[0114] S3, when the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined according to the target feature class.

[0115] Alternatively, as those skilled in the art will understand, Figure 5 The structure shown is for illustrative purposes only; the electronic device can be any terminal device. Figure 5 This does not limit the structure of the aforementioned electronic devices. For example, the electronic device may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, etc.), or having the same Figure 5 The different configurations shown.

[0116] The memory 502 can be used to store software programs and modules, such as the program instructions / modules corresponding to the monitoring method and device for intelligent devices in this embodiment of the invention. The processor 504 executes various functional applications and data processing by running the software programs and modules stored in the memory 502, thereby realizing the signal sorting method based on feature tree clustering described above. The memory 502 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 502 may further include memory remotely located relative to the processor 504, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. Specifically, the memory 502 may be used, but is not limited to, to store information such as feature classes, feature trees, target feature classes, and signal sorting results. As an example, such as Figure 5 As shown, the memory 502 may include, but is not limited to, the conversion unit 402, clustering unit 404, and determination unit 404 in the signal sorting device based on feature tree clustering. Furthermore, it may include, but is not limited to, other module units in the signal sorting device based on feature tree clustering, which will not be elaborated upon in this example.

[0117] Optionally, the transmission device 506 described above is used to receive or send data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 506 includes a Network Interface Controller (NIC), which can be connected to other network devices and a router via a network cable to communicate with the Internet or a local area network. In another example, the transmission device 506 is a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0118] In addition, the above-mentioned electronic device also includes: a display 508 for displaying the target feature class and signal sorting results; and a connection bus 510 for connecting the various module components in the above-mentioned electronic device.

[0119] In other embodiments, the aforementioned terminal device or server can be a node in a distributed system, wherein the distributed system can be a blockchain system, which is a distributed system formed by connecting multiple nodes through network communication. The nodes can form a peer-to-peer (P2P) network, and any form of computing device, such as a server, terminal, or other electronic device, can become a node in the blockchain system by joining this peer-to-peer network.

[0120] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations of the feature tree clustering-based signal sorting described above. The computer program is configured to execute the steps in any of the above method embodiments at runtime.

[0121] Optionally, in this embodiment, the computer-readable storage medium described above may be configured to store a computer program for performing the following steps:

[0122] S1, after processing the target feature parameters of each radar pulse signal into standard data, convert the standard data of each radar pulse signal into a corresponding feature class, where the feature class is a node of the feature tree;

[0123] S2, based on the merging conditions of feature classes, perform layer-by-layer clustering on the current feature classes until all feature classes are merged into the same feature class, to obtain a feature tree, where the top layer of the feature tree is the same feature class;

[0124] S3, when the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined according to the target feature class.

[0125] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0126] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0127] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0128] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0129] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0130] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0131] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0132] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A signal sorting method based on feature tree clustering, characterized in that, include: After processing the target feature parameters of each radar pulse signal into standard data, the standard data of each radar pulse signal is converted into a corresponding feature class, wherein the feature class is a node of the feature tree; Based on the merging conditions of the feature classes, the current feature classes are clustered layer by layer until all the feature classes are merged into the same feature class, thus obtaining the feature tree, wherein the top layer of the feature tree is the same feature class; When a target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold, the signal sorting result is determined according to the target feature class. After determining the target feature class, the method further includes: updating the feature tree based on the received new radar pulse signal and feature classes that have exceeded the time window, including: searching in the feature tree for the candidate feature class that is closest to the new feature class corresponding to the new radar pulse signal; replacing the candidate feature class with the new feature class when the distance between the candidate feature class and the new feature class is less than the radius of the target feature class, and merging the upper-level classes of the candidate feature class to update the feature tree; searching in the feature tree for expired feature classes that have exceeded the time window based on the input time of the corresponding radar pulse signal; sequentially deleting the expired feature classes from the upper-level classes that include the feature class at three or more levels; connecting the feature class closest to the expired feature class with the third-level class corresponding to the second-level class of the expired feature class, and deleting the expired feature class and the second-level class of the expired feature class to update the feature tree, wherein the time window is a preset radar signal sorting duration.

2. The method according to claim 1, characterized in that, Before processing the target characteristic parameters of each radar pulse signal into standard data, the following steps are also included: Convert the received radar pulse signal into a pulse descriptor; The information content of each feature in the pulse description is determined using information entropy. The target feature parameters are determined from the features based on the information content of each feature.

3. The method according to claim 1, characterized in that, The step of performing layer-by-layer clustering on the current feature class according to the merging conditions of the feature class includes: Calculate the class distance between any two feature classes in all feature classes, and determine the two feature classes with the closest distance; Merge the two feature classes that are closest in distance to obtain a two-level class; The two feature classes are replaced by the two-level class to continue merging and clustering based on class distance.

4. The method according to claim 1, characterized in that, After obtaining the feature tree, the process also includes: If the target feature class is not determined on the feature tree based on the target feature class radius and / or the target sample threshold, the search for radar pulse signals continues.

5. A signal sorting device based on feature tree clustering, employing the method described in any one of claims 1 to 4, characterized in that, include: The conversion unit processes the target feature parameters of each radar pulse signal into standard data, and then converts the standard data of each radar pulse signal into a corresponding feature class, wherein the feature class is a node of the feature tree. A clustering unit is used to perform layer-by-layer clustering on the current feature class according to the merging conditions of the feature class, until all the feature classes are merged into the same feature class to obtain the feature tree, wherein the top layer of the feature tree is the same feature class; The determining unit is used to determine the signal sorting result based on the target feature class when the target feature class is determined on the feature tree based on the target feature class radius and / or the target sample threshold.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method described in any one of claims 1 to 4.

7. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 4 through the computer program.

Images