A Single Radiation Source Determination Method Based on Temporal Information Entropy

Abstract

This application belongs to the field of radar reconnaissance technology and provides a single-source radiation determination method based on temporal information entropy. It first calculates the Time Difference of Arrival (TDOA) of all adjacent pulses, then sets a clustering threshold and a first-class cluster center, and uses a clustering algorithm to obtain multiple clusters based on the TDOA. Next, it reduces the number of clusters and removes invalid elements by merging and removing harmonics. The distribution probability of each TDOA is calculated based on the number of indices in the merged clusters after harmonic removal. The temporal information entropy is calculated based on the distribution probability of each TDOA. A set information entropy discrimination threshold is set; if the temporal information entropy does not exceed the threshold, the pre-sorted signal is considered to be composed of a single radiation source; otherwise, it is composed of multiple radiation sources. This method can accurately determine the composition of radiation source signals in the radar pre-sorting clustering results, avoiding further PRI main sorting processing of single-source signals and reducing redundant sorting calculations.

Description

Technical Field

[0001] This application belongs to the field of radar reconnaissance technology, and specifically relates to a method for determining a single radiation source based on temporal information entropy. Background Technology

[0002] Radar signal sorting is a crucial technical branch in radar electronic reconnaissance. The current mainstream processing architecture is pre-sorting clustering + PRI-based main sorting. Pre-sorting clustering performs coarse classification of signals based on features such as frequency and angle of arrival. Then, each coarse classification result undergoes PRI-based main sorting processing to ultimately sort the radar signals. However, in many cases, the pre-sorting results have already separated the signals of individual radar sources from overlapping signal streams. Performing further main sorting at this point would lead to redundant computations, reducing algorithm efficiency, and would also introduce batching issues for modulation types such as PRI grouping, requiring subsequent batching and increasing the system complexity of the sorting process. Therefore, it is necessary to determine the signal composition of the pre-sorting results and adopt different processing methods for single-source and multi-source pre-sorting to effectively improve signal sorting efficiency. Summary of the Invention

[0003] The purpose of this application is to provide a single radiation source determination method based on temporal information entropy, so as to solve the problems of redundant calculation and additional batching caused by existing radar signal sorting.

[0004] The technical solution of this application is: a single radiation source determination method based on temporal information entropy, comprising:

[0005] The arrival time (TOA) of all pulses is obtained from the pre-sorting results, and then the arrival time difference (TDOA) of all adjacent pulses is calculated based on the TOA value.

[0006] Clustering is initialized based on the arrival time difference (TDOA) values ​​of all adjacent pulses. Then, a clustering threshold and the cluster center of the first class are set. Starting from the second TDOA, it is determined whether the difference between the TDOA and the cluster center is less than the clustering threshold. If so, the TDOA is assigned to the corresponding cluster, the cluster index of the TDOA is obtained, and the cluster center is recalculated. If not, the element is determined to be the new class center. This process continues until all TDOAs have been traversed, and the cluster centers and cluster indices of all classes are obtained.

[0007] The indexes are reordered according to the order of arrival from the cluster centers, and then the merged categories are initialized before merging adjacent clusters.

[0008] Determine whether the TDOA corresponding to all indices in the merged class is a harmonic; if so, delete the harmonic from the corresponding cluster.

[0009] Calculate the probability distribution of each TDOA based on the number of indexes in the merged class after harmonic removal;

[0010] Calculate the temporal information entropy based on the distribution probability of each TDOA;

[0011] If the temporal information entropy does not exceed the information entropy discrimination threshold, the pre-sorted signal is considered to be composed of a single radiation source; otherwise, it is composed of multiple radiation sources.

[0012] Preferably, the method for performing linear clustering is as follows:

[0013] Cluster initialization: M = 1, Center(1) = TDOA(1), Index1 = 1. Where M represents the number of clusters obtained, Center(i) represents the cluster center of the i-th cluster, and Index1 = 1. i This represents the TDOA index of the i-th class, i = 1, ..., M;

[0014] Then, starting from the second TDOA, if |Center(i)-TDOA(n)|≤ΔT exists, then the nth TDOA is considered to belong to the i-th class, and the TDOA index is added to the index of the i-th class. i In the middle, ΔT represents the clustering threshold; if it does not exist, the nth TDOA is considered to belong to a new class: M = M + 1, Index M =n.

[0015] Preferably, the method for merging adjacent classes is as follows:

[0016] Initialize the merged categories: K = 1, CIndex1 = Index1. Here, K represents the number of categories after merging, and CIndex1 = Index1. k This represents the TDOA index of the k-th class after merging, where k = 1, ..., K. Then, starting from Index2, traverse all indices. i If the following conditions are met:

[0017]

[0018] Then the index of the i-th class i Add the index to CIndex K If K = K + 1 is not satisfied, then CIndex K =Index i .in, represents the mean TDOA of the i-th class before merging, and std(·) represents the standard deviation calculation.

[0019] Preferably, the method for removing harmonics is as follows:

[0020] If there exist p, q ∈ [1, K] that satisfy the following condition:

[0021]

[0022] Then it is believed that CIndex q The corresponding TDOA is a harmonic and needs to be removed from CIndex.

[0023] Preferably, the probability distribution is calculated as follows:

[0024]

[0025] Where K0 represents the number of categories after removing harmonics, numel(CIndex) k ) indicates CIndex k The number of elements.

[0026] Preferably, the method for calculating the temporal information entropy is as follows:

[0027]

[0028] This application presents a single-source radiation determination method based on temporal information entropy. First, it calculates the Time Difference of Arrival (TDOA) of all adjacent pulses. Then, by setting a clustering threshold and a first-class cluster center, it uses a clustering algorithm to obtain multiple clusters based on the TDOA. Next, it reduces the number of clusters and removes invalid elements by merging and removing harmonics. The distribution probability of each TDOA is calculated based on the number of indices in the merged clusters after harmonic removal. The temporal information entropy is calculated based on the distribution probability of each TDOA. A set information entropy discrimination threshold is set. If the temporal information entropy does not exceed the threshold, the pre-sorted signal is considered to be composed of a single radiation source; otherwise, it is considered to be composed of multiple radiation sources. This method can accurately determine the composition of radiation source signals in radar pre-sorting clustering results, avoiding further PRI main sorting processing of single-source signals and reducing redundant sorting calculations. It also boasts short calculation time and high computational efficiency. Attached Figure Description

[0029] To more clearly illustrate the technical solutions provided in this application, the accompanying drawings will be briefly described below. Obviously, the drawings described below are merely some embodiments of this application.

[0030] Figure 1 This is a schematic diagram of the overall process of this application;

[0031] Figure 2 This is a schematic diagram of the statistical results of the arrival time difference distribution of linear clustering in this application;

[0032] Figure 3 This is a schematic diagram showing the statistical results of the arrival time difference after merging and deharmonicization in this application. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0034] A method for determining a single radiation source based on temporal information entropy is proposed. The single radiation source signal has the property that the arrival time difference is equivalent to PRI, and its arrival time difference has a relatively clustered distribution. This application utilizes this characteristic for determination.

[0035] like Figure 1 As shown, it includes the following steps:

[0036] Step S100: Obtain the arrival time (TOA) of all pulses based on the pre-sorting results, and then calculate the arrival time difference (TDOA) of all adjacent pulses based on the TOA values.

[0037] TDOA(n) = TDOA(n+1) - TDOA(n)

[0038] Where n = 1, 2, ..., N-1, and N represents the length of the pulse sequence obtained from the pre-sorting.

[0039] Step S200: Cluster initialization is performed based on the arrival time difference (TDOA) values ​​of all adjacent pulses. Then, a clustering threshold and the cluster center of the first class are set. Starting from the second TDOA, it is determined whether the difference between it and the cluster center is less than the clustering threshold. If so, the TDOA is assigned to the corresponding cluster, the cluster index of the TDOA is obtained, and the cluster center is recalculated. If not, the element is determined to be the new class center. This process continues until all TDOAs are traversed, and the cluster centers and cluster indices of all classes are obtained.

[0040] Preferably, the method for performing linear clustering is as follows:

[0041] Cluster initialization: M = 1, Center(1) = TDOA(1), Index1 = 1. Where M represents the number of clusters obtained, Center(i) represents the cluster center of the i-th cluster, and Index1 = 1. i This represents the TDOA index of the i-th class, i = 1, ..., M;

[0042] Then, starting from the second TDOA, if |Center(i)-TDOA(n)|≤ΔT exists, then the nth TDOA is considered to belong to the i-th class, and the TDOA index is added to the index of the i-th class.i In the middle, ΔT represents the clustering threshold; if it does not exist, the nth TDOA is considered to belong to a new class: M = M + 1, Index M =n.

[0043] For example, if we set the time difference (TDOA) of a set of pulse arrival times to 20 / 35 / 20.1 / 35 / 19.8 / 34.8... μs, we can set the clustering threshold to 0.5 μs. Then, through the above calculations, the first cluster is initially centered at 20, and the second cluster is initially centered at 35. We can divide 20 / 20.1 / 19.8... into one cluster, and 35 / 35 / 34.8... into another cluster. After clustering, we recalculate the cluster centers of all elements within each cluster.

[0044] Using the above clustering method, if there is only a single radiation source, fewer clusters will be formed; if there are multiple radiation sources, more clusters will be formed. Furthermore, by setting clustering thresholds, different data can be effectively distinguished to facilitate further judgment.

[0045] In step S300, in order to ensure the accuracy of clustering, the clustering threshold is set relatively small. However, due to the presence of interference in the actual received pulses, the TDOA value, which should be, for example, 20, becomes 19 or 21, which is outside the range of the clustering threshold, thus forming different clusters. Of course, TDOA of 19 or 21 may also be other interference, so it is necessary to further determine whether such clusters can be merged with the clusters with a cluster center of 20.

[0046] In other words: the indexes are reordered according to the order of cluster centers from smallest to largest, and then the merged categories are initialized, K=1, CIndex1=Index1. Here, K represents the number of merged categories, and CIndex1=Index1. k This represents the TDOA index of the k-th class after merging, where k = 1, ..., K. Then, starting from Index2, traverse all indices. i If the following conditions are met:

[0047]

[0048] Then the index of the i-th class i Add the index to CIndex K If K = K + 1 is not satisfied, then CIndex K =Index i .in, represents the mean TDOA of the i-th class before merging, and std(·) represents the standard deviation calculation.

[0049] By calculating the standard deviation, clusters with time differences exceeding the clustering threshold can be identified and merged. This allows elements that should belong to one cluster to be merged into one cluster, preventing misjudgments and allowing the index to be reconfigured.

[0050] Step S400: Harmonic suppression is performed.

[0051] In actual received pulses, pulse loss may also occur. For example, if the PRI of a certain pulse is 40μs, when a pulse is lost in the middle, the time difference TDOA at that position becomes 80μs, which leads to the existence of harmonics. Based on this problem, it is necessary to determine whether certain TDOAs are harmonics. If so, harmonic suppression is required.

[0052] Specifically, determine whether the TDOA corresponding to all indexes in the merged class is a harmonic; if so, delete the harmonic from the corresponding cluster.

[0053] Preferably, the method for removing harmonics is as follows:

[0054] If there exist p, q ∈ [1, K] that satisfy the following condition:

[0055]

[0056] Then it is believed that CIndex q The corresponding TDOA is a harmonic and needs to be removed from CIndex.

[0057] Step S500: Calculate the distribution probability of each TDOA based on the number of indices in the merged class after harmonic removal.

[0058] Preferably, the probability distribution is calculated as follows:

[0059]

[0060] Where K0 represents the number of categories after removing harmonics, numel(CIndex) k ) indicates CIndex k The number of elements.

[0061] After removing clutter, merging clusters, and removing harmonics, if it is a single-source signal, there is only one cluster with a probability of approximately 1; if it is a dual-source signal, there are two clusters with a probability of approximately 0.5 each; and so on.

[0062] Step S600: Calculate the temporal information entropy based on the distribution probability of each TDOA.

[0063] Preferably, the method for calculating temporal information entropy is as follows:

[0064]

[0065] Temporal information entropy reflects the focusing characteristics of time difference of arrival (TDOA) to some extent. The larger the temporal information entropy, the more types of TDOA there are; the smaller the temporal information entropy, the fewer types of TDOA there are.

[0066] Step S700: Set the information entropy discrimination threshold. Since this application only needs to determine whether the radiation source corresponding to the pulse is a single radiation source or multiple radiation sources, the information entropy discrimination threshold is generally set to 1. If the temporal information entropy does not exceed the information entropy discrimination threshold, that is, it is less than 1, then the pre-sorting signal is considered to be composed of a single radiation source; otherwise, it is composed of multiple radiation sources.

[0067] By calculating the temporal entropy of clusters, the type of radiation source can be determined simply and accurately.

[0068] This application, when determining a single radiation source, first calculates the Time Difference of Arrival (TDOA) of all adjacent pulses. Then, by setting a clustering threshold and a first-class cluster center, a clustering algorithm is used to obtain multiple clusters based on the TDOA. Next, the number of clusters is reduced and invalid elements are deleted by merging and removing harmonics. The distribution probability of each TDOA is calculated based on the number of indices in the merged clusters after harmonic removal. The temporal information entropy is calculated based on the distribution probability of each TDOA. A set information entropy discrimination threshold is set. If the temporal information entropy does not exceed the threshold, the pre-sorted signal is considered to be composed of a single radiation source; otherwise, it is composed of multiple radiation sources. This method can accurately determine the composition of radiation source signals in the radar pre-sorting clustering results, avoiding further PRI main sorting processing of single-source signals and reducing redundant sorting calculations. Simultaneously, the calculation time is short and the calculation efficiency is high.

[0069] The following example illustrates the situation using a specific parameter:

[0070] Assume a pre-sorting result consists of two radar radiation sources. Radar source 1 has an initial pulse time of 15 μs and a PRI of 40 μs, while radiation source 2 has an initial pulse time of 30 μs and a PRI of 37 μs. The pulse duration is 2 ms, and ΔT is 1 μs. The statistical distribution results of the pulses from the two radar radiation sources are as follows: Figure 2-3 As shown, the temporal information entropy of the pulse obtained using this method is 1.9764. Here, the information entropy discrimination threshold α is set to 1, and according to this method, it is determined that it is composed of multiple radar radiation sources. This conclusion is consistent with the simulation settings, proving the effectiveness of the method.

[0071] Finally, the following points should be noted: First, in the description of this application, it should be noted that, unless otherwise specified and limited, the terms "installation", "connection", and "linkage" should be interpreted broadly, and can be mechanical or electrical connections, or internal connections between two components, or direct connections. "Up", "down", "left", "right", etc. are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may change.

[0072] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0073] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for determining a single radiation source based on temporal information entropy, characterized in that, include: The arrival time (TOA) of all pulses is obtained from the pre-sorting results, and then the arrival time difference (TDOA) of all adjacent pulses is calculated based on the TOA value. Clustering is initialized based on the arrival time difference (TDOA) values ​​of all adjacent pulses. Then, a clustering threshold and the cluster center of the first class are set. Starting from the second TDOA, it is determined whether the difference between the TDOA and the cluster center is less than the clustering threshold. If so, the TDOA is assigned to the corresponding cluster, the cluster index of the TDOA is obtained, and the cluster center is recalculated. If not, the TDOA is determined to be the new cluster center. This process continues until all TDOAs have been traversed, and the cluster centers and cluster indices of all classes are obtained. The cluster indexes are reordered according to the order of arrival from the cluster centers, and then the merged categories are initialized before merging adjacent clusters. Determine whether the TDOA corresponding to all indices in the merged class is a harmonic; if so, delete the harmonic from the corresponding cluster. Calculate the probability distribution of each TDOA based on the number of indexes in the merged class after harmonic removal; Calculate the temporal information entropy based on the distribution probability of each TDOA; An information entropy discrimination threshold is set. If the temporal information entropy does not exceed the information entropy discrimination threshold, the pre-sorted signal is considered to be composed of a single radiation source; otherwise, it is composed of multiple radiation sources.

2. The single radiation source determination method based on temporal information entropy as described in claim 1, characterized in that, The method for performing linear clustering is as follows: Cluster initialization: , , ;in, This represents the number of clusters obtained from clustering. Indicates the first Cluster centers of classes Indicates the first Class index, ; Then from the second one Begin iterating; if it exists... Then it is considered that the first indivual Belonging to the Class, and will Add the index to the first Class In; among them, This represents the clustering threshold; if it does not exist, then the first cluster is considered to be clustered. indivual Belong to a new class: , .

3. The single radiation source determination method based on temporal information entropy as described in claim 1, characterized in that: The method for merging adjacent classes is as follows: Initialize the merged categories. , ;in, This indicates the number of categories after merging. Indicates the number after the merger Class index, ; and then from the first Start traversing all If the following conditions are met: ； Then the first Class Adding the index If not satisfied: , ;in, Indicates the number before the merger Class corresponding mean This indicates the calculation of standard deviation.

4. The single radiation source determination method based on temporal information entropy as described in claim 3, characterized in that: The methods for removing harmonics are as follows: If it exists It meets the following conditions: ； Then it is believed corresponding As harmonics, they need to be removed from... Delete it.

5. The single radiation source determination method based on temporal information entropy as described in claim 3, characterized in that: The method for calculating the probability distribution is as follows: ； in, This indicates the number of categories after removing harmonics. express The number of elements, This indicates the length of the pulse sequence obtained from the pre-sorting.

6. The single radiation source determination method based on temporal information entropy as described in claim 5, characterized in that: The method for calculating the temporal information entropy is as follows: 。

Images