Anti-jitter pulse repetition interval sorting method based on two-dimensional feature vectors
By using sine interpolation algorithm and two-dimensional feature vector matching statistics, the problem of simultaneously sorting uneven and jittery PRI modulated radar pulse signals in existing technologies has been solved, achieving efficient and accurate radar signal sorting.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIDIAN UNIV
- Filing Date
- 2023-02-20
- Publication Date
- 2026-05-26
Smart Images
Figure CN116087883B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of radar communication technology, and more specifically relates to a two-dimensional feature vector-based anti-jitter pulse repetition interval (PRI) sorting method in the field of electronic reconnaissance technology. This invention can be used to sort pulse signals received by radar during radar reconnaissance. Background Technology
[0002] Radar signal sorting is a crucial component of radar reconnaissance. In radar reconnaissance, it is necessary to sort the large number of radar pulse signals captured by the radar receiving system from the electromagnetic environment, providing accurate information for implementing jamming strategies against target radars. With the increasing diversification of radar modes and applications, the PRI (Primary Intensity Level) variations of radar pulse signals have become increasingly complex. Furthermore, the density of radar pulse signals has increased, leading to higher demands on sorting efficiency. Therefore, developing radar signal sorting methods to overcome these challenges is of paramount importance.
[0003] In his paper "Signal Sorting and Identification of Radar Radiation Sources in Complex Environments" (Chengdu: University of Electronic Science and Technology of China, 2022, Master's Thesis), Han Jinxin proposed an improved PRI transform method using a least-squares segmentation algorithm. This method first performs gradient analysis on the radar pulse signal and sets a gradient threshold. Then, it segments the pulse sequence based on points with large gradient changes. After segmentation, it uses a least-squares algorithm to reduce the impact of PRI jitter on the signal. Finally, it performs a PRI transform on the processed signal to identify potential PRI values, thus completing the sorting process. While this method reduces the impact of jittered PRI on the radar pulse signal using the least-squares algorithm and can sort out jittered PRI-modulated radar pulse signals, it disrupts the PRI variation pattern of unevenly modulated radar pulse signals. This means that the method cannot simultaneously sort out jittered and unevenly modulated radar pulse signals.
[0004] The Shandong Aerospace Electronic Technology Research Institute disclosed a non-uniform quantization sorting method for PRI jitter signal sequences in its patent application "A Non-Uniform Quantization Sorting Method for PRI Jitter Signal Sequence Difference" (application number 201711260049.0, authorization announcement number CN 108181613 B). The method involves quantizing the arrival time series difference results within a set range into integer values to obtain quantized TOA difference results. Then, histogram statistics are performed on the quantized TOA difference results, and values exceeding a preset threshold in the histogram are taken as potential PRI values. Sequence retrieval confirms the potential PRI values as true PRI values, and the pulse sequence corresponding to the true PRI values is extracted from the PRI jitter signal to achieve sorting of the PRI jitter signal sequence. This sorting method can obtain jitter PRI modulated signals with large PRI values, but when the PRI value is small, or when radar signal pulses are dense, the non-uniform quantization method is less effective, reducing the accuracy of PRI estimation and leading to a decrease in the correct sorting rate.
[0005] Xi'an University of Electronic Science and Technology disclosed a comprehensive radar signal sorting method in its patent application "Radar Signal Sorting Method, Device, Computer Equipment and Storage Medium" (Application No. 201910173974.2, Publication No.: CN 109683143 A). The method involves the following steps: using a two-dimensional feature vector matching algorithm to match the feature vectors of the pulse dataset and extracting the successfully matched pulse sequences; using an improved PRI transformation method to detect and sort the remaining pulse sequences that exceed a threshold, obtaining the corresponding pulse sequences to complete the sorting of the radar pulse signals; if the threshold is not exceeded, the remaining pulses are considered insufficient to sort out other PRI modulation type radiation sources, thus completing the sorting process. This method uses a dynamic two-dimensional feature vector matching algorithm to separate fixed PRI modulated radar pulse signals and staggered PRI modulated radar pulse signals, and then uses an improved PRI transform method to separate jittered PRI modulated radar pulse signals. It can simultaneously separate staggered PRI modulated radar pulse signals and jittered PRI modulated radar pulse signals. However, the method counts the time interval matching of all pulses during the sorting process, which increases the computational load of sorting and leads to a decrease in sorting efficiency.
[0006] In summary, for radar signal sorting methods in existing electronic reconnaissance applications, the existing methods do not achieve ideal sorting results. They are limited by the PRI modulation pattern, making it difficult to simultaneously sort staggered PRI modulation and jittered PRI modulation radar pulse signals, and the correct sorting rate for jittered PRI modulation radar pulse signals is low. At the same time, there is also the problem of excessively long sorting time. Summary of the Invention
[0007] The purpose of this invention is to address the shortcomings of the prior art by proposing a jitter-resistant PRI sorting method based on two-dimensional feature vectors. This method solves the problems of difficulty in simultaneously sorting staggered PRI-modulated and jitter-modulated PRI-modulated radar pulse signals, poor jitter resistance, and low sorting efficiency in existing sorting methods.
[0008] The specific approach to achieving the objective of this invention is as follows: This invention uses a sine interpolation algorithm to estimate the potential PRI value, and constructs a two-dimensional feature vector using two parameters: pulse arrival time and PRI estimate. The two-dimensional feature vector is then used to perform matching statistics on the pulse sequence. This statistical approach reduces the impact of jittered PRI modulation signals on sorting, thereby making the sorting method anti-jitter and capable of sorting fixed, uneven, and jittered PRI modulated radar pulse signals. It also achieves a high correct sorting rate, solving the problems of existing sorting algorithms, such as difficulty in simultaneously sorting uneven and jittered PRI modulated radar pulse signals and low correct sorting rates. In generating the two-dimensional feature vector, this invention uses the PRI estimate obtained by the sine interpolation algorithm as the time interval, reducing the matching amount during pulse sorting, thus lowering the time complexity of sorting and overcoming the problem of low sorting efficiency.
[0009] To achieve the above objectives, the technical solution adopted by the invention includes the following steps:
[0010] Step 1, perform PRI estimation on the arrival time series:
[0011] Step 1.1: Transform the arrival time series into a square sinusoidal interpolation function;
[0012] Step 1.2: Perform a fast Fourier transform on the square sinusoidal interpolation function, and select the reciprocal of the x-coordinate value corresponding to the maximum peak value in the transformed spectrum as a PRI estimate of the arrival time series.
[0013] Step 2, construct the two-dimensional feature vector:
[0014] Step 2.1: Generate N two-dimensional feature vectors composed of the PRI estimate and each arrival time in the arrival time series. Generate the time series to be matched corresponding to each two-dimensional feature vector. The middle time of each time series to be matched corresponds to the arrival time of the two-dimensional feature vector. Each time interval of each time series to be matched is equal to the PRI estimate of the two-dimensional feature vector. The value of N is equal to the total number of arrival times in the arrival time series.
[0015] Step 2.2: Compare each arrival time in the arrival time series with each time in each time series to be matched. If the difference between the pairwise element values of corresponding times in the two series is less than 1% of the PRI estimate, then the two times are considered to be matched successfully. Count the number of successful matches for each time series to be matched.
[0016] Step 3: Sorting pulses using the successful matching value of each time series to be matched:
[0017] Determine whether the number of successful matches for each time series to be matched is greater than the sorting threshold. If so, extract the midpoint of the time series to be matched and determine the pulse corresponding to the midpoint. Combine all the determined pulses into a radar pulse signal.
[0018] Compared with the prior art, the present invention has the following advantages:
[0019] First, this invention uses a square sinusoidal interpolation algorithm to estimate the PRI modulation of radar pulse signals. A two-dimensional feature vector is constructed by comparing the PRI estimate with the arrival time series of the radar pulse signals. This two-dimensional feature vector is then used to match and extract the arrival time series, enabling the simultaneous sorting of fixed, staggered, and jittered PRI modulated radar pulse signals. This overcomes the deficiency in existing sorting methods that struggle to simultaneously sort staggered and jittered PRI modulated radar pulse signals.
[0020] Secondly, by utilizing the statistical properties of two-dimensional feature vectors, this invention overcomes the shortcomings of existing sorting methods in terms of poor anti-jitter capability, enabling it to achieve good sorting performance when faced with jittery PRI-modulated radar pulse signals.
[0021] Third, when generating two-dimensional feature vectors, this invention uses the PRI estimation algorithm to obtain the PRI estimation as the time interval, which reduces the number of time sequences to be matched, thereby reducing the time used for subsequent matching. This overcomes the inefficiency of radar pulse signals when sorting, enabling this invention to sort radar pulse signals with longer durations when faced with dense radar pulse signals. Attached Figure Description
[0022] Figure 1 This is a flowchart of the present invention;
[0023] Figure 2 This is a schematic diagram of the sorting process of the present invention. Detailed Implementation
[0024] To more clearly illustrate the present invention, a further detailed description is provided below in conjunction with the accompanying drawings and embodiments. Obviously, the drawings and embodiments described below are merely some embodiments of the present invention, and not all embodiments. For those skilled in the art, other drawings can be obtained based on these drawings without any inventive effort. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0025] Reference Figure 1 The specific implementation steps of the present invention will be described in further detail below.
[0026] Step 1: Perform PRI estimation on the arrival time series.
[0027] Step 1: Transform the arrival time series into a sine square interpolation function, as follows:
[0028]
[0029] Where s(·) represents the sine interpolation function with an initial state of 0, ∑ is the summation operation, i represents the index of the arrival time, i = 1, 2, ..., N, and N represents the total number of arrival times in the arrival time series, Δt i This represents the first difference of the arrival time series, t. i Let represent the i-th arrival time in the arrival time series, sin(·) represents the sine function, and π represents pi.
[0030] Step 2: Perform a fast Fourier transform on the square sinusoidal interpolation function, and select the reciprocal of the x-coordinate value corresponding to the maximum peak value in the transformed spectrum as a PRI estimate of the arrival time series.
[0031] Step 2: Construct a two-dimensional feature vector.
[0032] Step 1: Generate N two-dimensional feature vectors consisting of the PRI estimate and each arrival time in the arrival time series. Generate a time series to be matched for each two-dimensional feature vector. The midpoint of each time series to be matched corresponds to the arrival time of the two-dimensional feature vector. Each time interval of each time series to be matched is equal to the PRI estimate of the two-dimensional feature vector. The value of N is equal to the total number of arrival times in the arrival time series.
[0033] Step 2: Compare each arrival time in the arrival time series with each time in each time series to be matched. If the difference between the pairwise element values of corresponding times in the two series is less than 1% of the PRI estimate, then the two times are considered to be matched successfully. Count the number of successful matches for each time series to be matched.
[0034] Step 3: Sorting pulses using the successful matching value of each time series to be matched.
[0035] Determine if the number of successful matches for each time series to be matched is greater than the sorting threshold. If so, extract the midpoint of the time series to be matched and determine the pulse corresponding to that midpoint. Combine all the determined pulses into a radar pulse signal. The sorting threshold is as follows:
[0036]
[0037] Where FP represents the pulse sorting threshold, γ represents the jitter rate of the radar pulse signal, and P represents the total number of moments in the time series to be matched.
[0038] Combination Figure 2 The method of sorting pulses using the successful matching value of each time series to be matched is further described.
[0039] Figure 2 (a) is a schematic diagram of the arrival time series. Figure 2 In (a), the horizontal axis represents time, the nine upward arrows represent pulses, and the t below the pulse... i Let t3 represent the arrival time of the i-th pulse, where i = 1, 2, ..., 9. The arrival time of the 3rd pulse in the arrival time series is compared with the two-dimensional feature vector generated by the PRI estimate. The t3 in the generated two-dimensional feature vector corresponds to... Figure 2 The arrival time of the third pulse in (a) corresponds to the PRI estimate obtained in step 1. Figure 2 (b) is a schematic diagram of the time series to be matched generated corresponding to the two-dimensional feature vector at time t3. Figure 2 (b) The horizontal axis represents time, the five upward arrows represent pulses, the midpoint time corresponds to the eigenvalue t3 of the two-dimensional eigenvector, and the interval time PRI corresponds to the PRI estimate of the two-dimensional eigenvector. Figure 2 (a) arrival time series and Figure 2 (b) Match the time series to be matched. The two series correspond to time t3. If the number of successful matches is greater than the sorting threshold, then the pulse at time t3 belongs to the radar with PRI value estimated by PRI.
[0040] The effects of this invention will be further illustrated below with simulation experiments:
[0041] 1. Simulation experimental conditions:
[0042] The hardware platform for the simulation experiment of this invention is: an Intel i5 6300HQ CPU with a main frequency of 2.3GHz and 16GB of memory.
[0043] The software platform for the simulation experiment of this invention is: Windows 10 operating system and MATLAB R2021a.
[0044] The input parameters applicable to the simulation experiment of this invention are as follows:
[0045] In the simulation experiment of this invention, three airborne multi-function radars were set up, and the parameters of each radar are shown in Table 1.
[0046] Table 1. Pulse signal parameters of the three radars
[0047] Radar number PRI modulation style Pulse start time / μs Pulse repetition period / μs jitter rate Radar 1 Fixed PRI modulation 23 100 0.05％ Radar 2 Jitter PRI modulation 12 135 6％ Radar 3 Staggered PRI modulation 34 77、88、95 0.05％
[0048] 2. Simulation content and result analysis:
[0049] The simulation experiment of this invention uses the present invention and a prior art (dynamic two-dimensional feature vector matching method) to perform PRI sorting on a radar pulse signal with a pulse sequence length of 1000 pulses from an airborne multi-function radar. The radar pulse signal used in the simulation experiment consists of pulse signals transmitted by three airborne multi-function radars, with PRI modulation types of fixed, jitter, and staggered PRI modulation, respectively. The sorting results of the simulated pulse signal are shown in Table 2.
[0050] The existing technology (dynamic two-dimensional feature vector matching method) used in the simulation experiment refers to the PRI sorting method using dynamic two-dimensional feature vector matching, which was proposed by Wang Yining in his paper "Research on Sorting Technology of Complex Radar Signals" (Xi'an University of Electronic Science and Technology, Master's Thesis, 2019).
[0051] The sorting results obtained from the simulation experiment of this invention are shown in Table 2.
[0052] Table 2. Summary of Pulse Sorting Results
[0053]
[0054] In Table 2, the PRI estimate refers to the estimated PRI value of the sorting method for radar pulse signals, the number of pulses refers to the total number of pulse signals emitted by a radar, the correct sorting refers to the number of pulses whose sorting results match the actual results after using the sorting method, and the sorting time refers to the time taken to sort radar pulse signals using the sorting method.
[0055] As shown in Table 2, the sorting results of the present invention, compared with those of the prior art, correctly sort a greater total number of pulses and exhibit better sorting performance for radar pulse signals modulated by jittered PRI, proving that the anti-jitter capability of the present invention is superior to that of the prior art sorting method. Furthermore, the sorting operation time of the present invention is significantly shorter than that of the prior art, resulting in higher sorting efficiency.
[0056] The simulation experiments above show that this method, using the sine interpolation algorithm, can extract the frame period of the radar pulse signal and perform matching statistics on the radar pulse signal using two-dimensional feature vectors. It can identify radar pulse signals with fixed, staggered, and jittered PRI modulation types, solving the defects of existing methods that are difficult to simultaneously sort radar pulse signals with fixed, staggered, and jittered PRI modulation types and have poor anti-jitter capabilities. Using the PRI estimation algorithm to obtain the PRI estimation as the time interval reduces the number of time series to be matched, thereby reducing the time used for subsequent matching and solving the defect of low efficiency in sorting radar pulse signals in existing technologies.
Claims
1. A method for sorting pulses with anti-jitter repetition intervals based on two-dimensional feature vectors, characterized in that, A two-dimensional feature vector is generated by using the estimated pulse repetition interval (PRI) obtained through an estimation algorithm and the arrival time series of the radar pulse signal. This two-dimensional feature vector is then used to generate a matching time series and an arrival time series for sorting. The steps of this sorting method include the following: Step 1, perform PRI estimation on the arrival time series: Step 1.1: Transform the arrival time series into a square sinusoidal interpolation function; Step 1.2: Perform a fast Fourier transform on the square sinusoidal interpolation function, and select the reciprocal of the x-coordinate value corresponding to the maximum peak value in the transformed spectrum as a PRI estimate of the arrival time series. Step 2, construct the two-dimensional feature vector: Step 2.1: Generate N two-dimensional feature vectors composed of the PRI estimate and each arrival time in the arrival time series. Generate the time series to be matched corresponding to each two-dimensional feature vector. The middle time of each time series to be matched corresponds to the arrival time of the two-dimensional feature vector. Each time interval of each time series to be matched is equal to the PRI estimate of the two-dimensional feature vector. The value of N is equal to the total number of arrival times in the arrival time series. Step 2.2: Compare each arrival time in the arrival time series with each time in each time series to be matched. If the difference between the pairwise element values of corresponding times in the two series is less than 1% of the PRI estimate, then the two times are considered to be matched successfully. Count the number of successful matches for each time series to be matched. Step 3: Sorting pulses using the successful matching value of each time series to be matched: Determine whether the number of successful matches for each time series to be matched is greater than the sorting threshold. If so, extract the midpoint of the time series to be matched and determine the pulse corresponding to the midpoint. Combine all the determined pulses into a radar pulse signal.
2. The anti-jitter pulse repetition interval sorting method based on two-dimensional feature vectors according to claim 1, characterized in that, The sine interpolation function mentioned in step 1.1 is as follows: Where s(·) represents the sine interpolation function with an initial state of 0, ∑ is the summation operation, i represents the index of the arrival time, i = 1, 2, ..., N, and N represents the total number of arrival times in the arrival time series, Δt i This represents the first difference of the arrival time series, t. i Let represent the i-th arrival time in the arrival time series, sin(·) represents the sine function, and π represents pi.
3. The anti-jitter pulse repetition interval sorting method based on two-dimensional feature vectors according to claim 1, characterized in that, The sorting threshold value mentioned in step 3 is obtained by the following formula: Where FP represents the pulse sorting threshold, γ represents the jitter rate of the radar pulse signal, and P represents the total number of moments in the time series to be matched.