A radar signal pulse repetition time blind estimation method and device and a storage medium

Abstract

The application discloses a radar signal pulse repetition time blind estimation method and device and a storage medium. With the aid of the characteristics of optimal phase correlation accumulation data, the application can adaptively solve the PRT of pulses and the start and end time of data under ultra-low signal-to-noise ratio and is not limited by waveforms. In addition, the application can keep the data dimension from expanding and can effectively adapt to parallel acceleration calculation. The application groups, corrects and accumulates the data pulse string to be processed, uses the signal-to-noise ratio after accumulation, average accumulation efficiency and energy concentration to design a target function to search for an optimal grouping length, finally solves the PRT of pulses, improves the adaptability of electronic reconnaissance technology in a non-cooperative and low signal-to-noise ratio environment, provides strong data enhancement support for subsequent target radiation source fine intra-pulse modulation feature recognition and extraction technology, and can be widely applied to the technical field of computers.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a blind estimation method, apparatus and storage medium for radar signal pulse repetition time. Background Technology

[0002] Radar is widely used in military and civilian fields such as target detection, tracking, and imaging, bringing great convenience to people's work and life. Since radar itself is a product driven by military purposes and intentions, countering radar requires a focus on developing radar radiation source reconnaissance signal processing technology. In the current radar electronic reconnaissance signal processing environment, reconnaissance equipment is often within the sidelobe illumination range of the radar radiation source, and the transmission distance is relatively long, so low signal-to-noise ratio (SNR) of the received signal is the norm. Therefore, to improve our own electronic reconnaissance signal processing capabilities, the ability to perceive information and extract parameters from pulse signals under low SNR conditions is a pressing issue that needs to be addressed. Among these, pulse repetition time (PRT), as the most critical and typical radar signal modulation parameter, has always been a parameter that the reconnaissance adversary hopes to accurately obtain.

[0003] Traditional PRT parameter estimation uses the radar's pulse arrival time (TOA) as a reference, estimating the target signal's PRT through first-order difference estimation of the TOA. This method requires the radar pulse to be detected; however, in real-world electronic reconnaissance environments, due to low signal-to-noise ratios and non-cooperative targets, radar pulses and their TOA are difficult to obtain directly. Therefore, PRT estimation algorithms are often accompanied by radar pulse signal accumulation detection algorithms. Commonly used methods include time-domain cross-correlation, time-domain shifted autocorrelation accumulation detection, and time-frequency analysis. The time-domain cross-correlation method involves cross-correlating a segment of data with the original data, using signal correlation to compress energy and obtain the cross-correlation peak, and then using the difference obtained by detecting the cross-correlation peak to estimate the pulse time-to-arrival (TOA). The time-domain shift autocorrelation accumulation detection method first accumulates the pulse signal envelope based on the signal correlation to improve the signal-to-noise ratio (SNR) of the data to be processed, and then uses the high SNR pulse envelope to estimate the pulse time-to-arrival (TOA) and obtain the PRT. The time-frequency analysis method utilizes the gain of the data after Fourier transform, and by combining the two-dimensional distribution of time and frequency, it enhances the target signal in the noise to achieve effective signal detection and PRT estimation.

[0004] However, existing technologies have the following drawbacks: 1. For time-domain cross-correlation methods, parameters such as the length and position of the intercepted data will affect the cross-correlation results and performance. In electronic reconnaissance environments targeting non-cooperative targets, the lack of prior information makes it difficult to guarantee that the selection of these parameters will achieve optimal detection results. Furthermore, with the decrease in signal-to-noise ratio and improper selection of intercepted parameters, the algorithm's performance will rapidly decline. 2. For time-domain shifted autocorrelation accumulation detection methods, the accumulation effect of these methods depends on the signal-to-noise ratio of the original data. When the signal-to-noise ratio of the original data is low (<-0dB), the performance of this algorithm will suffer. Poor accumulation results will also affect the accuracy of subsequent pulse detection and PRT estimation. Moreover, these algorithms also depend on the target's transmitted waveform; when the transmitted waveform is a discrete frequency coded (DFC) signal, the algorithm's performance will also be compromised. 3. For time-frequency analysis algorithms, these algorithms have poor noise resistance. When the signal-to-noise ratio is low, it will significantly affect the subsequent pulse detection performance. In addition, these algorithms transform the original one-dimensional data into a two-dimensional distribution of time and frequency, which greatly increases the computational load of data processing and places higher demands on radar hardware. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a blind estimation method, apparatus and storage medium for radar signal pulse repetition time, which can efficiently perform blind estimation of radar signal pulse repetition time.

[0006] On one hand, embodiments of the present invention provide a blind estimation method for radar signal pulse repetition time, including:

[0007] Obtain the original pulse train and preset initial particle parameters, and use the initial particle parameters as the first particle parameters; the first particle parameters include pulse repetition time, start time and end time;

[0008] The original pulse train is truncated and grouped using the first particle parameter to obtain pulse groups; each pulse group includes several pulse signals.

[0009] A reference signal is obtained from the pulse group, pulse shift correction is performed on each pulse signal based on the reference signal, and phase compensation processing is performed on each pulse signal based on the reference signal; then the pulse signals are superimposed to obtain the accumulation result.

[0010] The objective function value of the accumulation result is obtained by weighted summation of the target parameters of the accumulation result; the target parameters include the signal-to-noise ratio after accumulation, the average accumulation efficiency, and the energy concentration degree.

[0011] Based on the objective function value, the first particle parameter is updated to obtain the updated first particle parameter; then, the original pulse train is truncated and grouped using the first particle parameter to obtain the pulse group, until the preset convergence condition is reached; then, the objective function value obtained based on each first particle parameter is compared, and the target particle parameter is obtained based on the comparison result. The blind estimation result of the pulse repetition time is extracted from the target particle parameter.

[0012] Optionally, the original pulse train is truncated and grouped according to the first particle parameter to obtain pulse groups, including:

[0013] The original pulse train is truncated using the start and end times to obtain the pulse train to be processed;

[0014] The pulse train to be processed is grouped according to the pulse repetition time interval to obtain pulse groups.

[0015] Optionally, pulse shift correction is performed on each pulse signal based on the reference signal, including:

[0016] Based on the reference signal, the time shift difference between each pulse signal and the reference signal is obtained by cross-correlation peak detection;

[0017] Based on the time shift difference, the corresponding pulse signal is time-shifted and corrected to obtain the pulse signal after pulse shift correction.

[0018] Optionally, based on the time shift difference, the corresponding pulse signal is time-shifted to obtain a pulse signal with pulse shift correction, including:

[0019] The pulse signal is subjected to a Fourier transform, and then the result of the Fourier transform of the pulse signal is shifted and compensated by the corresponding time shift difference of the pulse signal.

[0020] The result of the shift compensation is subjected to inverse Fourier transform to obtain the pulse signal after pulse shift correction.

[0021] Optionally, phase compensation processing is performed on each pulse signal based on the reference signal, including:

[0022] Based on the product of the reference signal and the conjugate signal of each pulse signal, the phase difference between each pulse signal and the reference signal is obtained by taking the phase.

[0023] Based on the phase difference, phase compensation is performed on the corresponding pulse signal to obtain the phase-compensated pulse signal.

[0024] Optionally, the method further includes:

[0025] The accumulation results are sorted according to data power. Based on the sorting results, the noise average power estimate and the average power estimate of the accumulation results are obtained. Then, the signal-to-noise ratio after accumulation is obtained by calculating the exponent of the ratio of the noise average power estimate and the average power estimate.

[0026] The average accumulation efficiency is obtained by the ratio of the accumulated signal-to-noise ratio to the number of pulse signals in the pulse group.

[0027] The energy concentration degree is determined based on the number of pulses within the repetition time of a single pulse in the accumulated results, combined with preset conditions.

[0028] Optionally, the parameters of the first particle are updated based on the objective function value, including:

[0029] The first particle parameters are updated by comparing the objective function values ​​obtained from the first particle parameters in each historical iteration with those obtained from the first particle parameters in the current iteration, combined with a random matrix and a learning factor.

[0030] On the other hand, embodiments of the present invention provide a blind estimation device for radar signal pulse repetition time, comprising:

[0031] The first module is used to acquire the original pulse train and the preset initial particle parameters, and to use the initial particle parameters as the first particle parameters; the first particle parameters include the pulse repetition time, the start time and the end time;

[0032] The second module is used to truncate and group the original pulse train according to the first particle parameter to obtain a pulse group; the pulse group includes several pulse signals.

[0033] The third module is used to obtain a reference signal from the pulse group, perform pulse shift correction on each pulse signal based on the reference signal, and perform phase compensation processing on each pulse signal based on the reference signal; then, the pulse signals are superimposed to obtain the accumulation result.

[0034] The fourth module is used to perform a weighted summation of the target parameters of the accumulation results to obtain the objective function value of the accumulation results; the target parameters include the signal-to-noise ratio after accumulation, the average accumulation efficiency, and the energy concentration degree.

[0035] The fifth module is used to update the first particle parameters based on the objective function value to obtain the updated first particle parameters; then return to the second module until the preset convergence condition is reached; then compare the objective function values ​​obtained based on each first particle parameter, obtain the target particle parameters based on the comparison results, and extract the blind estimation result of the pulse repetition time from the target particle parameters.

[0036] Optionally, the device further includes:

[0037] The sixth module is used to sort the accumulation results according to the data power, and obtain the noise average power estimate and average power estimate of the accumulation results based on the sorting results; then, the signal-to-noise ratio after accumulation is obtained by calculating the exponent of the ratio of the noise average power estimate and the average power estimate.

[0038] The seventh module is used to obtain the average accumulation efficiency based on the ratio of the accumulated signal-to-noise ratio to the number of pulse signals in the pulse group.

[0039] The eighth module is used to determine the energy concentration degree based on the number of pulses within the repetition time of a single pulse in the accumulation results, combined with preset conditions.

[0040] On the other hand, embodiments of the present invention provide an electronic device, including a processor and a memory;

[0041] Memory is used to store programs;

[0042] The processor executes the program as described above.

[0043] On the other hand, embodiments of the present invention provide a computer-readable storage medium storing a program that is executed by a processor to implement the method described above.

[0044] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0045] This invention first acquires the original pulse train and preset initial particle parameters, using the initial particle parameters as the first particle parameters. The first particle parameters include pulse repetition time, start time, and end time. The original pulse train is truncated and grouped using the first particle parameters to obtain pulse groups. Each pulse group includes several pulse signals. A reference signal is acquired from the pulse group, and pulse shift correction is performed on each pulse signal based on the reference signal. Phase compensation processing is also performed on each pulse signal based on the reference signal. The pulse signals are then superimposed to obtain an accumulation result. The target parameters of the accumulation result are weighted and summed to obtain the target function value of the accumulation result. The target parameters include the signal-to-noise ratio after accumulation, average accumulation efficiency, and energy concentration. The first particle parameters are updated based on the target function value to obtain the updated first particle parameters. Then, the process of truncating and grouping the original pulse train using the first particle parameters to obtain pulse groups is repeated until a preset convergence condition is reached. Finally, the target function values ​​obtained based on each first particle parameter are compared, and the target particle parameters are obtained based on the comparison results. The blind estimation result of the pulse repetition time is extracted from the target particle parameters. This invention, through iterative iteration of particle parameters and leveraging the characteristics of the coherently accumulated data, can adaptively solve for the pulse repetition time and start / end time of data under ultra-low signal-to-noise ratio conditions, without being limited by waveform influence. Furthermore, the method maintains the data dimension without expansion and effectively adapts to parallel accelerated computation. This invention enables efficient blind estimation of radar signal pulse repetition time. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 A flowchart illustrating a blind estimation method for radar signal pulse repetition time provided in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram comparing signal and noise power under different signal-to-noise ratios provided in an embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of a received pulse train under non-cooperative detection provided in an embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of the overall process of the blind estimation method for radar signal pulse repetition time provided in an embodiment of the present invention;

[0051] Figure 5(a) is a schematic diagram of the time-domain waveform of the simulated signal provided in an embodiment of the present invention;

[0052] Figure 5(b) is a schematic diagram of the time and frequency of the simulated signal provided in the embodiment of the present invention;

[0053] Figure 5(c) is a schematic diagram of the time-domain waveform of a partial simulated signal provided in an embodiment of the present invention;

[0054] Figure 5(d) is a schematic diagram of the time and frequency of some simulated signals provided in the embodiment of the present invention;

[0055] Figure 6 The iterative curve of the particle swarm optimization algorithm provided in the embodiments of the present invention;

[0056] Figure 7 A schematic diagram illustrating the pulse accumulation results under PRT estimation provided in an embodiment of the present invention;

[0057] Figure 8 A schematic diagram of a blind estimation device for radar signal pulse repetition time provided in an embodiment of the present invention;

[0058] Figure 9 This is a schematic diagram of the frame of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0060] First, it should be noted that, for the sake of understanding the technical solution of this invention, the technical terms that may appear in this invention will be explained:

[0061] Pulse Repetition Time (PRT);

[0062] Pulse arrival time (TOA);

[0063] Discrete Frequency Coded (DFC) signal;

[0064] Signal-to-noise ratio after accumulation (SNRAA);

[0065] Average Accumulation Efficiency (AAE);

[0066] Energy Concentration (EC);

[0067] Linear Frequency Modulation (LFM) signal;

[0068] Particle Swarm Optimization (PSO).

[0069] On the one hand, such as Figure 1 As shown, an embodiment of the present invention provides a blind estimation method for radar signal pulse repetition time, including:

[0070] S100: Obtain the original pulse train and the preset initial particle parameters, and use the initial particle parameters as the first particle parameters;

[0071] It should be noted that the first particle parameters include the pulse repetition time, start time, and end time.

[0072] In some specific embodiments, such as Figure 2 The diagram shows a schematic of the pulse train emitted by the target radiation source. When the SNR (Signal-to-Noise Ratio) ≤ 0dB, the real pulse signal will be submerged in the noise environment due to its low power. In this embodiment of the invention, the data pulse train is grouped and accumulated according to the PRT estimate under this condition, and the PRT estimate result corresponding to the data pulse train is obtained through optimization iteration.

[0073] Because the target radiation source is a non-cooperative target and is limited by hardware sampling equipment (fixed-length sampling, timed sampling), the intercepted target pulse train may have the following characteristics: Figure 3 The following are several scenarios. Under optimal truncation, the start and end of the data occur during the idle period of the pulse. Grouping the data according to the actual PRT based on the current start and end times achieves the best accumulation effect. However, when the data string has pre-pulse and post-pulse truncation, grouping the data according to the actual PRT based on the current start and end times will split the pulse at both ends of the data. Although a good signal-to-noise ratio accumulation effect can be obtained, a complete pulse signal cannot be obtained, which is detrimental to subsequent signal analysis and processing. In addition, when redundant truncation occurs, many data points without pulses will also be accumulated, resulting in excessive noise accumulation, which is not conducive to improving the signal-to-noise ratio and ultimately affects the optimization search performance of the algorithm's PRT. Therefore, in the iterative optimization design of the algorithm, not only should the target PRT estimate be considered, but the start and end times of data truncation should also be appropriately adjusted as parameters to obtain the best estimation result.

[0074] Therefore, based on the above analysis, the particle parameter p is defined as (PRT, T). s ,T e ), where PRT is the pulse repetition time corresponding to the current particle, and T s T represents the start time of the data. e The deadline for data is set. The variation range of each parameter is defined, and the variation space of each particle is shown in equation (1). Before the first optimization search, the particles are initialized, resulting in the initialized particle p0 = (PRT0, T s0 ,T e0 ), and then the updated particle parameters p are used.

[0075] p min =(PRT) min ,T smin ,T emin )≤p≤p max =(PRT) max ,T smax ,T emax (1)

[0076] Where, p min =(PRT) min ,T smin ,T emin ) represents the minimum value of the parameter particle vector, p max =(PRT) max ,T smax ,T emax ) represents the maximum value of the parameter particle vector.

[0077] S200: The original pulse train is truncated and grouped according to the first particle parameter to obtain pulse groups;

[0078] It should be noted that the pulse group includes several pulse signals; in some embodiments, step S200 may include: truncating the original pulse train using the start time and end time to obtain the pulse train to be processed; and grouping the pulse train to be processed into pulse groups with the pulse repetition time as the interval.

[0079] In some specific embodiments, assuming the detected data pulse train is f(t), f(t) is first truncated using the particle parameter p, and the pulse train to be processed under the current particle is obtained according to equation (2). p (t).

[0080] f p (t)=f(T s :T e (2)

[0081] Subsequently, according to the interval PRT, f pGrouping (t) into data groups yields S = [s1(t), s2(t), ..., s2(t)]. m (t),L,s M [(t)], where m = 1, 2, 3, L, M, and M is the maximum number of groups under the current PRT.

[0082] S300: Obtain a reference signal from the pulse group, perform pulse shift correction on each pulse signal based on the reference signal, and perform phase compensation processing on each pulse signal based on the reference signal; then perform superposition processing on each pulse signal to obtain the accumulation result;

[0083] It should be noted that, in some embodiments, pulse shift correction of each pulse signal based on the reference signal may include: obtaining the time shift difference between each pulse signal and the reference signal by cross-correlation peak detection based on the reference signal; and performing time shift correction on the corresponding pulse signal according to the time shift difference to obtain the pulse signal after pulse shift correction.

[0084] In some embodiments, time shift correction is performed on the corresponding pulse signal based on the time shift difference to obtain a pulse signal after pulse shift correction. This may include: performing a Fourier transform on the pulse signal, and then performing shift compensation on the result of the Fourier transform of the pulse signal based on the corresponding time shift difference of the pulse signal; performing an inverse Fourier transform on the result of the shift compensation to obtain a pulse signal after pulse shift correction.

[0085] In some embodiments, performing phase compensation processing on each pulse signal based on a reference signal may include: obtaining the phase difference between each pulse signal and the reference signal by taking the phase difference through a phase taking operation based on the product of the reference signal and the conjugate signal of each pulse signal; and performing phase compensation on the corresponding pulse signal according to the phase difference to obtain the phase-compensated pulse signal.

[0086] In some specific embodiments, step S300 can be implemented through the following process:

[0087] After data grouping, to improve the tolerance for PRT errors during optimization, pulse shift correction and phase compensation are performed before pulse accumulation to achieve the optimal accumulation effect under the current PRT estimate. s is selected from pulse group S. i (t) serves as a reference signal, and in this embodiment of the invention, i = 1 by default.

[0088] Due to the presence of PRT error, the pulse signals within each group are not strictly aligned after data grouping. To achieve better accumulation, time-shift alignment correction must be performed on the data. Considering only the effect of time shift, the pulses s of other groups of data... j (t) and reference signal s i The relationship between (t) can be expressed by equation (3).

[0089] s i (t)=s j (t-Δt j (3)

[0090] Where j = 1, 2, 3, L, M, j ≠ i, Δt j denoted as the pulse time shift difference between pulse group signal j and reference pulse group signal i.

[0091] As can be seen from equation (3), there is only a time shift difference Δt between other pulse group signals and the reference signal. j By performing Δt on other pulse group signals j Shift correction processing can yield perfectly aligned grouped data. Shift correction can be achieved using cross-correlation and Fourier transform. Due to the correlation between signals, for other pulse group signals s... j (t) and reference signal s i After performing the cross-correlation processing as shown in equation (4), when τ=Δt j At that time, R ij (τ) reaches its peak value. Therefore, the pulse group signal s can be obtained by cross-correlation peak detection. j (t) and reference signal s i The time shift difference Δt between (t) j .

[0092]

[0093] in, s j The conjugate signal of (t).

[0094] Then, using the estimated time shift difference Δt j For s j (t) Time shift correction is performed, and this technique employs a more accurate frequency domain correction method. Utilizing the properties of the Fourier transform, the time shift is corrected for s. j After shift compensation (phase compensation) is performed on the Fourier transform result of (t), the pulse is accurately shifted through the inverse Fourier transform. The calculation method can be expressed by equation (5).

[0095]

[0096] Among them, S j (ω) is s j The Fourier transform result of (t), s′ j (t) is s j (t) The result after shift correction.

[0097] Let j = 1, 2, 3L, M, j ≠ i, and perform time-shift correction processing on each pulse group sequentially. After obtaining accurately aligned pulse data for each group, phase compensation processing is needed for each group of data to achieve optimal accumulation effect, so that coherent accumulation can be achieved when the pulse data are superimposed. For ease of description, the following will use s... j (t) represents other pulse group data that have undergone time-shift correction.

[0098] The actual received data consists of waveform signals and noise. The pulse group data s(t) calculated above can be expressed in the form of equation (6).

[0099]

[0100] Where A represents the signal amplitude. Let n(t) be the modulation phase of the signal, and n(t) be Gaussian white noise.

[0101] For signal s j (t) and reference signal s i In terms of (t), the phase between them and They are not necessarily completely identical, therefore it is necessary to compare s j (t) Further phase compensation is performed. First, the phase difference of the signal is solved using equation (7).

[0102]

[0103] Where PH(·) represents the phase take operation, s j The conjugate signal of (t).

[0104] Then, the obtained phase difference is used to analyze the signals s of each pulse group. j (t), j=1,2,3L,M,j≠i are sequentially subjected to phase compensation processing, and the signal phase compensation operation can be represented by equation (8).

[0105]

[0106] After phase compensation, the signals s′ of each pulse group j The phase of (t) and the reference signal s i (t) remains consistent. At this point, the grouped data S is superimposed to obtain the accumulated result f under this PRT parameter. a (t).

[0107] S400. Perform a weighted summation of the target parameters of the accumulated results to obtain the target function value of the accumulated results;

[0108] It should be noted that the target parameters include the accumulated signal-to-noise ratio, average accumulation efficiency, and energy concentration. In some embodiments, the method may further include: sorting the accumulation results according to data power, obtaining the noise average power estimate and the average power estimate of the accumulation results based on the sorting results; then calculating the accumulated signal-to-noise ratio based on the exponent of the ratio of the noise average power estimate and the average power estimate; obtaining the average accumulation efficiency based on the ratio of the accumulated signal-to-noise ratio to the number of pulse signals in the pulse group; and determining the energy concentration based on the number of pulses within the repetition time of a single pulse in the accumulation results, combined with preset conditions.

[0109] In some specific embodiments, the method of the present invention introduces accumulated signal-to-noise ratio (SNRAA), average accumulation efficiency (AAE), and energy concentration degree (EC) as parameter factors in the design objective function. The design concept and definitions are as follows:

[0110] SNRAA: The optimization objective of this invention is to obtain the best coherent accumulation result, which is best reflected in the signal-to-noise ratio (SNR) of the accumulated data. Incorrect accumulation may not necessarily result in the detection of a valid pulse envelope; therefore, this invention uses the ratio of peak power to median power to approximate the SNR after accumulation. The median power of the accumulated data is taken as the noise average power estimate P. ne Then, the accumulated data power is sorted from largest to smallest, and the average of the first X power values ​​is taken as the average power estimate P of the accumulated signal. se Then the signal-to-noise ratio after data accumulation can be represented by equation (9).

[0111]

[0112] AAE: When the search PRT is 1 / 2, 1 / 3, etc. of the real PRT, a high value can be obtained under equation (9), which can easily cause the optimization search algorithm to fall into a local optimum. To avoid this situation, this technical solution introduces the average accumulation efficiency to balance the relationship between the accumulated signal-to-noise ratio and the number of accumulations. Assuming that a total of M groups of data are obtained under the current PRT, the average accumulation efficiency can be expressed as equation (10). When the search PRT is 1 / 2, 1 / 3, etc. of the subharmonic components of the real PRT, although an approximate accumulated signal-to-noise ratio result can be obtained, the average accumulation efficiency will be significantly reduced due to the increased number of groups, thereby avoiding the search for PRT from falling into the local optimum of the subharmonic components.

[0113]

[0114] EC: When accumulating pulses using PRT grouping, the concept of energy concentration is introduced for evaluation in order to obtain accurate data start and end times. Energy concentration is defined as the ability to detect only one pulse component within one PRT after accumulation. This index factor avoids pulse splitting caused by improper data truncation during the optimization process. The definition of energy concentration is shown in Equation (11), where n represents the number of pulses detected.

[0115]

[0116] Finally, the objective function of the particle swarm optimization algorithm is defined as follows:

[0117] Object=-(w1·SNRAA+w2·AAE-w3·EC) (12)

[0118] Where w1, w2, and w3 are the weights of each indicator factor, and their values ​​are determined by the specific task. In this technical solution, w1 = w2 = w3 = 1 is used by default.

[0119] Under this objective function, the particle parameters will tend to maximize the signal-to-noise ratio (SNRAA) after coherent accumulation, maximize the average accumulation efficiency (AAE), and optimize only one pulse signal, ultimately obtaining the optimal parameter result desired by this technical method.

[0120] S500. Based on the objective function value, update the first particle parameter to obtain the updated first particle parameter; then return to the step of truncating and grouping the original pulse train through the first particle parameter to obtain the pulse group, until the preset convergence condition is reached; then compare the objective function value obtained based on each first particle parameter, obtain the target particle parameter based on the comparison result, and extract the blind estimation result of the pulse repetition time from the target particle parameter.

[0121] It should be noted that, in some embodiments, updating the first particle parameter based on the objective function value may include: updating the first particle parameter by combining the objective function value obtained in each historical iteration with the objective function value obtained in the current iteration, and combining the result with the random matrix and the learning factor.

[0122] In some specific embodiments, after accumulation is achieved under the particle parameters, the accumulation result is evaluated, and the objective function value is calculated using the previously designed index factor according to equation (12).

[0123] a. If the maximum number of iterations has not been reached or the objective function value does not meet the minimum threshold, the particle parameters are updated. The update process is as follows:

[0124] 1) First, calculate the objective function value for each particle and its best position p. best By comparison, if the objective function value of the current particle is better, then the parameters of the current particle are updated to the optimal position p. best It's important to note that in particle swarm optimization (PSO), N particles are randomly scattered into a parameter space. Each particle's position represents a parameter vector. An objective function is calculated based on this parameter vector. By comparing the objective functions of all current particles, the direction and velocity of each particle's motion can be further determined, leading to the next iteration. Ultimately, one particle will reach its optimal parameter position. The best position for each particle refers to the optimal objective function value obtained by that particle within its historical trajectory; this is the best position for a single particle.

[0125] 2) Secondly, for each particle and the global optimal particle position g best The objective function value is compared; if the objective function value is better, the particle parameters are updated to reflect the global optimum position g. best The global optimal particle position refers to the historical optimal position of all particles. The particle position refers to the position of the particle in the parameter space, which is the parameter vector corresponding to the particle. The optimal position is calculated based on the objective function value corresponding to the particle position.

[0126] 3) Update the particle velocity and position using equations (13) and (14);

[0127] 4) End particle update and proceed with a new round of objective function calculation.

[0128] The formulas for updating the particle's position and velocity are shown below.

[0129] x i =x i +v i (13)

[0130]

[0131] Where, x i It is the position of the particle, v i ω is the particle's velocity, ω is the inertia factor with a non-negative value, rand(·) is a random matrix with the same dimension as the particle vector, and its random values ​​are between (0,1). c1 and c2 are learning factors.

[0132] After obtaining the new particle position, proceed to step S200 and its specific implementation process, recalculate using the current particle parameters, until the next judgment is made on whether the convergence condition has been met.

[0133] b. If the maximum number of iterations is reached or the objective function value meets the minimum set threshold, the optimization stops, and the current optimal particle parameters are given as the estimation result. The optimal PRT estimate, the start time and the end time of the pulse train are obtained, and the blind estimation of the signal pulse repetition time under low signal-to-noise ratio is successfully realized.

[0134] To illustrate the technical principles of the embodiments of the present invention in detail, the present invention will be further described below with reference to the accompanying drawings and some specific embodiments. It should be understood that the following is an explanation of the present invention and should not be regarded as a limitation of the present invention.

[0135] To address the shortcomings of existing technologies, the present invention aims to perform blind estimation of the pulse response time (PRT) of radiation source data under ultra-low signal-to-noise ratio (SNR) conditions, without any prior information or waveform limitations. The technical approach of this invention is as follows: when the PRT of a certain data segment is known, the data can be grouped according to the PRT, and then phase-aligned accumulation of each group of data is performed to achieve coherent accumulation, thereby obtaining pulse data with high SNR quality. During the group accumulation process, when the PRT is too small, the grouping result may result in pulse splitting, no pulse grouping, etc., leading to a decrease in accumulation efficiency and effect; while when the PRT is too large, the grouping result may result in multiple pulses in one group, leading to a negligible improvement in SNR. Therefore, only with an accurate PRT can the optimal pulse accumulation effect be achieved. Therefore, the estimation of PRT can be achieved by iteratively solving for the optimal coherent accumulation result. During the group accumulation process, the length of the group (the estimated value of PRT) and the data start time of the group (T) are considered. s ) and deadline (T) e These three parameters are key factors affecting the accumulation effect, and are set as independent variables in the solution process. To evaluate the accumulation effect, three indicators are introduced to design the objective function: signal-to-noise ratio after accumulation (SNRAA), average accumulation efficiency (AAE), and energy concentration (EC). Since the parameters have a large range of variation, in order to quickly solve for the optimal parameter estimation results, this technique introduces the particle swarm optimization algorithm for fast solution, effectively achieving accurate estimation of PRT.

[0136] To achieve the above objectives, such as Figure 4 As shown, the method flow adopted in this embodiment of the invention is as follows:

[0137] Step 1: Design the particle parameter variation space and build a particle swarm optimization processing framework;

[0138] Step 2: Under individual particle parameters, the data is truncated and grouped;

[0139] Step 3: Select a reference signal for the grouped data, perform shift alignment and phase compensation, and then accumulate the data from each group.

[0140] Step 4: Calculate the objective function of the coherent accumulation effect under the current particle, perform particle iterative optimization, and solve for the optimal PRT, data start time, and data end time.

[0141] It should be noted that in some specific embodiments, the effectiveness of this technical solution is further verified through simulation examples. The radar simulation waveform parameters are shown in Table 1. The radar waveform data obtained under these parameters is as follows: Figures 5(a) to 5(d) As shown in Figures 5(a) and 5(c), at an extremely low signal-to-noise ratio of -10dB, the envelope fluctuation of the pulse signal in the time domain is not significant, and the PRT of the target signal cannot be directly estimated. In Figures 5(b) and 5(d), the time-frequency ridge of the simulated waveform can be obtained using time-frequency analysis algorithms; however, in this high-noise environment, locating the pulse and estimating the PRT using the time-frequency analysis results is also not robust.

[0142] Table 1

[0143] Parameter name Parameter value Sampling rate 300MHz Simulated waveforms LFM Pulse repetition time 100us Pulse width 20us Starting frequency 0MHz Pulse bandwidth 100MHz Number of pulses 50 Signal-to-noise ratio -10dB

[0144] Using the method of this invention, PRT blind estimation processing was performed on the above signal data. The PRT parameter prediction range was set to [30µs~300µs], the data start time range was the first 1 / 10 of the data length, and the data end time range was the last 1 / 10 of the data length. The particle swarm size was set to 50, and the maximum number of iterations was set to 20. The estimated pulse repetition time of the data obtained through iterative calculation was 100.59µs, with a relative error rate of 0.59%, which showed high estimation accuracy. The final particle swarm optimization iteration curve and the pulse accumulation result under PRT estimation are shown below. Figure 6 and Figure 7 As shown. From Figure 6 and Figure 7 As can be seen, the Particle Swarm Optimization (PSO) algorithm can iteratively find the optimal pulse pulse response (PRT) parameters. Experimental results show that even at such a low signal-to-noise ratio (SNR), accurate pulse PRT can be obtained without significant computational power or prior information, demonstrating the feasibility and practical application value of the proposed method. Furthermore, by accumulating the PRT and start / end times of the data obtained using the proposed method, a complete pulse signal with a high SNR can be obtained. This overcomes the difficulties in processing signals from non-cooperative target radiation sources at ultra-low SNR, facilitating subsequent refined pulse feature parameter extraction and providing more information support in actual combat.

[0145] In summary, this invention addresses the problems of conventional algorithms by proposing a blind estimation method for radar signal pulse repetition time (PRT) for low signal-to-noise ratio (SNR) electronic reconnaissance signals. This method requires no prior parameter settings and, leveraging the characteristics of data after optimal coherent accumulation, can adaptively solve for the PRT and start / end times of the data under ultra-low SNR conditions, unaffected by waveform limitations. Furthermore, this method maintains the data dimension without expansion and effectively adapts to parallel accelerated computation. Compared to existing technologies, this method does not require extensive prior information on the radar hardware system and the target radiation source, and can achieve blind PRT estimation in low SNR environments where no target radiation source signal is detected—a significant advantage over traditional methods. This method involves grouping, correcting, and accumulating the data pulse train, and using the accumulated SNR, average accumulation efficiency, and energy concentration to design an objective function to search for the optimal grouping length, ultimately obtaining the PRT. This improves the adaptability of electronic reconnaissance technology in non-cooperative, low SNR environments and provides strong data augmentation support for subsequent refined intra-pulse modulation feature identification and extraction techniques for target radiation sources.

[0146] In radar electronic reconnaissance environments, received reconnaissance signals often exhibit low signal-to-noise ratio (SNR), making it difficult for reconnaissance processors to effectively detect and extract parameters from target radiation sources. This invention addresses this issue by utilizing the repetitive transmission characteristics of pulse-based radar radiation sources. It employs cross-correlation shift alignment and coherent accumulation techniques to perform pulse grouping and blind accumulation processing on low SNR received reconnaissance signals. A particle swarm optimization algorithm is then used to quickly and accurately determine the optimal grouping length, which is the pulse repetition period of the radar pulse signal. Obtaining the accurate pulse repetition period of the target radiation source without any prior information facilitates subsequent refined feature extraction, analysis, and countermeasure processing of the radiation source pulses.

[0147] On the other hand, such as Figure 8As shown, an embodiment of the present invention provides a blind estimation device 600 for radar signal pulse repetition time, comprising: a first module 610, used to acquire an original pulse train and preset initial particle parameters, and use the initial particle parameters as first particle parameters; the first particle parameters include pulse repetition time, start time and end time; a second module 620, used to truncate and group the original pulse train according to the first particle parameters to obtain pulse groups; the pulse groups include a plurality of pulse signals; and a third module 630, used to acquire a reference signal from the pulse groups, perform pulse shift correction on each pulse signal based on the reference signal, and perform phase compensation on each pulse signal based on the reference signal. The process involves: processing the pulse signals, superimposing them to obtain an accumulation result; a fourth module 640, which performs a weighted summation of the target parameters of the accumulation result to obtain the target function value of the accumulation result; the target parameters include the signal-to-noise ratio after accumulation, the average accumulation efficiency, and the energy concentration degree; a fifth module 650, which updates the first particle parameters based on the target function value to obtain the updated first particle parameters; then returning to the second module until the preset convergence condition is met; then comparing the target function values ​​obtained based on each first particle parameter, obtaining the target particle parameters based on the comparison results, and extracting the blind estimation result of the pulse repetition time from the target particle parameters.

[0148] In some embodiments, the apparatus may further include: a sixth module, configured to sort the accumulation results according to data power, obtain an average noise power estimate and an average power estimate of the accumulation results based on the sorting results; and then calculate the signal-to-noise ratio after accumulation based on the exponent of the ratio of the average noise power estimate and the average power estimate; a seventh module, configured to obtain the average accumulation efficiency based on the ratio of the accumulated signal-to-noise ratio to the number of pulse signals in the pulse group; and an eighth module, configured to determine the energy concentration degree based on the number of pulses within the repetition time of a single pulse in the accumulation results, combined with preset conditions.

[0149] The content of the method embodiments of the present invention is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above methods.

[0150] like Figure 9 As shown, another aspect of the present invention provides an electronic device 700, including a processor 710 and a memory 720;

[0151] Memory 720 is used to store programs;

[0152] The processor 710 executes the program as described above.

[0153] The content of the method embodiments of the present invention is applicable to the embodiments of the present electronic device. The specific functions implemented by the embodiments of the present electronic device are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above methods.

[0154] Another aspect of this invention provides a computer-readable storage medium storing a program that is executed by a processor to implement the method described above.

[0155] The content of the method embodiments of the present invention is applicable to the computer-readable storage medium embodiments. The specific functions implemented by the computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above methods.

[0156] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0157] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0158] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0159] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a 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 a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0160] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution means, apparatus, or device (such as a computer-based device, a processor-including device, or other means that can fetch and execute instructions from, or in conjunction with, an instruction execution means, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution means, apparatus, or device.

[0161] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0162] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution device. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0163] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0164] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0165] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A blind estimation method for radar signal pulse repetition time, characterized in that, include: Obtain the original pulse train and the preset initial particle parameters, and use the initial particle parameters as the first particle parameters; The first particle parameters include pulse repetition time, start time, and end time; The original pulse train is truncated and grouped using the first particle parameter to obtain a pulse group; the pulse group includes several pulse signals. A reference signal is obtained from the pulse group, pulse shift correction is performed on each pulse signal based on the reference signal, and phase compensation processing is performed on each pulse signal based on the reference signal; then, the pulse signals are superimposed to obtain an accumulation result. The objective function value of the accumulation result is obtained by weighted summation of the objective parameters of the accumulation result. The target parameters include the signal-to-noise ratio after accumulation, the average accumulation efficiency, and the energy concentration. Based on the objective function value, the first particle parameters are updated to obtain the updated first particle parameters; Then, return to the step of truncating and grouping the original pulse train using the first particle parameters to obtain pulse groups, until the preset convergence condition is reached; then compare the objective function values ​​obtained based on each of the first particle parameters, obtain the target particle parameters based on the comparison results, and extract the blind estimation result of the pulse repetition time from the target particle parameters; The method further includes: The accumulation results are sorted according to data power, and the noise average power estimate and average power estimate of the accumulation results are obtained based on the sorting results; then, the signal-to-noise ratio after accumulation is obtained by calculating the exponent of the ratio of the noise average power estimate and the average power estimate. The average accumulation efficiency is obtained based on the ratio of the accumulated signal-to-noise ratio to the number of pulse signals in the pulse group. The energy concentration degree is determined based on the number of pulses within a single pulse repetition time in the accumulation result, combined with preset conditions.

2. The blind estimation method for radar signal pulse repetition time according to claim 1, characterized in that, The step of truncating and grouping the original pulse train using the first particle parameter to obtain pulse groups includes: The original pulse train is truncated using the start time and the end time to obtain the pulse train to be processed; The pulse train to be processed is grouped according to the pulse repetition time to obtain the pulse group.

3. The blind estimation method for radar signal pulse repetition time according to claim 1, characterized in that, The pulse shift correction of each pulse signal based on the reference signal includes: Based on the reference signal, the time shift difference between each pulse signal and the reference signal is obtained by cross-correlation peak detection; Based on the time shift difference, the corresponding pulse signal is time-shifted and corrected to obtain the pulse signal after pulse shift correction.

4. The blind estimation method for radar signal pulse repetition time according to claim 3, characterized in that, The step of performing time shift correction on the corresponding pulse signal based on the time shift difference to obtain the pulse shift-corrected pulse signal includes: The pulse signal is subjected to a Fourier transform, and then the result of the Fourier transform of the pulse signal is shifted and compensated by the corresponding time shift difference of the pulse signal. The result of the shift compensation is subjected to inverse Fourier transform to obtain the pulse signal after pulse shift correction.

5. The blind estimation method for radar signal pulse repetition time according to claim 1, characterized in that, The phase compensation processing of each pulse signal based on the reference signal includes: Based on the product of the reference signal and the conjugate signal of each pulse signal, the phase difference between each pulse signal and the reference signal is obtained by taking the phase. Based on the phase difference, phase compensation is performed on the corresponding pulse signal to obtain the pulse signal after phase compensation processing.

6. The blind estimation method for radar signal pulse repetition time according to claim 1, characterized in that, The step of updating the first particle parameters based on the objective function value includes: The first particle parameters are updated by comparing the objective function values ​​obtained corresponding to the first particle parameters in each historical iteration with the objective function values ​​obtained corresponding to the first particle parameters in the current iteration, combined with a random matrix and a learning factor.

7. A blind estimation device for radar signal pulse repetition time, characterized in that, The apparatus, applied to the method of claim 1, comprises: The first module is used to acquire the original pulse train and preset initial particle parameters, and use the initial particle parameters as the first particle parameters; the first particle parameters include pulse repetition time, start time and end time; The second module is used to truncate and group the original pulse train according to the first particle parameters to obtain a pulse group; the pulse group includes several pulse signals. The third module is used to obtain a reference signal from the pulse group, perform pulse shift correction on each pulse signal based on the reference signal, and perform phase compensation processing on each pulse signal based on the reference signal; then, perform superposition processing on each pulse signal to obtain an accumulation result. The fourth module is used to perform a weighted summation of the target parameters of the accumulation result to obtain the target function value of the accumulation result; the target parameters include the signal-to-noise ratio after accumulation, the average accumulation efficiency, and the energy concentration degree. The fifth module is used to update the first particle parameters based on the objective function value to obtain the updated first particle parameters; then return to the second module until a preset convergence condition is reached; then compare the objective function values ​​obtained based on each first particle parameter, obtain the target particle parameters based on the comparison results, and extract the blind estimation result of the pulse repetition time from the target particle parameters.

8. An electronic device, characterized in that, Including the processor and memory; The memory is used to store programs; The processor executes the program to implement the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The storage medium stores a program that is executed by a processor to implement the method as described in any one of claims 1 to 6.

Images