Airborne bistatic radar clutter suppression and target accumulation detection combined implementation method

Abstract

The invention discloses an airborne bistatic radar clutter suppression and target accumulation detection combined implementation method, which is applied to the technical field of radars, and aims to solve the problems of range migration, Doppler migration and strong clutter noise background when an airborne bistatic radar detects a low-altitude maneuvering target. According to the method, non-stationary clutters are effectively suppressed and range migration and Doppler migration are corrected and compensated by utilizing two-stage sliding sub-aperture space-time processing, target signal reconstruction and sub-aperture segmented accumulation, so that coherent accumulation of target multi-pulse energy is realized; finally, the purposes of improving the echo signal-clutter-noise ratio and improving the maneuvering target detection performance of the bistatic radar are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar, and particularly relates to a technology for detecting maneuvering targets in a complex clutter environment. Background Art

[0002] Airborne bistatic radars have shown advantages in anti-jamming, anti-stealth, and anti-destruction capabilities, and have received extensive attention and research. An airborne bistatic radar consists of a separated transmitter and receiver. When detecting low-altitude maneuvering targets such as missiles and unmanned aerial vehicles on different airborne platforms, problems such as clutter non-stationarity / non-stationariness and range cell migration (RCM) / Doppler shift (DM) of target echoes will occur, resulting in a decline in the performance of pure clutter suppression or pure coherent integration detection.

[0003] Up to now, methods for single-channel clutter suppression have been widely studied, such as methods based on time-frequency analysis, Doppler filtering, and eigen-decomposition. Specifically, the method based on Doppler filtering realizes static clutter suppression and out-of-band detection of the clutter spectrum; the method based on time-frequency analysis uses the differences in the centroids and modulation frequencies of targets and stationary clutter to complete clutter suppression and target detection; the method based on eigen-decomposition can eliminate weak clutter energy points after echo reconstruction. The above single-channel clutter suppression methods have the advantages of simple hardware requirements and low operation complexity. However, due to the simultaneous broadening of clutter and targets, it is difficult for airborne radars to detect moving targets in main-lobe clutter through these methods.

[0004] Considering the limitations of the above single-channel methods in clutter suppression processing, multi-channel methods have been developed. Typical methods include Displaced Phase Center Antenna (DPCA) and Space-Time Adaptive Processing (STAP). DPCA can suppress stationary clutter and retain moving target signals by canceling the two-channel echo data. However, this method requires the platform speed, channel spacing, and Pulse Repetition Frequency (PRF) to meet strict requirements, that is, the DPCA condition. STAP makes full use of the spatio-temporal coupling characteristics of clutter signals for spatio-temporal joint processing, adaptively filters the echo data, and maximizes the output Signal-to-Clutter-plus-Noise Ratio (SCNR). However, the calculation of the optimal weight vector of the space-time adaptive filter and the accuracy problem of covariance matrix estimation have always accompanied its development, restricting its practical application.

[0005] In the research of coherent accumulation detection, the KT transform and MLRT are proposed to correct the first-order range migration. Coherent integration and target detection can be achieved by subsequent slow-time Fourier transform. Different from the foregoing methods, the Radon Fourier transform (RFT) realizes coherent integration by extracting the target energy trajectory through parameter search. However, this method is only applicable to coherent detection when there is a relative velocity between the radar and the target. When there is relative acceleration motion, the coherent detection performance of the foregoing three methods will degrade. The research on coherent accumulation detection technology for maneuvering targets with acceleration mainly focuses on quadratic RCM and DM compensation. Based on RFT, the Generalized RFT (GRFT) method is proposed through high-dimensional parameter search; the method based on transform-matched filtering (KT-MFP) is proposed by KT-correction and jointly searching for the folding factor and acceleration. This method obtains the accumulation detection result through multi-dimensional parameter search, with high computational complexity and difficulty in dealing with the influence of clutter.

[0006] Generally speaking, there is little research considering the joint implementation of clutter suppression and coherent accumulation detection, which is very important for the detection of maneuvering targets in complex clutter environments. The joint implementation based on sub-aperture STAP and KT-LVD is proposed. This method can simultaneously achieve the effects of clutter suppression and coherent accumulation detection, but it is difficult to solve the problems of non-stationary clutter and effective detection of airborne bistatic radars. Therefore, it is urgent to study the method for the joint implementation of clutter suppression and coherent accumulation detection. Summary of the Invention

[0007] To solve the above technical problems, the present invention proposes a method for jointly implementing clutter suppression and target accumulation detection for an airborne bistatic radar, which can simultaneously achieve clutter suppression and coherent accumulation detection and improve the output signal-to-clutter-plus-noise ratio.

[0008] The technical solution adopted by the present invention is as follows: A method for jointly implementing clutter suppression and target accumulation detection for an airborne bistatic radar, including:

[0009] S1. The bistatic radar transmits multi-pulse linear frequency modulation signals, then receives the multi-pulse time-domain echo signals returned to the radar, and performs pulse compression on the multi-pulse time-domain echo signals.

[0010] S2. Perform clutter correlation time sliding window processing on the multi-pulse time-domain echo signals after pulse compression to obtain a two-dimensional correlation time-domain echo signal regarding range cells and the number of pulses. Then, perform secondary sliding window space-time processing on the two-dimensional correlation time-domain echo signal to obtain a series of clutter suppression output results.

[0011] S3. Superimpose the clutter suppression output results with the same phase to reconstruct a time-domain echo signal including the target envelope and phase.

[0012] S4. Evenly divide the reconstructed time-domain echo signal into several uniform sub-apertures, and then perform echo discretization processing on the sub-apertures. Then, use the modified coordinate rotation transformation to correct the first-order range migration caused by the target velocity in each sub-segment, and then perform coherent integration on the target energy within the corrected sub-aperture by performing a fast Fourier transform on slow time.

[0013] S5. Perform coherent integration on the echo signals between sub-apertures. Then, perform an inverse fast Fourier transform along the fast-time frequency direction to obtain the coherent integration result of the energy of all time segments. Detect the target based on the coherent integration result. When the integration peak is higher than the threshold value, the target is detected.

[0014] Advantages of the present invention: The present invention provides a method for clutter suppression and target detection of an airborne bistatic radar based on sub-aperture processing. Aiming at the problems of non-stationary range of clutter signals and range / Doppler migration of target echoes when detecting maneuvering targets under the complex clutter background of an airborne bistatic radar, the above problems are effectively solved through the joint implementation of two-stage sub-aperture sliding window space-time processing and sub-aperture segmented coherent integration, suppressing the clutter background while correcting and compensating for range / Doppler migration, achieving coherent integration of target energy under low signal-to-clutter ratio conditions, and thus improving the detection performance of the airborne bistatic radar for maneuvering ultra-high-speed targets. Description of the Drawings

[0015] Figure 1 It is a flowchart of an embodiment of the present invention.

[0016] Figure 2 It is the echo signal after pulse compression in an embodiment of the present invention.

[0017] Figure 3 It is the reconstructed echo signal after the second sliding window processing in an embodiment of the present invention.

[0018] Figure 4 It is a diagram of the sub-aperture segmented coherent integration result in an embodiment of the present invention.

[0019] Figure 5 It is a diagram of the coherent integration result between sub-apertures in an embodiment of the present invention.

[0020] Figure 6 It is a diagram of the coherent integration result of KT-MFP.

[0021] Figure 7 It is a diagram of the coherent integration result of ARFT.

[0022] Figure 8 It is a diagram of the coherent integration result of GRFT based on sub-aperture STAP. Detailed Embodiment

[0023] The present invention mainly uses the scientific computing software Matlab R2023b to conduct simulation experiments to verify its correctness. The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0024] Please refer to Figure 1 , a method for jointly implementing clutter suppression and target accumulation detection of an airborne bistatic radar proposed by the present invention is specifically implemented through the following steps:

[0025] Step 1, the bistatic radar transmits multi-pulse linear frequency modulation signals, then receives the echo signals returned to the radar, and performs pulse compression on the multi-pulse time-domain echo signals.

[0026] In this embodiment, the airborne bistatic radar transmits a linear frequency modulation signal s tran (τ, t m ), where τ and t m are the fast time and slow time respectively; the sum of the bistatic distances of the target between the airborne bistatic radars is where R 0,tr , v eq,tr and a eq,tr are the initial bistatic distance sum, equivalent radial velocity, and equivalent radial acceleration between the radar and the target respectively; and the target time-domain echo signal after pulse compression is denoted as S tar (r, t m ), 2r represents the range history variable of the echo signal, the clutter time-domain echo signal is denoted as S c (r, t m ), the noise time-domain echo signal is denoted as n(r, t m ), where τ = 2r / c and c represents the speed of light. The received time-domain echo signal is denoted as S(r, t m ) = S tar (r, t m ) + S c (r, t m ) + n(r, t m ), as shown in Figure 2 .

[0027] The expressions of R 0,tr , v eq,tr and a eq,tr are respectively:

[0028]

[0029]

[0030] where, ||·||2 represents the 2-norm, respectively represent the initial position coordinates of the target, the signal transmitting platform, and the signal receiving platform during the detection of the bistatic radar, They are the corresponding initial velocities respectively. represents the acceleration of the target. Additionally

[0031] S tar (r, t m ) has the expression:

[0032] S tar (r, t m ) = [S tar,1 (r, t m ), S tar,2 (r, t m ), …, S tar,N (r, t m )] T

[0033] where N is the number of array elements or channels, and the echo signal expressions corresponding to each array element or channel after pulse compression of the meridian are:

[0034]

[0035] A1 represents the amplitude of the target echo after pulse compression.

[0036] In this embodiment, the system parameters adopted are: the initial distance unit of the target echo is 633, the equivalent radial velocity of the target is 81.6899 m / s, and the equivalent radial acceleration is 51.5252 m / s 2 , the carrier frequency of the radar transmitted signal is 0.2 GHz, the signal bandwidth is 5 MHz, the sampling frequency is 10 MHz, the pulse repetition frequency of the radar is 500 Hz, the pulse duration is 5 μs, the number of pulses included in one coherent integration time is 1500, and the signal-to-clutter ratio after pulse compression is -17 dB.

[0037] Step 2: Perform clutter correlation time sliding window processing on the multi-pulse time-domain echo signal after pulse compression to obtain a two-dimensional correlation time-domain echo signal regarding the range cell and the number of pulses. Subsequently, perform secondary sliding window space-time processing on the two-dimensional correlation time-domain echo signal to obtain a series of clutter suppression output results.

[0038] In this embodiment, first, according to the correlation time, perform the first sliding window processing on the echo signal S(r, t m ) to obtain the i-th first-level sliding window sub-aperture data that is, the two-dimensional correlation time-domain echo signal regarding the range cell and the number of pulses. Among them, h(t m ) represents the window function, whose length is determined by the correlation time, and δ i represents the middle moment of the sliding window sub-aperture. The type of sliding window function adopted for the first sliding window is determined according to the actual data type. Commonly used ones include rectangular window, Hanning window, etc.

[0039] Then, perform a secondary sliding window on the two-dimensional correlation time-domain echo signal. The secondary sliding window uses a rectangular window. The echo data of the sub-aperture of the k-th secondary sliding window is written in the form of spatio-temporal snapshots as m where the superscript T represents transpose,

[0040]

[0041] is an element in and n = 1, 2, …, N. The output of the sub-aperture of the secondary sliding window after spatio-temporal processing is denoted as

[0042] where [·] represents conjugate transpose, H denotes the spatio-temporal filtering weight vector of the sub-aperture, is a scalar, denotes the target steering vector,

[0043] s s (f s,tar ) represents the spatio-temporal steering vector of the target signal, represents the initial phase of the target signal; f s,tar denotes the target space normalized frequency, represents the Kronecker product, is the clutter covariance matrix estimated from a small sample of clutter within the sub-aperture.

[0044] k m The value range of k is determined by the length of the first sliding window and the length of the second sliding window. For example, if the length of the first sliding window is 10 and the length of the second sliding window is assumed to be 3, then the value range of k m is [1, 10 - 3 + 1]. In practical applications, the length of the second sliding window is preferably selected as the pulse length when there is no range walk for the target signal.

[0045] Step 3: Superimpose the clutter suppression results with the same phase and reconstruct the time-domain echo signal including the target envelope and phase, as shown in Figure 3 .

[0046] In this embodiment, superimpose the clutter suppression results of the sub-aperture of the secondary sliding window with the same phase to reconstruct the time-domain echo signal including the target envelope and phase; that is:

[0047]

[0048] where s rec (r, t m) represents the reconstructed echo signal, θ represents a variable constant and is an integer. represents the i-th first sliding window, the output of the θ-i-th second sliding window sub-aperture space-time processing, m represents the pulse number variable after zero-padding the original echo data, δ i represents the middle time of the sliding window sub-aperture, M h represents the length of the first pulse sliding window, K m represents the number of times of the second sliding window, PRI represents the pulse repetition time.

[0049] Step 4: Divide the reconstructed time-domain echo signal into several uniform sub-apertures on average, ensuring that the second-order range migration and Doppler migration caused by the equivalent radial acceleration within each sub-aperture can be ignored, and then perform echo discretization processing on the sub-apertures; then, use the modified coordinate rotation transformation to correct the first-order range walk caused by the target velocity in each sub-segment, and then perform coherent accumulation on the target energy within the corrected sub-aperture by performing a fast Fourier transform on the slow time, as Figure 4 shown.

[0050] In this embodiment, Step 4 includes the following processes:

[0051] First, ensure that the second-order range migration and Doppler migration caused by the equivalent radial acceleration within each sub-aperture can be ignored, that is where M s is the number of pulses in each sub-aperture segment, a max represents the maximum possible acceleration value of the target, f s represents the range sampling frequency, λ represents the wavelength. Divide the reconstructed time-domain echo signal into several uniform sub-apertures on average, and the echo signal of the m'-th discrete sub-aperture is expressed as s rec,m′ (n,m s ). n represents the number of discrete range cells, m s represents the number of pulses.

[0052] Then, use the rotation transformation formula to perform first-order range walk correction and phase alignment on the discrete sub-aperture echo signal, and the rotation transformation formula is

[0053]

[0054] When the search rotation angle ε' is equal to the true angle, the range walk correction result s rec,m′ (n′,m s ′; ε′) is obtained.

[0055] Subsequently, perform an M s -point Fourier transform (FT) along the slow time direction to obtain the coherent result within the m'-th sub-aperture represents the reconstructed echo Doppler frequency variable.

[0056] Step 5: Perform coherent integration on the echo signals between sub-apertures after correcting and compensating for the second-order range migration and Doppler migration, and then perform an inverse fast Fourier transform along the fast-time frequency direction to obtain the coherent integration result of the energy of all time segments, as Figure 5 shown; detect the target according to the coherent integration result, and detect the target when the integration peak is higher than the threshold value.

[0057] In this embodiment, the coherent result within the sub-aperture in the frequency domain is denoted as f n′ representing the range-frequency variable, and construct the frequency-domain matched filter H m′ (f n′ ; v e ′ q , a e ′ q ), and its expression is as follows:

[0058]

[0059] Perform frequency-domain compensation on the envelope and phase differences between sub-apertures to obtain whose expression is as follows:

[0060]

[0061] where f c represents the carrier frequency, M r represents the number of segmented sub-apertures, v e ′ q and a e ′ q represent the target echo trajectory search parameters. When v e ′ q = v eq,tr , a e ′ q = a eq,tr , the range walk and phase between sub-apertures are corrected and compensated. At this time, the integration result between sub-apertures is expressed as Finally, perform an inverse fast Fourier transform on the range dimension to obtain the coherent integration result of all sub-apertures represents the target component in the processing result, represents the residual clutter and noise components after processing. Detect the target according to the coherent integration result, where the false alarm rate is set to 10 -6 , and obtain the preset threshold through Monte Carlo simulation. Detect the target when the integration peak is higher than the preset threshold value, otherwise the target cannot be detected.

[0062] To illustrate the effectiveness of this method, Figure 6 ,Figure 7 and Figure 8 show the coherent integration results of KT-MFP and ARFT in the traditional method and the GRFT algorithm based on sub-aperture STAP of the present invention. Due to the influence of clutter background and range / Doppler migration, compared with the coherent integration results obtained by the present invention, the signal-to-clutter-plus-noise ratio of the existing method is much lower than the output signal-to-clutter-plus-noise ratio accumulation peak of the present invention, and the performance of the existing method drops significantly.

[0063] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principles of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, various changes and modifications can be made to the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the scope of the claims of the present invention.

Claims

1. A joint implementation method for airborne bistatic radar clutter suppression and target accumulation detection, characterized in that, Including: S1. The bistatic radar transmits multi-pulse linear frequency modulation signals, then receives the multi-pulse time-domain echo signals returned to the radar, and performs pulse compression on the multi-pulse time-domain echo signals; S2. Perform clutter correlation time sliding window processing on the multi-pulse time-domain echo signals after pulse compression to obtain two-dimensional correlated time-domain echo signals regarding range cells and pulse numbers. Then, perform secondary sliding window space-time processing on the two-dimensional correlated time-domain echo signals to obtain a series of clutter suppression output results; S3. Superimpose the clutter suppression output results with the same phase to reconstruct the time-domain echo signal containing the target envelope and phase; S4. Evenly divide the reconstructed time-domain echo signal into several uniform sub-apertures, and then perform echo discretization processing on the sub-apertures. Then, use the modified coordinate rotation transformation to correct the first-order range migration caused by the target speed in each sub-segment, and then perform coherent integration on the target energy within the corrected sub-aperture by performing fast Fourier transform on slow time; S5. Perform coherent integration on the echo signals between sub-apertures. Then, perform inverse fast Fourier transform along the fast-time frequency direction to obtain the coherent integration result of the energy of all time segments. Detect the target according to the coherent integration result. When the accumulation peak is higher than the threshold value, the target is detected; 2. The joint implementation method for airborne bistatic radar clutter suppression and target accumulation detection according to claim 1, wherein In step S2, perform the first sliding window processing on the multi-pulse time-domain echo signals after pulse compression according to the correlation time; 3. The joint implementation method of airborne bistatic radar clutter suppression and target accumulation detection according to claim 2, wherein In step S2, the sliding window function used for the secondary sliding window space-time processing is a rectangular window, and the length of the secondary sliding window is selected as the pulse length size when there is no range migration for the target signal; 4. The joint implementation method of airborne bistatic radar clutter suppression and target accumulation detection according to claim 3, characterized in that In step S3, the expression for reconstructing the time-domain echo signal containing the target envelope and phase is: where s rec (r, t m ) represents the reconstructed echo signal, θ represents a variable constant and is an integer, represents the i-th first sliding window, the output of the θ - i-th second sliding window sub-aperture space-time processing, m represents the pulse number variable after zero-padding the original echo data, δ i represents the middle time of the sliding window sub-aperture, M h represents the length of the first pulse sliding window, K m represents the number of times of the second sliding window, and PRI represents the pulse repetition time.

5. The joint implementation method for airborne bistatic radar clutter suppression and target accumulation detection according to claim 4, wherein, In step S4, the number of pulses in each sub-aperture satisfies the following requirements: Among them, M s is the number of pulses for each sub-aperture, c represents the speed of light, a max represents the maximum possible acceleration value of the target, f s represents the range sampling frequency, λ represents the wavelength, and PRI represents the pulse repetition time.

6. The joint implementation method for airborne bistatic radar clutter suppression and target accumulation detection according to claim 5, characterized in that Step S5 specifically includes the following sub-steps: Let the frequency domain of the coherent accumulation result of the target energy in the sub-aperture be denoted as Construct the frequency domain matched filter H m′ (f n′ ; v e ′ q , a e ′ q ), and its expression is as follows: Perform frequency-domain compensation on the envelope and phase differences between sub-apertures to obtain The expression is as follows: Among them, M r represents the number of segmented sub-apertures, and v e ′ q and a e ′ q represent the target echo trajectory search parameters; When v e ′ q = v eq,tr and a e ′ q = a eq,tr , the distance walk and phase between sub-apertures are corrected and compensated. At this time, the accumulation result between sub-apertures is expressed as S53. Perform fast inverse Fourier transform on the range dimension to obtain the coherent integration results of all sub-apertures; S54. Detect the target according to the coherent integration results obtained in step S53, set the false alarm rate, and obtain the preset threshold through Monte Carlo simulation. When the accumulation peak is higher than the preset threshold value, the target is detected, otherwise the target cannot be detected.

Images