Airborne radar clutter suppression and high-speed target coherent accumulation combined processing method based on Radon-STAP

Abstract

The invention discloses an airborne radar clutter suppression and high-speed target coherent accumulation combined processing method based on Radon-STAP, is applied to the technical field of radars, and aims at solving the problems that a high-speed target has range walk in a fast time dimension, the clutter suppression performance is reduced by directly using space-time adaptive processing, and target energy cannot be effectively accumulated. The method comprises the following steps: converting an echo signal to a frequency domain, converting target range walk into phase change, designing a corresponding walk-Doppler joint compensation function according to time-frequency characteristics of the echo signal in a fast time, and applying an improved space-time two-dimensional filter to non-stationary clutter suppression and target coherent accumulation detection; according to the method, good clutter suppression and high-speed target coherent accumulation can be realized under the condition that clutter correlation of different distance units is poor.

Description

Technical Field

[0001] The invention belongs to the field of radar technology, and in particular relates to a radar clutter suppression and high-speed target coherent accumulation technology. Background Art

[0002] Airborne radar plays a vital role in air detection, long-range early warning and other scenarios. Since the carrier platform itself has a certain height and speed, the clutter area illuminated by the airborne radar beam is expanded. At the same time, the different clutter scattering points on the ground or sea surface have great differences in azimuth and speed relative to the airborne radar, resulting in the center frequency shift and spectrum broadening of the clutter Doppler spectrum. At this time, if the target Doppler frequency falls within the clutter Doppler frequency range, it will bring difficulties to the effective detection of the airborne radar. Traditional clutter suppression methods only process in the time and frequency domains, which is difficult to effectively solve the above problems, resulting in difficulties in detecting weak targets in strong clutter. Space-time adaptive processing (STAP) technology utilizes the space-time coupling characteristics of airborne radar clutter, and can effectively suppress strong clutter through two-dimensional adaptive filtering in the space and time domains.

[0003] However, there may be range movement in the echo signal of high-speed targets, which may lead to the problem of space-time steering vector mismatch in array signal processing, and reduce the clutter suppression performance of STAP. Coherent accumulation of target signals can effectively improve the signal-to-noise ratio of target echo signals. Traditional coherent accumulation mainly performs envelope correction and phase compensation through parameter search or non-parameter search to improve the signal-to-noise ratio of echo signals. Among them, typical parameter search coherent accumulation methods include Keystone transform (KT) and Radon Fourier transform (RFT). However, the above coherent accumulation methods focus more on achieving target accumulation detection and fail to combine clutter suppression processing.

[0004] Up to now, the adaptive Radon Fourier transform (ARFT) can achieve clutter suppression and target signal accumulation at the same time based on the maximum signal-to-noise ratio criterion and RFT algorithm, and realize the effective detection of "low observable" targets under clutter background. However, this method only uses the echo information of a single channel and does not consider the fusion of airborne multi-channel radar echoes. In addition, this method divides the coherent pulse samples into sub-apertures, which limits the ability to suppress non-stationary clutter. The time-division STAP target refocusing algorithm under strong clutter background realizes robust clutter suppression and target energy accumulation based on a half-search mechanism. However, this method requires prior information or estimates of the target spatial frequency and Doppler frequency. At the same time, the target may still move in the sliding window after division, resulting in a decrease in algorithm performance. Therefore, it is urgent to study the joint processing method of airborne radar clutter suppression and high-speed target coherent accumulation when the clutter correlation of different distance units is poor. Summary of the invention

[0005] To solve the above technical problems, the present invention proposes a joint processing method of airborne radar clutter suppression and high-speed target coherent accumulation based on Radon-STAP, which can achieve good clutter suppression and high-speed target coherent accumulation when the correlation of clutter units at different distances is poor.

[0006] The technical solution adopted by the present invention is: a joint processing method of airborne radar clutter suppression and high-speed target coherent accumulation based on Radon-STAP, comprising:

[0007] S1. Assume that the antenna of the airborne radar is an N-element equidistant linear array with an element spacing of d. The elevation and azimuth angles of the high-speed target relative to the radar are θ and In the absence of interference signals, the space-time steering vector of the airborne radar echo signal is:

[0008] X=aS+X c +n

[0009] Where a is the complex amplitude of the target signal, X c and n represent the space-time steering vectors of clutter and noise respectively, and S is the space-time steering vector of the target signal;

[0010] S2. The discretization expression after transforming the airborne radar echo signal space-time steering vector of step S1 into the frequency domain is:

[0011] SF r (f) = F PC (f)+X c (f)+n

[0012] Among them, F PC (f) represents the frequency domain representation of S, X c (f) indicates X c Frequency domain representation of ;

[0013] S3, extracting a plurality of continuous distance units as units to be detected at the same time, and then using the data of the reference unit as training samples to estimate the clutter covariance matrix;

[0014] S4. For each search speed and search space frequency, design a corresponding walk-Doppler joint compensation function to rewrite the target space-time steering vector;

[0015] S5, using the clutter covariance matrix of step S3 and the target space-time steering vector rewritten in step S4, an improved space-time two-dimensional filter is constructed, and the optimal solution of the weight of the improved space-time two-dimensional filter is expressed as:

[0016]

[0017] S6, performing space-time adaptive filtering on the signal of the unit to be detected after discretization in step S2, and obtaining:

[0018] SF r (f) = W(f) H SF r (f)

[0019] S7. Perform inverse Fourier transform on the filtered signal to obtain the radial velocity and spatial frequency v respectively after clutter suppression. t and The distance dimension accumulation result of the echo signal at time:

[0020] S r (t) = IFT(SF r (f))

[0021] The beneficial effects of the present invention are as follows: the present invention aims at the problem of reduced clutter suppression and target coherent accumulation performance caused by the movement of high-speed targets and the poor correlation of clutter in different distance units. By transforming the echo signal into the frequency domain, the movement of the target distance is converted into a phase change, and a corresponding movement-Doppler joint compensation function is designed according to the time-frequency characteristics of the echo signal in the fast time. The improved space-time two-dimensional filter is used to achieve good clutter suppression and high-speed target coherent accumulation. The process included in the present invention can be implemented using the Radon-STAP algorithm, which is conducive to engineering implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of an embodiment of the present invention.

[0023] Figure 2 It is the low signal-to-noise ratio echo data received in the embodiment of the present invention.

[0024] Figure 3 A schematic diagram of a distance unit, a unit to be detected, and a reference unit provided in an embodiment of the present invention.

[0025] Figure 4 It is the clutter suppression and target coherent accumulation result of the method proposed in the embodiment of the present invention.

[0026] Figure 5 The clutter suppression and target coherence accumulation results of the STAP algorithm.

[0027] Figure 6 Accumulate the clutter suppression and target coherence results of the RFT algorithm. DETAILED DESCRIPTION

[0028] To facilitate those skilled in the art to understand the technical content of the present invention, the present invention is further explained below with reference to the accompanying drawings.

[0029] The present invention mainly uses scientific computing software Matlab R2022b to perform simulation experiments to verify its correctness. The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0030] See also Figure 1 The present invention proposes a method for joint processing of airborne radar clutter suppression and high-speed target coherent accumulation based on Radon-STAP, which is specifically implemented by the following steps:

[0031] Step 1: Perform Fourier transform processing on the fast time dimension of the echo signal to obtain the target echo in the fast time frequency domain, and convert the target distance movement into phase change.

[0032] In this embodiment, it is assumed that the antenna of the airborne radar is an N-element equidistant linear array, the element spacing is d, and the elevation angle and azimuth angle of the high-speed target relative to the radar are θ and In the absence of interference signals, the space-time steering vector of the airborne radar echo signal can be written as:

[0033] X=aS+X c +n

[0034] Where a is the complex amplitude of the target signal, X c and n represent the space-time steering vectors of clutter and noise respectively, S is the space-time steering vector of the target signal, and its expression is:

[0035]

[0036] in represents the Kronecker product, and there is and Respectively represent the time domain steering vector and spatial domain steering vector of the target:

[0037]

[0038] in, represents the radial Doppler shift of the target relative to the radar, represents the spatial frequency of the target, v t represents the radial velocity of the target relative to the radar, T r represents the pulse repetition period, λ is the wavelength of the transmitted signal, M and N represent the number of array elements and the number of pulses respectively. When the target moves across the distance unit in the slow time dimension due to its fast speed, the target echo signal can be written as:

[0039]

[0040] Where t represents the fast time variable, m and n represent the mth array element and the nth pulse respectively, B represents the signal bandwidth, and R 0is the initial distance of the target, c is the speed of light, A t is the signal envelope amplitude.

[0041] Due to the linear characteristics of Fourier transform (FT), FT processing of the fast time dimension of the echo signal will not affect the phase change characteristics of the clutter signal in the slow time and spatial domain. The expression of the target echo in the frequency domain can be obtained as follows:

[0042]

[0043] Among them, rect is a rectangular function, which combines the phase terms containing n:

[0044]

[0045] At this point, the different envelope offsets of the target echo signal in the distance dimension are converted into different phase change rates in the frequency domain, which can be compensated by the weight vector in STAP without affecting the properties of the clutter space-time steering vector.

[0046] In this embodiment, the system parameters used are: the number of airborne radar array elements M is 8, the linear frequency modulation signal is transmitted, the number of coherent pulses N is 40, and the pulse repetition period T r The signal bandwidth B is 15 MHz, the wavelength λ is 0.15 m, and the array element spacing d is half the wavelength. The target radial moving speed v t is 1200m / s, the initial distance unit R 0 is 2000, and the cone angle of the target airspace is 60°. In a strong clutter environment, the input signal-to-noise ratio is -32.21dB. In a low signal-to-noise ratio, the target track is submerged by clutter.

[0047] The low signal-to-noise ratio echo data received in the embodiment of the present invention is as follows: Figure 2 shown.

[0048] Step 2: extract multiple consecutive distance units as units to be detected at the same time, and then use the data X of adjacent reference units as training samples to estimate the clutter covariance matrix.

[0049] like Figure 3 As shown, the distance unit is a unit divided according to the radar resolution on the radar detection distance; the unit to be detected is the distance unit where the target is located. Since the target moves at a fast speed, the unit to be detected in this embodiment is several consecutive distance units where the target is located; the reference unit is several distance units near the unit to be detected, and the data in the reference unit is the training sample.

[0050] Different from the traditional STAP, the unit to be detected based on Radon-STAP needs to extract K adjacent distance units at the same time as the unit to be detected. Then the data of the reference unit is used as a training sample to estimate the clutter covariance matrix. Assuming there are L training samples, the clutter covariance matrix can be obtained according to the maximum likelihood estimation criterion:

[0051]

[0052] Where X l represents the space-time steering vector of the lth training sample, and H represents the conjugate transpose operation in the matrix operation.

[0053] In this embodiment, K is 200, the units to be detected are 1900 to 2099 distance units, and L is 800.

[0054] Searching for speed and spatial frequency is a necessary operation for STAP. When searching for the speed and spatial frequency of the target, a peak will appear at the speed and spatial frequency where the target is located to indicate the speed and angle of the target.

[0055] Step 3: According to the time-frequency characteristics of the target echo signal in the fast time, for each search speed and search space frequency, design the corresponding walk-Doppler joint compensation function and rewrite the target space-time steering vector.

[0056] In this embodiment, after the signal of the unit to be detected is transformed into the fast time frequency domain, for each frequency unit f, when the search speed is v t , the search space frequency is When , the space-time weight vector of the target can be written as:

[0057]

[0058] in represents the Kronecker product, and the expressions of the time-space compensation functions are:

[0059]

[0060] Here, T represents the "transpose" of the matrix.

[0061] Step 4: Use the improved space-time two-dimensional filter to perform space-time adaptive filtering on the echo signal in the detection unit.

[0062] The received space-time two-dimensional data is subjected to the linearly constrained minimum variance criterion (LCMV, linearly constrained

[0063] minimum-variance) to solve the weight vector, that is, to solve the following convex quadratic optimization problem:

[0064]

[0065] W(f) is the STAP space-time filter weight vector;

[0066] In this embodiment, the optimal solution of the STAP space-time filter weights can be expressed as:

[0067]

[0068] Among them, the non-zero complex constant for The inverse matrix of .

[0069] The discretization expression of the signal of the unit to be detected after being transformed into the frequency domain is:

[0070] SF r (f) = F PC (f)+X c (f)+n

[0071] By performing space-time adaptive filtering on the signal of the unit to be detected, we can obtain:

[0072] SF r (f)=W(f) H SF r (f)

[0073] Step 5, perform inverse Fourier transform on the filtered signal, and the maximum value is the target coherent accumulation result of the corresponding radial velocity and spatial frequency respectively corresponding to the search value after clutter suppression.

[0074] In this embodiment, the inverse Fourier transform is performed on the filtered signal to obtain the signal after clutter suppression when the search speed and search space frequency are v respectively. t and The distance dimension accumulation result of the echo signal at time:

[0075] S r (t) = IFT(SF r (f))

[0076] In order to illustrate the effectiveness of this method, the same target and clutter background echoes under the same signal-to-noise ratio (-32.21dB) are processed. Figure 4 The clutter suppression and target coherent accumulation results of the method proposed in this invention are given. Figure 5 and Figure 6 The results of clutter suppression and target coherent accumulation of STAP and RFT algorithms in traditional methods are shown. Due to the influence of low signal-to-noise ratio of received signals, the clutter suppression and target coherent accumulation performance of the existing methods are much lower than those of the method of the present invention.

[0077] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A Radon-STAP-based joint processing method for airborne radar clutter suppression and high-speed target coherent integration, characterized in that: include: S1. Assume that the antenna of the airborne radar is an N-element equidistant linear array with an element spacing of d. The elevation and azimuth angles of the high-speed target relative to the radar are θ and In the absence of interference signals, the space-time steering vector of the airborne radar echo signal is obtained as follows: X=aS+X c +n Where a is the complex amplitude of the target signal, X c and n represent the space-time steering vectors of clutter and noise respectively, and S is the space-time steering vector of the target signal; S2. The discretization expression after transforming the airborne radar echo signal space-time steering vector of step S1 into the frequency domain is: SF r (f)=F PC (f)+X c (f)+n Among them, F PC (f) represents the frequency domain representation of S, X c (f) indicates X c Frequency domain representation of ; S3, extracting a plurality of continuous distance units as the units to be detected at the same time, and then taking a number of distance units adjacent to and before the units to be detected as reference units, and using the data of the reference units as training samples to estimate the clutter covariance matrix; S4. For each search speed and search space frequency, design a corresponding walking-Doppler joint compensation function to rewrite the target space-time steering vector; S5, using the clutter covariance matrix of step S3 and the target space-time steering vector rewritten in step S4, construct an improved space-time two-dimensional filter, and obtain the optimal solution of the weight of the improved space-time two-dimensional filter, which is expressed as: Where W(f) represents the improved space-time two-dimensional filter weight, h(f) represents the space-time weight vector of the target, represents the clutter covariance matrix, μ represents a non-zero complex constant, and f is a frequency unit; S6, performing space-time adaptive filtering on the signal of the unit to be detected after discretization in step S2, and obtaining: SF r (f)=W(f) H SF r (f) Where H represents the conjugate transpose operation; S7. Perform inverse Fourier transform on the filtered signal to obtain the radial velocity and spatial frequency v after clutter suppression. t and The echo signal distance dimension accumulation result at that time.

2. The method of combining airborne radar clutter suppression and high-speed target coherent integration based on Radon-STAP according to claim 1, characterized in that: For each frequency unit f, when the search speed is v t , the search space frequency is When , the space-time weight vector of the target is expressed as: in, represents the Kronecker product, h t (f) h s (f) are the temporal and spatial compensation functions respectively.

3. The method of combining Radon-STAP-based airborne radar clutter suppression and high-speed target coherent integration according to claim 2, characterized in that: h t (f) h s (f) The expressions are: Among them, v t represents the radial velocity of the target relative to the radar, T r represents the pulse repetition period, λ is the wavelength of the transmitted signal, and c is the speed of light.

4. The method of combining airborne radar clutter suppression and high-speed target coherent integration based on Radon-STAP according to claim 3 is characterized in that: The weight vector is solved by using the linear constrained minimum variance criterion for the received space-time two-dimensional data, that is, by solving the optimization problem The optimal solution of the improved space-time two-dimensional filter weights is obtained.

5. The method of combining Radon-STAP-based airborne radar clutter suppression and high-speed target coherent integration according to claim 4 is characterized in that: It is obtained according to the maximum likelihood estimation criterion.

6. The method of combining airborne radar clutter suppression and high-speed target coherent integration based on Radon-STAP according to claim 4 is characterized in that:

Images