CN107703490A - Range ambiguity clutter suppression method based on FDA MIMO radars - Google Patents

Abstract

The invention discloses a local joint dimensionality reduction range ambiguity clutter suppression method based on FDA-MIMO radar, which mainly solves the poor detection performance of the existing range ambiguity clutter suppression method and the large amount of computation and high requirements for the number of independent and identically distributed samples The problem. The implementation steps are: (1) use the transmitted waveform to perform matched filtering on the echo data of the radar; (2) perform distance-dependent compensation on the data after the matched filtering; (3) construct a local joint dimensionality reduction matrix and process the received data Dimensionality reduction processing; (3) Estimating the clutter covariance matrix with the reduced dimension data; (4) Obtaining the optimal weight vector according to the minimum variance undistorted response beamforming; (5) Using the optimal weight to reduce the dimensionality Data weighting, suppressing range ambiguity clutter, and detecting target signals. Compared with the existing range ambiguous clutter suppression method, the present invention has the advantages of low computational complexity, low requirements on the number of independent and identically distributed samples and good clutter suppression performance, and realizes range ambiguous clutter suppression for airborne radar, It can be applied to airborne radar ground moving target detection.

Description

Range ambiguity clutter suppression method based on FDA-MIMO radar

technical field

The invention belongs to the technical field of radar, in particular to a range fuzzy clutter suppression method, which can be used for detecting moving targets.

Background technique

When the airborne early warning radar is in the looking-down working state, the clutter Doppler will spread, causing the weak target signal to be submerged by the clutter. Space-time two-dimensional adaptive processing STAP joint space and time two-dimensional information can effectively improve the ability of clutter suppression and the performance of moving target detection. The traditional STAP method assumes that there is no range ambiguity. When range ambiguity occurs, the target has to compete with both unambiguous and ambiguous clutter, which greatly reduces the detection performance of STAP.

T.B.Hale et al. in the paper "Localized three-dimensional adaptive spatial-temporal processing for airborne radar" (IEE Radar Sonar and Navig., vol.150, no.1, pp.18-22, Feb.2003) and "Clutter suppression using "elevation interferometryfused with space-time adaptive processing" (Electron.Lett., vol.37, no.12, pp.793-794, Jun.2001) proposed a three-dimensional SATP method, by introducing the degree of freedom of the pitch dimension to Solve the problem of distance blur clutter. However, this method requires a lot of calculation and requires a large amount of training data, which is difficult to apply in actual situations.

D.Cristallini et al. introduced a direct data domain in the paper "A robust direct data domain approach for STAP" (IEEE Trans.Signal Process., vol.60, no.3, pp.1283-1294, Mar.2012) The STAP method obtains second-order training samples through the smoothing of a single snapshot in the spatial and temporal domains to suppress range ambiguity clutter, but this method sacrifices a certain degree of freedom, resulting in a decline in detection performance.

Contents of the invention

The object of the present invention is to address the shortcomings of the above-mentioned prior art, and propose a range ambiguity clutter suppression method based on FDA-MIMO radar, so as to improve the detection performance of moving targets in the case of range ambiguity. The technical scheme of the present invention is: by adopting multiple-input multiple-output MIMO radar technology, the virtual launch degree of freedom can be effectively realized, and the range-angle dependence characteristics of the transmit steering vector of the frequency diversity array are used to separate and suppress the range ambiguous clutter, and at the same time, through The local joint dimension reduction method reduces the computational complexity and clutter suppression requirements for the number of independent and identically distributed clutter samples, and performs real-time processing on the ground target detection process. The implementation steps include the following:

(1) Utilize the side-looking FDA-MIMO radar mode, use the transmit waveforms of the M transmit antenna array elements for each snapshot data of the N antennas at the radar receiver end Perform matched filtering; connect the echo data of each receiving antenna matched and filtered end to end to obtain the space-time data vector x _l of the l-th range gate in MNK×1 dimension, where m=1,2,..., M, superscript * means conjugate, K is the number of pulses in a coherent processing interval;

(2) Perform distance-dependent compensation on the space-time data vector _xl , and obtain the compensated space-time data vector:

in, is the compensation vector, r _l is the principal value distance corresponding to the range gate to be detected, g _T (r _l ) is the compensation vector at the transmitting end, 1 _N is the column vector of all 1s in N dimension, and 1 _K is the column of all 1s in K dimension vector, symbol Represents the Kronecker product, the superscript H represents the conjugate transpose, and diag represents the diagonalization of the compensation vector;

(3) For the compensated space-time data vector and the orientation vector of the target Perform dimensionality reduction to obtain the data vector after dimensionality reduction and the reduced goal-directed vector

Among them, T represents the local joint dimensionality reduction matrix, p ₀ represents the index number of the distance region where the target is located, ψ ₀ is the angle of the target, and v ₀ is the speed of the target;

(4) Dimensionality-reduced data vectors using L range gates Estimate the covariance matrix of the range gate where the target is located:

in, is the covariance matrix of clutter plus noise after dimensionality reduction;

(5) According to the results of step (3) and step (4), the optimal weight vector is obtained by the minimum variance undistorted response beamformer:

in, Estimated values for the covariance matrix the inverse matrix;

(6) Use the optimal weight vector w to reduce the dimensionality of the data vector Carry out weighted summation, filter out the clutter in the echo data, and detect the moving target.

Compared with the prior art, the present invention has the following advantages:

(a) The present invention utilizes the degree of freedom of the FDA radar distance dimension, and utilizes MIMO technology at the same time to separate the range-ambiguous clutter in the transmitting-receiving space through distance-dependent compensation, and then suppresses the clutter in different distance regions, thereby improving the dynamic range. Target detection performance.

(b) The present invention uses the local joint dimensionality reduction processing method to reduce the dimensionality of the data after distance-dependent compensation, which not only reduces the amount of calculation, reduces the complexity of calculation, but also reduces the requirement for the number of independent and identically distributed data samples, Real-time detection of ground moving targets can be realized.

The purpose, features and advantages of the present invention can be described in detail by the following drawings and examples.

Description of drawings

Fig. 1 is the realization flowchart of the present invention;

Fig. 2 is the model schematic diagram of the side-looking airborne FDA-MIMO radar adopted in the present invention;

Figure 3 is the clutter power spectrum of the FDA-MIMO radar without distance-dependent compensation;

Fig. 4 is the clutter power spectrum after carrying out distance-dependent compensation to FDA-MIMO radar in the present invention;

Fig. 5 is the transmit-receive-Doppler three-dimensional processor response figure that suppresses clutter with the present invention;

FIG. 6 is a slice diagram of FIG. 5 on a space-time two-dimensional plane.

Detailed ways

The front and side looking airborne FDA-MIMO radar model used in the present invention is shown in Figure 2, and the origin O of the coordinate system is the projection point of the platform on the horizontal plane, and both the transmitting array and the receiving array adopt a uniform linear array, and the number of transmitting antenna array elements is M , the number of receiving antenna elements is N, d _T is the distance between transmitting antenna elements, d _R is the distance between receiving antenna elements, the x-axis is the direction of platform motion velocity υ _p , H is the height of the platform, θ _q is the clutter point azimuth, φ _l,p is the pitch angle of the clutter point, r _l,p is the distance from the clutter point to the platform, ψ _l,p,q are the angles between the clutter point and the platform, ξ _{l, p and q} are the scattering coefficients of the clutter points.

With reference to Fig. 1, the concrete realization steps of the present invention are as follows:

Step 1: Perform matched filtering on the echo data of the radar.

Using the side-looking FDA-MIMO radar mode, the echo data of each receiving antenna and the conjugate of the transmitting waveform The inner product is used for matching filtering, and the echo data after matching filtering are connected end-to-end to obtain the MNK×1-dimensional echo data as follows:

Among them, the superscript * represents the conjugate, m=1,2,...,M, K is the number of pulses within a coherent processing time, ξ _t is the reflection coefficient of the target, Direction vector for target, symbol represents the Kronecker product, a _T (f _T (r ₀ ,ψ ₀ )) is the launch steering vector of the target, f _T (r ₀ ,ψ ₀ ) is the launch spatial frequency of the target, a _R (f _R (ψ ₀ )) is the receiving steering vector of the target, f _R (ψ ₀ ) is the receiving spatial frequency of the target, b(f _d (ψ ₀ ,v ₀ )) is the Doppler steering vector of the target, f _d (ψ ₀ ,v ₀ ) is the Doppler frequency of the target, r ₀ is the distance of the target, ψ ₀ is the angle of the target, v ₀ is the velocity of the target, N _p is the range ambiguity number, N _c is the number of statistically independent clutter points in each range gate , is the steering vector of the clutter point, a _T (f _T (r _l,p ,ψ _l,p,q )) is the steering vector of the clutter emission, f _T (r _l,p ,ψ _l,p,q ) ＝-2Δfr _l,p /c+d _T cos(ψ _l,p,q )/λ ₀ is the normalized emission spatial frequency of the clutter point, a _R (f _R (ψ _l,p,q )) is The receiving steering vector of clutter, f _R (ψ _l,p,q )=d _R cos(ψ _l,p,q )/λ ₀ is the normalized receiving spatial frequency of clutter, b(f _d (ψ _{l ,p,q} )) is the Doppler steering vector of the clutter, f _d (ψ _l,p,q )=2v _p T cos(ψ _l,p,q )/λ ₀ is the normalized multiple Δf is the step frequency, c is the speed of light, λ ₀ is the carrier wavelength, T is the pulse repetition period, n _l is the space-time data vector of Gaussian white noise.

The described positive and side-looking FDA-MIMO radar mode means that the carrier frequency of the transmitting signal of each transmitting antenna unit has a linear step, and the flight direction of the platform is perpendicular to the normal direction of the antenna, and multiple transmitting antennas are used at the transmitting end Multiple transmission channels are generated by transmitting mutually orthogonal signals, and echo signals are received by multiple antennas at the receiving end.

Step 2: Perform distance-dependent compensation on the data after matched filtering.

Since the frequency diversity array FDA radar introduces the degree of freedom in the distance dimension through the frequency increment of the carrier frequency between the array elements, the launch steering vector has a distance dependence, and distance dependence compensation is required. The compensation vector is The compensated data vector is expressed as:

Among them, r _l is the principal value distance of the range gate to be detected, the superscript H indicates the conjugate transpose, diag indicates the diagonalization of the compensation vector, g _T (r _l )=[1,exp{-j4πΔfr _l /c} ,…,exp{-j4πΔf(M-1)r _l /c}] ^T represents the compensation vector at the transmitting end, the superscript T represents the transpose, j represents the imaginary number, 1 _N and 1 _K are N-dimensional and K-dimensional all-ones respectively column vector;

The spatial frequency of the compensated clutter emission is where p is the index number of the distance region, r _u is the maximum unambiguous distance, and the receiving spatial frequency and Doppler frequency constant; is the compensated target-oriented vector, where, Indicates the target emission spatial frequency after compensation, p ₀ indicates that the target is in the p _0th distance zone;

Step 3: Use the local joint dimensionality reduction matrix to reduce the dimensionality of the data.

Due to the large amount of computing and the required number of independent and identically distributed clutter samples for direct echo data processing, real-time processing cannot be realized, resulting in a decline in detection performance. Therefore, before estimating the clutter covariance matrix, it is necessary to use the local joint dimensionality reduction method to reduce the dimensionality of the received data. The dimensionality reduction steps are as follows:

3a) Construct the dimensionality reduction matrix of the launch space:

in, Indicates the Q ₁ auxiliary channels around the target launch space channel;

3b) Construct the dimensionality reduction matrix of the receiving space:

in, Indicates the Q ₂ auxiliary channels around the target receiving space channel;

3c) According to 3a) and 3b), the joint dimensionality reduction matrix of the transmit-receive space is obtained as:

3d) Construct the dimensionality reduction matrix of the Doppler domain as:

in, Indicates the Q ₃ auxiliary channels around the target Doppler channel;

3e) According to 3c) and 3d), the joint dimensionality reduction matrix of the transmit-receive-Doppler three-dimensional space is obtained as:

3f) Multiply the compensated space-time data vector by the dimensionality reduction matrix T(p ₀ ,ψ ₀ ,v ₀ ) and the orientation vector of the target Get the reduced data vector and the reduced goal-directed vector

Step 4: Estimate the clutter covariance matrix.

Average the clutter covariance matrix of L range gates to obtain the estimated value of the covariance matrix of the target range gate

in, Represents the dimensionality-reduced data vector of the l-th range gate.

Step 5: Calculate the optimal weight vector.

According to the reduced clutter covariance matrix and the goal-directed vector The optimal weight vector is calculated by the minimum variance undistorted response beamformer:

in Estimated value for the clutter covariance matrix the inverse matrix of .

Step 6: Perform weighting processing on the data after dimensionality reduction.

Use the optimal weight vector w to reduce the dimensionality of the data vector Perform weighting processing to suppress the clutter in the echo data and detect the target.

The effects of the present invention can be further illustrated by the following simulation experiments.

1. Experimental environment

With reference to Fig. 2, various parameters used in the example of the present invention are as table 1

Table 1 Front and side looking FDA-MIMO radar parameters

parameter name specific value parameter name specific value The number of transmitting array elements M 6 Platform height H 6000m Number of receiving array elements N 6 distance fuzzy number N _p 4 Related pulse number L 8 CNR 50dB Reference carrier frequency f ₀ 1GHz Sample sizeL 300 Pulse repetition frequency f _PRF 2500Hz Target distance r ₀ 15000m Array element spacing d _T and d _R 0.15m Target speed ν ₀ 50m/s platform speed v _p 150m/s target angle ψ ₀ 90 ^° (relative to the x-axis)

2. Simulation content and results

Under the simulation conditions, the following experiments were performed.

In simulation experiment 1, the clutter spectrum simulation without range-dependent compensation is performed on the side-looking FDA-MIMO radar, and the results are shown in Figure 3.

In simulation experiment 2, the clutter spectrum simulation of the side-looking FDA-MIMO radar after distance-dependent compensation is performed, and the results are shown in Figure 4.

From the comparison of Figure 3 and Figure 4, it can be seen that the clutter spectrum without compensation has distance dependence, and after distance-dependent compensation, the clutter from different distance regions is separated.

Simulation experiment 3, using the method of the present invention to suppress range ambiguity clutter, the result is shown in Figure 5,

Slicing the three-dimensional response map in Figure 5 on the space-time two-dimensional plane, the result is shown in Figure 6.

It can be seen from Fig. 5 and Fig. 6 that the response notch of the transmit-receive-Doppler three-dimensional processor is aligned with the clutter spectrum, indicating that the present invention effectively suppresses the range ambiguous clutter.

To sum up, the present invention is based on the side-looking FDA-MIMO radar mode. First, the FDA introduces the degree of freedom of the distance dimension to realize the separation of range ambiguity clutter, and then uses the local joint dimensionality reduction method to reduce the dimensionality of the echo data. It not only reduces the computational complexity and the requirements for the number of independent and identically distributed sample data, but also maintains the good performance of range ambiguity clutter suppression, and realizes the effective detection of moving targets.

Claims (3)

1. A range ambiguity clutter suppression method based on FDA-MIMO radar, comprising: (1) Utilize the side-looking FDA-MIMO radar mode, use the transmit waveforms of the M transmit antenna array elements for each snapshot data of the N antennas at the radar receiver end Perform matched filtering; connect the echo data of each receiving antenna matched and filtered end to end to obtain the space-time data vector x _l of the l-th range gate in MNK×1 dimension, where m=1,2,..., M, superscript * means conjugate, K is the number of pulses in a coherent processing interval; (2) Perform distance-dependent compensation on the space-time data vector _xl , and obtain the compensated space-time data vector: <mrow><msub><mover><mi>x</mi><mo>^</mo></mover><mi>l</mi></msub><mo>=</mo><mi>d</mi><mi>i</mi><mi>a</mi><mi>g</mi><mo>{</mo><mi>g</mi><msup><mrow><mo>(</mo><msub><mi>r</mi><mi>l</mi></msub><mo>)</mo></mrow><mi>H</mi></msup><mo>}</mo><msub><mi>x</mi><mi>l</mi></msub><mo>,</mo></mrow> in, is the compensation vector, r _l is the principal value distance corresponding to the range gate to be detected, g _T (r _l ) is the compensation vector at the transmitting end, 1 _N is the column vector of all 1s in N dimension, and 1 _K is the column of all 1s in K dimension vector, symbol Represents the Kronecker product, the superscript H represents the conjugate transpose, and diag represents the diagonalization of the compensation vector; (3) For the compensated space-time data vector and the orientation vector of the target Perform dimensionality reduction to obtain the data vector after dimensionality reduction and the reduced goal-directed vector <mrow><msub><mover><mi>x</mi><mo>~</mo></mover><mi>l</mi></msub><mo>=</mo><msup><mi>T</mi><mi>H</mi></msup><msub><mover><mi>x</mi><mo>^</mo></mover><mi>l</mi></msub><mo>,</mo></mrow> <mrow><mover><mi>t</mi><mo>~</mo></mover><mrow><mo>(</mo><msub><mi>p</mi><mn>0</mn></msub><mo>,</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>,</mo><msub><mi>v</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>=</mo><msup><mi>T</mi><mi>H</mi></msup><mover><mi>t</mi><mo>^</mo></mover><mrow><mo>(</mo><msub><mi>p</mi><mn>0</mn></msub><mo>,</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>,</mo><msub><mi>v</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>,</mo></mrow> Among them, T represents the local joint dimensionality reduction matrix, p ₀ represents the index number of the distance region where the target is located, ψ ₀ is the angle of the target, and v ₀ is the speed of the target; (4) Dimensionality-reduced data vectors using L range gates Estimate the covariance matrix of the range gate where the target is located: <mrow><mover><mi>R</mi><mo>~</mo></mover><mo>=</mo><munderover><mo>&amp;Sigma;</mo><mrow><mi>l</mi><mo>=</mo><mn>1</mn></mrow><mi>L</mi></munderover><msub><mover><mi>x</mi><mo>~</mo></mover><mi>l</mi></msub><msubsup><mover><mi>x</mi><mo>~</mo></mover><mi>l</mi><mi>H</mi></msubsup><mo>,</mo></mrow> in, is the covariance matrix of clutter plus noise after dimensionality reduction; (5) According to the results of step (3) and step (4), the optimal weight vector is obtained by the minimum variance undistorted response beamformer: <mrow><mi>w</mi><mo>=</mo><mi>&amp;mu;</mi><msup><mover><mi>R</mi><mo>~</mi>mo></mover><mrow><mo>-</mo><mn>1</mn></mrow></msup><mover><mi>t</mi><mo>~</mo></mover><mrow><mo>(</mo><msub><mi>p</mi><mn>0</mn></msub><mo>,</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>,</mo><msub><mi>v</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>,</mo></mrow> in, Estimated values for the covariance matrix the inverse matrix; (6) Use the optimal weight vector w to reduce the dimensionality of the data vector Carry out weighted summation, filter out the clutter in the echo data, and detect the moving target. 2. The method according to claim 1, wherein the positive side-looking FDA-MIMO radar mode in step (1) means that the carrier frequency of the transmitting signal of each transmitting antenna unit has a linear step, and the flight direction of the platform Perpendicular to the normal direction of the antenna, multiple transmitting antennas transmit mutually orthogonal signals at the transmitting end to generate multiple transmitting channels, and multiple antennas are used at the receiving end to receive echo signals. 3. The method according to claim 1, wherein the local joint dimensionality reduction matrix T in the step (3) refers to selecting several auxiliary beams around the main beam pointing to the target in the transmit-receive-Doppler three-dimensional beam space channel, forming a local joint dimensionality reduction matrix in the transmit-receive-Doppler three-dimensional space, which is expressed as follows: <mrow><mi>T</mi><mrow><mo>(</mo><msub><mi>p</mi><mn>0</mn></msub><mo>,</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>,</mo><msub><mi>v</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>=</mo><mi>A</mi><mrow><mo>(</mo><msub><mi>p</mi><mn>0</mn></msub><mo>,</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>&amp;CircleTimes;</mo><mi>B</mi><mrow><mo>(</mo><msub><mi>&amp;psi;</mi><mn>0</mn></msub><mo>,</mo><msub><mi>v</mi><mn>0</mn></msub><mo>)</mo></mrow><mo>,</mo></mrow> Among them, T(p ₀ ,ψ ₀ ,v ₀ ) represents the local joint dimensionality reduction matrix of the transmit-receive-Doppler three-dimensional space, A(p ₀ ,ψ ₀ ) represents the joint dimensionality reduction matrix of the transmit-receive space, B(ψ ₀ , v ₀ ) represents a dimensionality reduction matrix in Doppler space.