SAR (synthetic aperture radar) robust waveform design method for extended target in frequency band limited environment

By constructing error models and applying spectrum constraints in SAR waveform design, the Dinkelbach algorithm and Lagrangian dual method are used to solve the problem of poor refined perception of the expansion target, and multi-radio frequency equipment in a frequency band restricted environment is realized The SAR waveform design with coexistence and high resolution characteristics improves the signal-to-missibility ratio of the expansion target.

CN118962677AActive Publication Date: 2024-11-15YANSHAN UNIV +1

Patent Information

Application Number
CN202410959565.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-15
Estimated Expiration
2044-07-17

AI Technical Summary

Technical Problem

When the existing SAR waveform faces the expansion target, inaccurate auxiliary knowledge estimates lead to performance degradation, and it is difficult to coexist with other RF devices in an environment of restricted spectrum resources, resulting in poor refined perception of the expansion target.

Method used

Using the SAR robust waveform design method for the expansion target under the band-constrained environment, by selecting the SAR echo domain signal-to-missibility ratio as optimization criteria, an error model of auxiliary knowledge is constructed, spectrum constraints, energy constraints and similarity constraints are applied, and the Dinkelbach algorithm is used. The duality of Lagrangian transforms the problem into a semi-definite planning problem, solves the closed solution of the target scattering characteristics and background scattering characteristics statistics, and realizes waveform design.

Benefits of technology

Significantly improves the signal-to-missive ratio (SCR) value of the expansion target, improves the refined perception performance of the expansion target, and allows multiple RF devices to coexist in a band-constrained environment, reduces the requirements for auxiliary knowledge estimation accuracy, and improves the waveform Design robustness and high resolution characteristics.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118962677A_ABST
    Figure CN118962677A_ABST
Patent Text Reader

Abstract

The invention discloses an SAR robust waveform design method for an extended target in a frequency band limited environment, and belongs to the field of signal processing, and the method comprises the steps: S1, selecting a signal-to-clutter ratio as an optimization criterion, building an error model of auxiliary knowledge, applying various constraints, and constructing an optimization problem of SAR robust waveform design; s2, solving a minimization problem of an inner layer, and deducing a closed-form solution of target and background scattering characteristic statistics; s3, constructing a sub-optimization problem about the waveform, and decoupling a numerator and a denominator of the target function; s4, converting a sub-problem in an iteration process into a semi-definite programming problem; s5, judging whether a stop condition is met or not; if yes, executing S6, and if not, continuing iteration; and S6, judging whether an optimality condition is met or not, if so, ending the solving process, and if not, solving the sub-problem obtained in the S3 by adopting semi-definite relaxation and rank-I decomposition until a stop condition is met. According to the method, the signal-to-clutter ratio of the SAR image can be remarkably improved, and the scattering characteristic of an extended target is effectively highlighted.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of signal processing, in particular to a SAR robust waveform design method for extended targets in a band-limited environment. Background Art

[0002] Synthetic aperture radar (SAR) is an active microwave imaging system. With its technical advantages of all-day, all-weather, and high resolution, it is widely used in various fields such as national defense and people's livelihood. The performance of SAR waveform plays a significant role in improving image quality, and the optimization design of SAR waveform has always been one of the main themes in this research field.

[0003] The existing SAR waveform has a large bandwidth characteristic. After matched filtering, it can form a correlation function with narrow main lobe and low side lobe characteristics, thereby ensuring the resolution and detection performance of point targets. Today, with the significant improvement of SAR spatial resolution, SAR images can obtain more details, causing most high-value targets to appear as extended targets in the image, thus creating conditions for the refined perception of extended targets. However, since extended targets cannot be regarded as a linear superposition of point targets, the point target model cannot accurately describe the characteristics of extended targets, resulting in a mismatch between the existing SAR waveform for point target detection and the application requirements of refined perception of extended targets, resulting in poor refined perception of extended targets, especially for dim targets. Therefore, it is urgent to optimize the design of SAR waveforms based on the extended target scattering model.

[0004] Although the research on radar waveform design for extended targets is not uncommon, the design process requires the combination of auxiliary knowledge about the target, and the problem of performance degradation caused by inaccurate auxiliary knowledge estimation has not been fully solved. At the same time, SAR waveforms usually require a large bandwidth to ensure high resolution, and the coexistence of broadband systems and other RF devices in an environment with limited spectrum resources faces challenges.

[0005] Therefore, how to design auxiliary knowledge error tolerance, compatibility with other RF equipment, and high-resolution and low-sidelobe SAR waveforms based on the extended target model has important scientific significance and application value. Summary of the invention

[0006] In order to improve the imaging observation effect of SAR on extended targets of interest, the present invention proposes a SAR robust waveform design method for extended targets in a band-limited environment, which can relax the estimation accuracy of auxiliary knowledge and achieve coexistence with other RF devices in the observation scene, while ensuring the high resolution and low sidelobe characteristics of SAR in the range direction.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0008] A SAR robust waveform design method for extended targets in a band-limited environment includes the following steps:

[0009] S1, select the SAR echo domain signal-to-clutter ratio as the optimization criterion, build the error model of auxiliary knowledge, impose spectrum constraints, energy constraints and similarity constraints, and construct the optimization problem of SAR robust waveform design;

[0010] S2, solve the minimization problem of the inner layer and derive the closed-form solution of the statistics of the target scattering characteristics and the background scattering characteristics;

[0011] S3, construct a sub-optimization problem about the waveform, and use the Dinkelbach algorithm to decouple the numerator and denominator of the fractional objective function;

[0012] S4, based on Lagrange duality and Schur complement theorem, the sub-problems in the iteration process are transformed into semidefinite programming problems;

[0013] S5, determine whether the stop condition is met, if so, execute S6; if not, continue iterating;

[0014] S6, determine whether the optimality condition is met. If so, the solution process ends; if not, use semidefinite relaxation and rank-one decomposition algorithm to solve the subproblem obtained in S3 until the stopping condition is met and output the solution result.

[0015] A further improvement of the technical solution of the present invention is that in S1, the following contents are specifically included:

[0016] The signal-to-clutter ratio (SCR) in the SAR echo domain is characterized based on the extended target scattering model; x(t) is used to represent the radar transmission signal, G(t) and B(t) represent the target scattering characteristics and background scattering characteristics respectively, and N(t) is the noise. The echo signal Y(t) received at any azimuth and time is expressed as:

[0017]

[0018] The convolution operator is realized by constructing Toeplitz matrix multiplication, and the discrete signal model is:

[0019] y=Gx+Bx+n (2)

[0020] In the formula, G and B represent the Toeplitz matrices formed after the discretization of G(t) and B(t), respectively. Represent the discretized vectors of x(t), Y(t), and N(t) respectively; let R G and R B are the correlation matrices of G and B respectively, then the SCR of SAR range direction is expressed as:

[0021]

[0022] In the formula, R G and R B It is auxiliary knowledge estimated by historical data, empirical models, and electromagnetic measurement methods, so its error model is established as:

[0023]

[0024] In the formula, and is the ideal error-free correlation matrix, ε G and ε B It is used to limit the error range; in order to improve the robustness of the optimized waveform to the auxiliary knowledge error, the SCR corresponding to the worst auxiliary knowledge is maximized as the objective function, that is:

[0025]

[0026] To ensure the waveform can coexist with other RF systems, spectrum constraints are imposed:

[0027] x H R I x≤E I (6)

[0028] In the formula, E I is the energy threshold that other RF devices within the frequency band can tolerate, and represents the weight assigned to the interfering device, Ω q The (m,l) element of is:

[0029]

[0030] In the formula, f 1 q and represents the frequency band range of the qth RF device; to ensure that x also has high resolution characteristics, similarity constraints are introduced to establish the maximization and minimization problem

[0031]

[0032] In the formula, E x is the energy of the maximum transmitted waveform, and ε is used to define the similarity between the transmitted waveform and the reference waveform c.

[0033] A further improvement of the technical solution of the present invention is that in S2, it specifically includes:

[0034] Introduce two auxiliary variables ΔR G , Make The inner minimization problem is equivalent to:

[0035]

[0036] The objective function of the above formula is equivalent to:

[0037]

[0038] Based on this, we can get the following about ΔR G and ΔR B The linear programming problem is:

[0039]

[0040] Where I is a matrix of all 1s; the above linear programming problem has a closed-form solution, namely and

[0041]

[0042] A further improvement of the technical solution of the present invention is that: in S3, specifically including: forming an optimization problem about x according to the closed-form solution of the auxiliary knowledge:

[0043]

[0044] The Dinkelbach algorithm is used to decouple the numerator and denominator of the fractional objective function through an iterative process; assuming that the iteration reaches k steps, the following quadratic constrained quadratic programming problem is obtained:

[0045]

[0046] Through iteration Able to form a monotonically increasing sequence of the objective function And if each iteration can obtain The global optimal solution of , when the iteration stops, can be obtained The global optimal solution of .

[0047] The further improvement of the technical solution of the present invention is that: in S4, specifically including: constructing sub-problems The dual problem of , first write the Lagrangian function, that is:

[0048]

[0049] In the formula, A 1 =A 2 =I, A 3 =R I , b 0 =b 1 =b 3 =0, b2 =-c,c 0 =0, c 1 =-E x , c 3 =-E I , Based on this, we get the dual problem:

[0050]

[0051] In the formula, represents the value space of the matrix A(λ), and

[0052]

[0053] In the formula, A(λ) (1) represents the pseudo-inverse matrix of A(λ);

[0054] Based on Schur's complement theorem The equivalent transformation is a semidefinite programming problem:

[0055]

[0056] Solving for x (k) =-A(λ * ) (1) b(λ * ), It is a convex problem and is solved in polynomial time using the interior point method.

[0057] A further improvement of the technical solution of the present invention is that in S5, the stop condition of the iterative process is set to:

[0058] f(x (k) )-f(x (k-1) )≤e (18)

[0059] Where e represents the minimum change allowed for the objective function f(x) between two adjacent iterations; if the stopping condition is met, the iteration is terminated.

[0060] A further improvement of the technical solution of the present invention is that in S6, specifically including:

[0061] According to the complementary relaxation theorem, the sufficient condition for global optimality is derived as follows: is the solution of λ when the stopping condition is met; if the optimality condition is met, it means that the duality gap is 0, and each iteration can obtain The global optimal solution; now it can be proved that: when the stopping condition is met, if e is arbitrarily small, then the iteration stops when x (k) for The global optimal solution of If not satisfied, then solve based on SDR algorithm At this time, it is necessary to obtain at the cost of greater computational complexity The global optimal solution of The SDR form is:

[0062]

[0063] In the formula, It is a convex problem; the solution is Then, for X * Perform rank-one decomposition and introduce the following rank-one decomposition theorem:

[0064] Assume X * The dimension of B is greater than or equal to 3. 1 , B 2 , B 3 and B 4 For X * Hermitian matrix of the same dimension, X * Denotes the semi-positive definite matrix to be decomposed, let rank[X * ]=r, then:

[0065] 1) If r ≥ 3, we can find a range space to which a non-zero vector s belongs in polynomial time Make

[0066] s H A i s=tr[B i X], i=1,2,3,4 (20)

[0067] 2) If r = 2, for any Not There exists a vector s belonging to the space Make

[0068] s H A i s=tr[B i X], i=1,2,3,4 (21)

[0069] 3) If r = 1, then directly perform eigenvalue decomposition;

[0070] According to the rank-one decomposition theorem, let B 2 =E, B 4 =R I , so X * It can achieve rank-one decomposition in polynomial time, and also ensure Get the global optimum.

[0071] Due to the adoption of the above technical solution, the technical progress achieved by the present invention is:

[0072] 1. The present invention allows multiple radio frequency devices to coexist in a band-limited environment, and can significantly improve the SCR of the extended target and the performance of the refined perception of the extended target.

[0073] 2. The present invention relaxes the estimation accuracy of auxiliary knowledge, creating conditions for the transformation of SAR waveform design from theoretical demonstration to engineering realization.

[0074] 3. The solution algorithm proposed in the present invention has global convergence, and the solution method based on Lagrangian duality has lower time complexity than the SDR algorithm, saves running time, and is more suitable for application scenarios of large-scale waveform design. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0076] Figure 1 It is a flow chart of a SAR robust waveform design method for extended targets in a band-limited environment provided by the present invention;

[0077] Figure 2 It is an algorithm flow chart of a SAR robust waveform design method for extended targets in a band-limited environment provided in an embodiment of the present invention;

[0078] Figure 3(a) shows the objective function sequence corresponding to different parameters ε Line chart;

[0079] Figure 3(b) shows the different parameters E I The corresponding objective function sequence Line chart;

[0080] Figure 4 Indicates different R G and R B Schematic diagram of the corresponding robust and non-robust objective functions;

[0081] Figure 5(a) shows the corresponding The minimum eigenvalue curve of ;

[0082] Figure 5(b) shows the different E I Corresponding The minimum eigenvalue curve of ;

[0083] Figure 6 It is different E I Energy Spectral Density (ESD) curve of the optimized waveform below;

[0084] Figure 7 Compare the solution based on Lagrange duality and the solution based on SDR algorithm Time complexity comparison chart;

[0085] FIG8( a ) is a SAR image generated by the algorithm proposed in the present invention;

[0086] Figure 8(b) is a SAR image generated by a suboptimal waveform;

[0087] Figure 8(c) is the SAR image generated by the Chirp signal;

[0088] Figure 8(d) is the SAR image generated by the NLFM waveform;

[0089] Fig. 9 It is a schematic diagram of target detection probability and false alarm probability. DETAILED DESCRIPTION

[0090] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0091] The present invention is further described in detail below with reference to the accompanying drawings and embodiments:

[0092] like Figure 1-2 As shown in the figure, a SAR robust waveform design method for extended targets in a band-limited environment can relax the estimation accuracy of auxiliary knowledge, achieve coexistence with other RF devices in the observation scene, and ensure high resolution and low sidelobe characteristics of SAR in the range direction. The specific steps include:

[0093] S1, select SAR echo domain signal-to-clutter ratio (SCR) as the optimization criterion, build an error model of auxiliary knowledge, impose spectrum constraints, energy constraints and similarity constraints, and construct the optimization problem of SAR robust waveform design;

[0094] Specifically, it includes: characterizing the SCR in the SAR echo domain based on the extended target scattering model. Using x(t) to represent the radar transmission signal, G(t) and B(t) to represent the target scattering characteristics and background scattering characteristics respectively, and N(t) to represent the noise, the echo signal Y(t) received at any azimuth time is expressed as:

[0095]

[0096] The convolution operator is realized by constructing Toeplitz matrix multiplication, and the discrete signal model is:

[0097] y=Gx+Bx+n (2)

[0098] In the formula, G and B represent the Toeplitz matrices formed after the discretization of G(t) and B(t), respectively. Denote the discretized vectors of x(t), Y(t), and N(t) respectively. Let R G and R B are the correlation matrices of G and B respectively, then the SCR of SAR range direction is expressed as:

[0099]

[0100] In the formula, R G and R B It is usually auxiliary knowledge estimated by historical data, empirical models, electromagnetic measurements, etc., so its error model is established as:

[0101]

[0102] In the formula, and is the ideal error-free correlation matrix, ε G and ε B To improve the robustness of the optimized waveform to the auxiliary knowledge error, the SCR corresponding to the worst auxiliary knowledge is maximized as the objective function, that is:

[0103]

[0104] To ensure the waveform can coexist with other RF systems, spectrum constraints are imposed:

[0105] x H R I x≤E I (6)

[0106] In the formula, E I is the energy threshold that other RF devices within the frequency band can tolerate, and represents the weight assigned to the interfering device, Ω qThe (m,l) element of is:

[0107]

[0108] In the formula, f 1 q and represents the frequency band range of the qth RF device; to ensure that x also has high resolution characteristics, similarity constraints are introduced to establish the maximization and minimization problem

[0109]

[0110] In the formula, E x is the energy of the maximum transmitted waveform, and ε is used to define the similarity between the transmitted waveform and the reference waveform c.

[0111] S2, solve the minimization problem of the inner layer and derive the closed-form solution of the statistics of the target scattering characteristics and the background scattering characteristics;

[0112] Specifically include: introducing two auxiliary variables ΔR G , Make

[0113] The inner minimization problem is equivalent to:

[0114]

[0115] The objective function of the above formula can be equivalently transformed into:

[0116]

[0117] Based on this, we can get the following about ΔR G and ΔR B The linear programming (LP) problem is:

[0118]

[0119] Where I is a matrix of all 1s; the above linear programming problem has a closed-form solution, namely and

[0120]

[0121] S3, construct a sub-optimization problem about the waveform, and use the Dinkelbach algorithm to decouple the numerator and denominator of the fractional objective function;

[0122] Specifically, based on the closed-form solution of auxiliary knowledge, an optimization problem about x is formed:

[0123]

[0124] The Dinkelbach algorithm is used to decouple the numerator and denominator of the fractional objective function through an iterative process. Assuming that the iteration is k steps, the following quadratic constrained quadratic programming (QCQP) problem is obtained:

[0125]

[0126] Possible proof: By iterating Able to form a monotonically increasing sequence of the objective function And if each iteration can obtain The global optimal solution of , when the iteration stops, can be obtained The global optimal solution of .

[0127] S4, based on Lagrange duality and Schur complement theorem, the subproblems in the iteration process are transformed into semidefinite programming (SDP) problems;

[0128] Specifically include: Constructor problem The dual problem of , first write the Lagrangian function, that is:

[0129]

[0130] In the formula, A 1 =A 2 =I, A 3 =R I , b 0 =b 1 =b 3 =0, b 2 =-c,c 0 =0, c 1 =-E x , c 3 =-E I , Based on this, we can get the dual problem:

[0131]

[0132] In the formula, represents the value space of the matrix A(λ), and

[0133]

[0134] In the formula, A(λ) (1) represents the pseudo-inverse matrix of A(λ);

[0135] Based on Schur's complement theorem The equivalent transformation is a semidefinite programming (SDP) problem:

[0136]

[0137] Solving for x (k) =-A(λ * ) (1) b(λ * ), It is a convex problem and is solved in polynomial time using the interior point method.

[0138] S5, determine whether the stop condition is met, if so, execute S6; if not, continue iterating;

[0139] Specifically include:

[0140] Set the stopping condition of the iteration process to:

[0141] f(x (k) )-f(x (k-1) )≤e (18)

[0142] Where e represents the minimum change allowed for the objective function f(x) between two adjacent iterations; if the stopping condition is met, the iteration is terminated.

[0143] S6, determine whether the optimality condition is met. If so, the solution process ends; if not, the semidefinite relaxation (SDR) and rank-one decomposition algorithm are used to solve the subproblem obtained in S3 until the stopping condition is met and the solution result is output.

[0144] Specifically, according to the complementary relaxation theorem, the sufficient condition for global optimality can be derived as follows: is the solution of λ when the stopping condition is met. If the optimality condition is met, it means that the duality gap is 0, and each iteration can obtain The global optimal solution of . Now it can be proved that: when the stopping condition is met, if e is arbitrarily small, then x when the iteration stops (k) for The global optimal solution of . If not satisfied, then solve based on SDR algorithm At this time, it is necessary to obtain at the cost of greater computational complexity The global optimal solution of . The SDR form is:

[0145]

[0146] In the formula, It is a convex problem; the solution is Then, for X *Perform rank-one decomposition and introduce the following rank-one decomposition theorem:

[0147] Assume X * The dimension of B is greater than or equal to 3. 1 , B 2 , B 3 and B 4 For X * Hermitian matrix of the same dimension, X * Denotes the semi-positive definite matrix to be decomposed, let rank[X * ]=r, then:

[0148] 1) If r ≥ 3, a non-zero vector s can be found in polynomial time to belong to the range space Make

[0149] s H A i s=tr[B i X], i=1,2,3,4 (20)

[0150] 2) If r = 2, for any Not There exists a vector s belonging to the space Make

[0151] s H A i s=tr[B i X], i=1,2,3,4 (21)

[0152] 3) If r=1, then eigenvalue decomposition can be performed directly.

[0153] According to the rank-one decomposition theorem, let B 2 =E, B 4 =R I , so X * It can achieve rank-one decomposition in polynomial time, and also guarantees Get the global optimum.

[0154] Example

[0155] A. Theoretical Derivation and Verification

[0156] The theoretical derivations involved in the present invention are verified through numerical experiments, including the convergence of the proposed algorithm, the robustness of the waveform to auxiliary knowledge errors, the optimality conditions, and the effectiveness of the constraints.

[0157] 1) Convergence of the proposed algorithm

[0158] Set N = 120, e = 10-4 , E x =E 0 =N, Reference signal x 0 Chirp signal. By setting different ε or E I Get different feasible domains and construct Solving the problem Get the initial iteration point x (0) , start solving The iterative steps.

[0159]

[0160] Calculate the objective function value for each iteration, such as Figure 3(a)-3(b) As shown. It can be seen that the objective function sequence is monotonically increasing, indicating that the solution algorithm has convergence. (k) ) tends to be stable after 3 iterations, triggering the stop condition, indicating that the proposed algorithm converges quickly. In addition, by increasing the parameter ε or E I , which expands the feasible domain and increases the objective function value, which is consistent with expectations.

[0161] 2) Waveform robustness

[0162] In order to evaluate the robustness of the optimized waveform to the auxiliary knowledge error, the calculations for different R G and R B The objective function value of the sample. Randomly select 100 different groups of R G and R B Samples are used to calculate the objective function value of the robust waveform. At the same time, the proposed algorithm is used to design a non-robust waveform, that is, R is not considered. G and R B The uncertainty of G and R B The objective function value of the sample is Figure 4 As shown in the figure, the optimized waveform is obtained by maximizing R G and R B The worst-case f(x) on the uncertainty set is used to obtain robustness, so the objective function value of the optimized waveform is always maintained at a high level, and the objective function values ​​corresponding to different samples are relatively stable. At the same time, the objective function value of the robust waveform in the worst case is significantly higher than the objective function value of the non-robust waveform in the worst case, further indicating that the SAR robust waveform design method for extended targets in a band-limited environment provided by the present invention can ensure that the optimized waveform has good robustness.

[0163] 3) Optimality conditions

[0164] Whether the global optimal solution of the waveform can be obtained by solving the Lagrangian dual problem, thereby achieving global convergence with lower computational complexity, it is necessary to verify the optimality condition, that is, to verify It is equivalent to eig min [·] represents the minimum eigenvalue of the matrix. I The value of , the different feasible domains constructed, are given The minimum eigenvalue of Figure 5(a)-5(b) As shown in the figure, no matter what the feasible domain is, It is always established, indicating that the optimal conditions are met and global convergence can be achieved with lower computational complexity.

[0165] 4) Frequency band compatibility

[0166] To verify whether the optimized waveform has frequency band compatibility, it is assumed that there are two RF devices working in the normalized frequency band Ω 1 =[0.27,0.37] and Ω 2 =[0.27,0.37], and the frequency band of the radar system coexisting with it is [0,1]. I To limit the energy distribution of the waveform in the stopband, the energy spectral density (ESD) of the optimized waveform is given, such as Figure 6 Here, the reference signal x 0 As shown in Figure 2, the ESD of the optimized waveform can be significantly attenuated in the stop band through spectrum constraint, preventing the emission energy from mixing into the working frequency band of other devices, causing resource waste and mutual interference. I The smaller it is, the less energy is distributed in the stop band, which is consistent with expectations, indicating that the waveform designed by the present invention allows multiple radio frequency devices to work simultaneously in a band-limited environment.

[0167] 5) Computational complexity

[0168] As mentioned above, although the SDR-based algorithm can also solve However, it will increase the time and memory burden. In order to compare the complexity of the solution method based on Lagrange duality and SDR, different sampling points of the waveform are set and the CPU running time of the two algorithms is given, as shown in Figure 7 As shown, Figure 7 It shows that the solution algorithm based on Lagrange duality has significantly lower complexity, especially when the number of waveform sampling points is large, this algorithm can greatly reduce the CPU running time, which is consistent with expectations.

[0169] B. Practical Applications in SAR

[0170] In order to verify whether the optimized waveform can improve the imaging quality of SAR for extended targets, a simple airborne SAR system is simulated to obtain SAR images of typical extended targets and conduct performance analysis. The main parameters of the SAR system are shown in Table 1:

[0171] Table 1 Main parameters of SAR system

[0172]

[0173]

[0174] The existing SAR waveforms are introduced for performance comparison, including suboptimal waveforms, Chirp signals, and NLFM signals. First, it is ensured that the different waveforms have the same transmission energy and bandwidth. The Boeing 737 passenger plane is used as the target to be measured. The target impulse response G(t) and the background scattering characteristics B(t) are filled in the specified area to generate the scattering characteristics of the observation scene. The echo signal is obtained by SAR system observation, and then the SAR image is generated by the RD imaging algorithm, such as Figure 8(a)-8(d) Compared with other images, the SAR image corresponding to the present invention has a better presentation effect for extended targets, which is manifested in a greater contrast between the target and the background. Since the present invention uses SCR as the objective function, the SCR of the optimized waveform can be improved by 3.5dB compared with the Chirp signal.

[0175] Since SCR improvement can improve target detection performance, the obtained SAR images are used to carry out target detection experiments. The amplitude of the SAR image signal is traversed in turn and set as the detection threshold; a binary image is generated according to each threshold, and the detection probability P is calculated. D and false alarm probability P FA ,like Fig. 9 shown.

[0176] It can be seen that increasing the SCR will improve the SAR image quality and further achieve a more ideal target detection result. This embodiment verifies that the waveform proposed by the present invention has better imaging observation capability for extended targets in a band-limited environment.

[0177] In summary, the present invention imposes energy constraints, similarity constraints and spectrum constraints, and highlights the scattering characteristics of extended targets by maximizing the signal-to-noise ratio corresponding to the worst auxiliary knowledge, thereby helping to improve the refined perception performance of SAR image targets; a closed-form solution to the statistics of target scattering characteristics and background scattering characteristics is derived, and the minimax problem is converted into a non-convex quadratic constrained quadratic programming problem; the non-convex quadratic constrained quadratic programming problem is decomposed into a series of solvable semidefinite programming problems by using the Dinkelbach algorithm and constructing a dual problem; a sufficient condition for global convergence is given, and at the same time, a supplementary solution method is given when the condition is not met, which will obtain the global optimal solution with greater computational complexity; it can achieve the improvement of the SCR of extended targets in a spectrum-constrained environment, and at the same time relax the estimation accuracy of auxiliary knowledge, laying a foundation for the implementation of the waveform trend optimization project.

[0178] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robust waveform design method for SAR for extended targets in a band-limited environment, characterized in that: The following steps are involved: S1, select the SAR echo domain signal-to-clutter ratio as the optimization criterion, build the error model of auxiliary knowledge, impose spectrum constraints, energy constraints and similarity constraints, and construct the optimization problem of SAR robust waveform design; S2, solve the minimization problem of the inner layer and derive the closed-form solution of the statistics of the target scattering characteristics and the background scattering characteristics; S3, construct a sub-optimization problem about the waveform, and use the Dinkelbach algorithm to decouple the numerator and denominator of the fractional objective function; S4, based on Lagrange duality and Schur complement theorem, the sub-problems in the iteration process are transformed into semidefinite programming problems; S5, determine whether the stop condition is met, if so, execute S6; if not, continue iterating; S6, determine whether the optimality condition is met. If so, the solution process ends; if not, use semidefinite relaxation and rank-one decomposition algorithm to solve the subproblem obtained in S3 until the stopping condition is met and output the solution result.

2. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S1, the following are included: The signal-to-clutter ratio (SCR) in the SAR echo domain is characterized based on the extended target scattering model; x(t) is used to represent the radar transmission signal, G(t) and B(t) represent the target scattering characteristics and background scattering characteristics respectively, and N(t) is the noise. The echo signal Y(t) received at any azimuth and time is expressed as: The convolution operator is realized by constructing Toeplitz matrix multiplication, and the discrete signal model is: y=Gx+Bx+n (2) In the formula, G and B represent the Toeplitz matrices formed after the discretization of G(t) and B(t), respectively. Represent the discretized vectors of x(t), Y(t), and N(t) respectively; let R G and R B are the correlation matrices of G and B respectively, then the SCR of SAR range direction is expressed as: In the formula, R G and R B It is auxiliary knowledge estimated by historical data, empirical models, and electromagnetic measurement methods, so its error model is established as: In the formula, and is the ideal error-free correlation matrix, ε G and ε B It is used to limit the error range; in order to improve the robustness of the optimized waveform to the auxiliary knowledge error, the SCR corresponding to the worst auxiliary knowledge is maximized as the objective function, that is: To ensure the waveform can coexist with other RF systems, spectrum constraints are imposed: x H R I x≤E I (6) In the formula, E I is the energy threshold that other RF devices within the frequency band can tolerate, and γ q ≥0 indicates the weight assigned to the interfering device, Ω q The (m,l) element of is: In the formula, and represents the frequency band range of the qth RF device; to ensure that x also has high resolution characteristics, similarity constraints are introduced to establish the maximization and minimization problem In the formula, E x is the energy of the maximum transmitted waveform, and ε is used to define the similarity between the transmitted waveform and the reference waveform c.

3. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S2, it specifically includes: Introduce two auxiliary variables Make The inner minimization problem is equivalent to: The objective function of the above formula is equivalent to: Based on this, we can get the following about ΔR G and ΔR B The linear programming problem is: Where I is a matrix of all 1s; the above linear programming problem has a closed-form solution, namely and 4. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S3, it specifically includes: Based on the closed-form solution of auxiliary knowledge, an optimization problem about x is formed: The Dinkelbach algorithm is used to decouple the numerator and denominator of the fractional objective function through an iterative process; assuming that the iteration reaches k steps, the following quadratic constrained quadratic programming problem is obtained: Through iteration Able to form a monotonically increasing sequence of the objective function And if each iteration can obtain The global optimal solution of , when the iteration stops, can be obtained The global optimal solution of .

5. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S4, it specifically includes: Constructor problem The dual problem of , first write the Lagrangian function, that is: In the formula, A1=A2=I,A3=R I , b0=b1=b3=0, b2=-c, c0=0, c1=-E x , c3=-E I , Based on this, we get the dual problem: In the formula, represents the range space of the matrix A(λ), and In the formula, A(λ) (1) represents the pseudo-inverse matrix of A(λ); Based on Schur's complement theorem The equivalent transformation is a semidefinite programming problem: Solving for x (k) = -A(λ * ) (1) b(λ * ), It is a convex problem and is solved in polynomial time using the interior point method.

6. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S5, the stopping condition of the iteration process is set as: f(x (k) )-f(x (k-1) )≤e (18) Where e represents the minimum change allowed for the objective function f(x) between two adjacent iterations; if the stopping condition is met, the iteration is terminated.

7. The method for designing a SAR robust waveform for extended targets in a band-limited environment according to claim 1, characterized in that: In S6, it specifically includes: According to the complementary relaxation theorem, the sufficient condition for global optimality is derived as follows: is the solution of λ when the stopping condition is met; if the optimality condition is met, it means that the duality gap is 0, and each iteration can obtain The global optimal solution of ; At this time, it can be proved that: when the stopping condition is met, if x is arbitrarily small, then the iteration stops when x (k) for The global optimal solution of If not satisfied, then solve based on SDR algorithm At this time, it is necessary to obtain at the cost of greater computational complexity The global optimal solution of The SDR form is: In the formula, It is a convex problem; the solution is Then, for X * Perform rank-one decomposition and introduce the following rank-one decomposition theorem: Assume X * The dimension is greater than or equal to 3, B1, B2, B3 and B4 are * Hermitian matrix of the same dimension, X * Denotes the semi-positive definite matrix to be decomposed, let rank[X * ]=r, then: 1) If r ≥ 3, we can find a range space to which a non-zero vector s belongs in polynomial time Make s H A i s=tr[B i X], i=1,2,3,4 (20) 2) If r = 2, for any Not There exists a vector s belonging to the space Make s H A i s=tr[B i X], i=1,2,3,4 (21) 3) If r = 1, then directly perform eigenvalue decomposition; According to the rank-one decomposition theorem, let B2=E, B4=R I , so X * It can achieve rank-one decomposition in polynomial time, and also ensure Get the global optimum.

Citation Information

Patent Citations

  • Joint optimal criterion based waveform design method of cognitive synthetic aperture radar (SAR)

    CN110146887A

  • Radar sparse target detection and imaging integrated algorithm based on CFAR-ADMM

    CN116500616A

  • Complementary transmitted waveform design method for realizing SAR target enhancement

    CN118330640A

  • High resolution SAR imaging using non-uniform pulse timing

    JP2012215568A

  • Lorentz constraint angle estimation method and system in non-gaussian environment

    US11994610B1

Cited By

  • FEKO-based synthetic aperture radar (SAR) waveform and range direction filter joint design method

    CN119740372A