A high-resolution radar range imaging waveform design method and imaging method

Abstract

This invention relates to the field of radar signal processing technology, and more particularly to waveform design and imaging methods for high-resolution range imaging of sparsely structured targets within radar imaging technology. The method comprises the following steps: selecting radar waveform and filter parameters; performing target structure analysis and high-resolution range imaging modeling; optimizing the waveform and filter; and performing radar high-resolution range imaging. Compared to current linear frequency modulation signal imaging, the waveform and filter designed in this invention achieve high range resolution for range imaging of sparse targets while effectively preventing the phenomenon of strong scattering points masking weak scattering points. The resulting imaging more closely approximates the target's true shape.

Description

Technical Field

[0001] The present invention belongs to the field of radar signal processing technology, and in particular relates to a waveform design method and an imaging method for high-resolution range imaging of a target with a sparse structure in radar imaging technology. Background Art

[0002] In recent years, information perception technology has become increasingly important in national strategic security and the national economy. Radar, a key component of radio information perception systems, radiates electromagnetic signals into space and receives reflected echoes to detect targets of interest. With the advancement of radar technology, more and more radars are focusing on target imaging and recognition.

[0003] Currently, one-dimensional range imaging radars generally use a wide-bandwidth signal to achieve range-dimensional imaging of targets. The most representative example is a linear frequency modulation (LFM) signal. On the one hand, the range resolution capability of a LFM signal is inversely proportional to its bandwidth, necessitating a very large transmission bandwidth to improve target resolution. In today's increasingly crowded electromagnetic spectrum, a large transmission bandwidth is susceptible to radio frequency interference at certain frequency bands (or points). On the other hand, the maximum sidelobe level of a LFM signal after matched filtering is approximately -13dB, making it easy for a strong scattering target to overwhelm an adjacent weaker scattering target. Both of these issues arise because the use of LFM signals for one-dimensional range imaging of targets of interest fails to account for the potential for unique target structures. When such structures are present, the LFM signal's one-dimensional range imaging performance can degrade dramatically. With the continuous advancement of electronic information technology, radar waveform generators are now capable of generating a variety of transmit waveforms, and non-matched filtering can be employed for imaging processing. At the same time, when detecting and imaging targets of interest, radars often have prior information about the targets. Using radar transmission waveforms designed according to the target's unique structure and corresponding signal processing methods can significantly improve the radar's imaging capabilities.

[0004] In view of this, according to the special structure of the target of interest, the one-dimensional range imaging waveform suitable for the target of interest and the corresponding imaging processing method are studied, which can effectively make up for the shortcomings of the traditional linear frequency modulation signal and matched filter imaging methods in imaging capabilities, and improve the imaging performance and recognition ability of the target of interest. Summary of the Invention

[0005] In response to the problems that traditional linear frequency modulation signals cannot adapt to target structures, have low resolution, and strong targets mask weak targets, the present invention provides a high-resolution radar range imaging waveform design and imaging method that can match the sparse structure of targets. This waveform can fully utilize the sparse structure of the target and, combined with an appropriate imaging method, effectively solves the problem of strong targets masking weak targets, and has high target range resolution.

[0006] The technical solution adopted by the present invention is a high-resolution radar range imaging waveform design method that matches the sparse structure of the target. The implementation steps of this method are as follows:

[0007] S1. Select radar waveform and filter parameters;

[0008] S2. Target structure analysis and high-resolution range imaging modeling;

[0009] S3. Optimize waveform and filter.

[0010] Further;

[0011] S11, range imaging radar transmits phase coded waveform;

[0012] S12. Determine the coding length of the radar waveform and the length of the receiving filter weight coefficient according to the position and structure of the target of interest in the scene;

[0013] S13. Determine the radar waveform coding rate (radar system sampling rate) based on the target range resolution expected by the radar.

[0014] Furthermore, the range imaging radar transmits a phase coded waveform, and its discrete baseband waveform is represented by a vector s = [s(1), s(2), ..., s(N)] T express;

[0015] Where N represents the symbol length of the transmitted waveform, s(i) represents the i-th sampling point of the radar transmitted waveform, and the superscript T represents the vector transpose operation;

[0016] The filter weight coefficients of the radar receiver are expressed as w = [w(1), w(2), ..., w(N)] T Indicates that w(i) represents the i-th weight coefficient of the filter; the sampling rate of the radar system is f s , then the total time width of the radar transmission waveform is N / f s , the distance resolution corresponding to each sampling point is c / (2f s ), where c represents the speed of light;

[0017] The parameters N and f are selected based on the actual size of the target of interest and the desired resolution of the radar system. s,The selection method is: On the one hand, the larger the actual size of the target, the larger the parameter N needs to be, to ensure that the radar system has multiple degrees of freedom,

[0018] Assuming the target size is L radar resolution units, the code element length N is set to 10L;

[0019] On the other hand, the higher the range resolution that the radar expects to obtain, the higher the sampling rate f of the system s Also larger to ensure that c / (2f s ) is smaller than the desired radar range resolution.

[0020] Further;

[0021] S21. Using non-matched filtering to perform target range imaging, the ideal waveform and filter cross-correlation sequence is a nail-shaped impulse function.

[0022] S22, using the ideal cross-correlation sequence as a template, analyzing the position of the target of interest in the scene, and utilizing the sparsity structure of the target to determine the weight of the cross-correlation sequence, and matching the template using a weighted least squares method;

[0023] S23, weighting the area where the target of interest exists and the background area of ​​no interest is set.

[0024] S24. Establish a joint optimization model based on the imaging waveform and the filter weights.

[0025] Further;

[0026] The shape of the cross-correlation sequence between the transmit waveform s and the filter weight coefficient w determines the performance of radar range imaging. After completing the selection of radar waveform parameters, the cross-correlation sequence y between the transmit waveform s and the filter weight coefficient w is a column vector of length 2N-1, expressed as:

[0027] y=w*s,

[0028] Where * represents the convolution operator, and the above formula is expressed in matrix form:

[0029] y=Ws=Sw

[0030] in,

[0031]

[0032] The ideal waveform and filter cross-correlation sequence is a nail-shaped impulse function, set as y0, the Nth element of the cross-correlation sequence y0 is 1, and the rest of the elements are 0;

[0033] In order to achieve high imaging performance, the cross-correlation sequence y between the waveform and the filter should be close to y0, and the 2-norm of the Euclidean space is used to measure the difference between y and y0. On the other hand, considering that the target of interest has a sparse structure in space, the weight vector α = [α(1), α(2), ..., α(2N-1)] is used. T The difference between y and y0 is weighted to ensure that the target area of ​​interest receives more attention and the background area of ​​no interest is ignored. That is, the 2L+1 weights α(NL) to α(N+L) are set to large values, and the rest of the weights are set to small values.

[0034] The design of the radar waveform and filter aims to minimize the difference between the weighted cross-correlation sequence y and the ideal cross-correlation sequence y0, which is achieved by optimizing the following equation:

[0035]

[0036] Where Λ = diag(α) is a diagonal matrix with weight vector α arranged along the diagonal, and ||·||2 is the 2-norm of the Euclidean space;

[0037] In addition, considering that the radar waveform uses a phase-coded signal with a constant modulus value, after adding a constant modulus constraint to the radar waveform during optimization, the radar high-resolution range imaging waveform and filter design are modeled as the following optimization problem:

[0038]

[0039] Further;

[0040] S31, transmit waveform optimization and filter weight vector optimization;

[0041] S32, alternately looping until convergence, in the transmit waveform optimization step, the filter weight vector remains fixed, and the optimization variable is the phase code element of each fast time domain sampling point of the transmit waveform;

[0042] S33. In the filter weight vector optimization step, the transmit waveform remains fixed, and the optimization variables are the amplitude and phase of the filter weight vector.

[0043] Further;

[0044] Transmit waveform optimization

[0045] In the kth optimization emission waveform s (k) When , the filter weight vector is fixed to the value w at the k-1th iteration (k -1) , the optimization problem is reformulated as

[0046]

[0047] in,

[0048]

[0049] Use the coordinate descent method to solve the equation, that is, update s each time (k) An element in , the i-th element s (k) The update expression of (i) is as follows

[0050]

[0051] Where exp{·} represents the exponential function, j is the imaginary unit, arg(·) represents the phase operation, v=(W (k-1) ) H Λ H Λy0 and U=(W (k-1) ) H Λ H ΛW (k-1) , the superscript H represents the conjugate transpose operation of the matrix;

[0052] According to the update expression given by the formula, traverse s in turn (k) Each element in , and loop until convergence, and get the emission waveform s after the kth iteration optimization (k) ;

[0053] Filter weight vector optimization

[0054] In the kth optimization filter weight vector w (k) When , the emission waveform is fixed at the value s at the kth iteration (k) , the optimization problem is reformulated as

[0055]

[0056] in:

[0057]

[0058] From the structure of the analytical formula, it can be found that the solution is equivalent to solving the least squares solution of the following linear variance:

[0059] ΛS (k) w=Λy0,

[0060] Therefore, at the kth iteration, the optimization expression of the filter weight vector is:

[0061] w (k) =(ΛS (k) ) + Λy0

[0062] Among them, (ΛS (k) )+ Represents the matrix ΛS (k) The Moore-Penrose generalized inverse matrix of .

[0063] The present invention also provides a high-resolution radar range imaging method, comprising:

[0064] After step S3, the transmit waveform and filter weight vector have converged to the optimized transmit waveform s * and the filter weight vector w * , when imaging the target of interest, s * As the actual transmission waveform of the radar, the received target echo is recorded as r and the filter weight vector w * After performing convolution processing and taking amplitude calculation, the range image of the target is obtained.

[0065] Beneficial effect: Compared with the current linear frequency modulation signal imaging, the waveform and filter designed in the present invention can effectively solve the problem of strong scattering points covering weak scattering points when performing distance imaging on sparse targets while achieving high distance resolution, and the imaging result is closer to the true shape of the target. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 The high-resolution radar range imaging waveform design and imaging method proposed by the present invention that matches the sparse structure of the target;

[0067] Figure 2 A schematic diagram of the simulated sparse target scattering intensity provided by an example of the present invention;

[0068] Figure 3 A schematic diagram of a radar transmission waveform phase symbol provided by an example of the present invention;

[0069] Figure 4 The imaging results of the present invention provided by the examples of the present invention;

[0070] Figure 5 This is the imaging result of the traditional linear frequency modulation waveform of the present invention. DETAILED DESCRIPTION

[0071] The above embodiments are provided to specifically describe the present invention and are intended only to further illustrate the present invention. They are not to be construed as limiting the scope of protection of the present invention. Any non-essential improvements and adjustments made by those skilled in the art based on the contents of the present invention fall within the scope of protection of the present invention. To better illustrate the present invention and facilitate understanding, the present invention is described in detail below with reference to the accompanying drawings and through specific implementation methods.

[0072] Example 1:

[0073] A high-resolution radar range imaging waveform design method and imaging method, comprising the following steps:

[0074] 1. Radar waveform and filter parameter selection

[0075] The range imaging radar transmits a phase-coded waveform, and its discrete baseband waveform is represented by the vector s = [s(1), s(2), ..., s(N)] T express;

[0076] Where N represents the symbol length of the transmitted waveform, s(i) represents the i-th sampling point of the radar transmitted waveform, and the superscript T represents the vector transpose operation;

[0077] The filter weight coefficients of the radar receiver are expressed as w = [w(1), w(2), ..., w(N)] T Indicates that w(i) represents the i-th weight coefficient of the filter; the sampling rate of the radar system is f s , then the total time width of the radar transmission waveform is N / f s , the distance resolution corresponding to each sampling point is c / (2f s ), where c represents the speed of light;

[0078] The parameters N and f are selected based on the actual size of the target of interest and the desired resolution of the radar system. s ,The selection method is: On the one hand, the larger the actual size of the target, the larger the parameter N needs to be, to ensure that the radar system has multiple degrees of freedom,

[0079] Assuming the target size is L radar resolution units, the code element length N is set to 10L;

[0080] On the other hand, the higher the range resolution that the radar expects to obtain, the higher the sampling rate f of the system s Also larger to ensure that c / (2f s ) is smaller than the desired radar range resolution.

[0081] 2. Target structure analysis and high-resolution range imaging modeling

[0082] The shape of the cross-correlation sequence between the transmit waveform s and the filter weight coefficient w determines the performance of radar range imaging. After completing the selection of radar waveform parameters, the cross-correlation sequence y between the transmit waveform s and the filter weight coefficient w is a column vector of length 2N-1, expressed as:

[0083] y=w*s,

[0084] Where * represents the convolution operator, and the above formula is expressed in matrix form:

[0085] y=Ws=Sw

[0086] in,

[0087]

[0088] The ideal waveform and filter cross-correlation sequence is a nail-shaped impulse function, set as y0, the Nth element of the cross-correlation sequence y0 is 1, and the rest of the elements are 0;

[0089] In order to achieve high imaging performance, the cross-correlation sequence y between the waveform and the filter should be close to y0, and the 2-norm of the Euclidean space is used to measure the difference between y and y0. On the other hand, considering that the target of interest has a sparse structure in space, the weight vector α = [α(1), α(2), ..., α(2N-1)] is used. T The difference between y and y0 is weighted to ensure that the target area of ​​interest receives more attention and the background area of ​​no interest is ignored. That is, the 2L+1 weights α(NL) to α(N+L) are set to large values, and the rest of the weights are set to small values.

[0090] The design of the radar waveform and filter aims to minimize the difference between the weighted cross-correlation sequence y and the ideal cross-correlation sequence y0, which is achieved by optimizing the following equation:

[0091]

[0092] Where Λ = diag(α) is a diagonal matrix with weight vector α arranged along the diagonal, and ||·||2 is the 2-norm of the Euclidean space;

[0093] In addition, considering that the radar waveform uses a phase-coded signal with a constant modulus value, after adding a constant modulus constraint to the radar waveform during optimization, the radar high-resolution range imaging waveform and filter design are modeled as the following optimization problem:

[0094]

[0095] 3. Transmit waveform optimization

[0096] In the kth optimization emission waveform s (k) When , the filter weight vector is fixed to the value w at the k-1th iteration (k -1) , the combined formula reformulates the optimization problem as:

[0097]

[0098] in,

[0099]

[0100] Use the coordinate descent method to solve the equation, that is, update s each time(k) An element in , the i-th element s (k) The update expression of (i) is as follows:

[0101]

[0102] Where exp{·} represents the exponential function, j is the imaginary unit, arg(·) represents the phase operation, v=(W (k-1) ) H Λ H Λy0 and U=(W (k-1) ) H Λ H ΛW (k-1) , the superscript H represents the conjugate transpose operation of the matrix;

[0103] According to the update expression given by the formula, traverse s in turn (k) Each element in , and loop until convergence, and get the emission waveform s after the kth iteration optimization (k) ;

[0104] 4. Filter weight vector optimization

[0105] In the kth optimization filter weight vector w (k) When , the emission waveform is fixed at the value s at the kth iteration (k) , the combined formula reformulates the optimization problem as:

[0106]

[0107] in:

[0108]

[0109] From the structure of the analytical formula, it can be found that the solution is equivalent to solving the least squares solution of the following linear variance:

[0110] ΛS (k) w=Λy0,

[0111] Therefore, the optimization expression of the filter weight vector at the kth iteration is:

[0112] w (k) =(ΛS (k) ) + Λy0

[0113] Among them, (ΛS (k) ) + Represents the matrix ΛS (k) The Moore-Penrose generalized inverse matrix of .

[0114] The present invention also provides a high-resolution radar range imaging method. After the S3 step, the transmission waveform and the filter weight vector have converged to the optimized transmission waveform s * and the filter weight vector w * , when imaging the target of interest, s * As the actual transmission waveform of the radar, the received target echo is recorded as r and the filter weight vector w * After performing convolution processing and taking amplitude calculation, the distance direction of the target is obtained. Figure 4 and Figure 5 This is a schematic diagram comparing the imaging results provided by the present invention with the current imaging results, where Figure 4 The imaging results provided by the present invention are Figure 5 This is the imaging result of the traditional linear frequency modulation waveform.

[0115] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. 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 make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A high-resolution radar range imaging waveform design method, characterized in that: The following steps are involved: S1. Select radar waveform and filter parameters; S2, performing target structure analysis and high-resolution range imaging modeling; said step S2 includes: S21, using non-matched filtering to perform target range imaging, the ideal waveform and filter cross-correlation sequence is a nail-shaped impact function, S22, using the ideal cross-correlation sequence as a template, by analyzing the position of the target of interest in the scene and using the sparsity structure of the target, determining the weight of the cross-correlation sequence, and using the weighted least squares method to match the template, S23, weighting the area where the target of interest exists and the background area of ​​no interest is set. S24, establishing a joint optimization model based on the imaging waveform and the filter weight; S3. Optimize waveform and filter.

2. The high-resolution radar range imaging waveform design method according to claim 1, wherein: The step S1 comprises: S11, range imaging radar transmits phase coded waveform; S12. Determine the coding length of the radar waveform and the length of the receiving filter weight coefficient according to the position and structure of the target of interest in the scene; S13. Determine the coding rate of the radar waveform based on the target range resolution expected by the radar.

3. The high-resolution radar range imaging waveform design method according to claim 2, wherein: The range imaging radar transmits a phase-coded waveform, and its discrete baseband waveform is represented by the vector s = [s(1), s(2), ..., s(N)] T express; Where N represents the symbol length of the transmitted waveform, s(i) represents the i-th sampling point of the radar transmitted waveform, and the superscript T represents the vector transpose operation; The filter weight coefficients of the radar receiver are expressed as w = [w(1), w(2), ..., w(N)] T Indicates that w(i) represents the i-th weight coefficient of the filter; the sampling rate of the radar system is f s , then the total time width of the radar transmission waveform is N / f s , the distance resolution corresponding to each sampling point is c / (2f s ), where c represents the speed of light; The parameters N and f are selected based on the actual size of the target of interest and the desired resolution of the radar system. s ,The selection method is: On the one hand, the larger the actual size of the target, the larger the parameter N needs to be, to ensure that the radar system has multiple degrees of freedom, Assuming the target size is L radar resolution units, the code element length N is set to 10L; On the other hand, the higher the range resolution that the radar expects to obtain, the higher the sampling rate f of the system s Also larger to ensure that c / (2f s ) is smaller than the desired radar range resolution.

4. The high-resolution radar range imaging waveform design method according to claim 1, wherein: The shape of the cross-correlation sequence between the transmit waveform s and the filter weight coefficient w determines the performance of radar range imaging. After completing the selection of radar waveform parameters, the cross-correlation sequence y between the transmit waveform s and the filter weight coefficient w is a column vector of length 2N-1, expressed as: y=w*s, Where * represents the convolution operator, and the above formula is expressed in matrix form: y=Ws=Sw in, The ideal waveform and filter cross-correlation sequence is a nail-shaped impulse function, set as y0, the Nth element of the cross-correlation sequence y0 is 1, and the rest of the elements are 0; In order to achieve high imaging performance, the cross-correlation sequence y between the waveform and the filter should be close to y0, and the 2-norm of the Euclidean space is used to measure the difference between y and y0. On the other hand, considering that the target of interest has a sparse structure in space, the weight vector α = [α(1), α(2), ..., α(2N-1)] is used. T The difference between y and y0 is weighted to ensure that the target area of ​​interest receives more attention and the background area of ​​no interest is ignored. That is, the 2L+1 weights α(NL) to α(N+L) are set to large values, and the rest of the weights are set to small values. The design of the radar waveform and filter aims to minimize the difference between the weighted cross-correlation sequence y and the ideal cross-correlation sequence y0, which is achieved by optimizing the following equation: Where Λ = diag(α) is a diagonal matrix with weight vector α arranged along the diagonal, and ||·||2 is the 2-norm of the Euclidean space; In addition, considering that the radar waveform uses a phase-coded signal with a constant modulus value, after adding a constant modulus constraint to the radar waveform during optimization, the radar high-resolution range imaging waveform and filter design are modeled as the following optimization problem:

5. The high-resolution radar range imaging waveform design method according to claim 4, wherein: The step S3 comprises: S31, transmit waveform optimization and filter weight vector optimization; S32, alternately looping until convergence, in the transmit waveform optimization step, the filter weight vector remains fixed, and the optimization variable is the phase code element of each fast time domain sampling point of the transmit waveform; S33. In the filter weight vector optimization step, the transmit waveform remains fixed, and the optimization variables are the amplitude and phase of the filter weight vector.

6. The high-resolution radar range imaging waveform design method according to claim 5, characterized in that: include: Transmit waveform optimization In the kth optimization emission waveform s (k) When , the filter weight vector is fixed to the value w at the k-1th iteration (k-1) , the optimization problem is reformulated as in, Use the coordinate descent method to solve the equation, that is, update s each time (k) An element in , the i-th element s (k) The update expression of (i) is as follows Where exp{·} represents the exponential function, j is the imaginary unit, arg(·) represents the phase operation, v=(W (k-1) ) H Λ H Λy0 and U=(W (k-1) ) H Λ H ΛW (k-1) , the superscript H indicates the conjugate transpose operation of the matrix, p is the loop variable when summing, traversing from 1 to N, but skipping the case of p = i (that is, excluding the current item i itself); According to the update expression given by the formula, traverse s in turn (k) Each element in , and loop until convergence, and get the emission waveform s after the kth iteration optimization (k) ; Filter weight vector optimization In the kth optimization filter weight vector w (k) When , the emission waveform is fixed at the value s at the kth iteration (k) , the optimization problem is reformulated as in: From the structure of the analytical formula, it can be found that the solution is equivalent to solving the least squares solution of the following linear variance: ΛS (k) w=Λy0, Therefore, at the kth iteration, the optimization expression of the filter weight vector is: w (k) =(ΛS (k) ) + Λy0 Among them, (ΛS (k) ) + Represents the matrix ΛS (k) The Moore-Penrose generalized inverse matrix of .

7. A high-resolution radar range imaging method according to claim 6, characterized in that: Perform radar high-resolution range imaging processing; including: After step S3, the transmit waveform and filter weight vector have converged to the optimized transmit waveform s * and the filter weight vector w * , when imaging the target of interest, s * As the actual transmission waveform of the radar, the received target echo is then compared with the filter weight vector w * After performing convolution processing and taking amplitude calculation, the range image of the target is obtained.

Images