Through-the-wall radar grating lobe suppression method based on multi-bit phase quantization
By quantizing the multi-bit phase of the through-wall radar image to generate phase weights, suppressing the orientation gate lobe and side lobe, the problems of image distortion and false alarms in the through-wall radar are solved, and more accurate target detection and image quality improvement are achieved.
Patent Information
- Application Number
- CN202411115531.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-12-03
AI Technical Summary
Due to the limited frequency range of ultra-wideband signals in the wall-passing radar, the radar image generates distance to side lobes and directional grid lobes, which affects the accuracy of target detection and false alarm rate. The existing suppression methods include main lobe widening and side lobe lifting. , the impact of target detection, noise sensitivity and target fragmentation, etc.
Using a multi-bit phase quantization method, the imaging results of each channel of the radar image are quantized, and the phase weight is generated, and the original image is weighted using the quantized phase consistency to suppress the orientation gate lobe and side lobe.
Effectively suppress the orientation gate lobes and side lobes in wall-through radars, improve the accuracy and image quality of target detection, and are suitable for wall-through radars and other radar imaging fields, and can adjust the suppression effect by adjusting the quantization number of bits.
Smart Images

Figure CN119066295A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of through-wall radar imaging detection, and in particular relates to a through-wall radar grating lobe suppression method based on multi-bit phase quantization. Background Art
[0002] Through-the-wall imaging radar utilizes the penetrating characteristics of L / S band electromagnetic waves to detect targets in shielded spaces and is widely used in disaster relief, etc.
[0003] In through-the-wall imaging radar, the transmitting antenna transmits an ultra-wideband signal (such as a stepped frequency signal or a linear frequency modulation signal) to detect the target. However, due to the limited frequency range of the ultra-wideband signal, the radar image produces range sidelobes. It can be removed by windowing. In order to balance the contradiction between azimuth resolution and data acquisition rate, the array of through-the-wall imaging radar often adopts sparse MIMO distribution. Therefore, the radar image distortion leads to false alarms in target detection due to azimuth grating lobes and high sidelobes.
[0004] In order to suppress azimuth grating lobes and side lobes, many methods have been proposed. First, some scholars proposed a method to optimize the position of the array and the gain of each array element using constraints to suppress azimuth grating lobes, but such methods will cause the main lobe to widen or the side lobe to rise, and their practicality is limited. In addition, some scholars proposed a method of compressed sensing imaging to circumvent the grating lobe problem, but this method will affect the detection of weak targets and is sensitive to noise. In addition, neural networks have been used for grating lobe suppression in recent years, but their generalization in different radars needs further discussion. On the other hand, the difference between azimuth grating lobes and side lobes and targets in some aspects can be used to suppress azimuth grating lobes and side lobes while retaining targets. The MA method uses the consistency of ultra-wideband signal frequency to suppress grating lobes in through-wall radar images, but this method still has residual grating lobes. Tu scholars used the correlation of different subarrays to achieve better grating lobe and sidelobe suppression effects. The grating lobe suppression method based on phase coherence factor uses the consistency of the phase of each channel to suppress grating lobes, but the real-time performance of this method needs to be improved. As an extreme variant of the PCF method, SCF improves the real-time performance of the method, but it also causes the problem of target fragmentation. Summary of the invention
[0005] To solve the above problems, the present invention provides a through-wall radar grating lobe suppression method based on multi-bit phase quantization, which effectively suppresses the azimuth grating lobe and side lobe by weighting the original radar image.
[0006] A through-wall radar grating lobe suppression method based on multi-bit phase quantization comprises the following steps:
[0007] S1: Calculate the imaging result I(C, k) of each radar channel at the imaging grid point C based on the radar echo, where the value range of the channel number k is [1, Q×P], where Q is the number of receiving antennas on the radar and P is the number of transmitting antennas on the radar;
[0008] S2: According to the set quantization bit number n q The phase of the imaging result I(C, k) of each channel is quantized to obtain the quantized phase θ of each channel. C,k :
[0009]
[0010] in, is the phase of the imaging result I(C, k) of each channel, i Indicates the number of quantized bits n in the phase region [0,2π) q Evenly divided phase intervals, with interval number i = 1, ..., n q ;
[0011] S3: According to the quantized phase θ of each channel C,k Obtaining a weight W(C) for performing grating lobe suppression;
[0012] S4: Suppress the grating lobe of each imaging result I(C,k) according to the weight W(C) to obtain the image I′(C) after the azimuth grating lobe and side lobe are suppressed:
[0013] I′(C)=I(C)·W ν (C)
[0014] Where, ν is the set weighted control factor, the image before azimuth grating lobe and side lobe suppression
[0015] Furthermore, in step S1, the imaging result I(C, k) is obtained by:
[0016] Taking the radar center as the origin, assuming the coordinates of the pth transmitting antenna are (x p ,0,z p ), the coordinates of the qth receiving antenna are (x q ,0,z q );
[0017] The radar transmitting antenna transmits an ultra-wideband stepped frequency signal containing N frequency points by time division multiplexing, where the ultra-wideband stepped frequency signal x trans (t) is:
[0018] x trans (t)=exp(-j2π(f0+n△f)t),n=0,…,N-1
[0019] Wherein, f0 is the initial frequency of the ultra-wideband stepped frequency signal, △f is the frequency step interval, j represents the imaginary part, and t represents the time;
[0020] The kth channel is coupled to the ultra-wideband stepped frequency signal x trans (t) After down-mixing and frequency sampling, the target echo r is obtained k (n):
[0021] r k (n) = exp(-j2π(f0+n△f)τ k )
[0022] Among them, the kth channel refers to the ultra-wideband stepped frequency signal x trans (t) is transmitted by the pth transmit antenna and received by the qth receive antenna, τ k =R k / c+△τ k represents the signal propagation time delay of the kth channel, R k represents the distance between the kth channel and the target, △τ k represents the additional delay of the kth channel;
[0023] The target echo r is projected using the coherent backscattering method. k (n) is processed to obtain the imaging result I(C,k):
[0024]
[0025] Among them, τ C,k represents the time delay between the kth channel and the imaging grid point C.
[0026] Furthermore, the weight W(C) in step S3 is calculated as follows:
[0027] W(C)=1-std(θ C )
[0028] Among them, the phase sequence θ C =[θ C,1 … θ C,QP ] T , std(·) represents the standard deviation.
[0029] Furthermore, in step S2, the phase of the imaging result I(C, k) of each channel The calculation method is:
[0030]
[0031] Here, angle(·) represents a phase operation.
[0032] Further, in step S2, the phase interval is i The division method is:
[0033]
[0034] in, is a set of integers.
[0035] Beneficial effects:
[0036] The present invention provides a through-wall radar grating lobe suppression method based on multi-bit phase quantization. Firstly, a back-projection algorithm is used to calculate the imaging results of each channel for a radar receiving signal; then, the radar imaging phase of each channel is quantized according to a specified number of quantization bits; finally, a weighted image is generated by using the quantized phase to weight the original image, thereby suppressing grating lobes and side lobes. The present invention has better grating lobe and side lobe suppression performance for through-wall radar imaging than the existing suppression method, and the present invention can adjust the suppression effect by adjusting the number of quantization bits, so the present invention is not only applicable to the through-wall radar imaging detection field, but also applicable to various radar imaging, microwave imaging and other fields. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 It is a schematic diagram of the through-wall radar detection scenario and radar array;
[0038] Figure 2 It is a flow chart of the method for suppressing grating lobes and side lobes of through-wall radar based on multi-bit phase quantization of the present invention;
[0039] Figure 3 These are the imaging results of different algorithms. (a) BP (b) AD (c) PCF (d) SCF (e) The algorithm proposed by the present invention (the number of quantization bits is 8) (f) The algorithm proposed by the present invention (the number of quantization bits is 57);
[0040] Figure 4 The calculation results of indicators of different algorithms vary with the number of quantization bits. (a) Peak sidelobe ratio (b) Image entropy (c) Image contrast;
[0041] Figure 5 The calculation results of indicators of different algorithms change with the input signal-to-noise ratio. (a) Peak sidelobe ratio (b) Image entropy (c) Image contrast;
[0042] Figure 6 It is a schematic diagram of the experimental scene;
[0043] Figure 7 This is the performance of the present invention in the actual measurement results. (a) Original radar image (b) Radar image of the present invention (the number of quantization bits is 3). DETAILED DESCRIPTION
[0044] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.
[0045] The present invention provides a through-wall radar grating lobe suppression method based on multi-bit phase quantization, which is applied to through-wall radar imaging detection. First, the phase of each channel of the same imaging grid point is quantized according to the specified number of quantization bits. Then, the weighted value of the imaging grid point is obtained according to the consistency of the quantized phase. Finally, the original radar image is weighted to achieve azimuth grating lobe and sidelobe suppression. This method can not only be used for through-wall radar imaging, but also can be used in fields such as synthetic aperture radar imaging. It is a universal radar image grating lobe and sidelobe suppression method.
[0046] Specifically, a through-wall radar grating lobe suppression method based on multi-bit phase quantization includes the following steps:
[0047] S1: Calculate the imaging result I(C, k) of each radar channel at the imaging grid point C based on the radar echo, where the value range of the channel number k is [1, Q×P], where Q is the number of receiving antennas on the radar and P is the number of transmitting antennas on the radar;
[0048] Furthermore, the imaging result I(C, k) is obtained by:
[0049] A P-transmit Q-receive MIMO radar is placed against a wall. Figure 1 As shown in a. Taking the radar center as the origin of coordinates, assuming that the coordinates of the pth transmitting antenna are (x p ,0,z p ), the coordinates of the qth receiving antenna are (x q ,0,z q );
[0050] The radar transmitting antenna transmits an ultra-wideband stepped frequency signal containing N frequency points by time division multiplexing, where the ultra-wideband stepped frequency signal x trans (t) is:
[0051] x trans (t)=exp(-j2π(f0+n△f)t),n=0,…,N-1
[0052] Wherein, f0 is the initial frequency of the ultra-wideband stepped frequency signal, △f is the frequency step interval, j represents the imaginary part, and t represents the time;
[0053] It should be noted that the radar echo contains wall clutter and target echo, wherein the wall clutter can be suppressed by some preprocessing methods. The subsequent process of the present invention assumes that the clutter has been removed by the preprocessing method; based on this, the kth channel performs a multi-channel ... trans (t) After down-mixing and frequency sampling, the target echo r is obtained k (n):
[0054] r k (n) = exp(-j2π(f0+n△f)τ k )
[0055] Among them, the kth channel refers to the ultra-wideband stepped frequency signal x trans (t) is transmitted by the pth transmit antenna and received by the qth receive antenna, τ k =R k / c+△τ k represents the signal propagation time delay of the kth channel, R k represents the distance between the kth channel and the target, △τ k represents the extra delay of the kth channel, which is caused by the propagation of electromagnetic waves through walls and can be accurately calculated by some methods;
[0056] The target echo r is projected using the coherent backscattering method. k (n) is processed to obtain the imaging result I(C,k) of the kth channel at the imaging grid point C:
[0057]
[0058] Among them, τ C,k represents the time delay between the kth channel and the imaging grid point C, which is calculated in the same way as τ k Consistent.
[0059] It should be noted that in through-wall imaging radar, the coherent backscatter projection method is often used to generate radar images. Based on the imaging results I(C, k) of each channel, the imaging result at the imaging grid point C can be expressed as
[0060]
[0061] S2: According to the set quantization bit number n q The phase of the imaging result I(C, k) of each channel is quantized to obtain the quantized phase θ of each channel. C,k ;
[0062] The specific steps are as follows:
[0063] Get the phase of the imaging result I(C,k) of each channel
[0064]
[0065] Here, angle(·) represents a phase operation.
[0066] Then according to the number of quantization bits n q The phase of the imaging result I(C,k) of each channel To quantify:
[0067]
[0068] in, is the phase of the imaging result I(C, k) of each channel, i Indicates the number of quantized bits n in the phase region [0,2π) q Evenly divided phase intervals, with interval number i = 1, ..., n q θ C,k is the quantized phase of each channel;
[0069] When the number of quantization bits is nq, the phase region [0,2π) can be evenly divided into n q intervals, where the i-th th (i=1,…,n q ) intervals can be expressed as:
[0070]
[0071] in, is a set of integers.
[0072] S3: According to the quantized phase θ of each channel C,k Get the weight W(C) for grating lobe suppression:
[0073] W(C)=1-std(θ C )
[0074] Among them, the phase sequence θ C =[θ C,1 … θ C,QP ] T represents a vector composed of the phases of each channel at the grid point C, and std(·) represents the standard deviation. In the present invention, the standard deviation is used to measure the consistency of the phases of each channel after quantization.
[0075] S4: Suppress the grating lobe of each imaging result I(C,k) according to the weight W(C) to obtain the image I′(C) after the azimuth grating lobe and side lobe are suppressed:
[0076] I′(C)=I(C)·W ν (C)
[0077] Among them, ν is the set weighted control factor, which can control the influence of the weight on the original image, the image before azimuth grating lobe and sidelobe suppression
[0078] That is to say, step S4 forms a weighted image to suppress grating lobes or side lobes according to the consistency of the quantized phase, that is, the original image is weighted by the weight value, so as to obtain the image I′(C) after the azimuth grating lobe and side lobe are suppressed.
[0079] Assuming that the grid point where the target is located is T, from the geometric relationship, we know that τ k =τ T,k Therefore, for all channels, The phases of all channels are located in the first phase quantization interval, and the quantized phases are consistent. Therefore, the target can be completely retained. When the radar echo is contaminated by noise, the phases of each channel at the target position will be offset, resulting in a decrease in phase consistency. The first advantage of quantizing the phase is that it can improve the phase consistency at the target position and retain the target image to a higher degree. For the grid points where the azimuth grating lobe or side lobe is located, due to the position difference of each channel array element, there is τ k -τ C,k ≠τ γ -τ C,γ (γ=1,...,QP andγ≠k). Therefore, the phases of each channel are inconsistent, resulting in lower weights for the azimuth grating lobe or side lobe positions, thereby suppressing them. The second advantage of phase quantization is that it can magnify the phase difference of non-target positions. For example, for two phases with a very small difference but spanning the quantization interval, quantization can aggravate the inconsistency between the two phases.
[0080] The method proposed in the present invention is essentially a generalized PCF method. q =2, the method proposed in this paper is consistent with SCF. q →∞, the method proposed by the present invention is consistent with the PCF method. The algorithm flow of the method proposed by the present invention is as follows Figure 2 shown.
[0081] The performance of the present invention will be compared with the original imaging results, the PCF method, the SCF method and the AD method. The imaging performance of each method will be measured by three indicators: peak sidelobe ratio, image entropy and image contrast, and the calculation method can be expressed as
[0082]
[0083] A 10-transmit 10-receive ultra-wideband MIMO radar detects targets behind a wall. The array structure is as follows: Figure 1(b) shows. The radar transmits a stepped frequency signal with an initial frequency of 1.9 GHz and a frequency interval of 2 MHz. Taking the center of the array as the origin of the coordinate system, the target behind the wall is located at (0m, 5m, 0m). The wall echo has been removed by preprocessing, and the additional delay of the electromagnetic wave generated by the wall can be accurately estimated. The radar images obtained using different algorithms are as follows: Figure 3 shown.
[0084] Figure 3 Radar images processed by different algorithms. Figure 3 (a) It can be seen that the original radar image is seriously affected by the range sidelobes and azimuth sidelobes, resulting in increased false alarms in target detection. The range sidelobes can be removed by other methods. This paper focuses on the suppression of azimuth grating lobes and sidelobes. After the AD algorithm, the residual grating lobes will still affect the radar image. PCF and SCF can effectively suppress azimuth grating lobes and sidelobes, but the SCF method will face target segmentation. Figure 3 (e) and Figure 3 (f) It can be seen that the proposed method can also correctly suppress the azimuth sidelobes. Figure 3 (e) is the 8-bit quantization result, and the target splitting phenomenon is alleviated compared with the SCF method; Figure 3 (f) is the 57-bit quantization result, which is basically consistent with PCF. The index calculation results of different algorithms are as follows: Figure 4 shown.
[0085] Figure 4 It can be seen that the performance of the algorithm proposed in the present invention is affected by the number of quantization bits. When the number of quantization bits is 2, the performance of the proposed algorithm is consistent with the SCF method. When the number of quantization bits is small, the performance of the proposed algorithm fluctuates around the performance of the PCF algorithm. Some image performances corresponding to the number of quantization bits are better than the PCF method, for example, when the number of quantization bits is 3, 5, and 8. When the number of quantization bits gradually increases, the performance of the proposed algorithm is consistent with PCF. Figure 4 It can be concluded that the algorithm proposed in the present invention can maintain the best performance when the number of quantization bits is certain. A smaller number of quantization bits will lead to target fragmentation, and a larger number of quantization bits will lead to an increase in the amount of calculation.
[0086] No noise is added to the above simulation results, but in actual situations, radar data will be contaminated by noise. Figure 5 The effects of different methods on input signal-to-noise ratio are shown. The input signal-to-noise ratio varies from -20dB to 20dB, and each input signal-to-noise ratio is subjected to 200 Monte Carlo experiments. The results of the proposed algorithm are the best values under different quantization bits to show the optimal performance that the proposed algorithm can achieve. Figure 5It can be seen that the proposed method can achieve the maximum PSLR, the minimum image entropy and the maximum image contrast under different input signal-to-noise ratios. Therefore, the proposed algorithm has the ability to outperform the existing algorithms as long as the number of quantization bits is selected appropriately.
[0087] Furthermore, the performance of the method proposed in the present invention was simply tested through an outdoor experiment. The experimental scene is shown in FIG. Figure 6 The parameters of the radar array and the ultra-wideband signal are consistent with the simulation. The radar is placed close to a 24cm thick air brick wall, and the radar height is consistent with the human chest. Taking the center of the radar as the origin, the human body behind the wall is approximately located at (0m, 4m, 0m). The imaging results of the algorithm proposed in this invention are as follows Figure 7 As shown in Figure 2. After testing, when the number of quantization bits is 3, the performance of the proposed algorithm is the best. Figure 7 As shown in (b), the proposed algorithm can effectively suppress grating lobes and side lobes. However, due to the small quantization phase, it also faces the problem of target segmentation. Table 1 shows the index calculation results of the proposed algorithm. It can be seen that after the processing of the proposed algorithm, the peak sidelobe ratio of the image is improved, the image entropy is reduced, and the image contrast is improved.
[0088] Table 1 Calculation results of indicators of the proposed method in the measured data
[0089] PSLR / dB Entropy Contrast BP 14.16 10.11 5.16 Proposed 39.52 8.16 12.32
[0090] It can be seen that the present invention discloses a through-wall radar grating lobe suppression method based on multi-bit phase quantization. First, the phase of each channel of the same imaging grid point is quantized according to the specified number of quantization bits. Then, according to the consistency of the quantized phase, the weighted value of the imaging grid point is obtained. Finally, the original radar image is weighted to achieve azimuth grating lobe and sidelobe suppression. Through simulation and measured implementation cases, the effectiveness of the method disclosed in the present invention is proved. This method is not only applicable to the field of through-wall radar imaging detection, but can also be applied to synthetic aperture radar imaging, microwave imaging and other fields. It is a universal image grating lobe and sidelobe suppression method.
[0091] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may certainly make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.
Claims
1. A method for suppressing grating lobes of through-wall radar based on multi-bit phase quantization, characterized in that: The following steps are involved: S1: Calculate the imaging result I(C, k) of each radar channel at the imaging grid point C based on the radar echo, where the value range of the channel number k is [1, Q×P], where Q is the number of receiving antennas on the radar and P is the number of transmitting antennas on the radar; S2: According to the set quantization bit number n q The phase of the imaging result I(C, k) of each channel is quantized to obtain the quantized phase θ of each channel. C,k : in, is the phase of the imaging result I(C, k) of each channel, i Indicates the number of quantized bits n in the phase region [0,2π) q Evenly divided phase intervals, with interval number i = 1, ..., n q ; S3: According to the quantized phase θ of each channel C,k Obtaining a weight W(C) for performing grating lobe suppression; S4: Suppress the grating lobe of each imaging result I(C,k) according to the weight W(C) to obtain the image I′(C) after the azimuth grating lobe and side lobe are suppressed: I′(C)=I(C)·W ν (C) Where, ν is the set weighted control factor, the image before azimuth grating lobe and side lobe suppression 2. A through-wall radar grating lobe suppression method based on multi-bit phase quantization as claimed in claim 1, characterized in that: In step S1, the imaging result I(C, k) is obtained by: Taking the radar center as the origin, assuming the coordinates of the pth transmitting antenna are (x p ,0,z p ), the coordinates of the qth receiving antenna are (x q ,0,z q ); The radar transmitting antenna transmits an ultra-wideband stepped frequency signal containing N frequency points by time division multiplexing, where the ultra-wideband stepped frequency signal x trans (t) is: x trans (t)=exp(-j2π(f0+nΔf)t),n=0,...,N-1 Wherein, f0 is the initial frequency of the ultra-wideband stepped frequency signal, Δf is the frequency step interval, j represents the imaginary part, and t represents time; The kth channel is coupled to the ultra-wideband stepped frequency signal x trans (t) After down-mixing and frequency sampling, the target echo r is obtained k (n): r k (n)=exp(-j2π(f0+nΔf)τ k ) Among them, the kth channel refers to the ultra-wideband stepped frequency signal x trans (t) is transmitted by the pth transmit antenna and received by the qth receive antenna, τ k =R k / c+Δτ k represents the signal propagation time delay of the kth channel, R k represents the distance between the kth channel and the target, Δτ k represents the additional delay of the kth channel; The target echo r is projected using the coherent backscattering method. k (n) is processed to obtain the imaging result I(C,k): Among them, τ C,k represents the time delay between the kth channel and the imaging grid point C.
3. The through-wall radar grating lobe suppression method based on multi-bit phase quantization according to claim 1, characterized in that: The calculation method of the weight W(C) in step S3 is: W(C)=1-std(θ C ) Among them, the phase sequence θ C =[θ C,1 …θ C,QP ] T , std(·) represents the standard deviation.
4. The through-wall radar grating lobe suppression method based on multi-bit phase quantization according to claim 1, characterized in that: In step S2, the phase of the imaging result I(C, k) of each channel The calculation method is: Here, angle(·) represents a phase operation.
5. The through-wall radar grating lobe suppression method based on multi-bit phase quantization according to claim 1, characterized in that: In step S2, the phase interval is i The division method is: in, is a set of integers.
